Introduction
The detention and compulsory hospital admission for treatment are intended and regulated for psychiatric disorders [Reference Zinkler and Priebe1]. While raising legal and ethical concerns, it is also a socially desired and regulated practice [Reference Zinkler and Priebe1–Reference Pescosolido, Monahan, Link, Stueve and Kikuzawa3]. In most European countries, the legal framework of involuntary hospitalization usually requires the presence of a psychiatric disorder coupled with either self-harm or danger for others and the need for care [Reference Rusch, Muller, Lay, Corrigan, Zahn and Schonenberger4], whenever less restrictive (e.g., ambulatory or day-clinic) treatment options are not feasible. Compulsory admission and treatment infringe on personal autonomy, they stigmatize both the patient and psychiatric practices and can be detrimental [Reference Rusch, Muller, Lay, Corrigan, Zahn and Schonenberger4,Reference Wasserman, Apter, Baeken, Bailey, Balazs and Bec5]. Previous studies have detected several risk factors for involuntary admission. These are not limited to sociodemographic and clinical characteristics of the patients but encompass several external conditions such as national legislation, community structure, or mental health resources [Reference Rossler6]. Identified sociodemographic factors related to higher rates of compulsory admission are gender (male), civil status (unmarried, widow, single, or divorced), employment status (unemployed or welfare recipients), background, and ethnicity (immigrants and minorities) [Reference Barnett, Mackay, Matthews, Gate, Greenwood and Ariyo7,Reference Keown, McBride, Twigg, Crepaz-Keay, Cyhlarova and Parsons8]. Age is related to the risk of compulsory admissions, although under different circumstances [Reference Keown, McBride, Twigg, Crepaz-Keay, Cyhlarova and Parsons8–Reference Hotzy, Hengartner, Hoff, Jaeger and Theodoridou11]. Organic, psychotic, and bipolar disorders are related to a higher rate of compulsory admissions [Reference Silva, Golay and Morandi10–Reference Arnold, Moeller, Hochstrasser, Schneeberger, Borgwardt and Lang14], while affective and substance use disorders are related to a higher rate of voluntary admissions [Reference Walker, Mackay, Barnett, Sheridan Rains, Leverton and Dalton-Locke12]. Several accompanying clinical features, such as aggression, disruptive behavior, suicidal thoughts, and cognitive impairment, have been considered risk factors. Previous involuntary hospitalizations, police involvement, and referral by on-call physicians are also related to a higher number of compulsory admissions [Reference Hustoft, Larsen, Auestad, Joa, Johannessen and Ruud9]. Countries with higher healthcare spending and more inpatient beds also have higher rates of involuntary psychiatric hospitalizations [Reference Sheridan Rains, Zenina, Dias, Jones, Jeffreys and Branthonne-Foster15].
Compulsory admission orders affect the patients’ autonomy and disturb the therapeutic relationship (and shared decision-making) [Reference Theodoridou, Schlatter, Ajdacic, Rossler and Jager16]. Therapists have to reconcile the therapeutic needs and preferences of the patient, their wellbeing, and the reasons that lead to the compulsory admission itself [Reference Hoff17,Reference Huber and Schneeberger18]. Furthermore, they also affect the patients’ willingness and cooperation, to the extent that coercion or force might be required to treat the patient, directly hampering the therapeutic process and potentially diminishing the efficacy of the interventions [Reference Luciano, Sampogna, Del Vecchio, Pingani, Palumbo and De Rosa19]. Affected patients hold differing views regarding the need and success of compulsory admission orders [Reference Priebe, Katsakou, Amos, Leese, Morriss and Rose20,Reference Katsakou, Rose, Amos, Bowers, McCabe and Oliver21]. The use of coercive measures, on the other hand, is ubiquitously disapproved and recognized as potentially traumatizing [Reference McLaughlin, Giacco and Priebe22]. The risk factors, diagnoses, and clinical profiles related to a compulsory admission order and the involuntary treatment itself are determinants for treatment selection, and therefore outcomes. Despite its profound personal, therapeutic, ethical, and legal implications, compulsory admission order outcomes have not been thoroughly explored, which impedes a proper risk–benefit assessment of involuntary hospitalization as an intervention [Reference Huber and Schneeberger18,Reference Luciano, Sampogna, Del Vecchio, Pingani, Palumbo and De Rosa19,Reference Chieze, Hurst, Kaiser and Sentissi23,Reference Kersting, Hirsch and Steinert24]. Therefore, we aim to explore factors associated with compulsory admission and, using propensity score matching, to assess the impact of compulsory admission on treatment, outcome, and service use. The latter is of critical importance due to the restriction of patients’ freedom and autonomy, requiring ethical reasoning and research on the benefits of involuntary hospitalization (and coercive measures) beyond security aspects.
