Diff for /imach096d/doc/imach.htm between versions 1.1.1.1 and 1.5

version 1.1.1.1, 2000/12/28 18:49:54 version 1.5, 2002/03/04 10:01:45
Line 1 Line 1
 <html>  <html>
   
 <head>  <head>
 <meta http-equiv="Content-Type"  <meta http-equiv="Content-Type"
 content="text/html; charset=iso-8859-1">  content="text/html; charset=iso-8859-1">
 <meta name="GENERATOR" content="Microsoft FrontPage Express 2.0">  <meta name="ProgId" content="Word.Document">
 <title>Computing Health Expectancies using IMaCh</title>  <meta name="Originator" content="Microsoft Word 9">
 </head>  <meta name="GENERATOR" content="Microsoft FrontPage Express 2.0">
   <title>Computing Health Expectancies using IMaCh</title>
 <body bgcolor="#FFFFFF">  <link rel="File-List" href="./imach_fichiers/filelist.xml">
   <link rel="Edit-Time-Data" href="./imach_fichiers/editdata.mso">
 <hr size="3" color="#EC5E5E">  <!--[if !mso]>
   <style>
 <h1 align="center"><font color="#00006A">Computing Health  v\:* {behavior:url(#default#VML);}
 Expectancies using IMaCh</font></h1>  o\:* {behavior:url(#default#VML);}
   w\:* {behavior:url(#default#VML);}
 <h1 align="center"><font color="#00006A" size="5">(a Maximum  .shape {behavior:url(#default#VML);}
 Likelihood Computer Program using Interpolation of Markov Chains)</font></h1>  </style>
   <![endif]-->
 <p align="center">&nbsp;</p>  <!--[if gte mso 9]><xml>
    <o:DocumentProperties>
 <p align="center"><a href="http://www.ined.fr/"><img    <o:Author>agnes lievre</o:Author>
 src="logo-ined.gif" border="0" width="151" height="76"></a><img    <o:Template>Normal</o:Template>
 src="euroreves2.gif" width="151" height="75"></p>    <o:LastAuthor>agnes lievre</o:LastAuthor>
     <o:Revision>23</o:Revision>
 <h3 align="center"><a href="http://www.ined.fr/"><font    <o:TotalTime>311</o:TotalTime>
 color="#00006A">INED</font></a><font color="#00006A"> and </font><a    <o:Created>2002-03-02T16:20:00Z</o:Created>
 href="http://euroreves.ined.fr"><font color="#00006A">EUROREVES</font></a></h3>    <o:LastSaved>2002-03-03T21:50:00Z</o:LastSaved>
     <o:Pages>15</o:Pages>
 <p align="center"><font color="#00006A" size="4"><strong>March    <o:Words>6119</o:Words>
 2000</strong></font></p>    <o:Characters>34882</o:Characters>
     <o:Lines>290</o:Lines>
 <hr size="3" color="#EC5E5E">    <o:Paragraphs>69</o:Paragraphs>
     <o:CharactersWithSpaces>42837</o:CharactersWithSpaces>
 <p align="center"><font color="#00006A"><strong>Authors of the    <o:Version>9.4402</o:Version>
 program: </strong></font><a href="http://sauvy.ined.fr/brouard"><font   </o:DocumentProperties>
 color="#00006A"><strong>Nicolas Brouard</strong></font></a><font  </xml><![endif]-->
 color="#00006A"><strong>, senior researcher at the </strong></font><a  <!--[if gte mso 9]><xml>
 href="http://www.ined.fr"><font color="#00006A"><strong>Institut   <w:WordDocument>
 National d'Etudes Démographiques</strong></font></a><font    <w:HyphenationZone>21</w:HyphenationZone>
 color="#00006A"><strong> (INED, Paris) in the &quot;Mortality,   </w:WordDocument>
 Health and Epidemiology&quot; Research Unit </strong></font></p>  </xml><![endif]-->
   <style>
 <p align="center"><font color="#00006A"><strong>and Agnès  <!--
 Lièvre<br clear="left">   /* Font Definitions */
 </strong></font></p>  @font-face
           {font-family:Wingdings;
 <h4><font color="#00006A">Contribution to the mathematics: C. R.          panose-1:5 0 0 0 0 0 0 0 0 0;
 Heathcote </font><font color="#00006A" size="2">(Australian          mso-font-charset:2;
 National University, Canberra).</font></h4>          mso-generic-font-family:auto;
           mso-font-pitch:variable;
 <h4><font color="#00006A">Contact: Agnès Lièvre (</font><a          mso-font-signature:0 268435456 0 0 -2147483648 0;}
 href="mailto:lievre@ined.fr"><font color="#00006A"><i>lievre@ined.fr</i></font></a><font   /* Style Definitions */
 color="#00006A">) </font></h4>  p.MsoNormal, li.MsoNormal, div.MsoNormal
           {mso-style-parent:"";
 <hr>          margin:0cm;
           margin-bottom:.0001pt;
 <ul>          mso-pagination:widow-orphan;
     <li><a href="#intro">Introduction</a> </li>          font-size:12.0pt;
     <li>The detailed statistical model (<a href="docmath.pdf">PDF          font-family:"Times New Roman";
         version</a>),(<a href="docmath.ps">ps version</a>) </li>          mso-fareast-font-family:"Times New Roman";}
     <li><a href="#data">On what kind of data can it be used?</a></li>  h1
     <li><a href="#datafile">The data file</a> </li>          {margin-right:0cm;
     <li><a href="#biaspar">The parameter file</a> </li>          mso-margin-top-alt:auto;
     <li><a href="#running">Running Imach</a> </li>          mso-margin-bottom-alt:auto;
     <li><a href="#output">Output files and graphs</a> </li>          margin-left:0cm;
     <li><a href="#example">Exemple</a> </li>          mso-pagination:widow-orphan;
 </ul>          mso-outline-level:1;
           font-size:24.0pt;
 <hr>          font-family:"Times New Roman";
           mso-font-kerning:18.0pt;
 <h2><a name="intro"><font color="#00006A">Introduction</font></a></h2>          font-weight:bold;}
   h2
 <p>This program computes <b>Healthy Life Expectancies</b> from <b>cross-longitudinal          {margin-right:0cm;
 data</b>. Within the family of Health Expectancies (HE),          mso-margin-top-alt:auto;
 Disability-free life expectancy (DFLE) is probably the most          mso-margin-bottom-alt:auto;
 important index to monitor. In low mortality countries, there is          margin-left:0cm;
 a fear that when mortality declines, the increase in DFLE is not          mso-pagination:widow-orphan;
 proportionate to the increase in total Life expectancy. This case          mso-outline-level:2;
 is called the <em>Expansion of morbidity</em>. Most of the data          font-size:18.0pt;
 collected today, in particular by the international <a          font-family:"Times New Roman";
 href="http://euroreves/reves">REVES</a> network on Health          font-weight:bold;}
 expectancy, and most HE indices based on these data, are <em>cross-sectional</em>.  h3
 It means that the information collected comes from a single          {margin-right:0cm;
 cross-sectional survey: people from various ages (but mostly old          mso-margin-top-alt:auto;
 people) are surveyed on their health status at a single date.          mso-margin-bottom-alt:auto;
 Proportion of people disabled at each age, can then be measured          margin-left:0cm;
 at that date. This age-specific prevalence curve is then used to          mso-pagination:widow-orphan;
 distinguish, within the stationary population (which, by          mso-outline-level:3;
 definition, is the life table estimated from the vital statistics          font-size:13.5pt;
 on mortality at the same date), the disable population from the          font-family:"Times New Roman";
 disability-free population. Life expectancy (LE) (or total          font-weight:bold;}
 population divided by the yearly number of births or deaths of  h4
 this stationary population) is then decomposed into DFLE and DLE.          {margin-right:0cm;
 This method of computing HE is usually called the Sullivan method          mso-margin-top-alt:auto;
 (from the name of the author who first described it).</p>          mso-margin-bottom-alt:auto;
           margin-left:0cm;
 <p>Age-specific proportions of people disable are very difficult          mso-pagination:widow-orphan;
 to forecast because each proportion corresponds to historical          mso-outline-level:4;
 conditions of the cohort and it is the result of the historical          font-size:12.0pt;
 flows from entering disability and recovering in the past until          font-family:"Times New Roman";
 today. The age-specific intensities (or incidence rates) of          font-weight:bold;}
 entering disability or recovering a good health, are reflecting  h5
 actual conditions and therefore can be used at each age to          {margin-right:0cm;
 forecast the future of this cohort. For example if a country is          mso-margin-top-alt:auto;
 improving its technology of prosthesis, the incidence of          mso-margin-bottom-alt:auto;
 recovering the ability to walk will be higher at each (old) age,          margin-left:0cm;
 but the prevalence of disability will only slightly reflect an          mso-pagination:widow-orphan;
 improve because the prevalence is mostly affected by the history          mso-outline-level:5;
 of the cohort and not by recent period effects. To measure the          font-size:10.0pt;
 period improvement we have to simulate the future of a cohort of          font-family:"Times New Roman";
 new-borns entering or leaving at each age the disability state or          font-weight:bold;}
 dying according to the incidence rates measured today on  a:link, span.MsoHyperlink
 different cohorts. The proportion of people disabled at each age          {color:blue;
 in this simulated cohort will be much lower (using the exemple of          text-decoration:underline;
 an improvement) that the proportions observed at each age in a          text-underline:single;}
 cross-sectional survey. This new prevalence curve introduced in a  a:visited, span.MsoHyperlinkFollowed
 life table will give a much more actual and realistic HE level          {color:blue;
 than the Sullivan method which mostly measured the History of          text-decoration:underline;
 health conditions in this country.</p>          text-underline:single;}
   p
 <p>Therefore, the main question is how to measure incidence rates          {margin-right:0cm;
 from cross-longitudinal surveys? This is the goal of the IMaCH          mso-margin-top-alt:auto;
 program. From your data and using IMaCH you can estimate period          mso-margin-bottom-alt:auto;
 HE and not only Sullivan's HE. Also the standard errors of the HE          margin-left:0cm;
 are computed.</p>          mso-pagination:widow-orphan;
           font-size:12.0pt;
 <p>A cross-longitudinal survey consists in a first survey          font-family:"Times New Roman";
 (&quot;cross&quot;) where individuals from different ages are          mso-fareast-font-family:"Times New Roman";}
 interviewed on their health status or degree of disability. At  pre
 least a second wave of interviews (&quot;longitudinal&quot;)          {margin:0cm;
 should measure each new individual health status. Health          margin-bottom:.0001pt;
 expectancies are computed from the transitions observed between          mso-pagination:widow-orphan;
 waves and are computed for each degree of severity of disability          tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt;
 (number of life states). More degrees you consider, more time is          font-size:10.0pt;
 necessary to reach the Maximum Likelihood of the parameters          font-family:"Courier New";
 involved in the model. Considering only two states of disability          mso-fareast-font-family:"Courier New";}
 (disable and healthy) is generally enough but the computer  @page Section1
 program works also with more health statuses.<br>          {size:595.3pt 841.9pt;
 <br>          margin:70.85pt 70.85pt 70.85pt 70.85pt;
 The simplest model is the multinomial logistic model where <i>pij</i>          mso-header-margin:35.4pt;
 is the probability to be observed in state <i>j</i> at the second          mso-footer-margin:35.4pt;
 wave conditional to be observed in state <em>i</em> at the first          mso-paper-source:0;}
 wave. Therefore a simple model is: log<em>(pij/pii)= aij +  div.Section1
 bij*age+ cij*sex,</em> where '<i>age</i>' is age and '<i>sex</i>'          {page:Section1;}
 is a covariate. The advantage that this computer program claims,   /* List Definitions */
 comes from that if the delay between waves is not identical for  @list l0
 each individual, or if some individual missed an interview, the          {mso-list-id:184488806;
 information is not rounded or lost, but taken into account using          mso-list-type:hybrid;
 an interpolation or extrapolation. <i>hPijx</i> is the          mso-list-template-ids:-1529696740 -412605010 -907664984 -2032001966 -1219577922 -1989525672 1804215288 1798964568 2064445346 -921394074;}
 probability to be observed in state <i>i</i> at age <i>x+h</i>  @list l0:level1
 conditional to the observed state <i>i</i> at age <i>x</i>. The          {mso-level-number-format:bullet;
 delay '<i>h</i>' can be split into an exact number (<i>nh*stepm</i>)          mso-level-text:\F0B7;
 of unobserved intermediate states. This elementary transition (by          mso-level-tab-stop:36.0pt;
 month or quarter trimester, semester or year) is modeled as a          mso-level-number-position:left;
 multinomial logistic. The <i>hPx</i> matrix is simply the matrix          text-indent:-18.0pt;
 product of <i>nh*stepm</i> elementary matrices and the          mso-ansi-font-size:10.0pt;
 contribution of each individual to the likelihood is simply <i>hPijx</i>.          font-family:Symbol;}
 <br>  @list l1
 </p>          {mso-list-id:204831336;
           mso-list-type:hybrid;
 <p>The program presented in this manual is a quite general          mso-list-template-ids:-1844294122 94378308 -806846340 -1272687644 -333439258 -1126675816 1099701808 1453997410 1355317140 -183874942;}
 program named <strong>IMaCh</strong> (for <strong>I</strong>nterpolated  @list l1:level1
 <strong>MA</strong>rkov <strong>CH</strong>ain), designed to          {mso-level-number-format:bullet;
 analyse transition data from longitudinal surveys. The first step          mso-level-text:\F0B7;
 is the parameters estimation of a transition probabilities model          mso-level-tab-stop:36.0pt;
 between an initial status and a final status. From there, the          mso-level-number-position:left;
 computer program produces some indicators such as observed and          text-indent:-18.0pt;
 stationary prevalence, life expectancies and their variances and          mso-ansi-font-size:10.0pt;
 graphs. Our transition model consists in absorbing and          font-family:Symbol;}
 non-absorbing states with the possibility of return across the  @list l2
 non-absorbing states. The main advantage of this package,          {mso-list-id:441344838;
 compared to other programs for the analysis of transition data          mso-list-type:hybrid;
 (For example: Proc Catmod of SAS<sup>®</sup>) is that the whole          mso-list-template-ids:-1624363430 -1068334662 975101306 1309600228 -2116116870 769974826 -307843868 1028545738 868800422 -1496705886;}
 individual information is used even if an interview is missing, a  @list l2:level1
 status or a date is unknown or when the delay between waves is          {mso-level-number-format:bullet;
 not identical for each individual. The program can be executed          mso-level-text:\F0B7;
 according to parameters: selection of a sub-sample, number of          mso-level-tab-stop:36.0pt;
 absorbing and non-absorbing states, number of waves taken in          mso-level-number-position:left;
 account (the user inputs the first and the last interview), a          text-indent:-18.0pt;
 tolerance level for the maximization function, the periodicity of          mso-ansi-font-size:10.0pt;
 the transitions (we can compute annual, quaterly or monthly          font-family:Symbol;}
 transitions), covariates in the model. It works on Windows or on  @list l3
 Unix.<br>          {mso-list-id:628702260;
 </p>          mso-list-type:hybrid;
           mso-list-template-ids:-927709454 416302846 373988436 -1134147144 -1982968238 475822148 827730770 934571264 602550890 1800972154;}
 <hr>  @list l3:level1
           {mso-level-number-format:bullet;
 <h2><a name="data"><font color="#00006A">On what kind of data can          mso-level-text:\F0B7;
 it be used?</font></a></h2>          mso-level-tab-stop:36.0pt;
           mso-level-number-position:left;
 <p>The minimum data required for a transition model is the          text-indent:-18.0pt;
 recording of a set of individuals interviewed at a first date and          mso-ansi-font-size:10.0pt;
 interviewed again at least one another time. From the          font-family:Symbol;}
 observations of an individual, we obtain a follow-up over time of  @list l4
 the occurrence of a specific event. In this documentation, the          {mso-list-id:752507057;
 event is related to health status at older ages, but the program          mso-list-type:hybrid;
 can be applied on a lot of longitudinal studies in different          mso-list-template-ids:-1773518296 67895297 67895299 67895301 67895297 67895299 67895301 67895297 67895299 67895301;}
 contexts. To build the data file explained into the next section,  @list l4:level1
 you must have the month and year of each interview and the          {mso-level-number-format:bullet;
 corresponding health status. But in order to get age, date of          mso-level-text:\F0B7;
 birth (month and year) is required (missing values is allowed for          mso-level-tab-stop:36.0pt;
 month). Date of death (month and year) is an important          mso-level-number-position:left;
 information also required if the individual is dead. Shorter          text-indent:-18.0pt;
 steps (i.e. a month) will more closely take into account the          font-family:Symbol;}
 survival time after the last interview.</p>  @list l5
           {mso-list-id:818232419;
 <hr>          mso-list-type:hybrid;
           mso-list-template-ids:-1143944236 -95769308 -1426324944 1101845342 1451904852 -884162680 1945505468 -1163215946 -592140202 1700436632;}
 <h2><a name="datafile"><font color="#00006A">The data file</font></a></h2>  @list l5:level1
           {mso-level-number-format:bullet;
 <p>In this example, 8,000 people have been interviewed in a          mso-level-text:\F0B7;
 cross-longitudinal survey of 4 waves (1984, 1986, 1988, 1990).          mso-level-tab-stop:36.0pt;
 Some people missed 1, 2 or 3 interviews. Health statuses are          mso-level-number-position:left;
 healthy (1) and disable (2). The survey is not a real one. It is          text-indent:-18.0pt;
 a simulation of the American Longitudinal Survey on Aging. The          mso-ansi-font-size:10.0pt;
 disability state is defined if the individual missed one of four          font-family:Symbol;}
 ADL (Activity of daily living, like bathing, eating, walking).  @list l6
 Therefore, even is the individuals interviewed in the sample are          {mso-list-id:883836538;
 virtual, the information brought with this sample is close to the          mso-list-type:hybrid;
 situation of the United States. Sex is not recorded is this          mso-list-template-ids:32399718 -134710470 -1096777840 1309058016 1376137788 -1290646696 -883388418 -1972580442 -797425852 -2051513306;}
 sample.</p>  @list l6:level1
           {mso-level-number-format:bullet;
 <p>Each line of the data set (named <a href="data1.txt">data1.txt</a>          mso-level-text:\F0B7;
 in this first example) is an individual record which fields are: </p>          mso-level-tab-stop:36.0pt;
           mso-level-number-position:left;
 <ul>          text-indent:-18.0pt;
     <li><b>Index number</b>: positive number (field 1) </li>          mso-ansi-font-size:10.0pt;
     <li><b>First covariate</b> positive number (field 2) </li>          font-family:Symbol;}
     <li><b>Second covariate</b> positive number (field 3) </li>  @list l7
     <li><a name="Weight"><b>Weight</b></a>: positive number          {mso-list-id:904297090;
         (field 4) . In most surveys individuals are weighted          mso-list-type:hybrid;
         according to the stratification of the sample.</li>          mso-list-template-ids:651483934 -1330890910 68314940 -1090220572 161909366 1495697220 -847235596 603869070 -2107325174 -1292888626;}
     <li><b>Date of birth</b>: coded as mm/yyyy. Missing dates are  @list l7:level1
         coded as 99/9999 (field 5) </li>          {mso-level-number-format:bullet;
     <li><b>Date of death</b>: coded as mm/yyyy. Missing dates are          mso-level-text:\F0B7;
         coded as 99/9999 (field 6) </li>          mso-level-tab-stop:36.0pt;
     <li><b>Date of first interview</b>: coded as mm/yyyy. Missing          mso-level-number-position:left;
         dates are coded as 99/9999 (field 7) </li>          text-indent:-18.0pt;
     <li><b>Status at first interview</b>: positive number.          mso-ansi-font-size:10.0pt;
         Missing values ar coded -1. (field 8) </li>          font-family:Symbol;}
     <li><b>Date of second interview</b>: coded as mm/yyyy.  @list l8
         Missing dates are coded as 99/9999 (field 9) </li>          {mso-list-id:1063483777;
     <li><strong>Status at second interview</strong> positive          mso-list-type:hybrid;
         number. Missing values ar coded -1. (field 10) </li>          mso-list-template-ids:1396622820 67895297 67895299 67895301 67895297 67895299 67895301 67895297 67895299 67895301;}
     <li><b>Date of third interview</b>: coded as mm/yyyy. Missing  @list l8:level1
         dates are coded as 99/9999 (field 11) </li>          {mso-level-number-format:bullet;
     <li><strong>Status at third interview</strong> positive          mso-level-text:\F0B7;
         number. Missing values ar coded -1. (field 12) </li>          mso-level-tab-stop:36.0pt;
     <li><b>Date of fourth interview</b>: coded as mm/yyyy.          mso-level-number-position:left;
         Missing dates are coded as 99/9999 (field 13) </li>          text-indent:-18.0pt;
     <li><strong>Status at fourth interview</strong> positive          font-family:Symbol;}
         number. Missing values are coded -1. (field 14) </li>  @list l9
     <li>etc</li>          {mso-list-id:1168903496;
 </ul>          mso-list-type:hybrid;
           mso-list-template-ids:-2008413052 -1461950228 -1012751362 -919308802 443596168 -1469662014 -359112926 504948540 1759957928 1612641564;}
 <p>&nbsp;</p>  @list l9:level1
           {mso-level-number-format:bullet;
 <p>If your longitudinal survey do not include information about          mso-level-text:\F0B7;
 weights or covariates, you must fill the column with a number          mso-level-tab-stop:36.0pt;
 (e.g. 1) because a missing field is not allowed.</p>          mso-level-number-position:left;
           text-indent:-18.0pt;
 <hr>          mso-ansi-font-size:10.0pt;
           font-family:Symbol;}
 <h2><font color="#00006A">Your first example parameter file</font><a  @list l10
 href="http://euroreves.ined.fr/imach"></a><a name="uio"></a></h2>          {mso-list-id:1190214980;
           mso-list-type:hybrid;
 <h2><a name="biaspar"></a>#Imach version 0.63, February 2000,          mso-list-template-ids:934709936 -291345380 -81906558 -968867310 665229182 1336730126 -1107941388 -20391304 -674328264 -1574639962;}
 INED-EUROREVES </h2>  @list l10:level1
           {mso-level-number-format:bullet;
 <p>This is a comment. Comments start with a '#'.</p>          mso-level-text:\F0B7;
           mso-level-tab-stop:36.0pt;
 <h4><font color="#FF0000">First uncommented line</font></h4>          mso-level-number-position:left;
           text-indent:-18.0pt;
 <pre>title=1st_example datafile=data1.txt lastobs=8600 firstpass=1 lastpass=4</pre>          mso-ansi-font-size:10.0pt;
           font-family:Symbol;}
 <ul>  @list l11
     <li><b>title=</b> 1st_example is title of the run. </li>          {mso-list-id:1384715951;
     <li><b>datafile=</b>data1.txt is the name of the data set.          mso-list-type:hybrid;
         Our example is a six years follow-up survey. It consists          mso-list-template-ids:-515744014 -566093190 799967300 770599756 -1594063690 -869741144 1377056636 315393812 -1370061484 1511570004;}
         in a baseline followed by 3 reinterviews. </li>  @list l11:level1
     <li><b>lastobs=</b> 8600 the program is able to run on a          {mso-level-number-format:bullet;
         subsample where the last observation number is lastobs.          mso-level-text:\F0B7;
         It can be set a bigger number than the real number of          mso-level-tab-stop:36.0pt;
         observations (e.g. 100000). In this example, maximisation          mso-level-number-position:left;
         will be done on the 8600 first records. </li>          text-indent:-18.0pt;
     <li><b>firstpass=1</b> , <b>lastpass=4 </b>In case of more          mso-ansi-font-size:10.0pt;
         than two interviews in the survey, the program can be run          font-family:Symbol;}
         on selected transitions periods. firstpass=1 means the  @list l12
         first interview included in the calculation is the          {mso-list-id:1593661621;
         baseline survey. lastpass=4 means that the information          mso-list-type:hybrid;
         brought by the 4th interview is taken into account.</li>          mso-list-template-ids:-1035417432 1271449484 -308236740 -1122210034 -380844018 1478807872 266132728 -1091829500 812926462 -1442827238;}
 </ul>  @list l12:level1
           {mso-level-number-format:bullet;
 <p>&nbsp;</p>          mso-level-text:\F0B7;
           mso-level-tab-stop:36.0pt;
 <h4><a name="biaspar-2"><font color="#FF0000">Second uncommented          mso-level-number-position:left;
 line</font></a></h4>          text-indent:-18.0pt;
           mso-ansi-font-size:10.0pt;
 <pre>ftol=1.e-08 stepm=1 ncov=2 nlstate=2 ndeath=1 maxwav=4 mle=1 weight=0</pre>          font-family:Symbol;}
   @list l13
 <ul>          {mso-list-id:1636450504;
     <li><b>ftol=1e-8</b> Convergence tolerance on the function          mso-list-type:hybrid;
         value in the maximisation of the likelihood. Choosing a          mso-list-template-ids:-711022678 -1038569226 1304059700 -837663288 -1699980300 571783806 -231993906 -744861656 1958002196 -1476655198;}
         correct value for ftol is difficult. 1e-8 is a correct  @list l13:level1
         value for a 32 bits computer.</li>          {mso-level-number-format:bullet;
     <li><b>stepm=1</b> Time unit in months for interpolation.          mso-level-text:\F0B7;
         Examples:<ul>          mso-level-tab-stop:36.0pt;
             <li>If stepm=1, the unit is a month </li>          mso-level-number-position:left;
             <li>If stepm=4, the unit is a trimester</li>          text-indent:-18.0pt;
             <li>If stepm=12, the unit is a year </li>          mso-ansi-font-size:10.0pt;
             <li>If stepm=24, the unit is two years</li>          font-family:Symbol;}
             <li>... </li>  @list l14
         </ul>          {mso-list-id:1752386307;
     </li>          mso-list-type:hybrid;
     <li><b>ncov=2</b> Number of covariates to be add to the          mso-list-template-ids:-347696224 -386773934 1871641532 667840386 1914592500 1728978276 -196066776 1566372654 -755335742 341755130;}
         model. The intercept and the age parameter are counting  @list l14:level1
         for 2 covariates. For example, if you want to add gender          {mso-level-number-format:bullet;
         in the covariate vector you must write ncov=3 else          mso-level-text:\F0B7;
         ncov=2. </li>          mso-level-tab-stop:36.0pt;
     <li><b>nlstate=2</b> Number of non-absorbing (live) states.          mso-level-number-position:left;
         Here we have two alive states: disability-free is coded 1          text-indent:-18.0pt;
         and disability is coded 2. </li>          mso-ansi-font-size:10.0pt;
     <li><b>ndeath=1</b> Number of absorbing states. The absorbing          font-family:Symbol;}
         state death is coded 3. </li>  @list l14:level2
     <li><b>maxwav=4</b> Maximum number of waves. The program can          {mso-level-number-format:bullet;
         not include more than 4 interviews. </li>          mso-level-text:o;
     <li><a name="mle"><b>mle</b></a><b>=1</b> Option for the          mso-level-tab-stop:72.0pt;
         Maximisation Likelihood Estimation. <ul>          mso-level-number-position:left;
             <li>If mle=1 the program does the maximisation and          text-indent:-18.0pt;
                 the calculation of heath expectancies </li>          mso-ansi-font-size:10.0pt;
             <li>If mle=0 the program only does the calculation of          font-family:"Courier New";
                 the health expectancies. </li>          mso-bidi-font-family:"Times New Roman";}
         </ul>  @list l15
     </li>          {mso-list-id:1756245288;
     <li><b>weight=0</b> Possibility to add weights. <ul>          mso-list-type:hybrid;
             <li>If weight=0 no weights are included </li>          mso-list-template-ids:531934386 67895297 67895299 67895301 67895297 67895299 67895301 67895297 67895299 67895301;}
             <li>If weight=1 the maximisation integrates the  @list l15:level1
                 weights which are in field <a href="#Weight">4</a></li>          {mso-level-number-format:bullet;
         </ul>          mso-level-text:\F0B7;
     </li>          mso-level-tab-stop:36.0pt;
 </ul>          mso-level-number-position:left;
           text-indent:-18.0pt;
 <h4><font color="#FF0000">Guess values for optimization</font><font          font-family:Symbol;}
 color="#00006A"> </font></h4>  @list l16
           {mso-list-id:1839273133;
 <p>You must write the initial guess values of the parameters for          mso-list-type:hybrid;
 optimization. The number of parameters, <em>N</em> depends on the          mso-list-template-ids:-556523634 -715873828 -243865004 563531560 -898876536 640947630 967865102 1305671924 1810678544 -1115658030;}
 number of absorbing states and non-absorbing states and on the  @list l16:level1
 number of covariates. <br>          {mso-level-number-format:bullet;
 <em>N</em> is given by the formula <em>N</em>=(<em>nlstate</em> +          mso-level-text:\F0B7;
 <em>ndeath</em>-1)*<em>nlstate</em>*<em>ncov</em>&nbsp;. <br>          mso-level-tab-stop:36.0pt;
 <br>          mso-level-number-position:left;
 Thus in the simple case with 2 covariates (the model is log          text-indent:-18.0pt;
 (pij/pii) = aij + bij * age where intercept and age are the two          mso-ansi-font-size:10.0pt;
 covariates), and 2 health degrees (1 for disability-free and 2          font-family:Symbol;}
 for disability) and 1 absorbing state (3), you must enter 8  @list l17
 initials values, a12, b12, a13, b13, a21, b21, a23, b23. You can          {mso-list-id:1841849959;
 start with zeros as in this example, but if you have a more          mso-list-type:hybrid;
 precise set (for example from an earlier run) you can enter it          mso-list-template-ids:2053128728 -543362536 926470224 151426154 998932566 84972724 844683600 1807279286 -841218426 -1132452502;}
 and it will speed up them<br>  @list l17:level1
 Each of the four lines starts with indices &quot;ij&quot;: <br>          {mso-level-number-format:bullet;
 <br>          mso-level-text:\F0B7;
 <b>ij aij bij</b> </p>          mso-level-tab-stop:36.0pt;
           mso-level-number-position:left;
 <blockquote>          text-indent:-18.0pt;
     <pre># Guess values of aij and bij in log (pij/pii) = aij + bij * age          mso-ansi-font-size:10.0pt;
 12 -14.155633  0.110794           font-family:Symbol;}
 13  -7.925360  0.032091   @list l18
 21  -1.890135 -0.029473           {mso-list-id:1848639524;
 23  -6.234642  0.022315 </pre>          mso-list-type:hybrid;
 </blockquote>          mso-list-template-ids:638092306 940881202 -784414886 1026841176 1011505968 -653358884 -269310374 2133217052 1173680566 -1995784172;}
   @list l18:level1
 <p>or, to simplify: </p>          {mso-level-number-format:bullet;
           mso-level-text:\F0B7;
 <blockquote>          mso-level-tab-stop:36.0pt;
     <pre>12 0.0 0.0          mso-level-number-position:left;
 13 0.0 0.0          text-indent:-18.0pt;
 21 0.0 0.0          mso-ansi-font-size:10.0pt;
 23 0.0 0.0</pre>          font-family:Symbol;}
 </blockquote>  ol
           {margin-bottom:0cm;}
 <h4><font color="#FF0000">Guess values for computing variances</font></h4>  ul
           {margin-bottom:0cm;}
 <p>This is an output if <a href="#mle">mle</a>=1. But it can be  -->
 used as an input to get the vairous output data files (Health  </style>
 expectancies, stationary prevalence etc.) and figures without  <!--[if gte mso 9]><xml>
 rerunning the rather long maximisation phase (mle=0). </p>   <o:shapedefaults v:ext="edit" spidmax="1027"/>
   </xml><![endif]-->
 <p>The scales are small values for the evaluation of numerical  <!--[if gte mso 9]><xml>
 derivatives. These derivatives are used to compute the hessian   <o:shapelayout v:ext="edit">
 matrix of the parameters, that is the inverse of the covariance    <o:idmap v:ext="edit" data="1"/>
 matrix, and the variances of health expectancies. Each line   </o:shapelayout></xml><![endif]-->
 consists in indices &quot;ij&quot; followed by the initial scales  <!-- Changed by: Agnes Lievre, 12-Oct-2000 -->
 (zero to simplify) associated with aij and bij. </p>  </head>
   
