Background: Involuntary care and coercive measures are frequently present in mental healthcare for adolescents. The purpose of this study was to examine to what extent adolescents perceive or experience coercion during inpatient mental health care, and to examine predictors of experienced coercion. Methods: A cross-sectional sample of 96 adolescent inpatients from 10 Norwegian acute and combined (acute and sub-acute) psychiatric wards reported their experienced coercion on Coercion Ladder and the Experienced Coercion Scale in questionnaires. Staff reported use of formal coercion, diagnoses, and psychosocial functioning. We used two tailed t-tests and mixed effects models to analyze the impact from demographics, alliance with parents, use of formal coercion, diagnostic condition, and global psychosocial functioning. Results: High experienced coercion was reported by a third of all patients. In a mixed effects model, being under formal coercion (involuntary admission and / or coercive measures); a worse relationship between patient and parent; and lower psychosocial functioning, significantly predicted higher experienced coercion. Twenty-eight percent of the total sample of patients reported a lack of confidence and trust both in parents and staff. Conclusions: Roughly one third of patients in the sample reported high experienced coercion. Being under formal coercion was the strongest predictor. The average scores of experienced coercion in subgroups are comparable with adult scores in similar care situations. There was one exception: Adolescents with psychosis reported low experienced coercion and almost all of them were under voluntary care. Keywords: Adolescent psychiatry, Involuntary admission, Cross-sectional studies, Perceived coercion Background order to prevent harm to people or property. Some stud- Coercion in mental health care remains controversial. ies found that coercive measures are used more fre- Research is increasingly focused on the use, effect, and quently for younger adolescents [3, 4]. The reported rate patient’s perception or experience of coercive treatment of inpatients subjected to one or more of these coercive forms, but little is published on adolescent patients. Co- measures ranged from 30% for inpatients in New York ercion is present in adolescent mental health care: in  and Finland  to 6.5% in Norway . Norway, 20% of admissions among 16- to 17-year-olds Other staff activities, such as inducements, interper- was involuntary , although it was 36.5% in a German sonal leverage, show of force, threats and house rules, sample . Adolescent inpatients may also be subjected can be experienced as coercive and regarded as informal to forced medication or nutrition for treatment pur- pressure or informal coercion . For adolescents, the poses, as well as other coercive measures, such as hold- magnitude of age, status, and knowledge differences vis ing, mechanical restraints, seclusion, and medication, in a vis the staff can increase the influence of informal pressure. Furthermore, adolescents are usually materi- * Correspondence: firstname.lastname@example.org ally, financially, and emotionally dependent on parents R&D department, Division of Mental Health Services, Akershus University or guardians , so that control and pressure may relate Hospital, PB 1000, 1478 Lørenskog, Norway to care, trust, and family loyalty. There is also a risk that Institute of Clinical Medicine, University of Oslo, Oslo, Norway Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Nyttingnes et al. BMC Health Services Research (2018) 18:389 Page 2 of 10 some adolescent patients lack or lose trust in parents to the clinician is associated with lower perceived coer- and staff during hospitalization, and consequently feel cion . Research on the impact of demographic and isolated in the ward. Coercion is often accompanied or clinical characteristics displayed small and inconclusive followed by feeling rejected, aggrieved, punished  effects . Thus, we lack a clear and documented un- disempowered or terrified . Some prospective studies derstanding of the interrelation between the main ex- have connected experienced coercion to lower quality of planatory variables of perceived coercion, such as patient life , and worse alliance and follow up of care , characteristics, care regimen, alliance, and procedural but a review found small or absent effects on variables justice. Qualitative studies indicate that patients do not such as psychosocial functioning, readmissions, or ser- equate freedom restrictions to perceived coercion, but vice engagement . Given frequent use of formal co- restricts the coercion concept to negatively viewed re- ercion, the potential for informal pressure or coercion, strictions, such as the humiliating ones . In addition, and the vulnerable adolescent years, experienced coer- they described coercion as a broader experience affecting cion in adolescents should be an important research self-image, and sometimes with existential consequences topic . [30, 31]. For patients, coercion seems to be more of a A review of adolescent experiences with mental health negative experience than merely a perception, making care found few inpatient studies, and the main topic was experienced coercion the preferred concept. patient satisfaction . One recent American interview Our main study aims were to establish the level of expe- study with inpatients found that rigidity and confine- rienced coercion and test candidate predictor variables in ment were the most frequently disliked aspects of care a sample of hospitalized adolescents. Based on existing . A few interview studies have reported how adoles- findings for adults [32, 33] and how formal coercion is cent and young adult inpatients with anorexia view used for adolescents [1, 5] we hypothesized that younger treatment: Patients are aware of staff strategies for influ- age, use of formal coercion (involuntary care, coercive ence, such as persuasion and use of patient privileges. treatment or measures), eating disorders, and lower global Some patients attempt to resist or