The incubation period to AIDS in injecting drug users estimated from prevalent cohort data, accounting for death prior to an AIDS diagnosis

HIV Disease Progression Among Antiretroviral Therapy Patients in Zimbabwe: A Multistate Markov Model

Background: Antiretroviral therapy (ART) impact has prolonged survival of people living with HIV. We evaluated HIV disease progression among ART patients using routinely collected patient-level data between 2004 and 2017 in Zimbabwe. Methods: We partitioned HIV disease progression into four transient CD4 cell counts states: state 1 (CD4 ≥ 500 cells/μl), state 2 (350 cells/μl ≤ CD4 < 500 cells/μl), state 3 (200 cells/μl ≤ CD4 < 350 cells/μl), state 4 (CD4 < 200 cells/μl), and the absorbing state death (state 5). We proposed a semiparametric time-homogenous multistate Markov model to estimate bidirectional transition rates. Covariate effects (age, gender, ART initiation period, and health facility level) on the transition rates were assessed. Results: We analyzed 204,289 clinic visits by 63,422 patients. There were 24,325 (38.4%) patients in state 4 (CD4 < 200) at ART initiation, and 7,995 (12.6%) deaths occurred by December 2017. The overall mortality rate was 3.9 per 100 person-years. The highest mortality rate of 5.7 per 100 person-years (4,541 deaths) was from state 4 (CD4 < 200) compared to other states. Mortality rates decreased with increase in time since ART initiation. Health facility type was the strongest predictor for immune recovery. Provincial or central hospital patients showed a diminishing dose-response effect on immune recovery by state from a hazard ratio (HR) of 8.30 [95% confidence interval (95% CI), 6.64-10.36] (state 4 to 3) to HR of 3.12 (95% CI, 2.54-4.36) (state 2 to 1) compared to primary healthcare facilities. Immune system for male patients was more likely to deteriorate, and they had a 32% increased mortality risk (HR, 1.32; 95% CI, 1.23-1.42) compared to female patients. Elderly patients (45+ years) were more likely to immune deteriorate compared to 25-34 years age group: HR, 1.35; 95% CI, 1.18-1.54; HR, 1.56; 95% CI, 1.34-1.81 and HR, 1.53; 95% CI, 1.32-1.79 for states 1 to 2, state 2 to 3, and states 3 to 4, respectively. Conclusion: Immune recovery was pronounced among provincial or central hospitals. Male patients with lower CD4 cell counts were at a higher risk of immune deterioration and mortality, while elderly patients were more likely to immune deteriorate. Early therapeutic interventions when the immune system is relatively stable across gender and age may contain mortality and increase survival outcomes. Interventions which strengthen ART services in primary healthcare facilities are essential.

HIV treatment as prevention among people who inject drugs - a re-evaluation of the evidence

Background

Population-level associations between community measures of HIV viral load and HIV incidence have been interpreted as evidence for HIV anti-retroviral treatment (ART) as prevention among people who inject drugs (PWID). However, investigation of concurrent HCV and HIV incidence trends allows examination of alternative explanations for the fall in HIV incidence. We estimate the contribution of ART and reductions in injecting risk for reducing HIV incidence in Vancouver between 1996 and 2007.

Methods

A deterministic model of HIV and HCV transmission among PWID was calibrated to the baseline (1996) HIV and HCV epidemic among PWID in Vancouver. While incorporating parameter uncertainty, the model projected what levels of ART protection and decreases in injecting risk could reproduce the observed reduction in HIV and HCV incidence for 1996-2007, and so what impact would have been achieved with just ART or just reductions in injecting risk.

Results

Model predictions suggest the estimated reduction (84%) in HCV incidence for 1996-2007 required a 59% (2.5-97.5 percentile range 49-76%) reduction in injecting risk, which accounted for nine-tenths of the observed decrease in HIV incidence; the remainder was achieved with a moderate ART efficacy for reducing sexual HIV infectivity (70%, 51-89%) and an uncertain ART efficacy for reducing injection-related HIV infectivity (44%, 0-96%). Despite this uncertainty, projections suggest that the decrease in injecting risk reduced HIV incidence by 76% (63-85%) and ART further reduced HIV incidence by 8% (2-19%), or on its own by 3% (-34-37%).

Conclusions

Observed declines in HIV incidence in Vancouver between 1996 and 2007 should be seen as a success for intensive harm reduction, whereas ART probably played a small role.

