Aggressive behaviour of psychiatric patients with mild and borderline intellectual disabilities in general mental health care.

PURPOSE: Little is known about the associations between mild intellectual disability (MID), borderline intellectual functioning (BIF) and aggressive behaviour in general mental health care. The study aims to establish the association between aggressive behaviour and MID/BIF, analysing patient characteristics and diagnoses.
METHOD: 1174 out of 1565 consecutive in-and outpatients were screened for MID/BIF with the Screener for Intelligence and Learning Disabilities (SCIL) in general mental health care in The Netherlands. During treatment, aggressive behaviour was assessed with the Staff Observation Aggression Scale-Revised (SOAS-R). We calculated odds ratios and performed a logistic and poisson regression to calculate the associations of MID/ BIF, patient characteristics and diagnoses with the probability of aggression.
RESULTS: Forty-one percent of participating patients were screened positive for MID/BIF. Patients with assumed MID/BIF showed significantly more aggression at the patient and sample level (odds ratio (OR) of 2.50 for aggression and 2.52 for engaging in outwardly directed physical aggression). The proportion of patients engaging in 2-5 repeated aggression incidents was higher in assumed MID (OR = 3.01, 95% CI 1.82-4.95) and MID/BIF (OR = 4.20, 95% CI 2.45-7.22). Logistic regression showed that patients who screened positive for BIF (OR 2,0 95% CL 1.26-3.17), MID (OR 2.89, 95% CI 1.87-4.46), had a bipolar disorder (OR 3.07, 95% CI 1.79-5.28), schizophrenia (OR 2.75, 95% CI 1.80-4.19), and younger age (OR 1.69, 95% CI 1.15-2.50), were more likely to have engaged in any aggression. Poisson regression underlined these findings, showing a SCIL of 15 and below (β = 0.61, p<0.001) was related to more incidents.
CONCLUSIONS: We found an increased risk for aggression and physical aggression in patients with assumed MID/BIF. We recommend screening for intellectual functioning at the start of treatment and using measures to prevent and manage aggressive behaviour that fits patients with MID/BIF.

Introduction

Mild intellectual disability (MID) and borderline intellectual functioning (BIF) are highly prevalent in general mental health care but often stay unnoticed [1, 2]. Our research group has previously shown that in the Netherlands, the prevalence of MID/BIF increases by setting, from 27% in outpatient settings, to 40% in Flexible Assertive Community Treatment (FACT) teams and admission wards, to 67% in long-stay wards [3]. Furthermore, in the admissions wards, patients with MID or BIF were found to have increased risks of having been involuntarily admitted in the past (OR 2.71) and being subjected to coercive measures (OR 3.95) [1]. Aggressive and dangerous behaviours are the main reason for involuntary admissions and seclusion in the Netherlands. The severity and dangerousness of disruptive behaviour perceived by treating staff influence the decisions to use restrictive measures [4]. These measures are widely recognised as interventions that potentially have severe negative consequences for the patient, including trauma [5]. Aggression is often called "challenging behaviour" (CB) in intellectual disability (ID) services, and the use of coercive measures also has a significant impact on staff and healthcare workers. On average, 62% of nurses in different countries indicate they have experienced physical violence over the course of a year [6]. Health care workers [7-9] and workers in ID services [10, 11] experience psychological and emotional consequences of aggression such as post-traumatic stress, depression, and a negative impact on work functioning and job satisfaction.

From studies in institutions for people with ID [12, 13], we know that CB is a common problem. However, Bowring and colleagues [13] noted no agreed consensual, conceptual, or operational definition of CB. In population studies, considerable variation in CB prevalence is found (4%-22%, [14]). Communication problems, the severity of the ID, and psychopathology are associated with a higher risk of CB [14-16]. In a large Dutch study of an inpatient ID service covering 421 patients, 20% of the patients involved in aggression incidents were responsible for 50% of the verbal and 80% of the physical, aggressive incidents [17]. This study showed that the more severe the disability, the higher the possibility of repeated incidents in a single patient. Such patterns of incidence showing repetitive aggression in patients with more severe intellectual disability can also be expected to occur in general psychiatry.

A review of 424 studies conducted in general psychiatry in various settings across 11 countries showed that 32.4% of patients admitted to psychiatric facilities engaged in aggressive behaviour or violence and generated 182.8 events per 100 admissions [6]. Studies also show that a small subgroup of patients is generally responsible for a large proportion of violent incidents [6, 18]. Many previous publications have been single-centre reports, making comparison and generalising conclusions difficult. One indirect measure of aggression that considers whole country data is the UK NHS staff survey. In 2019, 48% of the 1.1 million NHS staff participated in the survey. 14.7% of all respondents reported having personally experienced violence from patients, relatives, or public members. This figure rose to 20.2% in the staff working in mental health services and 34% in those working for the ambulance services. It decreased to as low as 5.5% in acute services and 7.5% in non-psychiatric community services. Five years trends are remarkably stable in all measured groups [19]. Aggression and coercive measure are closely linked. Using whole country data, coercion figures were remarkably similar across four European countries [18]. An analysis of Welsh coercion data from this study across all Welsh Health Boards demonstrated twice as many coercive measures when ID services are included (2013 total incidents, Wales: 3735) compared to when ID services are not included (2013 incidents without ID, Wales: 1886). The results also showed that the number of patients affected by coercive measures per 100 occupied bed days was not affected by adding the ID data, but the number of coercive measures was. This suggests that those patients with ID who were affected by coercive measures were coerced multiple times and more often than the non-ID population [20]. This is similar to what we know from aggression data.

In a study on admission wards [1], we showed that patients with BIF/ MID had an increased risk of involuntary admission (OR 2.71; SD 1.28–5.70) and coercive measures (OR 3.95, SD 1.47–10.54). These findings were confirmed in nationwide data gathered in 2014, where intellectual impairment also showed an association with increased risk of seclusion and other coercive measures [21]. Internationally, there is evidence that patients with BIF/ID account for more and more prolonged seclusion and restraint events [2, 3].

Until now, however, the level of cognitive function has hardly been studied as a potential ’predictor’ of aggression, although MID /BIF is much more prevalent in general mental health care than previously assumed [3]. Therefore, in this study, we examined the associations between MID/BIF and aggressive behaviours in a sample of psychiatric inpatients and outpatients. We hypothesised that:

In mental health care services, patients suspected to have MID/BIF are more often engaged in aggression incidents

Patients suspected to have MID/BIF are more often involved in outwardly directed physical aggression and have more incidents per person than patients not suspected to have MID/BIF patients.

Method

Setting

We collected a consecutive sample of patients treated with four different types of care in a mental health care trust in the east of the Netherlands, covering a catchment area of 630000 inhabitants.

These four types of care concerned were:

Outpatient psychiatric clinics, in this context, are the services the general practitioner refers to patients for initial mental health care. This service provides acute crisis interventions, outpatient psychological and psychiatric treatment, and support.

Flexible Assertive Community Treatment (FACT) teams specialised in daily (outpatient) support and treatment for patients with serious mental illness (SMI). In the Netherlands, FACT teams are multidisciplinary outpatient teams with 8–10 professionals, such as psychiatrists, psychologists, nurses, and social workers, generally caring for 200 patients with SMI.

General admission wards admit first-onset patients and patients referred from FACT teams or outpatient clinics. Patients at these wards were eligible for inclusion in the current study when they resided on the ward for at least six days.

Long stay wards, providing residential care for patients with SMI. Patients all have a long history of receiving professional support and treatment, primarily in FACT teams.

The study was conducted and reported in accordance with the STROBE guidelines for reporting observational studies [22]. Screening for potential ID and data collection for aggressive incidents was done from May 2014 until January 2019. All patients treated in participating settings were asked to join the study to screen for potential IDnd participants who agreed to participate provided written informed consent for this.

Measures

The Staff Observation Aggression Scale-Revised (SOAS-R) was used to register aggression and is a widely used instrument to document the nature and severity of aggressive incidents. The SOAS-R records the following five aspects of aggressive incidents: (a) the apparent provocation, which led to the aggressive event, (b) the means used by the patient during the aggressive event, (c) the target of aggression, (d) the consequence(s) for the victim(s) of the aggression, and (e) the measures taken to stop the aggression, such as seclusion.

The inter-observer reliability of SOAS and SOAS-R aggression observations is acceptable, with a Cohen’s kappa of 0.61 and 0.74, respectively, and a Pearson product-moment correlation coefficient between independent raters of 0.87 [23]. The SOAS and SOAS-R severity scores correlate significantly with various other aggression measurement methods (i.e., correlations from 0.38 to 0.81) [24]. The scale is quick to complete, and there is no need for staff to be trained to use it.

We used the SCreener for Intelligence and Learning disability (SCIL) to detect patients with MID or BIF [23, 24]. Translation for use in English is in preparation. The SCIL is a test consisting of 14 questions, including educational level and small tasks intended to screen for patients’ overall cognitive abilities [24]. It was developed specifically to detect MID/BIF (IQ 50–85) in people in a range of settings, such as (mental) healthcare or social service settings and police stations and shelters for people experiencing homelessness. The reliability of the SCIL, as expressed in Cronbach’s alpha in the initial validation study, was good (0.83 in 318 adult subjects). The AUC value for detecting MID/BIF was 0.93, which is excellent. With 19 or lower as a cut-off score, the SCIL accurately classified 82% of people with MID/BIF. Of the ten people without MID/BIF, 9 (89%) were classified correctly as having no MID/BIF. In accordance with the SCIL manual, administering the SCIL requires no specific clinical skills.