 <ul>  <body bgcolor="#FFFFFF" link="#0000FF" vlink="#0000FF" lang="FR"
     <li>If mle=1 you can enter zeros:</li>  style="tab-interval:35.4pt">
 </ul>  
   <hr size="3" noshade color="#EC5E5E">
 <blockquote>  
     <pre># Scales (for hessian or gradient estimation)  <h1 align="center" style="text-align:center"><span lang="EN-GB" style="color:#00006A;
 12 0. 0.   mso-ansi-language:EN-GB">Computing Health
 13 0. 0.   Expectancies using IMaCh</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"><o:p></o:p></span></h1>
 21 0. 0.   
 23 0. 0. </pre>  <h1 align="center" style="text-align:center"><span lang="EN-GB" style="font-size:
 </blockquote>  18.0pt;color:#00006A;mso-ansi-language:EN-GB">(a Maximum
   Likelihood Computer Program using Interpolation of Markov Chains)</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"><o:p></o:p></span></h1>
 <ul>  
     <li>If mle=0 you must enter a covariance matrix (usually  <p align="center" style="text-align:center"><span lang="EN-GB" style="mso-ansi-language:
         obtained from an earlier run).</li>  EN-GB">&nbsp;<o:p></o:p></span></p>
 </ul>  
   <p align="center" style="text-align:center"><a
 <h4><font color="#FF0000">Covariance matrix of parameters</font></h4>  href="http://www.ined.fr/"><span style="text-decoration:none;text-underline:none"><img src="logo-ined.gif" border="0"
   width="151" height="76" id="_x0000_i1026"></span></a><img
 <p>This is an output if <a href="#mle">mle</a>=1. But it can be  src="euroreves2.gif" width="151" height="75" id="_x0000_i1027"></p>
 used as an input to get the vairous output data files (Health  
 expectancies, stationary prevalence etc.) and figures without  <h3 align="center" style="text-align:center"><a
 rerunning the rather long maximisation phase (mle=0). </p>  href="http://www.ined.fr/"><span lang="EN-GB" style="color:#00006A;mso-ansi-language:EN-GB">INED</span><span lang="EN-GB" style="color:#00006A;mso-ansi-language:EN-GB"></a> and </span><a
   href="http://euroreves.ined.fr"><span lang="EN-GB" style="color:#00006A;
 <p>Each line starts with indices &quot;ijk&quot; followed by the  mso-ansi-language:EN-GB">EUROREVES</span><span lang="EN-GB" style="mso-ansi-language:
 covariances between aij and bij: </p>  EN-GB"><o:p></o:p></span></a></h3>
   
