Does primary care level psychotherapy associate with later use of long-term psychotherapy – longitudinal register-based cohort study among Finnish occupationally active individuals Authors: Heli Järnefelt, Jaakko Airaksinen, Erkki Heinonen, Aki Vuokko, Ari Väänänen, Sanna Selinheimo Published: 2026 ABSTRACT AbstractBackgroundOne key question for current psychotherapy research is how limited resources should be used efficiently to optimise treatment pathways. We examined how primary care level (PCL) psychotherapy, a history of mental disorder, and sociodemographic factors are associated with the later probability of receiving long-term rehabilitative psychotherapy.MethodsThis longitudinal observational cohort study based on a register data from the Gaps in Mental Health-Related Work Disability and Treatment Outcomes (GapMind) project, which combines data from national registers (Statistics Finland and the Social Insurance Institution of Finland) and the largest occupational health service (OHS) provider in Finland. The study population comprised Finnish occupationally active individuals aged 15–65 (N = 1,261,320). Data on the use of PCL psychotherapy in the OHS between 2019 and 2020 were combined with data on age, gender, socioeconomic status, education, history of mental disorders (including psychotropic medication purchases, sickness absences, pensions, and use of OHS due to mental disorders), and somatic disabling comorbidities and four-model logistic regression analyses were conducted. The primary outcome was the use of long-term psychotherapy during 2020–2022.ResultsParticipation in PCL psychotherapy (OR 7.05–15.51) and a history of mental disorders (OR 1.67–4.25) were associated with a higher probability of receiving subsequently long-term rehabilitative psychotherapy. Higher age, female gender, and higher socioeconomic status were also associated with an elevated likelihood of engaging in long-term rehabilitative psychotherapy.ConclusionsParticipation in PCL psychotherapy, and various clinical and sociodemographic factors were associated with an increased probability of receiving later long-term psychotherapy. These findings may support healthcare systems in leveraging this information to better anticipate the need for long-term psychotherapy.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1186/s13033-026-00720-1.Keywords: Primary care level psychotherapy, Long-term psychotherapy, History of mental disorder, Sociodemographic factors Introduction Depression and anxiety disorders are leading causes of disability worldwide, significantly impacting working life through reduced performance, increased absenteeism, long-term disability, and early retirement [1–3]. A considerable body of research shows that effective psychotherapies are available for most mental disorders [4], and there is initial evidence that they have a positive effect also on work-related outcomes [5]. Accordingly, our earlier register-based studies have shown that psychotherapies may decrease work disability [6–8]. In relation to the societal and economic aspects, a key question for current psychotherapy research is how limited resources can be used efficiently. At least two theoretical and practical approaches have been suggested. One is the stepped care model [9, 10], in which most patients are considered to benefit from low-intensity treatments, and only those who do not respond are referred to more intensive treatment. The other is a stratified care model, in which treatment intensity is initially matched based on differing needs of patients, e.g., varying indicators of overall case complexity [11]. Disability, functional impairment, and comorbid symptoms, for example, have been shown to be associated with the persistence of symptoms after low-intensity treatment, and to have a cumulative negative effect on treatment response [12]. Various challenges in personality functioning, as well as limited capacities for psychotherapeutic work (e.g., lower capacities for self-reflection) [13–15] have also been shown to predict a greater need for long-term psychotherapy. Recently, stratified care has been shown to improve treatment outcomes at a modest additional cost [16]. The question of how much psychotherapy is needed to achieve clinically meaningful changes relates to the stepped versus stratified care models. According to the dose-response perspective, there is an optimal dosage of psychotherapy that remains consistent across patients [17]. In contrast, the Good-Enough Level (GEL) perspective proposes that this optimal dosage varies between patients [18]. Various demographic and socio-economic factors are also associated with the overall utilisation of psychotherapy. Studies indicate that individuals in more advantaged socioeconomic positions, characterised by higher levels of education, income, and occupational status, are more likely