A simple clinical risk nomogram to predict mortality-associated geriatric complications in severely injured geriatric patients.

ABSTRACT

BACKGROUND: This study aimed to (1) identify inpatient complications associated with the greatest differential mortality risk between young and old patients with traumatic injury and (2) identify older patients at elevated risk for mortality-associated complications. METHODS: Secondary analysis of more than 280,000 patients hospitalized for traumatic injury in 2001 to 2005 collected by the National Trauma Data Bank was performed. Predictor variables include 21 hospital complications. We used each complication to predict odds of hospital mortality, stratified by old (65+ years) versus young (18-64 years) age, controlling for age, sex, and preexisting condition count. We defined mortality-associated geriatric complications (MGCs) as complications associated with more than two times risk of mortality in older patients compared with younger patients. We then used age, comorbidity, and sex to predict development of MGCs or death. RESULTS: We defined seven infectious and six noninfectious complications as MGCs (adjusted relative risk reduction for death associated with old age [aRRR] with 95% confidence interval [CI]) abscess (aRRR, 4.1; 95% CI, 3.6-4.5), wound infection (aRRR, 3.5; 95% CI, 3.2-3.9), empyema (aRRR, 3.4; 95% CI, 3.1, 3.8), urinary tract infection (aRRR, 3.3; 95% 3.0-3.6), pneumonia (aRRR, 3.1; 95% CI, 2.8-3.5), bacteremia (aRRR, 2.8; 95% CI, 2.6-3.0), aspiration pneumonia (aRRR, 2.6; 95% CI, 2.2-3.0), reduction/fixation failure (aRRR, 3.6; 95% CI, 3.3-3.9), pressure ulcer (aRRR, 3.3; 95% CI, 3.1-3.6), deep venous thrombosis (aRRR, 3.2; 95% CI, 2.9-3.6), pneumothorax (aRRR, 3.1; 95% CI, 2.5-3.7), and compartment syndrome (aRRR, 2.2; 95% CI, 1.5-2.9). We developed a graphical nomogram based on sex, age, and number of preexisting conditions to predict risk of MGCs (c statistic, 0.74). CONCLUSION: Older patients at risk for MGC development should be considered for targeted interventions to improve quality of care. LEVEL OF EVIDENCE Prognostic/epidemiologic study, level III.