Uncertainties in predicting the demographic impact of AIDS.

A model giving the demographic impact of AIDS is analysed to examine the sensitivity of the projections when various complicating features are included. The model deals with age and sexual partner change rate as continuous variables and uses a device to specify arbitrary correlations between the ages of the people who form sexual partnerships. The device ensures consistency, in that the amount of partner formation is the same regardless of whether the partnerships are counted from the point of view of males or females. Arbitrary correlation between partner change rate and fertility is also permitted. The results show the uncertainty in model predictions that population growth will reduce over the next 20 years to approximately zero in parts of East Africa severely affected by the AIDS epidemic. The main sources of uncertainty in the model predictions are assumptions concerning the correlation between ages in a partnership, the correlation between partner change rate and fertility, the incubation period of AIDS, and the variability of the female partner change rate.

Assessing the National HIV/AIDS Strategy evaluation.

The report is a useful compilation of data on Australia's HIV/AIDS epidemic and the response to it. The report produces firm evidence for endorsing Australia's success in limiting the spread of HIV by a number of routes. Australia has been successful in preventing iatrogenic spread, spread through commercial sex and spread through injecting drug use. In the latter case however, the hepatitis C epidemic cautions against complacency. The report, unfortunately, does not recognise very serious weaknesses in the evidence it uses as a basis for its endorsement of current efforts against HIV in the male homosexual community and the general heterosexual community. The report recognises failure of current policies in Aboriginal and Torres Strait Islander communities and recognises the threat posed here by HIV but recommends no major policy change. There is some recognition of the limitation of the data, but no recommendation for increased surveillance. The report ignores methods of prevention other than public health education. In my view, there is sufficient evidence to recommend routine neonatal circumcision, at least in Aboriginal and Torres Strait Islander communities. There is a clear need for a more effective approach to STD control, particularly in these communities. Adequate STD and HIV control may well require contact tracing and surveillance using traditional public health methods. Measurement of objective indicators of the success of HIV prevention campaigns needs to be improved with more comprehensive collection of data on HIV and STD incidence and condom sales, stratified by relevant covariates. Survey evidence of behavioural change should be collected from those too young to be affected by the selective mortality factor. The international comparisons await a proper statistical study which may be able to identify the elements of an effective approach to AIDS. In future reviews of the effectiveness of Australia's response to AIDS, all methods of limiting the spread of the epidemic should be considered objectively. It is unacceptable for the AIDS program to be declared optimal simply because it accords with current public health ideology.

The shape of the distribution of the number of sexual partners.

Information on the number of sexual partners that people have is necessary for predicting the likely long term course of the AIDS epidemic. More information is required than is provided by simply stating the mean and variance of the number of partners. It is useful to find a family of curves which approximate data on the proportion of people in the various 'number of partners' categories. Two Australian surveys of the sexual behaviour of first year university behavioural science students were analysed. It was found here that log-normal distributions gave good approximations to the distribution of number of partners amongst people who had more than one partner. Gamma and negative binomial distributions gave less satisfactory fits and the truncated normal and Poisson distributions were clearly unable to match the data.

The impact of sexual mixing patterns on the spread of AIDS.

It has previously been demonstrated that the distribution of sexual partner change rates may not be sufficient behavioral information to predict the spread of AIDS. A more adequate description of the sexual network through which AIDS may spread will require information on who mixes with whom. Previous models indicated that a highly "assortive" pattern of sexual mixing, in which people of high partner change rates nearly always partner similar people, would result in a relatively rapid epidemic but one that would ultimately be rather small as it would be largely confined to those with high partner change rates. However, in earlier models it was necessary to arbitrarily vary either the partnering patterns or the distribution of partner change rates as the epidemic progressed in order to ensure consistency in the partnering patterns. Consequently, the "pure" effects of assortiveness on the size of the epidemic could not be assessed. In this paper a new model is described in which consistency in partnering patterns is maintained as the epidemic selectively depletes those with higher partner change rates. However, this model maintains consistency without arbitrary adjustments to either the assortive nature of partnering or important features of the distribution of the partner change rate. It is then possible to test the effects of assortiveness on the outcome of the epidemic without important confounding effects from the mathematical device used to maintain consistency in partnering. A practical sensitivity test on the impact of assortiveness on model predictions shows that, contrary to expectations, increasing assortiveness can, under certain circumstances, lead to larger epidemics.

Modelling AIDS reduction strategies.

