|
Long-term effects of a collaborative care intervention in persistently depressed primary care patients |
Katon W, Russo J, Von Korff M, Lin E, Simon G, Bush T, Ludman E, Walker E |
|
|
Record Status This is a critical abstract of an economic evaluation that meets the criteria for inclusion on NHS EED. Each abstract contains a brief summary of the methods, the results and conclusions followed by a detailed critical assessment on the reliability of the study and the conclusions drawn. Health technology A multifaceted, stepped, collaborative care intervention developed for the management of persistently depressed patients was examined. The intervention targeted patients, physicians, and the process of care. Patients received a book and companion videotape. Sessions with psychiatrists and telephone follow-up were offered.
Economic study type Cost-effectiveness analysis.
Study population The study population comprised patients aged between 18 and 80 years who received antidepressant medication for the diagnosis of depression or anxiety. The patients were required to have at least four residual major depressive symptoms and a mean score of 1.0 or more on the symptom check list (SCL)-20 depression items, or less than four symptoms but with a mean SCL-20 score of 1.5 or more, or have recurrent depression (at least two prior episodes) or dysthymia. Patients were excluded if they scored 2 or more on the CAGE alcohol-screening questionnaire, were pregnant or currently nursing, or were currently seeing a psychiatrist. They were also excluded if they had recently used lithium or antipsychotic medication.
Setting The setting was primary care. The economic study was carried out in the USA.
Dates to which data relate The dates when the resource use and effectiveness data were collected were not provided. The price year was not reported.
Source of effectiveness data The effectiveness evidence was derived from a single study.
Link between effectiveness and cost data The costing was performed prospectively on a sub-group of patients included in the effectiveness study.
Study sample Power calculations were not reported. The patients were identified using Group Health Cooperative of Puget Sound (GHC) automated registration, pharmacy and visit data. Five weeks after the prescription of antidepressants, patients received an approach letter from their primary care physicians to participate into the study. At 8 weeks, team members contacted patients by telephone to identify those eligible. First- and second-stage interviews were carried out. Recruitment for a parallel relapse prevention study was performed simultaneously.
Of the 2,699 letters initially sent, 336 (12.4%) refused to be interviewed and 312 (11.5%) were either ineligible or could not be contacted. Therefore, 2,051 patients (76.1%) agreed to participate in the study. This group did not differ statistically in age and gender from those who refused to participate. Of this group, 272 (13.3%) were eligible for the study based on having 4 or more residual depressive symptoms, while 694 (33.9%) were eligible for the parallel relapse prevention study. Of the 272 eligible patients, 159 were included in the final study sample. The patients not included were comparable in age and gender to those who participated. Of the 694 patients eligible for the relapse prevention study, 69 were included in the current study based on an SCL score of 1.5 or more. Again, the patients not included were comparable in age and gender to those who participated. Thus, the final study sample comprised 228 patients, 114 in the intervention group and 114 in the UC group. The patients in the intervention group had a mean age of 47.2 (+/ 14) years and 67.5% were female. The patients in the UC group had a mean age of 46.7 (+/ 13.4) years and 81.6% were female.
Study design This was a prospective, randomised, blinded, clinical trial that was carried out at four primary care clinics of GHC and HMO in Western Washington. Randomisation was stratified by severity of disease based on the symptom check-list (SCL)-20. The patients were stratified into severe depression (SCL>2.0; n=79) and moderate depression (SCL 1.0 to 2.0; n=149) groups. Within each stratum, the patients were randomised in groups of 8. The patients were followed for 28 months. The rate of patients completing the follow-up evaluations was 91.7% at 1 month, 84.6% at 3 months, 84.2% at 6 months, and 75% at 28 months. A telephone survey team blinded to the patients' randomisation status assessed the outcomes.
Analysis of effectiveness Depression and function was analysed on the basis of intention to treat, while adherence was assessed using treatment completers only. The outcome measures used in the analysis were:
SCL depression score;
functional impairment, measured using the Sheehan disability scale, with a global score reported on a 0 to 10 Likert scale; and
adherence to adequate antidepressant medications.
