[Intraobserver reliability and internal consistency of the Behavioral Pain Scale in mechanically-ventilated patients]

Abstract

Background: The Behavioral Pain Scale (BPS) is a tool of pain assessment that often gives contradictory results when used by different raters.

Objective: To assess internal consistency and interrater reliability of BPS scale in the pain assessment performed by intensives care nurses.

Methods: A prospective observational study in 34 mechanically-ventilated patients, carried out in an Intensive Care Unit from April to June 2012. Variables analyzed included demographic characteristics, diagnosis of referral, clinical status, pain and sedation level. Pain was assessed by two nurses independently at rest (T1) and during a mobilization procedure (T2) using the BPS scale. Internal consistency was calculated by Cronbach's alpha, and intraobserver reliability was determined with the intraclass correlation coefficient (ICC), with a confidence interval (CI) of 95%. This study was approved by the Ethical Committee for Clinical Research.

Results: One-hundred and twenty-eight pain assessments were performed. The Cronbach's alpha of total BPS score at rest was 0.66 (95%CI: 0.33 to 0.83) and during mobilization of 0.73 (95%CI: 0.47 to 0.87). The CCI of total BPS score was 0.50 (95%CI: 0.19 to 0.71) at rest and 0.58 (95%CI: 0.31 to 0.77) during mobilization.

Conclusions: The level of internal consistency of the scale is appropriate and moderate interrater agreement. For the BPS useful in clinical practice, it is imperative that nurses have prior experience with a regulated use of this tool.

Keywords: Behavioral Pain Scale; Enfermería; Fiabilidad; Intensive critical unit; Nursing; Reliability; Servicio de medicina intensiva.