Methods
Study setting and legal regulations
The Department of Psychiatry, Psychotherapy, and Psychosomatics, as part of the Psychiatric University Hospital of Zurich, is responsible for the psychiatric inpatient treatment of adult patients in the City of Zurich, Switzerland, and its surroundings, with a catchment area of approximately 500,000 inhabitants. The Canton of Zurich provides around 55 psychiatrists and 80 psychiatric beds per 100,000 inhabitants. The number of psychiatrists is above, and the number of psychiatric beds is marginally below the Swiss mean. Nevertheless, both rank higher than other Organisation for Economic Cooperation and Development (OECD) Countries [25–28]. Furthermore, the number of compulsory admissions in the Canton of Zurich is 2.2 per 1,000 inhabitants, which is high compared to national and international benchmarks [Reference Sheridan Rains, Zenina, Dias, Jones, Jeffreys and Branthonne-Foster15,Reference Schuler, Tuch and Peter29].
The Swiss Civil Code and the Law for Protection of Children and Adults regulate the issue of a compulsory admission order and involuntary admission to a psychiatric institution [30,31]. Admission orders can be issued by either the competent authority or any practicing physician and, unless prolonged by the competent authority, run out after 6 weeks. The patient is entitled to demand their discharge and appeal the compulsory admission order in court at any time. Compulsory admission orders require a “mental disorder, mental disability or serious self-neglect if the needed treatment or care cannot be provided otherwise” (Article 426 Swiss Civil Code). In contrast to other countries, impending danger to the patients themselves or others is not essential for a compulsory admission order. However, it is required in the case of a compulsory retention order for voluntarily admitted patients (Article 427 Swiss Civil Code). Compulsory retention orders last up to 72 hours, during which they must be confirmed by the competent authority or an independent board-certified psychiatrist; otherwise, the patient must be discharged. Compulsory measures like isolation, restraint, or forced medication are usually used only in the presence of severe and imminent danger to patients themselves or others. They are regulated for psychiatric emergencies (Article 435 Swiss Civil Code). Under certain circumstances, forced treatment may electively be ordered to avert impending or repetitive complications resulting from non-treatment (Article 434 Swiss Civil Code).
Study design, data sources
Our study retrospectively analyzed electronic health records of all first admissions and discharges from our hospital between January 1, 2013, and December 31, 2019. To record service use one year after discharge, we extended the collection period until December 31, 2020. We extracted routine clinical data from electronic health records for the present study. The Ethics Committee of the Canton of Zurich authorized the use of the anonymized data for research and publication purposes (BASEC: 2018-01906).
We used sociodemographic, clinical, and service use variables for the present analysis. Sociodemographic variables included age, gender, civil and educational status, German language proficiency, and migration status. We used the main treatment diagnoses according to the WHO-ICD-10 criteria. We used the Clinical Global Impression (CGI) scale and the Health of the Nation Outcome Scales (HoNOS) for the clinical evaluation. In addition, we extracted the pharmacologic and nonpharmacologic treatments prescribed during the hospitalization from the clinical records (Information Box 1). Service use variables included the type of admission, change of legal status during hospitalization, duration of treatment, type of discharge (i.e., regular or irregular in case of discharge against medical advice, court decision, death, or suicide of inpatients), and same hospital readmissions within one year after discharge.
Information Box 1. Description of pharmacologic and nonpharmacologic treatments prescribed during hospitalization.
We classified the treatment diagnosis upon discharge in nine diagnostic groups according to the WHO-ICD-10 categories [32], in order to obtain representative and sufficiently large groups of patients: Dementia (F00–F04), Neurocognitive Disorders (F0X, X denotes the remaining categories), Alcohol Use Disorder (F10), Substance Use Disorders (F11–F19), Schizophrenia Spectrum Disorders (F2), Mania and Bipolar Disorder (F30–F31), Major Depressive Disorder (F3X), Anxiety and Stress-Related Disorders (F4–F5), and Personality Disorders (F6). Furthermore, we recorded the presence of comorbid alcohol and substance use (F10-F19) and personality disorders (F6).
The CGI Scales and the HoNOS were rated upon admission and discharge. The CGI is an easily applicable measurement instrument to assess severity (CGI-S) and improvement or deterioration during hospitalization (CGI-I). CGI-S is rated on a seven-point Likert scale from 1 (“normal”) to 7 (“extremely ill”). The CGI-I evaluates changes in comparison to the previous CGI evaluation. It ranges from 1 (“very much improved”) to 7 (“very much worse”), whereby a score of 4 indicates no change [Reference Guy33,Reference Busner and Targum34]. The HoNOS is a measurement instrument used to assess the severity of psychiatric disorders in 12 different domains covering behavior, symptomatology, impairment, and psychosocial functioning. Each item is rated on a five-point Likert scale from 0 (“no problem”) to 4 (“severe to very severe problem”). We evaluated the HoNOS at scale level (i.e., sum score ranging from 0 to 48) and item level [Reference Wing, Curtis and Beevor35–Reference James, Painter, Buckingham and Stewart38]. We considered HoNOS Items rated three or four as clinically significant and as an integral part of the patients’ care plan [Reference James, Painter, Buckingham and Stewart38].
Statistical analysis
According to the principle of independence, the analysis only included the first admission between January 1, 2013, and December 31, 2019. Descriptive statistics (mean, standard deviation, median, interquartile range—IQR, and percentages) were used to characterize the whole sample according to their admission status (i.e., voluntary vs. compulsory). The propensity score represents the probability of individual cases to be compulsory admitted, conditional on their observed characteristics. Using logistic regression, we determined the relationship between sociodemographic and clinical characteristics and compulsory admission. Odds ratios (OR) were calculated with a 99% confidence interval (CI). Therefore, categorical variables were dichotomized, allowing to assess the risk associated with a single condition in contrast to all others not sharing this specific condition. The propensity score was calculated using logistic regression with the variables measured at admission. The model included sociodemographic, diagnostic, and clinical characteristics of the patients and the service use aspects (Information Box 2). Conditional on the propensity score, the distribution of observed baseline covariates will be similar between compulsory and voluntary admitted patients, allowing to assess the unbiased effect of compulsory admission order [Reference Austin39].
Information Box 2. Variables (and their levels) used to calculate the propensity score.
Each compulsorily admitted patient was matched in a 1:1 ratio to their unique nearest voluntarily admitted neighbor on the propensity score scale, with the smallest absolute, averaged propensity score distance across all included subjects [Reference Olmos and Govindasamy40,Reference Ho, Imai, King and Stuart41]. If no matching pair was found, cases were excluded to guarantee similar distribution of variables in the secondary dataset. To assess the balance between the groups (before and after matching), we used the standardized mean difference (SMD) for continuous variables, the Chi-square (χ 2) test for proportions, as well as propensity score distribution before and after matching (Table A1). We conducted an equivalence test for statistically different variables with a low effect size to determine whether the observed effect was smaller than our smallest effect size of interest (SD = 0.50). We chose a half standard deviation since it is consistently considered as a minimally important difference in health outcomes [Reference Norman, Sloan and Wyrwich42,Reference Lakens, Scheel and Isager43]. Two separated one-sided tests were performed to determine if the observed effect was larger than the lower bound (i.e., SD > −0.50) and less than the upper bound (i.e., SD < +0.50). Equivalence can be stated when the confidence interval rests within the equivalence boundaries [Reference Lakens, Scheel and Isager43–Reference Lakens45].
All subsequent analyses were conducted with the propensity score matched sample. Variables measured at discharge were used to estimate the differences in treatment prescribed and outcomes between compulsorily and voluntarily admitted patients. We used Student’s t-test to assess differences in continuous variables and the Chi-square test (χ 2) for differences in proportions. If assumptions about the distribution were not met, we additionally used an alternative nonparametric test (i.e., the Wilcoxon Signed-Rank Test). For changes in HoNOS sum scores, from admission to discharge, a single-factor independent group analysis of covariance (ANCOVA) was used to test for differences according to the admission status (i.e., voluntary vs. compulsory), thereby controlling for variability in scores upon admission. Two Kaplan–Meier time-to-event curves representing time to discharge (i.e., duration of treatment) and time to same hospital readmission were calculated; for testing the statistical significance, we used the log-rank p-value.
All tests of significance were two-tailed. Due to the large sample size, p-values less than .01 were considered significant. For significant results, SMD was used to evaluate effect sizes. For the analysis of the single HoNOS items, a Bonferroni correction for repeated measurements was performed. Because all remaining analyses were considered exploratory, no further correction for multiple comparisons was performed. Statistical analyses and figures were conducted using RStudio (2021.09.1 + 372); the statistical software R (4.1.2); and the R packages: tidyverse (1.3.1), TOSTER (0.3.4), MatchIt (4.3.1), survival (v 3.2–13), and survminer (0.4.9).
Results
Demographic and clinical characteristics of the study population
Between January 1, 2013, and December 31, 2019, 35,311 direct admissions and discharges occurred; 17,290 were individual first admissions. The mean age was 46.3 (19.4) years, with 49.0% (n = 8,474) females. Low German language proficiency occurred in 13.6% (n = 2,343) of the population. Almost one third of all admissions (30.4%, n = 5,250) were compulsory admissions. The most common diagnoses were major depressive disorder (29.2%, n = 5,042), anxiety and stress-related disorders (17.0%, n = 2,943), Schizophrenia spectrum disorders (15.8%, n = 2,729), and alcohol use disorders (11.2%, n = 1,935), accounting altogether for almost three quarters (73.2%, n = 12,649) of first admissions (Table 1).
Table 1. The sample’s demographic and diagnostic characteristics according to admission status (i.e., voluntary vs. compulsory), before and after propensity score matching.
Abbreviation: SMD, standardized mean difference.
Relationship between sociodemographic and clinical variables with compulsory admission
Patients aged over 75 years (OR: 3.63, 99%CI: 3.19–4.13) and unmarried (OR: 1.49, 99%CI: 1.36–1.63) had an increased risk of being compulsorily admitted. Adults between 25 and 50 had a lower risk. Patients with low German language proficiency (OR: 1.79, 99%CI: 1.59–2.02), those who completed regular school education (OR: 1.89, 99%CI: 1.73–2.06), and nonresidential population (OR: 2.62, 99%CI: 1.70–2.33) were also at higher risk. Patients with dementia (OR: 5.23, 99%CI: 4.01–6.89), neurocognitive disorders (OR: 5.40, 99%CI: 4.55–6.43), or a schizophrenia spectrum disorder (OR: 2.42, 99%CI: 2.17–2.70) showed an increased risk for compulsory admission, while those with a major depressive disorder showed a decreased risk (OR: 0.31, 99%CI: 0.27–0.34). Patients showing “Aggressive and Disruptive Behavior” (HoNOS Item 1: OR: 5.23, 99%CI: 4.71–5.81), “Self-Harm” (Item 02: OR: 2.53, 99%CI: 2.21–2.89), “Cognitive problems” (Item 4: OR: 2.99, 99%CI: 2.71–3.29), “Hallucinations and Delusions” (Item 6: OR: 3.05, 99%CI: 2.75–3.38), and “Problems with Living Conditions” (Item 11: OR: 2.08, 99%CI: 1.90–2.29) had a higher probability of being compulsorily admitted, while those with “Depressed Mood” (HoNOS Item 7: OR: 0.50, 99%CI: 0.46–0.55) had a lower probability. For a full description of the variables, see Appendix.
Propensity score matched paired sample
Using propensity score matching, we obtained a matched sample of 9,140 patients, 4,570 compulsorily and voluntarily admitted patients each. The mean age of the paired sample was 47.6 (20.6) years with 47.9% (n = 4,381) females. The more frequent diagnoses among compulsorily admitted patients were schizophrenia spectrum disorders (26.5%, n = 1,209), major depressive disorder (16.6%, n = 757), and anxiety and stress-related disorders (15.6%, n = 712). The balancing parameters of the matched pairs sample improved (Table A1). Upon admission, compulsorily admitted patients showed a higher HoNOS sum score (20.40 ± 6.82 vs. 20.88 ± 7.50; t(4,569) = 4.29, p < 0.001, SMD = 0.070) additional to a higher count of other clinically relevant items (4.00 ± 2.32 vs. 4.23 ± 2.47, t(4,569) = 4.65, p < 0.001, SMD = 0.095). However, the HoNOS sum score (t(8,949.45) = 20.63, p < 0.001) and the count of clinically relevant items (t(9,102.37) = 19.13, p < 0.001) were statistically equivalent. Severity gradings according to the CGI-S were (4.85 ± 1.07 vs. 4.94 ± 1.09; t(4,569) = 4.29, p < 0.001, SMD = 0.090) statistically different, and statistically equivalent (t(9,134.87) = 19.91, p < 0.001). No matched pair could be found for 680 (12.9%) of all compulsorily admitted patients. These were mostly patients aged 75 years (n = 515, 75.7%) or older and those diagnosed with dementia or neurocognitive disorders (n = 569, 83.7%) (see Appendix).
Treatment, clinical outcomes, and service use parameters
The duration of treatment was shorter for compulsorily admitted patients (24.44 ± 31.12 vs. 28.50 ± 28.70 days, t(4,569) = 6.56, p < 0.001). While hospitalized, the main treatment offered to compulsorily admitted patients was crisis intervention (64.6 vs. 76.7%, χ 2(1) = 161.2, p < 0.001). Involuntarily hospitalized patients were less frequently assigned to other treatments, such as individual psychotherapy (31.5 vs. 21.6%, χ 2(1) = 113.7, p < 0.001) or group psychotherapy (17.1 vs. 9.5%, χ 2(1) = 112.3, p < 0.001), occupational therapies (41.3 vs. 32.3%, χ 2(1) = 112.3, p < 0.001), with slightly lower rates of counseling (47.2 vs. 40.9%, χ 2(1) = 36.8, p < 0.001), observation (13.6 vs. 11.3%, χ 2(1) = 10.9, p < 0.001), and psychopharmacologic treatment (71.2 vs. 67.7%, χ 2(1) = 12.4, p < 0.001) (Figure 1A). Psychopharmacologic therapy was overall less frequently prescribed to compulsorily admitted patients, especially antidepressants (35.5 vs. 25.1%, χ 2(1) = 114.9, p < 0.001), mood stabilizers (8.8 vs. 6.7%, χ 2(1) = 12.9, p < 0.001), stimulants (2.0 vs. 1.1%, χ 2(1) = 11.2, p < 0.001), and opioids (3.7 vs. 2.3%, χ 2(1) = 14.2, p < 0.001). Similar prescription rates were found for antipsychotics (49.6 vs. 48.2%, χ 2(1) = 1.7, p = 0.18), LAI antipsychotics (1.7 vs. 1.9%, χ 2(1) = 0.6, p = 0.43), anxiolytics/hypnotics (40.3 vs. 42.1%, χ 2(1) = 2.74, p = 0.09), other psychotropics (4.8 vs. 4.0%, χ 2(1) = 2.84, p = 0.09), and other medication (24.1 vs. 22.0%, χ 2(1) = 5.7, p = 0.02) (Figure 1B). Compulsory patients were much more likely to be exposed to coercive measures, either as forced medication (2.3 vs. 8.7%, χ 2(1) = 177.6, p < 0.001) or seclusion or restraint (2.7 vs. 9.3%, χ 2(1) = 179.7, p < 0.001) (Figure 1C).
Figure 1. Odds ratios and 99% confidence intervals for treatment prescribed; the probability was calculated dichotomizing each variable. (A) Nonpharmacologic treatment. (B) Pharmacologic treatment. (C) Coercive treatment.
The HoNOS sum score improved for both groups from admission to discharge (F(3,13,000) = 960.3, p < 0.001), demonstrating that both groups experienced a significant improvement during hospitalization. Upon discharge, the HoNOS sum score (p = 0.23), and the number of clinically relevant items were similar (p = 0.57). Compulsorily admitted patients had a higher sum score difference (7.96 ± 7.33 vs. 8.63 ± 7.84, t(4,569) = 4.21, p < 0.001, SMD = 0.09), the difference was statistically equivalent (t(9,096.97) = 19.68, p < 0.001). Both groups had a similar percentage of change in the HoNOS sum score (41.5 ± 29.5% vs. 40.7 ± 39.6%, t(4,569) = 1.04, p = 0.29). According to CGI-I, compulsorily admitted patients experienced more improvement (2.57 ± 0.99 vs. 2.51 ± 0.99, t(4,569) = 4.65, p = 0.01; SMD = 0.05), although absolute differences remained small, and the results can be considered statistically equivalent (t(9,138) = 21.48, p < 0.001) (Table 2).
Table 2. The propensity score matched sample’s clinical and subsequent service use characteristics according to the admission status.
Abbreviation: SMD, standardized mean difference.
The distribution of the duration of treatment was right-skewed for both groups (voluntary admissions: median: 21; IQR: 35 days; compulsory admissions: median: 14; IQR: 30 days), with compulsory admissions having a shorter length of stay (T = 5,706,014, z = 4.49, p < 0.001). The duration of treatment curve showed a significant difference between both groups. This difference becomes larger during the second and fourth weeks after admission (Figure 2A). The percentage of patients readmitted was higher for the voluntarily admitted patients (38.3 vs. 34.2%, χ 2(1) = 43.34, p < 0.001). While the time to readmission did not differ (p = 0.34), the time to readmission curve (Figure 2B) was parallel for both groups. The number of readmissions was similar between both groups (2.30 ± 1.33 vs. 2.22 ± 1.24, t(3,304) = 1.95, p = 0.05). Although more irregular discharges were recorded for compulsory patients (3.2, n = 151 vs. 5.8%, n = 273, χ 2(1) = 43.34, p < 0.001), almost a quarter (23%, n = 1,050) expressed the willingness to remain voluntarily in further inpatient treatment. On the other hand, 4.2% (n = 186) of all voluntarily admitted patients were retained against their will. Moreover, deaths (including suicides), although overall low, occurred almost three times more often in the compulsorily admitted group (0.2%, n = 9 vs. 0.5%, n = 21, χ 2(1) = 4.04, p = 0.04) (Table 2).
Figure 2. Kaplan-Meier time-to-event curves. (A) Duration of treatment. (B) Time to readmission.
Discussion
The main finding of our study is that compulsorily admitted patients achieved a clinical improvement similar to voluntarily admitted patients in a shorter length of stay. During the 12 months following discharge, those initially voluntarily admitted had a higher readmission rate. This lower rate combined with the similar time to readmission suggests a robust and sustainable improvement in those compulsorily admitted. Furthermore, patients with a compulsory admission order were more frequently admitted for crisis intervention. Regarding their pharmacologic treatment, compulsorily admitted patients had an overall lower prescription rate, although they had higher rates of forced medication and seclusion or restraint. This finding is rather unsurprising since (imminent danger excluded) compulsory admission orders are legal requirements for coercive treatment [Reference Lay, Nordt and Rossler46].
The duration of treatment was shorter for compulsorily admitted patients, corroborating previous national findings [Reference Habermeyer, De Gennaro, Frizi, Roser and Stulz47]. This is opposed to international findings, where involuntary hospitalization has been related to a longer length of stay [Reference Gopalakrishna, Ithman and Malwitz48]. We consider this an effect of differences in legal stipulations between countries [Reference Habermeyer, De Gennaro, Frizi, Roser and Stulz47,Reference Dimitri, Giacco, Bauer, Bird, Greenberg and Lasalvia49], with the Swiss legislation mandating to release the patients as soon as possible [30,31]. The shorter duration of treatment in compulsorily admitted patients cannot be attributed to previously identified sociodemographic or clinical characteristics [Reference Gopalakrishna, Ithman and Malwitz48] since these were balanced out through the propensity score matching.
Considering the overall lower prescription rates of pharmacologic and nonpharmacologic treatments in compulsorily admitted patients, outcomes and the shorter duration of treatment are not explained by the prescribed treatment. We consider that, in contrast to other medical specialties, psychiatric hospitalization is an intervention in itself, capable of averting danger and modifying the course of illness, thus resulting in a reduced burden of disease [Reference Kortrijk, Staring, van Baars and Mulder50]. Furthermore, the improvement seems sustainable since the readmission rate to the same hospital is lower for compulsorily admitted patients, and the time to readmission is similar. However, the readmission rate should not be interpreted unconditionally as treatment success. A compulsory admission might have been such a disturbing experience [Reference Chieze, Hurst, Kaiser and Sentissi23], with the potential to elicit an avoidant behavior and a tendency to seek less help even if psychiatric inpatient care is needed [Reference Swartz, Swanson and Hannon51]. Thus, it partially explains the lower readmission rate of compulsorily admitted patients within one year.
The clinical psychiatric evaluation requires sufficient observation time to verify the stability of an improvement [Reference Walker, Mackay, Barnett, Sheridan Rains, Leverton and Dalton-Locke12,Reference Lay, Kawohl and Rossler52,Reference Setkowski, van der Post, Peen and Dekker53]. Therefore, this might explain the differences in length of stay observed principally between the second and fourth weeks of treatment. Although an observation period of a few days might be insufficient for accurate clinical appraisal and sustainable intervention, a more extended treatment duration may indicate a more demanding clinical case [Reference Schmitz-Buhl, Gairing, Rietz, Haussermann, Zielasek and Gouzoulis-Mayfrank54–Reference Katsakou and Priebe56]. Furthermore, as suggested by previous findings [Reference Arnold, Moeller, Hochstrasser, Schneeberger, Borgwardt and Lang14], compulsorily admitted patients with a longer length of stay usually chose to do this voluntarily when recommended to do so, leading to lower rates of discharge against medical advice as well as fewer appeals against the compulsory admission order.
We consider the observation time of one year after discharge appropriate since two-thirds of involuntary readmissions seem to occur within six months after the first hospitalization, an observation we also encountered in our population. Nonetheless, the same hospital readmission rate is a controversial measure for service use since the patients have the freedom to choose a specific institution and may not approach a hospital to which they were compulsorily admitted in the past [Reference Nasir, Lin, Bueno, Normand, Drye and Keenan57]. In contrast, involuntary patients do not have this freedom of choice since the catchment areas determine the institution responsible for inpatient treatment. Therefore, changes in institutions are unlikely in these cases.
Regarding the risk factors for involuntary hospitalization, our results align with previous studies [Reference Silva, Golay and Morandi10,Reference Silva, Gholam, Golay, Bonsack and Morandi58]. Older ages were associated with an increased risk of compulsory admission orders in our sample, while adults faced an overall lower risk. Older age might relate to the effects of aging, while younger age may be related to the peak of thought disorders [Reference Keown, McBride, Twigg, Crepaz-Keay, Cyhlarova and Parsons8–Reference Hotzy, Hengartner, Hoff, Jaeger and Theodoridou11]. Other cultural and social factors associated with an increased compulsory admission rate included low (German) language proficiency, lack of vocational/professional training, and psychotic symptomatology. Psychiatric diagnoses characterized by cognitive impairment (i.e., neurocognitive and neurodegenerative disorders) or thought disorders (schizophrenia, mania, and bipolar disorder) increased the risk for compulsory admission [Reference Keown, McBride, Twigg, Crepaz-Keay, Cyhlarova and Parsons8,Reference Lay, Nordt and Rossler46]. The clinical HoNOS profile of compulsorily admitted patients showed higher rates in domains relating to harm and danger, cognitive impairment, and psychotic symptoms [Reference Rabinowitz, Massad and Fennig59].
The main strength of our study is the large clinical sample collected under the same legal framework over a long period [Reference Hotzy, Hengartner, Hoff, Jaeger and Theodoridou11]. When analyzing, interpreting, and comparing our results, the interplay between patient characteristics, local mental health services peculiarities, and legal regulations must be considered. They regulate and determine the patients’ access to treatment and directly influence therapeutic interventions [Reference Rossler6,Reference Sheridan Rains, Zenina, Dias, Jones, Jeffreys and Branthonne-Foster15,Reference Smith, Gate, Ariyo, Saunders, Taylor and Bhui60–Reference Salize and Dressing62]. In Switzerland, the rate of compulsory admission orders is high compared to other countries, suggesting a relatively low threshold for their use [Reference Hotzy, Hengartner, Hoff, Jaeger and Theodoridou11,Reference Salize and Dressing62]. In contrast to previous findings, this might explain why danger or harm to oneself was also related to a compulsory admission order [Reference Silva, Golay and Morandi10]. We consider propensity score matching a valid method to control confounding variables and reduce the potential bias by indication [Reference Benedetto, Head, Angelini and Blackstone63,Reference Andrade64]. Using propensity score matching, we could balance the modifying effects of sociodemographic variables, diagnosis, and severity over treatment selections and, therefore, outcomes [Reference Zhang, Kim, Lonjon and Zhu65].
To reduce the possible flaws of propensity score matching, we selected the 1:1 ratio and unique nearest neighbor, a robust propensity score matching approach [Reference Ho, Imai, King and Stuart41,Reference King and Nielsen66]. The balancing indices substantially improved after matching. Nonetheless, the clinical rating scales HoNOS and CGI-S still showed a statistically significant difference, although they had low effect size and were statistically equivalent [Reference Lakens, Scheel and Isager43–Reference Lakens45]. We consider the statistically significant difference an artifact of the large sample size [Reference Kaplan, Chambers and Glasgow67,Reference Lantz68]. Unfortunately, matching was not feasible for the whole sample, with 12.9% (n = 680) of compulsorily admitted patients ending up without a counterpart, mainly comprising patients older than 75 years with a neurocognitive disorder. Due to the unique characteristics of this group and their influence on treatment selection, their exclusion led to a reduction in bias.
One main limitation of our study is the inability to deduce the individual’s perception of neither impaired personal integrity, nor legitimation and usefulness, as patients’ treatment satisfaction is prognostic for future involuntary admissions [Reference Priebe, Katsakou, Amos, Leese, Morriss and Rose20]. Nevertheless, we can infer that a large proportion of compulsorily admitted patients accept hospitalization as a reasonable option since almost a quarter of patients agrees to remain voluntarily in treatment, and irregular discharge is just two percentage points higher than in voluntarily admitted patients. Another point to consider is that patients suffering from critical, life-threatening medical conditions are usually transferred to a general hospital before being referred to a psychiatric institution (e.g., in case of severe intoxication or anorexia). Thus, the rate of deceased patients could be inaccurate, limiting comparability [Reference Kersting, Hirsch and Steinert24]. Finally, from our data, we cannot account for informal coercion or suggestion before or during treatment, though other research suggests its relevant influence [Reference Sheridan Rains, Zenina, Dias, Jones, Jeffreys and Branthonne-Foster15,Reference Salize and Dressing62,Reference Hotzy and Jaeger69].
In summation, the presence of danger to oneself or others in combination with factors that impair communication, collaboration, and bonding impede outpatient treatment. Thus, the lack of compliance and commitment to treatment undermines the efficacy of outpatient compulsory treatment orders [Reference Lay, Nordt and Rossler46,Reference Barnett, Matthews, Lloyd-Evans, Mackay, Pilling and Johnson70]. In circumstances when the need for care is urgent, a compulsory admission order becomes an imminent alternative [Reference Rabinowitz, Massad and Fennig59,Reference Sederer and Summergrad71]. Our results show that compulsorily admitted patients are more likely to present harmful or dangerous behavior, coupled with impaired communication and bonding. They experience forced medication and seclusion or restraint more frequently. However, they show a robust clinical improvement within a shorter time than voluntary patients. The lower readmission rate coupled to similar time to readmission suggests a sustainable improvement. From our analysis, compulsory admission orders leading to involuntary hospitalization appear to be a meaningful intervention to reduce dangerous and harmful behavior. However, due to its ethical and legal implications, the threshold for such an enormous impairment of a person’s rights and freedom must be carefully outweighed. Therefore, both in routine clinical practice and future studies, the patients’ opinions and the notion of treatment have to be considered even more.
Data Availability Statement
The data supporting the findings of this study are available from the corresponding author upon reasonable request. The data are not publicly available due to privacy or ethical restrictions.
Author Contributions
Conceptualization: E.S., S.V., S.T.E.; Data curation: A.H., S.T.E.; Methodology: A.H., V.N., S.T.E.; Project administration: E.S., S.V.; Supervision: E.S., S.V.; Validation: A.H., L.H., V.N.; Writing—review and editing: A.H., L.H., H.S., S.T.E., V.N.
Financial Support
This research received no specific grant from any funding agency, commercial, or not-for-profit sectors.
Conflict of Interest
The authors declare none.
Appendix
Figure A1. Propensity score distribution for the matched and nonmatched samples.
Figure A2. Pre- and post-matching propensity score distribution.
Table A1. Relation between the single variables and compulsory admission.
Table A2. Demographic and clinical characteristics of the sample without a matching pair.
Open access
Comments
No Comments have been published for this article.