Table I.

Overview of the selected randomized clinical trials.

	Setting	Definition Fas (kind of counted contacts)	Population	Identification	Number of intervention /control patients	Intervention	Follow-up	Outcomes	Results
Simon8	Primary care clinic, 3 prepaid health plans in Midwest, Northwest and New England (USA)	Top15% attenders during 2 consecutive years – outpatient medical visits	Age: 23–63	Electronic data SCID: Pos MDD or MDD pos. last 2 yrs (=1475 pat) HDRS >14 163 General practices: Usual care: 81 Intervention: 82	Intervention: 218 Usual care: 189	Depression management programme (DMP): – 2 h training Evaluation contact Antidepressant medication (AD) Information material Treatment coordinator	1 year after randomization.	Depression-free days Costs	More depression-free days (229– > 182) More costs (+$51.84 per additional depression-free day)
Katzelnick9	Primary care clinic, 3 prepaid health plans in Midwest, Northwest and New England (USA)	Top15% attenders during 2 consecutive years – outpatient medical visits	Age: 25–63	Electronic data SCID: Pos MDD or MDD pos last 2 yrs (= 1475 pat) HDRS > 14 163 General practices: Usual care: 81 Intervention: 82	Intervention: 218 Usual care: 189	Depression management programme (DMP): – 2 h training Evaluation contact – antidepressant medication – information material – treatment coordinator	1 year after randomization	HDRS SF-20 score Use of antidepressant medication Attendance	Improvement HDRS (13.6–> 9.9 at 1 year) More use of AD (69.3% of DMP patients and 18.5% of usual-care patients with at least 3 prescriptions in 0.5 year) Better SF-20 scores for social function, mental health, gen. health perceptions More attendance in year after inclusion (+3, 2)
Katon10	Primary care clinics of HMO, Washington State (USA)	Top 10% attenders in 1 year for sex and age – ambulatory healthcare visits	Age: 18–75 –18 GPs out HMO –300 000 patients	>2 Years in practice Selection from electronic data SCL-R one standard deviation above mean –> 339 patients – 251 Accepted randomization	Intervention: 124 Control: 127	DIS by psychiatrist Interview by the psychiatrist with the GP present Jointly formulated treatment plan Written protocol of treatment for GP	1 year after randomization	Use of antidepressant med. (AD) Rate of anxiety/depression Use (psych) healthcare	More AD (+38%) No better psych. State No lower use of healthcare and costs
Olbrisch11	Primary healthcare for students: Florida State University (USA)	> 4 face-to-face contacts in first quarter of study year – outpatient medical visits	Freshmen, sophomores and juniors: students university health centre	400 students – 300 got letter – 129 agreed – 112 randomized Plus “no contact” group	Intervention: 34 Control: 40 No contact: 30	Brief educational programme (group of 3–8 students)	1 year after intervention	Number of contacts primary care Use of other healthcare	Lower use of primary care in short term. Convergence towards same utilization during follow-up No differences in number of visits to other healthcare providers
Christensen12	Primary care out-of-hours service, County of Northern Jutland (Denmark)	> 4 Out-of-hours contacts one year before inclusion – consultations, home visits and telephone calls	No age restriction	Consecutive patients Randomization per practice	Intervention practices: 83 GPs; 3500 patients Control practices: 93 GPs 4635 patients	Status consultation by GP Education of participating GPs Questionnaire patients Economic incentives GP	1 year after the intervention	Number of contacts with the out-of-hours service Daytime contacts with the GP; hospital admissions; visits to hospital outpatient clinics	No convincing effect

Notes: SCID = Structured Clinical Interview DSM; MDD = major depressive disorder; HDRS = Hamilton Depression Rating Scale; SF = social functioning; SCL-R Symptom Checklist – Revised; DIS = Diagnostic Interview Schedule.