Table 1.
Characteristics of included studies
| Study type | Setting | Sample size | Sample description | Number of involuntary admissions (% of all inpatients) | Social and clinical correlates extracted for analysis | Quality of study | |
|---|---|---|---|---|---|---|---|
| Aguglia et al (2016)26 | Cohort | Italy | 730 | Consecutive admissions to the psychiatric inpatient unit of the San Luigi Gonzaga Hospital, Orbassano, Italy, from September, 2013, to August, 2015 | 112 (15·3%) | Age, gender, education level, relationship status, diagnosis | Moderate |
| Balducci et al (2017)27 | Cohort | Italy | 848 | Consecutive admissions to the psychiatric inpatient unit of the general teaching hospital of Santa Maria della Misericordia, Perugia, Italy, from June, 2011, to June, 2014 | 309 (36·4%) | Age, gender, relationship status, diagnosis, taking medication at the time of admission, more than one hospitalisation, risk to self | High |
| Bauer et al (2007)28 | Cohort | Israel | 34 799 | National psychiatric case registry of the Israeli Ministry of Health used to identify all adult inpatient psychiatric admissions to hospital between 1991 and 2000 | 11 156 (32·1%) | Gender, diagnosis, relationship status, years of education, risk to self | High |
| Beck et al (1984)29 | Cohort | USA | 300 | Random sample of voluntary and involuntary admissions to three adult inpatient units in the US State of Missouri over three periods (January, 1978, to June, 1978; January, 1979, to June, 1979; and January, 1980, to June, 1980) | 150 (selected control group) | None | Low |
| Bindman et al (2002)20 | Ecological | England | About 1·71 million | Purposive sample of eight mental health provider trusts in England from 1998 to 1999 | 1507 (voluntary admissions not recorded) | Number of inpatient beds, availability of less restrictive care, area deprivation | Moderate |
| Blank et al (1989)30 | Cohort | USA | 274 | All patients aged 55 years and older admitted to an old age psychiatric unit in a non-profit teaching hospital in the US State of New York from November, 1984, to December, 1985 | 75 (27·3%) | Gender, relationship status, living situation, diagnosis, risk to others, risk to self, presentation | High |
| Bonsack and Borgeat (2005)31 | Cross-sectional | Switzerland | 87 | Self-completed questionnaire given to all inpatients of the psychiatric hospital of the University of Lausanne, Lausanne, Switzerland, on May 10, 2002 (response rate 96%) | 30 (34·5%) | Gender, diagnosis | Low |
| Bruns (1991)32 | Cohort | Germany | 628 | Patients who were involuntarily admitted into the psychiatric unit of Hospital Bremen-Ost in Germany, and 300 randomly chosen controls who were voluntarily admitted between 1984 and 1985 | 328 (selected control group) | Gender, relationship status | Low |
| Burnett et al (1999)33 | Cohort | England | 100 | First admissions with psychosis within southeast London in England from April, 1991, to March, 1993 | 28 (28%) | Pathways to care | Moderate |
| Canova Mosele et al (2018)17 | Cohort | Brazil | 137 | Admissions to the psychiatry service of the University Hospital of Santa Maria in Brazil from August, 2012, to January, 2013 | 71 (51·8%) | Gender, living situation, occupation, relationship status, presentation, pathways to care, risk to self, risk to others, education level | High |
| Casella and Loch (2014)34 | Cohort | Brazil | 169 | Consecutive discharges from the Philippe Pinel Psychiatric Hospital in Brazil from May, 2009, to August, 2009; those with diagnoses other than psychosis or bipolar affective disorder were excluded | 81 (48%) | Gender, relationship status, diagnosis, previous admission, presentation, correct use of medication before admission, risk to others, risk to self, social support | Moderate |