The SCIL has recently been validated in patients with SMI in FACT teams [25]. The Cronbach’s alpha of the SCIL in that sample was 0.73. The AUC value for detecting MID/BIF and MID was 0.81, with percentages of correctly classified subjects of 73% and 79%, respectively. We used two cut-off scores: 19 and 15. Above 19 implies no MID/BIF, and 19 and below implies a (suspected) MID/BIF. The cut-off point of 15 and below implies a (suspected) MID [26]. In the following descriptions, we use two cut-off points, 19 for MID or BIF and 15 for MID only. The SCIL assessments used in the current study were performed between 2014 and 2018 [3]. We included all SOAS-R incidents reported in routine care between 2014 and 2019.

Patients were excluded from screening for potential ID with the SCIL based on (1) an inadequate grasp of the Dutch language, (2) lack of cooperation, (3) an inability, in the assessor’s opinion, to concentrate for at least 20 minutes in order to engage in the test as outlined in the instruction [26].

Nurses in inpatient and outpatient settings were trained to administer the SCIL. According to the questionnaire instructions, the SCIL was administered by a person not involved in the treatment. In the mental health trust where the study was carried out, the SOAS-R has been used since 2007 as a standard tool for nurses to log incidents and medical incident reports in inpatient and outpatient settings.

Demographic data and diagnosis were extracted from the electronic medical charts (EMC): age, gender, psychiatric diagnosis (DSM-IV-TR, as assessed by the psychiatrist), and Global Assessment of Functioning (GAF) score.

Statistical analyses

At the level of the patient, we identified whether a patient had shown an aggressive incident and whether a patient had shown outwardly directed physical aggression incidents against persons (so not against themselves). The total number of SOAS-R incidents per patient reported between 2014 and 2019 was also counted. Differences in the number of incidents between patients with or without MID/ BIF were tested using the Kruskal-Wallis rank order test because of extremely skewed frequencies. ​As mentioned earlier, the SCIL outcomes were categorised in scores of 19 and less, representing assumed MID/BIF and scores of 15 and less, representing assumed MID. BIF, MID and patient characteristics were cross-tabulated with having shown aggression incidents and physical aggression incidents against persons. We calculated chi-square statistics and Odds ratios to investigate the significance of the differences and the increased risk of showing (physical) aggression in relation to patient characteristics.

We also performed a logistic regression analysis to understand the association of these variables with having shown any aggression or physical aggression corrected for one another. A forward entry and backward deselection procedure were used. All variables selected from the EMC were entered in the analysis. Thus gender, age categories, diagnosis, MID or BIF as assessed with the SCIL. For the forward selection, variables with associations having a p-value of <0.2 were included in the logistic regression analysis, following the relevance criterion proposed by Hosmer and Lemeshow [27]. These were entered in 3 blocks: the demographic variables, the diagnoses, and the response categories in the SCIL.

Next, Poisson regression was applied to the number of incidents as we may expect a skewed distribution, and the number of incidents represents a count. Before applying the regression, the distribution of the number of incidents was tested. We applied forward entry and backward deselection to investigate which patient characteristics predicted the number of aggression incidents. We present the β, which as a rate ratio can be interpreted as a growth or downturn rate [28].

Ethical considerations

Ethical approval for the study was provided in 2014 by the ethical board of the University of Twente, Enschede, The Netherlands. All procedures performed in the current study were in accordance with the Helsinki Declaration of 1975, revised in 2008, and with comparable ethical standards. Data were analysed based on fully anonymised data that allowed none of the cases to be traced to an individual.

Results

SOAS-R score in general

In total, we found 1472 aggressive incidents in 196 (16.7%) of the 1565 patients. Most of the registered incidents occurred in inpatients. Only 36 outpatients were involved (18.3% of the 196, 2.2% of the complete sample). Of the 196 patients with an incident of aggression, 47 were involved in one incident, 84 patients between two and five incidents, and 65 were involved in over six incidents. 23 (11.7% of 196) patients were responsible for 751 aggression incidents (51.0% of 1472). The mean number of incidents was 7.53 per patient, with a maximum of 78 incidents. Of the 1565 patients, 105 patients were engaged in 269 physical, outwardly aggressive incidents (18.3% of the 1472 incidents). Of these 105 patients, 46 were involved in one incident, 51 in between two and five, and 8 in over six physically aggressive incidents. 20 (7.4%) of these patients were responsible for 137 (50.9%) of the 269 incidents. Both analyses show that approximately 10% of the patients account for half of the aggression incidents.

Sample and SCIL

We asked 1565 consecutive patients to participate. We obtained a SCIL score in 1174 cases (75.0%). 481 (41.0%) of the 1174 included patients showed a SCIL score of 19 and below (assumed MID/BIF). 239 (20.4%) showed a SCIL score of 15 and lower (assumed MID). In the various settings, the response was comparable with 71.5% at the outpatient services, 73.1% at the FACT teams, 75.5% at the long-stay wards and 78.9% at the admission wards [3]. The distribution of diagnoses was comparable in the participants compared to the non-responders, discarding selection bias by diagnosis.

SOAS-R and SCIL score, univariate analyses

Table 1 presents the number of aggression incidents over the SCIL negative or positive groups for MID and BIF. It shows that the proportion of patients engaging in (repeated) violent behaviour, in general, is higher in patients assumed to have MID or BIF. Furthermore, the table indicates that outwardly directed physical aggression occurred more often in patients with assumed MID. The odds ratios show that these increase in the higher categories above two incidents per patient. In general, the odds ratios for MID are higher than those for BIF, implying that an increasing number of incidents is associated with BIF but even more frequent in MID patients (Table 1.)

Table 1

BIF and MID compared to the number of aggression incidents.

		Aggression in general						Physical aggression
		N =	No aggression	One incident	2–5 incidents	> 5 incidents	P =	No aggression	One incident	2–5 incidents	> 5 incidents	P =
SCIL	above 19	693	636 (91.8%)	18 (2.6%)	19 (2.7%)	20 (2.9%)	<0.001	665 (96.0%)	13 (1.9%)	12 (1.7%)	3 (0.4%)	<0.001
	19 and below (BIF/MID)	481	385 (80.0%)	20 (4.2%)	51 (10.6%)	25 (5.2%)		432 (89.8%)	24 (5.0%)	23 (4.8%)	2 (0.4%)
OR	category / else		0.24 (0.17–0.36)	1.62 (0.85–3.11)	4.20 (2.45–7.22)	1.84 (1.10–3.35)		0.37 (0.22–0.59)	2.74 (1.38–5.45)	2.84 (1.40–5.78)	0.96 (0.16–5.77)
P =			< 0.001	0.148	< 0.001	0.045		< 0.001	0.004	0.004	0.965
SCIL	above 15	935	839 (89.7%)	25 (2.7%)	41 (4.4%)	30 (3.2%)	<0.001	903 (96.6%)	8 (0.9%)	9 (1.0%)	3 (0.3%)	<0.001
	15 and below (MID)	239	182 (76.2%)	13 (5.4%)	29 (12.1%)	15 (6.3%)		222 (92.9%)	16 (6.7%)	14 (5.9%)	2 (0.8%)
OR	category / else		0.36 (0.24–0.51)	2.09 (1.05–4.15)	3.01 (1.82–4.95)	2.02 (1.06–3.81)		0.46 (0.25–0.84)	8.31 (3.51–19.67)	6.40 (2.73–14.97)	2.62 (0.43–15.77)
P =			< 0.001	0.037	< 0.001	0.030		0.013	< 0.001	< 0.001	0.292

Table 2 presents the SCIL outcomes, patient characteristics, and aggression frequencies. A SCIL outcome of 19 and below (assumed BIF or MID) was associated with more aggression in general (OR = 2.50), as well as with more physical aggression (OR = 2.52). A SCIL outcome of 15 and below (assumed MID) was associated with more aggression in general (OR = 2.74), as well as with more physical aggression (OR = 3.06) (Table 2).

Table 2

Association between aggression BIF, MID and patient characteristics.

