Model uncertainty quantification in Cox regression.

We consider covariate selection and the ensuing model uncertainty aspects in the context of Cox regression. The perspective we take is probabilistic, and we handle it within a Bayesian framework. One of the critical elements in variable/model selection is choosing a suitable prior for model parameters. Here, we derive the so-called conventional prior approach and propose a comprehensive implementation that results in an automatic procedure. Our simulation studies and real applications show improvements over existing literature. For the sake of reproducibility but also for its intrinsic interest for practitioners, a web application requiring minimum statistical knowledge implements the proposed approach.

A comprehensive study of the epidemiology of haematological malignancies in North Queensland.

BACKGROUND: There is an absence of clinically relevant epidemiological data in regional Australia pertaining to haematological malignancies. AIM: To determine the incidence and geographical variation of haematological malignancies in North Queensland using a clinically appropriate disease classification. METHODS: Retrospective, observational study of individual patient data records of all adults diagnosed with a haematological malignancy between 2005 and 2014 and residing within The Townsville Hospital Haematology catchment region. We report descriptive summaries, incidence rates and incidence-rate ratios of haematological malignancies by geographic regions. RESULTS: One thousand, five hundred and eighty-one haematological malignancies (69% lymphoid, 31% myeloid) were diagnosed over the 10-year study period. Descriptive data are presented for 58 major subtypes, as per the WHO diagnostic classification of tumours of haemopoietic and lymphoid tissues. The overall median age at diagnosis was 66 years with a male predominance (60%). We demonstrate a temporal increase in the incidence of haematological malignancies over the study period. We observed geographical variations in the age-standardised incidence rates per 100 000 ranging from 0.5 to 233.5. Our data suggest an increased incidence rate ratio for haematological malignancies in some postcodes within the Mackay area compared with other regions. CONCLUSION: The present study successfully reports on the incidence of haematological malignancies in regional Queensland using a clinically meaningful diagnostic classification system and identifies potential geographic hotspots. We advocate for such contemporary, comprehensive and clinically meaningful epidemiological data reporting of blood cancer diagnoses in wider Australia. Such an approach will have significant implications towards developing appropriate data-driven management strategies and public health responses for haematological malignancies.

Transmission Dynamics in Tuberculosis Patients With Human Immunodeficiency Virus: A Systematic Review and Meta-analysis of 32 Observational Studies.

BACKGROUND: There are large knowledge gaps on the transmission dynamics of Mycobacterium tuberculosis in settings where both tuberculosis and human immunodeficiency virus (HIV) are endemic. We aimed to assess the infectiousness of tuberculosis patients coinfected with HIV. METHODS: We systematically searched for studies of contacts of both HIV-positive and HIV-negative tuberculosis index cases. Our primary outcome was Mycobacterium tuberculosis infection in contacts. Data on sputum smear and lung cavitation status of index cases were extracted from each study to assess effect modification. Secondary outcomes included prevalent tuberculosis and HIV in contacts of HIV-positive and HIV-negative index cases. RESULTS: Of 5255 original citations identified, 32 studies met inclusion criteria, including 25 studies investigating M. tuberculosis infection (Nparticipants = 36 893), 13 on tuberculosis (Nparticipants = 18 853), and 12 on HIV positivity (Nparticipants = 18 424). Risk of M. tuberculosis infection was lower in contacts of HIV-positive index cases (odds ratio [OR], 0.67, 95% confidence interval [CI], .58-.77) but was heterogeneous (I2 = 75.1%). Two factors modified this relationship: the lung cavitary status of the index case and immunosuppression (measured through CD4 counts or HIV or acquired immunodeficiency syndrome diagnoses) among index people living with HIV. Rates of HIV were consistently higher in contacts of coinfected index cases (OR, 4.9; 95% CI, 3.0-8.0). This was modified by whether the study was in sub-Saharan Africa (OR, 2.8; 95% CI, 1.6-4.9) or in another global region (OR, 9.8; 95% CI, 5.9-16.3). CONCLUSIONS: Tuberculosis patients coinfected with HIV are less infectious than HIV-uninfected cases when they have severe immunosuppression or paucibacillary disease. Contacts of coinfected index cases are almost 5 times more likely to also have HIV.

Excess Risk of Tuberculosis Infection Among Extra-household Contacts of Tuberculosis Cases in an African City.