 <pre>  <p align="center" style="text-align:center"><strong><span lang="EN-GB" style="font-size:13.5pt;color:#00006A;mso-ansi-language:EN-GB">Version 0.7,
    121 Var(a12)   February 2002</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"><o:p></o:p></span></strong></p>
    122 Cov(b12,a12)  Var(b12)   
           ...  <hr size="3" noshade color="#EC5E5E">
    232 Cov(b23,a12)  Cov(b23,b12) ... Var (b23) </pre>  
   <p align="center" style="text-align:center"><strong><span lang="EN-GB" style="color:#00006A;mso-ansi-language:EN-GB">Authors of
 <ul>  the program: </span></strong><a href="http://sauvy.ined.fr/brouard"><strong><span lang="EN-GB" style="color:#00006A;
     <li>If mle=1 you can enter zeros. </li>  mso-ansi-language:EN-GB">Nicolas
 </ul>  Brouard</span><span lang="EN-GB" style="color:#00006A;mso-ansi-language:EN-GB"></strong></a><strong>, senior researcher at the </span></strong><a
   href="http://www.ined.fr"><strong><span lang="EN-GB" style="color:#00006A;mso-ansi-language:EN-GB">Institut National d'Etudes
 <blockquote>  Démographiques</span><span lang="EN-GB" style="color:#00006A;
     <pre># Covariance matrix  mso-ansi-language:EN-GB"></strong></a><strong> (INED, Paris) in the
 121 0.  &quot;Mortality, Health and Epidemiology&quot; Research Unit </span><span lang="EN-GB" style="mso-ansi-language:EN-GB"><o:p></o:p></span></strong></p>
 122 0. 0.  
 131 0. 0. 0.   <p align="center" style="text-align:center"><strong><span lang="EN-GB" style="color:#00006A;mso-ansi-language:EN-GB">and Agnès
 132 0. 0. 0. 0.   Lièvre</span></strong><b><span lang="EN-GB" style="color:#00006A;mso-ansi-language:EN-GB"><br clear="left"
 211 0. 0. 0. 0. 0.   style="mso-special-character:line-break">
 212 0. 0. 0. 0. 0. 0.   </span><span lang="EN-GB" style="mso-ansi-language:EN-GB"><o:p></o:p></span></b></p>
 231 0. 0. 0. 0. 0. 0. 0.   
 232 0. 0. 0. 0. 0. 0. 0. 0.</pre>  <h4><span lang="EN-GB" style="color:#00006A;mso-ansi-language:EN-GB">Contribution to the mathematics: C. R. Heathcote </span><span lang="EN-GB" style="font-size:
 </blockquote>  10.0pt;color:#00006A;mso-ansi-language:EN-GB">(Australian
   National University, Canberra).</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"><o:p></o:p></span></h4>
 <ul>  
     <li>If mle=0 you must enter a covariance matrix (usually  <h4><span style="color:#00006A">Contact: Agnès Lièvre (</span><a href="mailto:lievre@ined.fr"><i><span style="color:#00006A">lievre@ined.fr</span><span style="color:#00006A"></i></a>)
         obtained from an earlier run).<br>  </span></h4>
         </li>  
 </ul>  <hr>
   <span style="font-size:12.0pt;font-family:&quot;Times New Roman&quot;;mso-fareast-font-family:
 <h4><a name="biaspar-l"></a><font color="#FF0000">last  &quot;Times New Roman&quot;;mso-ansi-language:FR;mso-fareast-language:FR;mso-bidi-language:
 uncommented line</font></h4>  AR-SA">
   <ul type="disc">
 <pre>agemin=70 agemax=100 bage=50 fage=100</pre>      <li class="MsoNormal"
       style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
 <p>Once we obtained the estimated parameters, the program is able       mso-list:l17 level1 lfo3;tab-stops:list 36.0pt"><a
 to calculated stationary prevalence, transitions probabilities          href="#intro"><span lang="EN-GB" style="mso-ansi-language:EN-GB">Introduction</span><span style="mso-ansi-language:EN-GB"></a> <span lang="EN-GB"><o:p></o:p></span></span></li>
 and life expectancies at any age. Choice of age ranges is useful      <li class="MsoNormal"
 for extrapolation. In our data file, ages varies from age 70 to      style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
 102. Setting bage=50 and fage=100, makes the program computing       mso-list:l17 level1 lfo3;tab-stops:list 36.0pt"><a
 life expectancy from age bage to age fage. As we use a model, we          href="#data"><span lang="EN-GB" style="mso-ansi-language:EN-GB">On what kind of data can it be used?</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"><o:p></o:p></span></a></li>
 can compute life expectancy on a wider age range than the age      <li class="MsoNormal"
 range from the data. But the model can be rather wrong on big      style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
 intervals.</p>       mso-list:l17 level1 lfo3;tab-stops:list 36.0pt"><a
           href="#datafile"><span lang="EN-GB" style="mso-ansi-language:EN-GB">The data file</span><span style="mso-ansi-language:EN-GB"></a> <span lang="EN-GB"><o:p></o:p></span></span></li>
 <p>Similarly, it is possible to get extrapolated stationary      <li class="MsoNormal"
 prevalence by age raning from agemin to agemax. </p>      style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
        mso-list:l17 level1 lfo3;tab-stops:list 36.0pt"><a
 <ul>          href="#biaspar"><span lang="EN-GB" style="mso-ansi-language:EN-GB">The parameter file</span><span style="mso-ansi-language:EN-GB"></a> <span lang="EN-GB"><o:p></o:p></span></span></li>
     <li><b>agemin=</b> Minimum age for calculation of the      <li class="MsoNormal"
         stationary prevalence </li>      style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
     <li><b>agemax=</b> Maximum age for calculation of the       mso-list:l17 level1 lfo3;tab-stops:list 36.0pt"><a
         stationary prevalence </li>          href="#running"><span lang="EN-GB" style="mso-ansi-language:EN-GB">Running Imach</span><span style="mso-ansi-language:EN-GB"></a> <span lang="EN-GB"><o:p></o:p></span></span></li>
     <li><b>bage=</b> Minimum age for calculation of the health      <li class="MsoNormal"
         expectancies </li>      style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
     <li><b>fage=</b> Maximum ages for calculation of the health       mso-list:l17 level1 lfo3;tab-stops:list 36.0pt"><a
         expectancies </li>          href="#output"><span lang="EN-GB" style="mso-ansi-language:EN-GB">Output files and graphs</span><span style="mso-ansi-language:EN-GB"></a> <span lang="EN-GB"><o:p></o:p></span></span></li>
 </ul>      <li class="MsoNormal"
       style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
 <hr>       mso-list:l17 level1 lfo3;tab-stops:list 36.0pt"><a
           href="#example">Exemple</a> </li>
 <h2><a name="running"></a><font color="#00006A">Running Imach  </ul>
 with this example</font></h2>  </span>
   <hr>
 <p>We assume that you entered your <a href="biaspar.txt">1st_example  
 parameter file</a> as explained <a href="#biaspar">above</a>. To  <h2><a name="intro"><span lang="EN-GB" style="color:#00006A;mso-ansi-language:EN-GB">Introduction</span><span style="mso-bookmark:intro"></span><span lang="EN-GB" style="mso-ansi-language:
 run the program you should click on the imach.exe icon and enter  EN-GB"><o:p></o:p></span></a></h2>
 the name of the parameter file which is for example <a  
 href="C:\usr\imach\mle\biaspar.txt">C:\usr\imach\mle\biaspar.txt</a>  <p style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">This program computes <b>Healthy
 (you also can click on the biaspar.txt icon located in <br>  Life Expectancies</b> from <b>cross-longitudinal data</b> using
 <a href="C:\usr\imach\mle">C:\usr\imach\mle</a> and put it with  the methodology pioneered by Laditka and Wolf (1). Within the
 the mouse on the imach window).<br>  family of Health Expectancies (HE), Disability-free life
 </p>  expectancy (DFLE) is probably the most important index to
   monitor. In low mortality countries, there is a fear that when
 <p>The time to converge depends on the step unit that you used (1  mortality declines, the increase in DFLE is not proportionate to
 month is cpu consuming), on the number of cases, and on the  the increase in total Life expectancy. This case is called the <em>Expansion
 number of variables.</p>  of morbidity</em>. Most of the data collected today, in
   particular by the international </span><a href="http://euroreves/reves"><span lang="EN-GB" style="mso-ansi-language:EN-GB">REVES</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></a>
 <p>The program outputs many files. Most of them are files which  network on Health expectancy, and most HE indices based on these
 will be plotted for better understanding.</p>  data, are <em>cross-sectional</em>. It means that the information
   collected comes from a single cross-sectional survey: people from
 <hr>  various ages (but mostly old people) are surveyed on their health
   status at a single date. Proportion of people disabled at each
 <h2><a name="output"><font color="#00006A">Output of the program  age, can then be measured at that date. This age-specific
 and graphs</font> </a></h2>  prevalence curve is then used to distinguish, within the
   stationary population (which, by definition, is the life table
 <p>Once the optimization is finished, some graphics can be made  estimated from the vital statistics on mortality at the same
 with a grapher. We use Gnuplot which is an interactive plotting  date), the disable population from the disability-free
 program copyrighted but freely distributed. Imach outputs the  population. Life expectancy (LE) (or total population divided by
 source of a gnuplot file, named 'graph.gp', which can be directly  the yearly number of births or deaths of this stationary
 input into gnuplot.<br>  population) is then decomposed into DFLE and DLE. This method of
 When the running is finished, the user should enter a caracter  computing HE is usually called the Sullivan method (from the name
 for plotting and output editing. </p>  of the author who first described it).<o:p></o:p></span></p>
   
 <p>These caracters are:</p>  <p style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">Age-specific proportions of people
   disable are very difficult to forecast because each proportion
 <ul>  corresponds to historical conditions of the cohort and it is the
     <li>'c' to start again the program from the beginning.</li>  result of the historical flows from entering disability and
     <li>'g' to made graphics. The output graphs are in GIF format  recovering in the past until today. The age-specific intensities
         and you have no control over which is produced. If you  (or incidence rates) of entering disability or recovering a good
         want to modify the graphics or make another one, you  health, are reflecting actual conditions and therefore can be
         should modify the parameters in the file <b>graph.gp</b>  used at each age to forecast the future of this cohort. For
         located in imach\bin. A gnuplot reference manual is  example if a country is improving its technology of prosthesis,
         available <a  the incidence of recovering the ability to walk will be higher at
         href="http://www.cs.dartmouth.edu/gnuplot/gnuplot.html">here</a>.  each (old) age, but the prevalence of disability will only
     </li>  slightly reflect an improve because the prevalence is mostly
     <li>'e' opens the <strong>index.htm</strong> file to edit the  affected by the history of the cohort and not by recent period
         output files and graphs. </li>  effects. To measure the period improvement we have to simulate
     <li>'q' for exiting.</li>  the future of a cohort of new-borns entering or leaving at each
 </ul>  age the disability state or dying according to the incidence
   rates measured today on different cohorts. The proportion of
 <h5><font size="4"><strong>Results files </strong></font><br>  people disabled at each age in this simulated cohort will be much
 <br>  lower (using the example of an improvement) that the proportions
 <font color="#EC5E5E" size="3"><strong>- </strong></font><a  observed at each age in a cross-sectional survey. This new
 name="Observed prevalence in each state"><font color="#EC5E5E"  prevalence curve introduced in a life table will give a much more
 size="3"><strong>Observed prevalence in each state</strong></font></a><font  actual and realistic HE level than the Sullivan method which
 color="#EC5E5E" size="3"><strong> (and at first pass)</strong></font><b>:  mostly measured the History of health conditions in this country.<o:p></o:p></span></p>
 </b><a href="prbiaspar.txt"><b>prbiaspar.txt</b></a><br>  
 </h5>  <p style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">Therefore, the main question is how
   to measure incidence rates from cross-longitudinal surveys? This
 <p>The first line is the title and displays each field of the  is the goal of the IMaCH program. From your data and using IMaCH
 file. The first column is age. The fields 2 and 6 are the  you can estimate period HE and not only Sullivan's HE. Also the
 proportion of individuals in states 1 and 2 respectively as  standard errors of the HE are computed.<o:p></o:p></span></p>
 observed during the first exam. Others fields are the numbers of  
 people in states 1, 2 or more. The number of columns increases if  <p style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">A cross-longitudinal survey
 the number of states is higher than 2.<br>  consists in a first survey (&quot;cross&quot;) where individuals
 The header of the file is </p>  from different ages are interviewed on their health status or
   degree of disability. At least a second wave of interviews
 <pre># Age Prev(1) N(1) N Age Prev(2) N(2) N  (&quot;longitudinal&quot;) should measure each new individual
 70 1.00000 631 631 70 0.00000 0 631  health status. Health expectancies are computed from the
 71 0.99681 625 627 71 0.00319 2 627   transitions observed between waves and are computed for each
 72 0.97125 1115 1148 72 0.02875 33 1148 </pre>  degree of severity of disability (number of life states). More
   degrees you consider, more time is necessary to reach the Maximum
 <pre># Age Prev(1) N(1) N Age Prev(2) N(2) N  Likelihood of the parameters involved in the model. Considering
     70 0.95721 604 631 70 0.04279 27 631</pre>  only two states of disability (disable and healthy) is generally
   enough but the computer program works also with more health
 <p>It means that at age 70, the prevalence in state 1 is 1.000  statuses.<span style="mso-spacerun:
 and in state 2 is 0.00 . At age 71 the number of individuals in  yes">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; </span><br>
 state 1 is 625 and in state 2 is 2, hence the total number of  <br>
 people aged 71 is 625+2=627. <br>  The simplest model is the multinomial logistic model where <i>pij</i>
 </p>  is the probability to be observed in state <i>j</i> at the second
   wave conditional to be observed in state <em>i</em> at the first
 <h5><font color="#EC5E5E" size="3"><b>- Estimated parameters and  wave. Therefore a simple model is: log<em>(pij/pii)= aij +
 covariance matrix</b></font><b>: </b><a href="rbiaspar.txt"><b>rbiaspar.txt</b></a></h5>  bij*age+ cij*sex,</em> where '<i>age</i>' is age and '<i>sex</i>'
   is a covariate. The advantage that this computer program claims,
 <p>This file contains all the maximisation results: </p>  comes from that if the delay between waves is not identical for
   each individual, or if some individual missed an interview, the
 <pre> Number of iterations=47  information is not rounded or lost, but taken into account using
  -2 log likelihood=46553.005854373667    an interpolation or extrapolation. <i>hPijx</i> is the
  Estimated parameters: a12 = -12.691743 b12 = 0.095819   probability to be observed in state <i>i</i> at age <i>x+h</i>
                        a13 = -7.815392   b13 = 0.031851   conditional to the observed state <i>i</i> at age <i>x</i>. The
                        a21 = -1.809895 b21 = -0.030470   delay '<i>h</i>' can be split into an exact number (<i>nh*stepm</i>)
                        a23 = -7.838248  b23 = 0.039490    of unobserved intermediate states. This elementary transition (by
  Covariance matrix: Var(a12) = 1.03611e-001  month or quarter trimester, semester or year) is modeled as a
                     Var(b12) = 1.51173e-005  multinomial logistic. The <i>hPx</i> matrix is simply the matrix
                     Var(a13) = 1.08952e-001  product of <i>nh*stepm</i> elementary matrices and the
                     Var(b13) = 1.68520e-005    contribution of each individual to the likelihood is simply <i>hPijx</i>.
                     Var(a21) = 4.82801e-001  <o:p></o:p></span></p>
                     Var(b21) = 6.86392e-005  
                     Var(a23) = 2.27587e-001  <p style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">The program presented in this
                     Var(b23) = 3.04465e-005   manual is a quite general program named <strong>IMaCh</strong>
  </pre>  (for <strong>I</strong>nterpolated <strong>MA</strong>rkov <strong>CH</strong>ain),
   designed to analyse transition data from longitudinal surveys.
 <h5><font color="#EC5E5E" size="3"><b>- Transition probabilities</b></font><b>:  The first step is the parameters estimation of a transition
 </b><a href="pijrbiaspar.txt"><b>pijrbiaspar.txt</b></a></h5>  probabilities model between an initial status and a final status.
   From there, the computer program produces some indicators such as
 <p>Here are the transitions probabilities Pij(x, x+nh) where nh  observed and stationary prevalence, life expectancies and their
 is a multiple of 2 years. The first column is the starting age x  variances and graphs. Our transition model consists in absorbing
 (from age 50 to 100), the second is age (x+nh) and the others are  and non-absorbing states with the possibility of return across
 the transition probabilities p11, p12, p13, p21, p22, p23. For  the non-absorbing states. The main advantage of this package,
 example, line 5 of the file is: </p>  compared to other programs for the analysis of transition data
   (For example: Proc Catmod of SAS<sup>(r)</sup>) is that the whole
 <pre> 100 106 0.03286 0.23512 0.73202 0.02330 0.19210 0.78460 </pre>  individual information is used even if an interview is missing, a
   status or a date is unknown or when the delay between waves is
 <p>and this means: </p>  not identical for each individual. The program can be executed
   according to parameters: selection of a sub-sample, number of
 <pre>p11(100,106)=0.03286  absorbing and non-absorbing states, number of waves taken in
 p12(100,106)=0.23512  account (the user inputs the first and the last interview), a
 p13(100,106)=0.73202  tolerance level for the maximization function, the periodicity of
 p21(100,106)=0.02330  the transitions (we can compute annual, quarterly or monthly
 p22(100,106)=0.19210   transitions), covariates in the model. It works on Windows or on
 p22(100,106)=0.78460 </pre>  Unix.<o:p></o:p></span></p>
   
