Systematic review and meta-analysis suggest that varying prevalence of non-acute pain in critically ill infants may be due to different definitions

Aim: Our aim was to quantify the prevalence of non-acute pain in critically ill infants and to identify how non-acute pain was described, defined and assessed. Methods: This systematic review and meta-analysis used multiple electronic databases to search for papers published in any language to March 2018: 2029 papers were identified, and 68 full texts were screened. Studies reporting the prevalence of non-acute pain in infants younger than 2 years and admitted to critical care units were included. The extracted data included the use of non-acute pain descriptions, definitions and pain assessment tools. Results: We included 11 studies published between 2002 and 2018 that comprised 1204 infants from Europe, the USA, Canada and India. They were prospective observational (n = 7) and retrospective observational (n = 1) studies and randomised controlled trials (n = 3). The prevalence of non-acute pain was 0%-76% (median 11%). Various pain assessment tools were used, and only two could be pooled. This gave a pooled prevalence of 3.7%-39.8%. A number of different descriptors were used for non-acute pain, and all of these were poorly defined. Conclusion: The prevalence of non-acute pain in infants admitted to critical care units varied considerably. This could have been because all the studies used different definitions of non-acute pain.