circumvent treatment, psychosocial functioning would predict higher experi- i.e., some play by the rules to get out, and some attenu- enced coercion. However, as eating disorders and ate staff authority by questioning their competence . lower psychosocial functioning  are associated with in- Patients spoke about formal coercion and informal pres- creased use of formal coercion, we expected these clinical sure, with some saying that coercion and restrictions variables to lose significance when controlled for use of could at times be justified and helpful . In one study, formal coercion. Additional study aims were to explore: adolescents with eating disorders reported more experi- enced coercion than adults . Whether a good relation to the parent or guardian The literature on coercion in adults is far more exten- would predict higher or lower experienced coercion; sive, with subfields such as outpatient coercion with Whether pressure for admission from parents would community treatment orders, coercive measures, and have different effects on experienced coercion for perceived or experienced coercion. In 1993, the develop- voluntary vs involuntary patients; ment of the Admission Experience Survey and its sub- What proportion of patients would report lack of scale the MacArthur Perceived Coercion Scale  trust or closeness towards both parents and staff. stimulated a series of studies on perceived coercion. These studies found that involuntary care predicts per- For voluntary patients, we expected that pressure from ceived coercion, although approximately 35% of involun- parents (re: admission) would lead to higher experienced tary patients in acute wards reported low perceived coercion compared to patients without such pressure. coercion in several studies [22, 23]. Conversely, the However, for involuntary patients, such pressure could number of voluntarily admitted patients who reported a be insufficient to add to experienced coercion, and high perceived coercion score, ranged from 10% in the might contribute to a sense of necessity and legitimate original MacArthur studies  to 48% in a smaller Eng- care, with less experienced coercion. lish study . Across studies, the odds ratio of patients under involuntary care reporting high perceived coer- Methods cion compared to voluntary patients was 8.6 . Use of The study context physical force or threats of social consequences for treat- In Norway, per 100,000 underage persons (aged 0–17), ment also predicts higher perceived coercion in patients there are 26 mental health inpatient beds used yearly by . A higher level of perceived procedural justice – i.e., 180 patients in 249 admissions . The adolescent feeling that you had a say in the decision and considered wards in this study accept patients from 13 to 17 years. the admission process to be fair – are associated with This age group uses approximately 75% of the total lower perceived coercion . Also, a positive relation underage inpatient capacity , indicating that in 2014, Nyttingnes et al. BMC Health Services Research (2018) 18:389 Page 3 of 10 0.5% of the adolescent population received inpatient Measurements mental health care. Adolescent inpatients in Norway are We used paper forms filled out by patients and thera- thus a highly select group, expected to have severe men- pists to measure the variables selected for this study. tal health problems, which services consider difficult to Members of the adolescent group of the Norwegian administer proper care in outpatient settings. Norwegian Acute Psychiatric Network suggested clinically-relevant adolescent acute and sub-acute units are small but variables as well as their wording. well-staffed, usually with 10 or fewer beds per ward, and with staffing (including administrative) of more than 4 Experienced coercion employees per bed . No measure of experienced coercion has been validated According to the Norwegian Mental Health Act, pa- for adolescents, so we chose two measures developed for tients 16 and above can be involuntarily admitted and adults with complementary strengths, and we report and treated according to the same rules as adults. Patients compare both. The Coercion Ladder (CL) is a one-item, less than 16 are admitted based on parental consent, and self-anchoring visual analogue scale based on the Cantril are thus formally seen as voluntary . The ward shall Ladder , measuring one’s recent experience of being notify the Control Commission (a tribunal board for coerced. The score range is 1–10 and the respondents complaints about involuntary mental health care) when- are instructed that the lowest and highest scores should ever an admitted patient under 16 disagrees with the correspond to the lowest and highest level of experi- parents’ decision. enced coercion they can imagine. The participant’s un- derstanding of the word ‘coercion’ is the anchor. This may sacrifice reliability, as found in other iterations of Design Cantril’s approach , but should be directly applicable We conducted a cross-sectional study of adolescent in- to adolescent mental health care and adolescents’ under- patients from 10 Norwegian acute and combined acute standing of the word ‘coercion.’ The Experienced Coer- and sub-acute psychiatric wards. Data were collected cion Scale (ECS) has 15 agreement-rated five-point from patients, staff, and clinical records. Likert items, and the score range is 0–4. Items are ap- plicable across care phases, care settings, and forms of Recruitment of wards coercion, focusing on patients’ negative evaluations and We sent an invitation to participate to all 16 Norwegian feelings . We calculated average sumscore from valid adolescent wards (acute and combined acute and inter- item responses. For both scales, we defined high experi- mediate inpatient) approved for involuntary care. Ten enced coercion as a score above the midpoints [> 5 on out of these wards participated in the study. CL, > 2 on the ECS). Patients also noted if they agreed with the admission and if they thought their parents or Patient inclusion and data collection other parties agreed with it too. Data collection took place in 2015. Each participating ward chose a start-up day for recruitment. At this point, Use of formal coercion all admitted patients in the ward regardless of care for- Involuntary admission was coded ‘yes’ if the adolescent was mality, were considered for eligibility. Patients’ inclusion 16 to 17 years old, and involuntary admitted according to criteria were being 13- to 17-years-old, competency to clinical records. This variable was also coded ‘yes’ for youn- consent by understanding the consequences of partici- ger patients who disagreed to being admitted, warranting a pating, and the ability to comprehend a two-page ques- notification to the Control Commission. Data about coer- tionnaire. Patients were approached by local clinicians, cive measures, such as involuntary medication, involuntary who gave them information about the study and re- nutrition, restraints, and open door seclusion, which hap- quested consent to participate. For patients under the pened during the last three weeks of admission, was re- age of 16, parents were also asked for consent. The pa- ported by staff. Patients under formal coercion were those tient was asked to fill out a form with questions and who had experienced any involuntary admission or coercive statements (see Additional file 1), preferably in private, measure described in this paragraph. and to enclose it in an envelope themselves. Staff assisted with reading or explanations if needed. The pa- Clinical status tient’s primary contact or responsible clinician also filled Diagnosis was measured as the main psychiatric disorder out a form about the patient and treatment based on the using Axis One (clinical psychiatric syndromes) in the patient’s record and past care (see Additional file 2). Re- multiaxial ICD-10 Classification of Child and Adolescent cruitment procedures were repeated weekly for Psychiatric Disorders from the World Health newly-admitted patients until the ward reached its goal, Organization . This was found in the patient’s record based on ward size, or gave up recruitment. during data collection. Global psychosocial functioning Nyttingnes et al. BMC Health Services Research (2018) 18:389 Page 4 of 10 was measured using the units’ routine application of the Children’s Global Assessment Scale (CGAS)  at ad- mission, and by asking the clinician to rate the CGAS at the time of the patient’s response. Staff rated the patient with Health of the Nation Outcome Scales – Children and Adolescents (HoNOSCA) for use of alcohol or drugs  in the last 6 months. HoNOSCA defines non-problematic use as no use or use within age norms. Length of stay and the living situation from which the adolescent was admitted were rated by staff using the patient records. Relation between patient and parent/guardian The quality of the patient’s relation to parent and staff was measured with a set of agreement-rated Likert-items. In this section, we rewrote and adapted the expectation of help from mother/father in the Conflict Behaviour Questionnaire . The patient rated parent Fig. 1 Flow chart of recruitment, exclusions, and refusals of patients and staff on separate items. We were similarly informed of the theme of openness and trust from the Scale to As- sess the Therapeutic Relationship , and coined an Statistical analyses item of hiding inner feelings, which the patient rated for We analysed data with SPSS 23. CL scores showed a parents and staff. Also, staff rated the relation between skewed distribution, so we utilized Spearman’s rank patient and parent or guardian. We calculated a com- order for correlation with this scale. Parametric tests bined measure of patients’ relation towards the parent as were restricted to ECS scores, in which we studied the the average score on two patient-rated items and one predictive value of use of formal coercion and diagnosis staff-rated item, where higher scores indicated better re- with two tailed t-tests, Cohen’s d effect size, and lations. An item was added after the pilot interviews, ac- ANOVA. We used ECS sum scores as the dependent knowledging the nuance between hiding one’s inner variable and estimated a linear mixed effects model. feelings from the parent due to lack of trust or in order Here we entered age, relation to parent or guardian, glo- to spare them from knowing convoluted feelings or situ- bal psychosocial functioning, eating disorders, and use ations. Staff rated the degree of informal pressure from of formal coercion as fixed effects and estimated a ran- parents on a self-made 5-point Likert item. dom intercept for the effect of wards. Non-dichotomous Gender was marked by the patients, and age was re- predictors were centered at their grand mean. In a sec- ported by the staff. Ethnic or immigrant backgrounds ond model, we explored whether informal pressure for were not recorded. admission from the parents influenced experienced coer- We piloted the patient questionnaire with a cognitive cion differently for voluntary patients compared to pa- validation interview , with three patients at two sites, tients under formal coercion, using an interaction and inquired how items were understood, how the pa- variable. tient reasoned, and how he/she thought other patients would reason when answering the form. Pilot interviews Results indicated that patient items, including experienced coer- The sampling procedure resulted in adolescent inpa- cion scales, were understood. tients with characteristics shown in Table 1. More girls (68.8%) than boys and more older (65.6%) than younger adolescents participated. Staff reported that 81 (86.2%) Study sample patients had non-problematic use of alcohol and drugs. Among 132 patients considered for participation, data from 96 (73%) were included in the analyses, as shown Experienced coercion among adolescent inpatients in Fig. 1. We excluded three cases with more than 20% The patients’ mean score on CL was 4.7 (SD = 2.9, me- missing ECS items. For remaining participants, CL had dian score = 5). The mean score for patients under for- no missing data. ECS items had