Estimates of disease incidence in women based on antenatal or neonatal seroprevalence data: HIV in New York City

Estimating age specific prevalence and force of infection in Zimbabwe using combined cross-sectional surveys from 2005 to 2015

Objective

Age structured sexual mixing patterns have been noted to be associated with HIV prevalence and force of infection. Therefore, this study aimed to estimate the age dependent HIV force of infection using survey cross-sectional data from Zimbabwe.

Methods

We fit generalized additive models namely; linear, semi-parametric, non-parametric and non-proportional hazards models. Using the 2005-06, 2010-11 and 2015 Zimbabwe Demographic Health Surveys data. The Akaike Information Criteria was used to select the best model. The best model was then used to estimate the age dependent HIV prevalence and force-of-infection.

Results

Based on birth year cohort-specific prevalence, the female HIV prevalence reaches the highest peak at around 29 years of age, then declines thereafter. Males have a lower cohort specific prevalence between 15 and 30 years than females. Male cohort-specific prevalence slightly decreases between the ages of 33 and 39, then peaks around the age of 40. The cohort-specific FOI is greater in females than in males throughout all age categories. In addition, the cohort-specific HIV FOI peaked at ages 22 and 40 for females and males, respectively. The observed 18-year age difference between the HIV FOI peaks of males and females.

Conclusion

Our model was appealing because we did not assume that the FOI is stationary over time; however, we used serological survey data to distinguish the FOI's age-and-time effect. The cohort-specific FOI peaked 18 years earlier in females than males, indicative of age-mixing patterns. We recommend interventions that target younger females so as to reduce HIV transmission rates.

A general HIV incidence inference scheme based on likelihood of individual level data and a population renewal equation

We derive a new method to estimate the age specific incidence of an infection with a differential mortality, using individual level infection status data from successive surveys. The method consists of a) an SI-type model to express the incidence rate in terms of the prevalence and its derivatives as well as the difference in mortality rate, and b) a maximum likelihood approach to estimate the prevalence and its derivatives. Estimates can in principle be obtained for any chosen age and time, and no particular assumptions are made about the epidemiological or demographic context. This is in contrast with earlier methods for estimating incidence from prevalence data, which work with aggregated data, and the aggregated effect of demographic and epidemiological rates over the time interval between prevalence surveys. Numerical simulation of HIV epidemics, under the presumption of known excess mortality due to infection, shows improved control of bias and variance, compared to previous methods. Our analysis motivates for a) effort to be applied to obtain accurate estimates of excess mortality rates as a function of age and time among HIV infected individuals and b) use of individual level rather than aggregated data in order to estimate HIV incidence rates at times between two prevalence surveys.

Estimating the distribution of the window period for recent HIV infections: a comparison of statistical methods

In the past few years a number of antibody biomarkers have been developed to distinguish between recent and established Human Immunodeficiency Virus (HIV) infection. Typically, a specific threshold/cut-off of the biomarker is chosen, values below which are indicative of recent infections. Such biomarkers have attracted considerable interest as the basis for incidence estimation using a cross-sectional sample. An estimate of HIV incidence can be obtained from the prevalence of recent infection, as measured in the sample, and knowledge of the time spent in the recent infection state, known as the window period. However, such calculations are based on a number of assumptions concerning the distribution of the window period. We compare two statistical methods for estimating the mean and distribution of a window period using data on repeated measurements of an antibody biomarker from a cohort of HIV seroconverters. The methods account for the interval-censored nature of both the date of seroconversion and the date of crossing a specific threshold. We illustrate the methods using repeated measurements of the Avidity Index (AI) and make recommendations about the choice of threshold for this biomarker so that the resulting window period satisfies the assumptions for incidence estimation.

