Related Experiment Video
Updated: Jul 25, 2026

Inverse Probability of Treatment Weighting (Propensity Score) using the Military Health System Data Repository and National Death Index
Published on: January 8, 2020
Deaths and injuries due to the earthquake in Armenia: a cohort approach
H K Armenian1, A Melkonian, E K Noji
1Department of Epidemiology, School of Hygiene and Public Health, Johns Hopkins University, Baltimore, MD 21205, USA.
Background:
This is the first population-based study of earthquake injuries and deaths that uses a cohort approach to identify factors of high risk. As part of a special project that collected data about the population in the aftermath of the earthquake that hit Northern Armenia on 7 December 1988, employees of the Ministry of Health working in the earthquake zone on 7 December 1988, and their families, were studied as a cohort to assess the short and long term impact of the disaster. The current analysis assesses short term outcomes of injuries and deaths as a direct result of the earthquake.
Methods:
From an unduplicated list of 9017 employees, it was possible to contact and interview 7016 employees or their families over a period extending from April 1990 to December 1992. The current analysis presents the determinants of 831 deaths and 1454 injuries that resulted directly from the earthquake in our study population of 32,743 people (employees and their families).
Results:
Geographical location, being inside a building during the earthquake, height of the building, and location within the upper floors of the building were risk factors for injury and death in the univariate analyses. However, multivariate analyses, using different models, revealed that being in the Spitak region (odds ratio [OR] = 80.9, 95% confidence interval [CI]: 55.5-118.1) and in the city of Gumri (OR = 30.7, 95% CI: 21.4-44.2) and inside a building at the moment of the earthquake (OR = 10.1, 95% CI: 6.5-15.9) were the strongest predictors for death. Although of smaller magnitude, the same factors had significant OR for injuries. Building height was more important as a factor in predicting death than the location of the individual on various floors of the building except for being on the ground floor of the building which was protective.
Conclusions:
Considering that most of the high rise buildings destroyed in this earthquake were built using standard techniques, the most effective preventive effort for this disaster would have been appropriate structural approaches prior to the earthquake.
Related Concept Videos
Prevalence and Incidence
Prevalence indicates the proportion of individuals in a population who have a specific disease or health condition at a...
Study Designs in Epidemiology
Observational studies are those where the researcher does not intervene but rather observes natural variations. They include cross-sectional, cohort, and case-control studies.
Bias in Epidemiological Studies
Statistical Methods for Analyzing Epidemiological Data
Actuarial Approach
Consider the example of a high-risk surgical procedure with significant early-stage mortality. A two-year clinical study is conducted,...
Hazard Ratio
For example, in a clinical trial evaluating a...