Identification of risk factors for involuntary psychiatric hospitalization: using environmental socioeconomic data and methods of machine learning to improve prediction

Involuntary psychiatric hospitalisation - differences and similarities between patients detained under the mental health act and according to the legal guardianship legislation

BACKGROUND

Involuntary psychiatric hospitalisation occurs under different legal premises. According to German law, detention under the Mental Health Act (MHA) is possible in cases of imminent danger of self-harm or harm to others, while detention according to the legal guardianship legislation (LGL) serves to prevent self-harm if there is considerable but not necessarily imminent danger. This study aims to compare clinical, sociodemographic and environmental socioeconomic differences and similarities between patients hospitalised under either the MHA or LGL.

METHODS

We conducted a retrospective health records analysis of all involuntarily hospitalised cases in the four psychiatric hospitals of the city of Cologne, Germany, in 2011. Of the 1,773 cases, 87.3% were detained under the MHA of the federal state of North Rhine-Westphalia and 6.4% were hospitalised according to the federal LGL. Another 6.3% of the cases were originally admitted under the MHA, but the legal basis of detention was converted to LGL during the inpatient psychiatric stay (MHA→LGL cases). We compared sociodemographic, clinical, systemic and environmental socioeconomic (ESED) variables of the three groups by means of descriptive statistics. We also trained and tested a machine learning-based algorithm to predict class membership of the involuntary modes of psychiatric inpatient care.

RESULTS

Cases with an admission under the premises of LGL lived less often on their own, and they were more often retired compared to MHA cases. They more often had received previous outpatient or inpatient treatment than MHA cases, they were more often diagnosed with a psychotic disorder and they lived in neighbourhoods that were on average more socially advantaged. MHA→LGL cases were on average older and more often retired than MHA cases. More often, they had a main diagnosis of an organic mental disorder compared to both MHA and LGL cases. Also, they less often received previous psychiatric inpatient treatment compared to LGL cases. The reason for detention (self-harm or harm to others) did not differ between the three groups. The proportion of LGL and MHA cases differed between the four hospitals. Effect sizes were mostly small and the balanced accuracy of the Random Forest was low.

CONCLUSION

We found some plausible differences in patient characteristics depending on the legal foundation of the involuntary psychiatric hospitalisation. The differences relate to clinical, sociodemographic and socioeconomical issues. However, the low effect sizes and the limited accuracy of the machine learning models indicate that the investigated variables do not sufficiently explain the respective choice of the legal framework. In addition, we found some indication for possibly different interpretation and handling of the premises of the law in practice. Our findings pose the need for further research in this field.

Single-label machine learning classification revealed some hidden but inter-related causes of five psychotic disorder diseases

Psychotic disorder diseases (PDD) or mental illnesses are group of illnesses that affect the minds and impair the cognitive ability, retard emotional ability and obstruct the process of communication and relationship with others and are characterized by delusions, hallucinations and disoriented or disordered pattern of thinking. Prognosis of PDD is not sufficient because of the nature of the diseases and as such adequate form of diagnosis is required to detect, manage and treat the illness. This paper applied the single-label classification (SLC) machine learning approach in mining of electronic health records of people with PDD in Nigeria using eleven independent (demographic) variables and five PDD as target variables. The five PDDs are Insomnia, Schizophrenia, Minimal Brain dysfunction (MBD), which is also known as Attention-Deficit/Hyperactivity Disorder (ADHD), Vascular Dementia (VD) and Bipolar Disorder (BD). The aim of using SLC is that it would be easier to detect some PDDs that are related to each other without the loss of information, which is a plus over multi-label classification (MLC). ReliefF algorithm was used at each experiment to precipitate the order of importance of the independent variables and redundant variables were excluded from the analysis. The order of the variables in feature selection was matched with feature importance after the classifications and quantified using the Spearman rank correlation coefficient. The data was divided into: 70% for training and 30% for testing. Four new performance metrics adapted from the root mean square (RMSE) were proposed and used to measure the differences between the performance results of the 10 Machine learning models in terms of the training and testing and secondly, feature and without feature selection. The new metrics are close to zero which is an indication that the use of feature selection and cross validation may not greatly affects the accuracy of the SLC. When the PDDs are included as predictors for classifying others, there was a tremendous improvement as revealed by the four new metrics for classification accuracy (CA), precision and recall. Analysis of variance showed the four different metrics differs significantly for classification accuracy (CA) and precision. However, there were no significant difference between the CA and precision when the duo are compared together across the four evaluation metrics at p value less than 0.05.