Bayesian evidence synthesis for a transmission dynamic model for HIV among men who have sex with men

Understanding infectious disease dynamics and the effect on prevalence and incidence is crucial for public health policies. Disease incidence and prevalence are typically not observed directly and increasingly are estimated through the synthesis of indirect information from multiple data sources. We demonstrate how an evidence synthesis approach to the estimation of human immunodeficiency virus (HIV) prevalence in England and Wales can be extended to infer the underlying HIV incidence. Diverse time series of data can be used to obtain yearly “snapshots” (with associated uncertainty) of the proportion of the population in 4 compartments: not at risk, susceptible, HIV positive but undiagnosed, and diagnosed HIV positive. A multistate model for the infection and diagnosis processes is then formulated by expressing the changes in these proportions by a system of differential equations. By parameterizing incidence in terms of prevalence and contact rates, HIV transmission is further modeled. Use of additional data or prior information on demographics, risk behavior change and contact parameters allows simultaneous estimation of the transition rates, compartment prevalences, contact rates, and transmission probabilities.

Estimating the HIV incidence rate: recent and future developments

PURPOSE OF REVIEW

To describe the needs for information on the rate of new HIV infections (incidence) in epidemics and review developments in various methods for its estimation.

RECENT FINDINGS

Epidemiological methods for estimating incidence with models using prevalence data have been useful, but the expansion of antiretroviral treatment programmes could now challenge their reliability. Laboratory-based HIV incidence assays that can be used to measure HIV incidence using a cross-sectional survey, provide a promising concept, but current technologies have not been sufficiently accurate. New statistical methods have been developed that show that if the properties of the assay are properly measured then unbiased estimates of incidence can be derived, and that assays meeting certain criteria can produce estimates of acceptable accuracy and precision. Encouragingly, some new assays and algorithms show signs of potentially meeting those criteria. Among the next challenges will be the systematic evaluation of assay performance in many different types of specimen and the validation of those methods through comparison with other measurements of incidence.

SUMMARY

Recent developments in epidemiological and incidence assay-based methods of measuring incidence have been substantial and are likely to eventually lead to a revolution in the way that worldwide HIV epidemics are routinely tracked.

Estimating incidence from prevalence in generalised HIV epidemics: methods and validation

BACKGROUND

HIV surveillance of generalised epidemics in Africa primarily relies on prevalence at antenatal clinics, but estimates of incidence in the general population would be more useful. Repeated cross-sectional measures of HIV prevalence are now becoming available for general populations in many countries, and we aim to develop and validate methods that use these data to estimate HIV incidence.

METHODS AND FINDINGS

Two methods were developed that decompose observed changes in prevalence between two serosurveys into the contributions of new infections and mortality. Method 1 uses cohort mortality rates, and method 2 uses information on survival after infection. The performance of these two methods was assessed using simulated data from a mathematical model and actual data from three community-based cohort studies in Africa. Comparison with simulated data indicated that these methods can accurately estimates incidence rates and changes in incidence in a variety of epidemic conditions. Method 1 is simple to implement but relies on locally appropriate mortality data, whilst method 2 can make use of the same survival distribution in a wide range of scenarios. The estimates from both methods are within the 95% confidence intervals of almost all actual measurements of HIV incidence in adults and young people, and the patterns of incidence over age are correctly captured.

CONCLUSIONS

It is possible to estimate incidence from cross-sectional prevalence data with sufficient accuracy to monitor the HIV epidemic. Although these methods will theoretically work in any context, we have able to test them only in southern and eastern Africa, where HIV epidemics are mature and generalised. The choice of method will depend on the local availability of HIV mortality data.

Do Tests Devised to Detect Recent HIV-1 Infection Provide Reliable Estimates of Incidence in Africa?

