Table 2.
Characteristics of included studies.
| Article | Design and methods | Intervention vs comparator | Explored outcomes | Results and conclusions |
|---|---|---|---|---|
| Huf et al. 2012, Brazil (44) | - Unblinded RCT, 14-day follow-up - 105 agitated psychotic patients (54 secluded, 51 restrained) - Dg (restrained vs secluded): 82.3 vs 77.8% psychosis (SD or mania), 5.9 vs 11.1% psychological agitations, 11.8 vs 11.1% SU |
Seclusion vs restraint | - Effectiveness - Adverse events - Subjective perception |
- Negative effect - 2/3 secluded patients fully managed with seclusion, 1/3 changed to restraint - No significant difference between groups in effects, adverse events, or patients’ satisfaction - Ccl: Suggestion to begin with seclusion that seems not to harm or prolong coercion |
| Bergk et al. 2011, Germany (32) | - Unblinded RCT - 102 patients (12 randomized/48 nonrandomized secluded, 14 randomized/28 nonrandomized restrained Semi-structured interview - Dg (randomized vs nonrandomized secluded/randomized vs nonrandomized restrained): 50 vs 71/86 vs 50% SD, 50 vs 8/14 vs 25% AD, 0 vs 21/0 vs 25% PD |
Seclusion vs restraint | - Symptom intensity - Levels of needed medication - Adverse events - Subjective perception |
- Negative effect - No significant differences for adverse events and subjective experience - Levels of medication and aggressive symptoms are only significantly lower for nonrandomized secluded patients - Ccl: Clinical decisions should take patients’ preferences into account. RCTs on coercion are feasible |
| Vaaler et al. 2005, Norway (33) | - Non-inferiority RCT - 25 secluded patients in a traditional manner; 31 in a redecorated room - Dg (new interior vs traditional interior): 51.6 vs 24% SD, 16.1 vs 28% AD, 16.1 vs 24% SU, 6.5 vs 4% OD and 9.7 vs 2% O |
Seclusion | - Ward environment - Length of stay - Symptom intensity - Subjective perception |
- Negative and beneficial effects - No significant differences between groups - Ccl: No negative effects of a refurnished room on seclusion efficacy |
| Cashin 1996, Australia (45) | - Prospective quasi-experimental study - 53 involuntary admissions (27 secluded patients, 26 non-secluded) - No diagnostic information but no significant difference between groups |
Seclusion vs non-exposure | Time to emergency resolution Levels of needed medication |
Beneficial effect No significant differences between groups Ccl: Seclusion may be the most effective choice in some circumstances |
| Hafner et al. 1989, Australia (46) | - 38-weeks multi-center prospective study - 30 secluded and 60 non-secluded patients - Dg (secluded, no difference between groups): 46.3 (vs 23% non-secluded) SD, 12.2% BPD, manic state, 12.2% MDD, 9.8% OD, 7.3% PD, 9.8% SU, 2.4% BRP |
Seclusion vs non-exposure | - Levels of needed medication - Length of stay - Readmission rate |
- Negative and beneficial effects - 25% more neuroleptic medication for secluded patients, suggesting that seclusion did not permit to reduce the levels of medication required to manage psychiatric agitation - Less medication for non-secluded patients, suggesting that secluding agitated patients may reduce the unit level of dangerousness - No differences in length of stay or readmission rate, suggesting no adverse effect of seclusion |
| Georgieva et al. 2012, Netherlands (47) | - 3-year prospective study - 125 coerced patients (62 secluded, 18 forced medicated, 34 secluded and forced medicated, 11 secluded and restrained) - Structured questionnaires - Dg (secluded/involuntary treated/secluded and treated/secluded and restrained): 27/39/53/60% SD, 34/33/38/10% AD, 9/33/9/0% PD, 32/28/13/30% SU, 5/0/6/0% PTSD |
Seclusion and restraint vs other coercive measures | - Effectiveness - Adverse events - PTSD - Subjective perception |
