Session X - Geriatrics
Sun., 10/10/04 Geriatrics, Paper #58, 8:12 am
Hemiarthroplasty for Femoral Neck Fracture in the Elderly: Case Volume-Related Outcomes
Purpose: Hemiarthroplasty is the treatment of choice for displaced femoral neck fractures (OTA class 31-B) in the elderly less-active patient. Provider caseload volume has been shown to have a significant effect on postoperative mortality, complications, and length of stay (LOS) for a number of surgical procedures. The objectives of this study were to (1) determine the rates of in-hospital mortality, pulmonary embolism (PE), infection, and prolonged LOS in elderly patients with femoral neck fractures treated by hemiarthroplasty, (2) elucidate the patient characteristics that predict these occurrences, and (3) investigate the influence of surgeon and hospital volumes on these outcomes.
Methods: We obtained the Nationwide Inpatient Sample (NIS) hospital discharge database for the years 1988 through 2000. The NIS is a 20% random sample of hospitals in the United States stratified by geographic region, teaching status, hospital size, and other characteristics. Femoral neck fracture was coded according to the international classification of diseases, ninth revision, clinical modification (ICD-9-CM), and included ICD-9-CM 820.0-820.9. Hemiarthroplasty was identified by ICD-9-CM 81.52. Pertinent demographic and postoperative data, as well as unique hospital and surgeon identifiers, were obtained from the database. Patient age was defined as 65 to 84 years and 85 years. Race was categorized as white or nonwhite. Cardiovascular, pulmonary, hepatic, renal, and malignant comorbid conditions were identified. The lowest, low-intermediate, high-intermediate, and highest surgeon volume quartiles (SVQ) and hospital volume quartiles (HVQ) were determined. Primary outcome variables included in-hospital mortality and prolonged LOS. Secondary variables were PE, wound infection, urinary tract infection (UTI), and pneumonia (ICD-9-CM 41.51, 41.510, 41.511, 41.519, 599.0, 998.59, and 997.3). Prolonged LOS was defined as 10 postoperative days, which corresponded to the 75th percentile of all admissions. Multivariate analysis was used to test the association of mortality and prolonged LOS with surgeon and hospital volume, complications, and demographic variables. Univariate analysis was used for all other comparisons.
Results: Of the 173,508 patients identified, most were white (66.8%), female (58.9%), between the ages of 65 and 84 (61.8%), and without comorbid disease (73.1%). The in-hospital mortality rate was 3.1%. The lowest SVQ (odds ratio [OR] 1.18; 95% confidence interval [CI] 1.03 to 1.34; P<0.05) was associated with a significantly increased mortality when compared with the highest SVQ. Hospital volume did not predict mortality. The median LOS was 7 days. Prolonged LOS (>10 days) was reported for 21.3% of patients. The lowest (OR 1.37; 95% CI, 1.29 to 1.45; P < 0.001), low intermediate (OR 1.15; 95% CI, 1.09 to 1.22; P <0.001) and high intermediate (OR 1.11; 95% CI, 1.05 to 1.17; P < 0.001) SVQ were each associated with increased risk of prolonged LOS compared with the highest SVQ. The lowest (OR 1.56; 95% CI 1.46 to 1.66; P <0.001), low intermediate (OR 1.44; 95% CI, 1.37 to 1.52; P <0.001), and high intermediate (OR 1.22; 95% CI, 1.15 to 1.28; P <0.001) HVQ were each associated with increased risk of prolonged LOS compared with the highest HVQ. Other independent predictors of both increased mortality and prolonged LOS were male sex, white race, age 85, PE, wound infection, pneumonia, and increasing numbers of comorbidities (P <0.05). Prolonged LOS was associated with UTI, but it did not increase mortality (P < 0.001). The incidences of PE, wound infection, UTI, and pneumonia were 1.5%, 0.1%, 15.8%, and 2.8%, respectively. Increased hospital volume correlated with decreased rates of PE, UTI, and pneumonia (P < 0.001). Surgeon volume did not predict these complications.
Conclusion/Significance: The results of this study demonstrate that hospitals with low caseload volumes are associated with increased patient risk for prolonged LOS, PE, UTI, and pneumonia after hemiarthroplasty for femoral neck fracture (OTA class 31-B). Surgeons with low caseload volumes are associated with increased risk for mortality and prolonged LOS. The presence of effective "institutional structure and process variables" at high-volume centers has been suggested as the reason for improved outcomes. These variables have not been well defined. In our study, such variables may include operating room and nursing staff familiarity with the procedure and its potential complications, standardized patient care protocols, and competent ancillary departments. The reason for increased mortality after hemiarthroplasty performed by low-volume surgeons is not obvious. It makes sense, however, that a surgeon with less experience in performing this operation may indirectly cause increased physiologic stress (e.g. increased blood loss, longer anesthesia time) to the patient as a result of the potential for greater intraoperative difficulties. Future studies that determine specifically which structure and process variables are most important may be able to provide guidance for quality improvement at low-volume centers.