to utilise psychotherapeutic services [7, 19–23] and benefit from them [7, 24–26]. Some evidence suggests that a higher socioeconomic positions is also linked to a longer duration of psychotherapy [27], and that these socioeconomic differences persist even when the social security system covers the cost of treatment [20, 21, 28]. In addition, men and older adults have been shown to be less likely to receive psychotherapy compared to women and younger individuals [22, 29]. Rehabilitative psychotherapy is the most common form of rehabilitation of the Social Insurance Institution of Finland [30]. Between 2020 and 2022, it was granted to between 56,682 and 62,505 individuals annually. Use of rehabilitative psychotherapy in mental health problems has been relatively common in Finland likely because it is statutorily granted for all people aged 16–67 years who are at risk of becoming disabled from work or study due to a mental disorder [31]. Rehabilitative psychotherapy is usually long-term, ranging from one to three years, and it is granted for one year at a time. The Social Insurance Institution of Finland reimburses up to 80 sessions per year (usually one to two sessions per week) and 200 sessions over three years, enabling long-term, intensive psychotherapeutic treatment. Patients typically pay part of the costs of the rehabilitative psychotherapy themselves, as the grant does not cover all expenses. Rehabilitative psychotherapy may represent different psychotherapeutic approaches of which the three most common orientations in years 2014–2018 were psychodynamic and psychoanalytic psychotherapy, solution-focused psychotherapy, and cognitive psychotherapy [32]. Before rehabilitative psychotherapy, a minimum three-month follow-up assesses the effectiveness of first-line treatments (pharmacological treatment, short-term counselling and/or psychotherapy) for symptoms. If work disability persists, a psychiatrist may refer the individual to rehabilitative psychotherapy. Therefore, this treatment is not intended as a primary treatment for individuals with mental health conditions. In recent years, primary care level (PCL) psychological interventions and other lower-step treatments have become increasingly available in Finland [33, 34]. Occupational health services (OHS) has also started to offer PCL psychotherapy for employees. This trend corresponds with the hypothesis that low-intensity treatments will be sufficient for the majority of the patients, and these therapies will prevent the necessity for long-term psychotherapy, and thus save limited resources of mental health services. However, it is unclear how PCL interventions affect the subsequent need to seek and receive long-term rehabilitative psychotherapy. An earlier randomised trial showed that over half of those randomised to receive short-term psychotherapy (lasting 12–20 sessions) had sought significant long-term psychotherapy during the ten-year follow-up. By contrast, patients randomised to long-term psychotherapy used considerably less auxiliary treatment [35, 36]. This suggests that low-intensity psychotherapy did not meet the treatment needs of many patients. In this register-based study, we studied whether PCL psychotherapy in OHS, including the number of sessions (dosage), associate with the later use of long-term rehabilitative psychotherapy. As sociodemographic (age, gender) and socioeconomic factors (occupational status, level of education and income), and the history of mental disorders and related work disability (including psychotropic medication purchases, sickness absences, temporary disability pensions and previous use of OHS due to mental disorders) as well as somatic disabling comorbidity may be linked with the increased psychotherapy use, we assessed whether they modify the association between PCL psychotherapy and long-term rehabilitative psychotherapy among Finnish occupationally active individuals. In addition, we examined the interactions between the dosage of PCL psychotherapy, previous use of OHS due to mental disorders and the probability of receiving later long-term rehabilitative psychotherapy. Based on the stratified care and the GEL models, as well as on earlier studies, we hypothesised that those who have previously received PCL psychotherapy at OHS, and have a prior history of mental disorder, have an increased probability of seeking and receiving long-term psychotherapeutic rehabilitation. We hypothesised also that women, younger individuals, and those with more advantaged socioeconomic positions have a higher probability of receiving long-term psychotherapy. Materials and methods Study context: occupational health services In Finland, the Occupational Health Care Act [37] requires all employers to arrange OHS for their employees. Self-employed individuals are encouraged but not obligated to arrange OHS for themselves. In 