Search

Background: Estimating the required sample size and statistical power for a study is an integral part of study design. For standard designs, power equations provide an efficient solution to the problem, but they are unavailable for many complex study designs that arise in practice. For such complex study designs, computer simulation is a useful alternative for estimating study power. Although this approach is well known among statisticians, in our experience many epidemiologists and social scientists are unfamiliar with the technique. This article aims to address this knowledge gap. Methods: We review an approach to estimate study power for individual- or cluster-randomized designs using computer simulation. This flexible approach arises naturally from the model used to derive conventional power equations, but extends those methods to accommodate arbitrarily complex designs. The method is universally applicable to a broad range of designs and outcomes, and we present the material in a way that is approachable for quantitative, applied researchers. We illustrate the method using two examples (one simple, one complex) based on sanitation and nutritional interventions to improve child growth. Results: We first show how simulation reproduces conventional power estimates for simple randomized designs over a broad range of sample scenarios to familiarize the reader with the approach. We then demonstrate how to extend the simulation approach to more complex designs. Finally, we discuss extensions to the examples in the article, and provide computer code to efficiently run the example simulations in both R and Stata. Conclusions: Simulation methods offer a flexible option to estimate statistical power for standard and nontraditional study designs and parameters of interest. The approach we have described is universally applicable for evaluating study designs used in epidemiologic and social science research. [ABSTRACT FROM AUTHOR]

Background: For individuals with AIDS, data exist relatively soon after diagnosis to allow estimation of "early" survival quantiles (e.g., the 0.10, 0.15, 0.20 and 0.30 quantiles, etc.). Many years of additional observation must elapse before median survival, a summary measure of survival, can be estimated accurately. In this study, a new approach to predict AIDS median survival is presented and its accuracy tested using AIDS surveillance data. Methods: The data consisted of 96,373 individuals who were reported to the HIV/AIDS Reporting System of the California Department of Health Services Office of AIDS as of December 31, 1996. We defined cohorts based on quarter year of diagnosis (e.g., the "931" cohort consists of individuals diagnosed with AIDS in the first quarter of 1993). We used early quantiles (estimated using the Inverse Probability of Censoring Weighted estimator) of the survival distribution to estimate median survival by assuming a linear relationship between the earlier quantiles and median survival. From this model, median survival was predicted for cohorts for which a median could not be estimated empirically from the available data. This prediction was compared with the actual medians observed when using updated survival data reported at least five years later. Results: Using the 0.15 quantile as the predictor and the data available as of December 31, 1996, we were able to predict the median survival of four cohorts (933, 934, 941, and 942) to be 34, 34, 31, and 29 months. Without this approach, there were insufficient data with which to make any estimate of median survival. The actual median survival of these four cohorts (using data as of December 31, 2001) was found to be 32, 40, 46, and 80 months, suggesting that the accuracy for this approach requires a minimum of three years to elapse from diagnosis to the time an accurate prediction can be made. Conclusion: The results of this study suggest that early and accurate prediction of median survival time after AIDS diagnosis may be possible using early quantiles of the survival distribution. The methodology did not seem to work well during a period of significant change in survival as observed with highly active antiretroviral treatment, but results suggest that it may work well in a time of more gradual improvement in survival. [ABSTRACT FROM AUTHOR]

Background: More than 200 studies related to nucleic acid amplification (NAA) tests to detect Mycobacterium tuberculosis directly from clinical specimens have appeared in the world literature since this technology was first introduced. NAA tests come as either commercial kits or as tests designed by the reporting investigators themselves (in-house tests). In-house tests vary widely in their accuracy, and factors that contribute to heterogeneity in test accuracy are not well characterized. Here, we used meta-analytical methods, including meta-regression, to identify factors related to study design and assay protocols that affect test accuracy in order to identify those factors associated with high estimates of accuracy. Results: By searching multiple databases and sources, we identified 2520 potentially relevant citations, and analyzed 84 separate studies from 65 publications that dealt with in-house NAA tests to detect M. tuberculosis in sputum samples. Sources of heterogeneity in test accuracy estimates were determined by subgroup and meta-regression analyses. Among 84 studies analyzed, the sensitivity and specificity estimates varied widely; sensitivity varied from 9.4% to 100%, and specificity estimates ranged from 5.6% to 100%. In the meta-regression analysis, the use of IS6110 as a target, and the use of nested PCR methods appeared to be significantly associated with higher diagnostic accuracy. Conclusion: Estimates of accuracy of in-house NAA tests for tuberculosis are highly heterogeneous. The use of IS6110 as an amplification target, and the use of nested PCR methods appeared to be associated with higher diagnostic accuracy. However, the substantial heterogeneity in both sensitivity and specificity of the in-house NAA tests rendered clinically useful estimates of test accuracy difficult. Future development of NAA-based tests to detect M. tuberculosis from sputum specimens should take into consideration these findings in improving accuracy of in-house NAA tests. [ABSTRACT FROM AUTHOR]

