Compliance therapy: a randomised controlled trial in schizophrenia
BMJ 2003; 327 doi: https://doi.org/10.1136/bmj.327.7419.834 (Published 09 October 2003) Cite this as: BMJ 2003;327:834All rapid responses
Rapid responses are electronic comments to the editor. They enable our users to debate issues raised in articles published on bmj.com. A rapid response is first posted online. If you need the URL (web address) of an individual response, simply click on the response headline and copy the URL from the browser window. A proportion of responses will, after editing, be published online and in the print journal as letters, which are indexed in PubMed. Rapid responses are not indexed in PubMed and they are not journal articles. The BMJ reserves the right to remove responses which are being wilfully misrepresented as published articles or when it is brought to our attention that a response spreads misinformation.
From March 2022, the word limit for rapid responses will be 600 words not including references and author details. We will no longer post responses that exceed this limit.
The word limit for letters selected from posted responses remains 300 words.
EDITOR – O’Donnell et al report their failure to find benefit for
compliance therapy in schizophrenia (1). This is somewhat at odds with the
positive findings of an earlier controlled trial of this therapy in
psychotic patients (2,3) and O’Donnell et al discuss possible explanations
for the discrepancy. One factor relevant to the interpretation of both
studies is a possible unrecognised selection bias: those people for whom
such an intervention would be most clearly indicated, that is, those with
psychotic illness who are particularly poorly adherent to their medication
regimen, may be under-represented. There is evidence that consent to take
part in a treatment study and adherence to treatment are linked in
schizophrenia, with those who are not adherent being less likely to
consent to participate in a trial (4). This is relevant to the
generalisability of the findings, given that the proportion of eligible
patients who did not provide consent was 34% in O’Donnell et al (1) and
31% in Kemp et al (2).
We had experience of the impact of this association between adherence
and consent in a non-commercially funded, peer-reviewed study of
compliance therapy in schizophrenia that we embarked upon some years ago.
We sought to recruit only people with a demonstrable history of poor
adherence to medication for relapse prevention, and chose to study people
receiving depot antipsychotic medication as this form of drug
administration does not allow for covert non-adherence. However, this
approach proved unrewarding, as only a small proportion of eligible
subjects provided consent.
Large, pragmatic studies testing the effectiveness of compliance
therapy in schizophrenia may now be warranted. To generate generalisable
findings reflecting normal clinical practice, such studies will need to
develop strategies to engage people with schizophrenia whose persistent
non-adherence to their treatment regimens has had evident adverse clinical
consequences.
Competing interests: None declared
Thomas R. E. Barnes professor
Tom Sensky professor
t.r.barnes@imperial.ac.uk
Department of Psychological Medicine, Imperial College, Charing Cross
campus, London W6 8RP
Shôn Lewis professor
School of Psychiatry & Behavioural Sciences, University of Manchester,
Manchester M23 9LT
1. O’Donnell C, Donohoe G, Sharkey L, Owens N, Migone M, Harries R,
Kinsella A, Larin C, O’Callaghan E. Compliance therapy: a randomised
controlled trial in schizophrenia. BMJ 2003;327:834-836.
2. Kemp R, Hayward P, Applewhaite G, Everitt B, David A. Randomised
controlled trial of compliance therapy. BMJ 1996;312:345-349.
3. Kemp R, Kirov G, Everitt B, Hayward P, David A. Randomised
controlled trial of compliance therapy: 18-month follow-up. Br J
Psychiatry 1998;172:413-419.
4. Bowen J, Barnes TRE. The clinical characteristics of schizophrenic
patients consenting and not consenting to a placebo-controlled trial.
Human Psychopharmacology 1994; 9: 423-433.
Competing interests:
None declared
Competing interests: No competing interests
Dear Sir,
Regarding the study in discussion, it should be noted that sample size
available for analysis was inadequate to conclude that the compliance
therapy was no better than non-specific counseling. Multivariate analysis
like logistic regression as conducted in this study requires a larger
sample size than univariate analysis. However, the authors have not
accounted for this in their power analysis.
It is also possible that some exchange of information could have
taken place between the intervention and control groups as they came from
the same geographic location and could have belonged to the same social
groups. This could have caused dilution of the intervention effect and
result in showing no significant differences between the two groups.
Another reason for failing to get significant differences between the
intervention and control groups could be the quality of the intervention
itself. The authors should have described their intervention in more
detail providing additional information such as whether trained experts
conducted the sessions and what information was covered.
Non-compliance problems are usually a result of the medication side-
effects among schizophrenia patients1,2. The authors have not discussed
these issues and how their intervention aimed to deal with these issues.
