Procedural pain assessment in infants and young children: identifying a suitable behavioural assessment scale

Abstract

Infants and young children frequently experience pain as a consequence of medical procedures associated with their healthcare. Pain management is often suboptimal, and this is in part due to the difficulties associated with assessment of pain of infants and children too young to self-report pain intensity. Observable behaviours indicative of pain have long been considered a viable alternative and scales comprised of these behaviours have proliferated in the literature. However, it remains unclear which scales are best suited for procedural pain assessment and whether they are well supported by psychometric data.

The aims of this project were to: identify behavioural observation scales potentially suitable for procedural pain assessment, summarise available psychometric data and prospectively test the psychometric properties of potentially suitable scales when used to assess procedural pain in infants and young children. These aims were addressed in three phases of work: i) a thorough interrogation of the literature to identify scales considered potentially suitable for assessing procedural pain in infants and children, ii) a series of systematic reviews to summarise the evidence supporting the psychometric properties of the identified scales and iii) a prospective observational study to test the psychometric properties of these scales used to assess procedural pain in infants and young children.

Three scales, the Face, Legs, Activity, Cry and Consolability Scale (FLACC), the Modified Behavioral Pain Scale (MBPS) and the Visual Analogue Scale for observers (VASobs), met predefined criteria and were considered potentially suitable for inclusion in this project. The systematic reviews showed that available psychometric data was insufficient to recommend these scales for procedural pain assessment of infants and children. There was data to tentatively support the MBPS and to a lesser extent the VASobs for assessing immunisation related pain. The data regarding the FLACC scale was inconclusive.

The results of the prospective study confirmed that all scales were sensitive to pain. The FLACC scale and MBPS scores were reliable (intraclass correlation (ICC) 0.92 and 0.87, respectively) but VASobs scores were less reliable (ICC 0.55). The FLACC scores showed the highest sensitivity (94.9%) and specificity (72.5%) for procedure type (painful vs non-painful) at the lowest cut-off score (pain score 2, area under the curve (AUC 0.83)). Similar results were achieved at a MBPS cut-off score of 4 (sensitivity 91.5%, specificity 77.5%, AUC 0.85). The FLACC scale resulted in more incomplete scores (p < 0.000) and was changed more often than other scale scores. Reviewers liked the VASobs most, considered it the quickest and easiest to apply, but judged the FLACC scale and MBPS to be more likely to be useful.

In conclusion, three behavioural observational pain scales to assess procedural pain in infants and young children were identified and included in systematic reviews. This work culminated in a prospective study, the results of which support use of the FLACC scale, but not without reservation as there are practical limitations when used to assess procedural pain. These results build on promising existing evidence that suggests that the FLACC scale may currently be a suitable scale for procedural pain assessment in infants and young children.