BACKGROUND: Although households of tuberculosis (TB) cases represent a setting for intense transmission of Mycobacterium tuberculosis, household exposure accounts for <20% of transmission within a community. The aim of this study was to estimate excess risk of M. tuberculosis infection among household and extra-household contacts of index cases. METHODS: We performed a cross-sectional study in Kampala, Uganda, to delineate social networks of TB cases and matched controls without TB. We estimated the age-stratified prevalence difference of TB infection between case and control networks, partitioned as household and extra-household contacts. RESULTS: We enrolled 123 index cases, 124 index controls, and 2415 first-degree network contacts. The prevalence of infection was highest among household contacts of cases (61.5%), lowest among household contacts of controls (25.2%), and intermediary among extra-household TB contacts (44.9%) and extra-household control contacts (41.2%). The age-adjusted prevalence difference between extra-household contacts of cases and their controls was 5.4%. The prevalence of infection was similar among the majority of extra-household case contacts and corresponding controls (47%). CONCLUSIONS: Most first-degree social network members of TB cases do not have adequate contact with the index case to experience additional risk for infection, but appear instead to acquire infection through unrecognized exposures with infectious cases in the community.

Evaluation of the durability of long-lasting insecticidal nets in Guatemala.

BACKGROUND: Insecticide-treated bed nets (ITNs) are widely used for the prevention and control of malaria. In Guatemala, since 2006, ITNs have been distributed free of charge in the highest risk malaria-endemic areas and constitute one of the primary vector control measures in the country. Despite relying on ITNs for almost 15 years, there is a lack of data to inform the timely replacement of ITNs whose effectiveness becomes diminished by routine use. METHODS: The survivorship, physical integrity, insecticide content and bio-efficacy of ITNs were assessed through cross-sectional surveys conducted at 18, 24 and 32 months after a 2012 distribution of PermaNet® 2.0 in a malaria focus in Guatemala. A working definition of 'LLIN providing adequate protection' was developed based on the combination of the previous parameters and usage of the net. A total of 988 ITNs were analysed (290 at 18 months, 349 at 24 months and 349 at 32 months). RESULTS: The functional survivorship of bed nets decreased over time, from 92% at 18 months, to 81% at 24 months and 69% at 32 months. Independent of the time of the survey, less than 80% of the bed nets that were still present in the household were reported to have been used the night before. The proportion of bed nets categorized as "in good condition" per World Health Organization (WHO) guidelines of the total hole surface area, diminished from 77% to 18 months to 58% at 32 months. The portion of ITNs with deltamethrin concentration less than 10 mg/m2 increased over time. Among the bed nets for which bioassays were conducted, the percentage that met WHO criteria for efficacy dropped from 90% to 18 months to 52% at 32 months. The proportion of long-lasting insecticidal nets (LLINs) providing adequate protection was 38% at 24 months and 21% at 32 months. CONCLUSIONS: At 32 months, only one in five of the LLINs distributed in the campaign provided adequate protection in terms of survivorship, physical integrity, bio-efficacy and usage. Efforts to encourage the community to retain, use, and properly care for the LLINs may improve their impact. Durability assessments should be included in future campaigns.

Blood cytokine, chemokine and growth factor profiling in a cohort of pregnant women from tropical countries.

The immune status of women changes during and after pregnancy, differs between blood compartments at delivery and is affected by environmental factors particularly in tropical areas endemic for multiple infections. We quantified the plasma concentration of a set of thirty-one TH1, TH2, TH17 and regulatory cytokines, pro-inflammatory and anti-inflammatory cytokines and chemokines, and growth factors (altogether biomarkers), in a cohort of 540 pregnant women from five malaria-endemic tropical countries. Samples were collected at recruitment (first antenatal visit), delivery (periphery, cord and placenta) and postpartum, allowing a longitudinal analysis. We found the lowest concentration of biomarkers at recruitment and the highest at postpartum, with few exceptions. Among them, IL-6, HGF and TGF-ß had the highest levels at delivery, and even higher concentrations in the placenta compared to peripheral blood. Placental concentrations were generally higher than peripheral, except for eotaxin that was lower. We also compared plasma biomarker concentrations between the tropical cohort and a control group from Spain at delivery, presenting overall higher biomarker levels the tropical cohort, particularly pro-inflammatory cytokines and growth factors. Only IL-6 presented lower levels in the tropical group. Moreover, a principal component analysis of biomarker concentrations at delivery showed that women from Spain grouped more homogenously, and that IL-6 and IL-8 clustered together in the tropical cohort but not in the Spanish one. Plasma cytokine concentrations correlated with Plasmodium antibody levels at postpartum but not during pregnancy. This basal profiling of immune mediators over gestation and in different compartments at delivery is important to subsequently understand response to infections and clinical outcomes in mothers and infants in tropical areas.