Patients on typical antipsychotic generally have more non-compliance
problems than those on atypical antipsychotics1,2. Therefore, the study
should have given a brief description of the patients’ antipsychotic
medication history and adjusted for the type of antipsychotic the patient
was taking.
It is interesting that despite the lack of sample size, variables
like carer involvement and attitude towards treatment were significant in
the logistic model. The study underscores the importance of these
variables in developing interventions for improving compliance.
1.Amadio PB, Cross LB, Amadio P. New drugs for schizophrenia: an
update for family physicians. Am Fam Physician 1997;56:149–1156.
2.Leucht S, Barnes TR, Kissling W, Engel RR, Correll C, Kane JM. Relapse
prevention in schizophrenia with new-gereration antipsychotics: a
systematic review and exploratory meta-analysis of randomized, controlled
trials. Am J Psychiatry 2003;160:1209-22.
Competing interests:
None declared
Competing interests: No competing interests
EDITOR – I congratulate O’Donnell et al on their study of compliance
in schizophrenia – a notoriously difficult subject to research and a
notoriously difficult condition to manage.1 In their discussion of the
relationship between intervention, compliance and outcome, however, they
appear to have not yet reported the effects of treatment choice on
compliance and therefore on efficacy. Patients in this study were not
randomised on the basis of drug treatment and this might bear a
significant relationship to compliance itself.
It is common for innovative treatments for schizophrenia to be
related to improved compliance. In formal studies this is often expressed
as “discontinuation due to adverse events or inefficacy”. Recent meta-
analytical studies have confirmed that newer antipsychotic agents tend to
induce fewer side-effects than older drugs whilst retaining efficacy and
that the role of treatment adherence may be of great importance.2
Similarly, in the latest studies on the partial dopamine agonist
aripiprazole, the discontinuation rate was significantly lower in the
aripiprazole group than in a control group managed with haloperidol.3
In the O’Donnell paper, patients were not randomised on the basis of
treatment type. The lower compliance they observed in some patients may
therefore simply be a feature of the adverse effect profile of the
therapies themselves. Future research might focus profitably on the
relationship between the type of intervention and tolerability and the
medication given.
1 O’Donnell C, Donohoe G, Sharkey L, Owens N, Migone M, Harries R, et
al. Compliance therapy: a randomised controlled trial in schizophrenia.
BMJ 2003;327:834-7. (11 October.)
2 Leucht S, Barnes TR, Kissling W, Engel RR, Correll C, Kane JM.
Relapse prevention in schizophrenia with new-gereration antipsychotics: a
systematic review and exploratory meta-analysis of randomized, controlled
trials. Am J Psychiatry 2003;160:1209-22.
3 Kujawa M, Saha AR, Ingenito GG, Ali MW, Luo X, Archibald DG, et al.
Aripiprazole for long-term maintenance treatment of schizophrenia. Int J
Neuropsychopharmacol 2002;5(suppl 1):S186-7.
Competing interests:
None declared
Competing interests: No competing interests
Editor,
We were interested in a replication (1) of Kemp and colleagues’ work
on ‘compliance therapy’ (2). The original work delivered ‘compliance
therapy’ with an innovative philosophy not specifically identified in this
paper. Anxiety about the inter-use of the terms compliance, adherence and
concordance will be exacerbated by this omission.
Cognitive Behavioural Therapy (CBT) is a collaborative venture, where
pre-determined treatment decisions are anathemas. Compliance therapy
should be delivered with the philosophy that taking medication would be a
‘freely-chosen strategy to enhance quality of life’ and in the original
report was delivered by cognitive therapists. This study reports the use
of the manual, but does not report the training of the therapists,
evaluation of adherence to the manual or emphasise the philosophy
identified above. Consequently, the anxiety remains that this study
(partly supported by the pharmaceutical industry) did not truly subscribe
to a collaborative approach in its inception and was consequently
fundamentally flawed. One worrying conclusion might be that we only need
prescribe atypical anti-psychotics and patients will take them.
The development of Early Intervention in Psychosis (EIP) Teams in the
U.K., some of which include strong emphasis on a Recovery model are an
important development (3). The Recovery approach recognises and endorses
the role of individuals in making treatment decisions, one among which is
medication use. While neglecting the helpful role of medication could
never be promoted and teams would avoid colluding with potentially
detrimental decisions, team members tolerate the possibility that the
service user may choose not to take medication and are prepared to
continue working with them to address other areas of functioning.
