Dear Editor,

The interesting article by Sahebi et al. ^{Reference Sahebi, Nejati-Zarnaqi and Vasei1} compiled evidence on deep vein thrombosis (DVT) rates among earthquake survivors. They concluded that DVT rates are higher among earthquake victims than those of other disaster types. Their study offers valuable insights into this critical subject, but there is significant heterogeneity in both the overall analysis (I ² = 97.9%; P < 0.001) and subgroups, including the general population (I ² = 98.0%; P < 0.001) and patient survivors (I ² = 77.7%; P = 0.001). We have identified 2 factors that may explain at least some of this heterogeneity.

First, the study included data from 4 countries (Japan, Pakistan, China, and Nepal). Genetic variation in factors such as PROS1, PROC, SERPINC1, factor V Leiden, prothrombin G20210A, and protein S mutation K196E, ^{Reference Miyata, Sato and Ishikawa2} and differences in earthquake frequency, demographics, emergency preparedness, and health-care infrastructure may exist among these nations, contributing to variation in DVT rates, even within each country. Indeed, DVT rates ranged from 0.94% to 34.2% in Japan, 1.5% to 4.8% in Pakistan, 0.2% in China, and 6.0% in Nepal. It is important to note that only studies involving Asian subjects were included, and DVT rates of other ethnicities remain unknown.

Second, the study merged reports that covered different screening locations. For instance, Ueda et al. ^{Reference Ueda, Hanzawa and Shibata3} focused on 190 admitted patients diagnosed with pulmonary thromboembolism 1 mo after an earthquake (mean age, 70.1 y; female, 137 [73.3%]) and reported DVT rates of 34.2%. Rathore et al. ^{Reference Rathore, Hanif and New4} investigated 187 earthquake survivors with spinal cord injuries (mean age, 28.3 y; female, 107 [57.2%]) and found DVT rates of 4.8%. Additionally, factors such as the length of hospital stay and implementation of prophylactic measures were not standardized across these studies. ^{Reference Ueda, Hanzawa and Shibata3,Reference Rathore, Hanif and New4} Unlike these studies, Shibata et al. ^{Reference Shibata, Hanzawa and Ueda5} targeted 269 evacuees in shelters 1 mo after an earthquake (mean age, 70.6 y; female, 174 [64.7%]) and observed DVT rates of 24.2%.

Despite problems inherent to systematic reviews and meta-analyses, the study of Sahebi et al. ^{Reference Sahebi, Nejati-Zarnaqi and Vasei1} is valuable because it illuminates DVT risks among earthquake survivors. Future studies need to perform more detailed subgroup analyses, ie, meta-analysis and meta-regression, conducted by country or screening location. Factors such as hospitalization with spinal cord injury, hospitalization with pulmonary thromboembolism, earthquake magnitude, injuries, evacuation to temporary housing in houses or shelters, or sheltering in cars after an earthquake should be examined to further enhance our understanding of the complex relationship between earthquakes and DVT.