2023, OHS was arranged for 91% of employees [38]. The legislation regulates the multidisciplinary service infrastructure, objectives, stakeholders’ rights and obligations, and the qualifications of OHS personnel. OHS is divided into statutory preventive services, which are aimed at preventing work-related illnesses and promoting health, and voluntary medical care, such as specialised medical services. The Social Insurance Institution of Finland reimburses 60% of the costs of preventive services and 50% of the costs of medical care expenses if these are offered by employers [39]. Every worker, regardless of their contract type, is entitled to free OHS. In addition, employers can voluntarily include other services. PCL psychotherapy is one of such services. Typically, a referral from a medical doctor or nurse specialised in OHS is required to access psychotherapy. PCL psychotherapy in OHS is non-manualised and can represent any therapeutic approaches. The maximum duration is regulated by the employer’s OHS contract (usually 1 to 20 sessions). Study design and population The study is a longitudinal observational cohort study based on a register data. We used data from the Gaps in Mental Health-Related Work Disability and Treatment Outcomes (GapMind) project, which includes all occupational health care patients from Terveystalo Plc, the largest Finnish OHS provider, from 2017 to 2022. From the entire OHS patient population of Terveystalo (N = 1,609,762), we had information on individuals belonging to working population and aged 15–65 who had complete data on their age, gender, socioeconomic status, education, medication purchases, sickness absences, pensions, previous use of OHS due to mental disorders, and disabling comorbidity, and also information if they had begun PCL psychotherapy between 2019 and 2020. Those with missing information were excluded from the analyses (N = 348,442). In addition, those who had received rehabilitative psychotherapy from 2014 to 2018 (N = 37,277) were excluded because there is a five-year waiting period before an individual can receive rehabilitative psychotherapy again. After the exclusions the overall study population was 1,261,320 individuals. The ethics committee of the Finnish Institute of Occupational Health approved the study in July 2022. Variables The primary outcome of this study was the use of long-term rehabilitative psychotherapy from 2020 to 2022. This variable was dichotomised as 0 if an individual did not receive rehabilitative psychotherapy and 1 if an individual received any rehabilitative psychotherapy during the follow-up period. Information on the use of rehabilitative psychotherapy was obtained from the Social Insurance Institution register. The use of PCL psychotherapy was derived from the OHS register of Terveystalo. We counted the number of PCL psychotherapy sessions individuals received in 2019 and 2020. These were then recoded as follows: 0 or no psychotherapy, 1–5 sessions, 6–10 sessions and 11 + sessions. This measure was used as a proxy for PCL psychotherapy dosage. Sociodemographic and socioeconomic characteristics. We obtained data on the following characteristics from Statistics Finland and used them as covariates in our analysis: age, gender (men/women), level of education, occupational status and income. Data regarding these characteristics was collected from the PCL psychotherapy onset year for those attending PCL psychotherapy and from 2019 for those not attending PCL psychotherapy. Level of education was classified as high (tertiary or higher education), intermediate (upper secondary education), or low (primary education or lower secondary education). Occupational status was categorised into five groups: (1) upper white-collar employee, (2) lower white-collar employee, (3) manual worker, (4) entrepreneur and (5) other (students, retirees, unemployed and those whose socioeconomic status is not known). According to Terveystalo register, those in the “Other”-category had OHS in 2019 or 2020 (their status might have changed during the year) and thus they were included in the primary analyses. Income was divided into four quartiles: (1) highest income, (2) second highest income, (3) second lowest income and (4) lowest income. Clinical characteristics. Psychotropic medication purchases, sickness absences, work disability pensions, and previous use of OHS due to mental disorders were used as proxy measures for the history of mental disorders. These were operationalised according to mental and behavioural disorders in the International Classification of Diseases, Tenth Edition (ICD-10). For each of these, we dichotomised the variables: 0 if participants had no register data on these variables during 2014–2018 and 1 if such data existed. Information on the previous use of OHS due to mental disorders was obtained from the Terveystalo OHS register. Information