| Chang et al (2013)35 | Cohort | Brazil | 2289 | All adults hospitalised at the Institute of Psychiatry of the Clinical Hospital, University of San Paulo, Brazil, between 2001 and 2008 | 305 (13·3%) | Gender, employment, relationship status, education level, diagnosis, adherence to treatment before admission | Low |
| Chiang et al (2017)36 | Cohort | Taiwan | 26 611 | All first admissions with psychosis in Taiwan between 2004 and 2007, identified using the national health insurance database | 2540 (9·5%) | Gender, employment, previous admission | High |
| Cole et al (1995)37 | Cohort | England | 93 | People with first-onset psychosis in the catchment area for St Ann's Hospital in London, England, between July 1, 1991, and June 30, 1992 | 29 (31%) | Age, living situation, employment, pathways to care, social support | Moderate |
| Cougnard et al (2004)38 | Cohort | France | 86 | Consecutive admissions with psychosis in ten departments of psychiatry in the Bordeaux region of France between March, 2001, and March, 2002 | 53 (61·6%) | Age, gender, living situation, employment, relationship status, diagnosis, presentation, pathways to care, risk to self, criminal history, social support, educational level | High |
| Craw and Compton (2006)39 | Cohort | USA | 227 | Consecutively discharged patients from a large public sector hospital in the US State of Georgia from December, 2003, to July, 2004 | 171 (75·3%) | Age, gender, living situation, employment, relationship status, previous psychiatric hospitalisation, presentation | High |
| Crisanti and Love (2001)40 | Cohort | Canada | 1718 | Admissions to the Department of Psychiatry at the Calgary General Hospital in Alberta, Canada between April 1, 1987, and March 31, 1995 | 711 (41·4%) | Gender, diagnosis, criminal history | High |
| Curley et al (2016)16 | Cohort | Ireland | 1099 | All admissions to St Aloysius Ward, an acute adult psychiatric inpatient facility in north Dublin, Ireland, between Jan 1, 2008, and Dec 31, 2014 | 155 (14·1%) | Area deprivation (other variables repeated in Kelly et al [2018])65 | High |
| de Girolamo et al (2009)41 | Cross-sectional national survey | Italy | 1548 | All patients admitted to public or private inpatient facilities in Italy (excluding Sicily) during a 12-day period in 2004 | 196 (12·6%) | Gender, housing status, employment status, relationship status, diagnosis, availability of less restrictive care, presentation, referral pathway, risk to self, risk to others, criminal history, educational level | Moderate |
| Delayahu et al (2014)42 | Cohort | Israel | 24 | Men aged 18–60 years with a DSM-IV axis I diagnosis and substance abuse disorder who were hospitalised in an acute psychiatric dual diagnosis ward in Israel between February, 2004, and March, 2004, and between May, 2004, and June, 2004* | 9 (37·5%) | Age, relationship status, presentation on admission, risk to self, educational level | Moderate |
| Di Lorenzo et al (2018)43 | Cohort | Italy | 396 | All patients admitted to an acute psychiatric ward in northern Italy between Jan 1, 2015, and Dec 31, 2015 | 160 (40%) | Gender, living arrangements, diagnosis, employment situation, risk to self, risk to others | Moderate |
| Donisi et al (2016)44 | Cohort | Italy | 74 931 | All discharges from the 40 acute inpatient facilities in the Vento region of Italy between 2000 and 2007 | 3975 (5·3%) | Referral pathway | High |
| Emons et al (2014)45 | Cohort | Germany | 230 678 | All admissions to the largest provider of psychiatric services in Germany (Landschaftsverbands Westfalen-Lippe) from 2004 to 2009 | 17 206 (7·5%) | Area deprivation, availability of less restrictive care | Moderate |
| Eytan et al (2013)46 | Cohort | Switzerland | 2227 | All admissions to an acute psychiatric facility in Switzerland over an 8-month period in 2006 | 1422 (63·9%) | None | Moderate |
| Fok et al (2014)47 | Cohort | England | 14 233 | Adult patients with severe mental illness with and without co-morbid personality disorder between Jan 1, 2007, and Dec 31, 2011 | 3748 (26%) | None | High |