Predictors		Aggression incidents			OR	95% CI OR	Physical aggression			OR	95% CI OR
		No aggression	Aggression	P =			No aggression	Aggression	P =
Npatients =		1369	196				1460	105
Age		43.3	42.3	0.290			43.2	41.9
Age categories1a1	0–35	407 (29.7%)	73 (37.2%)	0.033	1.40	1.03–1.92	439 (30.1%)	41 (39.0%)	0.054	1.49	0.99–2.24
	36–59	840 (61.4%)	101 (51.5%)	0.009	0.67	0.49–0.90	887 (60.8%)	54 (51.4%)	0.059	0.68	0.46–1.00
	60+	122 (8.9%)	22 (11.2%)	0.179	1.29	0.79–2.09	134 (9.2%)	10 (9.5%)	0.906	1.04	0.53–2.05
Gender	Male	650 (47.5%)	108 (55.1%)	0.027	1.31	1.00–1.70	700 (47.9%)	58 (55.2%)	0.149	1.31	0.90–1.91
	Female	719 (52.5%)	88 (44.9%)				760 (52.1%)	47 (44.8%)
Scil Outcome	No SCIL	248 (18.1%)	43 (21.9%)	0.217	0.86	0.60–1.25	363 (24.9%)	28 (26.7%)	0.341	0.90	0.58–1.42
	SCIL	1121 (81.9%)	153 (78.1%)				1097(75.1%)	77(73.3%)
Scil Outcome	Scil > 19	637 (62.4%)	56 (36.6%)	0.000			665 (60.6%)	28 (26.7%)	0.000
	Scil 16–19 (BIF)	202 (19.8%)	40 (26.1%)				225 (20.5%)	17 (16.2%)
	Scil ≤15 (MID)	182 (17.6%)	57 (37.3%)				207 (18/9%)	32 (30.5%)
Borderline Intellectual Functioning (BIF)	SCIL > 19	637 (62.4%)	56 (36.6%)	0.000	2.49	1.83–3.39	665 (60.6%)	28 (36.4%)	0.000	2.52	1.61–3.95
	SCIL ≤ 19	384 (37.6%)	97 (63.4%)				432 (39.4%)	49 (63.6%)
Mild Intellectual Disability (MID)	SCIL >15	839 (82.2%)	96 (62.7%)	0.000	2.74	1.90–3.94	890 (81.1%)	45 (58.4%)	0.000	3.06	1.89–4.93
	SCIL ≤15	182 (17.8%)	57 (37.3%)				207 (18.9%)	32 (41.6%)
Diagnosis	Anxiety	205 (15.0%)	20 (10.2%)	0.043	0.65	0.39–1.05	215 (14.7%)	10 (9.5%)	0.142	0.61	0.31–1.18
	Depression	429 (31.3%)	39 (19.9%)	0.001	0.54	0.38–0.79	449 (30.8%)	19 (18.1%)	0.006	0.50	0.30–0.83
	Bipolar	122 (8.9%)	30 (15.3%)	0.005	1.85	1.20–2.84	135 (9.2%)	17 (16.2%)	0.020	1.89	1.10–3.28
	Psychotic disorder	223 (16.3%)	32 (16.3%)	0.529	1.00	0.67–1.50	282 (19.3%)	48 (45.7%)	0.762	0.92	0.53–1.59
	Schizophrenia	256 (18.7%)	74 (37.8%)	0.000	2.64	1.92–3.63	239 (16.4%)	16 (15.2%)	0.000	3.52	2.35–5.28
	Developmental disorder	173 (12.6%)	33 (16.8%)	0.104	1.40	0.93–2.10	173 (11.8%)	33 (31.4%)	0.725	1.12	0.63–1.95
	Alcohol and drug abuse disorder	171 (12.5%)	45 (23.0%)	0.000	2.09	1.44–3.02	188 (12.9%)	28 (26.7%)	0.000	2.46	1.56–2.89
	Personality disorder	570 (41.6%)	73 (37.2%)	0.2430.243	0.83	0.61–1.13	610 (41.8%)	33 (31.4%)	0.037	0.64	0.42–0.98
	Low GAF	437 (31.9%)	102 (52.0%)	0.000	2.32	1.71–3.16	488 (33.4%)	51 (52.4%)	0.003	1.83	1.23–2.74

1 For calculating the OR, the other categories are applied as reference, e.g. 0–35 is compared to ’else’

SOAS-R, SCIL score and patient characteristics, univariate analyses

Gender showed no significant association between aggression in general or more physical aggression. Only middle age showed an inverse and significant association with aggression (OR = 0.67, p = 0.009). Diagnosis of bipolar disorder (OR = 1.85, p = 0.005), schizophrenia (OR = 2.64, p<0.001), alcohol and drug abuse disorder (OR = 2.09, p<0.001) and a low GAF (OR = 2.32, P<0.001) were associated with an increased risk of aggression.

Schizophrenia (OR = 3.52, p<0.001), drug abuse disorder (OR = 2.46, p<0.001), and a low GAF (OR = 1.83, P<0.003) were associated with an increased risk of physical aggression. Only depressive disorders (OR = 0.54, p = 0.001) were associated with less aggression in general and less physical aggression (OR = 0.50, p = 0.006).

Logistic regression

The logistic regression analysis showed that patients who screened positive for BIF (OR = 2.00, p = 0.003) or MID (OR 2.89, p<0.001) were more at risk of showing aggressive incidents, as well as the patients with the diagnoses bipolar disorder (OR 3.07, p<0.001), schizophrenia (OR 2.75, p<0.001), and a low GAF (OR 1.72, p = 0.005). Logistic regression analysis with physical aggression as an outcome showed that patients with MID (OR 2.50, p<0.001), a bipolar disorder (OR 3.13, p = 0.007) or schizophrenia (OR 4.04, p<0.001) were more at risk of showing aggressive incidents.

Poisson regression

These findings were underlined by the Poisson regression of the number of physical aggression incidents per patient. This showed anxiety disorder (β = 0.62, p<0.001), bipolar disorder (β = 1.63, p<0.001), schizophrenia (β = 1.12, p<0.001), developmental disorder (β) = 0.69, p<0.001), drug abuse disorder (β = 1.18, p<0.001) and a SCIL of 15 and below (Assumed MID; β = 0.61, p<0.001) were all related to more incidents.

In short, screening positive for BIF and MID were both associated with significantly more aggression, and this association appears to be somewhat stronger for MID. Bipolar disorder, developmental disorders, schizophrenia and drug abuse disorders are associated with higher aggression rates (Table 3).

Table 3

Multivariable association between predictors and aggression.

	Predictors		Aggression incidents					Physical aggression
			B	SE	sig	Ex (b)	95% CI Ex (b)	B	SE	sig	Ex (b)	95% CI Ex (b)
Univariable associations	Male Gender1		0.046	0.197	0.815	1.05	0.71–1.54	0.235	0.267	0.378	1.26	0.75–2.13
	Age2	< 35	0.324	0.341	0342	1.38	0.71–2.69	0.721	0.485	0.137	2.06	0.79–5.32
		35–59	-0.248	0.325	0.445	0.78	0.41–1.47	0.134	0.470	0.776	1.14	0.45–2.87
	Anxiety disorder3		0.366	0.303	0.288	1.44	0.79–2.61	0.191	0.444	0.668	1.21	0.51–2.89
	Depressive disorder		0.255	0.266	0.338	1.29	0.76–2.17	0.135	0.377	0.721	1.14	0.55–2.39
	Bipolar disorder		1.218	0.305	<0.001	3.38	1.86–6.15	1.158	0.402	0.004	3.18	1.45–6.99
	Psychotic disorder		- 0.367	0.286	0.200	0.69	0.39–1.21	- 0.428	0.387	0.269	0.65	0.31–1.39
	Schizophrenia		1.122	0.263	<0.001	3.07	1.84–5.14	1.337	0.338	<0.001	3.81	1.96–7.39
	Developmental disorder		0.735	0.276	0.008	2.08	1.21–3.58	0.475	0.386	0.218	1.61	0.75–3.43
	Personality disorder		-0.004	0.210	0.983	0.99	0.66–1.50	- 0.481	0.300	0.109	0.62	0.34–1.11
	Alcohol and drug abuse disorder		0.643	0.240	0.007	1.90	1.19–3.05	0.703	0.308	0.022	2.02	1.11–3.69
	Low GAF		0.596	0.194	0.002	1.81	1.24–2.68	0.264	0.260	0.311	1.30	0.78–2.17
	SCIL4	16–19 (BIF)	0.716	0.239	0.002	2.07	1.30–3.31	0.379	0.333	0.254	1.46	0.76–2.80
		≤15 (MID)	1.115	0.228	<0.001	3.05	1.96–4.75	0.986	0.295	<0.001	2.68	1.50–4.78
Final model	Age	<35	0.562	0.199	0.005	1.75	1.18–2.59	0.632	0.257	0.014	1.88	1.14–3.11
	SCIL score between 16 and 19 (BIF)		0.684	0.237	0.004	1.98	1.25–3.15
	SCIL score 15 and below (MID)		1.080	0.223	<0.001	2.94	1.90–4.56	0.850	0.257	<0.001	2.34	1.41–3.88
	Bipolar disorder		1.075	0.279	<0.001	2.93	1.69–5.05	1.102	0.364	0.001	3.22	1.58–6.58
	Psychotic disorder		-0.463	0.273	0.090	0.63	0.37–1.07
	Schizophrenia		0.954	0.219	<0.001	2.59	1.69–3.99	1.370	0.268	<0.001	3.93	2.33–6.65
	Developmental disorder		0.616	0.261	0.018	1.85	1.10–3.09
	Alcohol and drug abuse disorder		0.560	0.231	0.015	1.75	1.11–2.75	0.638	0.286	0.026	1.89	1.08–3.32
	Low GAF		0.602	0.194	0.002	1.83	1.25–2.67

1 Female gender was used as reference category

2Elderly was used as reference category.

3 No diagnosis was used as reference category

4 SCIL > 19 was used as reference category.

Discussion

The current study indicates that patients with a (suspected) BIF or MID are more likely to display aggressive incidents than patients without ID. The odds ratios for aggression of any type and physical aggression toward others are 2.5 to 2.9, making it a highly significant finding. This is in line with other studies with people with ID showing that ID is associated with higher rates of aggression incidents [29, 30]. Eight percent of patients without an ID had been engaged in aggressive incidents, keeping with previous Dutch data from international studies [18]. However, in the BIF patient group, 20%, and in the MID patient group, 24% of patients showed aggressive incidents, roughly half of which were physical aggression. This represents a sizable contribution to the risk of being confronted with aggression in the wards.

Regarding the number of incidents per patient, about half of the patients involved in aggression incidents were involved in 2–5 incidents, and just a small group of patients were responsible for more than five incidents. This is in keeping with previous studies. For instance, Bowers et al. [6] found that on average, 45% of patients with violent behaviour were involved in more than one incident. In a study by Broderick et al. [31] in a Canadian multihospital state psychiatric system, just 1% of the study population participated in 28.7% of all violent assaults. Considering the current findings, it seems fair to conclude that earlier studies and reviews concerning aggression in mental health may have paid too little attention to the role of impairments in intellectual functioning as a potential determinant of aggressive behaviour. This was also recently concluded in Weltens et al.’s systematic review [32]. However, Tsiouris et al. [33] reported in a large sample of persons with an ID that "impulse control, mood dysregulation and perceived threat appear to underlie most of the aggressive behaviours reported" across various settings. Another study in a Forensic Psychiatric Hospital [34] also concluded that chronic violent behaviour was associated with cognitive impairment or brain damage. In a study by Verstegen, a clear association between impulsivity and aggression was found [35]. Our study confirmed that BIF/MID is an often unnoticed and undiagnosed factor that significantly contributes to physically aggressive behaviour, supporting our knowledge that patients with lower cognitive functioning, in general, may have more problems with impulse regulation.

