Risk factors associated with pressure ulcer development at a major western Australian teaching hospital from 1998 to 2000: secondary data analysis.

Objective: The purpose of this study was to identify risk factors associated with the presence of pressure ulcer development in adult patients at an acute care teaching hospital.

Design: The database was established by combining the results from 3 annual cross-sectional pressure ulcer prevalence studies conducted between 1998 and 2000.

Setting And Sample: The sample consisted of all medical and surgical patients who were inpatients on the day of the pressure ulcer surveys. The sample includes a total of 1394 patients.

Methods: This study undertook secondary data analysis with use of logistic regression and descriptive statistics.

Results: Overall pressure ulcer prevalence was calculated at 15.9%, with this figure decreasing slightly to 12.7% when pressure ulcers present on admission were excluded. Initial analysis identified several variables as significant risk factors for pressure ulcer development (age, Braden Scale risk category, and admission type and specialty). However, when entered into a final multivariate model, only 2 factors, age and Braden Scale risk category, were found to be significant. The odds ratio matrix revealed a consistent trend toward increasing odds ratios with increasing Braden Scale risk category within each age group. The magnitude of this trend was more pronounced in the younger age groups, highlighting the importance of undertaking Braden Scale assessments of younger patients, some of whom may be at greater risk of having a pressure ulcer develop than would otherwise be expected.

Conclusion: Pressure ulcer prevalence in this acute care setting is high when compared with other Australian data. However, comparisons of pressure ulcer prevalence across studies are difficult to interpret because of different reporting methods. Although predictor variables for pressure ulcers have been identified in a number of studies, longitudinal studies are needed to identify the cause-and-effect relationships for potential predictor variables. In addition, more sophisticated statistical analyses such as the use of the odds ratio matrix may help guide further research into interaction effects between predictor variables and how these might affect the psychometric properties of risk assessment tools such as the Braden Scale. Pressure ulcer interventions should be targeted at the sacrum and heel, because these appear to be the most frequently observed locations for pressure ulcers.