| Folnegovic-Smalc et al (2000)48 | Cohort | Croatia | 888 | All admitted patients to two acute facilities in Croatia from Jan 1, 1997, to June 30, 1997 | 173 (19%) | Gender, diagnosis | Moderate |
| Gaddini et al (2008)49 | Cross-sectional | Italy | 7984 | All adult inpatients in 369 psychiatric facilities across Italy (excluding Sicily) on May 8, 2003 | 305 (3·8%) | None | Moderate |
| Garcia Cabeza et al (1998)50 | Cross- sectional | Spain | 367 | All patients admitted to the acute unit at the psychiatric service of the hospital Gregorio Marañon in Madrid, Spain, in the first 4 months of 1994 | 67 (18%) | Gender, relationship status, employment status, living arrangements, diagnosis, pathways to care | Moderate |
| Gou et al (2014)51 | Cohort | China | 160 | Consecutive admissions to an acute psychiatric facility in China between July 26, 2012, and Sept 10, 2012 | 85 (53·1%) | Age, gender, employment, relationship status, diagnosis, presentation on admission, education | High |
| Gultekin et al (2013)52 | Cohort | Turkey | 504 | Patients admitted to an acute psychiatric facility in Turkey between May 1, 2010, and Oct 31, 2010, who had been discharged at the time of data collection | 66 (13·1%) | Gender, employment, relationship status, diagnosis, education | High |
| Hansson et al (1999)53 | Cohort | Denmark, Finland, Norway, and Sweden | 2834 | All new patients contacting the psychiatric services in seven catchments areas over a 1-year period | 219 (7·7%) | None | Moderate |
| Hatling et al (2002)54 | Cohort | Norway | 13 985 | Patients admitted to psychiatric facilities in general hospitals in Norway in 1996 | 6476 (46·3%) | Gender, employment, relationship status, diagnosis, availability of inpatient beds | Moderate |
| Hoffman et al (2017)55 | Cohort | Germany | 213 595 | All admissions to the largest provider of psychiatric services in Germany (Landschaftsverbands Westfalen-Lippe) from 2004 to 2009 | 17 206 (8·1%) | Gender, relationship status, diagnosis, referral pathway, previous admission | Moderate |
| Hotzy et al (2019)56 | Cohort | Switzerland | 31 508 | Includes all admissions to the University Hospital of Psychiatry in Zurich, Switzerland, between 2008 and 2016; the number of admissions per patient ranged from one to ten (median two [IQR one to three]) | 8843 (28·1%) | Gender, diagnosis, education level | Moderate |
| Houston et al (2001)57 | Cohort | USA | 487 | First admissions (unclear where to) between October, 1986, and December, 1990 | 282 (58%) | None | Low |
| Hugo (1998)58 | Cohort | Australia | 402 | Inpatient admissions to an acute ward in Australia over an 8-month period | 136 (34%) | Diagnosis, presentation, risk to self, risk to others | Low |
| Hustoft et al (2012)59 | Cohort | Norway | 3326 | Consecutive admissions to 20 acute psychiatric units in Norway from 2005 to 2006 | 1453 (44%) | Gender, housing stability, employment, relationship status, presentation on admission, referral pathway, education level, risk to self, risk to others | Moderate |
| Ielmini et al (2018)60 | Cohort | Italy | 200 | 200 adult psychiatric inpatients hospitalised at the General Hospital Psychiatric Ward in Varese, Italy, from January, 2014, to March, 2017 | 100 (selected control group) | Age, gender, housing stability, employment, relationship status, presentation on admission, risk to others, having social support | Moderate |
| Indu et al (2018)61 | Case-control | India | 300 | Consecutive compulsory admissions and the two following voluntary admissions to the Indian Government's mental health centre in Thiruvananthapuram, Kerala, from June, 2010, to February, 2011 | 100 (33%) | Gender, housing stability, employment status, relationship status, diagnosis, previous involuntary admission, presentation, compliance, having social support, education level | Moderate |