Impact of opioid substitution therapy on the HIV prevention benefit of antiretroviral therapy for people who inject drugs

OBJECTIVE

A recent meta-analysis suggested that opioid substitution therapy (OST) increased uptake of antiretroviral treatment (ART) and HIV viral suppression. We modelled whether OST could improve the HIV prevention benefit achieved by ART among people who inject drugs (PWID).

METHODS

We modelled how introducing OST could improve the coverage of ART across a PWID population for different baseline ART coverage levels. Using existing data on how yearly HIV-transmission risk is related to HIV plasma viral load, changes in the level of viral suppression across the population were used to project the relative reduction in yearly HIV-transmission risk achieved by ART, with or without OST, compared with if there was no ART - defined here as the prevention effectiveness of ART.

RESULTS

Owing to OST use increasing the chance of being on ART and achieving viral suppression if on ART, the prevention effectiveness of ART for PWID on OST (compared with PWID not on OST) increases by 44, 31, or 20% for a low (20%), moderate (40%), or high (60%) baseline ART coverage, respectively. Improvements in the population-level prevention effectiveness of ART are also achieved across all PWID, compared with if OST was not introduced. For instance, if OST is introduced at 40% coverage, the population-level prevention effectiveness of ART could increase by 27, 20, or 13% for a low (20%), moderate (40%), or high (60%) baseline ART coverage, respectively.

CONCLUSION

OST could improve the HIV prevention benefit of ART; supporting strategies that aim to concurrently scale-up OST with ART.

CD4(+) T-cell depletion in HIV infection: mechanisms of immunological failure

The hallmark of acquired immunodeficiency syndrome (AIDS) pathogenesis is a progressive depletion of CD4(+) T-cell populations in close association with progressive impairment of cellular immunity and increasing susceptibility to opportunistic infections (OI). Disease progression in untreated human immunodeficiency virus (HIV) infection can take many years, and it was originally hypothesized to be a consequence of slow, viral-mediated CD4(+) T-cell destruction. However, massive CD4(+) memory T-cell destruction is now known to occur quite early in infection, almost always without overt immunodeficiency. In most individuals, this initial destruction is countered by CD4(+) memory T-cell regeneration that preserves CD4(+) T-cell numbers and functions above the threshold associated with overt immunodeficiency. This regeneration, which occurs in the setting of chronic immune activation and immune dysregulation does not, however, restore all functionally important CD4(+) T-cell populations and is not stable over the long term. Ultimately, CD4(+) memory T-cell homeostasis fails and critical effector populations decline below the level necessary to prevent OI. Thus, the onset of overt immune deficiency appears to be intimately linked with CD4(+) memory T-cell dynamics and reflects the complex interplay of direct viral cytopathogenicity and the indirect effects of persistent immune activation on CD4(+) memory T-cell proliferation, differentiation, and survival.

Injecting drug use is associated with a more rapid CD4 cell decline among treatment naïve HIV-positive patients in Indonesia

Background

It remains unclear whether the natural course of human immunodeficiency virus (HIV) differs in subjects infected through injecting drug use (IDU) and no data have been published from low- or middle-income countries. We addressed this question in an urban cohort in Indonesia, which is experiencing a rapidly growing HIV epidemic strongly driven by IDU.

Methods

All antiretroviral treatment (ART) naïve HIV-positive patients who had at least two subsequent CD4 cell counts available before starting ART were included in this study. We examined the association between IDU and CD4 cell decline using a linear mixed model, with adjustment for possible confounders such as HIV viral load and hepatitis C antibodies.

Results

Among 284 HIV-positive ART naïve patients, the majority were male (56%) with a history of IDU (79% among men). People with a history of IDU had a statistically significant faster decline in CD4 cells (p<0.001). Based on our data, patients with a history of IDU would have an average 33% decline in CD4 cells after one year without ART, compared with a 22% decline among non-users. At two years, the decline would average 66 and 40%, respectively. No other factor was significantly associated with CD4 cell decline.

Conclusions

We show that a history of IDU is associated with a more rapid CD4 cell natural decline among HIV-positive individuals in Indonesia. These findings have implications for monitoring ART naïve patients with a history of IDU and for starting ART in this group.

