Clinical effectiveness of care managers in collaborative care for patients with depression in Swedish primary health care: a pragmatic cluster randomized controlled trial

Patients attending 23 different urban and rural PCCs, aged ≥ 18 years, diagnosed with a new (< 1 month) mild or moderate (according to Montgomery-Åsberg Depression Rating Scale- Self assessment (MADRS-S) < 35) depression (ICD-10 diagnosis F32, F33) [19] and not diagnosed with bipolar disorder, psychosis, addiction, or cognitive impairment were included. Those not speaking/understanding Swedish were excluded. Start of inclusion was December 2014.

Primary outcome was patient’s depressive symptoms (measured by MADRS-S [19, 20] and BDI-II [21]) at 6 months.

Secondary outcomes were patient’s quality of life (EuroQoL-5D 3 L scale, English tariff [22]), sick leave (days), percentage return to work (RTW), antidepressant medication, and patient satisfaction (from Psychiatric Outpatient Satisfaction Scale [23]) at 3 and 6 months.

Initial data collection included age, gender, socio-demographic and economic variables, alcohol consumption, physical activity, and ethnicity. In the 3 and 6 month follow-ups, sick leave status, RTW, data on medication, comorbidity and other treatment (psychological, counseling, or other) were collected. For somatic health reasons, blood pressure and p-glucose were monitored.

The 23 health care centers were stratified into two strata; rural (12 health care centers) and urban (11 health care centers). Each stratum was allocated into six blocks consisting of two health care centers, in which one was randomly assigned to implement the care manager function.

At the intervention PCCs, a nurse devoted around 20-25% of working time as care manager for management of care for the patients with depression. Before patient recruitment began, GPs and the nurse/district nurse (care manager) participated in sessions (for GPs 2 one-day sessions and for care managers 1 three-days session before start of intervention + 2 one-day sessions during initial part of intervention) for training in providing clinical services. During the study every PCC, both intervention and control PCCs, was regularly, at least once a week, visited by the research assistants (all specialized nurses) and in addition, solely for intervention PCCs, regular follow-up meetings were held every second month where all care managers participated and where difficulties, obstacles, and successes were discussed and dealt with together with the research team and the Region’s implementation team.

Clinical services for the care manager consisted of creating an individual care plan (1 h session with patient) and further telephone contacts between nurse and patient at least 6-8 times (around 15-30 min each) during 12 weeks, with person-centered communication around depressive symptoms based on the patient’s current depression symptom assessment with a self-assessment instrument in connection with the regular telephone call, as well as behavioral activation [19, 24]. Thus, the intervention group received care as usual plus the intervention. All patients could directly contact the care manager between the scheduled telephone contacts if they needed. The care manager had direct and regular contact with the General Practitioner (GP), therapist, or other PCC personnel who were involved in the care of the patient. The care manager did not include any type of psycho-therapy in her/his care of the patient, but supported the patient and increased the accessibility and continuity of the PCC’s care for the patient, coupled with organizational changes that would facilitate care for the patient with depression (see Table 1).

Participants at the control PCCs received care as usual (CAU) according to standard protocol and procedures. The Swedish National Guidelines for Depression and Anxiety Disorders recommend high accessibility and continuity, early next appointment, guided self-help, cognitive behavior therapy (CBT) (face-to-face or internet delivered), interpersonal therapy, and/or antidepressants as first and second steps in a stepped care model [3].

All patients ≥ 18 years seeking care at the PCCs and who were judged to have a probable diagnosis of new (< 1 month) depression in connection with a visit to a doctor, nurse, or therapist and not judged to have any of the exclusion criteria were asked and informed about participation. The patients who accepted to participate immediately received a visit to the PCC’s care manager (intervention PCC), alternatively to a research nurse (control PCC), for diagnosis confirmation and research baseline data purposes. The depression diagnosis was confirmed by use of PRIME-MD (depression module) in accordance with DSM-IV criteria for mild/moderate depressive disorder [25].