a total of 15 missing an- mal coercion was 7.3 (SD = 2.6, median score = 8) while swers (1.04%). No participant missed more than two voluntary patients’ mean score was 4.1 (SD = 2.6, median item responses on the ECS. score = 4). The mean score on the ECS (scaled from 0 to Nyttingnes et al. BMC Health Services Research (2018) 18:389 Page 5 of 10 Table 1 Patient characteristics 13–15 years 16–17 years Total n= 33 n= 63 n= 96 n (%) n (%) n (%) Gender Female 22 (66.7) 44 (69.8) 66 (68.8) Male 11 (33.3) 19 (30.2) 30 (31.3) Diagnosis (ICD-10 codes) Psychosis (F20–31) 9 (27.3) 5 (7.9) 14 (14.6) Pervasive developmental disorder (F84) 1 (3) 6 (9.5) 7 (7.3) Eating disorders (F50) 5 (15.2) 9 (14.3) 14 (14.6) Depressive disorder (F32–34) 9 (27.3) 20 (31.7) 29 (30.2) Anxiety, dissociative disorders, PTSD (F40–44; F92–94) 5 (15.2) 9 (14.3) 14 (14.6) All other disorders (incl. missing) 4 (12.1) 14 (22.2) 18 (18.8) Living situation before admission Living at home 26 (78.8) 50 (79.4) 76 (79.2) Living in institution or foster care 6 (18.2) 9 (14.3) 15 (15.6) Not specified (other or missing) 1 (3) 4 (6.3) 5 (5.2) Length of stay at the time of data collection Short (1–4 days) 8 (24.2) 23 (36.5) 31 (32.3) Medium (5–21 days) 17 (51.5) 21 (33.3) 38 (39.6) Long (22 days or longer) 8 (24.2) 15 (23.8) 23 (24) Missing 4 (6.3) 4 (4.2) Involuntary admission No 29 (87.9) 49 (77.8) 78 (81.3) Yes 4 (12.1) 14 (22.2) 18 (18.8) Episode of coercive measure for last three weeks No 32 (97) 54 (85.7) 86 (89.6) Yes 1 (3) 8 (12.7) 9 (9.4) Missing 0 (0) 1 (1.6) 1 (1) Children’s Global Assessment Scale mean (sd) mean (sd) mean (sd) At admission 38.5 (8.8) 35.7 (13.8) 36.7 (12.3) At time of data collection 44.6 (9.1) 40.7 (13.8) 42.1 (12.5) ICD-10: International Statistical Classification of Diseases and Related Health Problems, 10th Revision sd standard deviation Personality disorder (F60), Hyperkinetic disorder (F90), Unspecified mental disorder (F99), Auditory hallucinations (R44.0), Suicidal ideation (R45.8), Observation for suspected mental or behavioral disorder (Z032) Coercive measures could include physical holding, mechanical restraints, medication, nutrition, isolation, or open door seclusion Missing data for 3 patients Missing data for 13 patients 4) was 1.7 (SD = 0.9). The correlation between CL scores adolescents under formal coercion, the percentage was and the ECS sum scores was r = .68. The distribution of 73.7, while 24.7% of the voluntary patients reported high both scales of experienced coercion is shown in Fig. 2. experienced coercion. The mean difference in experienced coercion as measured by the ECS, between patients under formal Predictors of experienced coercion coercion (2.4 points, n = 19) and the non-coerced patients Among the 96 patients, 46 (47.9%) agreed that they (1.5 points, n = 77) was 0.9 points [0.5, 1.3], with t (94) = ought to be treated on the ward. Of these, 12 patients 4.16, p <.001, d =1.01. (26.1%) nevertheless reported high experienced coercion In the total sample, 33 of the patients (34.4%) reported on the ECS. Fifty patients did not agree with treatment high experienced coercion (ECS score > 2). For on the ward, and 28 of these (56.0%) reported low Nyttingnes et al. BMC Health Services Research (2018) 18:389 Page 6 of 10 Fig. 2 Histogram of Coercion Ladder scores and Experienced Coercion Scale (ECS) average sumscores in the sample. N =96 experienced coercion according to the ECS. A majority patients with psychosis (p = .016). Other differences were of 62 patients (64.6%) thought their parents endorsed nonsignificant. Eight of 14 patients with eating disorders, the current stay. Only 16 patients (16.7%) disagreed with but only 1 of 14 patients with psychosis were under formal treatment on the ward and thought the parents did not coercion. endorse the current stay. Here, the adolescent consid- In the first step of multilevel modeling, we estimated a ered Child Protection Services (5 patients (5.2%)) and model not including any predictors, but accounting for Child and Adolescent Mental Health Care (10 patients the variation in experienced coercion between wards. (10.4%)) as proponents of their current treatment. The intraclass correlation (ICC) was 0.072. Akaikes in- We found significantly different levels of experienced formation criterion was 225.443 for this model. Then we coercion for patients in the diagnostic groups shown in added the predictors shown in Table 2 as fixed effect Table 1, as implied by ANOVA with F(5,90) = 2.570, variables in the model. In this model, ICC for ward was p = .032. A Tukey post hoc test revealed that the experi- 0.102, and the Akaike information criterion was 208.456, enced coercion score of 2.29 in patients with eating disor- indicating a smaller information loss when we included ders was significantly higher compared to that of 1.20 in the predictors. Being under formal coercion, having Table 2 Parameter estimates of predictors of Experienced Coercion Scale scores with a random intercept for wards in a mixed effects model Parameter Estimate 95% CI Fixed Intercept 1.466** [1.163, 1.769] Patient age − 0.114 [− 0.257, 0.028] Patients’ relation to parent or guardian − 0.258* [−0.425, − 0.091] Global psychosocial functioning (CGAS) − 0.021* [−0.039, − 0.003] Diagnosed with eating disorder (reference: no) 0.341 [−0.158, 0.840] Patient under formal coercion (reference: no coercion) 0.805** [0.353, 1.257] Covariance Residual standard deviation 0.546 [0.393, 0.798] Between wards standard deviation 0.062 [0.009, 0.450] CI confidence interval Non-dichotomous variables are grand mean centered * p ≤ .05. ** p ≤ .001 Nyttingnes et al. BMC Health Services Research (2018) 18:389 Page 7 of 10 lower psychosocial functioning and a worse relation to from adult samples. On the ECS, adolescents under for- parents or guardian were significant predictors of higher mal coercion scored 2.4 points, while patients under in- experienced coercion score in the model. Age and eating voluntary care in a Norwegian adult sample scored 2.2 disorders were non-significant. Parameter estimates with points . The