 <h5><font color="#EC5E5E" size="3"><b>- </b></font><a  <hr>
 name="Stationary prevalence in each state"><font color="#EC5E5E"  
 size="3"><b>Stationary prevalence in each state</b></font></a><b>:  <p><span lang="EN-GB" style="mso-ansi-language:EN-GB">(1) Laditka, Sarah B. and Wolf, Douglas A. (1998), &quot;New
 </b><a href="plrbiaspar.txt"><b>plrbiaspar.txt</b></a></h5>  Methods for Analyzing Active Life Expectancy&quot;. <i>Journal of
   Aging and Health</i>. </span>Vol 10, No. 2. </p>
 <pre>#Age 1-1 2-2   
 70 0.92274 0.07726   <hr>
 71 0.91420 0.08580   
 72 0.90481 0.09519   <h2><a name="data"><span lang="EN-GB" style="color:#00006A;mso-ansi-language:EN-GB">On what kind of data can it be used?</span><span style="mso-bookmark:data"></span><span lang="EN-GB" style="mso-ansi-language:EN-GB"><o:p></o:p></span></a></h2>
 73 0.89453 0.10547</pre>  
   <p style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">The minimum data required for a
 <p>At age 70 the stationary prevalence is 0.92274 in state 1 and  transition model is the recording of a set of individuals
 0.07726 in state 2. This stationary prevalence differs from  interviewed at a first date and interviewed again at least one
 observed prevalence. Here is the point. The observed prevalence  another time. From the observations of an individual, we obtain a
 at age 70 results from the incidence of disability, incidence of  follow-up over time of the occurrence of a specific event. In
 recovery and mortality which occurred in the past of the cohort.  this documentation, the event is related to health status at
 Stationary prevalence results from a simulation with actual  older ages, but the program can be applied on a lot of
 incidences and mortality (estimated from this cross-longitudinal  longitudinal studies in different contexts. To build the data
 survey). It is the best predictive value of the prevalence in the  file explained into the next section, you must have the month and
 future if &quot;nothing changes in the future&quot;. This is  year of each interview and the corresponding health status. But
 exactly what demographers do with a Life table. Life expectancy  in order to get age, date of birth (month and year) is required
 is the expected mean time to survive if observed mortality rates  (missing values is allowed for month). Date of death (month and
 (incidence of mortality) &quot;remains constant&quot; in the  year) is an important information also required if the individual
 future. </p>  is dead. Shorter steps (i.e. a month) will more closely take into
   account the survival time after the last interview.<o:p></o:p></span></p>
 <h5><font color="#EC5E5E" size="3"><b>- Standard deviation of  
 stationary prevalence</b></font><b>: </b><a  <hr>
 href="vplrbiaspar.txt"><b>vplrbiaspar.txt</b></a></h5>  
   <h2><a name="datafile"><span lang="EN-GB" style="color:#00006A;mso-ansi-language:
 <p>The stationary prevalence has to be compared with the observed  EN-GB">The data file</span><span style="mso-bookmark:datafile"></span><span lang="EN-GB" style="mso-ansi-language:EN-GB"><o:p></o:p></span></a></h2>
 prevalence by age. But both are statistical estimates and  
 subjected to stochastic errors due to the size of the sample, the  <p style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">In this example, 8,000 people have
 design of the survey, and, for the stationary prevalence to the  been interviewed in a cross-longitudinal survey of 4 waves (1984,
 model used and fitted. It is possible to compute the standard  1986, 1988, 1990). Some people missed 1, 2 or 3 interviews.
 deviation of the stationary prevalence at each age.</p>  Health statuses are healthy (1) and disable (2). The survey is
   not a real one. It is a simulation of the American Longitudinal
 <h6><font color="#EC5E5E" size="3">Observed and stationary  Survey on Aging. The disability state is defined if the
 prevalence in state (2=disable) with the confident interval</font>:<b>  individual missed one of four ADL (Activity of daily living, like
 vbiaspar2.gif</b></h6>  bathing, eating, walking). Therefore, even is the individuals
   interviewed in the sample are virtual, the information brought
 <p><br>  with this sample is close to the situation of the United States.
 This graph exhibits the stationary prevalence in state (2) with  Sex is not recorded is this sample.<o:p></o:p></span></p>
 the confidence interval in red. The green curve is the observed  
 prevalence (or proportion of individuals in state (2)). Without  <p><span lang="EN-GB" style="mso-ansi-language:EN-GB">Each line of the data set (named </span><a href="data1.txt"><span lang="EN-GB" style="mso-ansi-language:
 discussing the results (it is not the purpose here), we observe  EN-GB">data1.txt</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></a>
 that the green curve is rather below the stationary prevalence.  in this first example) is an individual record which fields are: <o:p></o:p></span></p>
 It suggests an increase of the disability prevalence in the  
 future.</p>  <ul type="disc">
       <li class="MsoNormal"
 <p><img src="vbiaspar2.gif" width="400" height="300"></p>      style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
        mso-list:l12 level1 lfo6;tab-stops:list 36.0pt"><b><span lang="EN-GB" style="mso-ansi-language:EN-GB">Index
 <h6><font color="#EC5E5E" size="3"><b>Convergence to the          number</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></b>: positive number (field 1) <o:p></o:p></span></li>
 stationary prevalence of disability</b></font><b>: pbiaspar1.gif</b><br>      <li class="MsoNormal"
 <img src="pbiaspar1.gif" width="400" height="300"> </h6>      style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
        mso-list:l12 level1 lfo6;tab-stops:list 36.0pt"><b><span lang="EN-GB" style="mso-ansi-language:EN-GB">First
 <p>This graph plots the conditional transition probabilities from          covariate</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></b> positive number (field 2) <o:p></o:p></span></li>
 an initial state (1=healthy in red at the bottom, or 2=disable in      <li class="MsoNormal"
 green on top) at age <em>x </em>to the final state 2=disable<em> </em>at      style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
 age <em>x+h. </em>Conditional means at the condition to be alive       mso-list:l12 level1 lfo6;tab-stops:list 36.0pt"><b><span lang="EN-GB" style="mso-ansi-language:EN-GB">Second
 at age <em>x+h </em>which is <i>hP12x</i> + <em>hP22x</em>. The          covariate</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></b> positive number (field 3) <o:p></o:p></span></li>
 curves <i>hP12x/(hP12x</i> + <em>hP22x) </em>and <i>hP22x/(hP12x</i>      <li class="MsoNormal"
 + <em>hP22x) </em>converge with <em>h, </em>to the <em>stationary      style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
 prevalence of disability</em>. In order to get the stationary       mso-list:l12 level1 lfo6;tab-stops:list 36.0pt"><a
 prevalence at age 70 we should start the process at an earlier          name="Weight"><b><span lang="EN-GB" style="mso-ansi-language:EN-GB">Weight</span><span style="mso-bookmark:Weight"></span><span lang="EN-GB" style="mso-ansi-language:
 age, i.e.50. If the disability state is defined by severe       EN-GB"></b></a>: positive number (field
 disability criteria with only a few chance to recover, then the          4) . In most surveys individuals are weighted according
 incidence of recovery is low and the time to convergence is          to the stratification of the sample.<o:p></o:p></span></li>
 probably longer. But we don't have experience yet.</p>      <li class="MsoNormal"
       style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
 <h5><font color="#EC5E5E" size="3"><b>- Life expectancies by age       mso-list:l12 level1 lfo6;tab-stops:list 36.0pt"><b><span lang="EN-GB" style="mso-ansi-language:EN-GB">Date
 and initial health status</b></font><b>: </b><a          of birth</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></b>: coded as mm/yyyy. Missing dates are coded
 href="erbiaspar.txt"><b>erbiaspar.txt</b></a></h5>          as 99/9999 (field 5) <o:p></o:p></span></li>
       <li class="MsoNormal"
 <pre># Health expectancies       style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
 # Age 1-1 1-2 2-1 2-2        mso-list:l12 level1 lfo6;tab-stops:list 36.0pt"><b><span lang="EN-GB" style="mso-ansi-language:EN-GB">Date
 70 10.7297 2.7809 6.3440 5.9813           of death</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></b>: coded as mm/yyyy. Missing dates are coded
 71 10.3078 2.8233 5.9295 5.9959           as 99/9999 (field 6) <o:p></o:p></span></li>
 72 9.8927 2.8643 5.5305 6.0033       <li class="MsoNormal"
 73 9.4848 2.9036 5.1474 6.0035 </pre>      style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
        mso-list:l12 level1 lfo6;tab-stops:list 36.0pt"><b><span lang="EN-GB" style="mso-ansi-language:EN-GB">Date
 <pre>For example 70 10.7297 2.7809 6.3440 5.9813 means:          of first interview</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></b>: coded as mm/yyyy. Missing dates
 e11=10.7297 e12=2.7809 e21=6.3440 e22=5.9813</pre>          are coded as 99/9999 (field 7) <o:p></o:p></span></li>
       <li class="MsoNormal"
 <pre><img src="exbiaspar1.gif" width="400" height="300"><img      style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
 src="exbiaspar2.gif" width="400" height="300"></pre>       mso-list:l12 level1 lfo6;tab-stops:list 36.0pt"><b><span lang="EN-GB" style="mso-ansi-language:EN-GB">Status
           at first interview</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></b>: positive number. Missing values
 <p>For example, life expectancy of a healthy individual at age 70          ar coded -1. (field 8) <o:p></o:p></span></li>
 is 10.73 in the healthy state and 2.78 in the disability state      <li class="MsoNormal"
 (=13.51 years). If he was disable at age 70, his life expectancy      style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
 will be shorter, 6.34 in the healthy state and 5.98 in the       mso-list:l12 level1 lfo6;tab-stops:list 36.0pt"><b><span lang="EN-GB" style="mso-ansi-language:EN-GB">Date
 disability state (=12.32 years). The total life expectancy is a          of second interview</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></b>: coded as mm/yyyy. Missing dates
 weighted mean of both, 13.51 and 12.32; weight is the proportion          are coded as 99/9999 (field 9) <o:p></o:p></span></li>
 of people disabled at age 70. In order to get a pure period index      <li class="MsoNormal"
 (i.e. based only on incidences) we use the <a      style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
 href="#Stationary prevalence in each state">computed or       mso-list:l12 level1 lfo6;tab-stops:list 36.0pt"><strong><span lang="EN-GB" style="mso-ansi-language:EN-GB">Status
 stationary prevalence</a> at age 70 (i.e. computed from          at second interview</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></strong> positive number. Missing
 incidences at earlier ages) instead of the <a          values ar coded -1. (field 10) <o:p></o:p></span></li>
 href="#Observed prevalence in each state">observed prevalence</a>      <li class="MsoNormal"
 (for example at first exam) (<a href="#Health expectancies">see      style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
 below</a>).</p>       mso-list:l12 level1 lfo6;tab-stops:list 36.0pt"><b><span lang="EN-GB" style="mso-ansi-language:EN-GB">Date
           of third interview</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></b>: coded as mm/yyyy. Missing dates
 <h5><font color="#EC5E5E" size="3"><b>- Variances of life          are coded as 99/9999 (field 11) <o:p></o:p></span></li>
 expectancies by age and initial health status</b></font><b>: </b><a      <li class="MsoNormal"
 href="vrbiaspar.txt"><b>vrbiaspar.txt</b></a></h5>      style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
        mso-list:l12 level1 lfo6;tab-stops:list 36.0pt"><strong><span lang="EN-GB" style="mso-ansi-language:EN-GB">Status
 <p>For example, the covariances of life expectancies Cov(ei,ej)          at third interview</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></strong> positive number. Missing
 at age 50 are (line 3) </p>          values ar coded -1. (field 12) <o:p></o:p></span></li>
       <li class="MsoNormal"
 <pre>   Cov(e1,e1)=0.4667  Cov(e1,e2)=0.0605=Cov(e2,e1)  Cov(e2,e2)=0.0183</pre>      style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
        mso-list:l12 level1 lfo6;tab-stops:list 36.0pt"><b><span lang="EN-GB" style="mso-ansi-language:EN-GB">Date
 <h5><font color="#EC5E5E" size="3"><b>- </b></font><a          of fourth interview</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></b>: coded as mm/yyyy. Missing dates
 name="Health expectancies"><font color="#EC5E5E" size="3"><b>Health          are coded as 99/9999 (field 13) <o:p></o:p></span></li>
 expectancies</b></font></a><font color="#EC5E5E" size="3"><b>      <li class="MsoNormal"
 with standard errors in parentheses</b></font><b>: </b><a      style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
 href="trbiaspar.txt"><font face="Courier New"><b>trbiaspar.txt</b></font></a></h5>       mso-list:l12 level1 lfo6;tab-stops:list 36.0pt"><strong><span lang="EN-GB" style="mso-ansi-language:EN-GB">Status
           at fourth interview</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></strong> positive number. Missing
 <pre>#Total LEs with variances: e.. (std) e.1 (std) e.2 (std) </pre>          values are coded -1. (field 14) <o:p></o:p></span></li>
       <li class="MsoNormal"
 <pre>70 13.42 (0.18) 10.39 (0.15) 3.03 (0.10)70 13.81 (0.18) 11.28 (0.14) 2.53 (0.09) </pre>      style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
        mso-list:l12 level1 lfo6;tab-stops:list 36.0pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">etc<o:p></o:p></span></li>
 <p>Thus, at age 70 the total life expectancy, e..=13.42 years is  </ul>
 the weighted mean of e1.=13.51 and e2.=12.32 by the stationary  
 prevalence at age 70 which are 0.92274 in state 1 and 0.07726 in  <p><span lang="EN-GB" style="mso-ansi-language:EN-GB">&nbsp;<o:p></o:p></span></p>
 state 2, respectively (the sum is equal to one). e.1=10.39 is the  
 Disability-free life expectancy at age 70 (it is again a weighted  <p style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">If your longitudinal survey do not
 mean of e11 and e21). e.2=3.03 is also the life expectancy at age  include information about weights or covariates, you must fill
 70 to be spent in the disability state.</p>  the column with a number (e.g. 1) because a missing field is not
   allowed.<o:p></o:p></span></p>
 <h6><font color="#EC5E5E" size="3"><b>Total life expectancy by  
 age and health expectancies in states (1=healthy) and (2=disable)</b></font><b>:  <hr>
 ebiaspar.gif</b></h6>  
   <h2><span lang="EN-GB" style="color:#00006A;mso-ansi-language:EN-GB">Your first example parameter file</span><a
 <p>This figure represents the health expectancies and the total  href="http://euroreves.ined.fr/imach"></a><a name="uio"><span lang="EN-GB" style="mso-ansi-language:EN-GB"><o:p></o:p></span></a></h2>
 life expectancy with the confident interval in dashed curve. </p>  
   <h2><a name="biaspar"><span lang="EN-GB" style="mso-ansi-language:EN-GB"></a>#Imach version 0.7, February 2002,
 <pre>        <img src="ebiaspar.gif" width="400" height="300"></pre>  INED-EUROREVES <o:p></o:p></span></h2>
   
