Can computers help in assessing children's postoperative pain? Initial validation of a computer-assisted interview

Abstract

Background and objective: Computers offer the potential for the assessment of children who have difficulties in communication and cannot describe pain using conventional approaches. Such approaches must be reliable and valid. As a preliminary step towards this goal, the validity and reliability of a computer-assisted pain assessment for children (MacInterview) was assessed using children with no known disabilities who had undergone surgery.

Methods: MacInterview uses body outlines and a range of different pain representations with scaling for size and intensity, and associated emotion. Following piloting with non-clinic children, the experience of acute postoperative pain was assessed for 30 children undergoing adenotonsillectomy or tonsillectomy the day following surgery using the MacInterview and three existing standardized assessment measures. Each child self-reported their current experience of postoperative pain on two occasions 30 min apart, and retrospectively on pain the previous evening in the immediate postoperative period.

Results: Analyses indicated good performance of MacInterview, showing positive correlations between 0.65 and 0.88 with standardized pain-intensity measures, and test-retest reliability was 0.9. Face validity was high, and children enjoyed the procedure.

Conclusions: The procedure shows promise and is likely to merit further development for children showing difficulties in communication.