Standard statistical methods were used for descriptive statistics. Continuous variables were analyzed by independent sample t-test or Mann-Whitney U test and categorical variables or frequencies by Pearson chi-square test. Means of intra-individual change of depressive symptoms, and quality of life (QoL) scores were compared between the intervention group and the TAU group by using mixed model analysis with repeated measures. These analyses were adjusted for the type of PCC, age, sex, education, antidepressants at inclusion, and response variable at baseline.

The primary variable was the level of depression (as measured by MADRS-S and BDI-II) and analyzed in an ANCOVA model with repeated measures. In order to detect an effect of 3 units in the difference between the two groups, with a power of 80% and a significance level of 10% (two-sided), around 200 patients were needed in each group. The underlying assumption was a standard deviation in the group of 10 units, a within-subject correlation of 0.4, and a within-cluster correlation of 0.1, i.e. a design effect of 1.9 to correct for having a cluster analysis.

A total of 34 patients did not participate in the 3 and 6 months follow-up, 29 in the intervention group and 5 in the control group. Patients lost to follow-up were not reached despite several contacts by mail and telephone. There were statistically significant differences (p < 0.05) at baseline concerning age (non- participants: mean age around 10 years younger), MADRS-S (non- participants: mean value around 4 units higher) and BDI-II (non- participants: mean value around 4 units higher), and in the intervention group, a greater proportion of the non-participants were students. There were no other significant differences between participants and non-participants concerning demographic data described in Table 2.

The course of depression and QoL is shown in Fig. 2a, b, and c. There was a substantial reduction of depression scores both in intervention and control groups, but the reduction was significantly greater in the intervention group compared to control group when measured with MADRS-S, and the difference still progressed during the period 4-6 months, although the care manager intervention was terminated at 3 months. Mean depression score measured by MADRS-S was 2.17 lower (95% CI [0.56; 3.79], p = 0.009) at 3 months and 2.27 lower (95% CI [0.59; 3.95], p = 0.008) at 6 months. The QoL also showed a steeper increase in the intervention group from baseline to 3 months (statistically significant difference between intervention and control at 3 months, p = 0.01), but this leveled off at the 6 months follow-up. Depression score reduction measured by BDI-II did not reach significance. Mean depression score measured by BDI-II was 0.44 lower (95% CI [− 1.62; 2.50], p = 0.67) at 3 months, and 1.96 lower (95% CI [− 0.19; 4.11], p = 0.07) at 6 months.

Table 3 shows remission frequency, defined as MADRS-S ≤ 12, in intervention and control group at 3 and 6 months follow-up. There was a statistically significant higher remission frequency in the intervention group at 6 months follow-up, 67% compared to 47% in the control CAU group, also illustrating the progress of depression symptom score reduction during the 4-6 months period in the intervention group.

Use of antidepressants was present at inclusion in 53% and 63% of the intervention and control group patients, respectively, and continued at the same level in the intervention group (51%), but significantly increased to 67% in the control group at 3 months follow-up (see Table 3). However, at 6 months follow-up, the antidepressant medication frequency in the control group was somewhat reduced to 61%, while the antidepressant medication frequency in the intervention group still remained stable at 51%.

At inclusion, 53% of the patients were on sick leave, with no significant difference between intervention and control patients (50.5% and 55%, respectively) (Fig. 3). During the period between baseline and 3 months follow-up, 55% of patients in the intervention group (83/152) and 51% (88/172) in the control group were on sick leave (ranging between 25 and 100% sick leave); mean number of sick leave days was 69.2 (intervention) and 66.2 (control). During the 4-6 months, 40% (59/147) of the intervention and 42% (64/152) of the control group patients were on sick leave for 60 and 62 mean days, respectively (Fig 3). However, significantly more patients in the intervention group returned to work during baseline to 3 months follow-up (62% vs 43%, p = 0.028), as significantly more patients in the intervention group returned via part-time sick leave (in Sweden 25, 50, and 75%) (Table 3, and Fig. 3).

