Personalising Psychotherapy for Chronic Primary Pain Using Network Analysis

Overview

The goal of this single case experimental design study with multiple baselines is to use network analysis to personalise cognitive behaviour therapy for chronic primary pain (CPP) and base the selection of individual treatment targets and interventions on data to avoid cognitive biases of the clinicians. The main questions it aims to answer are: * Is the study procedure accepted by and feasible for CPP patients as well as their therapists? * Does the personalised psychotherapy with databased clinical decisionmaking lead to significant improvement? Participants will go through several study phases: * Pretest and informational meeting with study management * Baseline 1: answering a questionnaire six times a day for 21 days in daily life on their mobilephone (EMA); this data will be used for the calculation of a network for each participant, that in turn will be used to select the treatment target and according treatment intervention as suggested by an algorithmic decisiontool * Probatory therapy phase: three weekly sessions with therapist; questionnaire three times a week * Baseline 2: questionnaire three times a week * Therapy phase: up to ten sessions with therapist; questionnaire three times a week * Post phase: posttest, two weeks of three weekly assessments, then another 21 days EMA; two monthly booster sessions with therapist * Follow-up: posttest and meeting with study management

Background and aims: The treatment of chronic primary pain (CPP) with Cognitive Behaviour Therapy (CBT) is effective, yet effects remain small to moderate. New approaches to the treatment of CPP are needed. They should include an orientation toward processes instead of symptoms and strategies for personalisation. This paradigm shift requests precisce, testable models that allow to link treatment procedures to psychological changes. Unfortunately, current models are rather vague about specific, especially temporal pathways by which its elements interact, or the pathways are not as simplistic as proposed. Therefore, multiple variables and pathways related to the development and persistence of pain-related disabilities should be considered. One approach to individualisation makes use of repeated measures in the daily lives of patients, which enable the usage of network analysis to investigate interactions between symptoms/processes of a patient. Process-oriented, individualised, network-based therapy for CPP (POINT Pain) combines process-oriented measurements to individualise therapy with network-based analysis approaches for target and treatment selection. We propose the following hypotheses: 1. In line with previous research, we assume that the implementation of POINT Pain is feasible for and accepted by patients as well as therapists. Feasibility and acceptance are indicated by mean ratings of at least 3.5 out of 5 across all items of one scale (each rated on 1-5 Likert scales) on a feedback form developed for this study and by positive comments to open ended questions. 2. Furthermore, we expect POINT Pain to lead to clinically significant improvements as compared to baseline, i.e. decreasing scores in the outcome measures. According to current recommendations, the improvement will be operationalised as effect-size estimated with Hedge's g, the significance with Tau-z. Procedure: Before participants are enrolled in the study, they will have two sessions with the study management for diagnostics, pretest, and individualisation of and instructions to the questionnaire. An ABA'C Multiple Baseline design is chosen to evaluate the effect of individually selected therapy interventions: Firstly, a baseline assessment (A) is conducted over the period of 21 days six times per day using EMA. In the EMA, a previously validated and piloted questionnaire with items on treatment relevant pain processes is implemented. It contains processes derived from the most prominent psychological models of chronic pain, as there is no overarching theoretical framework adressing the interaction of multi-dimensional pain processes yet. For the EMA, the waking hours of the participants will be divided into six equal time intervals within which the signals for the questionnaires will be randomised (semi randomisation). After completion of phase A, an individual network will be computed. Secondly, the probatory phase (B) is initiated after completion of phase A. After this phase, an application for psychotherapy is sent to the patient's insurance agency. As much information is already available from phase A, phase B will comprise three sessions. Participants will switch from EMA to SCED assessment in phase B, which means they will answer the same questionnaire as before three times a week, two of which during the week and one on the weekend. The timepoints for the questionnaires will again be semi-randomised during the waking hours of the participants. This procedure stays the same for the remaining phases. As there is often a natural break between the probatory and the treatment phases due to the psychotherapy application, another baseline (A') is implemented after phase B. The duration of this second baseline will be semi-randomised between one and three weeks. There are several possibilities for the selection of the most relevant or central variable in a network. The clinical decision regarding treatment target and module will be based on an algorithmic decision-tool which will is programmed based on experts' ratings (see preregistration: https://doi.org/10.17605/OSF.IO/BNFWY). Possible interventions will be limited to a pre-defined intervention pool that was developed based on common therapy manuals for chronic pain. Therapist and supervisor can only rule out the module suggested by the algorithmic decision-tool if there are serious concerns regarding potential harms (e.g. red flags rule out exposure in vivo). At last, the treatment phase (C) is realised. The selected intervention is carried out until clinically significant improvement is reached or until a maximum of ten sessions has been conducted. For all assessments using the EMA questionnaire (i.e. phases A to C), the software m-Path (https://m-path.io/landing/) published by KU Leuven will be used. In addition to the aforementioned assessments, participants will complete traditional outcome measures at three timepoints, i.e. before phase A (pretest), after phase C (posttest), and three months after finishing phase C (follow-up). After phase C, participants will have two more weeks of SCED, after which they will undergo another EMA phase of 21 days with a frequency of six assessments per day in order to calculate a post-therapy network. During this study phase, two monthly booster-sessions with their therapist will take place, but there will not be any further intervention, only monitoring. The results of the weekly SCED assessments will be analysed in a timely manner and discussed with the participants regularly. Feasibility and acceptance will be evaluated post-treatment and at follow-up. A feedback form developed for this study will be used for quantitative and qualitative assessment. Participants: Patients will be recruited via the waiting list of the university's outpatient clinic, in cooperation with other currently running studies at the same university recruiting chronic pain patients, and via various media (e.g. newspaper articles) and doctors' offices. Study therapists will be recruited in the university's outpatient clinic as well. Inclusion criteria for patients are at least 18 years of age, having access to a smartphone, and the main diagnosis of chronic pain. The diagnosis will be checked using the brief version of the Diagnostic Interview for Mental Disorders (Mini-DIPS). For patients recruited via the waiting list, screening for suitability will take place during the first consultation at the university psychotherapy training centre's outpatient clinic. Suitable participants will be informed about the study and referred if they agree to be contacted. Furthermore, patients that had to be excluded from other currently running studies will be referred if they agreed to be contacted as well. Analysis: In the context of SCEDs, each individual can be analysed separately. Especially regarding the qualitative measurements (acceptance, feasibility), it can be assumed that a sample of twelve will lead to a saturation in obtainable answers. Thus, all hypotheses can be answered with a sample of twelve: (1) We can evaluate whether the study procedure is feasible and accepted by patients and therapists, and (2) within SCED clinically significant change can be evaluated per individual. Of course, SCED results will not be generalizable to other individuals.