Non-cardiac Chest Pain: Effect of Cognitive Therapy Administered as Guided Self-help

This study aims to test an easily implementable web-based cognitive behaviour therapy (CBT) self-help intervention designed for non-cardiac chest pain patients, and compare effectiveness and cost-effectiveness to treatment as usual in an randomized controlled trial (RCT).

Most patients with chest pain referred to hospital do not have a cardiac illness. Non-cardiac chest pain (NCCP) is often followed by persistent distress and reduced quality of life, and societal costs are nearly equal to those of cardiac patients. Research suggests that face-to-face CBT is effective, but this has not been implemented as standard treatment. We plan to test an easily implementable web-based guided self-help intervention for NCCP patients. Patients will be recruited at the chest pain unit at Sørlandet Hospital, Kristiansand, and will be recruited after they have finished their cardiac examination. The intervention group will receive six web-based sessions, comprising information, exposure to physical activity, how worry can excess pain, physical reactions to pain and worry, consequences of avoidance, and specific panic treatment.

Eligible patients will be consecutively included and randomized to two groups, intervention group or control group. We will use a web-based randomisation procedure, conducted at a place remote from where the study takes place.

The intervention group will receive guided self-help comprising six web-based sessions, comprising information, exposure to physical activity, how worry can excess pain, physical reactions to pain and worry, consequences of avoidance, and specific panic treatment. The first session will be done at the hospital before discharge, the others at home. Between sessions there will be a brief telephone contact with a project worker.

This group will receive treatment as usual, which is no specific treatment. They can however use the general health system as they like.

CAQ is an 18-item self-report-questionnaire which measures hearth related anxiety/fear, attention, avoidance and safety seeking behavior. Each item is rated on a five-point Likert scale ranging from 0 to 4. Total range 0-72 where higher score represents more symptoms. Primary outcome is difference in mean score between intervention arm and control arm for CAQ at end of treatment (6 weeks after randomization for both arms). Difference in mean scores between arms will also be assessed at pretreatment/baseline, and at 3 months and 12 months after end of treatment (18 weeks and 58 weeks after randomization). There is no predefined clinical relevant effect for this measure, we use a distribution based method and define clinical relevant effect as improvement of more than 0.5 standard deviation compared to baseline.

BSQ is an 17-item self-report-questionnaire which measures fear of different bodily sensations. Each item is rated on a five-point Likert scale ranging from 1 to 5. Total range 17-85 where higher score represents more symptoms. Primary outcome is difference in mean score between intervention arm and control arm for BSQ at end of treatment (6 weeks after randomization for both arms). Difference in mean scores between arms will also be assessed at pretreatment/baseline, and at 3 months and 12 months after end of treatment (18 weeks and 58 weeks after randomization). There is no predefined clinical relevant effect for this measure, we use a distribution based method and define clinical relevant effect as improvement of more than 0.5 standard deviation compared to baseline.

BIPQ is an 8-item self-report-questionnaire which assess the patient's perception/beliefs about their symptoms. Each item is rated on a 0-10 scale. The items represent different components/dimensions. Secondary outcome is difference in mean score between intervention arm and control arm for item 1, 5, 6 and 8 at end of treatment (6 weeks after randomization for both arms). Difference in mean scores between arms will also be assessed for item 1, 5, 6 and 8 at pretreatment/baseline, and at 3 months and 12 months after end of treatment (18 weeks and 58 weeks after randomization). There is no predefined clinical relevant effect for these items, we use a distribution based method and define clinical relevant effect as improvement of more than 0.5 standard deviation compared to baseline.

PHQ-9 is an 9 item self-report-questionnaire which assess depression symptoms. Each item is rated on a 0-3 scale. Total range 0-27 where higher score represents more symptoms. Secondary outcome is difference in mean score between intervention arm and control arm for PHQ-9 at end of treatment (6 weeks after randomization for both arms). Difference in mean scores between arms will also be assessed at pretreatment/baseline, and at 3 months and 12 months after end of treatment (18 weeks and 58 weeks after randomization). We define clinical relevant effect as an absolute drop of 5 points on an individual level, and between groups as an improvement of 3 points.

GSE is an 10 item self-report-questionnaire which assess self-efficacy. Each item is rated on a 1-4 Lickert scale. Total range 10-40 where higher score represents more self-efficacy. Secondary outcome is difference in mean score between intervention arm and control arm for GSE at end of treatment (6 weeks after randomization for both arms). Difference in mean scores between arms will also be assessed at pretreatment/baseline, and at 3 months and 12 months after end of treatment (18 weeks and 58 weeks after randomization). There is no predefined clinical relevant effect for this measure, we use a distribution based method and define clinical relevant effect as improvement of more than 0.5 standard deviation compared to baseline. We will also investigate its ability to predict treatment effect.

EQ-5D-5L is an 5 item (plus a VAS scale) self-report-questionnaire assessing health related quality of life. Secondary outcome is difference in mean score between intervention arm and control arm for EQ-5D-5L at end of treatment (6 weeks after randomization for both arms). Difference in mean scores between arms will also be assessed at pretreatment/baseline, and at 3 months and 12 months after end of treatment (18 weeks and 58 weeks after randomization). Distribution based methods are commonly used, defining improvement of 0.2 or 0.5 SD as clinical relevant. We define a minimal clinical relevant improvement to be 0.2 SD in EQ-5D-5L.

Secondary outcome is to calculate the difference in healthcare utilization between intervention and control arm over a period of 12 months after end of treatment. We will collect data on both direct and indirect costs with self-report forms collected every third month, as well as register data. We will compare the costs between the two arms.

If the primary and secondary outcomes surprisingly do not show difference of statistical significance and clinical relevance, the study data will be searched for composite measures showing clinically and statistical relevance

Inclusion Criteria: Non-Cardiac Chest Pain Exclusion Criteria: Language difficulties Unable to perform at least moderate physical activity due to physical constraints Obvious cognitive impairment (e.g mentally retarded, psychotic, dementia or intoxicated) no regular access to a computer/tablet computer with internet connection

Thesen T, Himle JA, Martinsen EW, Walseth LT, Thorup F, Gallefoss F, Jonsbu E. Effectiveness of Internet-Based Cognitive Behavioral Therapy With Telephone Support for Noncardiac Chest Pain: Randomized Controlled Trial. J Med Internet Res. 2022 Jan 24;24(1):e33631. doi: 10.2196/33631.