 <p>Standard deviations (obtained from the information matrix of  <p><span lang="EN-GB" style="mso-ansi-language:EN-GB">This is a comment. Comments start with a '#'.<o:p></o:p></span></p>
 the model) of these quantities are very useful.  
 Cross-longitudinal surveys are costly and do not involve huge  <h4><span lang="EN-GB" style="color:red;mso-ansi-language:EN-GB">First uncommented line</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"><o:p></o:p></span></h4>
 samples, generally a few thousands; therefore it is very  
 important to have an idea of the standard deviation of our  <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">title=1st_example datafile=data1.txt lastobs=8600 firstpass=1 lastpass=4<o:p></o:p></span></pre>
 estimates. It has been a big challenge to compute the Health  
 Expectancy standard deviations. Don't be confuse: life expectancy  <ul type="disc">
 is, as any expected value, the mean of a distribution; but here      <li class="MsoNormal"
 we are not computing the standard deviation of the distribution,      style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
 but the standard deviation of the estimate of the mean.</p>       text-align:justify;mso-list:l1 level1 lfo9;tab-stops:list 36.0pt"><b><span lang="EN-GB" style="mso-ansi-language:EN-GB">title=</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></b>
           1st_example is title of the run. <o:p></o:p></span></li>
 <p>Our health expectancies estimates vary according to the sample      <li class="MsoNormal"
 size (and the standard deviations give confidence intervals of      style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
 the estimate) but also according to the model fitted. Let us       text-align:justify;mso-list:l1 level1 lfo9;tab-stops:list 36.0pt"><b><span lang="EN-GB" style="mso-ansi-language:EN-GB">datafile=</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></b>data1.txt
 explain it in more details.</p>          is the name of the data set. Our example is a six years
           follow-up survey. It consists in a baseline followed by 3
 <p>Choosing a model means ar least two kind of choices. First we          reinterviews. <o:p></o:p></span></li>
 have to decide the number of disability states. Second we have to      <li class="MsoNormal"
 design, within the logit model family, the model: variables,      style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
 covariables, confonding factors etc. to be included.</p>       text-align:justify;mso-list:l1 level1 lfo9;tab-stops:list 36.0pt"><b><span lang="EN-GB" style="mso-ansi-language:EN-GB">lastobs=</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></b>
           8600 the program is able to run on a subsample where the
 <p>More disability states we have, better is our demographical          last observation number is lastobs. It can be set a
 approach of the disability process, but smaller are the number of          bigger number than the real number of observations (e.g.
 transitions between each state and higher is the noise in the          100000). In this example, maximisation will be done on
 measurement. We do not have enough experiments of the various          the 8600 first records. <o:p></o:p></span></li>
 models to summarize the advantages and disadvantages, but it is      <li class="MsoNormal"
 important to say that even if we had huge and unbiased samples,      style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
 the total life expectancy computed from a cross-longitudinal       text-align:justify;mso-list:l1 level1 lfo9;tab-stops:list 36.0pt"><b><span lang="EN-GB" style="mso-ansi-language:EN-GB">firstpass=1</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></b>
 survey, varies with the number of states. If we define only two          , <b>lastpass=4 </b>In case of more than two interviews
 states, alive or dead, we find the usual life expectancy where it          in the survey, the program can be run on selected
 is assumed that at each age, people are at the same risk to die.          transitions periods. firstpass=1 means the first
 If we are differentiating the alive state into healthy and          interview included in the calculation is the baseline
 disable, and as the mortality from the disability state is higher          survey. lastpass=4 means that the information brought by
 than the mortality from the healthy state, we are introducing          the 4th interview is taken into account.<o:p></o:p></span></li>
 heterogeneity in the risk of dying. The total mortality at each  </ul>
 age is the weighted mean of the mortality in each state by the  
 prevalence in each state. Therefore if the proportion of people  <p
 at each age and in each state is different from the stationary  style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">&nbsp;<o:p></o:p></span></p>
 equilibrium, there is no reason to find the same total mortality  
 at a particular age. Life expectancy, even if it is a very useful  <h4
 tool, has a very strong hypothesis of homogeneity of the  style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="color:red;mso-ansi-language:EN-GB">Second
 population. Our main purpose is not to measure differential  uncommented line</span><a name="biaspar-2"><span lang="EN-GB" style="mso-ansi-language:EN-GB"><o:p></o:p></span></a></h4>
 mortality but to measure the expected time in a healthy or  
 disability state in order to maximise the former and minimize the  <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">ftol=1.e-08 stepm=1 ncov=2 nlstate=2 ndeath=1 maxwav=4 mle=1 weight=0<o:p></o:p></span></pre>
 latter. But the differential in mortality complexifies the  
 measurement.</p>  <ul type="disc">
       <li class="MsoNormal"
 <p>Incidences of disability or recovery are not affected by the      style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
 number of states if these states are independant. But incidences       text-align:justify;mso-list:l14 level1 lfo12;tab-stops:list 36.0pt"><b><span lang="EN-GB" style="mso-ansi-language:EN-GB">ftol=1e-8</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></b>
 estimates are dependant on the specification of the model. More          Convergence tolerance on the function value in the
 covariates we added in the logit model better is the model, but          maximisation of the likelihood. Choosing a correct value
 some covariates are not well measured, some are confounding          for ftol is difficult. 1e-8 is a correct value for a 32
 factors like in any statistical model. The procedure to &quot;fit          bits computer.<o:p></o:p></span></li>
 the best model' is similar to logistic regression which itself is      <li class="MsoNormal"
 similar to regression analysis. We haven't yet been sofar because      style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
 we also have a severe limitation which is the speed of the       text-align:justify;mso-list:l14 level1 lfo12;tab-stops:list 36.0pt"><b><span lang="EN-GB" style="mso-ansi-language:EN-GB">stepm=1</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></b>
 convergence. On a Pentium III, 500 MHz, even the simplest model,          Time unit in months for interpolation. Examples:<o:p></o:p></span></li>
 estimated by month on 8,000 people may take 4 hours to converge.      <li><ul type="circle">
 Also, the program is not yet a statistical package, which permits              <li class="MsoNormal"
 a simple writing of the variables and the model to take into              style="mso-margin-top-alt:auto;mso-margin-bottom-alt:
 account in the maximisation. The actual program allows only to        auto;text-align:justify;mso-list:l14 level2 lfo12;tab-stops:list 72.0pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">If
 add simple variables without covariations, like age+sex but                  stepm=1, the unit is a month <o:p></o:p></span></li>
 without age+sex+ age*sex . This can be done from the source code              <li class="MsoNormal"
 (you have to change three lines in the source code) but will              style="mso-margin-top-alt:auto;mso-margin-bottom-alt:
 never be general enough. But what is to remember, is that        auto;text-align:justify;mso-list:l14 level2 lfo12;tab-stops:list 72.0pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">If
 incidences or probability of change from one state to another is                  stepm=4, the unit is a trimester<o:p></o:p></span></li>
 affected by the variables specified into the model.</p>              <li class="MsoNormal"
               style="mso-margin-top-alt:auto;mso-margin-bottom-alt:
 <p>Also, the age range of the people interviewed has a link with        auto;text-align:justify;mso-list:l14 level2 lfo12;tab-stops:list 72.0pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">If
 the age range of the life expectancy which can be estimated by                  stepm=12, the unit is a year <o:p></o:p></span></li>
 extrapolation. If your sample ranges from age 70 to 95, you can              <li class="MsoNormal"
 clearly estimate a life expectancy at age 70 and trust your              style="mso-margin-top-alt:auto;mso-margin-bottom-alt:
 confidence interval which is mostly based on your sample size,        auto;text-align:justify;mso-list:l14 level2 lfo12;tab-stops:list 72.0pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">If
 but if you want to estimate the life expectancy at age 50, you                  stepm=24, the unit is two years<o:p></o:p></span></li>
 should rely in your model, but fitting a logistic model on a age              <li class="MsoNormal"
 range of 70-95 and estimating probabilties of transition out of              style="mso-margin-top-alt:auto;mso-margin-bottom-alt:
 this age range, say at age 50 is very dangerous. At least you        auto;text-align:justify;mso-list:l14 level2 lfo12;tab-stops:list 72.0pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">...
 should remember that the confidence interval given by the  <o:p></o:p></span>            </li>
 standard deviation of the health expectancies, are under the          </ul>
 strong assumption that your model is the 'true model', which is      </li>
 probably not the case.</p>      <li class="MsoNormal"
       style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
 <h5><font color="#EC5E5E" size="3"><b>- Copy of the parameter       text-align:justify;mso-list:l14 level1 lfo12;tab-stops:list 36.0pt"><b><span lang="EN-GB" style="mso-ansi-language:EN-GB">ncov=2</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></b>
 file</b></font><b>: </b><a href="orbiaspar.txt"><b>orbiaspar.txt</b></a></h5>          Number of covariates in the datafile. The intercept and
           the age parameter are counting for 2 covariates.<o:p></o:p></span></li>
 <p>This copy of the parameter file can be useful to re-run the      <li class="MsoNormal"
 program while saving the old output files. </p>      style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
        text-align:justify;mso-list:l14 level1 lfo12;tab-stops:list 36.0pt"><b><span lang="EN-GB" style="mso-ansi-language:EN-GB">nlstate=2</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></b>
 <hr>          Number of non-absorbing (alive) states. Here we have two
           alive states: disability-free is coded 1 and disability
 <h2><a name="example" </a><font color="#00006A">Trying an example</font></a></h2>          is coded 2. <o:p></o:p></span></li>
       <li class="MsoNormal"
 <p>Since you know how to run the program, it is time to test it      style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
 on your own computer. Try for example on a parameter file named <a       text-align:justify;mso-list:l14 level1 lfo12;tab-stops:list 36.0pt"><b><span lang="EN-GB" style="mso-ansi-language:EN-GB">ndeath=1</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></b>
 href="file://../mytry/imachpar.txt">imachpar.txt</a> which is a          Number of absorbing states. The absorbing state death is
 copy of <font size="2" face="Courier New">mypar.txt</font>          coded 3. <o:p></o:p></span></li>
 included in the subdirectory of imach, <font size="2"      <li class="MsoNormal"
 face="Courier New">mytry</font>. Edit it to change the name of      style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
 the data file to <font size="2" face="Courier New">..\data\mydata.txt</font>       text-align:justify;mso-list:l14 level1 lfo12;tab-stops:list 36.0pt"><b><span lang="EN-GB" style="mso-ansi-language:EN-GB">maxwav=4</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></b>
 if you don't want to copy it on the same directory. The file <font          Number of waves in the datafile.<o:p></o:p></span></li>
 face="Courier New">mydata.txt</font> is a smaller file of 3,000      <li class="MsoNormal"
 people but still with 4 waves. </p>      style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
        text-align:justify;mso-list:l14 level1 lfo12;tab-stops:list 36.0pt"><a
 <p>Click on the imach.exe icon to open a window. Answer to the          name="mle"><b><span lang="EN-GB" style="mso-ansi-language:EN-GB">mle</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></b></a><b>=1</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></b> Option for the
 question:'<strong>Enter the parameter file name:'</strong></p>          Maximisation Likelihood Estimation. <o:p></o:p></span></li>
       <li><ul type="circle">
 <table border="1">              <li class="MsoNormal"
     <tr>              style="mso-margin-top-alt:auto;mso-margin-bottom-alt:
         <td width="100%"><strong>IMACH, Version 0.63</strong><p><strong>Enter        auto;text-align:justify;mso-list:l14 level2 lfo12;tab-stops:list 72.0pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">If
         the parameter file name: ..\mytry\imachpar.txt</strong></p>                  mle=1 the program does the maximisation and the
         </td>                  calculation of health expectancies <o:p></o:p></span></li>
     </tr>              <li class="MsoNormal"
 </table>              style="mso-margin-top-alt:auto;mso-margin-bottom-alt:
         auto;text-align:justify;mso-list:l14 level2 lfo12;tab-stops:list 72.0pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">If
 <p>Most of the data files or image files generated, will use the                  mle=0 the program only does the calculation of
 'imachpar' string into their name. The running time is about 2-3                  the health expectancies. <o:p></o:p></span></li>
 minutes on a Pentium III. If the execution worked correctly, the          </ul>
 outputs files are created in the current directory, and should be      </li>
 the same as the mypar files initially included in the directory <font      <li class="MsoNormal"
 size="2" face="Courier New">mytry</font>.</p>      style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
        text-align:justify;mso-list:l14 level1 lfo12;tab-stops:list 36.0pt"><b><span lang="EN-GB" style="mso-ansi-language:EN-GB">weight=0</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></b>
 <ul>          Possibility to add weights. <o:p></o:p></span></li>
     <li><pre><u>Output on the screen</u> The output screen looks like <a      <li><ul type="circle">
 href="imachrun.LOG">this Log file</a>              <li class="MsoNormal"
 #              style="mso-margin-top-alt:auto;mso-margin-bottom-alt:
         auto;text-align:justify;mso-list:l14 level2 lfo12;tab-stops:list 72.0pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">If
 title=MLE datafile=..\data\mydata.txt lastobs=3000 firstpass=1 lastpass=3                  weight=0 no weights are included <o:p></o:p></span></li>
 ftol=1.000000e-008 stepm=24 ncov=2 nlstate=2 ndeath=1 maxwav=4 mle=1 weight=0</pre>              <li class="MsoNormal"
     </li>              style="mso-margin-top-alt:auto;mso-margin-bottom-alt:
     <li><pre>Total number of individuals= 2965, Agemin = 70.00, Agemax= 100.92        auto;text-align:justify;mso-list:l14 level2 lfo12;tab-stops:list 72.0pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">If
                   weight=1 the maximisation integrates the weights
 Warning, no any valid information for:126 line=126                  which are in field </span><a href="#Weight"><span lang="EN-GB" style="mso-ansi-language:EN-GB">4</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"><o:p></o:p></span></a></li>
 Warning, no any valid information for:2307 line=2307          </ul>
 Delay (in months) between two waves Min=21 Max=51 Mean=24.495826      </li>
 <font face="Times New Roman">These lines give some warnings on the data file and also some raw statistics on frequencies of transitions.</font>  </ul>
 Age 70 1.=230 loss[1]=3.5% 2.=16 loss[2]=12.5% 1.=222 prev[1]=94.1% 2.=14  
  prev[2]=5.9% 1-1=8 11=200 12=7 13=15 2-1=2 21=6 22=7 23=1  <h4
 Age 102 1.=0 loss[1]=NaNQ% 2.=0 loss[2]=NaNQ% 1.=0 prev[1]=NaNQ% 2.=0 </pre>  style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="color:red;mso-ansi-language:EN-GB">Covariates</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"><o:p></o:p></span></h4>
     </li>  
 </ul>  <p
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">Intercept
 <p>&nbsp;</p>  and age are systematically included in the model. Additional
   covariates can be included with the command <o:p></o:p></span></p>
 <ul>  
     <li>Maximisation with the Powell algorithm. 8 directions are  <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">model=<em>list of covariates<o:p></o:p></span></em></pre>
         given corresponding to the 8 parameters. this can be  
         rather long to get convergence.<br>  <ul type="disc">
         <font size="1" face="Courier New"><br>      <li class="MsoNormal"
         Powell iter=1 -2*LL=11531.405658264877 1 0.000000000000 2      style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
         0.000000000000 3<br>       text-align:justify;mso-list:l2 level1 lfo15;tab-stops:list 36.0pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">if<strong>
         0.000000000000 4 0.000000000000 5 0.000000000000 6          model=. </strong>then no covariates are included<o:p></o:p></span></li>
         0.000000000000 7 <br>      <li class="MsoNormal"
         0.000000000000 8 0.000000000000<br>      style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
         1..........2.................3..........4.................5.........<br>       text-align:justify;mso-list:l2 level1 lfo15;tab-stops:list 36.0pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">if
         6................7........8...............<br>          <strong>model=V1</strong> the model includes the first
         Powell iter=23 -2*LL=6744.954108371555 1 -12.967632334283          covariate (field 2)<o:p></o:p></span></li>
         <br>      <li class="MsoNormal"
         2 0.135136681033 3 -7.402109728262 4 0.067844593326 <br>      style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
         5 -0.673601538129 6 -0.006615504377 7 -5.051341616718 <br>       text-align:justify;mso-list:l2 level1 lfo15;tab-stops:list 36.0pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">if
         8 0.051272038506<br>          <strong>model=V2 </strong>the model includes the second
         1..............2...........3..............4...........<br>          covariate (field 3)<o:p></o:p></span></li>
         5..........6................7...........8.........<br>      <li class="MsoNormal"
         #Number of iterations = 23, -2 Log likelihood =      style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
         6744.954042573691<br>       text-align:justify;mso-list:l2 level1 lfo15;tab-stops:list 36.0pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">if
         # Parameters<br>          <strong>model=V1+V2 </strong>the model includes the first
         12 -12.966061 0.135117 <br>          and the second covariate (fields 2 and 3)<o:p></o:p></span></li>
         13 -7.401109 0.067831 <br>      <li class="MsoNormal"
         21 -0.672648 -0.006627 <br>      style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
         23 -5.051297 0.051271 </font><br>       text-align:justify;mso-list:l2 level1 lfo15;tab-stops:list 36.0pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">if
         </li>          <strong>model=V1*V2 </strong>the model includes the
     <li><pre><font size="2">Calculation of the hessian matrix. Wait...          product of the first and the second covariate (fields 2
 12345678.12.13.14.15.16.17.18.23.24.25.26.27.28.34.35.36.37.38.45.46.47.48.56.57.58.67.68.78          and 3)<o:p></o:p></span></li>
       <li class="MsoNormal"
 Inverting the hessian to get the covariance matrix. Wait...      style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
        text-align:justify;mso-list:l2 level1 lfo15;tab-stops:list 36.0pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">if
 #Hessian matrix#          <strong>model=V1+V1*age</strong> the model includes the
 3.344e+002 2.708e+004 -4.586e+001 -3.806e+003 -1.577e+000 -1.313e+002 3.914e-001 3.166e+001           product covariate*age<o:p></o:p></span></li>
 2.708e+004 2.204e+006 -3.805e+003 -3.174e+005 -1.303e+002 -1.091e+004 2.967e+001 2.399e+003   </ul>
 -4.586e+001 -3.805e+003 4.044e+002 3.197e+004 2.431e-002 1.995e+000 1.783e-001 1.486e+001   
 -3.806e+003 -3.174e+005 3.197e+004 2.541e+006 2.436e+000 2.051e+002 1.483e+001 1.244e+003   <h4
 -1.577e+000 -1.303e+002 2.431e-002 2.436e+000 1.093e+002 8.979e+003 -3.402e+001 -2.843e+003   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="color:red;mso-ansi-language:EN-GB">Guess
 -1.313e+002 -1.091e+004 1.995e+000 2.051e+002 8.979e+003 7.420e+005 -2.842e+003 -2.388e+005   values for optimisation</span><span lang="EN-GB" style="color:#00006A;mso-ansi-language:EN-GB"> </span><span lang="EN-GB" style="mso-ansi-language:EN-GB"><o:p></o:p></span></h4>
 3.914e-001 2.967e+001 1.783e-001 1.483e+001 -3.402e+001 -2.842e+003 1.494e+002 1.251e+004   
 3.166e+001 2.399e+003 1.486e+001 1.244e+003 -2.843e+003 -2.388e+005 1.251e+004 1.053e+006   <p
 # Scales  style="tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">You
 12 1.00000e-004 1.00000e-006  must write the initial guess values of the parameters for
 13 1.00000e-004 1.00000e-006  optimisation. The number of parameters, <em>N</em> depends on the
 21 1.00000e-003 1.00000e-005  number of absorbing states and non-absorbing states and on the
 23 1.00000e-004 1.00000e-005  number of covariates. <br>
 # Covariance  <em>N</em> is given by the formula <em>N</em>=(<em>nlstate</em> +
   1 5.90661e-001  <em>ndeath</em>-1)*<em>nlstate</em>*<em>ncov</em>&nbsp;. <br>
   2 -7.26732e-003 8.98810e-005  <br>
   3 8.80177e-002 -1.12706e-003 5.15824e-001  Thus in the simple case with 2 covariates (the model is log
   4 -1.13082e-003 1.45267e-005 -6.50070e-003 8.23270e-005  (pij/pii) = aij + bij * age where intercept and age are the two
   5 9.31265e-003 -1.16106e-004 6.00210e-004 -8.04151e-006 1.75753e+000  covariates), and 2 health degrees (1 for disability-free and 2
   6 -1.15664e-004 1.44850e-006 -7.79995e-006 1.04770e-007 -2.12929e-002 2.59422e-004  for disability) and 1 absorbing state (3), you must enter 8
   7 1.35103e-003 -1.75392e-005 -6.38237e-004 7.85424e-006 4.02601e-001 -4.86776e-003 1.32682e+000  initials values, a12, b12, a13, b13, a21, b21, a23, b23. You can
   8 -1.82421e-005 2.35811e-007 7.75503e-006 -9.58687e-008 -4.86589e-003 5.91641e-005 -1.57767e-002 1.88622e-004  start with zeros as in this example, but if you have a more
 # agemin agemax for lifexpectancy, bage fage (if mle==0 ie no data nor Max likelihood).  precise set (for example from an earlier run) you can enter it
   and it will speed up them<br>
   Each of the four lines starts with indices &quot;ij&quot;: <b>ij
 agemin=70 agemax=100 bage=50 fage=100  aij bij</b> <o:p></o:p></span></p>
 Computing prevalence limit: result on file 'plrmypar.txt'   
 Computing pij: result on file 'pijrmypar.txt'   <pre
 Computing Health Expectancies: result on file 'ermypar.txt'   style="margin-top:0cm;margin-right:36.0pt;margin-bottom:0cm;margin-left:
 Computing Variance-covariance of DFLEs: file 'vrmypar.txt'   36.0pt;margin-bottom:.0001pt;text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB"># Guess values of aij and bij in log (pij/pii) = aij + bij * age<o:p></o:p></span></pre>
 Computing Total LEs with variances: file 'trmypar.txt'   
 Computing Variance-covariance of Prevalence limit: file 'vplrmypar.txt'   <pre
 End of Imach  style="margin-top:0cm;margin-right:36.0pt;margin-bottom:0cm;margin-left:36.0pt;
 </font></pre>  margin-bottom:.0001pt;text-align:justify"><span lang="EN-GB" style="mso-ansi-language:
     </li>  EN-GB">12 -14.155633<span style="mso-spacerun: yes">&nbsp; </span>0.110794 <o:p></o:p></span></pre>
 </ul>  
   <pre
 <p><font size="3">Once the running is finished, the program  style="margin-top:0cm;margin-right:36.0pt;margin-bottom:0cm;margin-left:36.0pt;
 requires a caracter:</font></p>  margin-bottom:.0001pt;text-align:justify"><span lang="EN-GB" style="mso-ansi-language:
   EN-GB">13<span style="mso-spacerun: yes">&nbsp; </span>-7.925360<span style="mso-spacerun: yes">&nbsp; </span>0.032091 <o:p></o:p></span></pre>
 <table border="1">  
     <tr>  <pre
         <td width="100%"><strong>Type g for plotting (available  style="margin-top:0cm;margin-right:36.0pt;margin-bottom:0cm;margin-left:36.0pt;
         if mle=1), e to edit output files, c to start again,</strong><p><strong>and  margin-bottom:.0001pt;text-align:justify"><span lang="EN-GB" style="mso-ansi-language:
         q for exiting:</strong></p>  EN-GB">21<span style="mso-spacerun: yes">&nbsp; </span>-1.890135 -0.029473 <o:p></o:p></span></pre>
         </td>  
     </tr>  <pre
 </table>  style="margin-top:0cm;margin-right:36.0pt;margin-bottom:0cm;margin-left:36.0pt;
   margin-bottom:.0001pt;text-align:justify"><span lang="EN-GB" style="mso-ansi-language:
 <p><font size="3">First you should enter <strong>g</strong> to  EN-GB">23<span style="mso-spacerun: yes">&nbsp; </span>-6.234642<span style="mso-spacerun: yes">&nbsp; </span>0.022315 <o:p></o:p></span></pre>
 make the figures and then you can edit all the results by typing <strong>e</strong>.  
 </font></p>  <p
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">or,
 <ul>  to simplify: <o:p></o:p></span></p>
     <li><u>Outputs files</u> <br>  
         - index.htm, this file is the master file on which you  <pre
         should click first.<br>  style="margin-top:0cm;margin-right:36.0pt;margin-bottom:0cm;margin-left:
         - Observed prevalence in each state: <a  36.0pt;margin-bottom:.0001pt;text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">12 0.0 0.0<o:p></o:p></span></pre>
         href="..\mytry\prmypar.txt">mypar.txt</a> <br>  
         - Estimated parameters and the covariance matrix: <a  <pre
         href="..\mytry\rmypar.txt">rmypar.txt</a> <br>  style="margin-top:0cm;margin-right:36.0pt;margin-bottom:0cm;margin-left:36.0pt;
         - Stationary prevalence in each state: <a  margin-bottom:.0001pt;text-align:justify"><span lang="EN-GB" style="mso-ansi-language:
         href="..\mytry\plrmypar.txt">plrmypar.txt</a> <br>  EN-GB">13 0.0 0.0<o:p></o:p></span></pre>
         - Transition probabilities: <a  
         href="..\mytry\pijrmypar.txt">pijrmypar.txt</a> <br>  <pre
         - Copy of the parameter file: <a  style="margin-top:0cm;margin-right:
         href="..\mytry\ormypar.txt">ormypar.txt</a> <br>  36.0pt;margin-bottom:0cm;margin-left:36.0pt;margin-bottom:.0001pt;text-align:
         - Life expectancies by age and initial health status: <a  justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">21 0.0 0.0<o:p></o:p></span></pre>
         href="..\mytry\ermypar.txt">ermypar.txt</a> <br>  
         - Variances of life expectancies by age and initial  <pre
         health status: <a href="..\mytry\vrmypar.txt">vrmypar.txt</a>  style="margin-top:0cm;margin-right:36.0pt;margin-bottom:0cm;margin-left:36.0pt;
         <br>  margin-bottom:.0001pt;text-align:justify"><span lang="EN-GB" style="mso-ansi-language:
         - Health expectancies with their variances: <a  EN-GB">23 0.0 0.0<o:p></o:p></span></pre>
         href="..\mytry\trmypar.txt">trmypar.txt</a> <br>  
         - Standard deviation of stationary prevalence: <a  <h4
         href="..\mytry\vplrmypar.txt">vplrmypar.txt</a> <br>  style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="color:red;mso-ansi-language:EN-GB">Guess
         <br>  values for computing variances</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"><o:p></o:p></span></h4>
         </li>  
     <li><u>Graphs</u> <br>  <p
         <br>  style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">This
         -<a href="..\mytry\vmypar1.gif">Observed and stationary  is an output if </span><a href="#mle"><span lang="EN-GB" style="mso-ansi-language:EN-GB">mle</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></a>=1. But it can be used as
         prevalence in state (1) with the confident interval</a> <br>  an input to get the various output data files (Health
         -<a href="..\mytry\vmypar2.gif">Observed and stationary  expectancies, stationary prevalence etc.) and figures without
         prevalence in state (2) with the confident interval</a> <br>  rerunning the rather long maximisation phase (mle=0). <o:p></o:p></span></p>
         -<a href="..\mytry\exmypar1.gif">Health life expectancies  
         by age and initial health state (1)</a> <br>  <p
         -<a href="..\mytry\exmypar2.gif">Health life expectancies  style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">The
         by age and initial health state (2)</a> <br>  scales are small values for the evaluation of numerical
         -<a href="..\mytry\emypar.gif">Total life expectancy by  derivatives. These derivatives are used to compute the hessian
         age and health expectancies in states (1) and (2).</a> </li>  matrix of the parameters, that is the inverse of the covariance
 </ul>  matrix, and the variances of health expectancies. Each line
   consists in indices &quot;ij&quot; followed by the initial scales
 <p>This software have been partly granted by <a  (zero to simplify) associated with aij and bij. <o:p></o:p></span></p>
 href="http://euroreves.ined.fr">Euro-REVES</a>, a concerted  
 action from the European Union. It will be copyrighted  <ul type="disc">
 identically to a GNU software product, i.e. program and software      <li class="MsoNormal"
 can be distributed freely for non commercial use. Sources are not      style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
 widely distributed today. You can get them by asking us with a       text-align:justify;mso-list:l16 level1 lfo18;tab-stops:list 36.0pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">If
 simple justification (name, email, institute) <a          mle=1 you can enter zeros:<o:p></o:p></span></li>
 href="mailto:brouard@ined.fr">mailto:brouard@ined.fr</a> and <a  </ul>
 href="mailto:lievre@ined.fr">mailto:lievre@ined.fr</a> .</p>  
   <pre
 <p>Latest version (0.63 of 16 march 2000) can be accessed at <a  style="margin-top:0cm;margin-right:36.0pt;margin-bottom:0cm;margin-left:
 href="http://euroeves.ined.fr/imach">http://euroreves.ined.fr/imach</a><br>  36.0pt;margin-bottom:.0001pt;text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB"># Scales (for hessian or gradient estimation)<o:p></o:p></span></pre>
 </p>  
 </body>  <pre
 </html>  style="margin-top:0cm;margin-right:36.0pt;margin-bottom:0cm;margin-left:36.0pt;
   margin-bottom:.0001pt;text-align:justify"><span lang="EN-GB" style="mso-ansi-language:
   EN-GB">12 0. 0. <o:p></o:p></span></pre>
   
   <pre
   style="margin-top:0cm;margin-right:
   36.0pt;margin-bottom:0cm;margin-left:36.0pt;margin-bottom:.0001pt;text-align:
   justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">13 0. 0. <o:p></o:p></span></pre>
   
   <pre
   style="margin-top:0cm;margin-right:36.0pt;margin-bottom:0cm;margin-left:36.0pt;
   margin-bottom:.0001pt;text-align:justify"><span lang="EN-GB" style="mso-ansi-language:
   EN-GB">21 0. 0. <o:p></o:p></span></pre>
   
   <pre
   style="margin-top:0cm;margin-right:
   36.0pt;margin-bottom:0cm;margin-left:36.0pt;margin-bottom:.0001pt;text-align:
   justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">23 0. 0. <o:p></o:p></span></pre>
   
   <ul type="disc">
       <li class="MsoNormal"
       style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
        text-align:justify;mso-list:l11 level1 lfo21;tab-stops:list 36.0pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">If
           mle=0 you must enter a covariance matrix (usually
           obtained from an earlier run).<o:p></o:p></span></li>
   </ul>
   
   <h4
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="color:red;mso-ansi-language:EN-GB">Covariance
   matrix of parameters</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"><o:p></o:p></span></h4>
   
