RESUMO
The accurate prediction of pressure ulcer (PU) development among hospitalized elderly patients is a complex endeavour. A prospective, longitudinal, cohort study of 330 patients over age 65 in 2 Canadian tertiary-care teaching hospitals and 2 long-term-care facilities examined the association between risk-assessment scores, prevention strategies, and PU incidence. The overall PU incidence rate was 9.7%, with half of the subjects who developed a PU doing so in the first week of hospitalization. The incidence rate for "at risk" patients (10.1%) was similar to the rate for "not at risk" patients (9.3%). The number of prevention strategies used was related to risk-assessment scores and to PU development. Paradoxically, the incidence rate increased with the number of prevention strategies employed. The total risk-assessment score that appeared to have the best balance of sensitivity (69%) and specificity (55%) was 19. Four of the 6 risk-assessment subscales were associated with PU development. Logistic regression modelling confirmed the univariate results that the number of prevention strategies used was the best single predictor of PU development. The data confirm that predicting PU development for individual patients is difficult at best. Results suggest that use of a risk-assessment scale alone is not sufficient to accurately predict PU development. The clinical judgement and experience of nurses are required in providing supplementary information to standard measurement instruments.