Risk factors and prediction model for mental health in Chinese soldiers

Introduction

This study aimed to explore potential risk factors for mental health concerns, and the prediction model for mental health concerns in Chinese soldiers was constructed through combined eligible risk factors.

Methods

This cross-sectional study was performed on soldiers under direct command from Gansu, Sichuan, and Chongqing in China, and the soldiers were selected by cluster convenient sampling from 16 October 2018 to 10 December 2018. The Symptom Checklist-90 (SCL-90) and three questionnaires (Military Mental Health Status Questionnaire, Military Mental Health Ability Questionnaire, and Mental Quality Questionnaire for Army Men) were administered, including demographics, military careers, and 18 factors.

Results

Of 1,430 Chinese soldiers, 162 soldiers presented mental disorders, with a prevalence of 11.33%. A total of five risk factors were identified, including serving place (Sichuan vs. Gansu: OR, 1.846, 95% CI: 1.028-3.315, P = 0.038; Chongqing vs. Gansu: OR, 3.129, 95% CI, 1.669-5.869, P = 0.003), psychosis (OR, 1.491, 95% CI, 1.152-1.928, P = 0.002), depression (OR, 1.482, 95% CI, 1.349-1.629, P < 0.001), sleep problems (OR, 1.235, 95% CI, 1.162-1.311, P < 0.001), and frustration (OR, 1.050, 95% CI, 1.015-1.087, P = 0.005). The area under the ROC curve by combining these factors was 0.930 (95% CI: 0.907-0.952) for predicting mental disorders in Chinese soldiers.

Conclusion

The findings of this study demonstrate that mental disorders and onset in Chinese soldiers can be predicted on the basis of these three questionnaires, and the predictive value of the combined model was high.

Risk factors of involuntary referral by police to ER psychiatric services for patients with a severe mental illness: A GEE analysis

This study aimed to identify risk factors for involuntary referral by police to emergency room (ER) psychiatric services for community-based patients with a mental illness via a generalized estimating equation (GEE) analysis. The analysis was based on data from the Management Information System of Psychiatric Care (MISPC) system for patients with a severe mental illness in Taipei, Taiwan and registered referral records of the police. Data on 6378 patients aged ≥20 years were used in this study, including 164 patients who were involuntarily referred to the ER by the police and 6214 patients who were not during the period of January 1, 2018 to December 31, 2020. GEEs were utilized to explore possible risk factors of repeated involuntary referral to ER psychiatric services for patients with a severe mental illness. The logistic regressions indicated that patients defined as "severe" according to the Mental Health Act of Taiwan (crude odds ratio (OR): 3.840, 95 % confidence interval (CI): 2.407-6.126), with a disability (crude OR: 3.567, 95 % CI: 1.339-9.501), with two or more family members with a psychiatric disorder (crude OR: 1.598, 95 % CI: 1.002-2.548), with a history of a suicide attempt (crude OR: 25.582, 95 % CI: 17.608-37.167), and with a history of domestic violence (crude OR: 16.141, 95 % CI: 11.539-22.579) were positively associated with involuntary referral to ER psychiatric services. However, age (crude OR: 0.971, 95 % CI: 0.960-0.983) and the MISPC score (crude OR: 0.834, 95 % CI: 0.800-0.869) were inversely associated with involuntary referral to ER psychiatric services. After adjusting for demographics and potential confounders, we found that patients defined as "severe" (Exp (β): 3.236), with a disability (Exp (β): 3.715), with a history of a suicide attempt (Exp (β): 8.706), and with a history of domestic violence (Exp (β): 8.826), as well as age (Exp (β): 0.986) and the MISPC score (Exp (β): 0.902) remained significantly associated with repeated involuntary referral to ER psychiatric services. In conclusion, community-based mentally ill patients with a history of a suicide attempt, with a history of domestic violence, with a severe illness, and with a profound level of disability were highly associated with involuntary referral to ER psychiatric services. We suggest that community mental health case managers identify significant factors associated with involuntary referral to ER psychiatric services to accordingly arrange case management plans.

