Estimating Mortality After Natural Disasters In Low- And Middle-Income Countries: A Systematic Review
This thesis is restricted to Yale network users only. It will be made publicly available on 08/28/2020
The outcomes of natural disasters produce many complex uncertainties and statistics, many of which have no metrics to assess how accurate they are. In particular, mortality estimates in low- and middle-income countries encompass many variables in their inherently challenging physical, political, and logistical landscapes. Given that there has been no known systematic review of assessing mortality estimation methodologies, the author aimed to provide a systematic review of known methodologies and provide a framework to assess mortality estimation methodologies.
The author performed a systematic search of electronic databases from the past two decades up to January 2019. In addition, the author utilized study reference lists and expertise from two experts in humanitarian response. The author included studies that provided some type of methodology to estimate mortality after natural disasters. The author excluded studies that were not in low- or middle-income countries and studies that only focused on specific sub-populations (e.g. under-5 mortality).
Nineteen studies met the eligibility criteria. The studies used a range of methodologies, but consisted mainly of retrospective cohort studies, computer programs, and formulas and predictive models. Retrospective cohort studies (n=12) entailed the bulk of the results, while computer programs (n=3) and formulas and predictive models (n=4) entailed the rest.
This systematic review indicates that methodologies to estimate mortality after natural disasters are quite diverse and produce varying estimates. Even with the common use of retrospective cohort studies over time, the use of established and evidence-based epidemiological methods is rare. This review offers framework to assess mortality estimates after natural disasters but highlights the need to establish a reputable and accepted metric for policy makers, the media, and the public on how to interpret mortality estimates.