OTA 1997 Posters - Hip Fractures

The Effects of Nutritional Status on Outcome after Hip Fracture: A Prospective Study

Kenneth A. Egol, MD, Kenneth J. Koval, MD, Ed Su, BS, Gina Aharonoff, MPH, Joseph D. Zuckerman, MD

New York, New York, USA

Purpose: Studies have demonstrated that malnutrition is common in the elderly and is associated with a poorer outcome after surgery. Most of these studies, however, have used costly tests to determine patient nutritional status (e.g., skin antigen testing, nitrogen balance, prealbumin, or transferrin levels) and have not controlled for potential confounding variables. This prospective study determined the effects of nutrition on patient outcome after hip fracture in a community-dwelling elderly population using cost-effective laboratory and clinical parameters that are part of the perioperative screening process.

Methods: From January 1988 to June 1995, 702 community-dwelling patients over the age of 65 who had an operatively treated femoral neck or intertrochanteric fracture of nonpathologic origin were prospectively followed. All patients followed a similar postoperative protocol and were contacted at 3, 6, and 12 months after surgery. The outcomes studied were hospital length of stay, mortality (both in-house and 1 year), development of postoperative complications, and percent functional recovery at 3, 6, and 12 months as determined by an 11-item functional rating scale. The predictors were preoperative albumin level (< or 3.5), total lymphocyte count (< or 1 500 cells/ml), nutritional status score ( or >3), and nutritional risk score (< or 4). The nutritional status score was assigned by a nutritionist according to ICD-9 codes based on the patient's albumin level and variance from ideal and usual body weight; the nutritional risk score was determined by the type and number of associated comorbidities that could contribute to the development of malnutrition (e.g., diabetes, infection, inflammatory disease). The following were held as confounding variables: patient age, sex, fracture type, number of preinjury medical comorbidities, ASA rating of operative risk, prefracture function, and type of surgery. Data were analyzed by multiple logistic regression; a p-value < .05 was considered significant.

Results: Malnutrition at hospital admission was found in 18.4% of patients by albumin level, 42.6% by total lymphocyte count, and 22.9% by nutritional status; 35.6% of patients were at nutritional risk. Ninety-two patients (13.1%) developed an inhospital medical complication, and 20 patients (2.8%) died during hospitalization. The hospital length of stay averaged 22.9 days. Eighty-one patients (11.5%) died during the first year following hip fracture and an additional 32 patients refused to participate or were lost to follow-up. Therefore, 589 patients remained for whom complete follow-up information was available. Mean percent functional recovery at 3, 6 and 12 months was 69.8%, 79.0%, and 79.4%, respectively. In multivariate analyses, a lymphocyte count of < 1500 cells/ml (p < .01) and a nutritional risk score of 4 (p < .05) were predictors of increased one-year mortality; a nutritional risk score of 4 (p < .01) was also a predictor of increased hospital length of stay. None of the markers of nutritional depletion, however, had an effect on inhospital mortality, the development of an inhospital complication, or percent functional recovery at 3, 6, or 12 month follow-up.

Discussion and Conclusion: Patient malnutrition was found to be a predictor of poorer outcome after hip fracture in a community-dwelling elderly population. A preoperative lymphocyte count of < 1500 cells/ml and a nutritional risk score of 4 were found to be predictive of increased one-year mortality; an increased nutritional risk score was also predictive of increased hospital length of stay. The nutritional parameters studied are cost-effective: albumin level and total lymphocyte count are part of the preoperative screening process, and nutritional status and nutritional risk can be calculated after a simple patient interview. Although other laboratory parameters (e.g., prealbumin) transferring may be more sensitive indicators of nutritional status, their cost limits routine use.