In recent years, mental health has become increasingly significant from both medical and socio-economic perspectives (1–3). As health care systems worldwide face growing financial pressures, newly introduced treatment approaches must be both effective and cost-efficient. This is highly relevant for the Swiss healthcare system, where according to recent estimates, 14.3% of costs are due to mental illnesses (4). Home Treatment (HT) has been discussed as an alternative to inpatient treatment for decades (5–7). It has now been implemented in various health care systems throughout the globe (8–10). Multiprofessional teams visit the patient at home on up to six days per week and offer a 24-hour availability for emergencies. These teams consist of doctor, psychologist, social workers, nurses and occupational therapists. This form of treatment is intended to be equivalent to inpatient treatment while offering certain advantages, such as reduced stigma towards psychiatric care, better involvement of the patient’s social environment, and accessibility for specific patient groups like caregivers of children.

In 2015, HT was introduced at the Psychiatric Services Aargau AG (PDAG) accompanied by an initial 12 month long randomized controlled trial (RCT) (11), examining to what extent the HT service would reduce hospital bed use. Patients were randomized into the service models including optional HT or to a conventional service model with inpatient treatment only. With HT involvement, the study showed a reduction of inpatient treatment within 24 months after the index crisis necessitating hospital admission in the HT group by over 30% with similar treatment outcome measured among others by the Health of the Nation Outcome Scale (12). Treatment costs in the HT group remained lower, however, not significantly, compared to the control group.

A follow-up study (13) examined whether patient characteristics and outcomes changed after HT was implemented into routine care, allowing patients to access this form of treatment without unnecessary barriers such as randomization. While a reduction in hospital bed use was still observed, it was less pronounced than during the RCT, and overall treatment duration increased. Both studies also compared demographic and clinical profiles between HT and inpatient groups, providing insights into which patients might benefit most from HT. Compared to patients treated within the RCT framework, those receiving home treatment under routine care more often lived with others, had less often been admitted compulsorily to acute care and stress-related and anxiety disorders were more prevalent. There have been other studies on the implementation of HT (14, 15) at other Swiss hospitals.

In the first two years of HT (2015 – 2017) at the PDAG, patients were admitted to HT either after walk-in at our emergency department or being referred to HT exclusively from inpatient treatment within the PDAG. Since 2017, referrals to HT can be initiated directly by outpatient practitioners or an inpatient ward of PDAG. Thus, we suppose that the diversification of admission pathways may influence treatment trajectories, associated costs, and patient characteristics. The aim of this study was to evaluate whether HT, after being integrated into routine psychiatric care at PDAG for several years, continues to reduce inpatient treatment costs in real-life conditions. We hypothesized that inpatient treatment durations will decrease in all HT groups compared to their respective inpatient control groups. Additionally, we anticipate that different treatment pathways might affect the extent of cost reduction. Based on previous findings (11), we expect that HT combined with inpatient treatment will not yield significant cost advantages due to increased cumulative treatment durations. Finally, we expect that direct admission to HT will reduce overall treatment costs and result in shorter treatment durations compared to standard inpatient care.

The study was conducted at the Psychiatric Services Aargau AG (PDAG), one of the largest psychiatric clinics in Switzerland and the only facility in the Canton of Aargau providing mandated emergency psychiatric services (16). The study was designed as an observational study, including patients who received home treatment (HT) in the years 2019 or 2020. Patients were followed for two years, with the study concluding at the end of 2022. During this time, the number of inpatient and HT episodes and the associated costs were analyzed. This data was compared with a matched control group, based on age, gender, primary diagnosis, and the number of psychiatric comorbidities, which received solely inpatient treatment.

Patients’ sociodemographic, clinical data and service use were drawn from electronic medical records and from the case register using the hospital information system ORBIS NICE (Version: DACHL_08042700; Dedalus Healthcare group), which is routinely used in the PDAG. Data of all patients being admitted between January 1^st, 2019, and December 31^st, 2020, were analyzed over a follow-up period of two years. The first admission between January 1^st, 2019, and December 31^st, 2020, served as the starting point for the two-year follow-up, and provided sociodemographic data as well as the principal diagnosis. Patients with dementia, delirium, and substance use disorder (as a principal diagnosis) were excluded from HT and the study. Participants’ ages ranged from 18 to 66 years, and we also excluded those receiving services covered only by Swiss private or supplementary insurance. In line with the clinical care structures at PDAG, inpatient care during the study period was provided in psychiatric acute care units and in diagnosis-specific wards, ensuring differentiated treatment pathways for various patients’ needs.

