Predicting and preventing the emergence of antiviral drug resistance in HSV-2.

Genital herpes, caused by herpes simplex virus, is the most prevalent sexually transmitted disease worldwide. In many developing countries genital herpes is untreated, and in the United States only 10% of cases are treated. We present a mathematical model that we use as a health policy tool to predict the levels of antiviral drug resistance that would emerge, if treatment rates were increased, and to identify the key factors in determining the emergence of drug resistance. We use our results to suggest control measures for herpes epidemics that would prevent the emergence of substantial levels of antiviral drug resistance.

Global elimination of trachoma: how frequently should we administer mass chemotherapy?

The World Health Organization has recommended repeat mass drug administration as part of their global initiative to eliminate blinding trachoma by the year 2020. The efficacy of repeat treatment will be tested empirically, but the results will not be available for many years, and recommendations for the necessary frequency of treatment are needed immediately. We have developed a mathematical model that uses available epidemiological data from a variety of countries. We recommend, based on our analysis, that in areas where trachoma is moderately prevalent (<35% in children), it should be treated annually, but hyperendemic areas (>50% in children), it should be treated biannually.

The intrinsic transmission dynamics of tuberculosis epidemics.

In developed countries the major tuberculosis epidemics declined long before the disease became curable in the 1940s. We present a theoretical framework for assessing the intrinsic transmission dynamics of tuberculosis. We demonstrate that it takes one to several hundred years for a tuberculosis epidemic to rise, fall and reach a stable endemic level. Our results suggest that some of the decline of tuberculosis is simply due to the natural behaviour of an epidemic. Although other factors must also have contributed to the decline, these causal factors were constrained to operate within the slow response time dictated by the intrinsic dynamics.

Gut microbiome alteration in MORDOR I: a community-randomized trial of mass azithromycin distribution.

The MORDOR I trial1, conducted in Niger, Malawi and Tanzania, demonstrated that mass azithromycin distribution to preschool children reduced childhood mortality1. However, the large but simple trial design precluded determination of the mechanisms involved. Here we examined the gut microbiome of preschool children from 30 Nigerien communities randomized to either biannual azithromycin or placebo. Gut microbiome γ-diversity was not significantly altered (P = 0.08), but the relative abundances of two Campylobacter species, along with another 33 gut bacteria, were significantly reduced in children treated with azithromycin at the 24-month follow-up. Metagenomic analysis revealed functional differences in gut bacteria between treatment groups. Resistome analysis showed an increase in macrolide resistance gene expression in gut microbiota in communities treated with azithromycin (P = 0.004). These results suggest that prolonged mass azithromycin distribution to reduce childhood mortality reduces certain gut bacteria, including known pathogens, while selecting for antibiotic resistance.

Compartmental Model Diagrams as Causal Representations in Relation to DAGs.

Compartmental model diagrams have been used for nearly a century to depict causal relationships in infectious disease epidemiology. Causal directed acyclic graphs (DAGs) have been used more broadly in epidemiology since the 1990s to guide analyses of a variety of public health problems. Using an example from chronic disease epidemiology, the effect of type 2 diabetes on dementia incidence, we illustrate how compartmental model diagrams can represent the same concepts as causal DAGs, including causation, mediation, confounding, and collider bias. We show how to use compartmental model diagrams to explicitly depict interaction and feedback cycles. While DAGs imply a set of conditional independencies, they do not define conditional distributions parametrically. Compartmental model diagrams parametrically (or semiparametrically) describe state changes based on known biological processes or mechanisms. Compartmental model diagrams are part of a long-term tradition of causal thinking in epidemiology and can parametrically express the same concepts as DAGs, as well as explicitly depict feedback cycles and interactions. As causal inference efforts in epidemiology increasingly draw on simulations and quantitative sensitivity analyses, compartmental model diagrams may be of use to a wider audience. Recognizing simple links between these two common approaches to representing causal processes may facilitate communication between researchers from different traditions.

Modeling the impact of modified directly observed antiretroviral therapy on HIV suppression and resistance, disease progression, and death.

A simulation model that used Markov assumptions with Monte Carlo uncertainty analysis was evaluated 1500 times at 10,000 iterations. Modified directly observed therapy (MDOT) for human immunodeficiency virus was assumed to improve adherence to therapy to 90% of prescribed doses. The impact of MDOT interventions on modeled biological and clinical outcomes was compared for populations with mean rates of adherence (i.e., the mean percentage of prescribed doses taken by each member of the population who had not discontinued therapy) of 40%, 50%, 60%, and 70%. MDOT reduced the risk of virological failure, development of opportunistic infections, and death, yet increased the risk of drug resistance, for each adherence distribution among persons with detectable plasma virus loads. Over 1500 trials, for a population with 50% adherence to therapy and a 12-month period, MDOT increased the median rate of virological suppression from 13.2% to 37.0% of patients, decreased the rate of opportunistic infection from 5.7% to 4.3% of patients, and decreased the death rate from 2.9% to 2.2% of patients. In the same population, however, MDOT increased the rate of new drug resistance mutations from 1.00 to 1.41 per person during the 12-month period. The impact of MDOT was smaller in populations with higher levels of adherence. MDOT interventions will likely improve clinical outcomes in populations with low levels of adherence but may not be effective at preventing drug resistance in treatment-experienced populations. MDOT may be more effective in preventing drug resistance with potent regimens in treatment-naive patients.

Pooling of Chlamydia laboratory tests to determine the prevalence of ocular Chlamydia trachomatis infection.

With the Global Elimination of Trachoma by 2020 program underway, it has become increasingly important to identify the prevalence of ocular chlamydia infection in communities. DNA amplification tests are the gold standard, but are prohibitively expensive. In the present paper, we investigate whether pooling multiple specimens into a single test is feasible. The conjunctivae of 170 children in western Nepal were examined and swabbed. The prevalence of chlamydial infection was estimated in two ways using the ligase chain reaction: by testing all 170 specimens individually, and by testing 34 pools of 5 specimens each. We show that the confidence interval for 34 pooled specimens approaches that of doing all 170 specimens as the prevalence decreases. We also determine the optimal number of specimens to pool into a single test to minimize the confidence interval of the estimate. If the population prevalence is expected to be around 10%, then 14 specimens should be pooled per test. Even at 50% prevalence, costs can be reduced by pooling two samples per test.

Resistance to antimicrobial chemotherapy: a prescription for research and action.

The growing problem of resistance to antimicrobial chemotherapy was discussed by participants at the February 1995 workshop at Emory University on population biology, evolution, and control of infectious diseases. They discussed the nature and source of this problem and identified areas of research in which information is lacking for the development of programs to control of the emergence and spread of resistant bacteria. Particular attention was given to theoretical (mathematical modeling) and empirical studies of the within and between-host population biology (epidemiology) and the evolution of microbial resistance to chemotherapeutic agents. Suggestions were made about the kinds of models and data needed, and the procedures that could be employed to stem the ascent and dissemination of resistant bacteria. This article summarizes the observations and recommendations made at the 1995 meeting and in the correspondence between participants that followed. It concludes with an update on the theoretical and empirical research on the between- and within-host population biology and evolution of resistance to antimicrobial chemotherapy most of which has been done since that meeting.

Predictors of outcome in fungal keratitis.

PURPOSE: To analyse predictors of clinical outcome in fungal keratitis. METHODS: Data was collected during a prospective, randomized, controlled, double-masked clinical trial of treatment for fungal keratitis. Clinical features at presentation and demographics were collected at the enrollment visit for all patients. Pre-specified clinical outcomes included 3-month visual acuity and infiltrate/scar size, time to re-epithelialization, and corneal perforation. A separate multivariable model with each outcome as the dependent variable included all predictor variables. RESULTS: Predictors for worse 3-month visual acuity include older age (P=0.024), worse presentation visual acuity (P<0.001), larger infiltrate size at presentation (P<0.001), and pigmented ulcer (P=0.030). Larger infiltrate size at presentation was a significant predictor of worse 3-month infiltrate/scar size (P<0.001). Larger epithelial defect size was a significant predictor of perforation (P=0.0013). Predictors of longer time to re-epithelialization include infiltrate size at presentation (P<0.001) and older age (P=0.025). CONCLUSION: Ulcer severity at presentation is highly predictive of worse outcomes. Presentation of clinical characteristics such as baseline acuity and infiltrate scar can provide important information to clinicians about prognosis, and may help guide management and treatment decisions. Prevention of corneal ulcer remains important, as it is difficult to change the course of the ulcer once it has begun.

Microbiological cure times in acanthamoeba keratitis.

AIMS: The purpose of this study was to estimate the duration of treatment necessary for sequential acanthamoeba laboratory tests from corneal scrapings to become negative, and to assess predictors that affect this duration period. METHODS: We included all patients with at least one positive acanthamoeba culture or Giemsa stain at the F.I. Proctor Foundation Microbiology Laboratory from 1996 to 2009. A parametric survival analysis was performed among patients with repeat cultures to assess significant predictors for extended clearance time. Simulations were performed to estimate clearance time in the entire patient population, assuming imperfect sensitivity. RESULTS: Thirty-seven patients with laboratory evidence of acanthamoeba had testing at 69 time points. The median clearance time among eyes with repeat cultures was 42.5 days (interquartile range (IQR) 22.0-82.0 days; unadjusted parametric model). Initial visual acuity was the only predictor significantly associated with clearance time in univariate analyses (P<0.0001). Using initial visual acuity as a predictor for clearance time among the entire patient population, the estimated clearance time decreased to 38.7 days (95% confidence interval (CI) 27.9-53.5 days). When the imperfect sensitivity of the culture technique was also taken into account, the estimated clearance time was 44.1 days (95% CI 31.9-61.0 days). CONCLUSION: The duration of infection with acanthamoeba keratitis undergoing treatment has not been well characterized. In this report we estimate a median clearance time of approximately 6 weeks, with an IQR of 22-82 days.

A mathematical model of the ecology of Lyme disease.

A mathematical model of enzootic Lyme-disease transmission in a natural focus is presented. This model is based on the life history of the vector tick Ixodes scapularis Say and the primary reservoir host Peromyscus leucopus. Using this model, the threshold condition for the disease to be able to invade a nonenzootic region is determined as a function of the various possible transmission chains operating throughout the year. These expressions show that the transmission chain in which ticks acquire the disease from mice in the fall and transmit it back to mice as nymphs in the spring is the most important chain (contributing approximately 87% of the elasticity of the threshold for the parameter choices examined). Equilibrium disease levels were examined under the assumption of a constant tick population; these levels were determined as a function of tick and mouse density, the vertical transmission rate, the infectivity of mice, and the survivorship parameters of the ticks and of the tick-host contact rates. Vertical transmission has a disproportionately large effect, since unfed infected larval ticks have two opportunities to feed on mice, rather than only one opportunity (as for a newly infected unfed nymph). Finally, a global sensitivity analysis based on Latin hypercube sampling is performed, in which is shown the importance of quantifying the natural history of infection in mice, and of elucidating the contribution of other hosts for I. scapularis than mice.

Leprosy and tuberculosis: the epidemiological consequences of cross-immunity.

OBJECTIVES: This study tested the hypothesis, first proposed by Chaussinand, that individual-level immunity acquired from exposure to tuberculosis may have contributed to the disappearance of leprosy from western Europe. METHODS: The epidemiological consequences of cross-immunity were assessed by the formulation of a mathematical model of the transmission dynamics of tuberculosis and leprosy. RESULTS: The conditions under which Mycobacterium tuberculosis could have eradicated Mycobacterium leprae were derived in terms of the basic reproductive rates of the two infections and the degree of cross-immunity. CONCLUSIONS: If the degree of cross-immunity between two diseases within an individual is known, then the epidemiological consequences of this cross-immunity can be assessed with transmission modeling. The results of this analysis, in combination with previous estimates of the basic reproductive rate of tuberculosis and degree of cross-immunity, imply that tuberculosis could have contributed to the decline of leprosy if the basic reproductive rate of leprosy was low.

Quantifying the intrinsic transmission dynamics of tuberculosis.

Previously we have formulated transmission models of untreated tuberculosis epidemics (Blower et al., Nature, Medicine 1 (1995), 815-821); in this paper, we present time-dependent uncertainty and sensitivity analyses in order to quantitatively understand the transmission dynamics of tuberculosis epidemics in the absence of treatment. The time-dependent uncertainty analysis enabled us to evaluate the variability in the epidemiological outcome variables of the model during the progression of a tuberculosis epidemic. Calculated values (from the uncertainty analysis) for the disease incidence, disease prevalence, and mortality rates were approximately consistent with historical data. The time-dependent sensitivity analysis revealed that only a few of the model's input parameters significantly affected the severity of a tuberculosis epidemic; these parameters were the disease reactivation rate, the fraction of infected individuals who develop tuberculosis soon after infection, the number of individuals that an infectious individual infects per year, the disease death rate, and the population recruitment rate. Our analysis demonstrates that it is possible to improve our understanding of the behavior of tuberculosis epidemics by applying time-dependent uncertainty and sensitivity analysis to a transmission model.

Amplification dynamics: predicting the effect of HIV on tuberculosis outbreaks.

HIV affects the pathogenesis and the transmission of Mycobacterium tuberculosis. We used a discrete event simulation model to predict the potential impact of HIV on increasing the probability and the expected severity of tuberculosis outbreaks. Our predictions reveal that an HIV epidemic can significantly increase the frequency and severity of tuberculosis outbreaks, but that this amplification effect of HIV on tuberculosis outbreaks is very sensitive to the tuberculosis treatment rate. At moderate or low treatment rates, even a moderate HIV epidemic can cause the average size of tuberculosis outbreaks to almost double in comparison with the expected outbreak size when HIV is absent. However, we determined that the amplification effect of HIV can be substantially reduced if the treatment rate of tuberculosis is very high. We discuss the significant implications of these results for the global control of tuberculosis. Our results also reveal that occasionally a "normal-virulence" strain of M. tuberculosis can be expected to generate a large outbreak. We discuss the implications of these results in understanding the virulence of M. tuberculosis and in the planned elimination of tuberculosis in the United States.

Projected incidence of AIDS in San Francisco: the peak and decline of the epidemic.

To predict the incidence of AIDS from 1978 through 1998 in San Francisco, we developed a model that combined annual HIV seroconversion rates for homosexual and bisexual men and for heterosexual injecting drug users with estimates of the incubation period distribution between HIV seroconversion and AIDS diagnosis and with estimates of the size of the at-risk populations. Our model assumed the availability of antiretroviral therapy at the efficacy level of zidovudine monotherapy. The annual number of new AIDS cases is estimated to have peaked at 3332 in 1992, and is projected to decline to 1196 annually by 1998. Although the projected number of cases decreased steadily during this period for homosexual and bisexual men, the projected number of cases for injection drug users, women, and persons with other risks increased between 1993 and 1998. The decline in the incidence of AIDS in San Francisco reflects the dramatic reductions in new HIV infections that occurred a decade previously and that were achieved as a result of significant changes in high-risk behaviors, primarily among homosexual and bisexual men. Changes in HIV seroincidence must be factored in before attributing the decrease in AIDS incidence to more effective combination antiretroviral treatment.

Risk of infection from needle reuse at a phlebotomy center.

OBJECTIVES: This study determined infection risk for HIV, hepatitis B virus (HBV), and hepatitis C virus (HCV) from needle reuse at a phlebotomy center that possibly exposed 3810 patients to infection. METHODS: We used a model for the risk of infection per blood draw, supplemented by subsequent testing results from 1699 patients. RESULTS: The highest risk of transmission was for HBV infection: 1.1 x 10(-6) in the best case and 1.2 x 10(-3) in the (unlikely) worst case. Subsequent testing yielded prevalence rates of 0.12%, 0.41%, and 0.88% for HIV, HBV, and HCV, respectively, lower than National Health and Nutrition Examination Survey III prevalence estimates. CONCLUSIONS: The infection risk was very low; few, if any, transmissions are likely to have occurred.

The continuing HIV epidemic among men who have sex with men.

OBJECTIVES: This study characterized the AIDS epidemic among urban men who have sex with men (MSM). METHODS: A probability sample of MSM was obtained in 1997 (n = 2881; 18 years and older) from New York, Los Angeles, Chicago, and San Francisco, and HIV status was determined through self-report and biological measures. RESULTS: HIV prevalence was 17% (95% confidence interval = 15%, 19%) overall, with extremely high levels in African Americans (29%), MSM who used injection drugs (40%), "ultraheavy" noninjection drug users (32%), and less educated men (< high school, 37%). City-level HIV differences were non-significant once these other factors were controlled for. In comparing the present findings with historical data based on public records and modeling, HIV prevalence appears to have declined as a result of high mortality (69%) and stable, but high, incidence rates (1%-2%). CONCLUSIONS: Although the findings suggest that HIV prevalence has declined significantly from the mid-1980s, current levels among urban MSM in the United States approximate those of sub-Saharan countries (e.g., 14%-25%) and are extremely high in many population subsegments. Despite years of progress, the AIDS epidemic continues unabated among subsegments of the MSM community.