Our findings suggest that specific patient characteristics increase the risk of being involved in aggressive incidents. These include young ages of up to 35 years, which is in line with other studies with inpatients and reviews in adult psychiatry and ID populations [6, 29]. Other characteristics that showed an increased aggression risk were diagnosis of schizophrenia, harmful use of alcohol and drugs and bipolar affective disorder, and a GAF score below 45. The same patient characteristics were important when analysing only physical aggression incidents, showing that aggressive incidents per se appear to have similar patient risk factors. While this is one of the first comprehensive studies examining the association between ID and risk of aggression, schizophrenia and drug and alcohol use have commonly been associated with aggression in mental health care [6, 34, 35] and studies with people with ID. [12, 16, 29].

As we can learn from studies with people with ID based on interviews with people with intellectual disability [10, 36], people with ID often experience a lack of structure in their daily life, and staff may often place too many demands on them. This is coupled with the fact that these patients regularly have difficulty dealing with emotions, the complexity of social interaction, and other stressors. Challenging behaviour can also be related to a number of unmet needs that should be addressed, such as medical issues (e.g., pain) or communication difficulties, among others. It is also important to better understand patients’ capabilities by staff and others—both in terms of their intellectual, emotional and adaptive skills. It may be helpful to offer patients counselling or training to better cope with emotions and impulses to help to reduce the occurrence of CB. Other potential ways forward are functional analyses of earlier CB and positive behaviour support for patients. The results of a large-scale meta-analysis indicated that in patients/clients with ID, behavioural treatments based on Function Analysis tend to be more effective than pharmacological interventions [37]. Function Analysis derived from Behaviour Therapy to systematically identify the reinforcers of CB allows staff to mitigate the consequence and replace it with more prosocial behaviour [13, 38–41].

Clinical implications

Earlier studies examining the associations between aggressive behaviour and patient characteristics focused on diagnoses, psychiatric history, staff training, restraint and workload, the interaction between patient and staff, care processes, and ward architecture and environment [31, 42]. Based on the information from these studies, various suggestions and programs to reduce aggression were developed. We know from studies with people with ID that it is advisable to meet a person’s needs better (e.g., engaging in appropriate support and communication, presenting information in an accessible way, and approaching treatment from a biopsychosocial approach), all of which would likely result in fewer behaviours that challenge [43-46]. Until now, mental health services seem to be frequently not equipped enough to meet those needs. The current study reminds us of the importance of the intellectual functioning of each individual patient in general psychiatry to prevent aggressive incidents, especially those responsible for a large number of incidents. In light of our findings, we recommend screening patients for ID as part of any assessment at the start of treatment in order to support the prevention of aggressive incidents in psychiatric care. The SCIL can be helpful in psychiatric care [28, 47] and give a quick first impression. Treatment and support to the needs can be adapted according to the SCIL category. As such, avoid over-demanding and stress. The staff’s attitudes towards people with ID may also play a part in detecting and preventing aggressive incidents [16, 48].

Recommendations

In mental health care, future studies should examine the causes and reasons for aggressive incidents in patients with MID/BIF and learn more about how these patients differ or resemble those in other studies with people/patients with ID. For example, we do not yet know much about the mental problems in the patient group with ID in mental health care and how this affects behaviour. We know from another of our studies [2] that patients with assumed ID frequently experience neglect and (sexual)abuse, but this is not recognised enough in mental health care. We also do not know the influence of social factors between patients within ward settings, as the current study did not look into that level of detail.

Next, intervention studies in patients with MID and BIF are needed in mental health care to study which measures (such as positive behaviour support [48, 49]) we know are helpful in ID care and preventing aggressive incidents. Cooperation with ID services can therefore be helpful. Attention to this vulnerable, large patient group in mental health is of great importance and should get more attention in the training of professionals.

Limitations and strengths

In this study, we included outpatients and inpatients over a period of six years. This gave us a large sample of consecutive patients in a quite big mental healthcare trust. This large sample also provided the opportunity to study predictors of incidents of physical aggression as a subgroup of all aggression. Another strength is the high percentage of included patients, resulting in a representative sample. A limitation is that aggression incidents are likely to have been underreported with the SOAS-R.

(XLSX)

Click here for additional data file.

(XLSX)

Click here for additional data file.

(DOCX)

Click here for additional data file.

21 Mar 2022

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Robert Didden

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. PLOS requires an ORCID iD for the corresponding author in Editorial Manager on papers submitted after December 6th, 2016. Please ensure that you have an ORCID iD and that it is validated in Editorial Manager. To do this, go to ‘Update my Information’ (in the upper left-hand corner of the main menu), and click on the Fetch/Validate link next to the ORCID field. This will take you to the ORCID site and allow you to create a new iD or authenticate a pre-existing iD in Editorial Manager. Please see the following video for instructions on linking an ORCID iD to your Editorial Manager account: https://www.youtube.com/watch?v=_xcclfuvtxQ

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: No

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: No

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Thank you for the opportunity to review this manuscript, which examines the frequency of aggressive behaviours among people with mild intellectual disability/borderline intellectual functioning within a general mental health care setting. The large sample size is a strength of this study, but I have concerns about the novelty of the findings, the lack of depth of discussion of factors associated with behaviours that challenge, and the language used throughout. People with intellectual disability and co-occurring mental illness have multiple and complex support needs, and yet health and mental health service systems are frequently not equipped to meet those needs. It is already known that people with intellectual disability are more likely than people without intellectual disability to have behaviours that challenge including aggression towards self and others. The factors contributing to such behaviours can be complex, including difficulties with communication, stressors within the environment, and underlying physical and mental health conditions (e.g. physical pain, mental illness), among others. These factors can be proactively addressed but currently the manuscript lacks any discussion about how to equip mental health services to do so. I would also very much recommend that the authors review the language used throughout this manuscript. I have made some specific comments below.

Introduction

• It would be appropriate to use person-first language throughout, i.e. rather than “ID patients” say people with ID

• There is inconsistent use of ID and intellectual disability throughout

• Intellectual disability does not need to be capitalized, same with bipolar disorder

• Page 3: Plagued is a very emotive word and I would not consider it appropriate in this context. Suggest just saying that most previous studies have been single centre reports and say why this is a limitation.

Method

• Page 4: 630.000 should read 630,000

• Page 4: can outpatient psychiatric clinics be defined, as done for other types of care

• Page 4: expand SMI on first mention

• Page 6: expand GAF

• Page 6: more information is needed for the logistic regression analysis, what were the outcomes? What were the predictors? Were these included in a single model?

Results

• Page 7: How was missing SCIL data handled? E.g. in the SOAS-R data the proportion is reported with 1174 as the denominator (196/1174=16.7%), but for others the complete sample is used (36-1565=2.3%)

• Similar to above, suggest checking the rounding of decimal places as some of these are rounded incorrectly assuming a total sample of 1565

• Page 7, “to” is missing between “84 patients in two” and “up to five incidents”

• Figures 1 & 2: These appear to be cumulative frequencies and are not an appropriate or helpful way to present these data. This information is adequately described in text without figures.

• Table 2: Reference categories are not indicated and in some cases not provided (e.g. age categories). Why is age and SCIL included as both a continuous and categorical variable? I am also unsure of the appropriateness of having two separate variables for SCIL score (i.e. >19 vs ≤19 and then >15 vs ≤15); why not have a single variable with exclusive categories for people with no mild intellectual disability or borderline intellectual functioning, people with borderline intellectual functioning, and people with mild intellectual disability (SCIL scores >19, 16-≤19, and ≤15, respectively?)

• Table 3: reference categories are not provided but are needed for interpretation. Why do categories for things like age and SCIL change to those provided in Table 2? Why are only significant results presented?

• Page 8: did the authors perform a linear regression with the number of physical aggression incidents as the outcome? (“These findings were underlined by the regression of the number of physical aggression incidents per patient, which detected bipolar disorder (β=0.169, p=0.014), schizophrenia (β=0.144, p0.008), drug abuse disorder (β=0.195, p0.001) and a SCIL below 15 (β=0.138, p=0.009) as predictors.”) If so this is the first mention of it; this should have been described in the analysis section.

Discussion:

• I have concerns about the overall tone and language used throughout the manuscript but particularly the discussion. Terms such as “aggressive patients” and “violent patients” are stigmatising and not helpful. Similarly, describing people with intellectual disability as having an “inability to properly cope with emotions” is also not helpful, especially when considering the complex health and support needs of people with intellectual disability and co-occurring mental illness, and the systemic neglect and abuse people with intellectual disability frequently experience within health and mental health systems.

• Similar to the introduction, people first language is more appropriate- for example instead of saying “ID studies based on patient interviews”, say interviews with people with intellectual disability.

• The authors draw a number of associations in their discussion but on little basis – e.g. using their data as evidence of problems with impulse regulation; behaviours that challenge can also be related to a number of unmet needs that should be addressed such medical issues (e.g. pain), communication difficulties etc.