Service use preceding and following first referral for psychiatric emergency care at a short-stay crisis unit: A cohort study across three cities and one rural area in England

BACKGROUND

Internationally, hospital-based short-stay crisis units have been introduced to provide a safe space for stabilisation and further assessment for those in psychiatric crisis. The units typically aim to reduce inpatient admissions and psychiatric presentations to emergency departments.

AIMS

To assess changes to service use following a service user's first visit to a unit, characterise the population accessing these units and examine equality of access to the units.

METHODS

A prospective cohort study design (ISCTRN registered; 53431343) compared service use for the 9 months preceding and following a first visit to a short-stay crisis unit at three cities and one rural area in England. Included individuals first visited a unit in the 6 months between 01/September/2020 and 28/February/2021.

RESULTS

The prospective cohort included 1189 individuals aged 36 years on average, significantly younger (by 5-13 years) than the population of local service users (<.001). Seventy percent were White British and most were without a psychiatric diagnosis (55%-82% across sites). The emergency department provided the largest single source of referrals to the unit (42%), followed by the Crisis and Home Treatment Team (20%). The use of most mental health services, including all types of admission and community mental health services was increased post discharge. Social-distancing measures due to the COVID-19 pandemic were in place for slightly over 50% of the follow-up period. Comparison to a pre-COVID cohort of 934 individuals suggested that the pandemic had no effect on the majority of service use variables.

CONCLUSIONS

Short-stay crisis units are typically accessed by a young population, including those who previously were unknown to mental health services, who proceed to access a broader range of mental health services following discharge.

Involuntary psychiatric admission in Cyprus: A descriptive correlational study

INTRODUCTION

Studies on the clinical and socio-demographic characteristics of those involuntarily admitted to psychiatric settings could help professionals and researchers to develop effective, targeted interventions, alternative to compulsory psychiatric care.

AIM

The association between socio-demographic and clinical characteristics in adults under involuntary hospitalization for psychiatric treatment in the Republic of Cyprus was assessed.

METHOD

This was a descriptive, cross-sectional and correlational study. Data collection was achieved (December 2016 to February 2018) via a census sampling method. Socio-demographic and clinical data of individuals involuntarily admitted to the reference psychiatric hospital of Cyprus with psychotic symptomatology were recorded.

RESULTS

The sample encompassed 144 females and 262 males. The most frequent diagnosis was schizophrenia or a relevant psychotic disorder (72.9%). The most frequent cause of admission was "Disorganized behaviour" along with non-adherence to pharmacotherapy (53.7%). Approximately 42.8% of the participants confirmed positive substance use history, which was more frequently reported in males than in females (88.5% vs. 11.5%, respectively, p < 0.001). Additionally, males were more frequently admitted due to Disorganized behaviour with substance use compared to females (31.3% vs. 4.9%, respectively, p < 0.001), while females were more frequently admitted due to d"Disorganized behaviour with non-adherence to pharmacotherapy (70.1% vs. 44.7%, respectively, p < 0.001). Also, males were more frequently involuntarily hospitalized due to suicidal/self-harming behaviour compared to females (12.2% vs. 5.6%, respectively, p = 0.031).

