The study was a prospective longitudinal study, in which the effect of the programme in terms of return to work in the intervention group was compared to a control group. Data was collected at baseline, after four months, after one and two years. Independent variables were participation in the programme or participation in the control group, whilst the dependent variable was return to work. Inclusion criteria were age between 18 and 60 years, attachment to the labour market (few months' unemployment acceptable), a work-related stress induced adjustment disorder, and duration ≤ 1 year. Exclusion criteria were severe medical conditions, major mental disorders such as signs of organic brain disorders, substance use disorders, schizophrenia and bipolar affective disorders.
The intervention group
A total of 73 persons were referred to the Stress Clinic and were given the option of participating in a four-month intervention programme, followed by evaluations after one and two years. They were treated between November 2002 and November 2004. In total 63 persons completed the treatment, although seven did not attend the one-year evaluation. Data regarding their employment-situation was collected by phone or e-mail. Almost half of the 63 people who completed the treatment had been referred by their place of work, whereas an insurance company had referred 30%, and the unemployment benefit department of their local authority had referred the rest. As for the ten Stress Clinic referrals not included in the study, in three cases the initial interview resulted in a decision that the intervention should take place elsewhere due to long-term severe depression and only two of the referrals dropped out of the treatment during the first four months. The five remaining referrals were not offered a place on the intervention programme due to lack of attachment to the labour market or primary non-occupational causes of their stress condition. For the 63 people who completed the four-month intervention programme, the average number of sessions was six. In a few cases, the treatment was extended by a couple of months.
The control group
The control group consisted of the 34 referrals to the Clinic of Occupational Medicine, referred by their General Practitioner during the period from 1st January 2004 to 30th September 2004 on the basis of a stress-related illness. They all fulfilled the inclusion and exclusion criteria met by the intervention group. The control group subjects were given the same questionnaires as the patients at the Stress Clinic, and they had two sessions with a specialist in occupational medicine, the second four months after the first. The control group subjects were contacted by post one year after their first consultation, and again after two years, in order to identify their attachment to the labour market and to collect information on their symptoms. Data was successfully collected from everyone in the control group with regard to attachment to the labour market; but only 28 completed the questionnaires (82.4%).
We did not adhere to any specific psychological or psychiatric form of therapy. We followed the factors identified by Frank and his group [10, 11] which are common across the different treatment concepts such as (a) a plausible rationale for the patients' symptoms, (b) a treatment plan, (c) a therapeutic atmosphere of a caring and hope-inspiring relationship.
The programme included:
Before the initial interview, the participants filled out the following questionnaires:
Basic information regarding social conditions, exercise and health
The Stress Clinic General-wellbeing questionnaire, based on a questionnaire on stress symptoms developed by the National Institute of Occupational Health (http://www.ami.dk) and the general section of the SF-36, which deals with self-rated health domains 
The WHO depression questionnaire "Major Depression Inventory" (MDI) [12, 13]
The questionnaires were discussed with each participant prior to the initial interview. Based on the WHO depression questionnaire MDI, it was established whether or not the participant had a depression. Patients with an MDI score higher than 21 were referred to the Stress Clinic's psychiatric consultant. The consultant and the participant would agree on subsequent psychiatric treatment.
Depending on the anamnesis, a clinical medical examination was carried out; supplemented by para-clinical serological tests, x-rays or further examination when indicated when indicated.
Stress handling sessions
During a four month period, each participant had at least four sessions, each lasting 1-2 hours, where the therapists tried to convey to the participants an understanding of stress-inducing factors, the participants' own stress-level and possible ways of reducing stress, both in relation to work and private life. The participants were given homework to do between each session, e.g. listing tasks for the next six months, prioritising planned tasks or writing down stressful events, work-related as well as private. The homework was discussed each time at the subsequent interview.
Thus the stress handling sessions had the following objectives:
To make the participants aware that there condition imposed a certain level of stress on their bodies. The participants were informed about the transitive nature and positive prognosis of their condition with no increased risk of setback if the treatment were followed. The participants were also informed that the treatment would take several months.
To diminish the strain which had caused the current poor level of functioning in the participants, e.g. through:
Change of workplace
Change of job tasks
Reduced number of working hours
Sick leave, possibly on a part-time basis
Personal social networks were involved in the process. The participants were encouraged to discuss their treatment with their spouse or other family members as well as potentially stress-reducing measures which relatives could assist with. To contribute to this engagement, the participants received their updated record after each session.
During the sessions, the participants acquired tools to help them handle everyday stress-inducing incidents such as traffic, children's behaviour, etc. Relaxation exercises and breathing exercises were introduced.
The clinical significance of relaxation was emphasised. The participants were given a CD with a 15-minute relaxation programme, and all were encouraged to follow it every day for the duration of the treatment. The relaxation programme taught the participant to relax through concentrating on various parts of the body, one at a time, guided by the instructions on the CD. A few were also taught breathing exercises for use in panic attacks.
Participants were encouraged to exercise at least twice a week. In order to evaluate the effect of the exercise, the participants' blood pressure and maximum oxygen uptake while on an exercise bike were measured at the start of the intervention and after four months. The examination was carried out at the Clinical Physiological Department at Hillerød Hospital.
The participants were given the book "Stress", published by Denmark's Radio in connection with a series of health programmes . The objective was that the participants should use the book to refresh their memories regarding the information they were given during the sessions.
Contact with the workplace
During the intervention, the participants' place of work was contacted if adjustments to their tasks or responsibilities were needed. Such contact was only made if the participant agreed. One or more meetings would usually be held between the participant, the author of this article and the participant's supervisor in order to discuss a possible change in work allocation at the workplace. Finally, the participants were encouraged to let their work place know how they experienced their situation and the factors, which had brought it about.
The Stress Clinic target group were working people with a work-related long-term stress situation. The media were informed about the Stress Clinic and its website (http://www.stressklinik.dk), on which the treatment programme was described, together with information as to referral as well as advice about stress treatment. Treatment was paid for by the referring body, which could be an insurance company, a local authority, the person's employer, or - in one instance - the participant himself/herself. Each referral to the programme was evaluated in order to decide whether it would be appropriate to set up an initial interview. If so, a contract was sent to the referring body, and once the signed contract was received, the participant was invited to an initial interview. In some cases, further information was requested.
The initial interview lasted 1-1½ hours, and was used to assess whether there was a realistic chance that the participant would benefit from the treatment at the Stress Clinic. Inclusion criteria were labour market attachment, and stress symptoms related to working conditions. Exclusion criteria were major psychiatric disorder or other ongoing psychological or psychiatric treatment. A specialist in occupational medicine conducted this interview and the following sessions. In cases of severe depression, or where the person already participated in some kind of treatment, it was agreed with the participant that we would refer to psychiatric or other relevant treatment.
The study protocol was reviewed by the Committee System on Biomedical Research Ethics and found to be a quality development project not covered by the committee system based on the 'Guidelines about Notification of a Biomedical Research Project'. As a consequense of this informed consent was not nescesary.
Chi2-tests were used for description of the dichotomous data and t-tests for continuity data regarding baseline symptoms.
Furthermore, the odds ratio (OR) for a return to work was calculated using logistic regression analyses for the participants compared to the control group, both unadjusted and adjusted for relevant confounders: age, gender, MDI score and occupation. Finally ORs for possible predictive factors for RTW adjusted for the intervention were calculated. Statistics were calculated using SPSS, 13th edition.
The level of statistical significance was P ≤ 0.05.