This empowerment of people, in keeping with the philosophy of the
original work, may paradoxically improve their regularity of medication
use. Treating people in this way is also a powerful tool to social
inclusion and combat to stigma. A local EIP service that uses this
approach has reduced readmission bed days by 96% (4). The ability of
teams to offer other effective interventions prevents their own anxiety
and hopelessness, which is often at the heart of controlling responses in
over-stretched services. Moreover, true collaboration also mitigates
against a controlling response from clinicians. True collaboration also
takes time and EIP teams are staffed to have lower caseloads than some
other services. Yet this approach is one to which all mental health
services users are surely entitled.
1. O’Donnell, C, Donohoe, G, Sharkey, L, et al. Compliance Therapy: a
randomised trial in schizophrenia. BMJ 2003; 327: 834-7.
2. Kemp, R, Kirov, G, Everitt, B, et al. Randomised controlled trial of
compliance therapy. 18-month follow-up. Br J Psychiatry 1998; 172: 413-9
3. The NHS Plan. A plan for investment. A plan for reform. London:
Stationery Office, London, 2000.
4. Dodgson, G, Cramb, G, Mitford, E, & Proctor, S. Progress Report on
Northumberland Early Intervention Team. Internal report for Locality
Management Group, 2003.
Competing interests:
None declared
Competing interests: No competing interests
Editor -
Colin O’Donnell and colleagues (1) report a lack of significant
effect of compliance therapy, and so fail to replicate our original study
(2,3) and others like it (4,5). This may be due to their more homogeneous
patient group and lower statistical power. Another possibility is that
the interventions were not delivered with sufficient expertise and
fidelity. As in our original study the authors failed to define in detail
the control intervention of non-specific counselling.
However, by showing
outcome measures immediately post intervention we were at least able to
establish the initial advantage of the active treatment, which in that
case was maintained over the follow-up. O’Donnell et al do not provide
such data, hence, we are in the dark as to whether their intervention had
any effect at all, on for example, attitudes to medication. Hence, the
lack of effect in the long term becomes difficult to interpret. Our
compliance therapy intervention was delivered by ‘expert’ therapists, a
potential disadvantage if this cannot be generalised without training.
The same two individuals gave the control intervention, which was clearly
inferior to compliance therapy. We do not know whether the same is true
of the latest study.
References
1. O’Donnell C, Donohoe G, Sharkey L, Owens N, Migone M, Harries R,
Kinsella A, Larkin C, O’Callaghan E. Compliance therapy: a randomised
controlled trial in schizophrenia. BMJ 2003;327:834-6.
2. Kemp R, Applewhaite G, Hayward P, Everitt B, David, A. Compliance
therapy in psychotic disorders: randomised controlled trial. BMJ 1996;
312:345-349.
3. Kemp R, Kirov GB, Hayward P, David A. Randomised controlled trial
of compliance therapy – 18 month follow-up. Brit J Psychiat 1998;172: 413-
419.
4. Zygmunt A, Olfson M, Boyer CA, Mechanic, D. Interventions to
improve medication adherence in schizophrenia. Am J Psychiat 2002; 159:
1653-64.
5. Nosè M., Barbui C., Gray R. & Tansella M. Clinical
interventions for reducing treatment non-adherence in psychoses: meta-
analysis. Brit J Psychiat 2003;183:197-206.
Competing interests:
None declared
Competing interests: No competing interests
I was intrigued to read this paper but noted that the authors did not
make it clear whether they were, in fact, offering compliance therapy.
Surguladze et al (2002) demonstrated that, with a 2-day interactive
course, it is possible to change trainees attitudes to both patients and
compliance/concordance issues. However, it did take 2 days and a large
amount of participative learning. We know that reading manuals is a very
poor way of changing behaviour. We also know that maintaining models of
psychological intervention is not straightforward in practice. When
offering a psychological or behavioural intervention it would seem prudent
to check that it is "exactly what it says on the tin". It may well be that
what was delivered in this study was not compliance therapy, as described
in the manual, but business as usual.
Simon Surguladze, Philip Timms, and Anthony S. David
Teaching psychiatric trainees ‘compliance therapy’
Psychiatr. Bull., Jan 2002; 26: 12 - 15.
Competing interests:
I have run compliance therapy courses
Competing interests: No competing interests
Editor- O'Donnell et al indicate that compliance therapy is not of
benefit to patients with schizophrenia but on the basis of their
study I don't believe that they can conclude this. 1 Firstly, as they
have indicated in the limitations section of the paper, their sample
size was small. Secondly, they had the bare minimum number of
subjects (56) needed to power the study and, of these, 6 dropped
out. Consequently, a type 2 error is quite possible.