• Topics such as positive behviour support are mentioned only briefly in the discussion, yet this information is much more helpful for clinicians for preventing behaviours than other text provided in the discussion

• In the conclusion and clinical implications it would be more appropriate and helpful to state that screening for intellectual disability in a mental health setting may be helpful to inform strategies to better meet a person’s needs (e.g. engaging appropriate supports, presenting information in an accessible way, approaching treatment from a biopsychosocial approach, all of which would likely result in fewer behaviours that challenge).

Reviewer #2: This is interesting research that explores the frequency of aggressive behaviours and incidents among people suspected to have MID/BIF.

Introduction

- Are there any statistics that show the prevalence use of coercive or restrictive measures in Netherlands on people with MID and BIF?

- Explanation about the rationale for the aims of the research study is recommended. How would it useful for clinicians when they identified patients suspected to have MID/BIF in their work?

- Explanation about the use of screening tools to screen suspected patients with MID/BIF instead of using proper diagnosis as defined by ICD/DSM is recommended.

Methods

- Lack of details regarding the procedures. Does the SOAS-R administered by the same staff for the patients or administered by different staffs taking care of the patient?

- How does the screening of the participants using SCIL carried out?

- Does the staff administer both SCIL and SOAS-R for the same patient?

- Do the patients have both MID/BIF and co-morbid mental health conditions? How would the results take into account that the patient may display aggressive behaviours not because of their cognitive functioning and rather their mental health conditions?

- In the results table (Table 2 and 3) – how is developmental disorder defined? Some developmental disorders may include symptoms of MID/BIF, how will that affect the results?

Discussion

- Some patients are staying in long stay wards – does social factors play a part that may contribute to the aggressive behaviours instead of cognitive impairment

- How does screening patients for ID support the prevention of aggressive incidents as mentioned in clinical implications?

- Do the staff’s attitudes towards people with ID play a part to detect and prevent aggressive incidents?

- As the staff interacts with the patients for a long period, it is a possibility that their relationship with the patients may influence their ratings on the scales, would it become discriminatory tool to segregate people with suspected ID and those without?

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

20 May 2022

PONE-D-21-40785

Response to the reviewers' comments. We apply italics to show the reviewers questions and plain text for our answers:

Reviewer #1:

Question # 1 Thank you for the opportunity to review this manuscript, which examines the frequency of aggressive behaviours among people with mild intellectual disability/borderline intellectual functioning within a general mental health care setting. The large sample size is a strength of this study, but I have concerns about the novelty of the findings, the lack of depth of discussion of factors associated with behaviours that challenge, and the language used throughout.

Response #1

We thank the reviewer for this comment. Indeed, this study is one of the largest naturalistic empirical studies we know of. With respect to the concern of the novelty of the findings, in the discussion (page 10 line 316) we refer to Weltens et al of October 2021. They noted in a review of aggression factors on admission wards in Mental Health Care, that an IQ of 50 - 84 is a risk factor for coercive measures. They noted that this has not been studied as a direct factor for aggression before. They suggest that it should be a topic for future research. In our point of view, this is the main topic of the current study, which according to Weltens, has not been studied in detail before.

Question # 2. People with intellectual disability and co-occurring mental illness have multiple and complex support needs, and yet health and mental health service systems are frequently not equipped to meet those needs. It is already known that people with intellectual disability are more likely than people without intellectual disability to have behaviours that challenge, including aggression towards self and others.

The factors contributing to such behaviours can be complex, including difficulties with communication, stressors within the environment, and underlying physical and mental health conditions (e.g., physical pain, mental illness), among others. These factors can be proactively addressed, but currently the manuscript lacks any discussion about how to equip mental health services to do so. I would also very much recommend that the authors review the language used throughout this manuscript. I have made some specific comments below.

Response #2 Discussion: Thank you for this comment. We added text to the current draft to cover these discussion points (page 11, line 334-347):

As we can learn from ID studies based on interviews with people with intellectual disability [10, 37], people with ID often experience a lack of structure in their daily life, and staff may often be placing too many demands on them. This is coupled with the fact that these patients regularly have difficulty in dealing with emotions, the complexity of social interaction, and other stressors. Behaviours that challenge can also be related to a number of unmet needs that should be addressed such medical issues (e.g., pain), communication difficulties etc. It is also important to have a better understanding of patients' capabilities by staff and others - both in terms of their intellectual, emotional and adaptive skills. It may bel be helpful offering counselling or training to better cope with emotions and impulses to help to reduce the occurrence of CB. Other potential ways forward are functional analyses of earlier CB and positive behaviour support for patients. The results of a large-scale meta-analysis indicated that in clients with ID, behavioural treatments based on Function Analysis tend to be more effective than pharmacological interventions [38]. Function Analysis derived from Behaviour Therapy to systematically identify the reinforcers of CB allows staff to mitigate the consequence and replace it with more prosocial behaviour [13, 40-42].

Introduction

Question #3• It would be appropriate to use person-first language throughout, i.e., rather than “ID patients” say people with ID;

Response # 3 This is done.

Question #4 There is inconsistent use of ID and intellectual disability throughout;

Response #4 This is done. We used intellectual disability at first mention and ID throughout.

Question #5 Intellectual disability does not need to be capitalized, same with bipolar disorder.

Response#5 We adjusted this. As stated above we used the full word (intellectual disability) at first mention and the abbreviation (ID) later on.

Question #6 Page 3: Plagued is a very emotive word and I would not consider it appropriate in this context. Suggest just saying that most previous studies have been single centre reports and say why this is a limitation.

Response # 6/ We changed this, page 4, line 93: Many previous publications have been single centre reports, which makes comparison and generalizing conclusions difficult.

Method

Question #7 Page 4: 630.000 should read 630,000;

Response #7. This was changed, page 4, line 129.

Question #8 Page 4: can outpatient psychiatric clinics be defined, as done for other types of care;

Response #8. We explained this as follows (page 5, line 131): 1. Outpatient psychiatric clinics in this context are the services patients are referred to by the General Practitioner for initial mental health care. This service provides acute crisis interventions as well as outpatient psychological and psychiatric treatment and support.

We also refined the next sentence (page 5 line 134): 2. Flexible Assertive Community Treatment (FACT) teams specialised in daily (outpatient) support and treatment for Serious Mentally Ill (SMI) patients. In the Netherlands, FACT teams are multidisciplinary outpatient teams with 8–10 professionals, such as a psychiatrist, psychologist, nurses, and social workers, generally caring for 200 SMI patients.

Question #9 Page 4: expand SMI on first mention;

Response #9 Page 5 line 135: Serious Mentally Illness (SMI)

Question #10 Page 6: expand GAF;

Response #10 this was done page 6 line 192

Question # 11 Page 6: more information is needed for the logistic regression analysis, what were the outcomes? What were the predictors? Were these included in a single model?

Response # 11 We added the following section to the statistical analysis, page 6 line 207 onwards:

A forward entry, backward deselection procedure was used. All variables selected from the EMC were entered in the analysis, thus gender, age categories, diagnosis, MID or BIF as assessed with the SCIL. For the forward selection, variables with associations having a p-value of <0.2 were included in the logistic regression analysis, following the relevance criterion proposed by Hosmer and Lemeshow [26]. These were entered in 3 blocks, first the demographic variables, then the diagnoses and finally the response categories in the SCIL.

Next, a Poisson regression was applied to the number of incidents as we may expect a skewed distribution and the number of incidents represents a count. Before applying the regression, the distribution of the number of incidents was tested. We applied forward entry, backward deselection to investigate which patient characteristics predicted the number of aggression incidents. We present the Ex(β), which can be interpreted as a growth or downturn rate [27].

Results

Question # 12 Page 7: How was missing SCIL data handled? E.g., in the SOAS-R data the proportion is reported with 1174 as the denominator (196/1174=16.7%), but for others the complete sample is used (36-1565=2.3%).

Response #12, We thank the reviewer for this comment. We understand the confusion regarding the mentioning of various numbers in the first paragraphs of the results section. To clarify this, we first presented the aggression incidents in the full sample. We then present the number of SCIL’s assessed. We did not handle missing data, but described it in more detail in table 2. We did this deliberately to allow some insight into those patients without a SCIL outcome, as compared to those with a SCIL outcome. Especially to answer the question whether aggression occurs more in patients not assessed with the SCIL than in those without. This proved not to be the case, as can be seen in table 2, line 10, where we compared them. Please note that table 2 had the full sample as denominator. Table 1 shows the findings from all patients who have had a SCIL. We chose to describe what was assessed.

Question # 13 Similar to above, suggest checking the rounding of decimal places as some of these are rounded incorrectly assuming a total sample of 1565.

Response #13. We checked all tables, with respect to counters and rounding of percentages.

Question # 14 Page 7, “to” is missing between “84 patients in two” and “up to five incidents”.

Response #14. We corrected this.

Question # 15 Figures 1 & 2: These appear to be cumulative frequencies and are not an appropriate or helpful way to present these data. This information is adequately described in text without figures.

Response #15. We removed the figures and extended the text page 7, line 229 onwards:

In total, we found 1472 aggressive incidents in 196 (16.7%) of the 1565 patients. Most of the registered incidents occurred in inpatients. Only 36 outpatients were involved (18.3 % of the 196, 2.2% of the complete sample). Of the 196 patients with an incident of aggression, 47 were involved in one incident, 84 patients between two and five incidents, and 65 patients were involved in over six incidents. 23 (11.7% of 196) patients were responsible for 751 aggression incidents (51.0% of 1472). The mean number of incidents was 7.53 per patient, with a maximum of 78 incidents. Of the 1565 patients, 105 were engaged in 269 physical, outwardly aggressive incidents (18.3% of the 1472 incidents). Of these 105 patients, 46 were involved in one incident, 51 in between two and five incidents, and 8 in over six physically aggressive incidents. 20 (7.4%) of these patients were responsible for 137 (50.9%) of the 269 incidents. Both analyses show that approximately 10% of the patients account for half of the aggression incidents.

