Cost-effectiveness of treatment of depression in primary care
On June 13 2006 Judith Bosmans successfully defended her PhD thesis “Cost-effectiveness of
treatment of depression in primary care” at the VU University Medical Center. In her PhD thesis
Bosmans compared four treatments for depression in primary care.
1.
A pharmacy-based intervention to increase adherence to antidepressants
2.
A disease management program, that consisted of screening, diagnosis and treatment of
major depression in elderly patients.
3.
Interpersonal psychotherapy delivered by mental health workers in primary care
practices for depressed patients aged 55 years and older.
4.
Standardised usual care by General Practitioners with and without antidepressants.
No differences
There were no significant differences in costs or effects between the groups receiving the study
interventions and the groups receiving usual general practitioner care. Therefore, the overall
conclusion of the thesis is to continue usual general practitioner care for depression in primary care.
However, many patients had not recovered at the end of the follow-up indicating that there still is room
for improvement in the usual general practitioner care for depressed patients. Future research should
provide evidence on how to improve care for depressed primary care patients.
A summary of the research
Depression is a common disorder in primary care and is associated with decreases in functioning and
well-being, high levels of disability, and high work absenteeism and health care costs. However,
depression often remains unrecognised and consequently untreated in primary care.
Considering the high prevalence of depression, the undertreatment of depression in daily practice, the
enormous economic burden associated with depression, and the scarcity of health care resources it is
important to have information on both the costs and effects of depression treatments. The objective of
this thesis was to systematically review the cost-effectiveness of psychotherapy and counselling for
depression in primary care and to asses the cost-effectiveness of various treatment programs for
depression in primary care.
In Chapter 2 the literature on the cost-effectiveness of psychotherapy and counselling for adult primary
care patients with depression is systematically reviewed. A computer-assisted search of MEDLINE,
EMBASE, CINAHL, PsycINFO, and the Cochrane Library identified 1580 potentially relevant articles.
Of these, two independent reviewers selected 6 studies to be included in the review. The
methodological quality of the included studies was assessed by two independent reviewers, who also
extracted the data from the studies. All studies scored positively on at least 13 of the 19 items
contained by the methodological checklist we used. Cognitive behavioural therapy was associated
with insignificant differences in both costs and effects in comparison with usual care. Interpersonal
therapy (IPT) was significantly more effective and expensive than usual care. It remains uncertain
whether couple therapy and IPT are cost-effective in comparison with antidepressant therapy. There
was no evidence for the cost-effectiveness of counselling in comparison with usual care or
antidepressant treatment. Since all studies had methodological shortcomings and were underpowered,
no firm conclusions on the cost-effectiveness of psychotherapy and counselling in primary care can be
drawn. Given the large economic impact of depression and the reluctance of many patients to use
antidepressants, there is a need for well designed and sufficiently powered economic evaluations of
psychological treatments that have proved to be clinically effective for depressed primary care
patients.
The efficacy of antidepressants in the treatment of depression has been demonstrated in randomised
trials. However, non-adherence to antidepressant treatment is common. Therefore, we evaluated the
cost-effectiveness of a pharmacy-based intervention to improve adherence to antidepressants of adult
primary care patients in Chapter 3. The economic evaluation was conducted alongside a randomised
controlled trial in 19 pharmacies in The Netherlands. Patients who came to their pharmacy with a new
prescription for a non-tricyclic antidepressant were invited to participate in the study by the pharmacist.
Patients were randomly allocated to education and coaching by the pharmacist or to usual care. The
clinical outcomes of the study were adherence to antidepressants and improvement in depressive
symptoms over 6 months. Adherence was measured using an electronic pill container (eDEM) and
improvement in depressive symptoms by the Hopkins Symptom Checklist (SCL). Costs were
measured over 6 months from a societal perspective using questionnaires. Differences in costs and
cost-effectiveness between the treatment groups were tested using bootstrapping. Seventy patients
were randomised to the intervention group and 81 tot the usual care group. There were no significant
differences in adherence (mean difference 2.1, 95% CI -5.6 ; 9.8), improvement on the SCL
depression scale (mean difference -0.15, 95% CI -0.54 ; 0.23), and total costs (mean difference €315,
95%CI -1922 ; 2416). The incremental cost-effectiveness ratio was €149 per one percent improvement
in adherence and €2550 per point improvement in SCL depression mean item score. The costeffectiveness planes showed that there were no significant differences in cost-effectiveness. In
conclusion, in patients starting treatment with antidepressants, there are no significant differences in
costs and effects between usual care and education and coaching by pharmacists. However,
adherence in the usual care group was rather high, which may have been caused by the use of an
eDEM to measure adherence.