The objective of this study was to assess the performance of 4 biologic tests designed to detect recent HIV-1 infections in estimating incidence in West Africa (BED, Vironostika, Avidity, and IDE-V3). These tests were assessed on a panel of 135 samples from 79 HIV-1-positive regular blood donors from Abidjan, Côte d'Ivoire, whose date of seroconversion was known (Agence Nationale de Recherches sur le SIDA et les Hépatites Virales 1220 cohort). The 135 samples included 26 from recently infected patients (< or =180 days), 94 from AIDS-free subjects with long-standing infection (>180 days), and 15 from patients with clinical AIDS. The performance of each assay in estimating HIV incidence was assessed through simulations. The modified commercial assays gave the best results for sensitivity (100% for both), and the IDE-V3 technique gave the best result for specificity (96.3%). In a context like Abidjan, with a 10% HIV-1 prevalence associated with a 1% annual incidence, the estimated test-specific annual incidence rates would be 1.2% (IDE-V3), 5.5% (Vironostika), 6.2% (BED), and 11.2% (Avidity). Most of the specimens falsely classified as incident cases were from patients infected for >180 days but <1 year. The authors conclude that none of the 4 methods could currently be used to estimate HIV-1 incidence routinely in Côte d'Ivoire but that further adaptations might enhance their accuracy.

Modelling the force of infection for hepatitis B and hepatitis C in injecting drug users in England and Wales

Background

Injecting drug use is a key risk factor, for several infections of public health importance, especially hepatitis B (HBV) and hepatitis C (HCV). In England and Wales, where less than 1% of the population are likely to be injecting drug users (IDUs), approximately 38% of laboratory reports of HBV, and 95% of HCV reports are attributed to injecting drug use.

Methods

Voluntary unlinked anonymous surveys have been performed on IDUs in contact with specialist agencies throughout England and Wales. Since 1990 more than 20,000 saliva samples from current IDUs have been tested for markers of infection for HBV, HCV testing has been included since 1998. The analysis here considers those IDUs tested for HBV and HCV (n = 5,682) from 1998–2003. This study derives maximum likelihood estimates of the force of infection (the rate at which susceptible IDUs acquire infection) for HBV and HCV in the IDU population and their trends over time and injecting career length. The presence of individual heterogeneity of risk behaviour and background HBV prevalence due to routes of transmission other than injecting are also considered.

Results

For both HBV and HCV, IDUs are at greatest risk from infection in their first year of injecting (Forces of infection in new initiates 1999–2003: HBV = 0.1076 95% C.I: 0.0840–0.1327 HCV = 0.1608 95% C.I: 0.1314–0.1942) compared to experienced IDUs (Force of infection in experienced IDUs 1999–2003: HBV = 0.0353 95% C.I: 0.0198–0.0596, HCV = 0.0526 95% C.I: 0.0310–0.0863) although independently of this there is evidence of heterogeneity of risk behaviour with a small number of IDUs at increased risk of infection. No trends in the FOI over time were detected. There was only limited evidence of background HBV infection due to factors other than injecting.

Conclusion

The models highlight the need to increase interventions that target new initiates to injecting to reduce the transmission of blood-borne viruses. Although from the evidence here, identification of those individuals that engage in heightened at-risk behaviour may also help in planning effective interventions. The data and methods described here may provide a baseline for monitoring the success of public health interventions.

Estimating incidence of HIV infection in childbearing age African women using serial prevalence data from antenatal clinics

Ades and Medley provided the first flexible method for estimating age- and time-specific HIV incidence using HIV prevalence data collected among pregnant women and adjusting for the effect of differential selection between infected and uninfected women. This paper extends the approach proposed by these authors. We used a parametric model that allows the relative inclusion rate to depend on both age, calendar time, and duration of HIV infection. We developed a two dimensional penalized log-likelihood approach for estimating time- and age-specific incidence using a binomial likelihood function and a quadratic roughness penalty which allows smoothing over both age and time. Identifiability of the model parameters and effect of sample size are studied through simulations. The method is illustrated using prenatal HIV testing data recorded from 1995 to 2002 in Abidjan, Côte d'Ivoire, to estimate the HIV annual incidence rate among women aged 12-40 year old, from the beginning of the epidemic to 2002. We show that estimated incidence rates are highly dependent on hypotheses made to model the relative inclusion rate. Despite this dependency, the application of the method leads to new and accurate findings on HIV incidence qualitative features in Abidjan. We highlight the relevance of such a method in monitoring the dynamics of HIV epidemic in Africa which is essential for planning vaccine trials and future treatment needs, and for assessment of prevention policy.