CONCLUSION

Gender differences were noted in relation to clinical characteristics of the participants, highlighting the need for gender-specific interventions to decrease compulsory psychiatric care, including enhancement of adherence to therapy.

Determinants of compulsory hospitalisation at admission and in the course of inpatient treatment in people with mental disorders-a retrospective analysis of health records of the four psychiatric hospitals of the city of Cologne

BACKGROUND

We aimed to identify differences in predictors of involuntary psychiatric hospitalisation depending on whether the inpatient stay was involuntary right from the beginning since admission or changed from voluntary to involuntary in the course of in-patient treatment.

METHODS

We conducted an analysis of 1,773 mental health records of all cases treated under the Mental Health Act in the city of Cologne in the year 2011. 79.4% cases were admitted involuntarily and 20.6% were initially admitted on their own will and were detained later during the course of in-patient stay. We compared the clinical, sociodemographic, socioeconomic and environmental socioeconomic data (ESED) of the two groups. Finally, we employed two different machine learning decision-tree algorithms, Chi-squared Automatic Interaction Detection (CHAID) and Random Forest.

RESULTS

Most of the investigated variables did not differ and those with significant differences showed consistently low effect sizes. In the CHAID analysis, the first node split was determined by the hospital the patient was treated at. The diagnosis of a psychotic disorder, an affective disorder, age, and previous outpatient treatment as well as the purchasing power per 100 inhabitants in the living area of the patients also played a role in the model. In the Random Forest, age and the treating hospital had the highest impact on the accuracy and decrease in Gini of the model. However, both models achieved a poor balanced accuracy. Overall, the decision-tree analyses did not yield a solid, causally interpretable prediction model.

CONCLUSION

Cases with detention at admission and cases with detention in the course of in-patient treatment were largely similar in respect to the investigated variables. Our findings give no indication for possible differential preventive measures against coercion for the two subgroups. There is no need or rationale to differentiate the two subgroups in future studies.

Peer support for discharge from inpatient mental health care versus care as usual in England (ENRICH): a parallel, two-group, individually randomised controlled trial

BACKGROUND

High numbers of patients discharged from psychiatric hospital care are readmitted within a year. Peer support for discharge has been suggested as an approach to reducing readmission post-discharge. Implementation has been called for in policy, however, evidence of effectiveness from large rigorous trials is missing. We aimed to establish whether peer support for discharge reduces readmissions in the year post-discharge.

METHODS

We report a parallel, two-group, individually randomised, controlled superiority trial, with trial personnel masked to allocation. Patients were adult psychiatric inpatients (age ≥18 years) with at least one previous admission in the preceding 2 years, excluding those who had a diagnosis of any organic mental disorder, or a primary diagnosis of learning disability, an eating disorder, or drug or alcohol dependency, recruited from seven state-funded mental health services in England. Patients were randomly assigned (1:1) to the intervention (peer support plus care as usual) or control (care as usual) groups by an in-house, online randomisation service, stratified by site and diagnostic group (psychotic disorders, personality disorders, and other eligible non-psychotic disorders) with randomly permuted blocks of randomly varying length to conceal the allocation sequence and achieve the allocation ratio. The peer support group received manual-based, one-to-one peer support, focused on building individual strengths and engaging with activities in the community, beginning during the index admission and continuing for 4 months after discharge, plus care as usual. Care as usual consisted of follow-up by community mental health services within 7 days of discharge. The primary outcome was psychiatric readmission 12 months after discharge (number of patients readmitted at least once), analysed on an intention-to-treat basis. All patients were included in a safety analysis, excluding those who withdrew consent for use of their data. The trial is registered with the ISRCTN registry, ISRCTN10043328. The trial was complete at the time of reporting.