Evaluation of HIV incidence surveillance in New York City, 2006

In 2005, the New York City (NYC) Department of Health and Mental Hygiene implemented a standardized human immunodeficiency virus (HIV) incidence surveillance protocol based on the serologic testing algorithm for recent HIV seroconversion deployed nationwide by the Centers for Disease Control and Prevention (CDC). We evaluated four key attributes of NYC's HIV incidence surveillance system-simplicity, data quality, timeliness, and acceptability--using CDC's guidelines for surveillance system evaluation. The evaluation revealed that the system could potentially provide HIV incidence estimates stratified by borough and major demographic groups at about nine months after the period of interest. The system strengths include its relative simplicity and integration with routine HIV/acquired immunodeficiency syndrome surveillance. Weaknesses include lack of completeness of testing history information, a critical component of incidence estimation. Continued improvements in data completeness and timeliness will improve the currently available information to inform personnel who develop HIV-prevention programs and policy initiatives in NYC and nationally.

Impact of accelerated progression to AIDS on public health monitoring of late HIV diagnosis

Some patients develop AIDS within a year of HIV infection ("accelerated progression"). Classifying such cases as late HIV diagnosis may lead to inaccurate evaluation of HIV testing efforts. We sought to determine this group's contribution to overall late diagnosis rates. To identify cases of accelerated progression (development of AIDS within 12 months of a negative HIV test), we reviewed published HIV seroconverter cohort studies and used New York City's (NYC) HIV/AIDS surveillance registry. From the literature review, three seroconverter cohort studies revealed that 1.0-3.6% of participants had accelerated progression to AIDS. Applying this frequency estimate to the number of new infections in NYC (4762) for 2006 calculated by the Centers for Diseases Control and Prevention's incidence formula, we estimated that 3.6-13.0% of 1317 NYC HIV cases who are diagnosed with AIDS within 12 months of HIV diagnosis are accelerated progressors, not persons HIV infected for many years who did not test and present with AIDS (i.e., delayed diagnosis). In addition, our analysis of the 2006 NYC surveillance registry confirmed the occurrence of accelerated progression in a population-based setting; 67 accelerated progressors were reported and 9 (13%) could be confirmed through follow-up medical record review. With increased HIV testing initiatives, the irreducible proportion of AIDS cases with accelerated progression must be considered when interpreting late diagnosis data.

New challenges for mathematical and statistical modeling of HIV and hepatitis C virus in injecting drug users

Injecting drug users are not only driving blood-borne transmission of HIV and hepatitis C virus but also likely drive sexual transmission of HIV in large parts of the world. Mathematical and statistical modeling can provide important insights in these epidemiological processes and on the potential impact of interventions but have been little used to date. This review aims to discuss the potential areas of application of modeling in the field of viral infections in injecting drug users. After reviewing key examples of published modeling work on HIV and hepatitis C virus in injecting drug users, we discuss recent developments in the epidemiology, diagnosis, treatment and prevention of these infections. In particular, new methods for the diagnosis of early HIV infection, new antivirals for a more effective treatment of HIV, hepatitis B and hepatitis C virus infections, new concepts in design and surveillance of interventions for drug users and increasing possibilities of molecular typing of pathogens are changing the questions and decisions for public health policy makers who deal with drug-related infectious diseases. Research including mathematical modeling is needed to understand the impact of new diagnostic tools, new treatment options and combined intervention strategies on the epidemiology of viral infections in injecting drug users. Methodological advances in mathematical modeling are required to adequately approach some of the ensuing research questions. Modeling has much to offer for solving urgent policy questions, but current levels of funding in modeling research are insufficient and need to be scaled up to make better use of these possibilities.

Community coverage and HIV prevention: assessing metrics for estimating HIV incidence through syringe exchange

BACKGROUND

Evaluations of syringe-exchange programme effectiveness that attempt to measure "coverage" by determining the percentage of the at-risk population reached by a programme are insufficient since programmes must provide syringes on a continual basis. Determining the relationship between the extent of programme coverage and its impact (i.e., reductions in disease risk or incidence) is complicated by the lack of controlled trials with cohorts of drug users and instead has to be estimated by imputation, mathematical modelling, or ecological data analysis. This report offers an approach to determine community-wide impact and discusses the limitations of that approach.

