Economic evaluation of group CBT versus enhanced usual care for functional somatic syndromes

[hixxy]

Long-term economic evaluation of cognitive-behavioural group treatment versus enhanced usual care for functional somatic syndromes.
Schröder A, Ørnbøl E, Jensen JS, Sharpe M, Fink P.

Abstract

OBJECTIVE:
Patients with functional somatic syndromes (FSS) such as fibromyalgia and chronic fatigue syndrome have a poor outcome and can incur high healthcare and societal costs. We aimed to compare the medium-term (16months) cost-effectiveness and the long-term (40months) economic outcomes of a bespoke cognitive-behavioural group treatment (STreSS) with that of enhanced usual care (EUC).

METHODS:
We obtained complete data on healthcare and indirect costs (i.e. labour marked-related and health-related benefits) from public registries for 120 participants from a randomised controlled trial. Costs were calculated as per capita public expenses in 2010 €. QALYs gained were estimated from the SF-6D. We conducted a medium-term cost-effectiveness analysis and a long-term cost-minimization analysis from both a healthcare (i.e. direct cost) and a societal (i.e. total cost) perspective.

RESULTS:
In the medium term, the probability that STreSS was cost-effective at thresholds of 25,000 to 35,000 € per QALY was 93-95% from a healthcare perspective, but only 50-55% from a societal perspective. In the long term, however, STreSS was associated with increasing savings in indirect costs, mainly due to a greater number of patients self-supporting. When combined with stable long-term reductions in healthcare expenditures, there were total cost savings of 7184 € (95% CI 2271 to 12,096, p=0.004) during the third year after treatment.

CONCLUSION:
STreSS treatment costs an average of 1545 €. This cost was more than offset by subsequent savings in direct and indirect costs. Implementation could both improve patient outcomes and reduce costs.

Copyright © 2017 Elsevier Inc. All rights reserved.

KEYWORDS:
Bodily distress syndrome; Chronic fatigue syndrome; Cognitive-behavioural therapy; Cost-effectiveness; Economic evaluation; Fibromyalgia; Functional somatic syndromes

https://www.ncbi.nlm.nih.gov/pubmed/28183406
http://www.jpsychores.com/article/S0022-3999(17)30018-1/abstract

 

[Keith Geraghty]

the original study was highly flawed and biased - i dont want to waste time assessing this cost-evaluation given its build on the back of the flawed primary study

 

[A.B.]

Soon they will discover an even more cost effective way to deal with the unwanted patients:

 

[Dolphin]

The difference in QALYs seemed quite small: 0.035, which just reach statistical significance at p<0.05.

The QALY for the intervention group went from 0.58 at baseline to 0.6, 0.61 and 0.62 had 4 months, 10 months and 16 months respectively. Those numbers still seem quite low.
----

QALYs and clinical improvement
Quality-adjusted life years (QALYs) were generated on the basis of eleven items of the 36-
item Short Form Health Survey (SF-36)34 converted into SF-6D utility scores based on
weights of the general UK population according the method of Brazier.35

 

[Dolphin]

In weeks where no social benefits were registered or where patients were registered with transfer payments not related to reduced work ability or unemployment, such as maternity pay or state education grants, patients were regarded as self-supporting. They were assigned negative costs to reflect estimated public tax income, again according to advise from KORA.38

Weeks without transfer payments were assigned negative costs due to estimated public tax income; or zero in
sensitivity analyses without tax income

Note what they did with tax. They didn't have records of the tax the people actually paid. They just assumed they were paying tax if they weren't claiming disability benefits/payments. This seems a big assumption.

However there was still a difference even if one ignores the tax issue.

 

[Dolphin]

Diagnoses based on functional somatic symptoms in the past 2 years, according to SCANdiagnostic
interview and review of clinical records.

56 % in the treatment group and 62% in the control group were said to have CFS. It is unclear what criteria were used. Michael Sharpe was an author so maybe it was the Oxford criteria.

There was no subgroup analyses for different illnesses.

 

[Dolphin]

The treatment manual is available at:
http://funktionellelidelser.dk/en/for-specialists-researchers/psychologists/stress-manual/

 

[Dolphin]

Second, we relied on high-quality register data, which not only provided us with a complete dataset, but also avoided any recall- or response-bias. This is especially important as patients could not be blinded to treatment allocation, leading to a possibility of bias in self-reported outcomes.

This wasn't done in the PACE Trial.

 

[Dolphin]

The main results are in tables 2a and 2b which could be read without reading the paper. Tables 3 and 4 also have outcome data but they are a bit more abstract.

 

[Dolphin]

Note what they did with tax. They didn't have records of the tax the people actually paid. They just assumed they were paying tax if they weren't claiming disability benefits/payments. This seems a big assumption.

However there was still a difference even if one ignores the tax issue.

I do wonder whether an intervention like this could cause people either to:
(i) blame themselves for not getting better;
(ii) not see themselves as ill when they really are/see themselves as recovered when they are not;
(iii) lose confidence in the health and/or medical profession.

This might potentially cause people to be less inclined to claim disability payments as they might feel they are not worthy or they might feel no one would support them getting a claim.

Also they might go to medical and health professionals less often as they feel they can just get themselves better (other symptoms are just normal) or alternatively have given up on the current system helping them.

This makes me wonder whether for example one can assume people not claiming disability payments are definitely working.