Validation of pain measurements in Thai children aged 1-5 years within 24 hours following operation

Abstract

To find out the appropriate observational pain measurement with high specificity, sensitivity, reliability, validity and assess their practicality for Thai children aged 1-5 years within 24 hours following operation. This descriptive study validated the target tests: OPS(O), TPPPS(T), FLACC(F) with respect to the accepted reference tool: CHEOPS(C). All tools were translated and tested for reliability and content validity. Pain related behaviors in 167 children were videotaped in 5 periods: preoperative, before and after analgesics in recovery room, before and after analgesics in wards. Four observers, blinded to sequence of analgesics, rated the videotaped behaviors. One observer rated pain behaviors by using CHEOPS whereas the other 3 observers used target tests for rating. Satisfaction in pain scales were also evaluated in 30 nurses. All tools had excellent interrater reliability (ICC of C = .9164, O = .9754, T = .9092, F = .9265). Concerning contenvalidation, 7 items of CHEOPS and 1 item of TPPPS had IC less than 0.5, most of them were related to scale in ranking. All pain scales showed accepted discriminant validity from significantly different scores between before and after analgesics (p < .001). Predictive validity with respect to CHEOPS were tested. In recovery room, using cutoff point 3, OPS and FLACC yielded high specificity and sensitivity (OPS: specificity 92.7% sensitivity 90.7%, FLACC: specificity 95.1% sensitivity 89.8%). In wards, no tool could yield specificity and sensitivity as high as 90%. Provided cutoff point was lowered to 2, OPS and FLACC yielded the reasonable performance (OPS: specificity 81.7% sensitivity 80%, FLACC: specificity 80.3% sensitivity 81.4%). There was nonsuitable cutoff point for TPPPS. If previous cutoff point 3 was selected, sensitivity of all target tests were too low. OPS yielded specificity 87.3% sensitivity 68.6%, FLACC yielded specificity 84.5% sensitivity 72.9% and TPPS yielded specificity 88.7% sensitivity 72.9%. Concurrent validity of target tests with CHEOPS were compared. Agreement both in recovery room and wards were acceptable (Recovery room: Kappa O = .792, T = .619, F = .795 and Ward: Kappa O = .617, F = .617). Spear-man correlation of all target tests were moderate to good in recovery room (r: O = .799, T = .790, F = .765, p < .001) and wards (r: O = .798, T = .826, F = .804 p < .001). Practicality of scales were tested. Time consumed in rating CHEOPS was the most (59sec) and TPPPS was the least (40.1sec). CHEOPS was more satisfied than the target tests. This study indicated that among the three target tests, OPS and FLACC were appropriate especially in recovery room. The predictive capability of target tests were less efficient in wards