- Negative effect - Combined seclusion and restraint with higher psychological and physical burden than seclusion alone or seclusion and forced treatment - No significant difference in effectiveness - Ccl: Forced medication seems better tolerated. Seclusion and/or restraint could give revival of previous traumatism or PTSD |
| Soininen et al. 2013b, Finland (48) | - 1-year prospective study - 36 secluded or restrained (no distinction) patients, 228 non-exposed - Structured questionnaire - Dg (secluded vs non-secluded): 54 vs 33% SD, 31 vs 49% AD, 14 vs 18% O |
Seclusion and restraint vs non-exposure | Quality of life | - Beneficial effect - Exposed patients reported a better subjective quality of life at discharge compared to non-exposed patients - Ccl: Either seclusion and restraint had only short-term negative influence on quality of life, or the observed association may not be causal |
| McLaughlin et al. 2016, 10 European countries (34) | - Multi-center prospective study (EUNOMIA project) - 2,030 involuntary admissions, 770 with one or more coercive measures (84 secluded, 439 restrained, 556 forced medication). - 1,353 interviews - Dg (coerced vs non coerced): 68 vs 60% SD |
Seclusion and restraint vs other coercive measures | Length of stay | - Negative and beneficial effects - At 3 months, 843 involuntary admitted patients approved and 506 (37.4%) disapproved their previous admission. Forced medication was the only significant measure associated with admission disapproval - Seclusion and restraint were associated with increased length of stay (in multivariate analysis, only seclusion remains significant). Secluded patients’ symptom intensity did not fully explain the observed increase |
| Soloff et Turner 1981, US (49) | - 8-month prospective study - 59 secluded patients, 159 non-secluded - Structured questionnaire - Dg (secluded vs non-secluded): 42.4 vs 40.9% SD, 5.1 vs 1.9% BPD, 11.9 vs 11.3% other AD, 6.8 vs 4.4% OD, 8.5 vs 12.6% PD, 0 vs 11.3% neurosis, 23.7 vs 17.6% O (SU and MR) |
Seclusion vs non-exposure | Length of stay | - Beneficial effect - Length of stay associated with incidence of seclusion, but no influence of chronicity and legal status at admission - Initial postulate: Seclusion as therapeutic and control function for patient and ward milieu |
| Schwab et Lahmeyer 1979, US (50) | - 6-month prospective study - 52 secluded patients, 90 non-secluded - Dg (secluded vs non secluded): 29 vs 29% SD, 19 vs 7% BPD, manic state, 14 vs 14% psychotic MDD, 14 vs 32% neurosis, 8 vs 3% SU, 6 vs 3% PD, 10 vs 12% O |
Seclusion vs non-exposure | Length of stay | Negative effect Increased length of stay for secluded patients |
| Mattson et Sacks 1978, US (51) | - 1-year prospective study - 63 secluded patients, 160 non-secluded - Dg (secluded vs non secluded): 63 vs 38% SD, 17 vs 4% BPD, manic state, 10 vs 14% PD, 10 vs 44% O |
Seclusion vs non-exposure | Length of stay | - Negative effect - Increased length of stay for secluded patients - Effect no longer significant when focusing on patients less than 20 years of age |
| Hammill et al. 1989, US (52) | - Prospective study - 100 patients (26 secluded, 74 non-secluded) with SD or SAD - Semi-structured interview |
Seclusion vs non-exposure | - Length of stay - Subjective perception |
- Negative and beneficial effects - Increased length of stay for secluded patients - 13/17 secluded patients evaluated seclusion as necessary |
| Plutchik et al. 1978, US (53) | - 2 prospective studies - 1st: descriptive (118 secluded patients, 118 randomly assessed non-secluded) - 2nd: qualitative (30 secluded and 25 non-secluded patients) - Structured interview - Dg (secluded vs non secluded): 64 vs 45.8% SD, 2.5 vs 0% BPD, manic state, 3.4 vs 8.5% psychotic MDD, 10.2 vs 13.6% depressive neurosis, 0.8 vs 5.1% SU, 6.8 vs 13.6% PD, 5.9 vs 8.5% adjustment reactions, 3.4 vs 5.1% OD, 2.5 vs 0% MR |
Seclusion vs non-exposure | - Length of stay - Subjective perception |
- Negative and beneficial effects - 1st study: Increased length of stay for secluded patients - 2nd study: 40% secluded patients rated seclusion as not helpful. 60% reported feeling better after seclusion |