on purchases of psychotropic medications and sickness absences due to mental disorders (ICD-10 F-diagnoses) was obtained from the Social Insurance Institution registers. In Finland, all workers aged 16 to 67 are eligible for sickness absence compensation. If an employee has been with the same employer for at least one month, their employer will continue to pay their full salary from the first day of illness and for the following nine weekdays. After this period, the Social Insurance Institution begins to provide statutory sickness benefits, which gradually compensate for lost income. In this study, the number of these compensated sickness absences at baseline was used as one proxy variable for the history of mental disorders. Information on psychotropic medication purchases was derived from Anatomical Therapeutic Chemical (ATC) codes (N05–N06) [40]. Data on all purchases were sourced from the Finnish Prescription Register of the Social Insurance Institution. This register includes the dispensing date and the ATC code for medications reimbursed to Finnish residents in non-institutional settings. Since over-the-counter psychotropic medications are unavailable in Finland, the register information provides one sensitive measure of mental disorder history. Data on disability pensions due to mental disorders were obtained from the register of the Finnish Centre for Pensions. After a work disability has persisted for at least one year, the pension provider compensates for the loss of income with a disability pension. This pension may be granted either permanently or temporarily (as a fixed-term disability pension) and may be either full or partial. Information on disabling diseases other than mental disorders was derived from the Social Insurance Institution sickness allowance register. We counted the number of diagnoses involving compensated sickness absence among the individuals from 2014 to 2019 and coded the variable as none, one, two, three, or more diagnoses (based on the accuracy of the first two characters of ICD-10 diagnosis codes). This measure was used as a proxy for disabling-related comorbidity. National ID numbers, which are unique to each permanent resident of Finland, were used to link data from different registers. All data were pseudonymised before being made available to the researchers. Statistical methods First, we calculated descriptive statistics for baseline characteristics of those attending long-term rehabilitative psychotherapy and those who were not. Second, we examined how PCL psychotherapy predicted receiving long-term psychotherapy using logistic regression analysis. We ran four models, adding covariates each time: (1) a crude model with the use of PCL psychotherapy as the only independent variable, (2) we added sociodemographic and -economic variables to the model, (3) we further added variables related to health and work disability to the model, and (4) in the fourth model we added previous use of OHS due to mental disorders and also an interaction between the dosage of PCL psychotherapy (sessions) and previous use of OHS due to mental disorders. From the final model, we estimated the predicted probabilities of attending long-term psychotherapy using estimated marginal means, averaging over age and gender. In addition, we made sensitivity analyses where we excluded individuals under 24 years and “Other”-category of socioeconomic status from the data. This was made because there are uncertainties of the results regarding the work disability pensions among young people who have entered working life only recently. “Other”-category of socioeconomic status was excluded from the sensitivity analyses because we do not have accurate information on their primary status throughout the follow-up period. Analyses were performed using R 4.3.1 and the results were visualised using the ggplot2 package [41] for R (R Core Team, 2023). Results Table 1 presents the demographic and clinical characteristics of the study population who did and did not attend long-term rehabilitative psychotherapy. Of the overall study population (N = 1,261,320), 0,17% (N = 2,180) had received PCL psychotherapy in OHS in 2019 or 2020, while 2.70% (N = 34,078) had received long-term psychotherapy due to the risk of work disability caused by a mental disorder during 2020–2022. Of those who had received PCL psychotherapy, 442 patients (20.28%) participated in long-term psychotherapy as well. Table 1 shows that those who attended long-term psychotherapy were younger, more often women, their socioeconomic status was higher, and they had more often history of mental disorders (psychotropic medication purchases, and sickness absences, temporary pensions and previous use of OHS due to mental disorders) compared to those who did not attend long-term psychotherapy. Table 1Demographic and clinical characteristics of the