Patients were asked by a questionnaire distributed by the research personnel (postal questionnaire at the 6 months follow-up) about the helpfulness, perception of waiting time, information given, and whether they would recommend the treatment to family members and/or close friends [23]. There were statistically significant differences between intervention and control patients concerning recommending the treatment to close friends and relatives (Table 4), but on the whole most patients were satisfied with the care given at the PCCs.

The majority of all PCCs in a region with 1.6 million inhabitants were invited to this trial. The participating PCCs, constituting around 10% of all PCCs, were representative of the region as a whole, with both urban and rural PCCs scattered over the region. Thus, the participating PCCs can be regarded as representative for Swedish primary care (total number 1200 PCCs). The number of patients recruited by the PCCs was satisfactory for attaining the pre-determined level of statistical power. The results of this trial could consequently be generalizable and representative for Swedish primary care. The attrition rate was low, and through access to the electronic patient records, complementary data especially concerning medication and sick certification were also collected.

However, there were also limitations. Due to the comprehensive organizational changes that the establishment of a care manager entailed for a PCC, concealment of the intervention status of the PCC was not possible, and all personnel at the intervention PCCs were thoroughly informed about the aim of the study and the care manager organization. The 3 months assessment at the intervention PCCs was carried out by research personnel unknown to the patient, as it could have been a possible source of bias if the assessment was made by the local care manager. At the control PCCs, the 3 months assessment was administered by a specially trained research nurse. Further, we used both MADRS-S and BDI-II as measures of depression symptom outcomes, but only MADRS-S outcomes showed statistically significant differences between the intervention and control group, although BDI-II outcomes showed similar tendencies. However, MADRS-S is an instrument specially constructed to measure change in depression course, while BDI-II is an instrument developed primarily for measuring level of depression. The instruments are complementary and show good correspondence in primary care [26], but MADRS-S is more clinically applicable for primary care and more clinically relevant concerning measurement of depression level severity [26]. Another limitation is the follow-up duration, which only covered 6 months. Important health economic consequences concerning especially care consumption, sick leave duration, and RTW for mild/medium depressed individuals should preferably be evaluated within a longer time perspective. However, in a 6 month perspective, this PRIM-CARE RCT already has shown important significant effects concerning full recovery from depression and earlier RTW, despite lower antidepressant medication frequency.

The care manager’s function at the PCC is to combine patient contacts for increased accessibility and continuity for the patient as well as facilitate support for the depressed patient through organizational changes [11]. Swedish primary care has high quality concerning medical and psycho-therapeutic competence, which is shown by the relatively modest differences between the results in the intervention and control groups. However, cornerstones of primary care such as accessibility and continuity are not sufficiently met in Swedish primary care [27], and in that respect a care manager can make a difference, especially for the group of patients with depression and anxiety, who often have low access to care due to the symptoms of the disorder. Quality improvements in Swedish primary care should also strongly enhance non-psycho-therapeutic care components to support the patient’s recovery by facilitating the patient’s own course to remission [3, 14, 15]. Further, those individuals who do not improve or deteriorate in their depression course are earlier identified by the continuous care manager contact [14].

Recently, several trials and literature reviews have been published on collaborative care with care managers in primary care [16, 28, 29]. Studies show that mental illness often negatively affects other somatic conditions, and a care manager who coordinates the care by maintaining a close and regular contact with patients and aligns efforts for their individual needs is shown not only to generate improvement of the depression but also improvement of the physical health [16]. A recent systematic review of measures in primary care leading to improved RTW also showed, similar to our study, that adding a care manager to the PCC organization is one of few factors that increase RTW for the individual [9].

The distinguishing features of a collaborative care organization with a care manager in contrast to most other types of measures already undertaken in primary care are the provision of high accessibility and continuity for the individual but also the opportunity afforded for individuals to gain knowledge about their disease, as well as the increased ability to customize care interventions to the individual’s specific needs [1, 14]. This type of care organization also allows for a better engagement of the entire PCC in terms of resources and contacts with other health care levels and society in general and provides opportunities for the continuity of these contacts. Such an organization facilitates health care that is adapted to the patient’s specific needs over time in close collaboration with the patient [14, 28, 29] and thus allows for complexity, person-centered care, and interaction on multiple levels.