BACKGROUND: Mathematical models of the AIDS epidemic have not been able to give accurate predictions about the size of the epidemic because it is not possible to obtain sufficiently accurate measurements of the factors that enable HIV transmission. The uncertainties inherent in models of the AIDS epidemic appear to limit their relevance to epidemiologists. However, it is shown here that the uncertainties need not prevent models being used to make reliable decisions about which preventive strategy will be most effective. METHOD: A range of strategies are simulated in a model of the AIDS epidemic. The simulations are repeated as the value of what seems to be the most important uncertain factor, is varied. The effect of this variation on the effectiveness of each strategy is noted. In principle, the process could be repeated whilst all other uncertain factors are varied as well. RESULTS: Although varying one uncertain factor created enormous variation in the size of the epidemic, it is remarkable that for most preventive strategies the relative effectiveness of the strategies was barely altered. Hence for the most part the ranking of strategies in order of effectiveness is not affected by the area of uncertainty explored here. The results also highlight the potential effectiveness not only of general condom promotion, but also the use of circumcision and spermicides and general screening or targetted screening in sexually transmitted disease clinics. CONCLUSIONS: Epidemiological modelling may accurately rank the effectiveness of interventions although it may fail to predict the size of the epidemic.

Assessment of sexual mixing patterns.

The manner in which people with varying partner change rates choose their partners may have important implications for the future of the AIDS epidemic. A relatively rapid but small epidemic can be expected if sexual mixing patterns are highly assortative so that those with high partner change rates nearly always choose partners with similar rates. Direct estimation of sexual mixing is exceedingly difficult. On the other hand, information on the way people mix is implicitly contained in the epidemiology of other sexually transmitted diseases (STDs). If mixing is highly assortative, one might expect gonorrhea to be largely confined to a small subsection of the population who would frequently contract the disease repeatedly. In this paper a model of gonorrhea transmission in which partner change rates and preference are dealt with as continuous variables is used to extract information on mixing from data on an Australian heterosexual population. It was found that in this population mixing is not far from random with respect to partner change rates.

Moving muscles with microchips, mathematics and models: an update on the 'reverse' model.

This paper reviews the current status of our computer model for the management of strabismus and its future direction. Vector analysis was first used in the 1950s for the assessment of strabismus. Robinson's model was the first computer simulation of ocular motility. Using physiological principles and anatomical approximations, Robinson's model sought to predict the strabismic pattern to be expected from a given injury. The Kault/Stark 'reverse' model works in the opposite direction, to first simulate the given strabismic pattern and then advise the surgery required to restore orthophoria. The surgeon is able to 'trial' various operations and compare the expected postoperative results. An automated system is currently being developed to ease the difficulty in measuring the position of the eyes in all nine positions of gaze. This paper includes three illustrative case reports.

Modeling AIDS as a function of other sexually transmitted disease.

The spread of AIDS, as with any sexually transmitted disease, will depend on the pattern of sexual activity. Both the proportion of the population who have high partner exchange rates and the extent to which that proportion interacts with the remainder of the population are likely to be important determinants of the AIDS epidemic. However, it does not seem likely that surveys could obtain sufficiently reliable information of this nature for use in an accurate model of the AIDS epidemic. On the other hand, such information is implicitly contained in the epidemiology of other sexually transmitted diseases (STDs). Therefore a method is suggested of calculating the parameters of a model of the AIDS epidemic by comparing it with the epidemiology of another STD. The result is a model that predicts the likelihood of infection by the AIDS virus as a function of time and an individual's history of STD. It is suggested that further work along these lines may lead to a quantitative approach to assessing the importance of various STDs as cofactors in the spread of AIDS.

An automated system of strabismus management. A direct approach using a "reverse" model.

Until now computer models of strabismus have consisted of lengthy calculations based on oculomotor physiology. Such models have been able to demonstrate the pattern of strabismus that would arise from any given muscle abnormality and/or operation. However, it has been difficult to use such models to guide strabismus surgery because the surgeon requires a model that works in the opposite direction. The surgeon requires a model whose input is the pattern of strabismus and output is the muscle operation required. Such a model is described here. Instead of consisting of calculations based on physiology the new model consists of a store of most of the information capable of being generated by an existing strabismus model together with an efficient search procedure. In this paper the potential of the new model is demonstrated by its application to a case of paralytic strabismus.

Further investigation of a strabismus model.

Computer models of strabismus have been available for about ten years. Such models are computer programs which perform a set of calculations based on an interpretation of the oculomotor system as a system of two spheres each with six adjustable springs. The effects of pathology and surgery can be simulated by altering the parameters of the springs. However, such models have not yet met with wide practical acceptance. One reason for this lack of interest is that previous investigators have neglected to apply strabismus models to concomitant strabismus. In this paper the application of a strabismus model to concomitant strabismus is demonstrated and data are generated which should be clinically useful. Another reason for lack of acceptance of computer models results from the expectation that they may be unreliable because of biological variability. The model developed here is tested for its sensitivity to a wide range of biological variations and proves to be very robust. Therefore, information on concomitant squint surgery given by the model should be applicable to most patients.