Development of a Predictive Model for Postoperative Pulmonary Complications after Cholecystectomy

Jo Ann Brooks-Brunn
DOI: https://doi.org/10.1177/105477389200100207
IF: 1.724
1992-05-01
Clinical Nursing Research
Abstract:The purpose of this study was to develop a model to predict the occurrence of a postoperative pulmonary complication (PPC) following cholecystectomy. Seventeen potential risk factors were extracted from the literature by identifying and ranking those most frequently referenced. The study included only those risk factors available to the nurse in the preoperative, intraoperative, and immediate postoperative setting. Three institutions were used for data collection, and data were collected by a retrospective chart review of 300 randomly chosen subjects from a population of 720. Of the 300 subjects, 37 were omitted due to exclusion criteria. A PPC was present in 54 of the remaining 263 subjects (20.5%). Of the original 17 risk factors, 10 were included in model development. The 54 subjects with a PPC and 54 subjects without a PPC (randomly chosen from the remaining 209) were used to determine which combination of risk factors best predicted subject classification (PPC or no PPC). The direct entry discriminant function that provided the highest percentage of correct classification (PPC, no PPC) consisted of five variables: sex, age, smoking history, duration of anesthesia, and nasogastric tube. The resulting equation correctly classified 75% of the cases.
nursing
What problem does this paper attempt to address?