The outcome measures were the treatment costs, the total amount of treatment days, and number of readmissions during the two-year period. In line with Swiss DRG (Diagnosis-Related Groups) guidelines (17) and in accordance with Swiss hospital billing regulations, a readmission within the first 18 days after release was not counted as a readmission but as the same stay. The overall costs were calculated by multiplying the number of inpatient treatment days by the daily DRG base rate (CHF 695, approximately EUR 720) or by multiplying the number of HT days by the HT rate, which was negotiated between health insurers, the canton, and PDAG (CHF 410, approximately EUR 425). Finally, the costs of inpatient and home treatment over the two-year period were added up.

The regional ethics committee determined that this research project does not fall under the Human Research Act, as it only evaluated anonymized clinical routine data. However, the committee confirmed that the project meets general ethical and scientific standards for human research.

Among the 392 patients in the study, 306 (78.1%) were female, and 312 (79.6%) held Swiss citizenship. The median age was 39 years (IQR = 32 – 50, range = 18 – 66). The most common primary diagnosis were mood (64.8%) and anxiety disorders (19.9%). Most patients lived in shared households (67.3%), and 38.5% were married. Table 1 presents the sociodemographic and clinical characteristics of the different study groups. Matched characteristics such as gender, age, primary diagnosis, and number of secondary diagnoses did not show significant group differences. Table 2 provides a comparison of treatment costs and cost-relevant characteristics. Comparison between both HT subgroups is summarized in Supplementary Table S1 (in supplements).

Consistent with previous studies (20, 21), our findings indicate that, when implemented in routine clinical practice, home treatment (HT) does not result in a significant cost advantage over inpatient care. Specifically, HT patients experienced fewer inpatient days but had longer total treatment durations, which offset potential cost savings. Overall, costs did not differ significantly between groups. A similar pattern has been reported in studies on psychiatric day hospital treatment, another alternative to inpatient care, where greater symptom improvement and treatment satisfaction were observed, but treatment durations and associated costs were generally higher in the day hospital group (22).

This prompted us to further explore whether different trajectories of HT application have distinct effects on treatment costs. To this end, we conducted a cost analysis across subgroups, differentiating between patients who received HT as a primary treatment (HT - 1) and those who transitioned to HT following an initial inpatient stay (HT - 2). The results suggest that combining inpatient care with subsequent HT, as in the HT - 2 subgroup, leads to significantly longer cumulative treatment duration and, consequently, higher overall treatment costs. These results are in line with previous reports from the subsequent follow-up study after HT was implemented in routine care (13), where an increase in total treatment days was observed during the second year, after the initial RCT (12) was completed and HT was implemented into routine care. Recent studies primarily examined how HT as a main treatment could reduce inpatient capacity usage (8, 20 – 22), potentially lowering costs. The results from the HT - 1 subgroup support these findings. In contrast, combining inpatient care with subsequent HT, as in the HT - 2 subgroup, resulted in significantly longer cumulative treatment duration and, consequently, higher overall treatment costs. This pattern diverges from the findings of Stulz et al., who, in the initial RCT, reported a 30% reduction in inpatient days, leading to a shorter total treatment duration. Our current results suggest that, although HT holds potential for reducing inpatient stays, this effect is challenging to sustain in routine clinical practice, particularly when HT follows an initial inpatient treatment. However, HT as the primary mode of care (HT - 1) resulted in an 87.5% reduction in inpatient bed utilization and a 24% median cost reduction compared to controls over the same two-year period. Contrary to our initial hypothesis, which suggested that the HT - 1 group would have a shorter overall treatment duration, the total treatment duration (sum of all inpatient and HT days during the two-year follow-up) was non-significantly longer than that of the solely inpatient-treated controls. However, it was accompanied by a significantly lower readmission rate, suggesting potential long-term stabilizing effects of home treatment when applied early and proactively.