| Mann et al. 1993, US (54) | - 6-month prospective study - 50 secluded patients - Structured questionnaire - Dg: 24% MDD, 10% dysthymic disorders, 30% BPD, 2% SAD, 16% SD, 6% BRP, 8% SU, 4% none |
Seclusion | - Length of stay - Subjective perception |
- Negative and beneficial effects - Seclusion safe and secure (67%) - Feelings of constant attention and care from staff (45%) - Increased length of stay for secluded patients (compared to general unit mean) |
| Ishida et al. 2014, Japan (55) | - Prospective study - 190 restrained patients - Dg: 3.9% OD, 9.9% SU, 63.5% SD, 14.9% AD, 1.1% somatoform disorders, 6.6% PD |
Mechanical restraint | Adverse effects | - Negative effect - D-dimer augmentation for 72 restrained patients with prophylaxis. - US Doppler of lower extremities showed asymptomatic DVT in 21 patients (11.6%) - Incidence of DVT associated with excessive sedation, longer duration of restraint, lower antipsychotic dosage - Ccl: Probable underestimation of DVT in routine use of restraint |
| Steinert et al. 2013, Germany (56) | - Cross-sectional study, 1-year follow-up after Bergk et al. 2011 - 60 of 102 (59%) previous patients (31 secluded, 29 restrained) - Dgs: 63% SD, 23% BPD, 14% O |
-Seclusion vs restraint | - PTSD - Subjective perception |
- Negative and beneficial effects - Seclusion reported as less restrictive - 1 secluded and 2 restrained patients with symptoms fulfilling PTSD diagnosis - Ccl: The lower than expected incidence of PTSD may be due to natural resolution of symptoms or to the interviews conducted with the patients, which could have helped prevent PTSD |
| Guzmán-Parra et al. 2018, Spain (57) | - 2-year prospective study - 111 coerced patients (32 restrained, 41 forced medicated, 38 forced medicated and restrained) - Dg (restrained vs involuntary treated vs combined): 4.9 vs 9.4 vs 10.5% SU, 58.5 vs 50 vs 68.4% SD, 22 vs 28.1 vs 18.4% AD, 2.4 vs 3.1 vs 0% anxiety disorders, 7.3 vs 6.3 vs 0% PD, 4.9 vs 3.1 vs 2.6% O |
Mechanical restraint vs forced medication | - PTSD - Subjective perception |
- Negative effect - Higher perceived coercion with restraint (compared to forced medication). - Higher post-traumatic stress with forced medication - Combined forced medication and restraint associated with higher coercion perception and less treatment satisfaction (than restraint or forced medication alone) |
| Steinert et al. 2007, Germany (58) | - Prospective study - 117 involuntary admissions with history of seclusion or restraint, 18 secluded or restrained (no distinction) patients at present admission - Structured questionnaires - Dg: 79.5% SD 8.5% other psychotic disorders, 12% SAD |
Seclusion and restraint vs non-exposure | - Influence of history of life-threatening events on traumatic effects of intervention | - Negative effect - Bidirectional association of history of seclusion or restraint with life-threatening traumatic events. - Exposure to past traumatic events enhances the risk of revictimization and revival of previous traumatism during inpatient treatment - Ccl: Coercive measures may cause re-experienced traumatism |
| Wallsten et al. 2008, Sweden (37) | - 2-year prospective study - 115 patients (19 reported mechanically restrained but 8 false positives; 98 reported non-restrained but 4 false negatives); 15 truly restrained - Structured interview - Dg (true positives/true negatives/false positives/false negatives): 46/52/38/25% SD, 36/9/63/25% AD, 18/19/-/50% O |
Mechanical restraint vs non-exposure | - Discrepancy between objective and reported coercion - Subjective perception |
- Negative effect - 42% false positive and 4% false negative reports of restraint. - Causes are not clear [communication problem, memories failures (or false memories), or emotional traumatic reactivation] - Ccl: Subjective quality of reports of past traumatic events |