The study groups were comparable at baseline in terms of age, education, employment, race, chronic diseases, depression characteristics, and history of dysthymia. However, there were significantly more female participants in the UC group. Several regression procedures were carried out to account for potential confounding factors, which were entered as covariates in the regression analysis.
Effectiveness results The regression analysis for SCL depression resulted in a significant 3-way interaction of time, treatment and strata. After stratifying by baseline severity, the SCL depression adjusted mean at the end of the follow-up period for the moderate depression group was significantly lower in the intervention group (0.88 +/- 0.52) than in the UC group (1.23 +/- 0.62), (p=0.004). The corresponding figures in the severe depression group were 1.16 (+/- 0.85) in the intervention group and 1.19 (+/- 0.72) in the UC group, (p=0.88).
The regression model for the Sheehan disability scale also showed that there was a significant 3-way interaction of time, treatment and strata. Differences between the intervention and UC patients did not reach statistical significance, regardless of disease severity, (p=0.27 for moderate strata and p=0.76 for high strata).
The regression model for adequate dosage showed that the 3-way interaction of time, treatment and strata was not significant. Only the interaction between time and treatment was statistically significant. In the severe depression group, 72% of intervention patients and 40% of UC patients were adherent to an adequate dosage of medication during the first 6 months of treatment, (p<0.01). A similar trend was observed in the second 6-month period, with 70% of intervention patients and 37% of UC patients being adherent with medication, (p<0.05). In the moderate depression group, the intervention patients were more likely to adhere to 90 days or more of adequate dosage during the first 6-month period (76% versus 46%; p<0.05). However, differences in the other periods were not statistically significant.
Clinical conclusions The effectiveness analysis showed that the collaborative care intervention only led to long-term improvements in depression symptoms in patients with moderate disease. The differences in functional measures were comparable.
Measure of benefits used in the economic analysis The health outcomes were left disaggregated and no summary benefit measure was used in the economic analysis. In effect, a cost-consequences analysis was carried out.
Direct costs Discounting might have been relevant since the costs were incurred in more than 2 years (28 months). However, no discounting was applied to the cost estimates. The unit costs were not presented separately from the quantities of resources used. The health services included in the economic evaluation were antidepressant medications, specialty mental health visits, primary care visits with mental health diagnosis, intervention visits, depression treatment, and inpatient care. Depression treatment covered primary care visits without mental health diagnosis, diagnostic tests, specialty medical visits, emergency visits, other medications and other outpatient services. The cost/resource boundary of the HMO appears to have been adopted. Resource use was estimated using patient-level data derived from a sample of 187 individuals (95 in the intervention group and 92 in the UC group), which provided complete data for all patients included in the effectiveness trial. The costs were derived from the health plan computerised system, and included personnel, supplies, and overheads. Costs incurred in the 6 months preceding the inclusion in the trial and the 30 months after randomisation were considered. Annualised costs were presented. The price year was not reported.
Statistical analysis of costs Linear regression analyses were performed to test for intervention and control group differences in costs after adjusting for covariates (e.g. age, gender, co-morbidities, and 6-month costs prior to baseline). The data were not stratified when the costs were analysed, owing to the small sample of patients considered. The costs were presented as mean values with 95% confidence intervals (CIs).
Indirect Costs The indirect costs were not considered.
Sensitivity analysis Sensitivity analyses were not performed.
Estimated benefits used in the economic analysis See the 'Effectiveness Results' section.
Cost results The total unadjusted annualised costs were $9,799 (95% CI: 7,763 - 11,834) in the UC group and $9,192 (95% CI: 7,504 - 10,880) in the intervention group.
The cost of the programme would be approximately $67 per patient, which should be added to the estimated cost of the intervention.
Differences in costs were not statistically significant for total ambulatory costs, depression treatment costs, non-depression-related outpatient costs, and total health care costs.
Synthesis of costs and benefits A synthesis of the costs and benefits was not relevant since a cost-consequences analysis was, in effect, performed.
