Abstract
Pressure ulcers (a type of skin failure) have served as an indicator of care quality. The purpose of this study was to utilize data-mining techniques as a means of identifying risk factors related to different stages of pressure ulcers to demonstrate how this means of analysis might be used as a vehicle to guide improved care quality. Data were obtained from a Web-based incident reporting system at a regional hospital in Taiwan. A total of 4301 cases dating from March 2005 to May 2009 were collected. For data-cleaning purposes, data within 3 SDs were kept for further analysis. Data-mining techniques were applied to identify the predictors, and a logistic regression analysis was used for result comparison purposes. The results revealed that sacral ulcer was the most prevalent, and most ulcers were in stage I, followed by stages II to IV. Five predictors were identified including hemoglobin, weight, sex, height, and use of repositioning sheet. The study concluded that nurses could use data-mining technique to identify predictors to assist in guiding ulcer interventions such as those based on a patient's demographic profile and application of a repositioning sheet to prevent ulcer occurrence to minimize harm.