Abstract
Introduction
When treating a patient admitted with hip fracture it is useful to have a simple scoring system to predict outcomes, based on admission clerking and routine investigations. The Nottingham Hip Fracture Score (NHFS) is one such measure. Its use has been described by Wiles et al (Br J. Anaes. Jan. 2001) for risk stratification in predicting 30-day and 1-year mortality.
Objective
Our aim was to use the hip fracture database at Peterborough City Hospital, UK to conduct an independent validation study of the NHFS stratification system.
Method
After excluding patients with incomplete data there were 5062 records suitable for inclusion in the study. The NHFS for each patient was calculated using spreadsheet software and then stratified into low (score ≤4) and high risk (score ≥5) patients. The difference in mortality between the two groups was calculated by means of Kaplan-Meier analysis, including log rank comparison and hazard ratio.
Results
The difference between the two groups was highly statistically significant, both at 30 days and 1 year:• Those who died within 30 days: Median survival for low and high risk was 13.0 and 13.5 days respectively, a ratio of 0.963 (95% CI 0.18–1.75); Log rank comparison gave a chi-square of 134.9, p< 0.0001; Hazard ratio: 1.168 (95% CI 0.91–1.57).• Those who died within 1 year: Median survival for low and high risk was 104 and 78 days respectively, a ratio of 1.333 (95% CI 0.45–2.22); Log rank comparison gave a chi-square of 233.0, p< 0.0001; Hazard ratio was 0.406 (95% CI 0.39–0.48).
Discussion
Using the Peterborough hip fracture database, we have demonstrated independently the validity of risk stratification based on the NHFS as a predictor of mortality outcomes. This supports its further use in the management of patients with hip fracture to identify those at significant risk.