| Isohanni et al (1991)62 | Case-control | Finland | 1586 | Admissions to a closed psychiatric ward with modified therapeutic community principles in Oulu, Finland, between 1978 and 1987 | 215 (13·6%) | Age, diagnosis, previous admission | Moderate |
| Iversen et al (2002)63 | Cohort | Norway | 223 | All patients admitted to four acute wards in Norway from October, 1998, to November, 1999 | 150 (67%) | Gender, diagnosis, presentation | Moderate |
| Kelly et al (2004)64 | Cohort | Ireland | 78 | Patients with first-episode psychosis admitted to two psychiatric hospitals in Dublin, Ireland, over a 4-year period | 17 (22%) | Age, gender, presentation | Moderate |
| Kelly et al (2018)65 | Cohort | Ireland | 2940 | All adult admissions to three acute psychiatric hospitals in Dublin, Ireland, from 2008 to 2015 (Dublin Involuntary Admission Study) | 423 (14·4%) | Gender, employment, relationship status, diagnosis | Moderate |
| Keown et al (2016)66 | Cohort | England | Population of 138 primary care trusts | All adult psychiatric admissions in England in 2010 and 2011; data from the Mental Health Minimum Data Set | Unclear | Area deprivation | High |
| Lastra Martinez et al (1993)67 | Cross-sectional | Spain | 298 | Clinical records of patients admitted to the acute unit of the psychiatric service of a general hospital (San Carlos University Hospital) in Madrid, Spain, between March, 1990, and February, 1991 | 148 (voluntary group is a selected control group) | Gender, relationship status, risk to self, risk to others | Moderate |
| Lay et al (2011)68 | Cohort | Switzerland | 9698 | All patients admitted to psychiatric inpatient facilities in Zurich, Switzerland, in 2007 | 2406 (24·8%) | Age, gender, housing stability, employment, diagnosis, inpatient beds, education level, presentation on admission | Moderate |
| Lebenbaum et al (2018)69 | Cohort | Canada | 115 515 | All patients admitted to mental health beds in the Canadian Province of Ontario from 2009 to 2013 | 85 607 (74·1%) | Gender, housing stability, diagnosis, previous involuntary admission, referral pathways, risk to self, risk to others, presentation on admission | High |
| Leung et al (1993)70 | Case-control | USA | 44 | Admissions of Indochinese patients to a psychiatric facility in the US State of Oregon in 1985 and 1986; all involuntary admissions were included, and the same number of voluntary patients was selected randomly | 22 (selected control group) | Gender, housing stability, employment, relationship status, diagnosis, previous involuntary admission, previous admission, education level | Moderate |
| Lin et al (2019)71 | Case-control | Taiwan | 10 190 | All inpatients in Taiwan with a principal diagnosis of schizophrenia between 2007 and 2013; all involuntary patients were included and matched to four voluntary patients based on age, gender, and year of admission | 2038 (selected control group) | Risk to self, previous admission | Moderate |
| Lorant et al (2007)72 | Cohort | Belgium | 346 | Random sample of 1200 patients referred to one of six psychiatric inpatient units in Brussels, Belgium, in 2004 | 154 (44·5%) | Age, availability of less restrictive care, compliance with treatment before admission, risk to self, risk to others | High |
| Luo et al (2019)73 | Cross-sectional | China | 155 | All patients with a diagnosis of bipolar affective disorder admitted to 16 psychiatric institutions in China in an index month (March 15, 2013, to April 15, 2013) | 81 (52%) | Gender, employment status, relationship status, education level, previous outpatient treatment, previous hospitalisation, risk to self, risk to others, presentation | Moderate |
| Malla et al (1987)74 | Cohort | Canada | 5729 | Consecutive admissions to four psychiatric facilities in the Canadian Province of Ontario between October, 1975, and October, 1978 | 724 (12·6%) | Gender, employment, relationship status, referral pathways, diagnosis, risk to self, risk to others | Moderate |
