Pediatric Anesthesiology Section Editor: Peter J. Davis A Comparison of the Clinical Utility of Pain Assessment Tools for Children with Cognitive Impairment Terri Voepel-Lewis, MSN, RN* Shobha Malviya, MD* Alan R. Tait, PhD* Sandra Merkel, MS, RN* Roxie Foster, PhD, RN† Elliot J. Krane, MD‡ BACKGROUND: Difficulty assessing pain has been cited as one of the primary reasons for infrequent and inadequate assessment and analgesia for children with cognitive impairment (CI). Several behavioral observational pain tools have been shown to have good psychometric properties for pain assessment in this population; however, routine clinical use may depend largely on their pragmatic qualities. We designed this study to evaluate pragmatic attributes or clinical utility properties of three recently developed pain assessment tools for children with CI. METHODS: A sample of clinicians from three medical centers were asked to review 15 videotaped observations of children with CI, recorded during their first three postoperative days during participation in a previous study. Participants scored pain using the revised-Face, Legs, Activity, Cry, Consolability (r-FLACC) tool (individualized for the child during the previous study) for five observations, the noncommunicative Non-Communicating Children’s Pain Checklist-Postoperative Version (NCCPC-PV) for five, and the Nursing Assessment of Pain Intensity (NAPI) for five observations. After their review of all segments, participants completed the Clinical Utility Attributes Questionnaire (CUAQ) ranking three attributes of clinical utility; complexity, compatibility, and relative advantage. RESULTS: Five physicians and 15 nurses comprised the sample. There was excellent agreement between the coded pain scores (i.e., mild, moderate, severe pain) assigned using all tools and r-FLACC scores assigned by original observers (88%–98% exact agreement;  0.71– 0.96). The internal consistency or reliability of the CUAQ was supported by high  values for each of the subscales ( = 0.84 – 0.93). Subscale and total CUAQ scores were higher for the r-FLACC and NAPI compared with the NCCPC-PV. The r-FLACC had similar scores for complexity, but slightly higher scores for compatibility, relative advantage, and total utility compared with the NAPI. CONCLUSIONS: We found that clinicians rated the complexity, compatibility, relative advantage, and overall clinical utility higher for the r-FLACC and NAPI compared with the NCCPC-PV, suggesting that these tools may be more readily adopted into clinical practice. (Anesth Analg 2008;106:72–8) The simplest and most reliable measure of postop- erative pain assessment is self-report by the patient, but most individuals with cognitive impairment (CI) are unable to report or quantify pain severity. 1–3 Difficulty assessing pain has been cited as one of the primary reasons for infrequent and inadequate assess- ment and analgesia in this population. 4 Recently, observation of behaviors has been used to identify and quantify pain in patients with CI, 5–10 which has lead to the development and testing of pain tools specifically designed for this population. 11–17 Several of these tools have been found to have good inter-rater reli- ability, and construct and criterion validity in assess- ing pain in children with CI. 11–14 Although these psychometric properties are necessary to ensure accu- rate pain assessment, the ability to implement these tools into routine clinical practice may depend largely on their pragmatic qualities. 18 Indeed, it has been suggested that busy clinicians tend to be pragmatists, concerned mainly with the practicality of innovations when considering their translation into practice. 19 The Diffusion of Innovations model, 20 which is commonly used to evaluate translational strategies of technology and innovation in the public health sec- tor, 21,22 identifies several characteristics that affect or explain implementation of an innovation or idea into routine practice. These attributes include the innova- tion’s relative advantage compared with others (de- gree to which it is perceived as being better than the From the *University of Michigan Health Systems, Department of Anesthesiology, Ann Arbor, Michigan; †University of Colorado Health Sciences Center, School of Nursing, Denver, Colorado; ‡Stanford University Medical Center, Department of Anesthesiology, Stanford, California. Accepted for publication August 23, 2007. Supported entirely by the NICHD; Grant 5 RO3 HD043920-02. Address correspondence and reprint requests to Terri Voepel- Lewis, MSN, RN, C.S. Mott Children’s Hospital, University of Michigan Health Systems, F3900 Box 0211, 1500 E. Medical Center Dr., Ann Arbor, MI 48109-0211. Address e-mail to terriv@umich. edu. Copyright © 2007 International Anesthesia Research Society DOI: 10.1213/01.ane.0000287680.21212.d0 Vol. 106, No. 1, January 2008 72