study population who attended and did not attend long-term rehabilitative psychotherapy in 2020–2022Long-term rehabilitative psychotherapyNo (N = 1,227,242)Yes (N = 34,078) Demographic characteristics Mean age (SD, range)40 (13, 15–65)34 (10, 15–65)Gender – Men621,863 (51%)7,583 (22%)Socioeconomic statusUpper white-collar employee286,969 (23%)9,372 (28%)Lower white-collar employee415,301 (34%)12,810 (38%)Manual worker279,557 (23%)3,628 (11%)Entrepreneur45,741 (3.7%)956 (2.8%)Other199,674 (16%)7,312 (21%)EducationHigh445,746 (36%)17,922 (53%)Intermediate133,065 (11%)1,559 (4.6%)Low648,431 (53%)14,597 (43%)Income quartileHighest349,178 (28%)6,233 (18%)Second highest334,452 (27%)7,992 (23%)Second lowest317,906 (26%)9,703 (28%)Lowest225,706 (18%)10,150 (30%) Clinical characteristics Primary care level psychotherapyNone1,225,504 (100%)33,636 (99%)1–5 sessions622 (< 0.1%)112 (0.3%)6–10 sessions800 (< 0.1%)202 (0.6%)11 + sessions316 (< 0.1%)128 (0.4%)Range0–470–30Psychotropic medication purchases214,326 (17%)15,912 (47%)Sickness absences due to mental disorders64,700 (5.3%)6,206 (18%)Temporary pensions due to mental disorders10,515 (0.9%)2,332 (6.8%)Number of disabling comorbid somatic diseases0954,026 (78%)26,847 (79%)1193,025 (16%)5,127 (15%)254,937 (4.5%)1,454 (4.3%)3+25,254 (2.1%)650 (1.9%)Previous use of OHSa due to mental disorders58,736 (4.8%)6,163 (18%)aOccupational health services Demographic and clinical characteristics of the study population who attended and did not attend long-term rehabilitative psychotherapy in 2020–2022 Clinical characteristics Primary care level psychotherapy aOccupational health services Even after several adjustments, earlier participation in PCL psychotherapy (OR 3.84–15.51) was associated with a higher probability of participating in long-term psychotherapy, and greater the number of sessions, the higher the probability (Table 2, Models 1–4). Of the sociodemographic factors, older age (OR 0.63) and male gender (OR 0.31) were less likely to be associated with long-term psychotherapy. In addition, lower-level socioeconomic position in terms of occupational status (OR 0.51–0.93), education (OR 0.53–0.57), and income (OR 0.70–0.91) was associated with decreased odds of long-term psychotherapy (Table 2, Model 2). Previous psychotropic medication purchases (OR 3.93), and sickness absences (OR 1.67) and temporary disability pensions (OR 4.25) due to mental disorders were associated with an increased likelihood of participating in long-term psychotherapy, whereas the number of disabling comorbid somatic diseases had negligible or no association with the probability of receiving long-term rehabilitative psychotherapy (Table 2, Model 3). Table 2Results from the logistic regression analysis assessing factors associated with the probability of receiving long-term rehabilitative psychotherapy in 2020–2022Model 1 (N = 1,478,621)Model 2 (N = 1,261,320)Model 3 (N = 1,261,320)Model 4 (N = 1,261,320)OR95% CIOR95% CIOR95% CIOR95% CIPrimary care level (PCL) psychotherapy sessions in OHSa in 2019-2020 (Ref: 0 sessions)1–57.055.76, 8.545.524.46, 6.783.843.07, 4.764.953.80, 6.376–109.808.40, 11.387.796.61, 9.145.564.69, 6.577.506.16, 9.0911+15.5112.61, 18.9712.8310.30, 15.909.367.45, 11.7011.728.95, 15.24Age (by 10 years)0.630.62, 0.640.520.51, 0.530.520.52, 0.53Gender (Ref: women)Men0.310.30, 0.320.360.35, 0.370.360.35, 0.37Socioeconomic status (Ref: Upper w-cb employee)Lower w-cb employee0.800.77, 0.820.770.74, 0.790.760.74, 0.79Manual worker0.510.49, 0.540.500.48, 0.520.510.48, 0.53Entrepreneur0.880.82, 0.940.890.83, 0.960.900.84, 0.96Other0.930.89, 0.970.790.75, 0.820.780.75, 0.82Education (Ref: High)Intermediate0.570.54, 0.600.560.53, 0.600.560.53, 0.59Low0.530.52, 0.550.460.45, 0.470.460.44, 0.47Income quartile (Ref: Highest)Second highest0.910.88, 0.940.850.82, 0.880.820.79, 0.84Second lowest0.830.80, 0.860.850.82, 0.890.800.78, 0.84Lowest0.700.67, 0.730.770.74, 0.810.730.70, 0.76Previous psychotropic medication purchases (Ref: No)Yes3.933.83, 4.033.513.42, 3.61Previous sickness absencesc (Ref: No)Yes1.671.62, 1.731.421.37, 1.47Previous temporary pensionsc (Ref: No)Yes4.254.03, 4.494.213.99, 4.45Disabling comorbid somatic diseases (Ref: None)11.061.02, 1.091.051.02, 1.0821.061.01, 1.131.051.00, 1.123+1.020.93, 1.101.000.91, 1.08Previous use of OHSa (Ref: No)Yes2.232.15, 2.30PCL psychotherapy sessions (1–5) * Previous use of OHSa0.320.20, 0.51PCL psychotherapy sessions (6–10) * Previous use of OHSa0.260.18, 0.38PCL psychotherapy sessions (11+) * Previous use of OHSa0.330.20, 0.54aoccupational health servicesbwhite-collarcdue to mental disorders Results from the logistic regression analysis assessing factors associated with the probability of receiving long-term rehabilitative psychotherapy in 2020–2022 aoccupational health services bwhite-collar cdue to mental disorders Finally, we examined