Background: Conventional tests for tuberculous pleuritis have several limitations. A variety of new, rapid tests such as nucleic acid amplification tests -- including polymerase chain reaction -- have been evaluated in recent times. We conducted a systematic review to determine the accuracy of nucleic acid amplification (NAA) tests in the diagnosis of tuberculous pleuritis. Methods: A systematic review and meta-analysis of 38 English and Spanish articles (with 40 studies), identified via searches of six electronic databases, hand searching of selected journals, and contact with authors, experts, and test manufacturers. Sensitivity, specificity, and other measures of accuracy were pooled using random effects models. Summary receiver operating characteristic curves were used to summarize overall test performance. Heterogeneity in study results was formally explored using subgroup analyses. Results: Of the 40 studies included, 26 used in-house ("home-brew") tests, and 14 used commercial tests. Commercial tests had a low overall sensitivity (0.62; 95% confidence interval [CI] 0.43, 0.77), and high specificity (0.98; 95% CI 0.96, 0.98). The positive and negative likelihood ratios for commercial tests were 25.4 (95% CI 16.2, 40.0) and 0.40 (95% CI 0.24, 0.67), respectively. All commercial tests had consistently high specificity estimates; the sensitivity estimates, however, were heterogeneous across studies. With the in-house tests, both sensitivity and specificity estimates were significantly heterogeneous. Clinically meaningful summary estimates could not be determined for in-house tests. Conclusions: Our results suggest that commercial NAA tests may have a potential role in confirming (ruling in) tuberculous pleuritis. However, these tests have low and variable sensitivity and, therefore, may not be useful in excluding (ruling out) the disease. NAA test results, therefore, cannot replace conventional tests; they need to be interpreted in parallel with clinical findings and results of conventional tests. The accuracy of in-house nucleic acid amplification tests is poorly defined because of heterogeneity in study results. The clinical applicability of inhouse NAA tests remains unclear. [ABSTRACT FROM AUTHOR]

Background: Cryptosporidiosis, caused by Cryptosporidium, is an enteric illness that has received much attention as an infection of immunocompromised persons as well as in community outbreaks (frequently waterborne). There are, however, no studies of the risk factors for sporadic community-acquired cryptosporidiosis in the immunocompetent US population. We undertook a case-control study in the San Francisco Bay Area as part of a national study sponsored by the Centers for Disease Control and Prevention to ascertain the major routes of transmission for endemic cryptosporidiosis, with an emphasis on evaluating risk from drinking water. Methods: Cases were recruited from a population-based, active surveillance system and age-matched controls were recruited using sequential random-digit dialing. Cases (n = 26) and controls (n = 62) were interviewed by telephone using a standardized questionnaire that included information about the following exposures: drinking water, recreational water, food items, travel, animal contact, and person-to-person fecal contact, and (for adults) sexual practices. Results: In multivariate conditional logistic regression analyses no significant association with drinking water was detected. The major risk factor for cryptosporidiosis in the San Francisco Bay Area was travel to another country (matched odds ratio [95% confidence interval]: 24.1 [2.6, 220]). Conclusion: The results of this study do not support the hypothesis that drinking water is an independent risk factor for cryptosporidiosis among the immunocompetent population. These findings should be used to design larger studies of endemic cryptosporidiosis to elucidate the precise mechanisms of transmission, whether waterborne or other. [ABSTRACT FROM AUTHOR]

Field studies often use caregiver-reported diarrhoea and related symptoms to measure child morbidity. There are various vernacular terms to define diarrhoea that vary across the local cultural contexts. The relationship between vernacular definitions of diarrhoea and symptoms-based definitions is not well-documented. This paper describes the association of the vernacular Quechua term k'echalera with the symptoms-based standard definition of diarrhoea in rural Bolivian settings. During a cluster randomized trial in rural Bolivia, both signs and symptoms of diarrhoea and reports of k'echalera were collected for children aged less than five years. Reported k'echalera were found to be associated with important changes in stool frequency, consistency, and presence of blood and mucus. Reported k'echalera were highly related to three of four recorded categories of watery stool. The intermediate (milk-rice) stool consistency, which fits into the definition of watery stool, was not strongly related to k'echalera. Mucus in the stool was also associated with k'echalera; however, its presence in k'echalera-free days accounted for at least 50% of the possible false negatives. The sensitivity and specificity of the term k'echalera were estimated by Bayesian methods, allowing for both symptoms of diarrhoea and reports of k'echalera to be subject to diagnosis error. An average specificity of at least 97% and the sensitivity of at least 50% were obtained. The findings suggest that the use of k'echalera would identify fewer cases of diarrhoea than a symptom-based definition in rural Bolivia. [ABSTRACT FROM AUTHOR]