scores for voluntary patients were 1.5 confidence intervals (CIs) are in Table 2. and 1.3 points in the adolescent and adult samples, re- To assess if informal pressure from the parents influ- spectively. The correlation between CL and the ECS was enced experienced coercion differently in voluntary and in the same range in this study as in the ECS validation coerced patients, we estimated a second mixed effects study, with r = .68 in both studies. About 1/3 of the model with fever parameters but adding interaction be- sample, and 3/4 of the patients under formal coercion, tween informal pressure from parents and being under reported high experienced coercion. Adolescent in- formal coercion. In this model, ICC for ward was 0.088 patient stays may be formative for future alliance, con- and the Akaike information criterion was 231.895. Re- cordance with care plans, and possibly influence later moving the interaction parameter increased the Akaike treatment results. We therefore see a need for prospect- information criterion to 234.536. The parameter esti- ive studies examining the consequences of experienced mates are given in Table 3. coercion in adolescent mental health care. The analyses show that voluntary patients rate higher Experienced coercion varied with diagnosis. As ex- experienced coercion when there is more informal pres- pected, patients with eating disorders reported higher sure from parents or guardians. For patients under for- experienced coercion. Surprisingly, patients with psych- mal coercion, more informal pressure predicted lower osis reported low experienced coercion, and only one of experienced coercion. these patients was involuntary admitted or subjected to coercive measures. Psychosis and psychotic symptoms The patients’ trust in parents and staff have repeatedly been connected to more coercion and Fifty-six (58.9%) patients agreed or strongly agreed they higher experienced coercion in adult samples [32, 46]. In would not show their parents how they really felt, either Norway, 62% of all adult involuntary inpatient time was due to lack of trust (14 patients) or to spare the parents for patients with a main diagnosis of schizophrenia . (19 patients), or for both reasons (23 patients). Nevertheless, Norwegian adolescent inpatient wards Thirty-seven patients (38.5%) would not show staff how seemed able to care for most psychotic patients without they felt. Twenty-six patients (27%) would not show formal involuntary care or experienced coercion for how they felt to either parents or staff. them. More studies are needed to rule out bias in our results, and to investigate how non-coercive psychosis Discussion care is accomplished. To our knowledge, this is the first quantitative study of Although the strongest predictor of experienced coer- experienced coercion in adolescent mental health care. cion was being under formal coercion, approximately ¼ The study adds valuable knowledge regarding degree of patients under formal coercion reported low experi- and predictors of experienced coercion. enced coercion, and approximately ¼ of voluntary pa- The level of experienced coercion, as measured by the tients reported high experienced coercion. Another ECS and the CL, was in a similar range as in reports significant predictor in the mixed effects model was Table 3 Parameter estimates for effect of parent or guardian pressure on Experienced Coercion Scale scores with a random intercept for wards in a mixed effects model Variable Estimate 95% CI Fixed Intercept 1.543** [1.281, 1.805] Patients’ relation to parent or guardian −0.232* [− 0.392, − 0.072] Informal pressure from parent or guardian 0.222* [0.059, 0.385] Patient under formal coercion (reference: no coercion) 0.902** [0.489, 1.314] Informal pressure from parent or guardian x patient under formal coercion −0.358* [− 0.697, − 0.019] Covariance Residual standard deviation 0.573 [0.422, 0.777] Between wards standard deviation 0.056 [0.009, 0.3570] CI confidence interval Non-dichotomous variables are grand mean centered * p ≤ .05. ** p ≤ .001 Nyttingnes et al. BMC Health Services Research (2018) 18:389 Page 8 of 10 negative relations with parents, which may stem from them particularly lonely and vulnerable. If, for some rea- more relational problems in general. Age did not predict son, understanding, empathy, or care quality breaks experienced coercion, and this hypothesis was based on down, then the staff, the control system, or the parents findings of more frequent use of coercive measures for cannot rest assured that an adolescent will discuss it younger patients . In Table 1, we see no sign of such a with a parent. As implied in the pilot interviews, some tendency in this sample. Patients with eating disorders adolescents may hide their negative feelings and experi- reported high experienced coercion, but this may have ences from parents to spare them a burden. This may be been mediated by being under formal coercion, making the case if parents initiated or agreed with admission, eating disorders insignificant in the mixed effects model. and if the alliance or treatment results eventually The explanatory power of patient characteristics varies soured. between existing studies of experienced coercion. Our results indicate that sometimes a more restrictive care regimen may mediate the effect of patient variables on Limitations experienced coercion. For some variables, there may be The study sample is small, partly reflecting the small competing causal chains at work. The mixed effects adolescent wards. This sample size implies that findings model in Table 2, indicates that a worse psychosocial on subgroups should be treated with caution. On the functioning predicted higher experienced coercion. Bet- other hand, the rate of missing data was low from both ter global psychosocial functioning may protect from patients and staff. Ten out of 16 Norwegian adolescent some care restrictions, leading to lower experienced co- acute wards participated, and the participation rate on ercion. However, better psychosocial functioning indi- the wards was high. ICC for