Emerging Themes in Epidemiology

Background: An unprecedented number of nationwide tuberculosis (TB) prevalence surveys will be implemented between 2010 and 2015, to better estimate the burden of disease caused by TB and assess whether global targets for TB control set for 2015 are achieved. It is crucial that results are analysed using best-practice methods. Objective: To provide new theoretical and practical guidance on best-practice methods for the analysis of TB prevalence surveys, including analyses at the individual as well as cluster level and correction for biases arising from missing data. Analytic methods: TB prevalence surveys have a cluster sample survey design; typically 50-100 clusters are selected, with 400-1000 eligible individuals in each cluster. The strategy recommended by the World Health Organization (WHO) for diagnosing pulmonary TB in a nationwide survey is symptom and chest X-ray screening, followed by smear microscopy and culture examinations for those with an abnormal X-ray and/or TB symptoms. Three possible methods of analysis are described and explained. Method 1 is restricted to participants, and individuals with missing data on smear and/or culture results are excluded. Method 2 includes all eligible individuals irrespective of participation, through multiple missing value imputation. Method 3 is restricted to participants, with multiple missing value imputation for individuals with missing smear and/or culture results, and inverse probability weighting to represent all eligible individuals. The results for each method are then compared and illustrated using data from the 2007 national TB prevalence survey in the Philippines. Simulation studies are used to investigate the performance of each method. Key findings: A cluster-level analysis, and Methods 1 and 2, gave similar prevalence estimates (660 per 100,000 aged ≥ 10 years old), with a higher estimate using Method 3 (680 per 100,000). Simulation studies for each of 4 plausible scenarios show that Method 3 performs best, with Method 1 systematically underestimating TB prevalence by around 10%. Conclusion: Both cluster-level and individual-level analyses should be conducted, and individual-level analyses should be conducted both with and without multiple missing value imputation. Method 3 is the safest approach to correct the bias introduced by missing data and provides the single best estimate of TB prevalence at the population level. * Correspondence: [email protected] Equal contributors London School of Hygiene & Tropical Medicine, London, UK Full list of author information is available at the end of the article © 2013 Floyd et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Floyd et al. Emerging Themes in Epidemiology 2013, 10:10 Page 2 of 13 http://www.ete-online.com/content/10/1/10 Background National population-based surveys of the prevalence of pulmonary tuberculosis (TB) disease in adults can be used to measure the burden of disease caused by TB, to measure trends in this burden when repeat surveys are performed and to understand why people with TB have not been detected or diagnosed by national TB control programmes (NTPs). Surveys are of greatest relevance in countries with a high burden of TB in which surveillance data capture much less than 100% of cases. Global targets for reductions in disease burden set for 2015 include halving prevalence rates compared with their level in 1990; the other targets are that mortality rates should be halved between 1990 and 2015, and that TB incidence should be falling by 2015 [1]. The Global Task Force on TB Impact Measurement is hosted by the World Health Organization (WHO) with a mandate to ensure the best-possible assessment of whether 2015 global targets for reductions in TB disease burden are achieved [2]. The Task Force has strongly recommended national TB prevalence surveys in 22 global focus countries in the years leading up to 2015 [3,4]. Since 2008, there has been an unprecedented increase in the number of countries either implementing or planning to implement nationwide surveys. Between 2009 and 2015, approximately 23 countries including 20 of the global focus countries are expected to implement a survey, compared with a total of 7 countries in the period 2002– 2007 (Figure 1). Only four countries, all in Asia, implemented surveys between 1990 and 2001. The global investment in prevalence surveys will amount to around US$ 50 million between 2010 and 2015. Figure 1 Global progress with nationwide prevalence surveys of TB d nationwide surveys of the prevalence of TB disease, actual (2002–2012) and Analysis of results using best-practice methods is crucial. TB prevalence surveys have a cluster sample survey design, in which groups of individuals are sampled, with clusters selected at random from an area sampling frame with probability proportional to size (PPS). While the classic method of using each survey cluster as the unit of analysis has been carefully and thoroughly described for a TB prevalence survey [5,6], methods to implement an individual-level analysis, in which each eligible adult enumerated in the survey is the unit of analysis, have not. An individual-level analysis is valuable because it enables adjustment for differences between participants and non-participants and multiple imputation of missing data, while simultaneously allowing for clustering in the sampling design. Missing data in TB prevalence surveys can be observed in both the outcome (TB case or not) but also other covariates, for example due to non-participation of eligible individuals, unavailability of screening or diagnostic results due to human error, and loss of specimens at laboratories for reasons such as contamination. A prevalence estimate based on only individuals with complete data will be biased, except under the strong assumption that those with and without full information have the same prevalence of TB. Methods that incorporate missing value imputation are thus important for two reasons: to obtain a more valid estimate of pulmonary TB prevalence, and to assess the bias of simpler analytical approaches [7,8]. Moreover, while participation rates in recent surveys in Asia have been very high, the rates achieved in other surveys from 2012 onwards may be lower; accounting for missing data isease. Global progress in implementing field operations of expected (2013–2017). Floyd et al. Emerging Themes in Epidemiology 2013, 10:10 Page 3 of 13 http://www.ete-online.com/content/10/1/10 will become essential for production of robust results. Findings from national TB prevalence surveys completed in 2007 in the Philippines and Viet Nam have been published [9,10]. Other national surveys have either not followed the screening strategy now recommended by WHO [11,12], or the results have been disseminated in a survey report but not in a scientific journal. The analysis of the Philippines survey attempted to account for missing data using within-cluster mean imputation, stratified on age and sex, but did not include individual-level analysis. The analysis of the Vietnam survey used an individual-level analysis but did not formally account for missing data on smear and culture results, or age and sex differences between participants and non-participants. This paper (outlined in Figure 2) provides new theoretical and practical guidance on best-practice methods for analysis of data from a TB prevalence survey, notably methods for individual-level analyses that account for the cluster sample survey design and that allow correction for biases due to missing data. Methods are described and explained, and then illustrated and compared using data from the 2007 survey in the Philippines. We draw on material Figure 2 Paper outline. previously developed in 2010 by the authors in a WHO handbook [4] but provide much more explanation of the underlying principles and methods required to implement multiple imputation of missing data. This includes guidance based on insights gained in 2011 and 2012 through the analysis of prevalence surveys conducted in Myanmar (2010) [13], Ethiopia (2010/11) [14], and Cambodia (2010/11) [15]. We also place the analytical methods within a new conceptual framework [16,17].