| Mandarelli et al (2014)75 | Case-control | Italy | 60 | Consecutive involuntary admissions to a psychiatric inpatient unit in Rome, Italy, between October, 2009, and April 2010; each inpatient was matched for age and sex to a voluntarily admitted patient from the same hospital over the same period | 30 (selected control group) | Relationship status, diagnosis, presentation, risk to self | Moderate |
| Montemagni et al (2011)76 | Cohort | Italy | 119 | Patients with schizophrenia consecutively admitted to an emergency psychiatric ward in Turin, Italy, between December, 2007, and December, 2009 | 34 (28·5%) | Age, gender, employment, relationship status, previous involuntary admission, presentation | Moderate |
| Montemagni et al (2012)77 | Cohort | Italy | 848 | Consecutive admissions to an emergency psychiatric ward in Turin, Italy, between January, 2007, and December, 2008 | 146 (17%) | Age, diagnosis, education level, risk to self, presentation | Moderate |
| Myklebust et al (2012)78 | Cohort | Norway | 1963 | Admissions to a psychiatric hospital in northern Norway from 2003 to 2006 | 183 (9·3%) | Age, gender, diagnosis, presentation on admission, referral pathway | Moderate |
| Okin (1986)79 | Cross-sectional | USA | 198 | All admissions to seven state psychiatric hospitals in the US State of Massachusetts over a 2-week period in 1981 | 94 (47·5%) | Gender, housing stability, diagnosis, previous admission, relationship status, education, risk to self, risk to others | Low |
| Olajide et al (2016)80 | Cohort | England | 2087 | Patients referred for a Mental Health Act assessment in London, Birmingham, or Oxfordshire in England between the months of July and October in 2008–11 | 1396 (66·9%) | Age, diagnosis, risk to self, risk to others | Moderate |
| Opjordsmoen et al (2010)81 | Cohort | Norway | 217 | Inpatients with first-episode psychosis in four psychiatric facilities in Norway from January, 1997, to December, 2000 | 126 (58·1%) | Gender, relationship status, presentation, education level | Moderate |
| Opsal et al (2011)82 | Cross-sectional | Norway | 1187 | All patients with a history of substance abuse admitted to 39 acute psychiatric wards in Norway over a 3-month period in 2005–06 | 361 (30·4%) | Gender, housing stability, employment status, diagnosis, presentation, risk to self, referral pathways | Moderate |
| Polachek et al (2017)83 | Cohort | Israel | 5411 | All patients with a diagnosis of a psychotic disorder discharged from a mental health centre between January, 2010, and April, 2013 | 2109 (39%) | Gender | Low |
| Riecher et al (1991)84 | Cohort | Germany | 10 749 | All patients admitted to psychiatric hospital in Baden-Wurttemberg, Germany, between Jan 1, 1984, and June 30, 1986 | 517 (4·8%) | Gender, housing stability, employment, relationship status diagnosis, previous admission | Moderate |
| Ritsner et al (2015)85 | Cohort | Israel | 439 | All patients admitted to the Sh'ar Menashe mental health centre in Israel between March 1, 2012, and Feb 28, 2013 | 106 (24·1%) | Age, gender, diagnosis, presentation, risk to self | Low |
| Rodrigues et al (2019)7 | Cohort | Canada | 5191 | All patients from a cohort of young people (aged 16–35 years) with a diagnosis of non-affective psychosis who were hospitalised over a 2-year follow-up period from the initial diagnosis | 4208 (84%) | Gender, living arrangements, social support, risk to self, risk to others, presentation, previous hospitalisation, adherence to treatment before hospitalisation, pathways to care | High |
| Rooney et al (1996)86 | Case-control | Ireland | 101 | Consecutive involuntary admissions to an inpatient psychiatric unit in Dublin, Ireland, over 6 months were compared with a sample of voluntary patients in the same hospital | 58 (selected control group) | Gender, diagnosis, referral pathways, risk to self, risk to others | Low |