The current study reminds us of the importance of the intellectual functioning of each individual patient in general psychiatry in an attempt to prevent aggressive incidents, especially those responsible for a large number of incidents.

Question # 16 Table 2: Reference categories are not indicated and, in some cases, not provided (e.g., age categories). Why is age and SCIL included as both a continuous and categorical variable? I am also unsure of the appropriateness of having two separate variables for SCIL score (i.e., >19 vs ≤19 and then >15 vs ≤15); why not have a single variable with exclusive categories for people with no mild intellectual disability or borderline intellectual functioning, people with borderline intellectual functioning, and people with mild intellectual disability (SCIL scores >19, 16-≤19, and ≤15, respectively?).

Response #16. We chose to analyze the categories in the SCIL, identifying BIF and MID as these were developed to assist treatment decision processes (Seelen et al, 2019). As suggested we first present them in three groups (SCIL scores >19, 16-≤19, and ≤15) and then in two (19 vs ≤19 and then >15 vs ≤15). The last was done to calculate odds ratios.

Seelen-de Lang BL, Smits HJH, Penterman BJM, Noorthoorn EO, Nieuwenhuis JG, Nijman HLI. Screening for intellectual disabilities and borderline intelligence in Dutch outpatients with severe mental illness. J Appl Res Intellect Disabil. 2019 Sep;32(5):1096-1102. doi: 10.1111/jar.12599. Epub 2019 Apr 29.

Question # 17. Table 3: reference categories are not provided but are needed for interpretation. Why do categories for things like age and SCIL change to those provided in Table 2? Why are only significant results presented?

Response # 17. A forward entry and backward deselection procedure were used, and the final model was presented. We added the uncorrected univariable comparisons to the table to improve interpretation. The reference categories can now be checked in the new table 3, as they are noted below the table.

Question # 18. Page 8: did the authors perform a linear regression with the number of physical aggression incidents as the outcome? (“These findings were underlined by the regression of the number of physical aggression incidents per patient, which detected bipolar disorder (β=0.169, p=0.014), schizophrenia (β=0.144, p0.008), drug abuse disorder (β=0.195, p0.001) and a SCIL below 15 (β=0.138, p=0.009) as predictors.”) If so, this is the first mention of it; this should have been described in the analysis section.

Response # 18. No, we performed a Poisson regression analysis of the number of aggression incidents. We elaborated on this analysis on page 7 line 214 onwards, as follows:

Next, a Poisson regression was applied to the number of incidents as we may expect a skewed distribution and the number of incidents represents a count. Before applying the regression, the distribution of the number of incidents was tested. We applied forward entry, backward deselection to investigate which patient characteristics predicted the number of aggression incidents. We present the Ex(β), which can be interpreted as a growth or downturn rate [27].

Also in the results section, page 9 line 287 onwards:

Poisson regression

These findings were underlined by the regression of the number of physical aggression incidents per patient, which detected the following as predictors: anxiety disorder (Ex(β)=2.01, p<0.001), bipolar disorder (Ex(β)=5.14, p<0.001), schizophrenia (Ex(β)=3.11, p<0.001), developmental disorder (Ex(β)=2.11, p<0.001), drug abuse disorder (β=3.21, p0.001), and a SCIL below 15 (Ex(β)=1.74, p<0.001).

Discussion:

Question # 19 I have concerns about the overall tone and language used throughout the manuscript but particularly in the discussion. Terms such as “aggressive patients” and “violent patients” are stigmatising and not helpful;

Response # 19. We changed this to: involved in aggression incidents in lines 84, 308, 311 and in line 325 to patients with violent behaviour as the articles refer to violence.

Question # 20. Similarly, describing people with intellectual disability as having an “inability to properly cope with emotions” is also not helpful, especially when considering the complex health and support needs of people with intellectual disability and co-occurring mental illness, and the systemic neglect and abuse people with intellectual disability frequently experience within health and mental health systems;

Response 20 # We changed this to: these patients regularly have difficulty in dealing with emotions in line and mentioned the complex need in the discussion (lines 336 -341), as follows:

This is coupled with the fact that these patients regularly have difficulty in dealing with emotions, the complexity of social interaction, and other stressors. Behaviours that challenge can also be related to a number of unmet needs that should be addressed such medical issues (e.g., pain), communication difficulties etc. It is also important to have a better understanding of patients' capabilities by staff and others - both in terms of their intellectual, emotional and adaptive skills. It may bel be helpful offering counselling or training to better cope with emotions and impulses to help to reduce the occurrence of CB.

Question # 21 Similar to the introduction, people first language is more appropriate- for example instead of saying “ID studies based on patient interviews”, say interviews with people with intellectual disability;

Response #21 We changed in line 334 as follows: As we can learn from ID studies based on interviews with people with intellectual disability

Question #22. The authors draw a number of associations in their discussion but on little basis – e.g., using their data as evidence of problems with impulse regulation.

Response #22 We now used the words, eg in line 325:

Our findings suggest that specific patient characteristics increase the risk of being involved in aggressive incidents.

Later in the same paragraph we say that these findings are in line with other studies in psychiatry and in ID studies, line 331:

While this is one of the first comprehensive studies examining the association between ID and risk of aggression, schizophrenia and drug and alcohol use have commonly been associated with aggression in mental health care [ 6, 35, 36] and ID studies [12 16, 30].

We therefore think that we do not draw such conclusions from our own study.

Question # 23 Behaviours that challenge can also be related to a number of unmet needs that should be addressed such medical issues (e.g. pain), communication difficulties etc.

Response #23a We added this in line 336: Behaviours that challenge can also be related to a number of unmet needs that should be addressed such medical issues (e.g., pain) or communication difficulties among others.

Response #23b Next, we added in clinical implications, lines 354-357: We know from studies in people with ID that it is advisable to better meet a person’s needs with engaging appropriate supports, presenting information in an accessible way, approaching treatment from a biopsychosocial approach, all of which would likely result in fewer behaviours that challenge. Until now, mental health services frequently do not seem to be adequately equipped to meet those needs.

Response #23c In recommendations, we added some lines (368-372) for future research:

In mental health care, future studies should examine the causes and reasons are for aggression incidents in patients with MID/BIF, and to learn more about how these patients differ or resemble those from people in other ID studies. For example, we do not yet know much about the mental problems in the ID patient group in MHC, and how this affects behaviour. We know from another of our studies [2] that patients with ID frequently experience neglect and (sexual) abuse , but this is not necessarily recognized enough in mental health care. We also do not know the influence of social factors between patients within ward settings, as the current study did not look into that level of detail.

Question # 24 Topics such as positive behaviour support are mentioned only briefly in the discussion, yet this information is much more helpful for clinicians for preventing behaviours than other text provided in the discussion.

Response #24 We added several lines in the discussion and clinical implications, such as lines 354 – 357:

We know from studies in people with ID that, it is advisable to better meet a person’s needs (e.g., engaging appropriate supports and communication, presenting information in an accessible way, approaching treatment from a biopsychosocial approach), all of which would likely result in fewer behaviours that challenge. Until now mental health services systems seem to be frequently not equipped enough to meet those needs.

And lines 375- 379:

Next, intervention studies in patients with MID and BIF are needed in mental health care to study which measures (such as positive behaviour support) that we know are helpful in ID care, are effective in preventing aggressive incidents. Cooperation with ID services can therefore be useful. Attention to this vulnerable, large patient group in the mental health is of great importance and should get more attention in the training of professionals.

Question # 25 In the conclusion and clinical implications, it would be more appropriate and helpful to state that screening for intellectual disability in a mental health setting may be helpful to inform strategies to better meet a person’s needs (e.g., engaging appropriate supports, presenting information in an accessible way, approaching treatment from a biopsychosocial approach, all of which would likely result in fewer behaviours that challenge).

Response #25 This was addressed in line 358-365: The current study reminds us of the importance of the intellectual functioning of each individual patient in general psychiatry in an attempt to prevent aggressive incidents, especially those responsible for a large number of incidents. In light of our findings, we recommend screening patients for ID as part of any assessment at the start of treatment in order to support the prevention of aggressive incidents in psychiatric care. The SCIL can be helpful in psychiatric care [29,45] and give a quick first impression. Treatment and support to the needs can then be adapted according to the SCIL category. As such, avoid over-demanding and stress. The staff’s attitudes towards people with ID may also play a part to detect and prevent aggressive incidents.

Reviewer #2: This is interesting research that explores the frequency of aggressive behaviours and incidents among people suspected to have MID/BIF.

Introduction

Question # 26 Are there any statistics that show the prevalence use of coercive or restrictive measures in Netherlands on people with MID and BIF?

Response # 26. Yes, one of the first studies of the same research line investigated this. In the study of Nieuwenhuis et al, 2017 these Figures were presented: Nieuwenhuis JG, Noorthoorn EO, Nijman HL, Naarding P, Mulder CL. A Blind Spot? Screening for Mild Intellectual Disability and Borderline Intellectual Functioning in Admitted Psychiatric Patients: Prevalence and Associations with Coercive Measures. PLoS One. 2017 Feb 2;12(2):e0168847. doi: 10.1371/journal.pone.0168847.

Also, a nationwide study confirmed this association: 22. Noorthoorn EO, Voskes Y, Janssen WA, Mulder CL, van de Sande R, Nijman HL, Smit A, Hoogendoorn AW, Bousardt A, Widdershoven GA (2016). Seclusion Reduction in Dutch Mental Health Care: Did hospitals meet goals set? Psychiatric Services, 1;67(12):1321-1327 DOI: appips201500414.

