Table 2.
Univariate and multivariate Cox regression analyses
| Variable | Univariate analysis | Multivariate model | ||
|---|---|---|---|---|
| HR value (95% CI) | p Value | HR value (95% CI)—final model | p Value | |
| Age | 0.000* | 0.001* | ||
| 60–69 years | Reference | Reference | ||
| 70–79 years | 3.25 (0.97–10.91) | 0.056 | 3.07 (0.90–10.51) | 0.073 |
| 80 + years | 7.03 (2.21–22.41) | 0.001 | 6.48 (1.97–21.36) | 0.002 |
| Male gender | 2.20 (1.39–3.48) | 0.001 | 2.04 (1.26–3.30) | 0.004 |
| Level of education | 0.927* | |||
| Illiterate–4 years | 0.91 (0.43–1.96) | 0.818 | ||
| 5–8 years | 0.99 (0.45–2.19) | 0.989 | ||
| 9+ years | Reference | |||
| Site of fall (in the home vs outside) | 1.53 (0.93–2.52) | 0.095 | ||
| Marital status (married vs not married) | 0.89 (0.51–1.57) | 0.689 | ||
| Ethnicity | 0.204* | |||
| White | 0.90 (0.54–1.51) | 0.693 | ||
| Not black/not white | Reference | |||
| Black | 0.26 (0.06–1.14) | 0.075 | ||
| Weight | 0.607* | |||
| Normal (BMI = 22–27) | Reference | |||
| Underweight (BMI < 22.0) | 0.83 (0.52–1.31) | 0.417 | ||
| Overweight/obese (BMI > 27) | 1.14 (0.53–2.44) | 0.745 | ||
| Femur vs intertrochanteric fracture | 1.25 (0.80–1.96) | 0.325 | ||
| Lived alone before fracture vs not | 0.88 (0.46–1.70) | 0.704 | ||
| BOAS score | 0.001* | 0.000* | ||
| No limitations | Reference | |||
| 1–2 limitations | 1.62 (0.92–2.9) | 0.098 | 1.75 (1.00–3.09) | 0.052 |
| 3–4 limitations | 1.85 (1.00–3.40) | 0.048 | 1.18 (0.63–2.23) | 0.605 |
| 5+ limitations | 6.11 (3.21–11.65) | 0.0001 | 4.35 (2.25–8.40) | 0.000 |
| Chronic diseases at admission | ||||
| Diabetes mellitus | 1.85 (1.07–3.19) | 0.028 | ||
| Hypertension | 1.41 (0.92–2.17) | 0.118 | ||
| Chronic obstructive pulmonary disease | 1.04 (0.55–1.97) | 0.909 | ||
| Stroke | 2.20 (1.17–4.12) | 0.014 | ||
| Cardiovascular disease | 1.39 (0.78–2.47) | 0.260 | ||
| Depression | 1.69 (0.61–4.66) | 0.311 | ||
| Cataracts | 0.65 (0.37–1.15) | 0.137 | ||
| Glaucoma | 0.68 (0.17–2.75) | 0.584 | ||
| Hearing impairment | 0.85 (0.34–2.10) | 0.723 | ||
| Postural instability | 1.41 (0.87–2.29) | 0.161 | ||
| Urinary incontinence | 1.18 (0.60–2.30) | 0.636 | ||
| Medication use at admission | ||||
| Anticoagulants | 1.11 (0.60–2.06) | 0.733 | ||
| Inotropics | 1.73 (0.75–4.00) | 0.201 | ||
| AntiParkinson drugs | 1.83 (0.72–4.65) | 0.206 | ||
| Sedatives, antipsychotics, and antianxiolytics | 1.06 (0.62–1.80) | 0.841 | ||
| Oral glucose-lowering agents and insulin | 1.09 (0.43–2.76) | 0.852 | ||
| Antidepressants | 1.21 (0.38–3.90) | 0.744 | ||
| Antiinflammatory drugs | Not possible | |||
| Bronchodilators | 0.64 (0.89–4.67) | 0.664 | ||
| Antiarrhythmics | 1.01 (0.47–2.20) | 0.974 | ||
| Antihypertensives | 1.32 (0.86–2.03) | 0.208 | ||
| Hormones | 1.07 (0.26–4.37) | 0.929 | ||
| Hip arthroplasty vs osteosynthesis | 0.99 (0.98–1.01) | 0.566 | ||
| Use of traction | 0.83 (0.53–1.31) | 0.430 | ||
| Surgery risk | 0.001* | 0.018* | ||
| Low (ASA scores 1 and 2) | Reference | Reference | ||
| High (ASA scores 3 and 4) | 2.58 (1.56–4.27) | 0.000 | 2.15 (1.25–3.69) | 0.006 |
| No information | 1.34 (0.72–2.49) | 0.359 | 1.33 (0.70–2.53) | 0.378 |
| General anesthesia vs other | 0.54 (0.20–1.47) | 0.229 | ||
| Delirium during hospital stay | 1.40 (0.90–2.17) | 0.142 | ||
| Complications during/after surgery | 1.54 (0.91–2.59) | 0.105 | ||
| Physical therapy received | 0.25 (0.15–0.44) | 0.000 | 0.32 (0.18–0.57) | 0.000 |
| Medication use after surgery | ||||
| Anticoagulants | 0.90 (0.58–1.40) | 0.637 | ||
| Antiinflammatory drugs | 1.47 (0.46–4.75) | 0.520 | ||
| Analgesics | 0.62 (0.37–1.04) | 0.710 | ||
| Antihypertensives | 0.89 (0.52–1.51) | 0.656 | ||
| Antibiotics | 0.51 (0.33–0.78) | 0.002 | 0.51 (0.32–0.80) | 0.003 |
| Total length of stay | 0.627* | |||
| ≤ 7 days | Reference | |||
| 8–30 days | 1.18 (0.63–2.19) | 0.611 | ||
| > 30 days | 1.43 (0.68–3.00) | 0.345 | ||
| Time between fracture and admission | 0.605* | |||
| 0 days | Reference | |||
| 1–2 days | 0.77 (0.43–1.38) | 0.383 | ||
| 3+ days | 1.07 (0.64–1.81) | 0.788 | ||
Variables associated with mortality in univariate analysis (p < 0.10) were selected as candidate determinants for the multivariate analysis; a final model (age and gender adjusted) was constructed using manual backward selection method to delete nonsignificant variables from the final model using the −2 log likelihood test; *p value for the group dummy variables; HR = hazard ratio; CI = confidence interval; BMI = body mass index; BOAS = Brazil Old Age Schedule; ASA = American Society of Anesthesiologists.