METHODS

Easily created programme tracking data were maintained by exchanges in New Haven, CT, USA and Chicago, IL, USA. Data compiled by month quantified the number of unique participants visiting syringe-exchange programmes and the number of syringes distributed. "Coverage", defined either as the percentage of individuals reached or percentage of community syringe need met, was estimated by incorporating measures of the size of the injector population or injection frequency. These measures of coverage are placed in the context of changing programme operations to estimate the effect of these changes on coverage. Finally, data on AIDS cases from New Haven and Chicago were used to estimate the community-wide impact of syringe exchange.

RESULTS

Two mobile syringe-exchange programmes operated with very different exchange policies. Programme data revealed that coverage of individuals rarely exceeded 10% and was higher in New Haven than in Chicago. On the other hand, coverage measured as the percentage of syringe need met was higher at the Chicago exchange that employed the less restrictive policy. The impact of syringe exchange in the two cities was measured by comparing subsequent AIDS cases. The relative reduction in injection-related AIDS cases as a function of all new AIDS diagnoses was 21.7% in New Haven and 41.4% in Chicago.

CONCLUSIONS

A modest investment in the collection of programme data can yield reliable and interpretable information on the extent of programme reach and retention. Limitations to the approach result from the ecological nature of the data and from the need to use data from outside the programme that may be less reliable. For the cases presented here, coverage rates will vary as a function of the programme policies; however, even modest coverage rates - well below those recommended by UNAIDS - can have significant impacts on HIV epidemics. Restrictive policies appeared to increase the coverage if measured only by the proportion of monthly participants and not by the proportion of syringe need met by a programme. More generally, programmes can collect programmatic data and some rapid assessment data (estimates of IDU population and injection frequency) to estimate of the impact of their programmes.

Estimation and prediction of the HIV–AIDS-epidemic under conditions of HAART using mixtures of incubation time distributions

The estimation of the HIV-AIDS epidemic by means of back-calculation (BC) has been difficult since the introduction of highly active anti-retroviral therapy (HAART) because the incubation time distributions needed for BC were poorly known. Moreover, it has been assumed that if the general public is aware that effective treatments are available then the majority of infected people would be known, and therefore a hidden epidemic was assumed not to exist. Nevertheless, it was suspected that not every infected person would come to the attention of health-care providers, and therefore estimates independent of the patients' registration were necessary. In this paper, the incubation time distributions for HIV treated with the HAART regimen are derived from a cohort study. By using estimates of the proportion treated according to the HAART regimen and the incubation time distributions estimated in the era before the implementation of HAART (pre-HAART), new marginal population incubation time distributions for each of the three risk groups (homosexuals, drug users and others) were constructed. The BC was performed using an empirical Bayesian approach based on the latter incubation time distribution.

HIV progression and predictors of mortality in a community-based cohort of Zambian adults

This article describes immunological HIV progression, mortality, and its predictors in 974 Zambian adults. During 3138 person-years of follow-up, 281 deaths occurred, and the overall mortality rate was 9.0 per 100 person-years. Thirty-six percent of patients were dead within 5 years of enrollment. The median survival in patients with baseline CD4 count ≥500 cells/mm³ was 5.62 years, with CD4 count between 200 and 499 cells/mm³ 5.46 years, and with CD4 count <200 cells/mm³ 3.89 years. The mortality rate increased significantly with older age (6.9 in patients <25 years, 9.3 in individuals aged 25-39 years, 10.2 in patients ≥40 years) and was higher in women (rate ratio 1.29). The median annual change of progression markers was -29.6 cells/mm³ for CD4 count, -3.0% for CD4 count percentage, 1.2 nmol/L for neopterin, -1.9 g/L for hemoglobin, and -70 cells/mm³ for total lymphocyte count. Hemoglobin and neopterin were as accurate as CD4 count to predict mortality.

Antiretroviral HIV treatment and care for injecting drug users: an evidence-based overview

AIDS-related mortality and the rate of progression to AIDS have dramatically decreased since the advent of highly active antiretroviral treatment (HAART). The overall benefit from antiretroviral HIV treatment has, however, been lesser in HIV-infected injecting drug users (IDUs) than in other patient groups (e.g. men who have sex with men). Poorer outcomes in HIV-infected IDUs are related to a variety of factors, including increased rates of non-HIV-related deaths, hepatitis C, delayed access to effective treatment, lower adherence to care and treatment regimens, continuation of illicit drug use, depression and negative life events. The available evidence strongly suggests the need for the large-scale implementation of comprehensive treatment and care strategies for IDUs that include both treatment of drug dependence and HAART.