combined effects of previous use of OHS due to mental disorder and PCL psychotherapy dosage on the predicted probability of participating in long-term rehabilitative psychotherapy (Table 2, Model 4; Fig. 1). We found that previous use of OHS due to mental disorders (OR 2.23) increases the probability of participating in long-term psychotherapy. However, there was a trend that those PCL psychotherapy attendants who did not have previous use of OHS due to mental disorder had even higher likelihood of participating in long-term psychotherapy compared to those PCL psychotherapy attendants who did have previous use of OHS due to mental disorder (Fig. 1). The combined effects of previous use of occupational health services (OHS) due to mental disorder and the dosage of earlier primary care level (PCL) psychotherapy (number of sessions) on the predicted probability (including confidence intervals) of participating in long-term psychotherapy Our sensitivity analyses where we excluded individuals under 24 years and “Other”-category of socioeconomic status are presented in Supplementary Appendix 1. The results of the sensitivity analyses were consistent with those observed in the full sample (Supplemental Table 1 and 2, and Supplemental Fig. 1). Discussion In this comprehensive, register-based study of Finnish occupationally active individuals, we showed that participating in PCL psychotherapy was associated with an increased probability of receiving long-term rehabilitative psychotherapy. Among those who had received PCL psychotherapy in OHS, over a fifth of patients participated in long-term psychotherapy later on. According to the stepped care model, this could be interpreted to mean that the participation in low-intensity treatment is adequate for most of the patients, with only a proportion requiring more intensive care [9]. However, although an attractive proposition in policy terms, this conclusion would be premature. Firstly, a notable proportion of PCL patients ended up in long-term rehabilitative psychotherapy anyway, and secondly, it is not known whether these patients - if left untreated by psychotherapy in OHS - would have sought rehabilitative psychotherapy. Our results also confirmed the hypothesis that a history of mental disorder increases the probability of long-term psychotherapeutic treatment, thus supporting the stratified care model [11]. In our study, previous psychotropic medication purchases, and sickness absences, temporary disability pensions and earlier use of OHS due to mental disorders, were associated with an increased likelihood of receiving long-term psychotherapy. In line with our findings, disability and functional impairment have been recognised as factors associated with the persistence of symptoms after short-term treatments [11, 12]. These results suggest that these patients, over a fifth in this study, could be identified and referred to long-term psychotherapy earlier, in line with the stratified care and GEL models, which recommend considering treatment intensity at the outset [18, 42]. In our study, the number of disabling comorbid somatic diseases had no association with the probability of receiving long-term psychotherapeutic rehabilitation. Thus, even though comorbid diseases may decrease psychological resources [43], they did not appear to affect the ability to commit to long-term rehabilitation in this study. However, we must remember that our study reached only those comorbid diseases which had led to sickness absences over nine weekdays. Our results showed that the more PCL psychotherapy sessions a patient attended, the higher the probability that they would also receive long-term rehabilitative psychotherapy. There was also a trend that those PCL psychotherapy attendants who did not have previous use of OHS due to mental disorder had even higher likelihood of participating in long-term psychotherapy compared to those PCL psychotherapy attendants who did have previous use of OHS due to mental disorder. In other words, it seems that perhaps especially for patients without prior (or at least relatively recent) mental health issues, the need for lengthier or more intensive psychological intervention was more likely. Based on our results, especially longer PCL psychotherapy in OHS is the best indicator for long-term rehabilitative psychotherapy, and in many of these cases, an earlier referral for long-term psychotherapy might be more appropriate in both supporting recovery and saving resources in the long run than prolonging the PCL psychotherapy. In the interpretation of our results, we must remember that our results did not study how effective PCL psychotherapy was in improving mental symptoms, and that the availability of PCL psychotherapy for the individuals depended on their employer´s contract with the OHS. It should also be noted that the overall evidence for the difference