Defining adequate contact for transmission of Mycobacterium tuberculosis in an African urban environment.

BACKGROUND: The risk of infection from respiratory pathogens increases according to the contact rate between the infectious case and susceptible contact, but the definition of adequate contact for transmission is not standard. In this study we aimed to identify factors that can explain the level of contact between tuberculosis cases and their social networks in an African urban environment. METHODS: This was a cross-sectional study conducted in Kampala, Uganda from 2013 to 2017. We carried out an exploratory factor analysis (EFA) in social network data from tuberculosis cases and their contacts. We evaluated the factorability of the data to EFA using the Kaiser-Meyer-Olkin Measure of Sampling Adequacy (KMO). We used principal axis factoring with oblique rotation to extract and rotate the factors, then we calculated factor scores for each using the weighted sum scores method. We assessed construct validity of the factors by associating the factors with other variables related to social mixing. RESULTS: Tuberculosis cases (N = 120) listed their encounters with 1154 members of their social networks. Two factors were identified, the first named "Setting" captured 61% of the variance whereas the second, named 'Relationship' captured 21%. Median scores for the setting and relationship factors were 10.2 (IQR 7.0, 13.6) and 7.7 (IQR 6.4, 10.1) respectively. Setting and Relationship scores varied according to the age, gender, and nature of the relationship among tuberculosis cases and their contacts. Family members had a higher median setting score (13.8, IQR 11.6, 15.7) than non-family members (7.2, IQR 6.2, 9.4). The median relationship score in family members (9.9, IQR 7.6, 11.5) was also higher than in non-family members (6.9, IQR 5.6, 8.1). For both factors, household contacts had higher scores than extra-household contacts (p < .0001). Contacts of male cases had a lower setting score as opposed to contacts of female cases. In contrast, contacts of male and female cases had similar relationship scores. CONCLUSIONS: In this large cross-sectional study from an urban African setting, we identified two factors that can assess adequate contact between tuberculosis cases and their social network members. These findings also confirm the complexity and heterogeneity of social mixing.

P-value calibration in multiple hypotheses testing.

As p-values are the most common measures of evidence against a hypothesis, their calibration with respect to null hypothesis conditional probability is important in order to match frequentist unconditional inference with the Bayesian ones. The Selke, Bayarri and Berger calibration is one of the most popular attempts to obtain such a calibration. This relies on the theoretical sampling null distribution of p-values, which is the well-known Uniform(0,1), but arising only for specific sampling models. We generalize this calibration by considering a sampling null distribution estimated from the data. It is possible to obtain such an empirical null distribution, for instance, in the context of multiple testing in which many p-values come from the null model. Such a context is purely instrumental for the purposes of p-value calibration, and multiple testing still needs to be considered with appropriate techniques. The new calibration proposed here still remains a simple analytic formula like the original one under the Uniform(0,1) and basically provides a stronger interpretation framework for the widely used p-value. Copyright © 2017 John Wiley & Sons, Ltd.

Risk Analysis for Unintentional Slide Deployment During Airline Operations.

We present a risk analysis undertaken to mitigate problems in relation to the unintended deployment of slides under normal operations within a commercial airline. This type of incident entails relevant costs for the airline industry. After assessing the likelihood and severity of its consequences, we conclude that such risks need to be managed. We then evaluate the effectiveness of various countermeasures, describing and justifying the chosen ones. We also discuss several issues faced when implementing and communicating the proposed measures, thus fully illustrating the risk analysis process.

Comparison of objective Bayes factors for variable selection in parametric regression models for survival analysis.

This paper considers the problem of selecting a set of regressors when the response variable is distributed according to a specified parametric model and observations are censored. Under a Bayesian perspective, the most widely used tools are Bayes factors (BFs), which are undefined when improper priors are used. In order to overcome this issue, fractional (FBF) and intrinsic (IBF) BFs have become common tools for model selection. Both depend on the size, Nt , of a minimal training sample (MTS), while the IBF also depends on the specific MTS used. In the case of regression with censored data, the definition of an MTS is problematic because only uncensored data allow to turn the improper prior into a proper posterior and also because full exploration of the space of the MTSs, which includes also censored observations, is needed to avoid bias in model selection. To address this concern, a sequential MTS was proposed, but it has the drawback of an increase of the number of possible MTSs as Nt becomes random. For this reason, we explore the behaviour of the FBF, contextualizing its definition to censored data. We show that these are consistent, providing also the corresponding fractional prior. Finally, a large simulation study and an application to real data are used to compare IBF, FBF and the well-known Bayesian information criterion.