The study highlighted that factors such as baseline compliance
and attitude to treatment were indicators of compliance at one
year. Intuitively this makes sense. It's quite possible that those
who are going to comply with treatment will do so anyway and that
maybe we should be engaging the subgroup who don't with
compliance therapy specifically tailored towards them. In this
regard, it would be interesting to see what the effects would be of
spreading the therapy sessions over 6 months or a year (linked in
with outpatients appointments and included in patients' care
plans).
1 O'Donnell C, Donohoe G, Sharkey L, Owens N, Migone M,
Harries R, Kinsella A, Larkin C, O'Callaghan E. Compliance
therapy: a randomised controlled trial in schizophrenia. BMJ
2003;327:834-6
Competing interests:
None declared
Competing interests: No competing interests
EDITOR--It was interesting to read the article of O'Donnell et al (BMJ
2003),Compliance therapy:a randomised controlled trial in
schizophrenia(1). My comment is that Schizophrenic patients are well known to
be non-compliant with their medications, up to 50% are non-compliant,and
the majority of those will have a relapse of their illness. Naber &
Karow suggests that to improve compliance could be through a good doctor
patient relationship and the use of atypical antipsychotics, as one of the
main reasons for their non-compliance is the side effects of their
medications, which are minimal among the atypical anti-psychotic group(2).
Thanking you,
Yours sincerely,
AK.Al-Sheikhli
References
1.O'Donnell C,Donohoe G,Sharkey L etal,Compliance therapy:a randomised
controlled trial in schizophrenia,BMJ(2003);327:834.
2.Naber D& Karow A,Good tolerability equals good results:the
patient's perspective,Eur Neuropspchopharmacol,(2001),11 supp 4,S391-6.
Competing interests:
None declared
Competing interests: No competing interests
During the past few days there has been another 'schitzophrenic
killer' story splashed all over the country with usual lazy journalism
adding to the fear of those labelled with this tag.eg cut straight to the
Zeto Foundation which constantly refers to people who 'stop taking 'their'
medication'. No reference at all to RETHINK which represents many who
suffer from mental health problems and have the right to contribute to the
debate. Many make a considered choice not to take medication - especially
when they have been lied to in the past about it's negative effects,are
given shoddy drugs, especially those under the 80 IQ identified by this
study group, are not given proper advice about options and are not
routinely tested to ensure the safety of the medication they are coerced
into taking. Unless being used as research fodder hardly any will receive
appropriate psychological help if this is the treatment of THEIR choice.
This study constantly changes it's terms, 'compliance' 'adherence' or
'concordance' - they need to be open about the aims of the study rather
than blur it when the effects of such flawed projects may have serious
effects on others'lives. The drip drip of messages from such as this group
position people with serious mental illness diagnoses as needing to be
constantly surveilled and controlled and medicated. It highlights the
issue of IQ, a dodgy criterion in anybody's book, of 'refusal' by people
to participate but only 'guesses' at the reason. Many people are rightly
suspicious of having their information recorded, lack faith in the aims of
research carried out by those who perceive them in terms of IQ levels;
'compliance''refusals''survival in the community' of needing 'key
informants', of 'occupying hospital beds'.
As far back as 1996 (Journal of psychiatry) psychiatrists in cardiff were
suggesting that they should have the power to compulsorily medicate people
diagnosed as schizophrenic' - for life. Cardiff has subsequently had two
special investigations into it's psychiatric services but
individual clinicians still have too much individual freedom to withhold
informed choice and too many powers to coerce - which seems to getting
worse.There is still a long way to go in many areas before practitioners
learn the idea of partnership, concordance
Competing interests: ?
None declared
Competing interests: No competing interests
Efficacy of compliance therapy in schizophrenia
Dear Sir,
We read with great interest the trial by O’Donnell et al (2003)
evaluating the efficacy of compliance therapy, a potentially important
brief intervention, based on motivational interviewing and cognitive
behavioural therapy, to improve treatment adherence in people with
schizophrenia. The findings of O’Donnell et al (2003) contradict the
earlier study by Kemp et al (1996) and appear to suggest that Compliance
Therapy may not be of benefit to patients with schizophrenia. In some
respects the O’Donnel et al (2003) trial is more methodologically robust
than the Kemp et al (1996) study: assessors were blind to treatment
allocation and an impressive 89% of patients were followed up after one
year. However, we observed a number of potential limitations within the
study that may explain why Compliance Therapy in this trial failed to show
a positive treatment effect.
