An Open Book

Look at the new presentation of September 1995

How to read this "Open Book"

This book is a classical printed book, but it has a World-wide Web access (http://sauvy.ined.fr/popafsi/english). Thus, accessing it from the Web, gives you unpublished information, particularly graphs and tables, but you really need the book to understand all from these graphs unless your are a demographer and well aware of epidemic and population theory.

So, let us suppose that you already got a copy of the french book "African Populations and AIDS" ("Populations africaines et sida" in french) and have succesfully read chapter three entitled "Demographic Aspects and Consequences of the AIDS Epidemic in Africa".

Let me imagine that you aggree with the scientific arguments expressed in the chapter and now are willing to change our published scenario with your own.

On page 159 of the book one can find two figures: figure 26 for example at the bottom, gives on one hand (a) the survival function by age in Africa before the HIV epidemic, ie with a life expectancy at birth of 50 years and on the other hand (b) the same function with the assumption that the HIV epidemic affects 15% of the adult population.

Once you are connected, the original published figure 26 does appear on your screen. How to read it?

The epidemic affects the sexually active population which dies for AIDS after a median time of 10 years, and lowers the survival function around age 25-35 for the Women and around age 30-45 for the Men (this because of age differences between spouses in Africa).
Also, a baby born in a population where 15% of the adult population is infected, has about 15 chances out of 100 to have an infected mother. If his mother is infected, she can transmit the virus (mostly during the delivery, but also during pregnancy and breast feeding). Here we assumed that the "vertical transmission" was 30% (which is probably too high).
So the survival function is afflicted twice.

How to change the scenario?

On the top of your screen, you can easily change the HIV prevalence from the default value (15%) to whatever value. If you have heard of a much higher prevalence level (proportion of infected) in some urban areas you can enter 30 instead of 15. But you can also, try 0.4% percent with a life expectancy of 75 years, like in Western Countries for Women.

Then by clicking on "Submit" you will get a new screen after some seconds required for computation and transmission. An original page 159 bis, with both updated figures (legend and captions are also changed) will appear.
By cloning your browser window you can visually compare the different scenarios on your screen. If you prefer to delay your reactions, you can print your new pages and they will fit the format of the published book so you can insert them in your own private copy of the book!

Other figures and tables are planned. I hope that you will understand the book more easily. I thank François Gèze, director of "Les Éditions La Découverte" for his help in promoting this new service.

Nicolas Brouard
18 octobre 1994

Brouard@sauvy.ined.fr

PS: For the published version of the figures I have used my own programs usually written in C or Ratfor depending on how old the subroutines were. To let the figures interactive and accessible to the World-Wide Web I wrote an interface written in Perl with the aid of many netizens and in particular Steven Brenner S.E.Brenner@biochemistry.cambridge.ac.uk and Oscar Nierstrasz, oscar@iam.unibe.ch. I also use GNU-Plot from the Free Software Foundation.

Revised version (September 1995)

Some figures and tables couldn't be included in the book because of the limitation from the publisher. Also, the original (C) program was using other parameters, like the mother to child transmission ratio, which hopefully will decrease in the future and which, at least, is of main interest to change interactively. In addition, some technical explanations, not included in the book, are now included in a glossary. Here are the new possibilities offered through this web access of the book.

Stable populations

Age pyramids of an African country facing a stable epidemic were missing (results on the growth rate were included but no graph). Thus in the case of an African country with an HIV prevalence of 15% among the sexually active population (15+), the growth rate falls from 3% to 2% per year. This only litle change comes from the high fertility of women (6-7 children). If the age survival function in a cohort is highly affected by the AIDS epidemic, with two depressions in infancy and for most sexually active ages, the age pyramid is only slightly affected. The reason is that if the growth rate is still very high (2% per year), each single age new cohort is 2% bigger than the previous one, giving to the age pyramid (succession of cohorts) on fig 26-I (unpublished), a profile which is much more influenced by the exponential (exp(-0.02x)) curve than by the mortality curve (fig. 26).

It is also of interest to look at figure 26-II which gives the distribution of deaths by age at the stable state. Here the influence of the epidemic is terrific. Deaths from AIDS are highlighted to distinguish them from other deaths.

All these figures are interactive and values of parameters can be changed from the main menu.

In such wise, you can verify that you need a prevalence over 30% to have a low growth rate and then a marked pyramid.

Reproduction tables concerning the demographic situation before the epidemic and the demographic and epidemiological situations during the epidemic are also accessible on demand. What we mean by reproduction table is the collection of the life table by single age, of the fertility and cumulative fertility by age, of the distribution of prevalence, stable population, by age, etc. Many synthetic indices, like the life expectancy, are included in the header of each table.
Remark: The tables are written in HTML 3 (hypertext markup language version 3). If your browser does not display the table correctly, dowload a new one, like Arena, or better Nestcape. The tables can be printed in landscape mode.

Much more changeable parameters

The original programm was built on many more parameters than life expectancy and HIV prevalence. To make this programm more general, we decided to make most of parameters modifiable.
In that case the model can be used, for example, to simulate the HIV-2 epidemic, or even to have an idea of the demographic consequences of the heterosexual spread in Western or Asian Countries.
In such a way, the parameters concerning the demographic situation before the epidemic are not only the life expectancy but also the age profile from the so-called Coale and Demeny model life tables. For fertility, we also choosed two age-sex profiles called African and European patterns. The difference comes from the important age gap between spouses in most African countries and the existence of polygamy which makes the fertility of old men still very high.
Epidemiological parameters are more numerous. The mean incubation time of adults can be changed, either to simulate HIV-2 or for pedagogical reasons (with very short incubation time). The mean survival time of infected children which was estimated to 2 years in the book but can now be increased. Also the vertical transmission ratio, which is still lowering as more research are involved, can be changed (we use a ratio of 30%) to simulate the western conditions where AZT may be administered during pregnancy.
Tables and figures (published and unpublished) are updated according to the new entered values.
A glossary
Majority of technical terms used in this presentation are briefly explained in the glossary which is callable from numerous links in the main menu or figures.
Some details mathematics, not included in the book, are also described in the glossary.

A Bibliography
The bibliography of chapter 3 is now accessible.
A tutorial will be included if we have time.
Hope it will help you to understand this controversial problem of the demographic consequences of the AIDS epidemic in Africa.
Please, send me any comments, suggestions, reactions, lack of understanding. Up to know, we got many connections but very few comments.

Nicolas Brouard
September 20, 1995
Brouard@sauvy.ined.fr