A DECADE OF IN-HOSPITAL FRACTURE MORTALITY REVIEW: AN ANALYSIS OF 8700 PATIENTS

Abstract

Introduction: Fractures are common with well described morbidities. Few small studies have reported in-hospital mortality of patients with fractures. This study aims to describe the distribution of fractures, the incidence of in-hospital fracture mortality and its risk factors.

Materials and Methods: All patients with fractures related to the hip, pelvis, femur, tibia and spine over a 10 year period in a university hospital were identified using the ICD-9 codes upon discharge. Age, gender, race, length of hospital stay and in-hospital mortality data were collected from electronic records. Detailed analysis of the mortality data was performed with statistical analysis using SPSS software.

Results: 8709 fractures were identified, of which 30.3% were hip fractures [1422(54%) NOF and 1216(46%) IT fractures], 24.4% were spinal fractures [144 (7%) cervical; 558(26%) thoracic; 1038(49%) lumbar; 47(2.2%) sacral and 335(16%) unclassified spine fractures], 24.1% were tibial fractures [proximal and shaft], 14.4% were femoral fractures [1037(83%) shaft and 215(17%) supracondylar] and 6.9% were pelvic fractures [265(42%) non rami and 351(58%) rami fractures]. 25% of the fractures occurred between age 21–40 years and 65% fractures occurred in patients older than age 41 years. 53% of the patients were male. 69%, 12% and 11% of the fractures were found in Chinese, Malay and Indian respectively. 90% were closed fractures.

Overall in-hospital fracture mortality was 1.3% (117 fractures). Fracture specific in-hospital mortality was 2.8% (75 fractures) for hip fractures, 5% (13) for non rami pelvic fractures, 1% (20) for spinal fractures, 0.5% (6) for femoral fractures and near 0%(1) for tibial fractures. In-hospital mortality increased with age (0.4% mortality between 21–40 years, 0.6% between 41–60 years, 0.9% between 61–70 years, 1.7% between 71–80 and 4% between 81–90 years). Infection related causes of death were most common. The mean duration from hospital admission until death was 19 days (SD 20, range 1–34). More hip and spine fractures were seen in Chinese and more femoral and tibial fractures were seen in Malay and Indian patients, and this difference was statistically significant. Logistic regression analysis showed only increased patient age, male gender and fracture type as statistically significant risk factors for increased in-hospital mortality. Subgroup analysis showed a 30 and 20 times increased risk of in-hospital mortality for pelvic (p=0.001, 95% CI 4, 241) and hip (p=0.003, 95% CI 3, 159) fractures respectively.

Conclusion: The overall in-hospital fracture mortality was 1.3% (2.8 % for hip and 5% for pelvic fractures). Increased patient age, male gender, hip and pelvic fractures were found to be risk factors for increased mortality.