We added these references to the reference list and included the following text in the introduction, page 4, line 109:

In a study on admission wards [1], we showed patients with BIF/ MID had an increased risk of involuntary admission (OR 2.71; SD 1.28-5.70) and coercive measures (OR 3.95, SD 1.47-10.54). These findings were confirmed in nationwide data gathered in 2014, where intellectual impairment also showed an association with increased risk of seclusion and other coercive measures [22]. Internationally, there is evidence that patients with BIF/ID account for more and longer seclusion and restraint events [2,3].

Thus, the following references were added:

2. Turner KV & Mooney P A comparison of seclusion rates between intellectual disability and non-intellectual disability services: the effect of gender and diagnosis, The Journal of Forensic Psychiatry & Psychology, 2016, 27:2, 265-280, Doi 10.1080/14789949.2015. 1122822

3. Lepping P, Masood B, Flammer E, Noorthoorn EO. Comparison of restraint data from four countries. Soc Psychiatry Psychiatr Epidemiology. 2016 Sep;51(9):1301-9. doi: 10.1007/s00127-016-1203-x.

Question # 27. Explanation of the rationale for the aims of the research study is recommended. How would it useful for clinicians when they identified patients suspected to have MID/BIF in their work?

Response # 27 This was addressed in the lines 357-361: The current study reminds us of the importance of the intellectual functioning of each individual patient in general psychiatry in an attempt to prevent aggressive incidents, especially those responsible for a large number of incidents. In light of our findings, we recommend screening patients for ID as part of any assessment at the start of treatment in order to support the prevention of aggressive incidents in psychiatric care.

Question # 28. Explanation about the use of screening tools to screen suspected patients with MID/BIF instead of using proper diagnosis as defined by ICD/DSM is recommended.

Response #28. We added the following lines (362-366): The SCIL can be helpful in psychiatric care [29,45] and give a quick first impression. Treatment and support to the needs can then be adapted according to the SCIL category. As such, avoid over-demanding and stress. The staff’s attitudes towards people with ID may also play a part to detect and prevent aggressive incidents.

Also, the suggestion of the reviewer addresses an issue which is the motivation to perform a line of studies into MID and BIF. From these former studies, we identified an underreport in the ‘proper diagnoses’. In our point of view clinicians do not estimate the intellectual abilities of patients completely adequate, especially in the serious mentally ill patients. This was especially addressed in the publication : Nieuwenhuis JG, Lepping P, Mulder NL, Nijman HLI, Veereschild M, Noorthoorn EO. Increased prevalence of intellectual disabilities in higher-intensity mental healthcare settings. BJPsych Open. 2021 Apr 22;7(3):e83. doi: 10.1192/bjo.2021. 28. In the introduction this is addressed in the first sentences (page 3 line 63 - 67:

Mild Intellectual Disability (MID) and Borderline Intellectual Functioning (BIF) are highly prevalent in general Mental Health Care but often stay unnoticed [1,2]. Our research group has previously shown that in the Netherlands, the prevalence of MID/BIF increases by setting, from 27% in outpatient settings, to 40% in Flexible Assertive Community Treatment (FACT) teams and admission wards, to 67% in long-stay wards [3].

Methods

Question #29 Lack of details regarding the procedures. Does the SOAS-R administered by the same staff for the patients or administered by different staff taking care of the patient? How does the screening of the participants using SCIL carried out? Does the staff administer both SCIL and SOAS-R for the same patient?

Response # 29 These three questions were addressed in page 6, line 185-189:

Nurses working both in inpatient and outpatient settings were trained to administer the SCIL. According to the SCIL questionnaire instructions, the SCIL was administered by a person not involved in the treatment. In the Mental Health Trust where the study was carried out, the SOAS-R has been used since 2007 as a standard tool for nurses to log incidents and medical incident reports, both in inpatient and outpatient settings.

Question # 30 Do the patients have both MID/BIF and co-morbid mental health conditions? How would the results take into account that the patient may display aggressive behaviours not because of their cognitive functioning and rather their mental health conditions?

Response # 30 This was a study in a Mental Health Service. Patients were referred to the service to be treated for psychiatric illnesses, with medication or psychotherapy. Our main issue is that more knowledge of MID or BIF may lead to taking the cognitive ability of the patient into account in treating the illness. MID and BIF were determined with the SCIL, but all patients had one psychiatric illness or another, as may be seen in table 2. In the results, we of course take the association of both into account when looking for the association between MID/BIF and aggression. For this reason, in the logistic regression and in the Poisson regression, MID and BIF are included together with the demographic characteristics and the diagnoses. All are included in the regression models.

Question # 31 In the results table (Table 2 and 3) – how is developmental disorder defined? Some developmental disorders may include symptoms of MID/BIF, how will that affect the results?

Response # 31 As stated above, it was a study in a Mental Health Service. Patients could have more than one diagnosis. Developmental disorders predominantly concerned the ICD-10 F8 diagnoses and include the next ICD-10 codes: F81.0; F81.2; F81.8; F81.9; F82; F80.1; F80.2; F80.0; F80.9; F84.0; F84.2; F84.3; F84.5; F84.9. Also, the next DSM-IV tr codes were included in this category: 315.00; 315.1; 315.2; 315.9; 315.4; 315.31; 315.32; 315.39; 307.9; 299.00; 299.80; 299.10; 299.80; 299.80.

Even though one might expect some association between MID/BIF and developmental disorders, this was only partly the case, such as for Autism and Asperger syndrome. On the whole, the association between developmental disorder and MID/ BIF was not remarkable. To investigate this, we calculated a % agreement and a Kappa. In BIF the % agreement was 71%, and in MID it was 70%. The Kappa’s were below 0.10, the correlations as well. We investigated collinearity while performing the regression analyses, which showed no issues. Consequently we feel the concern of the reviewer is addressed in these extra analyses. As it does not concern the focus of the study, we did not present these findings in the publication.

Question # 32 Some patients are staying in long-stay wards – do social factors play a part that may contribute to the aggressive behaviours instead of cognitive impairment.

Response # 32 We added a line (373) in recommendations for future research because we do not know yet if or how this plays a role: We also do not know the influence of social factors between patients within ward settings, as the current study did not look into that level of detail.

Question # 33 How does screening patients for ID support the prevention of aggressive incidents as mentioned in clinical implications?

Response # 33 We added lines 362-365: The SCIL can be helpful in psychiatric care [29,45] and give a quick first impression. Treatment and support to the needs can then be adapted according to the SCIL category. As such, avoid over-demanding and stress.

Question # 34 Do the staff’s attitudes towards people with ID play a part to detect and prevent aggressive incidents?

Response # 34 Yes indeed, this was addressed in line 364:

The staff’s attitudes towards people with ID may also play a part to detect and prevent aggressive incidents.

Question # 35 As the staff interact with the patients for a long period of time, it is possible that their relationship with patients may influence their ratings on the scales, would it become a discriminatory tool to segregate people with suspected ID and those without?

Response # 35 We added line 186: According to the questionnaire instructions, the SCIL was administered by a person not involved in the treatment. So, this concern was not the case.

Submitted filename: 2022-04-20 PLOS ID aggression rebuttal .docx

Click here for additional data file.

7 Jun 2022

25 Jun 2022

PONE-D-21-40785R1

Aggressive behaviour of psychiatric patients with Mild and Borderline Intellectual Disabilities in general Mental Health Care

PLOS ONE

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the "Comments to the Author" section, enter your conflict of interest statement in the "Confidential to Editor" section, and submit your "Accept" recommendation.

Reviewer #1: (No Response)

________________________________________

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

________________________________________

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

________________________________________

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

________________________________________

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: No

________________________________________

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Thank you for the opportunity to re-review this manuscript. The authors have addressed most of my concerns. Remaining issues are listed below – I have also commented on some new sections of text.

Response #1We thank the reviewer for this comment. Also for the new comments. We think the article improved again a lot.

#1. Again I would encourage use of person-first language, e.g.

a) changing "MID/BIF patients" (pg 2, line 67) to "people with MID/BIF"; "

b) SMI patients" (page 5, line 137, line 142, pg 6 line 174) to "people with SMI";

c) "the homeless" (page 6, line 168) to "people experiencing homelessness";

d) "MID or BIF cases" (page 8, line 253) to "people with MID or BIF";

e) "MID patients" (page 8, line 257) to "people with MID"

#1. response: We have adopted the recommendations. The only difference is that this study was conducted in the mental health care, where we speak of patients. Therefore, we have used patients instead of people throughout the draft.

2. Page 2, line 87, suggest changing "the more severely intellectually impaired patients" to "people with more severe intellectual disability"

Response 2: we have changed this in ‘patients with more severe intellectual disability’.

3. Page 5, line 135, "Serious Mentally Ill patients' should read "people with serious mental illness (SMI)".

Response 3: Again we changed this, but in patients with serious mental illness

4. Again, terms such as mild intellectual disability (pg 2, line 63), borderline intellectual functioning (pg 2, line 63), mental health care (pg 2, line 64; page 4, line 120), challenging behaviour (pg 2, line 73) do not need to have the first letter of each word capitalised.

Response 4: Thank you for being so precise. We changed this all.

5. Page 11, line 336: intellectual disability should be abbreviated to ID in the following sentence- "As we can learn from ID studies based on interviews with people with intellectual disability [10, 37], people with ID often experience.."

Response 5: Again thank you for being so precise. We changed this.

6. Page 9, line 271: should "Patients with bipolar disorder… were associated with an increased risk of" read "Diagnoses of bipolar disorder… were associated with an increased risk of"?