| Whitecross et al. 2013, Australia (59) | - 9-month prospective study - 31 secluded patients - Dg: 51.6% SD, 32.3% SAD, 16.1% O |
Seclusion | PTSD | - Negative effect - 47% probable PTSD (IER-S >33) after seclusion |
| Fugger et al. 2015, Austria (35) | - 18-month prospective study - 47 mechanically restrained patients - Dg: 23.4% OD, 12.8% SU, 19.1% paranoid SD, 8.5% catatonic SD, 4.2% SAD, manic state, 14.9% BPD, manic episode, 2.1% BPD, mixed episode, 2.1% recurrent MDD, 6.4% anorexia, 6.4% PD |
Mechanical restraint after intervention | - PTSD - Subjective perception |
- Negative and beneficial effects - 50% high perceived coercion and 25% probable PTSD - Less memory event, more feeling of being healthy and more acceptance of restraint than rated by physicians |
| Palazzolo 2004, France (60) | - 6-month prospective study - 67 secluded patients - Semi-structured interview - Dg: 32.8% SD, 28.4% BPD, 14.9% BRP, 10.4% SAD, 5.9% anorexia, 4.6% somatoform disorders, 3% antisocial PD |
Seclusion | - Hallucinations - Subjective perception |
- Negative and beneficial effects - Anger was the most frequent reported emotion - 31% reported hallucinatory experience - 67% reported anxiety - 8% reported feeling better, and 8% the necessity of continuing treatment |
| Kennedy et al. 1994, US (61) | - Prospective study - 25 secluded patients with SD or SAD - Semi-Structured interview |
Seclusion | - Hallucinations - Subjective perception |
- Negative and beneficial effects - For 48%, seclusion was not helpful - 52% reported hallucinations during seclusion - 70% who experienced hallucinations during seclusion were hallucinating before seclusion but proportional increase of hallucinations during seclusion was not significant - Hallucinating patients had longer (but not significantly) seclusion stay, more therapeutic interaction (nurse-patient relationship) and levels of needed medication |
| Sagduyu et al. 1995, US (62) | - Prospective study - 25 secluded and 25 restrained patients - Semi-structured interview - 76% restrained and 80% secluded patients had a SD |
Seclusion vs Restraint | Subjective perception | - Negative and beneficial effects - 40% secluded and 20% restrained with positive evaluation - 71% secluded and 89% restrained remembered past traumatic experiences - 73% secluded and 81% restrained reported negative feelings |
| Krieger et al. 2018, Germany (36) | - 18-month prospective study, - 213 involuntary admitted patients (78 mechanically restrained, 32 secluded, 30 forced medicated, 20 video monitored) - 51 voluntarily admitted patients in a closed ward, - Structured interview - Dg (coerced vs control groups): 71.1 vs 51% SD, 10 vs 21.6% SU, 12.8 vs 19.6% AD, 3.3 vs 7.8% PD, 33.6 vs 45.1% of comorbidities with SU |
Seclusion and restraint vs other coercive measures | Subjective perception | - Negative and beneficial effects - Negative emotions associated with seclusion or restraint - Increasing understanding of use of seclusion or restraint during hospitalization - Seclusion preferred among all coercive measures, while restraint less accepted than the other measures |
| Gowda et al. 2018, India (63) | - Prospective study - 200 patients (40 mechanically or manually restrained, 36 secluded, 116 chemical restrained, 64 involuntarily treated, 29 ECT) - Dg: 48% SD, 43.5% AD, 18.5% O, 48.5% comorbidities with SU |
Seclusion and restraint vs other coercive measures | Subjective perception at admission and discharge | Negative effect Physical restraint associated with a greater perception of coercion, followed by involuntary treatment, chemical restraint, seclusion and finally ECT |
| Sorgaard 2004, Norway (64) | - 17-week prospective interventional study - 190 admissions (16% secluded, 160 non-secluded) - Standardized questionnaires - Dg (baseline vs project phase): 26.8 vs 28.6% SD, 53.6 vs 41.2% AD, 3.6 vs 5.0% PD, 8.9 vs 11.8% SU, 7.1 vs 13.6% O |
Seclusion vs non-exposure | - Adverse events - Subjective perception |