   <p
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">This
   is an output if </span><a href="#mle"><span lang="EN-GB" style="mso-ansi-language:EN-GB">mle</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></a>=1. But it can be used as
   an input to get the various output data files (Health
   expectancies, stationary prevalence etc.) and figures without
   rerunning the rather long maximisation phase (mle=0). <o:p></o:p></span></p>
   
   <p
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">Each
   line starts with indices &quot;ijk&quot; followed by the
   covariances between aij and bij: <o:p></o:p></span></p>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">&nbsp;<o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB"><span style="mso-spacerun: yes">&nbsp;&nbsp; </span>121 Var(a12) <o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB"><span style="mso-spacerun: yes">&nbsp;&nbsp;&nbsp;</span>122 Cov(b12,a12)<span style="mso-spacerun: yes">&nbsp; </span>Var(b12) <o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB"><span style="mso-spacerun: yes">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;</span>...<o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB"><span style="mso-spacerun: yes">&nbsp;&nbsp; </span>232 Cov(b23,a12)<span style="mso-spacerun: yes">&nbsp; </span>Cov(b23,b12) ... Var (b23) <o:p></o:p></span></pre>
   
   <ul type="disc">
       <li class="MsoNormal"
       style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
        text-align:justify;mso-list:l18 level1 lfo24;tab-stops:list 36.0pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">If
           mle=1 you can enter zeros. <o:p></o:p></span></li>
   </ul>
   
   <pre
   style="margin-top:0cm;margin-right:36.0pt;margin-bottom:0cm;margin-left:
   36.0pt;margin-bottom:.0001pt;text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB"># Covariance matrix<o:p></o:p></span></pre>
   
   <pre
   style="margin-top:0cm;margin-right:36.0pt;margin-bottom:0cm;margin-left:36.0pt;
   margin-bottom:.0001pt;text-align:justify"><span lang="EN-GB" style="mso-ansi-language:
   EN-GB">121 0.<o:p></o:p></span></pre>
   
   <pre
   style="margin-top:0cm;margin-right:
   36.0pt;margin-bottom:0cm;margin-left:36.0pt;margin-bottom:.0001pt;text-align:
   justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">122 0. 0.<o:p></o:p></span></pre>
   
   <pre
   style="margin-top:0cm;margin-right:36.0pt;margin-bottom:0cm;margin-left:36.0pt;
   margin-bottom:.0001pt;text-align:justify"><span lang="EN-GB" style="mso-ansi-language:
   EN-GB">131 0. 0. 0. <o:p></o:p></span></pre>
   
   <pre
   style="margin-top:0cm;
   margin-right:36.0pt;margin-bottom:0cm;margin-left:36.0pt;margin-bottom:.0001pt;
   text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">132 0. 0. 0. 0. <o:p></o:p></span></pre>
   
   <pre
   style="margin-top:0cm;margin-right:36.0pt;margin-bottom:0cm;margin-left:36.0pt;
   margin-bottom:.0001pt;text-align:justify"><span lang="EN-GB" style="mso-ansi-language:
   EN-GB">211 0. 0. 0. 0. 0. <o:p></o:p></span></pre>
   
   <pre
   style="margin-top:0cm;
   margin-right:36.0pt;margin-bottom:0cm;margin-left:36.0pt;margin-bottom:.0001pt;
   text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">212 0. 0. 0. 0. 0. 0. <o:p></o:p></span></pre>
   
   <pre
   style="margin-top:0cm;margin-right:36.0pt;margin-bottom:0cm;margin-left:36.0pt;
   margin-bottom:.0001pt;text-align:justify"><span lang="EN-GB" style="mso-ansi-language:
   EN-GB">231 0. 0. 0. 0. 0. 0. 0. <o:p></o:p></span></pre>
   
   <pre
   style="margin-top:
   0cm;margin-right:36.0pt;margin-bottom:0cm;margin-left:36.0pt;margin-bottom:
   .0001pt;text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">232 0. 0. 0. 0. 0. 0. 0. 0.<o:p></o:p></span></pre>
   
   <ul type="disc">
       <li class="MsoNormal"
       style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
        text-align:justify;mso-list:l7 level1 lfo27;tab-stops:list 36.0pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">If
           mle=0 you must enter a covariance matrix (usually
           obtained from an earlier run).<o:p></o:p></span></li>
   </ul>
   
   <h4
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="color:red;mso-ansi-language:EN-GB">Age
   range for calculation of stationary prevalences and health
   expectancies</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"><o:p></o:p></span></h4>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">agemin=70 agemax=100 bage=50 fage=100<o:p></o:p></span></pre>
   
   <p
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">Once
   we obtained the estimated parameters, the program is able to
   calculated stationary prevalence, transitions probabilities and
   life expectancies at any age. Choice of age range is useful for
   extrapolation. In our data file, ages varies from age 70 to 102.
   Setting bage=50 and fage=100, makes the program computing life
   expectancy from age bage to age fage. As we use a model, we can
   compute life expectancy on a wider age range than the age range
   from the data. But the model can be rather wrong on big
   intervals.<o:p></o:p></span></p>
   
   <p
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">Similarly,
   it is possible to get extrapolated stationary prevalence by age
   ranging from agemin to agemax. <o:p></o:p></span></p>
   
   <ul type="disc">
       <li class="MsoNormal"
       style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
        text-align:justify;mso-list:l13 level1 lfo30;tab-stops:list 36.0pt"><b><span lang="EN-GB" style="mso-ansi-language:EN-GB">agemin=</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></b>
           Minimum age for calculation of the stationary prevalence <o:p></o:p></span></li>
       <li class="MsoNormal"
       style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
        text-align:justify;mso-list:l13 level1 lfo30;tab-stops:list 36.0pt"><b><span lang="EN-GB" style="mso-ansi-language:EN-GB">agemax=</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></b>
           Maximum age for calculation of the stationary prevalence <o:p></o:p></span></li>
       <li class="MsoNormal"
       style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
        text-align:justify;mso-list:l13 level1 lfo30;tab-stops:list 36.0pt"><b><span lang="EN-GB" style="mso-ansi-language:EN-GB">bage=</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></b>
           Minimum age for calculation of the health expectancies <o:p></o:p></span></li>
       <li class="MsoNormal"
       style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
        text-align:justify;mso-list:l13 level1 lfo30;tab-stops:list 36.0pt"><b><span lang="EN-GB" style="mso-ansi-language:EN-GB">fage=</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></b>
           Maximum age for calculation of the health expectancies <o:p></o:p></span></li>
   </ul>
   
   <h4
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><a
   name="Computing"><span lang="EN-GB" style="color:red;mso-ansi-language:EN-GB">Computing</span><span lang="EN-GB" style="color:red;mso-ansi-language:EN-GB"></a> the observed prevalence</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"><o:p></o:p></span></h4>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">begin-prev-date=1/1/1984 end-prev-date=1/6/1988 <o:p></o:p></span></pre>
   
   <p
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">Statements
   'begin-prev-date' and 'end-prev-date' allow to select the period
   in which we calculate the observed prevalences in each state. In
   this example, the prevalences are calculated on data survey
   collected between 1 January 1984 and 1 June 1988. <o:p></o:p></span></p>
   
   <ul type="disc">
       <li class="MsoNormal"
       style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
        text-align:justify;mso-list:l3 level1 lfo33;tab-stops:list 36.0pt"><strong><span lang="EN-GB" style="mso-ansi-language:EN-GB">begin-prev-date=
   </span><span lang="EN-GB" style="mso-ansi-language:EN-GB">        </strong>Starting date (day/month/year)<o:p></o:p></span></li>
       <li class="MsoNormal"
       style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
        text-align:justify;mso-list:l3 level1 lfo33;tab-stops:list 36.0pt"><strong><span lang="EN-GB" style="mso-ansi-language:EN-GB">end-prev-date=
   </span><span lang="EN-GB" style="mso-ansi-language:EN-GB">        </strong>Final date (day/month/year)<o:p></o:p></span></li>
   </ul>
   
   <h4
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="color:red;mso-ansi-language:EN-GB">Population-
   or status-based health expectancies</span><span lang="EN-GB" style="mso-ansi-language:
   EN-GB"><o:p></o:p></span></h4>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">pop_based=0<o:p></o:p></span></pre>
   
   <p
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">The
   user has the possibility to choose between population-based or
   status-based health expectancies. If pop_based=0 then
   status-based health expectancies are computed and if pop_based=1,
   the programme computes population-based health expectancies.
   Health expectancies are weighted averages of health expectancies
   respective of the initial state. For a status-based index, the
   weights are the cross-sectional prevalences observed between two
   dates, as </span><a href="#Computing"><span lang="EN-GB" style="mso-ansi-language:EN-GB">previously explained</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></a>, whereas
   for a population-based index, the weights are the stationary
   prevalences.<o:p></o:p></span></p>
   
   <h4
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="color:red;mso-ansi-language:EN-GB">Prevalence
   forecasting </span><span lang="EN-GB" style="mso-ansi-language:EN-GB"><o:p></o:p></span></h4>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">starting-proj-date=1/1/1989 final-proj-date=1/1/1992 mov_average=0 <o:p></o:p></span></pre>
   
   <p
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">Prevalence
   and population projections are available only if the
   interpolation unit is a month, i.e. stepm=1. The programme
   estimates the prevalence in each state at a precise date
   expressed in day/month/year. The programme computes one
   forecasted prevalence a year from a starting date (1 January of
   1989 in this example) to a final date (1 January 1992). The
   statement mov_average allows to compute smoothed forecasted
   prevalences with a five-age moving average centred at the mid-age
   of the five-age period. <o:p></o:p></span></p>
   
   <ul type="disc">
       <li class="MsoNormal"
       style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
        text-align:justify;mso-list:l10 level1 lfo36;tab-stops:list 36.0pt"><strong><span lang="EN-GB" style="mso-ansi-language:EN-GB">starting-proj-date</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></strong>=
           starting date (day/month/year) of forecasting<o:p></o:p></span></li>
       <li class="MsoNormal"
       style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
        text-align:justify;mso-list:l10 level1 lfo36;tab-stops:list 36.0pt"><strong><span lang="EN-GB" style="mso-ansi-language:EN-GB">final-proj-date=
   </span><span lang="EN-GB" style="mso-ansi-language:EN-GB">        </strong>final date (day/month/year) of forecasting<o:p></o:p></span></li>
       <li class="MsoNormal"
       style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
        text-align:justify;mso-list:l10 level1 lfo36;tab-stops:list 36.0pt"><strong><span lang="EN-GB" style="mso-ansi-language:EN-GB">mov_average</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></strong>=
           smoothing with a five-age moving average centred at the
           mid-age of the five-age period. The command<strong>
           mov_average</strong> takes value 1 if the prevalences are
           smoothed and 0 otherwise.<o:p></o:p></span></li>
   </ul>
   
   <h4
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="color:red;mso-ansi-language:EN-GB">Last
   uncommented line : Population forecasting </span><span lang="EN-GB" style="mso-ansi-language:EN-GB"><o:p></o:p></span></h4>
   
   <pre><span lang="EN-GB" style="mso-ansi-language:EN-GB">popforecast=0 popfile=pyram.txt popfiledate=1/1/1989 last-popfiledate=1/1/1992<o:p></o:p></span></pre>
   
   <p
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">This
   command is available if the interpolation unit is a month, i.e.
   stepm=1 and if popforecast=1. From a data file including age and
   number of persons alive at the precise date &#145;</span><span lang="EN-GB" style="font-size:10.0pt;mso-bidi-font-size:12.0pt;font-family:&quot;Courier New&quot;;
   mso-ansi-language:EN-GB">popfiledate&#146;,
   </span><span lang="EN-GB" style="mso-ansi-language:EN-GB">you can forecast the number of persons in each state until date</span><span lang="EN-GB" style="font-size:10.0pt;mso-bidi-font-size:
   12.0pt;font-family:&quot;Courier New&quot;;mso-ansi-language:EN-GB">
   &#145;last-popfiledate&#146;. </span><span lang="EN-GB" style="mso-ansi-language:EN-GB">In this example, the popfile </span><a
   href="pyram.txt"><b><span lang="EN-GB" style="mso-ansi-language:EN-GB">pyram.txt</span><span style="mso-ansi-language:EN-GB"></b></a><b> </span><span lang="EN-GB" style="mso-ansi-language:
   EN-GB"><span style="mso-spacerun: yes"></b>&nbsp;</span>includes real
   data which are the Japanese population in 1989.<span style="mso-spacerun: yes">&nbsp; </span><o:p></o:p></span></p>
   
   <ul type="disc">
       <li class="MsoNormal"
       style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
        text-align:justify;mso-list:l10 level1 lfo36;tab-stops:list 36.0pt"><b><span lang="EN-GB" style="mso-ansi-language:EN-GB">popforecast=
           0</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></b> Option for population forecasting. If
           popforecast=1, the programme does the forecasting<b>.<o:p></o:p></span></b></li>
       <li class="MsoNormal"
       style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
        text-align:justify;mso-list:l10 level1 lfo36;tab-stops:list 36.0pt"><b><span lang="EN-GB" style="mso-ansi-language:EN-GB">popfile=
   </span><span lang="EN-GB" style="mso-ansi-language:EN-GB">        </b>name of the population file<o:p></o:p></span></li>
       <li class="MsoNormal"
       style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
        text-align:justify;mso-list:l10 level1 lfo36;tab-stops:list 36.0pt"><b><span lang="EN-GB" style="mso-ansi-language:EN-GB">popfiledate=</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></b>
           date of the population population<o:p></o:p></span></li>
       <li class="MsoNormal"
       style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
        text-align:justify;mso-list:l10 level1 lfo36;tab-stops:list 36.0pt"><b><span lang="EN-GB" style="mso-ansi-language:EN-GB">last-popfiledate</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></b>=
           date of the last population projection&nbsp;<o:p></o:p></span></li>
   </ul>
   
   <hr>
   
   <h2
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><a
   name="running"><span lang="EN-GB" style="color:#00006A;mso-ansi-language:EN-GB"></a>Running Imach with this example</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"><o:p></o:p></span></h2>
   
   <p
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">We
   assume that you entered your </span><a href="biaspar.imach"><span lang="EN-GB" style="mso-ansi-language:EN-GB">1st_example
   parameter file</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></a> as explained </span><a href="#biaspar"><span lang="EN-GB" style="mso-ansi-language:
   EN-GB">above</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></a>. To
   run the program you should click on the imach.exe icon and enter
   the name of the parameter file which is for example </span><a
   href="..\mle\biaspar.txt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">C:\usr\imach\mle\biaspar.txt</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></a> (you
   also can click on the biaspar.txt icon located in </span><a
   href="..\mle"><span lang="EN-GB" style="mso-ansi-language:EN-GB">C:\usr\imach\mle</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></a> and put it with the mouse on
   the imach window).<o:p></o:p></span></p>
   
   <p
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">The
   time to converge depends on the step unit that you used (1 month
   is cpu consuming), on the number of cases, and on the number of
   variables.<o:p></o:p></span></p>
   
   <p
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">The
   program outputs many files. Most of them are files which will be
   plotted for better understanding.<o:p></o:p></span></p>
   
   <hr>
   
   <h2
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><a
   name="output"><span lang="EN-GB" style="color:#00006A;mso-ansi-language:EN-GB">Output of the program and graphs</span><span style="mso-bookmark:output"><span lang="EN-GB" style="mso-ansi-language:EN-GB"></a> </span></span><span lang="EN-GB" style="mso-ansi-language:EN-GB"><o:p></o:p></span></h2>
   
   <p
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">Once
   the optimization is finished, some graphics can be made with a
   grapher. We use Gnuplot which is an interactive plotting program
   copyrighted but freely distributed. A gnuplot reference manual is
   available </span><a href="http://www.gnuplot.info/"><span lang="EN-GB" style="mso-ansi-language:EN-GB">here</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></a>. <br>
   When the running is finished, the user should enter a character
   for plotting and output editing. <o:p></o:p></span></p>
   
   <p
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">These
   characters are:<o:p></o:p></span></p>
   
   <ul type="disc">
       <li class="MsoNormal"
       style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
        text-align:justify;mso-list:l0 level1 lfo41;tab-stops:list 36.0pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">'c'
           to start again the program from the beginning.<o:p></o:p></span></li>
       <li class="MsoNormal"
       style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
        text-align:justify;mso-list:l0 level1 lfo41;tab-stops:list 36.0pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">'e'
           opens the </span><a href="biaspar.htm"><strong><span lang="EN-GB" style="mso-ansi-language:EN-GB">biaspar.htm</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></strong></a>
           file to edit the output files and graphs. <o:p></o:p></span></li>
       <li class="MsoNormal"
       style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
        text-align:justify;mso-list:l0 level1 lfo41;tab-stops:list 36.0pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">'q'
           for exiting.<o:p></o:p></span></li>
   </ul>
   
   <h5
   style="tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="font-size:18.0pt;mso-bidi-font-size:10.0pt;color:#00006A;
   mso-ansi-language:EN-GB">Results
   files</span><strong><span lang="EN-GB" style="font-size:13.5pt;mso-ansi-language:EN-GB"> </span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></strong><br>
   <br>
   </span><strong><span lang="EN-GB" style="font-size:12.0pt;color:#EC5E5E;
   mso-ansi-language:EN-GB">- </strong><a name="Observed_prevalence_in_each_state"><strong>Observed
   prevalence in each state</strong></a><strong> (and at first pass)</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></strong>:
   </span><a href="prbiaspar.txt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">prbiaspar.txt</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"><o:p></o:p></span></a></h5>
   
   <p
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">The
   first line is the title and displays each field of the file. The
   first column is age. The fields 2 and 6 are the proportion of
   individuals in states 1 and 2 respectively as observed during the
   first exam. Others fields are the numbers of people in states 1,
   2 or more. The number of columns increases if the number of
   states is higher than 2.<br>
   The header of the file is <o:p></o:p></span></p>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB"># Age Prev(1) N(1) N Age Prev(2) N(2) N<o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">70 1.00000 631 631 70 0.00000 0 631<o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">71 0.99681 625 627 71 0.00319 2 627 <o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">72 0.97125 1115 1148 72 0.02875 33 1148 <o:p></o:p></span></pre>
   
   <p
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">It
   means that at age 70, the prevalence in state 1 is 1.000 and in
   state 2 is 0.00 . At age 71 the number of individuals in state 1
   is 625 and in state 2 is 2, hence the total number of people aged
   71 is 625+2=627. <o:p></o:p></span></p>
   
   <h5
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="font-size:12.0pt;color:#EC5E5E;mso-ansi-language:EN-GB">-
   Estimated parameters and covariance matrix</span><span lang="EN-GB" style="mso-ansi-language:EN-GB">: </span><a
   href="rbiaspar.txt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">rbiaspar.txt</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"><o:p></o:p></span></a></h5>
   
