Abstract / Summary
Background and Objectives: Hip fracture in older adults is associated with high morbidity, and length of hospital stay is the simplest summary of the acute admission. We aimed to identify the clinical factors associated with prolonged hospital stay in older adults admitted with a hip fracture, and to test whether a weighted comorbidity index predicts stay better than a simple count. Materials and Methods: A retrospective cohort of 624 consecutive patients aged 65 years and older with a trochanteric or femoral neck fracture was studied. Length of stay (LOS) was strongly right-skewed (skewness 5.02) and was log-transformed; comparisons used Mann–Whitney and Kruskal–Wallis tests and Spearman correlations, with multivariable models fitted on 487 operated patients with complete data. Postoperative stay was analyzed as a secondary outcome because the preoperative interval forms part of the total stay. Results: Median stay was 10 days (IQR 8–13) and median wait for surgery was 4 days (IQR 2–6), with only 25.2% operated on within 48 h. In-hospital mortality was 5.3%. The number of comorbidities was the only patient characteristic consistently associated with stay (rho = 0.119, p = 0.003; 1.4% per comorbidity, p < 0.001). The Charlson Comorbidity Index (CCI) showed no association in the unweighted, age-adjusted or severity-graded form. Age, sex and admission hemoglobin effects did not survive non-parametric testing. Conclusions: A simple count of active comorbidities was associated with LOS whereas the CCI was not, and the model containing the count fitted modestly better (adjusted R2 0.385 vs. 0.365; index term non-significant). This hypothesis-generating finding suggests that the comorbidity count, which is available on admission, may serve as a pragmatic screening tool for selecting patients for comprehensive geriatric assessment; its use in place of the CCI would require validation in independent cohorts. The preoperative wait remains the dominant modifiable target, although its effect on total stay is largely structural.