Incidence and risk factors for travellers' diarrhoea among short-term international adult travellers from high-income countries: a systematic review with meta-analysis of cohort studies.

BACKGROUND: Travellers' Diarrhoea (TD) continues to be the most common travel-related medical event in international travellers. Updated incidence and risk factor data will improve pre-travel medical advice for travellers from high-income countries (HIC), providing an opportunity for disease prevention, and appropriate disease management. METHODS: A systematic search for cohort studies of TD incidence published between 1 January 1997 and 2 March 2023 was performed using Ovid Medline, SCOPUS, and Google Scholar databases. Study quality was assessed with a modified Newcastle-Ottawa scale (NOS). We extracted incidence data for adults travelling less than 100 days from HIC and available risk factor data. The overall random-effects pooled incidence and the corresponding 95% confidence intervals (95% CI) were estimated. Heterogeneity was assessed using the I2 statistic, tau and the 95% prediction intervals. Subgroup analyses were conducted to identify sources of heterogeneity. Risk factor studies were reviewed qualitatively and described. RESULTS: Ten studies were included in the meta-analysis, containing 8478 participants. Two of the studies measured as high quality and eight as good quality as assessed by the modified NOS. The TD incidence was 36.1% (95% CI 24-41%; I2 94%), with a prediction interval ranging from 20.3% to 55.8%. The pooled incidence of mild, moderate, and severe TD was 23.6%, 8.1% and 2.9%, respectively. Subgroup analysis showed that the incidence increased with increasing average data collection period. Risk factors for TD in travellers from HIC identified include younger age, longer travel periods, low and middle-income destinations, travelling for tourism, backpacking travel styles and pre-travel health status. CONCLUSION: It is estimated that between 20 to 56% of international travellers can expect to develop TD in travel of under 100 days. While most cases are mild, approximately 3% of all travellers will experience a disease that prevents usual activities or requires medical attention.

M. tuberculosis Infection Attributable to Exposure in Social Networks of Tuberculosis Cases in an Urban African Community.

Background: The persistence of tuberculosis today and its global disparity send a powerful message that effective tuberculosis control must respond to its regional epidemiology. Active case finding through contact investigation is a standard protocol used for tuberculosis control, but its effectiveness has not been established, especially in endemic areas. Methods: To quantify the potential effectiveness of contact investigation in Kampala, Uganda, we used a cross-sectional design to evaluate the social networks of 123 tuberculosis index cases and 124 controls without tuberculosis. Results: Tuberculous infection was present in 515 of 989 tuberculosis case contacts (52.1%) and 396 of 1026 control contacts (38.6%; adjusted prevalence ratio, 1.4; 95% CI, 1.3-1.6). The proportion of infected participants with known exposure within the social network of the tuberculosis case was 35%. The population-attributable fraction was 11.1% for any known exposure, with 7.3% attributable to household exposure and 3.4% attributable to extrahousehold exposure. Conclusions: This low population-attributable fraction indicates that contact tracing in the social networks of index cases will have only a modest effect in reducing tuberculous infection in a community. New approaches to community-level active case finding are needed.

Next-generation vaccines for tropical infectious diseases.

Tropical infectious diseases inflict an unacceptable burden of disease on humans living in developing countries. Although anti-pathogenic drugs have been widely used, they carry a constant threat of selecting for resistance. Vaccines offer a promising means by which to enhance the global control of tropical infectious diseases; however, these have been difficult to develop, mostly because of the complex nature of the pathogen lifecycles. Here, we present recently developed vaccine candidates for five tropical infectious diseases in the form of a catalog that have either entered clinical trials or have been licensed for use. We deliberate on recently licensed dengue vaccines, provide evidence why combination vaccination could have a synergistic impact on schistosomiasis, critically appraise the value of typhoid conjugate vaccines, and discuss the potential of vaccines in the efforts to eliminate vivax malaria and hookworms.

Patient satisfaction among persons living with HIV/AIDS and receiving antiretroviral therapy in urban Uganda: A factor analysis.

