Secondary data analysis on the duration of sick leave of depression patients in general practice and
psychiatric practices

In the area of incapacity for work (AU), according to Meyer and colleagues, various health insurance companies have noticed a significant increase in the importance of depression in recent years [7]. According to data from the current health report from the Techniker Krankenkasse (TK), there was a significant increase in absences due to psychiatric disorders by more than four fifths between 2006 and 2014. This absenteeism only stagnated at a high level in the years 2014 to 2017. An increase in 90% is reported between 2000 and 2017 [8]. Similar information can be found in the current health report of the German Employee Health Insurance Fund (DAK). DAK data shows that the duration of AU in mental disorders has continuously increased threefold over the last 20 years. In 2017, the number of days absent due to mental disorders increased, albeit only slightly, by 1.5% more than in the previous year [9]. With 16.7% of lost days, mental disorders rank second in terms of the volume of sick days. In the results of the report by the umbrella organization of company health insurance companies (BKK), mental disorders also rank second among the causes of AU with 16.6% of all AU days [10]. By far the most absent days were caused by depression [9, 10].

>> When considering individual diagnoses of mental disorders, unipolar depression accounted for 83.4 sick days and recurrent depressive disorder for 30.9 sick days [9]. People with depressive disorders have twice as many sick days compared to people without this diagnosis. The number of new disability pensions due to affective disorders more than doubled between 2000 and 2013 [2]. As with all psychiatric diagnoses, their share of sick leave cases is considerably lower than their share of sick days, meaning they are mostly long-term illnesses [9, 10]. Sick leave due to depression has rarely been investigated at the level of individual medical practices. The aim of this study was to determine the duration of sick leave due to depression and the influence of demographic factors and depression severity on the duration of sick leave.
method
database
This secondary data analysis uses data from the Disease Analyzer database (IQVIA). It allows anonymous access to a selected sample of private practices and patients [11]. The Disease Analyzer database data are generated directly from the practice's computer system via standardized interfaces. Patient data stored in the physician's computer system are transferred to IQVIA monthly. The data are encrypted before transfer and correspond in scope and level of detail to the patient's electronic health record [11]. Studies by Becher et al. [12] and Rathmann et al. [13] on the validity and representativeness of the Disease Analyzer database for use in pharmacoepidemiological and pharmacoeconomic studies have shown that the database can be considered representative and valid.
study population
The study selection process is shown in Figure 1. Patients aged 20 to 60 years with statutory health insurance, diagnosed with unipolar and recurrent depression of the known severity levels (F32.0-F32.3, F33.0-F33.3), and who received a sick note from a general practitioner (GP) or psychiatrist (PS) in 2017 were selected. This selection is based on the assumption that the individuals aged 20-60 years are employed if they do not have family insurance coverage or are retired. Employment is a prerequisite for the need for a sick note.

Statistical methods
The mean total duration of sick leave was calculated and analyzed by sex and age group. The influence of demographic variables on the risk of long sick leave (>42 calendar days) was determined using multivariate logistic regression (see Fig. 2), adjusted for age, sex, severity of depression, and medical specialty. Sick leave was limited to a maximum of 42 calendar days because, according to legal regulations regarding sick pay, after 42 calendar days the health insurance fund, not the employer, takes over payment [14].
Results
Patient characteristics
A total of 5,319 patients from psychiatric practices and 19,475 patients from general practitioner practices, for whom at least one sick leave certificate was documented in 2017, were included (Fig. 1). Patient characteristics are shown in Table 1. The mean age was 42.7 years. 58.91% of the patients were female. The severity of depression was 16.91% mild, 60.31% moderate, and 22.81% severe. Workplace problems were also documented for 5.31% of the patients.