FINDINGS

Between Dec 1, 2016, and Feb 8, 2019, 590 patients were recruited and randomly assigned, with 294 allocated to peer support (287 included in the analysis after withdrawals and loss to follow-up), and 296 to care as usual (291 in the analysis). Mean age was 39·7 years (SD 13·7; range 18-75). 306 patients were women, 267 were men, three were transgender, and two preferred not to say. 353 patients were White, 94 were Black, African, Caribbean, or Black British, 68 were Asian or Asian British, 48 were of mixed or multiple ethnic groups, and 13 were of other ethnic groups. In the peer support group, 136 (47%) of 287 patients were readmitted at least once within 12 months of discharge. 146 (50%) of 291 were readmitted in the care as usual group. The adjusted risk ratio of readmission was 0·97 (95% CI 0·82-1·14; p=0·68), and the adjusted odds ratio for readmission was 0·93 (95% CI 0·66-1·30; p=0·68). The unadjusted risk difference was 0·03 (95% CI -0·11 to 0·05; p=0·51) in favour of the peer support group. Serious adverse events were infrequent (67 events) and similar between groups (34 in the peer support group, 33 in the care as usual group). Threat to life (self-harm) was the most common serious adverse event (35 [52%] of 67 serious adverse events). 391 other adverse events were reported, with self-harm (not life threatening) the most common (189 [48%] of 391).

INTERPRETATION

One-to-one peer support for discharge from inpatient psychiatric care, plus care as usual, was not superior to care as usual alone in the 12 months after discharge. This definitive, high-quality trial addresses uncertainty in the evidence base and suggests that peer support should not be implemented to reduce readmission post-discharge for patients at risk of readmission. Further research needs to be done to improve engagement with peer support in high-need groups, and to explore differential effects of peer support for people from different ethnic communities.

FUNDING

UK National Institute for Health Research.

Impact of Compulsory Admission on Treatment and Outcome: A Propensity Score Matched Analysis

Background

Despite multiple ethical issues and little evidence of their efficacy, compulsory admission and treatment are still common psychiatric practice. Therefore, we aimed to assess potential differences in treatment and outcome between voluntarily and compulsorily admitted patients.

Methods

We extracted clinical data from inpatients treated in an academic hospital in Zurich, Switzerland between January 1, 2013 and December 31, 2019. Observation time started upon the first admission and ended after a one-year follow-up after the last discharge. Several sociodemographic and clinical characteristics, including Health of the Nation Outcome Scales (HoNOS) scores, were retrospectively obtained. We then identified risk factors of compulsory admission using logistic regression in order to perform a widely balanced propensity score matching. Altogether, we compared 4,570 compulsorily and 4,570 voluntarily admitted propensity score-matched patients. Multiple differences between these groups concerning received treatment, coercive measures, clinical parameters, and service use outcomes were detected.

Results

Upon discharge, compulsorily admitted patients reached a similar HoNOS sum score in a significantly shorter duration of treatment. They were more often admitted for crisis interventions, were prescribed less pharmacologic treatment, and received fewer therapies. During the follow-up, voluntarily admitted patients were readmitted more often, while the time to readmission did not differ.

Conclusions

Under narrowly set circumstances, compulsory admissions might be helpful to avert and relieve exacerbations of severe psychiatric disorders.

Sociodemographic, Circumstantial, and Psychopathological Predictors of Involuntary Admission of Patients with Acute Psychosis