wards explained less than cates that involuntary admission is less proportional, 10% of the variation in experienced coercion. Also, we which may lead to less acceptance of the admission, as received no reports of problems from the involved clini- found in an English study . In the former case a cians, such that the adolescents seem to have handled thorough multivariate control for all care restrictions the questionnaire well. should remove the significance of psychosocial function- Another limitation to this study is that the scales of ing. A main effect cannot be ruled out either, in which experienced coercion have not previously been applied lower psychosocial functioning may weaken a patient’s or validated in adolescent populations. We did not use ability to see the care situation from different perspec- the frequently-used MacArthur Perceived Coercion tives, creating a sense of more experienced coercion in Scale, as it was developed and validated for an adult ad- an otherwise comparable care situation. While studies mission process, with little regard for parent authority may control for formal coercion, it is difficult to rule out and involvement. We piloted the patient form, and in- that effects of patient variables are mediated by informal cluded two measures of experienced coercion. The cor- pressure or coercion. In order to resolve these questions, relation between these two measures was r = .68, as for a validated measure of informal pressure and restrictions adults. This similarity between the self-anchoring CL in care, preferably reported by sources other than the and ECS with items of negative valence indicates that patient, would seem to be necessary. adolescents delimit the coercion concept to freedom re- Given adolescent dependency on parents, how does in- strictions that are experienced negatively. Scale revisions formal pressure from caregivers predict experienced co- or separate development for adolescents is preferable, ercion? Our post hoc mixed effects model shows that however. Some other variables were also measured with pressure from parents predicted higher experienced co- items adapted or developed for this study, which have ercion on the ward for voluntary patients. But for pa- not yet been validated. For the patient-parent relation tients under formal coercion, informal pressure from we combined two patient-reported items and one parents was associated with lower experienced coercion, staff-reported item, and attributes of adolescents and although the subsamples were small in this model. We families with a certain rating is not known. In particular, speculate that this effect may be due to the parental the parent perspective is not included, and the findings legitimization of the involuntary care. must be considered as tentative. Diagnosis and CGAS How did inpatients assess their alliance and trust in was based on clinicians’ assessments, and not tested for staff and parents? Almost half the patients agreed to reliability. These clinical variables reflect the staff con- treatment in the ward. Nevertheless, 27% of inpatients siderations and perspective, and include the breadth of did neither report a good alliance with their parents nor the assessment, supplemental information and observa- the staff. The study sample is a highly select group based tion. While the diagnosis should have a broad base and on problem severity. Lack of trust in adult relations may be informed by the cooperation between professionals be a part of the situation for several adolescent inpa- on the ward and cooperating services, the CGAS-score tients. This may contribute to their problems, and make practice may vary more from one ward to another. Nyttingnes et al. BMC Health Services Research (2018) 18:389 Page 9 of 10 Generalizability is limited by the sample size and the Authors’ contributions TR, RN and ON planned and designed the study. ON organized data study context. The organisation of mental health ser- collection, conducted the statistical analyses and drafted the main vices for children and adolescents shows great variation manuscript. ON, JR, TR and KHB contributed to the interpretation of the data. across countries . In Norway, the proportion of All authors revised the work critically several times and approved the final manuscript. underage persons in contact with the outpatient division of Child and Adolescent Mental Health Services was 5.1 Ethics approval and consent to participate in 2014 . From 1998 to 2013, around 0.03% of The study was approved by Regional Committee for Medical and Health Research Ethics, Norway, project No 2011/2574/REK sør-øst. Informed underage persons were hospitalised each year , and consent for participation in the study was obtained in verbal form from all for Norwegian adolescent’s about 0.5% received a mental participants, with additional consent from parents of patients from 13 to health inpatient stay in 2014. 15 years. No identifiable patient information was demanded by the study design, and signing and collecting consent forms would increase The sample had a majority of girls (69%), close to the transparency, but also the risk for confidentiality breaks. The Ethical yearly national rates (65%) . Severe diagnoses, such committee deemed that a procedure where we asked patients for oral as psychosis, eating disorders, and pervasive develop- consent before they produced anonymous questionnaire information and allowed using anonymized patient record information was acceptable in this mental disorders made up 36.5% of this sample, while study. The same considerations applied for consent from parents of patients national all-year statistics for 2014 indicates that these under 16 years. disorders amounted to 21% . Our sampling was cross-sectional, and patients with more severe problems Competing interests The authors declare that they have no competing interests. often have longer stays and a greater likelihood for sam- pling than those with shorter stays. We think the reason for a low rate of externalizing behavioural disorders is Publisher’sNote Springer Nature