Authors' conclusions Compared with usual care (UC), the stepped, collaborative care interventions led to long-term improvements in depression symptoms in patients with moderate persistent depression, despite the fact that better drug adherence was observed in intervention patients only in the first 6 months of treatment. The cost analysis suggested that the collaborative care intervention did not lead to a significant increase in costs in comparison with UC.
CRD COMMENTARY - Selection of comparators The selection of the comparator appears to have been appropriate because it reflected the standard practice at the authors' clinics. You should decide whether this is a valid comparator in your own setting.
Validity of estimate of measure of effectiveness The analysis of effectiveness was based on a clinical trial, which was appropriate for the study question. The authors provided several details of the method used to select the sample and the randomisation procedure. The study groups were unbalanced at baseline in terms of the gender distribution of the study participants. The potential impact of confounding factors was investigated by performing regression analysis and by stratification of the patients. The clinical study was based on intention to treat for some of the outcome measures. The authors showed that the study participants were not significantly different from those who refused to participate. The evidence came from multiple centres and the study sample appears to have been representative of the patient population. Blinded researchers assessed the outcomes, and this represents a further strength of the analysis. The length of and loss to follow-up were reported. These issues tend to enhance the internal validity of the analysis. However, power calculations were not performed and there was no evidence that the sample size was appropriate.
Validity of estimate of measure of benefit No summary benefit measure was used in the analysis because, in effect, a cost-consequences analysis was conducted.
Validity of estimate of costs The perspective adopted in the study was implicitly reported and all the relevant categories of costs were included in the analysis. However, details of the unit costs and quantities of resources used were not provided. Similarly, the price year was not reported, which limits the possibility of replicating the study and performing reflation exercises in other settings. Appropriate statistical analyses, to test the statistical significance of differences in costs between the groups, were performed. Regression analyses were also conducted to account for the potential impact of bias. However, only unadjusted and annualised costs were presented. The use of discounting would have been interesting since the costs were incurred during a long timeframe.
Other issues The authors made some comparisons of their findings with those from other studies. They did not address the issue of the generalisability of the study results to other settings. Sensitivity analyses were not carried out and the cost estimates were specific to the study setting. This reduced the external validity of the analysis. The study referred to patients with persistent depression and this was reflected in the authors' conclusions. The authors noted some limitations to the validity of their study. First, there was a predominance of Caucasian patients. Second, the intervention was multifaceted, thus it was unclear which component had the greatest impact on the clinical outcomes. Third, the intervention could be difficult to implement in primary care networks that did not integrate psychiatric services. Finally, the study was underpowered to detect statistically significant differences in the main outcome measures.
Implications of the study In terms of clinical practice, the results supported the use of a stepped, collaborative care programme for patients with persistent depression from both clinical and economic points of view.
Source of funding Supported by grants from the National Institute of Mental Health Services Division in Bethesda, MD.
Bibliographic details Katon W, Russo J, Von Korff M, Lin E, Simon G, Bush T, Ludman E, Walker E. Long-term effects of a collaborative care intervention in persistently depressed primary care patients. Journal of General Internal Medicine 2002; 17(10): 741-748 Other publications of related interest Hunkeler E, Meresman J, Hargreaves W, et al. Efficacy of nurse telehealth care and peer support in augmenting treatment of depression in primary care. Archives of Family Medicine 2000;9:700-8.
Lin E, Simon G, Katon W, et al. Can enhanced acute-phase treatment of depression improve long-term outcomes. A report on randomized trials in primary care. American Journal of Psychiatry 1999;156:643-5.
Indexing Status Subject indexing assigned by NLM MeSH Adult; Aged; Antidepressive Agents /therapeutic use; Continuity of Patient Care /economics; Depressive Disorder /therapy; Disease Management; Female; Health Care Costs; Health Maintenance Organizations; Humans; Male; Middle Aged; Patient Compliance; Primary Health Care; Time Factors; Treatment Outcome AccessionNumber 22002001876 Date bibliographic record published 30/04/2005 Date abstract record published 30/04/2005 |
|
|
|