Latent Profile Analysis of Patients Entering Interdisciplinary Pain Treatment: Subgroup Patterns and Relations with Gender, Race, and Opioid Use

Abstract

This study aimed to utilize profiling analysis to explore subgroups of patients entering a pain rehabilitation program (PRP) based on pre-treatment physical and emotional functioning. A secondary aim was to examine relations between gender, race, and opioid use with subgroup "membership," with potential treatment implications. N=113 patients who entered a 3-week, interdisciplinary PRP completed baseline assessments of pain, mood, pain coping, opioid misuse, and a 6-minute walk test. Latent profile analysis was utilized to identify subgroup profiles. One profile was added at a time; fit indices were examined after each step. Within-group means were examined to describe groups. Multinomial logistic regression was utilized to examine relations between profile "membership" and gender, race, and opioid use. Fit indices (AIC/BIC) indicated a 4-profile solution. Classification quality was good (entropy=.89). Within-group means revealed patterns: Group 1 (n=50) had high pain and poorer physical functioning but low distress; Group 2 (n=14) had favorable scores across measures; Group 3 (n=21) had moderate pain, high interference and distress, poor pain coping, but higher physical functioning; Group 4 (n=28) had the least favorable scores across measures. Women had higher odds of membership in Group 2 or 4 than 1 (p<.01). Opioid use was associated with lower likelihood of membership in Group 2 compared to all others (p<.01). White participants had higher odds of being in Group 3 compared to Groups 1 or 4; African Americans had higher odds of being in Group 4 than 3 (p<.01). These analyses offer preliminary evidence that patients may enter interdisciplinary pain treatment with distinct "profile" patterns of physical and emotional functioning. Gender and race differences were observed. Opioid use was lowest in the highest functioning subgroup. A more nuanced understanding of subgroup patterns prior to pain treatment has implications for tailored interventions, including gender and disparity-informed treatment. Support from Duke Endowment Grant.