remains neutral with regard to jurisdictional claims in that inpatient care for this group is often mandated by published maps and institutional affiliations. the Norwegian Child Protection Services. Author details R&D department, Division of Mental Health Services, Akershus University Conclusions 2 Hospital, PB 1000, 1478 Lørenskog, Norway. Institute of Clinical Medicine, The level of experienced coercion in adolescent inpatient University of Oslo, Oslo, Norway. Work Research Institute, Oslo Metropolitan University, Oslo, Norway. Health Services Research Unit, Akershus University care found in this study was similar to comparable re- Hospital, Oslo, Norway. Centre for Care Research, University College of sults for adult inpatient care. Use of formal coercion is Southeast Norway, Notodden, Norway. the strongest predictor of experienced coercion, so use Received: 28 June 2017 Accepted: 15 May 2018 of coercion in adolescent mental health care should re- ceive similar attention as in research and policies for adults. Norwegian adolescent wards treated psychosis References with little use of formal coercion, and these patients also 1. Furre A, Heyerdahl S. Bruk av tvang i ungdomspsykiatriske akuttavdelinger [use of coercion in adolecent psychiatric acute wards]. Report: Regionsenter reported low experienced coercion. for barn og unges psykiske helse. Helseregion ØST og SØR; 2010. 2. Jendreyschak J, Illes F, Hoffmann K, Holtmann M, Haas C-R, Burchard F, Emons B, Schaub M, Armgart C, Schnieder H, et al. Voluntary versus Additional files involuntary hospital admission in child and adolescent psychiatry: a German sample. Eur Child Adolesc Psychiatry. 2014;23(3):151–61. Additional file 1: Patient report form. An English translation of the 3. Gullick K, McDermott B, Stone P, Gibbon P. Seclusion of children and patient report form utilized in the study. (PDF 16 kb) adolescents: psychopathological and family factors. Int J Ment Health Nurs. 2005;14(1):37–43. Additional file 2: Staff report form. An English translation of the staff 4. Sourander A, Ellilä H, Välimäki M, Piha J. Use of holding, restraints, seclusion report form utilized in the study. (PDF 23 kb) and time-out in child and adolescent psychiatric in-patient treatment. Eur Child Adolesc Psychiatry. 2002;11(4):162–7. 5. Pogge DL, Pappalardo S, Buccolo M, Harvey PD. Prevalence and precursors Abbreviations of the use of restraint and seclusion in a private psychiatric hospital: CGAS: Children’s Global Assessment Scale; CL: Coercion Ladder; comparison of child and adolescent patients. Admin Pol Ment Health. 2013; ECS: Experienced Coercion Scale; HoNOSCA: Health of the Nation Outcome 40(3):224–31. Scales – Children and Adolescents; ICC: Intraclass correlation; ICD- 6. Siponen U, Välimäki M, Kaltiala-Heino R. The use of coercive measures in 10: International Statistical Classification of Diseases and Related Health adolescent psychiatric inpatient treatment: a nation-wide register study. Soc Problems, 10th Revision Psychiatry Psychiatr Epidemiol. 2012;47(9):1401–8. 7. Furre A, Sandvik L, Friis S, Knutzen M, Hanssen-Bauer K. A nationwide study Acknowledgements of why and how acute adolescent psychiatric units use restraint. Psychiatry MSc Jonas Christoffer Lindstrøm assisted with statistical advice and analyses. Res. 2016;237:60–6. Thanks to the adolescent group of the Norwegian acute psychiatric network 8. Hotzy F, Jaeger M. Clinical relevance of informal coercion in psychiatric for discussions and practical assistance with coordination of the data treatment—a systematic review. Front Psychiatr. 2016;7:1-12. https://doi.org/ collection. 10.3389/fpsyt.2016.00197. 9. Group for the Advancement of Psychiatry, editor. Coercion in Child Availability of data and materials Psychiatric Treatment. In: Group for the Advancement of Psychiatry, editors. The dataset analyzed during the current study are available from the Forced into treatment: The role of coercion in clinical practice. Washington corresponding author on reasonable request. DC: American Psychiatric Press; 1994. p. 1–12. Nyttingnes et al. BMC Health Services Research (2018) 18:389 Page 10 of 10 10. Katsakou C, Priebe S. Patient's experiences of involuntary hospital admission 35. Helsedirektoratet. Samdata spesialisthelsetjenesten 2014. Norwegian [The and treatment: a review of qualitative studies. Epidemiol Psichiatr Soc. 2007; Norwegian Directorate of Health: Comparison data for Norwegian specialist 16(2):172–8. health services 2014] Report IS-2348.2015. https://helsedirektoratet.no/Lists/ 11. Seed T, Fox JR, Berry K. The experience of involuntary detention in acute Publikasjoner/Attachments/989/SAMDATA%202014%20IS-2348.pdf. Accessed psychiatric care. A review and synthesis of qualitative studies. Int J Nurs 10 May 2017. Stud. 2016;61:82-94. 36. Helsedirektoratet. Samdata spesialisthelsetjenesten 2013. Norwegian [The 12. Link B, Castille DM, Stuber J. Stigma and coercion in the context of Norwegian Directorate of Health: Comparison data for Norwegian specialist outpatient treatment for people with mental illnesses. Social science & health services 2013] Report IS-2194. 2014. https://helsedirektoratet.no/Lists/ medicine (1982). 2008;67(3):409–19. Publikasjoner/Attachments/615/Samdata-spesialisthelsetjenesten-2013-IS- 2194.pdf. Accessed 10 May 2017. 13. Kaltiala-Heino R, Laippala P, Salokangas RKR. Impact of coercion on 37. Mental Health Care Act: Act No. 62 of 2 July 1999 relating to the provision treatment outcome. Int J Law Psychiatry. 1997;20(3):311–22. and implementation of mental health care, with later amendments. In. 14. Newton-Howes G, Mullen R. Coercion in psychiatric care: systematic review Oversatte lover [Translated laws]. 1999. Retrieved from: http://app.uio.no/ub/ of correlates and themes. Psychiatr Serv. 2011;62(5):465–70. ujur/oversatte-lover/data/lov-19990702-062-eng.pdf. Accessed 10 May 2017. 15. Ellila HT, Sourander A, Valimaki M, Warne T, Kaivosoja M. The involuntary 38. Høyer G, Kjellin L, Engberg M, Kaltiala-Heino R, Nilstun T, Sigurjónsdóttir M, treatment of adolescent psychiatric inpatients–a nation-wide survey from Syse A. Paternalism and autonomy: A presentation of a Nordic study on the Finland. J Adolesc. 