Evaluation of a modified commercial assay in detecting antibody to hepatitis C virus in oral fluids and dried blood spots

Oral fluid testing is an effective alternative to serum antibody testing for surveillance of human immunodeficiency virus (HIV) and hepatitis B infections, and is being extended to hepatitis C infections. The objective of this study was to determine and compare the sensitivity and specificity of a modified commercial assay for the detection of antibody to hepatitis C virus (anti-HCV) in oral fluids collected by two different oral fluid collection devices (the Epitope OraSure trade mark and Sarstedt Salivette ) and in dried fingerprick blood spots. In this study, 253 anti-HCV seropositive patients and 394 blood donors (all anti-HCV negative) were recruited between August 2000 and January 2001. Each participant provided oral fluid specimens by OraSure and Salivette, and at least one dried blood spot. Serum specimens were collected from the patients whenever possible. For those injecting drug users who did not provide a serum specimen, HCV status was established on the basis of previous testing. All the nonserum samples were tested for the presence of anti-HCV, using a modified Ortho HCV 3.0 SAVe enzyme-linked immunosorbent assay (ELISA) protocol. The recommended preliminary cutoffs for the modified ELISA were suboptimal. Further, the sensitivity, specificity, and positive and negative predictive values could be improved by varying the cutoff and taking into account the likely prevalence of HCV in the population under investigation. For instance, given a population with a 50% prevalence of anti-HCV, the optimal sensitivities of the modified assay on OraSure, Salivette, and dried blood spots were 92%, 83%, and virtually 100%, respectively, in contrast to 83%, 59%, and 99% using the preliminary cutoffs. The respective optimal specificities were 99%, 93%, and 100%. In conclusion, oral fluids collected by the OraSure device provide an extremely useful method to conduct public health surveillance of not only HIV, but also hepatitis C, among injecting drug users. In addition, dried blood spot specimens may be useful for surveillance and could be employed as a first line diagnostic specimen.

HIV incidence estimates among women of childbearing age in the area around Paris, France: no evidence for any effect of age or time

OBJECTIVE

To explore whether there are any age or time effects on the HIV incidence rate among women of childbearing age in the area around Paris (France).

DESIGN

Three seroprevalence surveys were conducted among pregnant women in the Paris area (PREVAGEST) during three periods (1990-1991, 1992-1993, 1994-1995); their data were used to derive HIV incidence estimates with Ades and Medley's method. To assess the power of our study, data were also simulated with a demographic model under different assumptions for HIV-incidence rates.

RESULTS

No age or time effect was detected on HIV incidence in the Paris area during the period from 1990 to 1995. Analysis of simulated surveys showed that with the sample size of the PREVAGEST surveys, the method was able to detect an age or time effect with a relative risk of about 2.5 between age groups or periods.

CONCLUSION

We conclude that the method is applicable to our data. If any existed, age or time effect was quite likely no higher than 2.5. As we previously reported using another method, incidence of HIV infection among women of childbearing age in the Paris area can be estimated at 0.74/1,000 per year (95% confidence interval, 0.62-0.87) from 1990 through 1995.