Of major concern is the issue of treatment fidelity. The therapists
working on the O’Donnel et al (2003) trial delivered both compliance
therapy (using a manual provided by Kemp et al 1996) and non-specific
counselling (not manualised). However, the amount of skills training that
therapists received to enable them to deliver both interventions is not
reported. Given that Compliance Therapy is a new intervention it seems
essential that therapists are appropriately and adequately trained;
evidence should have been presented that they are able to deliver the
intervention to a high standard with fidelity to the manual. It is routine
in trials of other cognitive behavioural interventions (Turkington et al
2002) to demonstrate that therapists are delivering interventions in a
consistent way that is true to the model of therapy that is being tested.
Typically therapists tape a number of sessions with patients that are then
independently rated using a measure such as the cognitive therapy scale
(CTS). As these data are not reported, it is not possible to demonstrate
whether therapists were actually delivering compliance therapy and,
secondly, whether the two (compliance therapy and non-specific
counselling) interventions were actually different in respect of the
therapeutic approaches used.
Our second concern relates to statistical power. The trial is powered
to detect a greater than doubling in the proportion of patients who were
fully compliant. However, it is unclear how the authors derived this
estimate of the effect that Compliance Therapy should have on compliance.
In the Kemp et al Compliance Therapy trial a more modest improvement in
adherence was reported but this was not used as a basis for the power
calculation in this trial. As a consequence of the low statistical power,
the 95% confidence interval of the estimate of treatment effect extends
from -37% (favouring control intervention) to a 15% increase in compliance
at one year, an increase which surely would be worth having.
Third, we would make an observation about the use of compliance as
the primary outcome and the validity of the measure used. In both studies
compliance is a main outcome: O’Donnel used a four point informant rating
of compliance compared with a 7 point informant rating of compliance used
by Kemp et al. The use of compliance as the primary outcome is ethically
contentious (McDonald et al 2002) as interventions to improve adherence
should only be judged by their clinical benefits to patients. These are
secondary outcomes and the trial is not powered to detect these
differences.
Finally, although there is not a statistically significant difference
between the number of patients who were fully compliant at baseline, there
were almost twice as many in the compliance therapy group (35%) compared
to the control group (19%). This may be an important difference that
limits the opportunity for compliance therapy to exert an effect.
Meta-analyses of interventions for treatment non-adherence in
psychosis support the conclusion that adherence can be improved (Nosé et
al 2003), although psychoeducational interventions without accompanying
behavioural components might not be that effective (Zygmunt et al 2002).
Compliance Therapy is a popular approach for improving adherence and a
replication of the original, largely positive trial by Kemp et al (1996)
is to be welcomed. However, there is need to undertake an appropriately
powered multi-centred definitive randomised controlled trial. Such a trial
must address the methodological weaknesses in both the previous compliance
therapy trials. Specifically, both interventions must be manualised with
the active differences between the approaches defined; therapists must
receive appropriate skills training in both compliance therapy and control
intervention, with a close monitoring of treatment fidelity throughout the
trial; the primary outcome should be clinical (eg psychopathology or
quality of life) without the absence of cost-effectiveness insights.
Finally, the trial should be appropriately powered to detect a realistic
clinical effect.
Such a trial is well under way. A four country randomised controlled
trial (the QUATRO project) to evaluate the impact of a revised version of
compliance therapy, called adherence therapy, started in January 2003 and
we expect to report findings in the Spring of 2005. Only such a trial can
truly answer the question of whether Compliance Therapy is an effective
intervention for enhancing treatment adherence in people with psychosis.
References
O’Donnell C, Donohoe G, Sharkey L, Owens N, Migone M, Harries R, Kinsella
A, Larkin C, O’Callaghan E. Compliance therapy: a randomised controlled
trial in schizophrenia. BMJ 2003; 327: 834-837.
Kemp R, Hayward P, Applewhaite G, Everitt B, David A. Compliance
therapy in psychotic patients: randomised controlled trial. BMJ1996; 312:
345-349.
Turkington D, Kingdon D, Turner T; Insight into Schizophrenia
Research Group. Effectiveness of a brief cognitive-behavioural therapy
intervention in the treatment of schizophrenia. Br J Psychiatry 2002; 180:
523-527.
McDonald HP, Garg AX, Haynes RB. Interventions to enhance patient
adherence to medication prescriptions: scientific review. JAMA. 2002; 288:
2868-79.
Nosé M, Barbui C, Gray R, Tansella M. Meta-analysis of clinical
interventions for reducing treatment non-adherence in psychoses. British
Journal of Psychiatry, 183, 197-206, 2003
Zygmunt, A., Olfson, M., Boyer, C.A. & Mechanic D. Interventions
to improve medication adherence in schizophrenia. American Journal of
Psychiatry, 159, 1653-1664, 2002.
Competing interests:
None declared
Competing interests: No competing interests