Survival Analysis: Where, Why, What and How?

Durations of any event, such as duration of hospitalization, is usually found to have a highly skewed distribution and incomplete values due to dropouts and limited follow-up. The usual methods of statistical analysis are, therefore, not applicable. The method of survival analysis is a nonparametric method and is designed to overcome these problems. Survival is a generic term and is used for any time-toevent data. The entire survival pattern at different points in time is studied by the Kaplan-Meier method under certain conditions. Logrank method is used to compare survival pattern in two or more groups. Hazard is the rate of occurrence of an event per unit of time and studied by Cox method. The concept of survival and all these methods of survival analysis are briefly discussed in this short note in a non-mathematical format for medical audience.

Reporting of Basic Statistical Methods in Biomedical Journals: Improved SAMPL Guidelines

Statistical methods have become an essential component of all empirical biomedical research. Science requires that these methods arefully reported with complete accuracy so that the evidence base could be fully appraised for validity, reliability, and generalizability. Tomeet this objective, Statistical Analyses and Methods in Published Literature (SAMPL) guidelines have been prepared for statisticalreporting in biomedical publications. This communication proposes substantial improvement of these guidelines to make them morecomprehensive, organized, compact, and easier to adopt.

Evaluation of quality of multivariable logistic regression in Indian medical journals using multilevel modeling approach.

Background: Availability of user-friendly statistical software has increased the application of multivariable logistic regression (MLR) in the medical journal many fold. The reporting quality in terms of checking assumptions, model building strategies, proper coding, and report format need proper care and attention to communicate correct and reliable model results. Objective: The objective of this article is to evaluate the quality of MLR article based on 10-point well establish criteria and to study the factors that may influence the quality. Methods: Study included PubMed indexed Indian medical journals as on March 2010 and published at least ten original articles that applied MLR during 10 years was included in the study. Multilevel modeling was applied to assess the role of journal and article attributes on MLR quality. Results: Twelve out of 39 Indian PubMed indexed journals fulfilled the inclusion criterion. Of a total 5599 original articles in these journals, 262 (4.68%) applied MLR in their study. Conformity of linear gradient assumption for continuous covariate was the least fulfilled criterion. One-third of the MLR articles involved statistician or epidemiologist as co-author, and almost same number of MLR articles' first author was from outside India. The trend of 10-point criteria remained consistent although the number of MLR articles increased over the period. The average quality score was 3.78 (95% confidence interval: 2.97-4.60) out of a possible 10. Larger sample size, involvement of statistician as co-author, non-Indian as the first author, and use of SAS/STATA software increased the quality of MLR articles. Conclusions: The quality of MLR articles in Indian medical journals is lagging behind as compared to the quality of MLR articles published from the United States and Europe medical journals. Joint effort of editors, reviewers, and authors are required to improve the quality of MLR in Indian journals so that the reader gets the correct results.

Demystifying LMS and BCPE Methods of Centile Estimation for Growth and Other Health Parameters.

Lambda-Mu-Sigma and Box-Cox Power Exponential are popular methods for constructing centile curves but are difficult to understand for medical professionals. As a result, the methods are used by experts only. Non-experts use software as a blackbox that can lead to wrong curves. This article explains these methods in a simple non-mathematical language so that medical professionals can use them correctly and confidently.

Receiver Operating Characteristic (ROC) Curve for Medical Researchers.

Sensitivity and specificity are two components that measure the inherent validity of a diagnostic test for dichotomous outcomes against a gold standard. Receiver operating characteristic (ROC) curve is the plot that depicts the trade-off between the sensitivity and (1-specificity) across a series of cut-off points when the diagnostic test is continuous or on ordinal scale (minimum 5 categories). This is an effective method for assessing the performance of a diagnostic test. The aim of this article is to provide basic conceptual framework and interpretation of ROC analysis to help medical researchers to use it effectively. ROC curve and its important components like area under the curve, sensitivity at specified specificity and vice versa, and partial area under the curve are discussed. Various other issues such as choice between parametric and non-parametric methods, biases that affect the performance of a diagnostic test, sample size for estimating the sensitivity, specificity, and area under ROC curve, and details of commonly used softwares in ROC analysis are also presented.

Effect of vitamin A supplementation on childhood morbidity and mortality: critical review of Indian studies.

OBJECTIVE: To review all published randomized trials concerned with linkage of vitamin A supplementation with reduction of mortality and morbidity in Indian children. METHOD: The studies were identified by searching the PubMed, review articles, references of available meta-analyses and bibliography of pertinent references. Studies were extracted and the quality of each study was reviewed with regards to 10 categories of parameters that in our opinion, were important for a vitamin A prophylaxis trial. These included background indicators, subjects, design, intervention, proximal measures, field-work, sources of bias, data analysis, interpretation and documentation. RESULTS: Out of 12 studies satisfying the inclusion criteria, the available 11 were examined. Two of the trials were concerned with mortality, 6 with morbidity, and 2 with both mortality and morbidity; 1 study assessed the impact of vitamin A on pneumococcal colonization. Out of 4 mortality trials, only one could satisfactorily report a significant reduction (54 percent) in child mortality following vitamin A supplementation. Of 8 morbidity studies, only 3 indicated some beneficial effect of vitamin A supplementation. None of the studies was perfect in methodology. We could not locate any study that addressed the issue of cost-effectiveness or dietary modifications. The results were not unequivocal and findings for mortality and morbidity were not corroborative. CONCLUSION: There is no definite evidence as yet in favor or against substantial benefit of universal vitamin A supplementation to children in India. There is a clear need to undertake a comprehensive trial with adequate sample size and a standardized methodology that could give clear, unbiased, and convincing evidence on the role of routine vitamin A supplementation.

Estimates of the years-of-life-lost due to the top nine causes of death in rural areas of major states in India in 1995.

BACKGROUND: Years-of-life-lost (YLL) contribute nearly two-thirds of the disability-adjusted life-years (DALYs) worldwide and are especially Important for India where infant and child mortality is still high. These were estimated for India under the Global Burden of Disease study for the year 1990. No estimates are available for the different states of India. We aimed to prepare state-wise estimates of YLL for different causes of death in rural areas and to determine the causes responsible for a higher burden in different states. METHOD: Percentage deaths of the top 9 causes reported in the Registrar-General's Survey of Causes of Deaths (Rural)--1995 in 13 major states of India and different age groups was applied to the expected number of total deaths. The life lost according to the standard life-table was age-weighted and discounted using the methodology of the Global Burden of Disease 1990 study. The causes of death were based on lay reporting which otherwise seem reliable. RESULTS: The all-cause YLL in rural India in 1995 were 207 per 1,000 population. The minimum was 74 in Kerala and maximum 276 in Madhya Pradesh. Pneumonia was the top cause responsible for 15 YLL. The inter-state variation was high as Tamil Nadu had only 1.6 and Uttar Pradesh 30.5 YLL from this cause. Cancers were a uniform burden across the states. Heart attack, and bronchitis and asthma cut across the more and less developed states. Suicides were a heavy burden in Andhra Pradesh and vehicular accidents in Haryana and Rajasthan. Bihar, Gujarat, Madhya Pradesh, Orissa, Rajasthan and Uttar Pradesh had communicable and nutritional conditions as predominant causes while Kerala and Punjab had non-communicable diseases as the predominant cause of YLL due to premature mortality. CONCLUSION: These results provide a new perspective about the causes of death that need more attention in rural areas of different states of India. These will also help prioritize areas which require more inputs at the state-level and hence will be useful for health policymakers.