Higher set point plasma viral load and more-severe acute HIV type 1 (HIV-1) illness predict mortality among high-risk HIV-1-infected African women

BACKGROUND

There is limited information on the natural history of human immunodeficiency virus type 1 (HIV-1) infection in Africa, especially from individuals with well-defined dates of infection. We used data from a prospective cohort study of female sex workers in Mombasa, Kenya, who were followed up monthly from before the date of HIV-1 infection.

METHODS

Antiretroviral-naive women who had a well-defined date of HIV-1 infection were included in this analysis. The effects of set point plasma viral load (measured 4-24 months after infection), early CD4+ cell count, and symptoms of acute HIV-1 infection on mortality were assessed using Cox proportional hazards analysis.

RESULTS

Among 218 women, the median duration of follow-up after HIV-1 infection was 4.6 years. Forty women died, and at 8.7 years (the time of the last death), the cumulative survival rate was 51% by Kaplan-Meier analysis. Higher set point viral load, lower early CD4+ cell count, and more-symptomatic acute HIV-1 illness each predicted death. In multivariate analysis, set point viral load (hazard ratio [HR], 2.28 per 1 log10 copies/mL increase; P=.001) and acute HIV-1 illness (HR, 1.14 per each additional symptom; P=.05) were independently associated with higher mortality.

CONCLUSION

Among this group of African women, the survival rate was similar to that for HIV-1-infected individuals in industrialized nations before the introduction of combination antiretroviral therapy. Higher set point viral load and more-severe acute HIV-1 illness predicted faster progression to death. Early identification of individuals at risk for rapid disease progression may allow closer clinical monitoring, including timely initiation of antiretroviral treatment.

On the intra-host dynamics of HIV-1 infections

An extension of a previously proposed theory for the pathogenesis of AIDS is presented and analyzed using a mathematical modelling approach. This theory is based on the observation that human immunodeficiency virus type 1 (HIV-1) predominantly infects and replicates in (CD4+)-T cells, and that the infection process within an infected individual is characterized by ongoing generation and selection of HIV variants with increasing reproductive capacity. This evolutionary process is considered to be the reason for the gradual loss of immunocompetence and the final destruction of the immune system observed in most patients. The extension presented here incorporates the effect of the permanently increasing susceptibility of (CD4+)-T cell clones, as a result of the evolutionary process. The presented model reproduces and possibly explains a wide variety of findings about the HIV infection process. Numerical results indicate that the effect of the initial dose is minimal, and restricted to the primary phase of infection. According to the model predictions the impact of the HIV evolutionary speed is crucial for the progression to disease. An important progression determinant is the initial infection rate, being a component of the viral reproductive capacity. An influential role in disease progression seems to be played by the initial (CD4+)-T cell count.

The role of substance abuse in HIV disease progression: reconciling differences from laboratory and epidemiologic investigations

From the onset of the HIV/AIDS epidemic, the use of licit and illicit drugs has been investigated for its potential impact on HIV disease progression. Findings from a large number of laboratory-based studies indicate that drug abuse may exacerbate HIV disease progression; however, epidemiological studies have shown mixed results. This article presents a review of findings from both laboratory-based and epidemiologic investigations. In addition, we provide a careful evaluation of methodological strengths and limitations inherent to both study designs in order to provide a more nuanced understanding of how these findings may complement one another.

Anti-human immunodeficiency virus-1 antibody titers in injection drug users compared to sexually infected individuals

Sera from infected injection drug users (IDU) have shown to have antibodies against synthetic human immunodeficiency virus-1 (HIV-1) envelope peptides more frequently. In this study, reactivity of 48 IDU plasma were compared to 60 plasmas obtained from sexually infected individuals (S). The overall reactivity of plasma from IDU compared to S was higher, and the reactivity titers were much higher for IDU plasma than S. IDU plasma also showed a broader antibody response. The higher reactivity titers were observed mainly for the gp41 immunodominant epitope and V3 peptides corresponding to the consensus sequences of HIV-1 subtypes/variants prevalent in Brazil (B, F, C) indicating the specificity in the higher immune response of IDU.