INTRODUCTION: Patient satisfaction is an important predictor of health outcomes among patients in HIV/AIDS treatment and care, yet it is rarely measured in routine clinic settings in most of Africa. The aims of our study were to evaluate the internal validity and reliability of the Consumer Assessment of Healthcare Providers and Systems instrument for measuring satisfaction, assess the general level of patient satisfaction, and identify the factors associated with the level of satisfaction among patients receiving antiretroviral therapy in Uganda. MATERIALS AND METHODS: We conducted a cross-sectional study of 475 HIV/AIDS-infected patients from July to August 2015 in Kampala, Uganda. Eligible participants were 18 years or older, consented to the study and receiving antiretroviral therapy and outpatient care at the selected public health clinic. This study used a modified version of the validated Consumer Assessment of Healthcare Providers and Systems (CAHPS) instrument to assess the level of satisfaction among HIV/AIDS patients receiving outpatient care. We collected data on socio-demographics, clinical variables and 18-items adapted from the CAHPS instrument rating satisfaction with aspects of health services. We conducted an exploratory factor analysis to assess the internal validity of the 18 items and multiple linear regression analysis of factors associated with patient satisfaction with care. RESULTS: Majority of the respondents were females (76.8%), and the mean age was 37 years (SD = 10). The modified CAHPS instrument had high internal consistency (Cronbach's α = 0.94) for measuring satisfaction with HIV/AIDS care. Female sex (p = 0.016), perceived providers' technical and interpersonal skills (p = 0.022), emotional health (p = 0.032), and quality of reception services (p<0.001) were significantly associated with satisfaction in this urban HIV/AIDS public clinic. CONCLUSION: The reliability of the CAHPS instrument was high for measuring satisfaction. Providers' technical and interpersonal skills, and the quality of reception services are key to achieving patient satisfaction. Health system interventions to address the gaps identified will enhance the quality of patient-centered HIV/AIDS care in the Ugandan setting.

Plasmodium vivax congenital malaria in an area of very low endemicity in Guatemala: implications for clinical and epidemiological surveillance in a malaria elimination context.

This is a report of the first Plasmodium vivax congenital malaria case in Guatemala and the first case in Latin America with genotypical, histological and clinical characterization. The findings show that maternal P. vivax infection still occurs in areas that are in the pathway towards malaria elimination, and can be associated with detrimental health effects for the neonate. It also highlights the need in very low transmission areas of not only maintaining, but increasing awareness of the problem and developing surveillance strategies, based on population risk, to detect the infection especially in this vulnerable group of the population.

Characterization of the Proportion of Clustered Tuberculosis Cases in Guatemala: Insights from a Molecular Epidemiology Study, 2010-2014.

There is little information about the amount of recent tuberculosis transmission in low-income settings. Genetic clustering can help identify ongoing transmission events. A retrospective observational study was performed on Mycobacterium tuberculosis isolates from persons living with HIV (PLHIV) and HIV-seronegative participants who submitted samples to a referral tuberculosis laboratory in Guatemala City, Guatemala from 2010 to 2014. Genotyping results were classified according to the international spoligotyping database, SITVIT2. Spoligotype patterns were categorized as clustered or nonclustered depending on their genotype. The proportion of clustering and the index of recent transmission index (RTIn-1) were estimated. In the RTIn-1 method, clustered cases represent recent transmission, whereas nonclustered cases represent reactivation of older tuberculosis infections. As a secondary aim, the potential risk factors associated with clustering in isolates from the subset of participants living with HIV were explored. From 2010 to 2014, a total of 479 study participants were confirmed as culture-positive tuberculosis cases. Among the 400 available isolates, 71 spoligotype patterns were identified. Overall, the most frequent spoligotyping families were Latin American-Mediterranean (LAM) (39%), followed by T (22%) and Haarlem (14%). Out of the 400 isolates, 365 were grouped in 36 clusters (range of cluster size: 2-92). Thus, the proportion of clustering was 91% and the RTIn-1 was 82%. Among PLHIV, pulmonary tuberculosis was associated with clustering (OR = 4.3, 95% CI 1.0-17.7). Our findings suggest high levels of ongoing transmission of M. tuberculosis in Guatemala as revealed by the high proportion of isolates falling into genomic clusters.

A strategy analysis for genetic association studies with known inbreeding.