   <p
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">This
   file contains all the maximisation results: <o:p></o:p></span></p>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB"><span style="mso-spacerun: yes">&nbsp;</span>-2 log likelihood= 21660.918613445392<o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB"> Estimated parameters: a12 = -12.290174 b12 = 0.092161 <o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB"><span style="mso-spacerun: yes">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;</span><span style="mso-spacerun: yes">&nbsp;</span>a13 = -9.155590<span style="mso-spacerun: yes">&nbsp; </span>b13 = 0.046627 <o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB"><span style="mso-spacerun: yes">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;</span>a21 = -2.629849<span style="mso-spacerun: yes">&nbsp; </span>b21 = -0.022030 <o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB"><span style="mso-spacerun: yes">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;</span>a23 = -7.958519<span style="mso-spacerun: yes">&nbsp; </span>b23 = 0.042614<span style="mso-spacerun: yes">&nbsp; </span><o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB"><span style="mso-spacerun: yes">&nbsp;</span>Covariance matrix: Var(a12) = 1.47453e-001<o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB"><span style="mso-spacerun: yes">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; </span>Var(b12) = 2.18676e-005<o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB"><span style="mso-spacerun: yes">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; </span>Var(a13) = 2.09715e-001<o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB"><span style="mso-spacerun: yes">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; </span>Var(b13) = 3.28937e-005<span style="mso-spacerun: yes">&nbsp; </span><o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB"><span style="mso-spacerun: yes">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;</span>Var(a21) = 9.19832e-001<o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB"><span style="mso-spacerun: yes">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; </span>Var(b21) = 1.29229e-004<o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB"><span style="mso-spacerun: yes">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; </span></span><span lang="DE" style="mso-ansi-language:DE">Var(a23) = 4.48405e-001<o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="DE" style="mso-ansi-language:DE"><span style="mso-spacerun: yes">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; </span>Var(b23) = 5.85631e-005 <o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="DE" style="mso-ansi-language:DE"><span style="mso-spacerun: yes">&nbsp;</span><o:p></o:p></span></pre>
   
   <p
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">By
   substitution of these parameters in the regression model, we
   obtain the elementary transition probabilities:<o:p></o:p></span></p>
   
   <p
   style="tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><img
   src="pebiaspar1.gif" width="400" height="300" id="_x0000_i1037"></p>
   
   <h5
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="font-size:12.0pt;color:#EC5E5E;mso-ansi-language:EN-GB">-
   Transition probabilities</span><span lang="EN-GB" style="mso-ansi-language:EN-GB">: </span><a href="pijrbiaspar.txt"><span lang="EN-GB" style="mso-ansi-language:
   EN-GB">pijrbiaspar.txt</span><span lang="EN-GB" style="mso-ansi-language:
   EN-GB"><o:p></o:p></span></a></h5>
   
   <p
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">Here
   are the transitions probabilities Pij(x, x+nh) where nh is a
   multiple of 2 years. The first column is the starting age x (from
   age 50 to 100), the second is age (x+nh) and the others are the
   transition probabilities p11, p12, p13, p21, p22, p23. For
   example, line 5 of the file is: <o:p></o:p></span></p>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB"><span style="mso-spacerun: yes">&nbsp;</span>100 106 0.02655 0.17622 0.79722 0.01809 0.13678 0.84513 <o:p></o:p></span></pre>
   
   <p
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">and
   this means: <o:p></o:p></span></p>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">p11(100,106)=0.02655<o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">p12(100,106)=0.17622<o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">p13(100,106)=0.79722<o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">p21(100,106)=0.01809<o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">p22(100,106)=0.13678<o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">p22(100,106)=0.84513 <o:p></o:p></span></pre>
   
   <h5
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="font-size:12.0pt;color:#EC5E5E;mso-ansi-language:EN-GB">-
   <a name="Stationary_prevalence_in_each_state">Stationary
   prevalence in each state</span><span style="mso-bookmark:Stationary_prevalence_in_each_state"></span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></a>: </span><a href="plrbiaspar.txt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">plrbiaspar.txt</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"><o:p></o:p></span></a></h5>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">#Prevalence<o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">#Age 1-1 2-2<o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">&nbsp;<o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">#************ <o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">70 0.90134 0.09866<o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">71 0.89177 0.10823 <o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">72 0.88139 0.11861 <o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">73 0.87015 0.12985 <o:p></o:p></span></pre>
   
   <p
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">At
   age 70 the stationary prevalence is 0.90134 in state 1 and
   0.09866 in state 2. This stationary prevalence differs from
   observed prevalence. Here is the point. The observed prevalence
   at age 70 results from the incidence of disability, incidence of
   recovery and mortality which occurred in the past of the cohort.
   Stationary prevalence results from a simulation with actual
   incidences and mortality (estimated from this cross-longitudinal
   survey). It is the best predictive value of the prevalence in the
   future if &quot;nothing changes in the future&quot;. This is
   exactly what demographers do with a Life table. Life expectancy
   is the expected mean time to survive if observed mortality rates
   (incidence of mortality) &quot;remains constant&quot; in the
   future. <o:p></o:p></span></p>
   
   <h5
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="font-size:12.0pt;color:#EC5E5E;mso-ansi-language:EN-GB">-
   Standard deviation of stationary prevalence</span><span lang="EN-GB" style="mso-ansi-language:EN-GB">: </span><a
   href="vplrbiaspar.txt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">vplrbiaspar.txt</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"><o:p></o:p></span></a></h5>
   
   <p
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">The
   stationary prevalence has to be compared with the observed
   prevalence by age. But both are statistical estimates and
   subjected to stochastic errors due to the size of the sample, the
   design of the survey, and, for the stationary prevalence to the
   model used and fitted. It is possible to compute the standard
   deviation of the stationary prevalence at each age.<o:p></o:p></span></p>
   
   <h5
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="font-size:12.0pt;color:#EC5E5E;mso-ansi-language:EN-GB">-Observed
   and stationary prevalence in state (2=disable) with the confident
   interval</span><span lang="EN-GB" style="mso-ansi-language:EN-GB">: </span><a href="vbiaspar21.htm"><span lang="EN-GB" style="mso-ansi-language:EN-GB">vbiaspar21.gif</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"><o:p></o:p></span></a></h5>
   
   <p
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">This
   graph exhibits the stationary prevalence in state (2) with the
   confidence interval in red. The green curve is the observed
   prevalence (or proportion of individuals in state (2)). Without
   discussing the results (it is not the purpose here), we observe
   that the green curve is rather below the stationary prevalence.
   It suggests an increase of the disability prevalence in the
   future.<o:p></o:p></span></p>
   
   <p
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><img
   src="vbiaspar21.gif" width="400" height="300" id="_x0000_i1038"></p>
   
   <h5
   style="tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="font-size:12.0pt;color:#EC5E5E;mso-ansi-language:EN-GB">-Convergence
   to the stationary prevalence of disability</span><span lang="EN-GB" style="mso-ansi-language:EN-GB">: </span><a
   href="pbiaspar11.gif"><span lang="EN-GB" style="mso-ansi-language:EN-GB">pbiaspar11.gif</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></a><br>
   </span><img src="pbiaspar11.gif" width="400" height="300"
   id="_x0000_i1039"><span lang="EN-GB" style="mso-ansi-language:EN-GB"><o:p></o:p></span></h5>
   
   <p
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">This
   graph plots the conditional transition probabilities from an
   initial state (1=healthy in red at the bottom, or 2=disable in
   green on top) at age <em>x </em>to the final state 2=disable<em> </em>at
   age <em>x+h. </em>Conditional means at the condition to be alive
   at age <em>x+h </em>which is <i>hP12x</i> + <em>hP22x</em>. The
   curves <i>hP12x/(hP12x</i> + <em>hP22x) </em>and <i>hP22x/(hP12x</i>
   + <em>hP22x) </em>converge with <em>h, </em>to the <em>stationary
   prevalence of disability</em>. In order to get the stationary
   prevalence at age 70 we should start the process at an earlier
   age, i.e.50. If the disability state is defined by severe
   disability criteria with only a few chance to recover, then the
   incidence of recovery is low and the time to convergence is
   probably longer. But we don't have experience yet.<o:p></o:p></span></p>
   
   <h5
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="font-size:12.0pt;color:#EC5E5E;mso-ansi-language:EN-GB">-
   Life expectancies by age and initial health status</span><span lang="EN-GB" style="mso-ansi-language:EN-GB">: </span><a
   href="erbiaspar.txt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">erbiaspar.txt</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"><o:p></o:p></span></a></h5>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB"># Health expectancies <o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB"># Age 1-1 1-2 2-1 2-2 <o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">70 10.9226 3.0401 5.6488 6.2122 <o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">71 10.4384 3.0461 5.2477 6.1599 <o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">72 9.9667 3.0502 4.8663 6.1025 <o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">73 9.5077 3.0524 4.5044 6.0401 <o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">For example 70 10.9226 3.0401 5.6488 6.2122 means:<o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="DE" style="mso-ansi-language:DE">e11=10.9226 e12=3.0401 e21=5.6488 e22=6.2122<o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><img src="expbiaspar21.gif"
   width="400" height="300" id="_x0000_i1040"><img
   src="expbiaspar11.gif" width="400" height="300" id="_x0000_i1041"></pre>
   
   <p
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">For
   example, life expectancy of a healthy individual at age 70 is
   10.92 in the healthy state and 3.04 in the disability state
   (=13.96 years). If he was disable at age 70, his life expectancy
   will be shorter, 5.64 in the healthy state and 6.21 in the
   disability state (=11.85 years). The total life expectancy is a
   weighted mean of both, 13.96 and 11.85; weight is the proportion
   of people disabled at age 70. In order to get a pure period index
   (i.e. based only on incidences) we use the </span><a
   href="#Stationary prevalence in each state"><span lang="EN-GB" style="mso-ansi-language:EN-GB">computed or
   stationary prevalence</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></a> at age 70 (i.e. computed from
   incidences at earlier ages) instead of the </span><a
   href="#Observed prevalence in each state"><span lang="EN-GB" style="mso-ansi-language:
   EN-GB">observed prevalence</span><span lang="EN-GB" style="mso-ansi-language:
   EN-GB"></a>
   (for example at first exam) (</span><a href="#Health expectancies"><span lang="EN-GB" style="mso-ansi-language:EN-GB">see
   below</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></a>).<o:p></o:p></span></p>
   
   <h5
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="font-size:12.0pt;color:#EC5E5E;mso-ansi-language:EN-GB">-
   Variances of life expectancies by age and initial health status</span><span lang="EN-GB" style="mso-ansi-language:EN-GB">: </span><a
   href="vrbiaspar.txt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">vrbiaspar.txt</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"><o:p></o:p></span></a></h5>
   
   <p
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">For
   example, the covariances of life expectancies Cov(ei,ej) at age
   50 are (line 3) <o:p></o:p></span></p>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB"><span style="mso-spacerun: yes">&nbsp;&nbsp; </span></span><span lang="DE" style="mso-ansi-language:DE">Cov(e1,e1)=0.4776<span style="mso-spacerun: yes">&nbsp; </span>Cov(e1,e2)=0.0488=Cov(e2,e1)<span style="mso-spacerun: yes">&nbsp; </span>Cov(e2,e2)=0.0424<o:p></o:p></span></pre>
   
   <h5
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="font-size:12.0pt;color:#EC5E5E;mso-ansi-language:EN-GB">-
   <a name="Health_expectancies">Health expectancies</a> with
   standard errors in parentheses</span><span lang="EN-GB" style="mso-ansi-language:EN-GB">: </span><a href="trbiaspar.txt"><span lang="EN-GB" style="font-family:&quot;Courier New&quot;;
   mso-ansi-language:EN-GB">trbiaspar.txt</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"><o:p></o:p></span></a></h5>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">#Total LEs with variances: e.. (std) e.1 (std) e.2 (std) <o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">70 13.76 (0.22) 10.40 (0.20) 3.35 (0.14) <o:p></o:p></span></pre>
   
   <p
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">Thus,
   at age 70 the total life expectancy, e..=13.76years is the
   weighted mean of e1.=13.96 and e2.=11.85 by the stationary
   prevalence at age 70 which are 0.90134 in state 1 and 0.09866 in
   state 2, respectively (the sum is equal to one). e.1=10.40 is the
   Disability-free life expectancy at age 70 (it is again a weighted
   mean of e11 and e21). e.2=3.35 is also the life expectancy at age
   70 to be spent in the disability state.<o:p></o:p></span></p>
   
   <h5
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="font-size:12.0pt;color:#EC5E5E;mso-ansi-language:EN-GB">-Total
   life expectancy by age and health expectancies in states
   (1=healthy) and (2=disable)</span><span lang="EN-GB" style="mso-ansi-language:EN-GB">: </span><a href="ebiaspar1.gif"><span lang="EN-GB" style="mso-ansi-language:EN-GB">ebiaspar1.gif</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"><o:p></o:p></span></a></h5>
   
   <p
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">This
   figure represents the health expectancies and the total life
   expectancy with the confident interval in dashed curve. <o:p></o:p></span></p>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB"><span style="mso-spacerun: yes">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; </span></span><img
   src="ebiaspar1.gif" width="400" height="300" id="_x0000_i1042"></pre>
   
   <p
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">Standard
   deviations (obtained from the information matrix of the model) of
   these quantities are very useful. Cross-longitudinal surveys are
   costly and do not involve huge samples, generally a few
   thousands; therefore it is very important to have an idea of the
   standard deviation of our estimates. It has been a big challenge
   to compute the Health Expectancy standard deviations. Don't be
   confuse: life expectancy is, as any expected value, the mean of a
   distribution; but here we are not computing the standard
   deviation of the distribution, but the standard deviation of the
   estimate of the mean.<o:p></o:p></span></p>
   
   <p
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">Our
   health expectancies estimates vary according to the sample size
   (and the standard deviations give confidence intervals of the
   estimate) but also according to the model fitted. Let us explain
   it in more details.<o:p></o:p></span></p>
   
   <p
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">Choosing
   a model means at least two kind of choices. First we have to
   decide the number of disability states. Second we have to design,
   within the logit model family, the model: variables, covariables,
   confounding factors etc. to be included.<o:p></o:p></span></p>
   
   <p
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">More
   disability states we have, better is our demographical approach
   of the disability process, but smaller are the number of
   transitions between each state and higher is the noise in the
   measurement. We do not have enough experiments of the various
   models to summarize the advantages and disadvantages, but it is
   important to say that even if we had huge and unbiased samples,
   the total life expectancy computed from a cross-longitudinal
   survey, varies with the number of states. If we define only two
   states, alive or dead, we find the usual life expectancy where it
   is assumed that at each age, people are at the same risk to die.
   If we are differentiating the alive state into healthy and
   disable, and as the mortality from the disability state is higher
   than the mortality from the healthy state, we are introducing
   heterogeneity in the risk of dying. The total mortality at each
   age is the weighted mean of the mortality in each state by the
   prevalence in each state. Therefore if the proportion of people
   at each age and in each state is different from the stationary
   equilibrium, there is no reason to find the same total mortality
   at a particular age. Life expectancy, even if it is a very useful
   tool, has a very strong hypothesis of homogeneity of the
   population. Our main purpose is not to measure differential
   mortality but to measure the expected time in a healthy or
   disability state in order to maximise the former and minimize the
   latter. But the differential in mortality complexifies the
   measurement.<o:p></o:p></span></p>
   
   <p
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">Incidences
   of disability or recovery are not affected by the number of
   states if these states are independant. But incidences estimates
   are dependant on the specification of the model. More covariates
   we added in the logit model better is the model, but some
   covariates are not well measured, some are confounding factors
   like in any statistical model. The procedure to &quot;fit the
   best model' is similar to logistic regression which itself is
   similar to regression analysis. We haven't yet been so far
   because we also have a severe limitation which is the speed of
   the convergence. On a Pentium III, 500 MHz, even the simplest
   model, estimated by month on 8,000 people may take 4 hours to
   converge. Also, the program is not yet a statistical package,
   which permits a simple writing of the variables and the model to
   take into account in the maximisation. The actual program allows
   only to add simple variables like age+sex or age+sex+ age*sex but
   will never be general enough. But what is to remember, is that
   incidences or probability of change from one state to another is
   affected by the variables specified into the model.<o:p></o:p></span></p>
   
   <p
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">Also,
   the age range of the people interviewed has a link with the age
   range of the life expectancy which can be estimated by
   extrapolation. If your sample ranges from age 70 to 95, you can
   clearly estimate a life expectancy at age 70 and trust your
   confidence interval which is mostly based on your sample size,
   but if you want to estimate the life expectancy at age 50, you
   should rely in your model, but fitting a logistic model on a age
   range of 70-95 and estimating probabilities of transition out of
   this age range, say at age 50 is very dangerous. At least you
   should remember that the confidence interval given by the
   standard deviation of the health expectancies, are under the
   strong assumption that your model is the 'true model', which is
   probably not the case.<o:p></o:p></span></p>
   
   <h5
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="font-size:12.0pt;color:#EC5E5E;mso-ansi-language:EN-GB">-
   Copy of the parameter file</span><span lang="EN-GB" style="mso-ansi-language:
   EN-GB">: </span><a href="orbiaspar.txt"><span lang="EN-GB" style="mso-ansi-language:
   EN-GB">orbiaspar.txt</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"><o:p></o:p></span></a></h5>
   
   <p
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">This
   copy of the parameter file can be useful to re-run the program
   while saving the old output files. <o:p></o:p></span></p>
   
   <h5
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="font-size:12.0pt;color:#EC5E5E;mso-ansi-language:EN-GB">-
   Prevalence forecasting</span><span lang="EN-GB" style="mso-ansi-language:EN-GB">: </span><a href="frbiaspar.txt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">frbiaspar.txt</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"><o:p></o:p></span></a></h5>
   
   <p
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">First,
   we have estimated the observed prevalence between 1/1/1984 and
   1/6/1988. <span style="mso-spacerun:
   yes">&nbsp;</span>The mean date of interview (weighed average of
   the interviews performed between1/1/1984 and 1/6/1988) is
   estimated to be 13/9/1985, as written on the top on the file.
   Then we forecast the probability to be in each state. <o:p></o:p></span></p>
   
   <p
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">Example,
   at date 1/1/1989 : <o:p></o:p></span></p>
   
   <p class="MsoNormal"><span lang="DE" style="mso-ansi-language:DE"># StartingAge FinalAge P.1 P.2 P.3<o:p></o:p></span></p>
   
   <p class="MsoNormal"><span lang="EN-GB" style="mso-ansi-language:EN-GB"># Forecasting at date 1/1/1989 <o:p></o:p></span></p>
   
   <p class="MsoNormal"><span lang="EN-GB" style="mso-ansi-language:EN-GB">73 0.807 0.078 0.115 <o:p></o:p></span></p>
   
   <p
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">Since
   the minimum age is 70 on the 13/9/1985, the youngest forecasted
   age is 73. This means that at age a person aged 70 at 13/9/1989
   has a probability to enter state1 of 0.807 at age 73 on 1/1/1989.
   Similarly, the probability to be in state 2 is 0.078 and the
   probability to die is 0.115. Then, on the 1/1/1989, the
   prevalence of disability at age 73 is estimated to be 0.088.<o:p></o:p></span></p>
   
   <h5
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="font-size:12.0pt;color:#EC5E5E;mso-ansi-language:EN-GB">-
   Population forecasting</span><span lang="EN-GB" style="mso-ansi-language:EN-GB">: </span><a href="poprbiaspar.txt"><span lang="EN-GB" style="mso-ansi-language:
   EN-GB">poprbiaspar.txt</span><span lang="EN-GB" style="mso-ansi-language:
   EN-GB"><o:p></o:p></span></a></h5>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB"># Age P.1 P.2 P.3 [Population]<o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB"># Forecasting at date 1/1/1989 <o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">75 572685.22 83798.08 <o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">74 621296.51 79767.99 <o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">73 645857.70 69320.60 <o:p></o:p></span></pre>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB"># Forecasting at date 1/1/1990<o:p></o:p></span></pre>
   
   <pre style="text-align:justify">76 442986.68 92721.14 120775.48</pre>
   
   <pre style="text-align:justify">75 487781.02 91367.97 121915.51</pre>
   
   <pre style="text-align:justify">74 512892.07 85003.47 117282.76 </pre>
   
   <pre style="text-align:justify">&nbsp;<o:p></o:p></pre>
   
   <p class="MsoNormal"><span lang="EN-GB" style="mso-ansi-language:EN-GB">From the population file, we estimate the
   number of people in each state. At age 73, 645857 persons are in
   state 1 and 69320 are in state 2. One year latter, 512892 are
   still in state 1, 85003 are in state 2 and 117282 died before
   1/1/1990.<o:p></o:p></span></p>
   