between sufficiency of short- versus long-term psychotherapies on adult mental disorders is unclear, as few trials have compared them directly in a randomized design [44]. In a rare exception, that is furthermore relevant for having also been conducted in Finland, Knekt et al. [35, 36] found that especially patients randomized to a short-term therapy were likely to seek over a four-fold amount of additional, out-of-protocol sessions over the next 10 years. The general lack of such randomized studies may complicate comparing effectiveness and sufficiency of short- and long-term psychotherapies. Such long-term follow-up data is highly relevant also when evaluating the chaining of the psychotherapeutic interventions. As hypothesised, sociodemographic factors also influenced the probability of receiving long-term rehabilitative psychotherapy, and these associations remained even after adjusting for mental health status, which suggests that there are differences in service use between sociodemographic groups. Consistent with earlier studies [22, 29], older age and male gender were associated with a lower probability of receiving long-term psychotherapy. Mood and anxiety disorders are approximately twice as prevalent in women as in men [45], and are somewhat more prevalent among young adults than among older adults [46]. In addition, lower-level socioeconomic positions in terms of lower education, occupational status and income were associated with decreased odds of long-term psychotherapeutic rehabilitation, as have been shown in earlier studies [7, 19–23] and these factors have also been associated with benefiting less from the psychotherapeutic services [7, 24–26]. Multiple barriers may prevent individuals with low socioeconomic positions from seeking or completing a course of psychotherapy [47]. For instance, low-income jobs often offer little autonomy or control, creating practical barriers to participation. In addition, differing expectations and communication styles between mental health providers from higher socioeconomic backgrounds and lower-income individuals in need of care can create psychological barriers to engaging in psychotherapy. Participants in rehabilitative psychotherapy also must pay part of the costs themselves. This may create an economic barrier to participation. These results should be interpreted in the context of the Finnish mental health treatment structures, in which the use of statutorily granted rehabilitative psychotherapy has played a significant role. In addition, it should be noted that rehabilitative psychotherapy is only available to individuals who have previously received primary treatment for their symptoms. Therefore, it is evident that receiving PCL psychotherapy increases the likelihood of receiving long-term psychotherapy as shown also in other recent Finnish study estimating hazard ratio for rehabilitative psychotherapy following the initial psychological intervention [48]. In addition to the OHS, these primary psychotherapeutic services may be provided in the public or private healthcare sector [33, 34]. Strengths and limitations The strength of our study lies in the large, comprehensive sample of Finnish occupationally active individuals, combined with data from national registers, with no loss during the follow-up. Our study provides information about different factors associated with the likelihood of receiving long-term psychotherapy. These results could be utilised to optimise individual treatment pathways for those occupationally active individuals who have mental disorders. Our results can be generalised into healthcare contexts and systems where different intensities of psychotherapeutic services can be planned and delivered to match individual needs. Our findings should also be considered in light of study´s limitations. Overall, participation in PCL psychotherapy was relatively low in our population, and as stated earlier, its availability for employees depends on the OHS contract of employers. This may have resulted in differential access to PCL psychotherapy among socioeconomic position groups, as employers of individuals with a higher position may be more likely to offer psychotherapy services compared to employers of those with a lower position. According to Statistics Finland [48], in 2019 nearly 90% of working population worked as employees and the rest were entrepreneurs. Thus, it is likely that this study represents the employed population, whereas entrepreneurs (2.8% and 3.8% of our study population) were underrepresented. It must be noted that OHS coverage is low among sole entrepreneurs, who constitute approximately 70% of all entrepreneurs challenging the OHS study among them. We should also consider that some individuals may have applied for rehabilitative psychotherapy, but the Social Insurance Institute did not grant funding because the