Patient characteristics and contextual factors may also play a critical role in explaining the observed differences in treatment duration and costs. In particular, the proportion of involuntary admissions differed notably between the subgroups. Involuntary admission may indicate a severe level of illness due to acute risk to oneself or others. In this study, the HT - 1 subgroup had significantly fewer involuntary admissions than the HT - 2 subgroup. Along with a higher readmission rate in the HT - 2 subgroup, this suggests that HT - 1 patients may exhibit more favorable illness characteristics and therefore require shorter treatment durations. This interpretation aligns with Mötteli et al., who reported that lower symptom severity and current employment status are associated with a higher likelihood of successfully substituting inpatient treatment with HT (23). Additionally, legal formalities associated with involuntary admissions may potentially prolong inpatient treatment in certain cases. However, Swiss and cantonal regulations mandate that the criteria for involuntary treatment be reviewed daily. If these criteria are no longer met, patients should be discharged upon request, which is likely to reduce the potential prolongation mentioned earlier. Overall, these results support the need for subgroup differentiation, reinforcing that HT - 1 patients exhibit distinct characteristics compared to those in the HT - 2 group.

While patient characteristics and legal factors are important in understanding treatment duration and costs, from a clinical perspective, the transition from inpatient to outpatient treatment presents additional challenges and considerations. Sometimes, it can lead to a renewed worsening of symptoms and a possible readmission after initial discharge (24, 25). For these cases HT enables an intensive form of treatment, at the same time allowing patients to become reacquainted with their home setting. Our data show that this clinical practice does not necessarily offer significant cost advantages. But the involvement of HT could mean that the patient can be discharged earlier, thereby freeing up capacities in inpatient units. However, it must be considered that each individual case needs to be assessed to determine whether such an intensive form of post inpatient treatment is necessary or if other, less intensive outpatient treatment options would suffice. If feasible, a transfer to HT should be considered shortly after hospital admission.

Building on the clinical considerations discussed earlier, operational factors during the transition from inpatient care to home treatment, such as the change in treatment team, may also influence treatment duration and costs. Establishing a new therapeutic relationship requires time and may delay treatment progress. In this context, ensuring therapeutic continuity is not only essential for clinical outcomes but could also have economic implications. Repeated patient admissions and disruptions in care processes tend to prolong treatment episodes and increase overall costs, particularly as the initial days of hospitalization are associated with the highest per diem costs due to more intensive ancillary services (26). Therefore, an integrative approach, where the patient is continuously treated by the same team (both inpatient and HT), might potentially reduce treatment duration and thus lower the costs. This emphasis on team continuity aligns with the Open Dialogue approach, which promotes consistent therapeutic involvement across all phases of care. By maintaining a stable treatment team, Open Dialogue addresses fragmented care pathways and supports more efficient treatment processes in psychiatric care settings (27).

This study also confirmed the HT-patient characteristics previously described in other studies, one of them being mood disorders as the most common primary diagnosis (28). Secondly, while matching ensured, that both HT and control samples had an identical number of women, the HT sample characteristics indicate that women utilized HT more frequently (78% vs. 22% men), aligning with previous studies (23). This may be attributed to the frequent referrals of female patients with postpartum depression and psychosis, for which HT is well-suited. Anxiety disorders are more prevalent in women, with studies indicating that they are approximately twice as likely as men to experience these conditions (29). This higher prevalence may contribute to the increased referrals of anxiety disorders, as HT appears well suited for cognitive behavioral based interventions and for in vivo exposure embedded in the patients everyday environment (30).

As noted by Stulz et al. (13), a change in the characteristics of patients admitted to HT was observed after the transition from the randomized controlled trial phase in the first study year to non-randomized admissions in the second year. The authors of that study attributed this discrepancy to a potential Hawthorne effect (30), which describes the phenomenon where individuals alter their behavior due to the awareness of being observed or studied in an experiment, thereby impacting the outcomes. Since our study was conducted entirely under routine care conditions, without the structured procedures inherent to randomized trials, this difference appears to have persisted, particularly in the HT - 2 subgroup. While many studies have explored which patients benefit from HT (23, 28, 31, 32), future research could examine whether admission algorithms based on these criteria contribute to better outcomes and lower costs.

In accordance with previous studies, HT in the Swiss Canton of Aargau may effectively reduce treatment costs. Its efficacy, assessed in terms of treatment duration, cost efficiency, and readmission rates, is optimized when HT is utilized as a standalone intervention rather than in conjunction with inpatient treatment, which diminishes the inherent benefits of HT. From an economic viewpoint, we therefore suggest that HT be employed primarily as an inpatient-equivalent intervention, designated for cases where outpatient or day clinic care proves inadequate, thereby mitigating the risk of unintended expansion of service volumes. Subsequent studies should explore which patients or diagnostic groups may benefit from direct admission to HT without prior inpatient care, potentially enhancing its overall effectiveness. Furthermore, the implementation of a program that provides short-term inpatient crisis intervention followed by HT may yield significant benefits for both patients and healthcare systems.