   <pre style="text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">&nbsp;<o:p></o:p></span></pre>
   
   <hr>
   
   <h2
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><a
   name="example"><span lang="EN-GB" style="color:#00006A;mso-ansi-language:EN-GB"></a>Trying an example</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"><o:p></o:p></span></h2>
   
   <p
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">Since
   you know how to run the program, it is time to test it on your
   own computer. Try for example on a parameter file named </span><a
   href="..\mytry\imachpar.txt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">imachpar.txt</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></a> which is a copy of </span><span lang="EN-GB" style="font-size:10.0pt;font-family:&quot;Courier New&quot;;mso-ansi-language:EN-GB">mypar.txt</span><span lang="EN-GB" style="mso-ansi-language:EN-GB">
   included in the subdirectory of imach, </span><span lang="EN-GB" style="font-size:10.0pt;font-family:&quot;Courier New&quot;;
   mso-ansi-language:EN-GB">mytry</span><span lang="EN-GB" style="mso-ansi-language:
   EN-GB">. Edit it to change
   the name of the data file to </span><span lang="EN-GB" style="font-size:10.0pt;font-family:&quot;Courier New&quot;;mso-ansi-language:
   EN-GB">..\data\mydata.txt</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"> if you don't want
   to copy it on the same directory. The file </span><span lang="EN-GB" style="font-family:&quot;Courier New&quot;;mso-ansi-language:EN-GB">mydata.txt</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"> is a
   smaller file of 3,000 people but still with 4 waves. <o:p></o:p></span></p>
   
   <p
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">Click
   on the imach.exe icon to open a window. Answer to the question: '<strong>Enter
   the parameter file name:'<o:p></o:p></span></strong></p>
   
   <table border="1" cellpadding="0"
   style="mso-cellspacing:1.5pt;mso-padding-alt:
    0cm 0cm 0cm 0cm">
       <tr>
           <td width="100%"
           style="width:100.0%;padding:.75pt .75pt .75pt .75pt"><strong><span lang="EN-GB" style="mso-ansi-language:EN-GB">IMACH,
           Version 0.7</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"><o:p></o:p></span></strong><p style="text-align:justify"><strong><span lang="EN-GB" style="mso-ansi-language:
     EN-GB">Enter
           the parameter file name: ..\mytry\imachpar.txt</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"><o:p></o:p></span></strong></p>
           </td>
       </tr>
   </table>
   
   <p
   style="tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">Most
   of the data files or image files generated, will use the
   'imachpar' string into their name. The running time is about 2-3
   minutes on a Pentium III. If the execution worked correctly, the
   outputs files are created in the current directory, and should be
   the same as the mypar files initially included in the directory </span><span lang="EN-GB" style="font-size:10.0pt;font-family:&quot;Courier New&quot;;mso-ansi-language:EN-GB">mytry</span><span lang="EN-GB" style="mso-ansi-language:EN-GB">.<o:p></o:p></span></p>
   
   <pre
   style="margin-left:36.0pt;text-indent:-18.0pt;mso-list:l5 level1 lfo43"><span lang="EN-GB" style="font-family:Symbol;mso-ansi-language:EN-GB">·<span style="font:7.0pt &quot;Times New Roman&quot;">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; </span></span><u><span lang="EN-GB" style="mso-ansi-language:EN-GB">Output on the screen</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></u> The output screen looks like </span><a
   href="imachrun.LOG"><span lang="EN-GB" style="mso-ansi-language:EN-GB">this Log file</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"><o:p></o:p></span></a></pre>
   
   <pre style="margin-left:18.0pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">&nbsp;<o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">#title=MLE datafile=..\data\mydata.txt lastobs=3000 firstpass=1 lastpass=3<o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">ftol=1.000000e-008 stepm=24 ncov=2 nlstate=2 ndeath=1 maxwav=4 mle=1 weight=0<o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">Total number of individuals= 2965, Agemin = 70.00, Agemax= 100.92<o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">&nbsp;<o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">Warning, no any valid information for:126 line=126<o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">Warning, no any valid information for:2307 line=2307<o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt;text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">Delay (in months) between two waves Min=21 Max=51 Mean=24.495826<o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt;text-align:justify"><span lang="EN-GB" style="font-family:&quot;Times New Roman&quot;;mso-ansi-language:EN-GB">These lines give some warnings on the data file and also some raw statistics on frequencies of transitions.</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"><o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt;text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">Age 70 1.=230 loss[1]=3.5% 2.=16 loss[2]=12.5% 1.=222 prev[1]=94.1% 2.=14<o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt;text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB"> prev[2]=5.9% 1-1=8 11=200 12=7 13=15 2-1=2 21=6 22=7 23=1<o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt;text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">Age 102 1.=0 loss[1]=NaNQ% 2.=0 loss[2]=NaNQ% 1.=0 prev[1]=NaNQ% 2.=0 <o:p></o:p></span></pre>
   
   <ul type="disc">
       <li class="MsoNormal"
       style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
        mso-list:l6 level1 lfo46;tab-stops:list 36.0pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">Maximisation
           with the Powell algorithm. 8 directions are given
           corresponding to the 8 parameters. This can be rather
           long to get convergence.<br>
   </span><span lang="EN-GB" style="font-size:7.5pt;font-family:&quot;Courier New&quot;;
        mso-ansi-language:EN-GB">        <br>
           Powell iter=1 -2*LL=11531.405658264877 1 0.000000000000 2
           0.000000000000 3<br>
           0.000000000000 4 0.000000000000 5 0.000000000000 6
           0.000000000000 7 <br>
           0.000000000000 8 0.000000000000<br>
           1..........2.................3..........4.................5.........<br>
           6................7........8...............<br>
           Powell iter=23 -2*LL=6744.954108371555 1 -12.967632334283
           <br>
           2 0.135136681033 3 -7.402109728262 4 0.067844593326 <br>
           5 -0.673601538129 6 -0.006615504377 7 -5.051341616718 <br>
           8 0.051272038506<br>
           1..............2...........3..............4...........<br>
           5..........6................7...........8.........<br>
           #Number of iterations = 23, -2 Log likelihood =
           6744.954042573691<br>
           # Parameters<br>
           12 -12.966061 0.135117 <br>
           13 -7.401109 0.067831 <br>
           21 -0.672648 -0.006627 <br>
           23 -5.051297 0.051271 </span><span lang="EN-GB" style="mso-ansi-language:
        EN-GB"><o:p></o:p></span></li>
   </ul>
   
   <pre
   style="margin-left:36.0pt;text-align:justify;text-indent:-18.0pt;
   mso-list:l6 level1 lfo46"><span lang="EN-GB" style="font-family:Symbol;mso-ansi-language:EN-GB">·<span style="font:7.0pt &quot;Times New Roman&quot;">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; </span></span><span lang="EN-GB" style="mso-ansi-language:EN-GB">Calculation of the hessian matrix. Wait...<o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt;text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">12345678.12.13.14.15.16.17.18.23.24.25.26.27.28.34.35.36.37.38.45.46.47.48.56.57.58.67.68.78<o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt;text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">&nbsp;<o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt;text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">Inverting the hessian to get the covariance matrix. </span>Wait...</pre>
   
   <pre style="margin-left:18.0pt;text-align:justify">&nbsp;<o:p></o:p></pre>
   
   <pre style="margin-left:18.0pt;text-align:justify">#Hessian matrix#</pre>
   
   <pre style="margin-left:18.0pt"><span lang="DE" style="mso-ansi-language:DE">3.344e+002 2.708e+004 -4.586e+001 -3.806e+003 -1.577e+000 -1.313e+002 3.914e-001 3.166e+001 <o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt"><span lang="DE" style="mso-ansi-language:DE">2.708e+004 2.204e+006 -3.805e+003 -3.174e+005 -1.303e+002 -1.091e+004 2.967e+001 2.399e+003 <o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt"><span lang="DE" style="mso-ansi-language:DE">-4.586e+001 -3.805e+003 4.044e+002 3.197e+004 2.431e-002 1.995e+000 1.783e-001 1.486e+001 <o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt"><span lang="DE" style="mso-ansi-language:DE">-3.806e+003 -3.174e+005 3.197e+004 2.541e+006 2.436e+000 2.051e+002 1.483e+001 1.244e+003 <o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt"><span lang="DE" style="mso-ansi-language:DE">-1.577e+000 -1.303e+002 2.431e-002 2.436e+000 1.093e+002 8.979e+003 -3.402e+001 -2.843e+003 <o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt"><span lang="DE" style="mso-ansi-language:DE">-1.313e+002 -1.091e+004 1.995e+000 2.051e+002 8.979e+003 7.420e+005 -2.842e+003 -2.388e+005 <o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt"><span lang="DE" style="mso-ansi-language:DE">3.914e-001 2.967e+001 1.783e-001 1.483e+001 -3.402e+001 -2.842e+003 1.494e+002 1.251e+004 <o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt"><span lang="DE" style="mso-ansi-language:DE">3.166e+001 2.399e+003 1.486e+001 1.244e+003 -2.843e+003 -2.388e+005 1.251e+004 1.053e+006 <o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt;text-align:justify"><span lang="DE" style="mso-ansi-language:
   DE"># Scales<o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt;text-align:
   justify"><span lang="DE" style="mso-ansi-language:DE">12 1.00000e-004 1.00000e-006<o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt;text-align:justify"><span lang="DE" style="mso-ansi-language:
   DE">13 1.00000e-004 1.00000e-006<o:p></o:p></span></pre>
   
   <pre style="margin-left:
   18.0pt;text-align:justify"><span lang="DE" style="mso-ansi-language:DE">21 1.00000e-003 1.00000e-005<o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt;text-align:justify"><span lang="DE" style="mso-ansi-language:
   DE">23 1.00000e-004 1.00000e-005<o:p></o:p></span></pre>
   
   <pre style="margin-left:
   18.0pt;text-align:justify"><span lang="DE" style="mso-ansi-language:DE"># Covariance<o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt;text-align:justify"><span lang="DE" style="mso-ansi-language:
   DE"><span style="mso-spacerun: yes">&nbsp; </span>1 5.90661e-001<o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt;text-align:justify"><span lang="DE" style="mso-ansi-language:
   DE"><span style="mso-spacerun: yes">&nbsp; </span>2 -7.26732e-003 8.98810e-005<o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt;text-align:justify"><span lang="DE" style="mso-ansi-language:
   DE"><span style="mso-spacerun: yes">&nbsp; </span>3 8.80177e-002 -1.12706e-003 5.15824e-001<o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt;text-align:justify"><span lang="DE" style="mso-ansi-language:
   DE"><span style="mso-spacerun: yes">&nbsp; </span>4 -1.13082e-003 1.45267e-005 -6.50070e-003 8.23270e-005<o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt;text-align:justify"><span lang="DE" style="mso-ansi-language:
   DE"><span style="mso-spacerun: yes">&nbsp; </span>5 9.31265e-003 -1.16106e-004 6.00210e-004 -8.04151e-006 1.75753e+000<o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt;text-align:justify"><span lang="DE" style="mso-ansi-language:
   DE"><span style="mso-spacerun: yes">&nbsp; </span>6 -1.15664e-004 1.44850e-006 -7.79995e-006 1.04770e-007 -2.12929e-002 2.59422e-004<o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt;text-align:justify"><span lang="DE" style="mso-ansi-language:
   DE"><span style="mso-spacerun: yes">&nbsp; </span>7 1.35103e-003 -1.75392e-005 -6.38237e-004 7.85424e-006 4.02601e-001 -4.86776e-003 1.32682e+000<o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt;text-align:justify"><span lang="DE" style="mso-ansi-language:
   DE"><span style="mso-spacerun: yes">&nbsp; </span>8 -1.82421e-005 2.35811e-007 7.75503e-006 -9.58687e-008 -4.86589e-003 5.91641e-005 -1.57767e-002 1.88622e-004<o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt;text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB"># agemin agemax for lifexpectancy, bage fage (if mle==0 ie no data nor Max likelihood).<o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt;text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">&nbsp;<o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt;text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">&nbsp;<o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt;text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">agemin=70 agemax=100 bage=50 fage=100<o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt;text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">Computing prevalence limit: result on file 'plrmypar.txt' <o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt;text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">Computing pij: result on file 'pijrmypar.txt' <o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt;text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">Computing Health Expectancies: result on file 'ermypar.txt' <o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt;text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">Computing Variance-covariance of DFLEs: file 'vrmypar.txt' <o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt;text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">Computing Total LEs with variances: file 'trmypar.txt' <o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt;text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">Computing Variance-covariance of Prevalence limit: file 'vplrmypar.txt' <o:p></o:p></span></pre>
   
   <pre style="margin-left:18.0pt;text-align:justify"><span lang="EN-GB" style="mso-ansi-language:EN-GB">End of Imach<o:p></o:p></span></pre>
   
   <p
   style="text-align:justify;tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">Once
   the running is finished, the program requires a caracter:<o:p></o:p></span></p>
   
   <table border="1" cellpadding="0"
   style="mso-cellspacing:1.5pt;mso-padding-alt:
    0cm 0cm 0cm 0cm">
       <tr>
           <td width="100%"
           style="width:100.0%;padding:.75pt .75pt .75pt .75pt"><strong><span lang="EN-GB" style="mso-ansi-language:EN-GB">Type
           e to edit output files, c to start again, and q for
           exiting:</span><span lang="EN-GB" style="mso-ansi-language:
     EN-GB"><o:p></o:p></span></strong></td>
       </tr>
   </table>
   
   <p
   style="tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">First
   you should enter <strong>e </strong>to edit the master file
   mypar.htm. <o:p></o:p></span></p>
   
   <ul type="disc">
       <li class="MsoNormal"
       style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
        mso-list:l9 level1 lfo49;tab-stops:list 36.0pt"><u><span lang="EN-GB" style="mso-ansi-language:EN-GB">Outputs
           files</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></u> <br>
           <br>
           - Observed prevalence in each state: </span><a
           href="..\mytry\prmypar.txt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">pmypar.txt</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></a> <br>
           - Estimated parameters and the covariance matrix: </span><a
           href="..\mytry\rmypar.txt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">rmypar.txt</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></a> <br>
           - Stationary prevalence in each state: </span><a
           href="..\mytry\plrmypar.txt"><span lang="EN-GB" style="mso-ansi-language:
        EN-GB">plrmypar.txt</span><span lang="EN-GB" style="mso-ansi-language:
        EN-GB"></a> <br>
           - Transition probabilities: </span><a
           href="..\mytry\pijrmypar.txt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">pijrmypar.txt</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></a> <br>
           - Copy of the parameter file: </span><a
           href="..\mytry\ormypar.txt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">ormypar.txt</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></a> <br>
           - Life expectancies by age and initial health status: </span><a
           href="..\mytry\ermypar.txt"><span lang="EN-GB" style="mso-ansi-language:
        EN-GB">ermypar.txt</span><span lang="EN-GB" style="mso-ansi-language:
        EN-GB"></a> <br>
           - Variances of life expectancies by age and initial
           health status: </span><a href="..\mytry\vrmypar.txt"><span lang="EN-GB" style="mso-ansi-language:
        EN-GB">vrmypar.txt</span><span lang="EN-GB" style="mso-ansi-language:
        EN-GB"></a>
           <br>
           - Health expectancies with their variances: </span><a
           href="..\mytry\trmypar.txt"><span lang="EN-GB" style="mso-ansi-language:
        EN-GB">trmypar.txt</span><span lang="EN-GB" style="mso-ansi-language:
        EN-GB"></a> <br>
           - Standard deviation of stationary prevalence: </span><a
           href="..\mytry\vplrmypar.txt"><span lang="EN-GB" style="mso-ansi-language:
        EN-GB">vplrmypar.txt</span><span lang="EN-GB" style="mso-ansi-language:
        EN-GB"></a><br>
           - Prevalences forecasting: </span><a href="frmypar.txt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">frmypar.txt</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></a>
           <br>
           - Population forecasting (if popforecast=1): </span><a
           href="poprmypar.txt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">poprmypar.txt</span><span style="mso-ansi-language:EN-GB"></a> <span lang="EN-GB"><o:p></o:p></span></span></li>
       <li class="MsoNormal"
       style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;
        mso-list:l9 level1 lfo49;tab-stops:list 36.0pt"><u><span lang="EN-GB" style="mso-ansi-language:EN-GB">Graphs</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></u>
           <br>
           <br>
           -</span><a href="..\mytry\pemypar1.gif"><span lang="EN-GB" style="mso-ansi-language:
        EN-GB">One-step transition
           probabilities</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></a><br>
           -</span><a href="..\mytry\pmypar11.gif"><span lang="EN-GB" style="mso-ansi-language:
        EN-GB">Convergence to the
           stationary prevalence</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></a><br>
           -</span><a href="..\mytry\vmypar11.gif"><span lang="EN-GB" style="mso-ansi-language:
        EN-GB">Observed and stationary
           prevalence in state (1) with the confident interval</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></a> <br>
           -</span><a href="..\mytry\vmypar21.gif"><span lang="EN-GB" style="mso-ansi-language:
        EN-GB">Observed and stationary
           prevalence in state (2) with the confident interval</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></a> <br>
           -</span><a href="..\mytry\expmypar11.gif"><span lang="EN-GB" style="mso-ansi-language:EN-GB">Health life
           expectancies by age and initial health state (1)</span><span lang="EN-GB" style="mso-ansi-language:
        EN-GB"></a> <br>
           -</span><a href="..\mytry\expmypar21.gif"><span lang="EN-GB" style="mso-ansi-language:EN-GB">Health life
           expectancies by age and initial health state (2)</span><span lang="EN-GB" style="mso-ansi-language:
        EN-GB"></a> <br>
           -</span><a href="..\mytry\emypar1.gif"><span lang="EN-GB" style="mso-ansi-language:
        EN-GB">Total life expectancy by
           age and health expectancies in states (1) and (2).</span><span style="mso-ansi-language:EN-GB"></a> <span lang="EN-GB"><o:p></o:p></span></span></li>
   </ul>
   
   <p
   style="tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">This
   software have been partly granted by </span><a
   href="http://euroreves.ined.fr"><span lang="EN-GB" style="mso-ansi-language:EN-GB">Euro-REVES</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></a>, a concerted
   action from the European Union. It will be copyrighted
   identically to a GNU software product, i.e. program and software
   can be distributed freely for non commercial use. Sources are not
   widely distributed today. You can get them by asking us with a
   simple justification (name, email, institute) </span><a
   href="mailto:brouard@ined.fr"><span lang="EN-GB" style="mso-ansi-language:EN-GB">mailto:brouard@ined.fr</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></a> and </span><a
   href="mailto:lievre@ined.fr"><span lang="EN-GB" style="mso-ansi-language:EN-GB">mailto:lievre@ined.fr</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"></a> .<o:p></o:p></span></p>
   
   <p
   style="tab-stops:45.8pt 91.6pt 137.4pt 183.2pt 229.0pt 274.8pt 320.6pt 366.4pt 412.2pt 458.0pt 503.8pt 549.6pt 595.4pt 641.2pt 687.0pt 732.8pt"><span lang="EN-GB" style="mso-ansi-language:EN-GB">Latest
   version (0.7 of February 2002) can be accessed at </span><a
   href="http://euroreves.ined.fr/imach"><span lang="EN-GB" style="mso-ansi-language:EN-GB">http://euroreves.ined.fr/imach</span><span lang="EN-GB" style="mso-ansi-language:EN-GB"><o:p></o:p></span></a></p>
   </body>
   </html>

Removed from v.1.1.1.1  
changed lines
  Added in v.1.5


FreeBSD-CVSweb <freebsd-cvsweb@FreeBSD.org>