Lessons learned from the catastrophic Marmara earthquake: factors influencing the final outcome of renal victims

Abstract

Background: During catastrophic earthquakes, crush syndrome is the second most frequent cause of death after the direct impact of trauma. The Marmara earthquake, which struck Northwestern Turkey in August 1999, was characterized by 639 crush syndrome victims with acute renal problems. The factors influencing their final outcome have been the subject of this study. Patients methods: Within the first week of the disaster questionnaires asking about 63 clinical and laboratory variables were sent to 35 reference hospitals that treated the victims. Information obtained by means of these questionnaires, including the factors with a potential influence on outcome, was submitted to analysis. Results: Overall mortality rate was 15.2%. In univariate analysis, nonsurvivors were older (p = 0.048); the highest mortality rates were observed among the victims coming from the closest cities to the reference hospitals. Admission within the first 3 days of the disaster (p = 0.016), with oliguria (p = 0.042), lower figures for blood pressure (p < 0.001), platelets (p = 0.004) and serum albumin (p = 0.005) were associated with mortality. Also, higher body temperature (p = 0.013) and serum potassium (p < 0.001) as well as suffering from thoracic or abdominal traumas, extremity amputations and medical complications other than renal failure (for all 4: p < 0.0001) in addition to need of dialysis support (p = 0.015) and mechanical ventilation (p < 0.0001) indicated higher risk of death. In the multivariate analysis, age (p = 0.030, OR =1.02), presence of disseminated intravascular coagulation (p = 0.001, OR = 4.49), abdominal trauma (p 0.012, OR = 4.05) and amputations (p = 0.010, OR = 2.81) were predictors of mortality. Dialyzed patients were characterized by higher mortality rates than nondialyzed victims (17.2% versus 9.3%, p 0.015). Conclusion: Outcome of the renal victims of catastrophic earthquakes is influenced by the type of trauma, comorbid events and complications observed during the clinical course as well as epidemiological features such as age, distance to reference hospitals and time lapse between disaster and admission to reference hospitals.