Why do patients go off track? Examining potential influencing factors for being at risk of psychotherapy treatment failure

Abstract

Background Routine outcome monitoring can support clinicians to detect patients who deteriorate [not-on-track (NOT)] early in psychotherapy. Implemented Clinical Support Tools can direct clinicians’ attention towards potential obstacles to a positive treatment outcome and provide suggestions for suitable interventions. However, few studies have compared NOT patients to patients showing expected progress [on-track (OT)] regarding such obstacles. This study aimed to identify domains that have predictive value for NOT trajectories and to compare OT and NOT patients regarding these domains and the items of the underlying scales. Methods During treatment, 413 outpatients filled in the Hopkins-Symptom-Checklist-11 (depressive and anxious symptom distress) before every therapy session as a routine outcome measure. Further, the Assessment for Signal Clients, Affective Style Questionnaire, and Outcome Questionnaire-30 were applied every fifth session. These questionnaires measure the following domains, which were investigated as potential obstacles to treatment success: risk/suicidality, therapeutic alliance, motivation, social support and life events, as well as emotion regulation. Two groups (OT and NOT patients) were formed by defining a cut-off (failure boundary) as the 90% confidence interval (upper bound) of the respective patients’ expected recovery curves. In order to differentiate group membership based on the respective problem areas, multilevel logistic regression analyses were performed. Further, OT and NOT patients were compared with regard to the domains’ and items’ cut-offs by performing Pearson chi-square tests and independent samples t-tests. Results The life events and motivation scale as well as the risk/suicidality scale proved to be significant predictors of being not-on-track. NOT patients also crossed the cut-off significantly more often on the domains risk/suicidality, social support, and life events. For both OT and NOT patients, the emotion regulation domain’s cut-off was most commonly exceeded. Conclusion Life events, motivation, and risk/suicidality seem to be directly linked to treatment failure and should be further investigated for the use in clinical support tools.

Long‐term psychotherapy in tertiary care: A practice‐based benchmarking study

“…It is unclear why social role was the only sub-scale that continued to make improvements (albeit at a reducing rate) or why interpersonal and symptom distress scales did not. This pattern is different to patterns previously identified (Schilling et al, 2020;White et al, 2015) and may be more characteristic of patients who access UK tertiary care therapy. Interpersonal problems being a predictor of poor treatment outcomes were consistent with prior evidence (Probst et al, 2020).…”

Prospective evaluation of a clinical decision support system in psychological therapy.

“…Patients were considered not on track, when their HCSL-11 total score at a given session crossed the upper boundary of the 90% confidence interval of their dynamically recalculated expected recovery curve 1 based on their most similar already-treated patients (cf. Lutz et al, 2019; Schilling et al, 2020). Of the overall sample ( N = 538), 245 patients (45.54%) were not on track (NOT) at least once during the first 25 sessions.…”

“…Within a digital portal accessible to therapists, the navigation system combines two modules: One offering therapists a personalized treatment strategy recommendation for the first 10 sessions (problem-solving, motivationoriented, or a mix of both strategies) and one supporting therapists during treatment with feedback, adaptive recommendations, and clinical problem-solving tools for patients at risk for treatment failure. The system is based on psychometric questionnaires routinely collected every session (see Schilling et al, 2020). Both modules of the data-driven decision support system are described in detail in Supplemental Materials 1.…”

Using feedback tools to enhance the quality and experience of care

“…These studies illustrate that patients place value on information sharing and potentially also improved communication with their clinicians; but clinicians need to be open to using these systems, systems need to connect to their needs in the specific consultation context, and clinicians need to be trained in their use [41]. The special issue closes with a paper [42] reporting secondary analyses of a trial investigating the effectiveness of routine outcome monitoring in a mental health outpatient service offering cognitive-behavioral therapy. In addition to a detailed description of how session-by-session PROMs data is collected [43], the analyses suggest that in this setting session number, suicidality, therapy motivation, and the occurrence of life events were predictive of deterioration in the later therapy sessions.…”