Survival analysis of the factors affecting in mortality in injured patients requiring dialysis due to acute renal failure during the Marmara earthquake: Survivors vs non-survivors

Abstract

Background: We reviewed medical records of dialyzed patients admitted to our hospital after the Marmara earthquake and evaluated the factors affecting mortality in survivors and non-survivors according to the survival times. Patients and methods: Crush syndrome ( CS) was diagnosed in 110 patients. Dialysis treatment was initiated in 60 patients; 21 of all died. The patients were divided into 2 groups which consisted of 39 survivors ( Group A, 25 male, 14 female, mean age: 31 +/- 2.2 years) and 21 nonsurvivors ( Group B, 9 male, 12 female, mean age: 27 +/- 3.0 years). Victims treated by any form of renal replacement therapy, including daily or intermittant hemodialysis and/or continuous venovenous hemodiafiltration. Clinical and laboratory findings were recorded regularly. Statistical analysis was performed with Kaplan-Meier method, log rank test and Cox regression analysis for the survival functions. Results: APACHE II scores were 13.5 +/- 0.5 for Group A and 13 +/- 0.9 for Group B. Dialysis support was started to patients in Group A in a mean period of 2.8 +/- 0.2 days and in Group B in a mean period of 3.7 +/- 0.6 days after the earthquake ( p > 0.05). The most frequent site of trauma was lower extremity (61.5%) and upper + lower extremities (23%) in Group A, and lower extremity (38.1%) and trunk + lower extremity (23.8%) in Group B. The frequencies of abdominal trauma, pelvic fracture and thoracic trauma in Group B were 23.8%, 19% and 14.2%, respectively. Multiple trauma was more frequent in Group B than in Group A (42.8% vs 2.5%). The rates of fasciotomy, amputation and surgery were similar in both groups. The frequency of sepsis was higher in nonsurvivors. In our center, the overall mortality rate was 8%, mortality rate in CS was 21% and in dialyzed patients it was 35%. Mortality was mainly associated with sepsis. Survival periods (52.3 +/- 4.0 days) in Group A were longer than in Group B (17.3 +/- 2.5 days). With Cox regression analysis, the parameters such as systolic hypotension on admission, female gender, high serum peak creatine kinase (> 20,000 U/l) and multiple trauma including thoracic and abdominal regions, were factors increasing risk of mortality. Conclusion: As a result, sepsis, multiple trauma and severe crush injury were the main factors increasing mortality risk in dialyzed injuries after the earthquake.