- Negative effect - Seclusion as principal factor associated with perceived coercion (compared to age, sex, forced medication, or length of stay) |
| Martinez et al. 1999 (65) | - Cross-sectional study - 69 patients (53 secluded, 16 non-secluded) - Semi-structured interview - No diagnostic information |
Seclusion vs non-exposure | Subjective perception | - Negative and beneficial effects - Negative perception of seclusion (62% overuse, 76.5% punishment) - 56.2% reported seclusion as needed |
| Larue et al. 2013, Canada (66) | - 1-year prospective study - 50 secluded or restrained (no distinction) patients - Semi-structured interview - Dg: 66% SD, 30% AD, 2% PD, 2% anxious disorders |
Seclusion and restraint | Subjective perception | - Beneficial effect - 52% agreed with improved behavior after seclusion |
| Soininen et al. 2013a, Finland (67) | - 18-month multi-center prospective study - 90 secluded or restrained patients (no distinction) - Structured questionnaire - Dg: 12% SU, 60% SD, 20% AD, 6% PD |
Seclusion and restraint | Subjective perception after intervention | - Negative effect - Deny necessity and beneficence of seclusion or restraint - Dissatisfaction - Not enough dialogue |
| Keski-Valkama et al. 2010, Finland (68) | - 1-year prospective study - 38 secluded patients in general vs 68 in forensic wards - Structured interview - Dg in general wards: 71.1% SD, 10.5% SU, 15.8% AD, 2.6% O |
Seclusion | Subjective perception | - Negative and beneficial effects - Mostly negative feelings, loneliness - Need for interaction - Seclusion as necessary - 54% secluded patients perceived seclusion as a punishment |
| Stolker et al. 2006, Netherlands (69) | - 18-month prospective study - 78 secluded patients - Structured interview - Dg: 67% SD, 11% BPD, 11% cluster B PD |
Seclusion | - Ward environment - Subjective perception |
- Negative and beneficial effects - Staying in multi-bed rooms prior to seclusion associated with less negative views of seclusion |
| Richardson et al. 1987, US (70) | - 1-year prospective study - 52 secluded patients - Semi-structured interview - Dg: 36.5% SD, 28.8% SAD, 19.2% AD, 9.6% atypical psychosis, 1.9% borderline PD, 1.9% organic hallucinosis, 1.9% dementia |
Seclusion | Subjective perception | - Negative and beneficial effects - 31% patients reported anger, 58% felt punished - 50% reported seclusion as protection, 48% as necessary - 37% reported hallucinatory experience - 20/52 reported improvement after seclusion, 8/52 deterioration |
| Binder et McCoy 1983, US (71) | - 8-month prospective study - 27 secluded patients - Semi-structured interview - Dg: 45.8% SD, 33.3% AD, 8.3% SAD, 8.3% antisocial PD, 4.2% acute paranoid BRP |
Seclusion | Subjective perception | - Negative and beneficial effects - 4 patients rated seclusion as therapeutic, 12 as necessary - 11 rated beneficial aspects (7 hypostimulation) - 18 negative emotions - For 14, seclusion had no effect, 3 beneficial effect, 2 negative effect, 5 first negative effect changed to beneficial effect |
| Tooke et Brown 1992, US (72) | - 11-week prospective study - 19 secluded patients (11 locked rooms, 8 secluded area) - Structured questionnaire - Dg: 47.3% SD, 26.3% MDD or suicidal ideations |
Seclusion | Subjective perception | - Negative effect - 73% secluded patients (in locked rooms) felt punished - Strong negative feelings |
vs, versus; RCT, randomized controlled trial; Dg, diagnoses; Ccl, conclusions; SD, schizophrenic disorders; AD, affective disorders; PD, personality disorders; SAD, schizoaffective disorder; BRP, brief reactive psychosis; SU, substance use; O, others; BPD, bipolar disorders; MDD, major depressive disorder; OD, organic disorders; MR, mental retardation; PTSD, post-traumatic stress disorder; EUNOMIA, European Evaluation of Coercion in Psychiatry and Harmonization of Best Clinical Practice; US, ultrasound; DVT, deep vein thrombosis; IES-R, Impact of Event Scale-Revised; ECT, Electro-convulsive therapy.