Predictive validity and reliability of the Braden scale for risk assessment of pressure ulcers in an intensive care unit.

OBJECTIVE: Contribution to validation of the Braden scale in patients admitted to the ICU, based on an analysis of its reliability and predictive validity.

DESIGN: An analytical, observational, longitudinal prospective study was carried out.

SETTING: Intensive Care Unit, Hospital Virgen del Rocío, Seville (Spain).

PATIENTS: Patients aged 18years or older and admitted for over 24hours to the ICU were included. Patients with pressure ulcers upon admission were excluded. A total of 335 patients were enrolled in two study periods of one month each.

INTERVENTIONS: None.

VARIABLES OF INTEREST: The presence of gradei-iv pressure ulcers was regarded as the main or dependent variable. Three categories were considered (demographic, clinical and prognostic) for the remaining variables.

RESULTS: The incidence of patients who developed pressure ulcers was 8.1%. The proportion of gradei andii pressure ulcer was 40.6% and 59.4% respectively, highlighting the sacrum as the most frequently affected location. Cronbach's alpha coefficient in the assessments considered indicated good to moderate reliability. In the three evaluations made, a cutoff point of 12 was presented as optimal in the assessment of the first and second days of admission. In relation to the assessment of the day with minimum score, the optimal cutoff point was 10.

CONCLUSIONS: The Braden scale shows insufficient predictive validity and poor precision for cutoff points of both 18 and 16, which are those accepted in the different clinical scenarios.