HIV-1 infection in injection drug users

Injection drug use is an efficient and ongoing means of HIV transmission and is the principal mode of transmission in some parts of the world. In the United States, approximately 10,000 injection drug users are believed to acquire HIV each year. The US Public Health Service hopes to decrease all HIV transmission in the United States by 50% in the next 5 years, by promoting care and prevention services to infected persons. Subtle differences in the virology and immunopathogenesis of HIV between injection drug users and other groups at risk are still being investigated. So far such differences have no practical implication. Comparison of progression rates and survival with HIV across risk groups has been difficult because of the many competing causes of death unrelated to HIV among injection drug users, but overall HIV disease progression rates are similar across risk groups, after adjusting for age. Some AIDS-related opportunistic infections are more common (such as tuberculosis) or less common (such as Kaposi's sarcoma) among injection drug users, based on rates of exposure and latent infection. Other comorbidities, including chronic psychiatric disorders and hepatitis C disease, are more common among injection drug users than among others with HIV infection. Highly active antiretroviral treatment seems to be as effective in persons with a history of injected drug use as in others. Increasing the numbers of HIV-infected injection drug users who know their diagnosis, increasing their access to care and prevention services, and increasing their adherence to a therapeutic regimen are the current challenges in confronting the HIV-epidemic among injection drug users. To overcome these obstacles, clinicians must have both technical knowledge and skill in assisting patients with behavior change.

Improved methods and assumptions for estimation of the HIV/AIDS epidemic and its impact: Recommendations of the UNAIDS Reference Group on Estimates, Modelling and Projections

UNAIDS and WHO produce biannual country-specific estimates of HIV/AIDS and its impact. These estimates are based on methods and assumptions that reflect the best understanding of HIV epidemiology and demography at the time. Where significant advances are made in epidemiological and demographic research, the methods and assumptions must evolve to match these advances. UNAIDS established an Epidemiology Reference Group in 1999 to advise them and other organisations on HIV epidemiology and methods for making estimates and projections of HIV/AIDS. During the meeting of the reference group in 2001, four priority areas were identified where methods and assumptions should be reviewed and perhaps modified: a) models of the HIV epidemic, b) survival of adults with HIV-1 in low and middle income countries, c) survival of children with HIV-1 in low and middle income countries, and d) methods to estimate numbers of AIDS orphans. Research and literature reviews were carried out by Reference Group members and invited specialists, prior to meetings held during 2001-2. Recommendations reflecting the consensus of the meeting participants on the four priority areas were determined at each meeting. These recommendations were followed in UNAIDS and WHO development of country-specific estimates of HIV/AIDS and its impact for end of 2001.

Comparison of smoothing techniques for CD4 data in a Markov model with states defined by CD4: an example on the estimation of the HIV incubation time distribution

Multi-state models defined in terms of CD4 counts are useful for modelling HIV disease progression. A Markov model with six progressive CD4-based states and an absorbing state (AIDS) was used to estimate the cumulative probability of progressing to AIDS in 158 HIV-1 infected haemophiliacs with known seroconversion (SC) dates. A problem arising in such analysis is how to define CD4-based states, since this marker is subject to measurement error and short timescale variability. Four approaches were used: no smoothing, ad hoc smoothing (to move to a later/previous state two consecutive measurements to later/previous states are needed), kernel smoothing and random effects (RE) models. The estimates were compared with the Kaplan-Meier estimate based solely on data concerning time to AIDS. There was an apparent lack of agreement between the Kaplan-Meier and the "no smoothing" estimate. With the exception of the "no smoothing" method, "ad hoc", kernel and RE estimates fell within the range of the 95 per cent CIs of the Kaplan-Meier curve. Simulations demonstrated that the use of raw CD4 counts provides overestimated transition intensities. Compared to the kernel method, ad hoc is easier to implement and overcomes the problem of the choice of bandwidth. The RE approach leads to simple models, since it usually results in very few transitions to previous states, and can handle individuals with sparse data by smoothing their predictions towards the population mean. Ad hoc was the method that performed better, in terms of bias, than the other smoothing approaches.

Methods for estimating the AIDS incubation time distribution when date of seroconversion is censored

In most cohort studies on HIV infection and AIDS, data on time from seroconversion to AIDS or death are doubly censored, both at the time origin and at the endpoint of interest. In epidemiological research, the most frequently adopted approach is to restrict the analysis to persons with narrow seroconversion intervals and to impute the midpoint of this interval as date of seroconversion. For many cohort studies, the consequence is that a substantial proportion of the data is not used. We consider four methods that are expected to be less biased when all cohort data are used: two imputation methods, conditional mean and multiple imputation, and two likelihood maximization methods. We derive the likelihood structure of the cohort data and clarify its dependence on study design. All methods are applied to data from the Amsterdam cohort study among injection drug users. In a simulation study the data generation process of this cohort study is imitated. The performance of midpoint, conditional mean and multiple imputation are compared. With midpoint imputation, both an analysis using the full data set, as well as one restricted to the cases with small seroconversion intervals, is performed. Conditional mean imputation comes out as the preferred method. It gives best results with respect to mean squared error. Moreover, when confidence intervals are computed through standard methods that ignore the uncertainty in the imputed date of seroconversion, coverage probabilities are almost correct.

Survival estimates for patients with homozygous sickle-cell disease in Jamaica: a clinic-based population study

BACKGROUND

Information about life expectancy of patients with homozygous sickle-cell disease is needed for research and patient counselling. Our aim was to study two Jamaican populations, one clinic-based and one birth cohort and, by careful consideration of data quality and statistical analysis, to identify ways to increase the chances of obtaining valid and generalisable results.

METHODS

We investigated the survival experience of 3301 patients with homozygous sickle-cell disease attending the Jamaican sickle-cell clinic between Jan 1, 1987, and Dec 31, 1996. We applied and assessed a simulation technique for incorporating early life mortality using a birth cohort, and analysed the precision of this technique. Kaplan-Meier survival estimates are produced.

FINDINGS

290 of the 3301 patients died. Median survival calculated with the excess mortality rate simulation data was 53 years (95% CI 49.3-57.0) for men and 58.5 (55.1-67.5) for women.

INTERPRETATION

Our simulation technique, with realistic assumptions based on empirical evidence, offers a new estimate of median survival for patients with homozygous sickle-cell disease. We present the precision of these survival estimates, which introduces an important level of uncertainty. The inherent biases of clinically ascertained populations of patients, and the assumptions underlying analysis techniques are crucial features of survival studies in sickle-cell disease, and can modify summary statistics substantially.

Three stage AIDS incubation period: a worst case scenario using addict-needle interaction assumptions

In this paper we develop and analyse a model for the spread of HIV/AIDS amongst a population of injecting drug users. We start off with a brief literature survey and review; this is followed by the derivation of a model which allows addicts to progress through three distinct stages of variable infectivity prior to the onset of full blown AIDS and where the class of infectious needles is split into three according to the different levels of infectivity in addicts. Given the structure of this model we are required to make assumptions regarding the interaction of addicts and needles of different infectivity levels. We deliberately choose these assumptions so that our model serves as an upper bound for the prevalence of HIV under the assumption of a three stage AIDS incubation period. We then perform an equilibrium and stability analysis on this model. We find that there is a critical threshold parameter R(0) which determines the behaviour of the model. If R(0)< or =1, then irrespective of the initial conditions of the system HIV will die out in all addicts and all needles. If R(0)>1, then there is a unique endemic equilibrium which is locally stable if, as is realistic, the time scale on which addicts inject is much shorter than that of the other epidemiological and demographic processes. Simulations indicate that if R(0)>1, then provided that disease is initially present in at least one addict or needle it will tend to the endemic equilibrium. In addition we derive conditions which guarantee this. We also find that under calibration the long term prevalence of disease in our variable infectivity model is always greater than in an equivalent constant infectivity model. These results are confirmed and explored further by simulation. We conclude with a short discussion.

Back-calculation by birth cohort, incorporating age- specific disease progression, pre-AIDS mortality and change in European AIDS case definition. European Union Concerted Action on Multinational AIDS Scenarios

OBJECTIVES

To adapt and improve methodology for back-calculation of AIDS in Europe and to examine the feasibility of estimating past HIV incidence by birth cohort.

METHODS

Empirical Bayesian back-calculation (EBBC) used Markov disease progression models, modified to allow for three diseases added to the AIDS case definition in 1993 and for pre-AIDS mortality, and estimation by penalized maximum likelihood with a neighbour prior. EBBC by 5-year birth cohort assumed a minimum age at infection and age-dependent progression rates; three versions, with varying age effects, were investigated using AIDS cases diagnosed prior to the introduction of highly active antiretroviral therapies (HAART).

RESULTS

Compared with the no age-effect version, EBBC by birth cohort tended to produce flattened HIV incidence curves in country-exposure groups with < 1000 AIDS cases, reflecting effects of the neighbour prior when data become sparse. Otherwise, birth cohort analysis, with moderate effects of age on progression, gave initially increasing incidence curves and consistent patterns across countries, with the 1960-1964 cohort most affected. In the European Union, incidence is estimated to have peaked in 1983 among homosexual men and in 1988 among injecting drug users; 460000 persons were estimated to be living with HIV/AIDS at the end of 1995.

CONCLUSIONS

Our improved methodology deals effectively with the change in AIDS case definition and has allowed quantitative assessments of the HIV epidemic by birth cohort using all AIDS cases diagnosed before 1996, thus providing a sound basis for public health policy at a time when estimation of more recent prevalence is compromised by the effects of HAART.

Pre-AIDS mortality and its association with HIV disease progression in haemophilic men, injecting drug users and homosexual men

OBJECTIVE

To study pre-AIDS mortality and its association with HIV disease progression in different exposure groups with known intervals of HIV seroconversion.

DESIGN AND METHODS

The type and rate of pre-AIDS deaths were assessed in 111 HIV-infected haemophilic men followed in London, and 118 injecting drug users and 158 homosexual men followed in Amsterdam. In each group, the association between CD4+ T-cell count, HIV RNA and pre-AIDS mortality was studied using proportional hazards analysis.

RESULTS

By 10 years after seroconversion 7.3% of the haemophilic men had died without AIDS and 38.2% had developed AIDS. These figures were 20.2 and 30.5% for injecting drug users, and 8.0 and 55.0% for homosexual men. The major causes of pre-AIDS mortality appear to differ in the three exposure groups. The risk of pre-AIDS death tended to increase with decreasing CD4 cell count and increasing HIV RNA levels in injecting drug users and homosexual men. In men with haemophilia the associations were less obvious, although the log-transformed CD4 cell count was predictive for pre-AIDS death.

CONCLUSIONS

Pre-AIDS deaths occur and are at least partially related to HIV disease progression irrespective of how individuals became infected. Because of the longer life expectancy due to highly active antiretroviral therapy (HAART), pre-AIDS deaths are likely to show a further increase. Methods to incorporate these intermediate outcomes should be considered in the estimation of the size of the HIV epidemic and in the survival analysis of HIV-infected individuals. Prevention and treatment of non-AIDS infections, especially hepatitis C virus infection, and cancers will become increasingly important in HIV-infected individuals. The interaction between these therapies and HAART should be closely monitored.

Methods for estimating HIV prevalence: A comparison of extrapolation from surveys on infection rate and risk behaviour with back-calculation for the Netherlands

OBJECTIVES

To compare HIV prevalence estimates (total number infected) by using extrapolation from surveys on infection rate and risk behaviour (EIR) in specific segments of the population and back-calculation (BC) on reported AIDS cases. To discuss potential sources of bias and error, and to identify areas for improvement of the methodology.

DESIGN

Systematic comparison and epidemiological assessment of data input, underlying assumptions, and output.

METHODS

Low, possibly unbiased and high estimates of HIV prevalence as of January 1996 for homo/bisexual men, injecting drug users. heterosexual men and women with multiple partners, and blood transfusion recipients and haemophiliacs were derived from surveys and continuous data collections on HIV infection rate and risk behaviour in the Netherlands between 1992 and 1996. These were compared with estimates (point and 95 % CI) by empirical Bayesian BC on AIDS cases 1982-1995.

RESULTS AND CONCLUSIONS

The estimate of HIV prevalence by EIR was 13,806 with low and high estimates of 9619 and 17,700, respectively. The HIV prevalence estimate by BC was 8812 (95% CI: 7759-9867). The available data from EIR are too limited for accurate estimates of HIV prevalence. EIR estimates could be improved considerably with more precise data on prevalence of risk behaviours and HIV prevalence rate for homosexual men. More confidence can be put in the BC estimates, but these could be underestimates because of the age effect on incubation time, pre-AIDS treatment and relapse of risk behaviour. BC estimates could be improved by a better representation of the incubation time distribution (including the effect of age there-upon), better data on the effectiveness and uptake of pre-AIDS antiretroviral treatment and prophylaxis of opportunistic infections, and on the level of underreporting.