BACKGROUND: Association studies consist in identifying the genetic variants which are related to a specific disease through the use of statistical multiple hypothesis testing or segregation analysis in pedigrees. This type of studies has been very successful in the case of Mendelian monogenic disorders while it has been less successful in identifying genetic variants related to complex diseases where the insurgence depends on the interactions between different genes and the environment. The current technology allows to genotype more than a million of markers and this number has been rapidly increasing in the last years with the imputation based on templates sets and whole genome sequencing. This type of data introduces a great amount of noise in the statistical analysis and usually requires a great number of samples. Current methods seldom take into account gene-gene and gene-environment interactions which are fundamental especially in complex diseases. In this paper we propose to use a non-parametric additive model to detect the genetic variants related to diseases which accounts for interactions of unknown order. Although this is not new to the current literature, we show that in an isolated population, where the most related subjects share also most of their genetic code, the use of additive models may be improved if the available genealogical tree is taken into account. Specifically, we form a sample of cases and controls with the highest inbreeding by means of the Hungarian method, and estimate the set of genes/environmental variables, associated with the disease, by means of Random Forest. RESULTS: We have evidence, from statistical theory, simulations and two applications, that we build a suitable procedure to eliminate stratification between cases and controls and that it also has enough precision in identifying genetic variants responsible for a disease. This procedure has been successfully used for the beta-thalassemia, which is a well known Mendelian disease, and also to the common asthma where we have identified candidate genes that underlie to the susceptibility of the asthma. Some of such candidate genes have been also found related to common asthma in the current literature. CONCLUSIONS: The data analysis approach, based on selecting the most related cases and controls along with the Random Forest model, is a powerful tool for detecting genetic variants associated to a disease in isolated populations. Moreover, this method provides also a prediction model that has accuracy in estimating the unknown disease status and that can be generally used to build kit tests for a wide class of Mendelian diseases.

Clinical and microbiological comparison of three non-surgical protocols for the initial treatment of chronic periodontitis.

OBJECTIVE: To compare the clinical and microbiological effects of three protocols for nonsurgical periodontal therapy, including full-mouth scaling and root planing plus systemic antibiotics, on the treatment of chronic periodontitis patients. METHODS: Twenty-nine patients diagnosed with moderate to severe chronic periodontitis, selected according to specific criteria, were randomly assigned to one of three treatment groups: quadrant scaling, full-mouth scaling, and full-mouth scaling supplemented by systemic antibiotics. Antibiotic selection was based on the results of individual susceptibility testing. Oral hygiene instructions and reinforcement were given during the study. All patients received a clinical periodontal and microbiological examination at baseline and at reexamination, 4-6 weeks after therapy. Means and standard deviations were calculated and differences between groups were analyzed via the Kruskal-Wallis test, p < 0.05. RESULTS: The mean age of the study sample was 49.1 + 11.6 years old, and there were 17 men and 12 women. Patients treated with antibiotics showed antimicrobial susceptibility for amoxicillin and doxycycline. All study groups showed a similar significant improvement in periodontal parameters. Plaque scores were reduced in a range of 29.0% to 42.6%. Bleeding on probing was reduced by 34.8% to 55.0%; the reduction for the full-mouth scaling group was larger. Mean reduction in pocket depth was 1.2 to 1.3 mm in all groups. Mean bacterial counts were reduced in the groups receiving full-mouth treatment, but not in the quadrant treatment group. CONCLUSION: The three protocols for non-surgical periodontal treatment demonstrated a similar positive effect on clinical parameters; however, only full-mouth treatment groups showed a reduction in anaerobic microbial counts at re-examination.

Defining an intermediate category of tuberculin skin test: A mixture model analysis of two high-risk populations from Kampala, Uganda.

One principle of tuberculosis control is to prevent the development of tuberculosis disease by treating individuals with latent tuberculosis infection. The diagnosis of latent infection using the tuberculin skin test is not straightforward because of concerns about immunologic cross reactivity with the Bacille Calmette-Guerin (BCG) vaccine and environmental mycobacteria. To parse the effects of BCG vaccine and environmental mycobacteria on the tuberculin skin test, we estimated the frequency distribution of skin test results in two divisions of Kampala, Uganda, ten years apart. We then used mixture models to estimate parameters for underlying distributions and defined clinically meaningful criteria for latent infection, including an indeterminate category. Using percentiles of two underlying normal distributions, we defined two skin test readings to demarcate three ranges. Values of 10 mm or greater contained 90% of individuals with latent infection; values less than 7.2 mm contained 80% of individuals without infection. Contacts with values between 7.2 and 10 mm fell into an indeterminate zone where it was not possible to assign infection. We conclude that systematic tuberculin skin test surveys within populations at risk, combined with mixture model analysis, may be a reproducible, evidence-based approach to define meaningful criteria for latent tuberculosis infection.