Studies have consistently determined that patients with acute psychosis are more likely to be involuntarily admitted, although few studies examine specific risk factors of involuntary admission (IA) among this patient group. Data from all patients presenting in the psychiatric emergency department (PED) over a period of one year were extracted. Acute psychosis was identified using specific diagnostic criteria. Predictors of IA were determined using logistic regression analysis. Out of 2533 emergency consultations, 597 patients presented with symptoms of acute psychosis, of whom 118 were involuntarily admitted (19.8%). Involuntarily admitted patients were more likely to arrive via police escort (odds ratio (OR) 10.94) or ambulance (OR 2.95), live in a psychiatric residency/nursing home (OR 2.76), report non-adherence to medication (OR 2.39), and were less likely to suffer from (comorbid) substance abuse (OR 0.53). Use of mechanical restraint was significantly associated with IA (OR 13.31). Among psychopathological aspects, aggressiveness was related to the highest risk of IA (OR 6.18), followed by suicidal intent (OR 5.54), disorientation (OR 4.66), tangential thinking (OR 3.95), and suspiciousness (OR 2.80). Patients stating fears were less likely to be involuntarily admitted (OR 0.25). By understanding the surrounding influencing factors, patient care can be improved with the aim of reducing the use of coercion.

Predicting involuntary hospitalization in psychiatry: A machine learning investigation

Background

Coercion in psychiatry is a controversial issue. Identifying its predictors and their interaction using traditional statistical methods is difficult, given the large number of variables involved. The purpose of this study was to use machine-learning (ML) models to identify socio-demographic, clinical and procedural characteristics that predict the use of compulsory admission on a large sample of psychiatric patients.

Methods

We retrospectively analyzed the routinely collected data of all psychiatric admissions that occurred between 2013 and 2017 in the canton of Vaud, Switzerland (N = 25,584). The main predictors of involuntary hospitalization were identified using two ML algorithms: Classification and Regression Tree (CART) and Random Forests (RFs). Their predictive power was compared with that obtained through traditional logistic regression. Sensitivity analyses were also performed and missing data were imputed through multiple imputation using chain equations.

Results

The three models achieved similar predictive balanced accuracy, ranging between 68 and 72%. CART showed the lowest predictive power (68%) but the most parsimonious model, allowing to estimate the probability of being involuntarily admitted with only three checks: aggressive behaviors, who referred the patient to hospital and primary diagnosis. The results of CART and RFs on the imputed data were almost identical to those obtained on the original data, confirming the robustness of our models.

Conclusions

Identifying predictors of coercion is essential to efficiently target the development of professional training, preventive strategies and alternative interventions. ML methodologies could offer new effective tools to achieve this goal, providing accurate but simple models that could be used in clinical practice.

Factors Associated With Involuntary Psychiatric Hospitalization of Youths in China Based on a Nationally Representative Sample

Objective: This nationally representative sample investigates demographic, diagnostic and clinical features associated with both voluntary and involuntary psychiatric hospitalization among children and adolescents psychiatrically hospitalized in China. Method: As part of an official national survey, 41 provincial tertiary psychiatric hospitals in China were selected. Data from 196 children and adolescents who were discharged from these psychiatric hospitals from March 19 to 31, 2019 were retrieved and analyzed. Results: 1. Psychotic symptoms, depressive symptoms and self-injury/suicide were the most common reasons of admission. Girls were significantly likely to be admitted due to depressive symptoms, whereas boys were more likely to be admitted due to aggressive behaviors. 2. The overall rate of involuntary admission was 32.1% (N = 63). Compared to patients who were admitted voluntarily, those who were admitted involuntarily had lower GAF scores on admission, were older, were more likely to present with psychotic symptoms, manic symptoms or aggressive behavior as primary reason for admission, were less likely to present with depressive symptoms, had a significantly longer length of stay, were more likely to be diagnosed with schizophrenia and were less likely to be diagnosed as depressive disorder. 3. A logistic regression showed that depressive symptom as primary reason for admission was significantly associated with voluntary admission (OR = 0.159, p < 0.001), along with two other factors: age (p < 0.01) and a lower GAF score at admission (p < 0.001) were significantly associated with involuntary admission. Conclusion: The rate of involuntary psychiatric hospitalization among children and adolescents is higher in China than in other regions. Developing more specific and more operational criteria to guide involuntary psychiatric admission for child and adolescent patients is of urgency and great importance to ensure appropriate treatment of these patients and protect their rights.