| Schmitz-Buhl et al (2019)87 | Cohort | Germany | 5764 | All patients involuntarily hospitalised in the state of North Rhine-Westphalia in Cologne, Germany, in 2011; 3991 patients treated voluntarily in the same hospitals over the same period served as a control group | 1773 (voluntary group is a selected control group) | Education level, risk to self | Moderate |
| Schuepbach et al (2006)88 | Cohort | Switzerland | 86 | Inpatients with an acutely manic or mixed episode of bipolar disorder in the Swiss cohort of the EMBLEM study | 55 (64%) | Gender, relationship status, presentation on admission, compliance with medication before admission | Moderate |
| Schuepbach et al (2008)89 | Cross-sectional | 14 European countries | 1374 | A sample of inpatients with an acutely manic or mixed episode of bipolar disorder enrolled in the EMBLEM study | 561 (40·8%) | Gender, housing stability, relationship status, presentation, compliance with medication before admission, risk to self, education level | Moderate |
| Serfaty and McCluskey (1998)90 | Case series | England | 12 | A sample of 11 inpatients with a diagnosis of an eating disorder | 7 (58·3%) | Diagnosis, presentation | Low |
| Silva et al (2018)91 | Cohort | Switzerland | 5027 | All consecutive admissions to four psychiatric hospitals in the Canton of Vaud, Switzerland, between Jan 1, 2015, and Dec 31, 2015 | 1918 (38·2%) | Gender, relationship status, diagnosis, risk to self, risk to others, previous psychiatric hospitalisation, previous involuntary hospitalisation | High |
| Spengler (1986)92 | Cohort | Germany | 206 | Consecutive new contacts with the psychiatric emergency department who were admitted to public psychiatric hospitals in Hamburg, Germany, from January, 1980, to September, 1981 | 122 (59·2%) | Gender, housing stability, employment, relationship status, diagnosis, presentation, compliance with treatment before admission, risk to self | High |
| Stylianidis et al (2017)93 | Cohort | Greece | 715 | All patients admitted to the psychiatric hospital of Attica, Greece, from June, 2011, to October, 2011 | 427 (59·7%) | Age, gender, employment status, relationship status, diagnosis, previous admission, social support, education | High |
| Tørrissen (2007)94 | Cohort | Norway | 104 | All patients discharged from an acute ward in the Norwegian county of Hedmark from January, 2005, to June, 2005 | 49 (47%) | Age, diagnosis | Low |
| van der Post et al (2009)95 | Cohort | Netherlands | 7600 | Consecutive patients presenting to emergency psychiatric services in Amsterdam and admitted to an inpatient unit between Sept 15, 2004, and Sept 15, 2006 | 352 (46·3%) | Previous involuntary admission, referral pathway, presentation, risk to self, risk to others | Moderate |
| Wang et al (2015)96 | Cohort | Taiwan | 2777 | Admissions to psychiatric hospital from the emergency psychiatric service from January, 2009, to December, 2010 | 110 (4·0%) | Age, gender, diagnosis, presentation on admission, referral pathways, risk to self | Moderate |
| Watson et al (2000)97 | Cohort | USA | 397 | Consecutive patients with an eating disorder referred for admission in the University of Iowa hospital between July, 1991, and June, 1998 | 66 (16·6%) | Gender, relationship status | Low |
| Weich et al (2017)6 | Cross-sectional | England | 1 238 188 total sample; 104 647 inpatient admissions | All patients who received care at 64 NHS provider trusts in 2010–11; data from the Mental Health Minimum Data Set | 42 915 (3·5% of total sample, 41·0% of the inpatient sample) | Gender, area deprivation, inpatient beds, availability of less restrictive care | High |
EMBLEM=European Mania in Bipolar Longitudinal Evaluation of Medication. NHS=National Health Service.