Although major depression in elderly primary care patients is common, it often remains unrecognized
and undertreated in elderly people in primary care. Disease management programs, that usually
consist of screening, diagnosis and treatment, can improve quality of care and outcomes for patients
with depression. Chapter 4 describes the economic evaluation of such a disease management
program for major depression in elderly primary care patients in comparison with usual care. The
economic evaluation was conducted alongside a cluster randomised trial. All consecutive patients of
55 years and older visiting their GP were screened for depression using the Geriatric Depression
Scale (GDS). Patients with a GDS score of 5 or more, were further evaluated using the PRIMary care
Evaluation of Mental Disorders (PRIME-MD). Patients who were diagnosed as having major
depression according to the PRIME-MD were invited to participate in the trial. General practitioners in
the intervention group received training on how to implement the disease management program,
consisting of screening, patient education, drug therapy with paroxetine, and supportive contacts.
GPs in the usual care group were blind to the screening results. Treatment in this group was not
restricted in any way. Clinical outcome measures were improvement in severity of depression
(Montgomery Asberg Depression Rating Scale – MADRS), recovery from depression (PRIME-MD)
and quality adjusted life years (QALYs) gained (EuroQol). Resource use was measured over a 12
month period using interviews and was subsequently valued using cost prices. Bootstrapping was
used to evaluate the differences in costs and cost-effectiveness between the treatment groups. There
were no statistically significant differences in improvement in depression symptoms, recovery from
depression and QALYs between the intervention and the usual care group. Total costs were €1784 in
the intervention and €1898 in the usual care group (mean difference -€114, 95% CI: -€1003 ; €933).
Cost-effectiveness planes indicated that there were no statistically significant differences in costeffectiveness between the two groups. In conclusion, this disease management program for major
depression in elderly primary care patients had no impact on clinical outcomes, costs and costeffectiveness. Therefore, continuing usual care is recommended.
Many depressed patients prefer psychotherapy over antidepressants. Psychotherapy is mainly
provided in specialized mental health clinics, but many older depressed patients are reluctant to
accept mental health treatment in specialized mental health clinics. In the study described in Chapter
5, the cost-effectiveness of interpersonal psychotherapy (IPT) delivered by mental health workers in
primary care practices, for depressed patients aged 55 years and older who were identified by
screening, was evaluated. The economic evaluation was conducted alongside a pragmatic
randomised controlled trial. Patients who contacted their general practitioner (GP), were screened
using the Geriatric Depression Scale (GDS). Patients with a GDS score of 5 or more, were further
evaluated using the PRIMary care Evaluation of Mental Disorders (PRIME-MD). Patients who were
diagnosed as having major depression according to the PRIME-MD were invited to participate in the
trial. IPT patients were offered 10 sessions of IPT over a period of 5 months. GPs of care as usual
(CAU) patients were not informed which patients were included in the study. Clinical outcome
measures were the Montgomery Asberg Depression Rating Scale (MADRS), PRIME-MD, and the
EuroQol. Resource use was measured over a 12 month period using cost diaries and was
subsequently valued using cost prices. Missing cost and effect data were imputed using Multiple
Imputation. Bootstrapping was used to analyse differences in costs and cost-effectiveness. At 6
months there was a remarkable difference in recovery rate from depression according to the PRIMEMD between IPT and CAU, but not in the other outcome measures. At 12 months there were no
significant differences in clinical outcomes between IPT and CAU. The costs of providing IPT were on
average €656 per patient. Total costs at 12 months were €5753 in the IPT group and €4984 in the
CAU group (mean difference €769, 95% CI –2459 ; 3433). Cost-effectiveness planes indicated that
there were no significant differences in cost-effectiveness between the two groups. Thus, provision of
IPT in primary care to elderly depressed patients who were identified by screening, was not costeffective. Future research should focus on cheaper forms of psychotherapy, such as group therapy.
Although there is insufficient evidence for the effectiveness and no information on the costeffectiveness of antidepressants in patients with minor or mild-major depression, general practitioners
in the Netherlands regularly prescribe antidepressants to this group of patients. Prescription of
antidepressants in patients with minor or mild-major depression is only justified if they have additional
benefits over usual care without antidepressants. Therefore, the objective of the economic evaluation
described in Chapter 6 was to evaluate whether standardised usual care without antidepressants is
non-inferior to (i.e. at least as effective as and not more expensive than) standardised usual care with
antidepressants in patients with minor or mild-major depression. The economic evaluation was
conducted alongside a randomized controlled non-inferiority trial from a societal perspective with 52
weeks of follow-up. Adult primary care patients who had minor or mild-major depression (3-6 out of 9
depressive symptoms according to the Diagnostic and Statistical Manual of Mental Disorders (DSMIV)) were randomized to standardised usual care without or with antidepressant medication. Clinical
outcome measures were the Montgomery Asberg Depression Rating Scale (MADRS) to measure
severity of depressive symptoms and the EuroQol to measure quality of life and to calculate Quality
Adjusted Life Years (QALYs). Resource use was measured using cost diaries and was valued using
Dutch guideline prices. Differences in costs and cost-effectiveness were evaluated using
bootstrapping. For improvement in MADRS score the standardised usual care without antidepressants
group was non-inferior to the standardised usual care with antidepressants group at 6,13 and 52
weeks. For QALYs gained over 52 weeks non-inferiority could not be shown. Mean total costs were
€4668 in the standardised usual care without antidepressants group and €5418 in the standardised
usual care with antidepressants group (mean difference -€751, 95% CI -3601 ; 1522). Although noninferiority could not be shown in the costs, cost-effectiveness, and cost-utility analyses, costs in the
two treatment groups were very similar. Considering the non-inferior effects of standardised usual care
without antidepressants and the disadvantages of antidepressant use, we recommend general
practitioners in general to refrain from prescribing antidepressants in patients with minor or mild-major
depression.
For many diseases an effective treatment already exists. In these cases, the effectiveness of a new
treatment may be established by showing that the new treatment is as effective as (i.e. equivalent to)
or at least as effective as (i.e. non-inferior to) the old treatment. Chapter 7 gives practical guidelines for
economic evaluations conducted alongside such so called clinical equivalence or non-inferiority trials.
For an economic evaluation accompanying a clinical equivalence or non-inferiority trial, it is important
to decide before the start of the study on the appropriate research question. In many cases the
objective of the economic evaluation will be to show equivalence or non-inferiority of the costeffectiveness of the treatments. This has major implications for the design and analysis of the
economic evaluation. Most importantly, before the start of the study an acceptable difference in total
costs, θ, should be defined. θ has to be selected so that any economically unimportant difference in
costs is smaller than θ. To show equivalence of non-inferiority of costs, the confidence interval
approach is used. Equivalence is demonstrated, if the confidence interval around the difference in
mean total costs lies entirely between -θ and θ. Non-inferiority of costs is shown, if the upper limit of
the confidence interval is smaller than θ. The clinical (δ) and economic (θ) equivalence or noninferiority margins can be drawn in the cost-effectiveness plane. Equivalence with regard to costeffectiveness can be demonstrated, if 100-α% of the cost-effect pairs lies within the area within the
equivalence margins. Non-inferiority with regard to cost-effectiveness can be demonstrated, if 100-α%
of the cost-effect pairs lies right of the non-inferiority margin for effects and below the non-inferiority
margin for total costs. We recommend researchers to perform a sensitivity analysis in which θ is
varied.
Chapter 8 summarises the main findings and conclusions of this thesis. Also some methodological
considerations of the cost-effectiveness analyses included in this thesis are discussed. Arguments are
provided to support our conclusion that, since the interventions we evaluated in the studies included in
this thesis, were not more (cost-) effective than usual care, usual care should be the preferred
treatment. The chapter concludes with recommendations for clinical practice and future research.
Judith Bosmans, PhD
Amsterdam Free University
Faculty of Earth and Life Sciences
Institute of Health Sciences
De Boelelaan 1085
1081 HV Amsterdam
Room U-446
E-mail: judith.bosmans@falw.vu.nl