Therapeutic Interactive Voice Response for chronic pain reduction and relapse prevention

Abstract

We developed Therapeutic Interactive Voice Response (TIVR) as an automated, telephone-based tool for maintenance enhancement following group cognitive-behavioral therapy (CBT) for chronic pain. TIVR has four components: a daily self-monitoring questionnaire, a didactic review of coping skills, pre-recorded behavioral rehearsals of coping skills, and monthly personalized feedback messages from the CBT therapist based on a review of the patient's daily reports. The first three components are pre-recorded and all four can be accessed remotely by patients via touch-tone telephone on demand. Following 11 weeks of group CBT, 51 subjects with chronic musculoskeletal pain were randomized to one of two study groups. Twenty-six subjects participated in 4 months of TIVR, while a control group of 25 subjects received standard care only. The TIVR group showed maximum improvement over baseline at the 8-month follow-up for seven of the eight outcome measures; improvement was found to be significant for all outcomes (p<or=001). Between-group analysis of covariance (ANCOVA) revealed significantly greater improvement for the experimental group at both 4- and 8-month follow-ups for most of the outcomes. Results demonstrate that TIVR can be used to decrease pain, improve coping and decrease likelihood of relapse into pain behavior. Preliminary analysis of medication usage suggests that the superior outcome of the TIVR group was unlikely to be a consequence of differential medication use.

Figure 1

STUDY DESIGN

Figure 2

Mean scores for the control and TIVR groups at four time points. The two groups’ scores were similar at post-CBT but became significantly different by the end of the TIVR intervention and more so by the last follow-up since the TIVR group continued to improve. The p-values are reported for the between-group comparisons, adjusted for post-CBT differences (ANCOVA), at the last two assessments.

Figure 2

Mean scores for the control and TIVR groups at four time points. The two groups’ scores were similar at post-CBT but became significantly different by the end of the TIVR intervention and more so by the last follow-up since the TIVR group continued to improve. The p-values are reported for the between-group comparisons, adjusted for post-CBT differences (ANCOVA), at the last two assessments.

Figure 2

Mean scores for the control and TIVR groups at four time points. The two groups’ scores were similar at post-CBT but became significantly different by the end of the TIVR intervention and more so by the last follow-up since the TIVR group continued to improve. The p-values are reported for the between-group comparisons, adjusted for post-CBT differences (ANCOVA), at the last two assessments.

Figure 2

Mean scores for the control and TIVR groups at four time points. The two groups’ scores were similar at post-CBT but became significantly different by the end of the TIVR intervention and more so by the last follow-up since the TIVR group continued to improve. The p-values are reported for the between-group comparisons, adjusted for post-CBT differences (ANCOVA), at the last two assessments.