2008;31(3):407–19. use of coercion in the mental health care system. Int J Law Psychiatry. 2002; 16. Biering P. Child and adolescent experience of and satisfaction with 25(2):93–108. psychiatric care: a critical review of the research literature. J Psychiatr Ment 39. Larsen RJ, Diener E, Emmons RA. An evaluation of subjective well-being Health Nurs. 2010;17(1):65–72. measures. Soc Indic Res. 1985;17(1):1–17. 17. Moses T. Adolescents’ perspectives about brief psychiatric hospitalization: 40. World Health Organisation. Multiaxial classification of child and adolescent what is helpful and what is not? Psychiatr Q. 2011;82(2):121–37. psychiatric disorders: the ICD-10 classification of mental and behavioural 18. Boughtwood D, Halse C. Other than obedient: Girls' constructions of disorders in children and adolescents. Cambridge: Cambridge University doctors and treatment regimes for anorexia nervosa. J Community Appl Soc Press; 1996. Psychol. 2010;20(2):83–94. 41. Shaffer D, Gould MS, Brasic J, Ambrosini P, Fisher P, Bird H, Aluwahlia S. A children's 19. Tan JOA, Stewart A, Fitzpatrick R, Hope T. Attitudes of patients with anorexia global assessment scale (CGAS). Arch Gen Psychiatry. 1983;40(11):1228–31. nervosa to compulsory treatment and coercion. Int J Law Psychiatry. 2010; 42. Gowers S, Harrington R, Whitton A, Beevor A, Lelliott P, Jezzard R, Wing J. 33(1):13–9. Health of the nation outcome scales for children and adolescents (HoNOSCA). 20. Guarda AS, Pinto AM, Coughlin JW, Hussain S, Haug NA, Heinberg LJ. Glossary for HoNOSCA score sheet Br J Psychiatry. 1999;174(5):428–31. Perceived coercion and change in perceived need for admission in patients 43. Prinz RJ, Foster S, Kent RN, O'Leary KD. Multivariate assessment of conflict in hospitalized for eating disorders. Am J Psychiatry. 2007;164(1):108–14. distressed and nondistressed mother-adolescent dyads. J Appl Behav Anal. 21. Gardner W, Hoge SK, Bennett N, Roth LH, Lidz CW, Monahan J, Mulvey EP. 1979;12(4):691–700. Two scales for measuring patients' perceptions for coercion during mental 44. McGuire-Snieckus R, McCabe R, Catty J, Hansson L, Priebe S. A new scale to hospital admission. Behav Sci Law. 1993;11(3):307–21. assess the therapeutic relationship in community mental health care: STAR. 22. Iversen KI, Høyer G, Sexton H, Grønli OK. Perceived coercion among patients Psychol Med. 2007;37(01):85–95. admitted to acute wards in Norway. Nordic J Psychiatry. 2002;56(6):433–9. 45. Collins D. Pretesting survey instruments : an overview of cognitive methods. 23. Hoge SK, Lidz CW, Eisenberg M, Gardner W, Monahan J, Mulvey E, Roth L, Qual Life Res. 2003;12:229–38. Bennett N. Perceptions of coercion in the admission of voluntary and 46. Längle G, Renner G, Günthner A, Stuhlinger M, Eschweiler G, U'Ren R, involuntary psychiatric patients. Int J Law Psychiatry. 1997;20(2):167–81. Foerster K. Psychiatric commitment: patients' perspectives. Med Law. 2003; 24. Monahan J,Lidz CW, Hoge SK,MulveyEP,EisenbergMM,Roth LH,Gardner WP, 22(1):39–53. Bennett N. Coercion in the provision of mental health services: the MacArthur 47. Helsedirektoratet. Bruk av tvang i psykisk helsevern for voksne 2014. studies. In: Morrissey JP, Monahan J, editors. Coercion in mental health services - Norwegian [The Norwegian Directorate of Health: Use of Coercion in international perspectives. Grenwich CT: Jai Press Inc; 1999. p. 13–30. Norwegian Mental Health Care for adults 2014] Report IS-2452. 2016. 25. Sheehan KA, Burns T. Perceived coercion and the therapeutic relationship: a https://helsedirektoratet.no/Lists/Publikasjoner/Attachments/1161/ neglected association? Psychiatr Serv. 2011;62(5):471–6. Rapport%20om%20tvang%20IS-2452.pdf. Accessed 2 May 2017. 26. Newton-Howes G, Stanley J. Prevalence of perceived coercion among 48. Priebe S, Katsakou C, Amos T, Leese M, Morriss R, Rose D, Wykes T, Yeeles K. psychiatric patients: literature review and meta-regression modelling. Patients' views and readmissions 1 year after involuntary hospitalisation. Br J Psychiatrist. 2012;36(9):335–40. Psychiatry. 2009;194(1):49–54. 27. Lidz CW, Hoge SK, Gardner W, Bennett NS, Monahan J, Mulvey EP, Roth LH. 49. World Health Organization. Atlas: Child Adolesc Mental Health resources: Perceived coercion in mental hospital admission. Pressures and process. Global concerns, implications for the future. Geneva CH: World Health Arch Gen Psychiatry. 1995;52(12):1034–9. Organization; 2005. 28. Newton-Howes G. Coercion in psychiatric care: where are we now, what do we know, where do we go? Psychiatrist. 2010;34(6):217–20. 29. Svindseth MF, Dahl AA, Hatling T. Patients' experience of humiliation in the admission process to acute psychiatric wards. Nordic J Psychiatry. 2007; 61(1):47–53. 30. Norvoll R, Pedersen R. Exploring the views of people with mental health problems' on the concept of coercion: towards a broader socio-ethical perspective. Soc Sci Med. 2016;156:204–11. 31. Sibitz I, Scheutz A, Lakeman R, Schrank B, Schaffer M, Amering M. Impact of coercive measures on life stories: qualitative study. Br J Psychiatry. 2011; 199(3):239–44. 32. Fiorillo A, Giacco D, De Rosa C, Kallert TW, Katsakou C, Onchev G, Raboch J, Mastrogianni A, Del Vecchio V, Luciano M, et al. Patient characteristics and symptoms associated with perceived coercion during hospital treatment. Acta Psychiatr Scand. 2012:125;460-67. 33. Nyttingnes O, Holmén A, Rugkåsa J, Ruud T. The development, validation, and feasibility of the experienced coercion scale (ECS). Psychol Assess. 2017; 29(10):1210–20. 34. Furre A, Sandvik L, Heyerdahl S, Friis S, Knutzen M, Hanssen-Bauer K. Characteristics of adolescents subjected to restraint in acute psychiatric units in Norway: a case-control study. Psychiatr Serv. 2014;65:1367–72.
BMC Health Services Research – Springer Journals
Published: May 30, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
All the latest content is available, no embargo periods.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud