The study has four important findings: 1) when people in the workforce consult the GP with stress, it is often multifactorial and not only work-related. 2) sick leave, counselling, referral to psychologist, and medication are frequent management strategies often applied together. 3) stress is not managed equally in men and women. Compared to women, stress in men is less often considered work-related, more often managed with prescription of tranquilizers, and less often with sick leave, the latter only associated with female GPs. Finally, 4) very few cases of work-related stress are reported to the Occupational Safety- and Health- Administration or referred to a department of occupational medicine.
Strengths and limitations
The data were collected by the involved GP assisted by EMR and were used for voluntary personal quality development incentivising proper recording. Most GPs were experienced users of the standardised questionnaire layout probably reducing recording errors (20). The stress identification search did not include prescription records, preventing falsely increased medication frequencies.
Diagnostic coding of mental disorders has been mandatory for Danish GPs since 2014, and to improve stress identification, the GPs were given a checklist with physical, cognitive, and behavioural symptoms of stress. However, stress has no unique diagnostic code and even though several codes were used in the search, some eligible patients are likely to be missing from the study, e.g. if coded with bodily symptoms not included in the search. Overinclusion of patients seems less likely since the GPs only confirmed the stress and included the patient in around ¼ of the cases identified by the search (11). Furthermore, the prevalence and age- and sex-distribution of patients with stress in the study matches the large labour force surveys in the UK and Denmark (1, 11, 21, 22), indicating a representative sample for countries where free GP services is associated with high frequency of contacts due to mental problems (14).
Except for the higher proportion of female GPs, the characteristics of the participants did not differ substantially from that of all GPs in the region. However, it is likely that the participating GPs were more interested in stress and may manage it somewhat differently than non-participants. Nevertheless, the associations found between stress management, patient sex, and the assessed cause of stress are likely to be generalizable.
The study included both ongoing and completed cases. Some cases unquestionably had additional management elements after the audit date. Thus, the percentage usages of the elements are underestimated. The inclusion of ongoing cases was necessary because stress cases in general practice do not have a recorded end date.
In Denmark, psychologist care is partially remunerated if the patient is referred from the GP with depression at all ages and anxiety only until the age of 38 years. This may incentivise GPs to diagnose more stressed patients with depression and explain why more patients with depression was referred to psychologist.
Comparison with existing literature
Many studies have investigated management of common mental disorders (CMD) in general practice (14), but very few studies report specifically on stress. In this study, more than half of the patients were sick listed. The GPs had no valid information about the duration of sick leave. However, other studies indicate that these periods are generally long. In the UK on average 25.8 days are lost per episode of stress-related sick leave (1), and 54% of sick leaves lasts more than 3 weeks (6).
The GPs provided counselling to nearly half of the patients regardless of sex, age, assessed cause of stress, and coexisting mental disorders. Details on the counselling were not obtained, but the fee for counselling requires at least three and no more than seven sessions. Popular therapy forms include cognitive behavioural therapy and problem-solving therapy, both modestly effective (23). In addition to the counselling, 39% of the patients were referred to psychologist while only eight percent had benzodiazepines. GPs in Ireland have been criticized for prescribing benzodiazepines to patients with CMD rather than using counselling (14). An OECD report based on data from 2005 found that 30% of primary care mentally ill patients in Denmark received counselling and no medication compared to only 10% in the UK (24). During the latest decade benzodiazepine use in Denmark has been halved (www.medstat.dk), and in this study almost half of the patients had counselling and no medication indicating a considerable shift in strategy which could maybe serve as an inspiration to other countries.
With regard to gender equality, the Global Gender Gab Report 2018 ranked Denmark number 13 and the UK number 15 out of 149 countries (25). Nevertheless, this study indicates that men and women with stress are not perceived of and managed equally in Danish general practice. This inequality is substantial especially considering that CMD-related sick leave accounts for around 57% of all working days lost to ill health (1, 13). Our findings of sex-differences regarding sick leave and medication points in opposite directions suggesting that they cannot be explained by sex-differences in case severity. Internationally, women have more sick leave than men even when equally ill and under similar work- and family- requirements (17). It is generally believed that women are more tolerant of other women being on sick leave than of men. Concordantly, the found sex-difference regarding sick leave was only found if the GP was a woman. A recent Norwegian study on attitudes towards sick leave found that sick leave was more tolerated in workplaces heavily dominated by either one of the sexes (26). Stressed women in general and in this study are often employed in the women-dominated service and health sectors while stressed men are less often employed in male-dominated workplaces (22). So, the excess sick-leave in women might be due to sex-difference in workplaces. However, it was found only when the GP was female. Further supporting the existence of gender differences among GPs regarding attitudes towards sick-leave, a Swedish study found that in general female GPs sick-list more patients than male GPs do (27).
Implications for research and/or practice
Even among people in the workforce, the majority of stress is not only work-related. The spectra of causes and treatment options are wide why it is pivotal that GPs are curious and enquire into multiple areas of life when choosing how to manage the individual patient. Despite the multifactorial causation of stress, sick leave is often prescribed. We need to investigate benefits and harms of this management strategy and to uncover how sick leave may best be tailored to the individual patient. Furthermore, coexisting psychiatric disorder is common which makes the clinical evaluation of the patient’s overall mental health important. Counselling may be beneficial but should be explored with regard to content and effectiveness of the actual treatment provided in general practice.
GPs should strive to perceive and manage stressed men and women equally so that choices regarding important elements of management do not depend on gender.