Response 6: We changed this.

7. Page 9, line 276-77: check for typographical errors – e.g. "associated to" should read "associated with"

Response 7: Most were "associated with". With search and change we detected only one "associated to "in the whole article, and corrected this.

8. Page 9, line 280-285: inconsistent use of "patients who screened positive for BIF or MID were associated with . . . and "MID was associated with". It would be helpful for the reader to be consistent. I would also recommend this section be reviewed by a statistician or biostatistician to ensure the results are appropriately differentiated in text from those of the Poisson regression. E.g. if logistic regression would be determining risk for having at least one aggressive incident, while the Poisson would look at the number of incidents.

Response 8:

An experienced statistician was one of the authors (Noorthoorn) and carefully reviewed this section. We substantially reformulated the complete section, using for the logistic regression ‘being at risk’ in stead of the more vague ‘associated with’. With respect to the poisson regression we presented the beta (β).

Logistic regression

The logistic regression analysis showed that patients who screened positive for BIF (OR=2.00, p=0.003) or MID (OR 2.89, p<0.001) were more at risk to show aggressive incidents, as well as the patients with the diagnoses bipolar disorder (OR 3.07, p<0.001), schizophrenia (OR 2.75, p<0.001), and a low GAF (OR 1.72, p=0.005). Logistic regression analysis with physical aggression as outcome showed that patients with MID (OR 2.50, p<0.001), a bipolar disorder (OR 3.13, p=0.007) or schizophrenia (OR 4.04, p<0.001) were more at risk to show aggressive incidents.

Poisson regression

These findings were underlined by the poisson regression of the number of physical aggression incidents per patient. This showed anxiety disorder (β=0.62, p<0.001), bipolar disorder (β=1.63, p<0.001), schizophrenia (β=1.12, p<0.001), developmental disorder (β)=0.69, p<0.001), drug abuse disorder (β=1.18, p<0.001) and a SCIL below 15 (β=0.61, p<0.001) were all related to more incidents.

9. Page 11, line 339: if using the term challenging behaviour, then this should be used here as well – behaviours that challenge is an alternative expression preferred by some.

Response 9: We changed this to challenging behaviour.

10. Page 11, lines 356-360- The authors may wish to include some citations for statements included here.

Response 11:

Thank you for offering some references. We included a reference to The NICE guidelines and to the first two references (Whittle et al and Mesa et al. We did not citate because, in our opinion, this is beyond the scope of this article.

E.g. for recommendations regarding the prevention and management of behaviours that challenge: The NICE guidelines here https://www.nice.org.uk/guidance/ng11/resources/challenging-behaviour-and-learning-disabilities-prevention-and-interventions-for-people-with-learning-disabilities-whose-behaviour-challenges-1837266392005.

Regarding mental health services frequently being ill-equipped to meet the needs of people with intellectual disability:

Whittle, E.L.; Fisher, K.R.; Reppermund, S.; Lenroot, R.; Trollor, J. Barriers and Enablers to Accessing Mental Health Services for People With Intellectual Disability: A Scoping Review. J. Ment. Health Res. Intellect. Disabil. 2018, 11, 69–102.

Mesa, S.; Tsakanikos, E. Attitudes and self-efficacy towards adults with mild intellectual disability among staff in acute psychiatric wards: An empirical investigation. Adv. Ment. Health Intellect. Disabil. 2014, 8, 79–90.

We did not include the next publications suggested:

Evans, E.; Howlett, S.; Kremser, T.; Simpson, J.; Kayess, R.; Trollor, J. Service development for intellectual disability mental health: A human rights approach. J. Intellect. Disabil. Res. 2012, 56, 1098–1109.

Weiss, J.A.; Lunsky, Y.; Gracey, C.; Canrinus, M.; Morris, S. Emergency Psychiatric Services for Individuals with Intellectual Disabilities: Caregivers' Perspectives. J. Appl. Res. Intellect. Disabil. 2009, 22, 354–362.

Donner, B.; Mutter, R.; Scior, K. Mainstream In-Patient Mental Health Care for People with Intellectual Disabilities: Service User, Carer and Provider Experiences. J. Appl. Res. Intellect. Disabil. 2010, 23, 214–225.

11. Page 11, line 363, the I in ID has been deleted in track changes but should be retained

Response 11: Thank you for your attention. We changed this.

12. Page 11, like 367- can the authors include a reference to support the link between staff attitudes toward people with intellectual disability and the prevention and detection of aggressive incidents.

Response 12: We included the next references to support this link:

van den Bogaard KJHM, Nijman HLI, Embregts PJCM. Attributional styles of support staff working with people with intellectual disabilities exhibiting challenging behaviour. J Appl Res Intellect Disabil. 2020 May;33(3):465-474.

Knotter MH, Wissink IB, Moonen XM, Stams GJ, Jansen GJ. Staff's attitudes and reactions towards aggressive behaviour of clients with intellectual disabilities: a multi-level study. Res Dev Disabil. 2013 May;34(5):1397-407.

13. Page 12, line 370-372, it could be made clearer here that the authors are recommending further work specifically focusing on aggressive behaviour among people with intellectual disability within the mental health service setting, including how this compares to the broader population of people with intellectual disability (if that is what they are recommending).

Response 13:

Thank you for your comment. The lines 370-372 concern not further work (etc.) but concerns about giving staff and nurses a better opportunity for personal tailormade treatment and attitude toward patients with ID. We explained in lines 376-381 more precisely what kind of future research is recommended.

We changed the sentence and added an extra sentence ( and reference) to be more clear;

" In mental health care, future studies should examine the causes and reasons for aggression incidents in patients with MID/BIF and learn more about how these patients differ or resemble those of people in ID studies. For example, we do not yet know much about the mental problems in general in the patient group with ID in mental health care and how this affects behaviour, including how this compares to the broader population of people with intellectual disability and mental health problems [..].

14. Page 12- line 372- in the sentence starting "for example, we do not yet know much about…" it is not clear whether the authors are referring to mental health diagnosis profiles of their cohort specifically or mental health service users with intellectual disability more generally. This should be made more clear. Also, MHC should not be abbreviated here as in previous mentions it is expanded as mental health care.

Response 14:

As explained in our response 13, we tried to be more clear.

MCH has been changed in mental health care.

15. Page 12, line 377, suggest including references to supporting statements about the usefulness of positive behaviour

support.

Response 15: we added the next reference: Lewis N, Reynolds E, Vale M, Keenan E, Hartland A, Haines M, Davies BE. An evaluation of positive behavioural support implemented within an intensive community support service for people with learning disabilities who present with behaviours that challenge. J Intellect Disabil. 2021 Sep;25(3):348-356.

________________________________________

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose "no", your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Submitted filename: 15-06-2022 PLOS ONE ID agression rebuttal 2_.docx

Click here for additional data file.

4 Jul 2022

20 Jul 2022

Reviewer #1: Thank you for sending this manuscript for re-review. The edits related to the statistical analyses were very helpful. I would just suggest the statistician on the author team also review the abstract to ensure the results are represented appropriately there. I noted that the track changes in the document denote changes made in the first round of revision while the most recent edits are not tracked. I did not review the manuscript thoroughly as I think there may be some issues with the file:

#1a: We added three brief sections to the abstract highlighting the main findings of the analyses which were added in the process of the rebuttal. On page 2 line 46 we added: and poisson. On page 2 line 52 we added the confidence interval of the OR for BIF: OR 4.20, 95% CI 2.45-7.22. And on page 2 line 55-56 we added: Poisson regression underlined these findings, showing a SCIL below 15 (β=0.61, p<0.001) was related to more incidents. We chose not to discuss all findings of the poisson regression in the same detail as in the results section, as the poisson regression was primarily used as a confirmation of the findings of the logistic regression.

1. Typographical errors that the authors said have been fixed have not been; e.g. serious mentally ill on pg 15 should read patients with serious mental illness; the I in ID is still deleted on page 11, line 363; MHC is still abbreviated on page 12, line 373. I did not check the rest and recommend the authors ensure their response matches the manuscript document.

#1b. Thank you very much for your comments on typographical errors. Indeed there have been some issues with the file but we also find again some errors scattered throughout the text. These have been corrected.

2. Response 13 (re lines 30-381) does not appear to be incorporated into the manuscript. Nonetheless I still suggest the authors reword this as it is still not clear to me what they are recommending, and suggest the authors ensure use of people first language (i.e. studies with people with ID, not people in ID studies), and also to be consistent with abbreviating intellectual disability to ID.

#2 Thank you for persisting in using first language. We have again changed a number of sentences

Also, some minor typographical issues in the abstract:

3. Some words are incorrectly capitalized in the abstract; mild intellectual disability, borderline intellectual functioning, mental health care, behaviour, odds (or if this is to be capitalized then ratio should be capitalized as well), bipolar disorder. Also, “was higher in assumed MID” should read “in patients with assumed MID”

#3 Thank you for your comment. We have aligned the use of capital letters throughout the article as much as possible

________________________________________

Submitted filename: 15-06-2022 PLOS ONE ID agression rebuttal 2_.docx

Click here for additional data file.

21 Jul 2022

Aggressive behaviour of psychiatric patients with Mild and Borderline Intellectual Disabilities in general Mental Health Care

PONE-D-21-40785R3

Dear Dr. Nieuwenhuis,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Robert Didden

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

19 Sep 2022

PONE-D-21-40785R3

Aggressive behaviour of psychiatric patients with Mild and Borderline Intellectual Disabilities in general Mental Health Care

Dear Dr. Nieuwenhuis:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Professor Robert Didden

Academic Editor

PLOS ONE