Advances in medical statistics arising from the AIDS epidemic

Many statisticians have contributed to studies of the HIV epidemic and progression to AIDS. They have developed new statistical methodology, where needed, to address HIV-related issues. The transfer of methods from one area to another often involves a substantial delay. This paper points to methods that were developed in the HIV context and have either already found applications in other areas of medical research or have the potential for such applications, with the hope that this will promote a speedier transfer of the research methods. Among the new tools that HIV studies have placed firmly into the pool of statistical methods for medical research are the methods of back-calculation, methods for the analysis of retrospective ascertainment data and methods of analysis for the combined data from clinical trials and associated longitudinal studies. Notions that have been stimulated substantially are use of surrogate endpoints in clinical trials and screening blood products by the use of pooled serum samples. Research activity in many other areas has been boosted substantially through contributions motivated by HIV/AIDS studies. Noteworthy examples are analyses for doubly-censored lifetime data and methods for assessing vaccines for transmissible diseases.

A non-parametric method for the reconstruction of age- and time-dependent incidence from the prevalence data of irreversible diseases with differential mortality

A method is proposed for reconstructing the time and age dependence of incidence rates from successive age-prevalence cross sections taken from the sentinel surveys of irreversible diseases when there is an important difference in mortality between the infected and susceptible subpopulations. The prevalence information at different time-age points is used to generate a surface; the time-age variations along the life line profiles of this surface and the difference in mortality rates are used to reconstruct the time and age dependence of the incidence rate. Past attempts were based on specified parametric forms for the incidence or on the hypothesis of time-invariant forms for the age-prevalence cross sections. The proposed method makes no such assumptions and is thus capable of coping with rapidly evolving prevalence situations. In the simulations carried out, it is found to be resilient to important random noise components added to a prescribed incidence rate input. The method is also tested on a real data set of successive HIV age-prevalence cross sections from Burundi coupled to differential mortality data on HIV(+) and HIV(-) individuals. The often-made assumption that the incidence rate can be written as the product of a calendar time component and an age component is also examined. In this case, a pooling procedure is proposed to estimate the time and the age profiles of the incidence rate using the reconstructed incidence rates at all time-age points.

Estimating HIV incidence from age-specific prevalence data: comparison with concurrent cohort estimates in a study of male factory workers, Harare, Zimbabwe

OBJECTIVE

To compare HIV incidence estimates from cross-sectional age-specific prevalence data with concurrent cohort estimates and to examine the sensitivity of the estimates to changes in age-categorization and survivorship assumptions.

METHODS

Two previously described methods of estimating HIV incidence from cross-sectional prevalence data - the cumulative incidence and survival (CIS) and constant prevalence (CP) methods - are applied using data from a study of male factory workers in Harare, Zimbabwe. The methods are applied under two alternative groupings of the HIV prevalence data and under alternative survivorship assumptions: (a) Weibull distribution providing the best fit to the HIV prevalence data using the CIS method; (b) Weibull distribution matching data from an HIV natural history cohort study in Uganda; and (c) survivorship pattern as in (b) with survival periods reducing with increasing age at infection. Age-specific, age-standardized and cumulative HIV incidence estimates are calculated. The results are compared with concurrent longitudinal estimates from 3 years of follow-up of the Harare cohort (1993-1995).

RESULTS

Age-standardized HIV incidence was estimated at 2.02 per 100 man years (95% CI, 1.57-2.47) in the cohort study. There was evidence of recent variability in HIV incidence in these data. Estimates from the cross-sectional methods ranged from 1.98 to 2.74 per 100 man years and were sensitive to changes in age-categorization of the HIV prevalence data and changes in survivorship assumptions. The cross-sectional estimates were higher at central ages and lower at older ages than the cohort estimates. The age-specific estimates from the CIS method were less sensitive to changes in age grouping than those from the CP method.

CONCLUSIONS

HIV incidence remains high in Harare. Incidence estimates broadly consistent with cohort estimates can be obtained from single-round cross-sectional HIV prevalence data in established epidemics - even when the underlying assumption of stable endemic prevalence is not fully met. Estimates based on cross-sectional surveys should therefore be explored when reliable longitudinal estimates cannot be obtained. More data on post-HIV infection survivorship distributions in sub-Saharan Africa would facilitate the improvement of estimates of incidence based on cross-sectional surveys.