Validity of pain behaviors in persons with mild to moderate cognitive impairment

OBJECTIVES: To evaluate the validity of traditional pain behaviors (guarding, bracing, rubbing, grimacing, and sighing) in persons with and without cognitive impairment and chronic low back pain (CLBP). DESIGN: Prospective observational study. SETTING: Outpatient clinics. PARTICIPANTS: Thirty-seven cognitively intact and 40 cognitively impaired participants with and without CLBP. MEASUREMENTS: Frequency of traditional pain behaviors. RESULTS: Forty-six of the participants were pain free, and 31 had CLBP. The internal consistency reliability coefficient of the five pain behaviors was 0.32, suggesting that a unidimensional scale did not exist. Multivariate analysis of variance analysis according to the independent variables pain status (pain free vs CLBP) and cognitive status (intact vs impaired) with the dependent variable frequency of pain behaviors found significant differences according to pain status (F[5,61]=3.06, P=.02) and cognitive status (F[5,61]=5.41, P <.001) but without evidence of an interaction (F[5,61]=1.14, P=.35). Participants with CLBP exhibited significantly higher levels of grimacing (P <.001) and guarding (P=.02) than pain-free participants. Intact subjects exhibited fewer guarding (P=.02) and rubbing behaviors (P <.001) but a higher number of bracing behaviors (P=.03) than cognitively impaired participants. CONCLUSION: These results support the utility of facial grimacing in assessing pain in patients with mild to moderate cognitive impairment and call into question the validity of guarding and rubbing in assessing pain in persons with mild to moderate cognitive impairment.