rehabilitation was not timely. This could be due to the individual having a substance abuse disorder or other comorbid conditions, for example, which was assessed to be in poor treatment balance. In this register-based study, we also lacked information about other psychotherapeutic treatments that individuals may have received from the public or private healthcare sectors. Furthermore, many other unmeasured factors may influence the probability of participating in long-term psychotherapy among occupationally active individuals. Finally, this cohort represents only Finnish individuals in the workforce and therefore does not include groups totally outside employment [49], such as the unemployed, homemakers, or those retired for medical reasons, which may limit the generalizability of the findings. Further research is needed to identify and address the most effective ways to meet the needs of different patients with mental disorders. In addition, studies should examine the long-term effectiveness of various psychotherapy pathways in alleviating symptoms and improving work ability. These studies should also consider how factors such as personality functions and suitability for psychotherapy, which are known to affect response to psychotherapy [13–15], moderate the likelihood of seeking and benefiting from short-term versus long-term psychotherapy. Conclusions Our results suggest that matching the needs of patients has the potential to improve the efficiency of psychotherapeutic resource allocation, and to save limited mental health service resources. Our study showed that receiving long-term psychotherapy is associated both with earlier use and dosage of PCL psychotherapy in OHS and with sociodemographic and socioeconomic factors and mental disorder history of individuals. In future, these findings highlight the need to develop psychotherapy implementation strategies and care pathways that address the needs of diverse clinical populations on time. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Acknowledgements No applicable. Author contributions Conceptualization SS HJ; Methodology JA SS; Formal analysis JA; Funding acquisition SS; Project administration SS; Visualization JA; Writing-original draft HJ SS JA; Writing review and editing: HJ, SS, JA, EH, AVu, AVä; All authors have read and approved the final version of the manuscript. Funding information This study was funded by the Social Insurance Institution of Finland (4/26/2022 to SS). Data availability Data in this study were used under a license approved by Statistics Finland and the Social Security Institution of Finland. The data are not publicly available. The Finnish Social and Health Data Permit Authority (Findata) coordinates the permissions for similar datasets, and they are available to researchers under permission (https://findata.fi/en/). Declarations Ethics approval and consent to participate The ethics committee of the Finnish Institute of Occupational Health approved the study in July 2022. No patients were involved in this study, and the manuscript does not contain individual person’s data. Consent for publication The final version of the manuscript has been approved by all co-authors and they have given consent for publication. Competing interests HJ works as a self-employed psychotherapist (50%) at c/o Terapiakeskus Aurelia, providing both short-term psychotherapy and long-term rehabilitative psychotherapy. JA works as an occupational psychologist and data-analyst at Terveystalo Plc. EH works as a part-time self-employed psychotherapist at Terveystalo Plc. AVU has no competing interests. AVÄ was funded by independent research foundations (The Finnish Research Impact Foundation, The Research Council of Finland). As part of the funding conditions, collaboration with a private company was required. In this case, the health service company (Terveystalo Plc) contributed to the research by providing a portion of the research data and participating in collaborative aspects of the study. SS works as a part-time (10%) self-employed occupational health psychologist at Terveystalo Plc. Neither the authors’ institutions nor Terveystalo Plc influenced the design, execution, analysis, or reporting of the research. The researchers acted independently and followed academic research ethics. The content and conclusions of the study are not directed at or approved by either the authors’ institutions or Terveystalo Plc. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Material 1 Data Availability Statement Data in this study were used under a license approved by Statistics Finland and the Social Security Institution of Finland. The data are not publicly available. The Finnish Social and Health Data Permit Authority (Findata) coordinates the permissions for similar datasets, and they are available to researchers under permission (https://findata.fi/en/).