Association between defined variables and the risk of long-term sick leave
Figure 2 shows the results regarding the relationship or association of sick leave duration with demographic and other relevant variables (e.g. severity of depression, workplace problems).
The median duration of sick leave was 19 days for general practitioners (GPs) and 32 days for psychiatrists (PSs). 28.61 patients in the GP group and 44.51 patients in the PS group had sick leave exceeding 42 days (OR: 1.75, p<0.001). Thus, the duration of sick leave granted by psychiatrists was significantly higher than that granted by GPs. The severity of depression can also be associated with an increasing duration of sick leave. This is reflected in ascending medians and percentage distributions. This is most pronounced in patients with severe depression. These patients exhibit the highest values, both in terms of median (29) and with 391 patients having sick leave exceeding 42 days (OR: 2.13, p<0.001). The risk of being on sick leave for more than 42 calendar days increases with the age of the patients observed. This was most pronounced in the 51-60 age group, with a mean of 37.5%.
The increase in sick leave days can therefore be associated with the increasing age of those affected (OR: 1.71, p<0.001). Patients with documented workplace problems also had the largest proportion of sick leave days exceeding 42 days (39.6%) (OR: 1.55, p<0.001).
discussion
As we were able to show in our study, the median duration of sick leave was higher when using primary care (PC) and when sick leave exceeded 42 days, compared to when using general practitioners (GPs) (p<0.001). Furthermore, we were able to show that an increase in sick leave is associated with increasing severity (p<0.001). Both findings can be associated with the severity of the depression. It should be noted that, due to structural factors, the care of depression in Germany, across all severity levels, is more frequently provided by GPs than by specialists [15]. However, it can be assumed that the diagnosis of depression by specialists is considerably more targeted and accurate, and therefore the actual severity is also diagnosed more clearly [16]. The more thorough diagnosis by specialists may, in turn, lead to a longer estimated recovery time and thus a longer period of sick leave. Furthermore, it is known that patients with increasing severity of depression experience a greater degree of impairment in their ability to work and their quality of life [15, 17, 18]. The fact that depression is a major cause of long-term sick leave was described in the study by Koopmans et al. [19]. Furthermore, Schneider et al. showed that the duration of sick leave correlates with depression [20].
The association with age proved to be statistically significant (p<0.001). This was most pronounced in the 51-60 age group. This is also reflected in the health reports of national health insurance funds [9]. The longer average duration of sick leave among older insured individuals is due to the fact that long-term sick leave increases with age and accounts for a considerably larger proportion of overall sick leave. According to DAK data, the values are highest among insured individuals aged 60+. This age group accounts for almost 601 TP3T of sick leave due to long-term sick leave lasting more than 6 weeks [9]. While older employees have fewer sick leave cases overall, their sick leave days increase significantly [9]. This finding is therefore consistent with previous results. No significant influence of gender could be determined in our study (p=0.248). In contrast, the TK health report indicates that in 2017, mental disorders ranked second among men in terms of their contribution to sick leave. Among female workers, however, mental disorders accounted for the most reported sick leave days in 2017 [8]. It is an internationally established finding that women are approximately twice as likely to suffer from depression as men [1, 6]. Additionally, data from the DAK health report show that women have a higher rate of sick leave than men in all age groups [9]. Furthermore, patients for whom a physician documented workplace problems were on sick leave for significantly longer periods (p<0.001). This finding is consistent with previous research. As data from the DAK health report show, a poor work environment, which leads to workplace problems, is, alongside high workloads, a key factor influencing sick leave [9], thus explaining our findings. Rau and Henkel demonstrated that the higher the work intensity, the greater the risk of depression.
to become ill [21]. Furthermore, it has been shown that the lower the available scope for action, the lower the social support in the workplace, and the greater the imbalance between demands and expenditure of received recognition, the higher the risk of suffering from depression [21]. Schneider and colleagues found in their study that HA explored psychological burdens and workplace-related stressors in 85 cases. Depression was reported more frequently in these patients [20].
This retrospective study has several limitations. The first limitation concerns the patient population. Only individuals with at least one sick leave certificate were analyzed. Many people with depression, especially mild depression, are not signed off sick or do not want to be. The influence of other comorbidities on the duration of sick leave was not investigated. Furthermore, information about the patients' occupations and details of their work situations is lacking. On the other hand, this study included a large number of patients from a representative nationwide database and analyzed demographic and clinical variables simultaneously, which is one of its strengths.
According to health insurance reports [8, 9, 10], mental disorders show the most striking trends and the most significant increase in sick leave. In today's world, working life is characterized by constant availability – due to modern information and communication technologies. This form of on-call availability for employees has a significant impact on psychological stress [22]. Against the backdrop of increasing psychological strain, the modern, stress-related working environment presents a challenge for both employees and employers, to which both groups must adapt. In addition to legally mandated measures for employers, such as occupational health and safety and workplace reintegration management, strategies for workplace health promotion (WHP), such as stress management and functional coping, must be developed and implemented within the framework of workplace health management (WHM). Systematic implementation of WHP measures within WHM can contribute to reducing sick leave in companies. To achieve a reduction in long-term sick leave, separate programs for older employees are needed. Age diversity and sensitivity play a role here, both on the part of employees and employers.