Researching outcomes from forensic services for people with intellectual or developmental disabilities: a systematic review, evidence synthesis and expert and patient/carer consultation

Search strategy

The search strategy aimed to identify studies from a range of sources. Reports were included regardless of publication status. Searches were conducted on 1 June 2015. The electronic databases searched were MEDLINE, PsycINFO, EMBASE, Allied and Complementary Medicine, Health Management Information Consortium, the British Nursing Index and the Cumulative Index to Nursing and Allied Health Literature from inception to May 2015. A broad search strategy was employed, with intellectual disability terms being combined with forensic terms. For terms specific to intellectual and developmental disabilities, search terms employed were based on those used for a previous Cochrane review. 6 Forensic and offender terms search terms were adapted from a large review for the Department of Health by Duggan et al. 7 The full search terms including ‘explode’ terms, keywords and text words are in Appendix 1.

In addition, the grey literature (opengrey.eu) was searched online on the same date using the same keywords. References of suitable studies were examined to identify additional relevant articles. Finally, expert members of the project steering group were consulted in order to identify any key references not retrieved by the search strategy or in press or unpublished articles and service evaluation reports.

Inclusion and exclusion criteria

Study selection

After removal of duplicates, titles and abstracts of articles were screened against the inclusion criteria. In order to avoid false negatives, all studies that could not be shown to be outside the inclusion criteria were initially included. The full texts of the remaining articles were screened against inclusion and exclusion criteria in order to identify appropriate articles for the review. Two reviewers (CM and NG) independently rated each article using a screening tool developed for the purposes of the study (see Appendix 3). Any disagreements were resolved by a third reviewer (MF).

Data extraction

The articles were subjected to data extraction using a structured form (see Appendix 3). Details regarding the sample, design, service type, methods, outcome domains and related measures were obtained, coded and reported. The information is presented in the form of data summary tables (see Tables 1–6).

Hierarchy of evidence

On the advice of the systematic review consultant on the project, it was decided a priori that no structured quality assessment was to be undertaken. This decision was taken for a number of reasons. First, as noted above, the aim of the current review was to identify the range and type of outcome domains and related measures that have been used in FIDD services research. Its purpose was not to critique the methodology of the studies or to draw robust conclusions about the findings. Furthermore, it was recognised the studies were likely to be heterogeneous and there would be no single quality assessment tool that would be appropriate for included studies. Quality or risk-of-bias ratings would not have influenced the inclusion of studies as, arguably, a poor-quality study could nevertheless have presented outcome domains of interest for the review. We therefore decided simply to code study design using a hierarchy of evidence. The York Centre for Reviews and Dissemination hierarchy of evidence was considered, but it was found that included studies were difficult to classify using that hierarchy. Instead, study designs were coded with reference to the simpler hierarchy of evidence employed by Greenhalgh9 as follows:

1. systematic reviews and meta-analyses

2. randomised controlled trials with definitive results

3. randomised controlled trials with non-definitive results

4. cohort studies (follow participants over time)

5. case–control studies (matched control group)

6. cross-sectional surveys

7. case reports or case series.

Analysis

The broad inclusion criteria applied for the review meant that the included studies represented a wide variety of services, study designs, outcome domains and measures. A method of content analysis10 was used to synthesise the outcome domain types and related measures and indicators. This approach involved an iterative process of identifying categories to capture the outcome measures reported in the studies. Using the aforementioned three core outcome domains (effectiveness, patient safety and patient experience) as an a priori superordinate framework, the outcome measures from all included studies were listed. A process of refining and grouping similar subdomains together was then undertaken. This led to the construction of a ‘framework’ to describe the outcome domains that have been studied in the literature. The number of studies falling into each final outcome domain was also quantified.

Domain 1: effectiveness

A total of 53 studies (29 of which were cohort studies) presented data on at least one outcome that measured the effectiveness of FIDD services. Twelve outcome subdomains were identified as a result of content analysis.

Domain 2: patient safety

A total of 11 studies (see Tables 3 and 4) reported on some aspect of patient safety, which were classified into five subdomains. Two of these subdomains related to interventions that could be perceived as ‘restrictive practices’71: the first included physical restraint or periods of increased intensive nursing observations, and the second included seclusion/segregation in response to risky or ‘challenging’ behaviour. These were reported either in hours or in terms of the number of episodes. In parallel with recording numbers of violent incidents, mean rates per patient/month or year of treatment were typically reported, in order to control for length of stay. These outcomes could arguably also be construed as an effectiveness indicator if change in rates over time in treatment is measured. Only two studies37,38 reported on such change in seclusion rates in a cohort with follow-up. Only one study reported on the reason for seclusion, and on distress-related behaviours after seclusion, that is, the outcome measured had a specific emphasis on patient safety. 63

Premature death was reported in only one study, which differentiated between natural causes and suicide. 19 Physical health outcomes or access to services was reported by one study only. 56 This is surprising given the higher incidence of premature death and physical health-care problems in forensic psychiatric populations. 72 Use of pro re nata (PRN) medication, which could also be perceived as a restrictive practice, was monitored as an outcome in three studies. The number of episodes of PRN use was recorded in two secure service studies39,56 and via a case study demonstrating a reduction in use in a community patient. 16

Domain 3: patient experience

A total of 11 of the included studies reported on four patient experience-related subdomains (see Tables 5 and 6). Four studies addressed quality of life using a number of rating scales. These included the 10-item Personal Wellbeing Index,73 the only measure specifically designed for people with intellectual disabilities (S Trout, personal communication). Other measures employed were the Quality of Life Questionnaire,74 used among both a medium secure (E Marks, personal communication) and female community sample,64 and the Life Experience Checklist67 in a community sample. When change over time in reported quality of life was examined, this was in relation to a change of environment, that is, a service development. There were no community studies of quality-of-life maintenance over time in discharged FIDD patients.

Three studies addressed satisfaction with services. These related to a specific service development or move, in a community setting,64 a low secure unit65 and a high secure unit (S Trout, personal communication). A range of satisfaction instruments that were service specific and tailored for use with people with IDD were used. Only one study evaluated the number of patient complaints as a service outcome measure. 56

A fourth subdomain of patient experience identified from the literature was the therapeutic environment, reported in three studies using either the Essen Climate Evaluation Scale75 (EssenCES), which was rated by both patients and staff, or the Correctional Institutions Environment Scale. 76 There was no indication that these measures had been specifically adapted for IDD patients in any of the studies. These were typically cross-sectional studies66,68 or evaluated a specific change in environment after 3 months. 65 No identified studies reported long-term change over time in milieu quality, although it is feasible that such measures could be used as a repeated cross-sectional service outcome indicator. Only one community study reported on the level of support that patients received. 16

Finally, patient involvement as an outcome indicator was reported in only one study: Mansell et al. ,56 in their large study, reported the number of patients with an up-to-date and accessible copy of their care plan. This is one of the outcomes addressed by the current Transforming Care programme care and treatment reviews. 3

Effectiveness domain

Patient safety domain

Patient (and carer) experience domain

Participants

Members of the Clinical Research Group in FIDD services were invited to participate in the Delphi exercise. In addition, senior clinicians and commissioners in the field (identified by the Project Steering Group) were contacted. An intended sample size of 15–20 experts was sought in line with Delphi recommendations. 79 Participants were eligible to participate if they had clinical, academic, managerial or commissioning experience of working in FIDD services. No specific exclusion criteria were applied given the purposive sampling method. Individuals were asked to highlight their professional background, current area of work (i.e. clinical/academic/commissioning/managerial), type of FIDD service(s) by level of security in which they were currently working and number of years of experience working within such services.

Procedure

Participants were contacted by e-mail to explain the study and to seek participation. For round 1, 76 individuals were invited to take part. The invitation e-mail provided the rationale and intended aims of the Delphi exercise. The web link, closing date and a contact for further questions were also provided. On opening the link, participants were provided with the information sheet (see Appendix 6) and instructions for completing the survey.

If respondents consented to continue, they were asked to provide brief professional background information as described above. Next, participants rated how important each of the 31 outcome subdomains was for the evaluation of (FIDD) services, within the three superordinate domains of effectiveness (15 subdomains presented), patient safety (six subdomains) and patient experience (10 subdomains). A Likert scale anchored 1 (not at all important) to 5 (extremely important) was used to rate each item, with each score having operationalised definitions. For each subdomain participants were able to supply their preferred measures or indicators in a free-text box. Participants were then asked to specify any additional outcome subdomains or indicators which they considered as important from their expert knowledge. Following this, participants were asked to select up to five subdomains which they considered to be the most important/highest priority. This included subdomains already identified as well as any additional items. Finally, in order to fulfil the study’s second aim, all participants indicated the outcome measures used routinely to collect data on all patients in their own service(s). Respondents had 2 weeks to complete the first round. A reminder e-mail was sent 5 days before the closing date to those who had not responded.

Those who completed the first round were invited to take part in the second round (see Appendix 7). The e-mail was sent on an individual basis with the web link, closing date and the participant’s individual round 1 ratings in a portable document format (PDF) attachment. On continuing to the survey web page, further instructions were provided. It was outlined that the second round was an opportunity to reappraise ratings in the light of the average group response, which was provided adjacent to each subdomain. First, participants were asked to rate the importance of each subdomain using the same Likert scale. Second, participants were asked to select up to five subdomains that they perceived to be the most important/highest priority. Participants were reminded that they did not have to change their answer from their original response if they did not wish to do so. Round 2 was available to complete over a 2-week period, with a reminder e-mail sent at the halfway point.

Analysis

Descriptive methods were used to report findings. Frequency data were generated to describe the professional backgrounds of participants. Percentages and average figures were calculated to describe the ratings made in rounds 1 and 2. The consensus level for items rated using the Likert scale was set at an average rating of 4.0 or above (i.e. items rate as either ‘moderately’ or ‘extremely’ important). The percentage of participants who ranked each subdomain in their top five was calculated. Free-text answers were analysed using content analysis.

Characteristics of participants

A total of 17 respondents completed round 1 (22% response rate). The majority of participants were psychologists and psychiatrists who worked in a clinical role at the time of participation (Figures 2 and 3). Respondents had on average 12.5 years’ experience of working in FIDD services (range 1–25 years). Fifteen of these respondents participated in the second round (88% response rate).

FIGURE 2.

Area of work.

FIGURE 3.

Type of role.

Although all security levels of FIDD services were represented by respondents, the majority were working in medium and low secure care as shown in Figure 4. Seven respondents were working across either two or three different levels of security. Both the NHS and independent sectors were represented.

FIGURE 4.

Type of service currently working in. ASD, autism spectrum disorder; ID, intellectual disability; LD, learning disability.

Ratings

In round 1 respondents were asked to rate how important they thought each outcome subdomain was for the evaluation of FIDD services. The subdomains with an average group score from round 1 of 4.0 and above (the consensus threshold level) were presented in round 2. Participants rerated each item. The mean scores were then calculated for each subdomain and are presented in Table 8.

None of the items was rated as ‘not at all important’ or ‘slightly important.’ There were five items which did not reach the consensus level at round 1 (length of stay, security needs, adaptive functioning, clinician-rated quality of life and closeness to home area). A decision was made to retain one item (length of stay) as, paradoxically, this item was highly ranked in the ‘top five’ domains section. At round 2, two items (length of stay and physical health) did not reach consensus level.

The subdomains that gained the highest ratings at the end of round 2 were discharge outcome, treatment response/engagement, reoffending (effectiveness domain), premature death (patient safety domain), therapeutic milieu and access to meaningful activity (patient experience domain).

Top five outcomes

Participants selected up to five subdomains that they believed were the most important/highest priority for the evaluation of FIDD services. The percentages of respondents who endorsed each subdomain as one of their most important outcomes were calculated for both rounds and are presented in Table 9.

Suggested measures/indicators

Professionals were asked to suggest an ideal or preferred measure/indicator for each subdomain using a free-text response.

What are the most important/highest priority outcomes?

There were six subdomains which were clearly rated as greatest importance by respondents (shaded in Table 9). Clinical symptom severity, treatment response and reoffending were most frequently endorsed as being within the top five most important subdomains. In addition, recovery/progress on treatment goals and risk assessment measures and discharge outcome were also most commonly ranked as being the most important outcome domains. This broadly concurred with the Likert scale ratings. By the end of the second round, respondents favoured items under the superordinate domain of effectiveness only, with only a few respondents including items from patient safety or patient experience in their top five.

Although in round 1 there appeared to be little difference between the ratings of most domains, if the Likert rating and the top five are combined, the second round identified clear consensus for the nine most highly rated outcome subdomains, as shown in Box 2.

It is worth noting that in round 1, reoffending (formal reconvictions) and offender-like behaviour (activity that could be construed as an offence but for which no formal convictions or criminal justice system involvement followed) were similarly rated. One participant emphasised that the latter was ‘as important as formal reconvictions’ to measure. Another respondent highlighted ‘offender-like’ behaviour was of particular importance ‘to capture in an LD [learning disability] population’. In round 2, these two outcomes were therefore presented as one subdomain to reflect this.

Additional subdomains

Only one participant suggested an additional subdomain, degree of patient insight, which did not appear in the original list of items presented in round 1. This was combined with the ‘engagement’ item at round 2, as it was considered to be similar in theme.

Current use of outcome measures

In round 1, participants were asked to indicate any outcome measures used to collect routine data on all patients in the service(s) in which they worked. Table 10 provides a summary of responses. Services routinely measured levels of risk using structured clinical judgement methods (most commonly with the HCR-20). Change on risk measures (such as the HCR 20 and Short-Term Assessment of Risk and Treatability) was clearly considered important, despite the fact that the literature review indicated that there are very few studies that examine change over time in risk measures as a result of treatment. Those that do exist suggest that the measures are quite insensitive to change. It should also be noted that the authors of the HCR-20 caution against ‘totalling up’ of scores, which is the only way of measuring change in rated risk, unless only the summary risk rating (high, medium, low) is used. Totalling of scores is also complicated by the new ‘relevance’ ratings in the HCR-20 latest Version 3, which is now in use. The rated importance of risk measures as an outcome variable by professionals fails to acknowledge such problems and may be driven more by their availability owing to mandated use as part of Commissioning for Quality and Innovation targets.

There was not a clearly preferred measure of clinical change or treatment progress. Clinical symptoms were most commonly rated using HoNOS-secure or other versions of HoNOS. Patients’ views on their own progress were not routinely collected, despite their rated importance, although a small number of participants mentioned trust-led patient satisfaction surveys or patient self-risk ratings.

Incident data and associated restrictive practices were routinely monitored, but there was no evidence of a standardised coding system across services, which would be necessary for benchmarking purposes. This may reflect the multiplicity of service providers in this sector.

Additional issues raised

Respondents were given the opportunity to provide additional free-text responses in both rounds. The key themes are as follows:

• Barriers to measuring outcomes: participants reflected on the difficulties of measuring outcomes in FIDD services in general. One participant summarised the issue as ‘a lack of clear objective measures and outcomes for forensic intellectual disabilities services’ suggesting an absence of current guidance and paucity of appropriate outcome measures for the population. Another respondent indicated that ‘there is a huge reliance on clinical judgement and this can be flawed and biased’. As the literature review found, it is notable that the majority of collected data reflect professional opinion and there is a relative dearth of outcomes from the patient perspective.

• Additional subdomains: participants were given the opportunity to suggest any subdomains which were not presented in round 1. The limited number of extra subdomains suggested that the original list of subdomains was comprehensive. It therefore appeared that the relevant and important items were captured by the framework developed from stages 1 and 2.

• Discord around direction of ‘positive’ outcomes: with regard to length of stay, the comments of some participants highlighted ambivalence as to the desired direction of this outcome indicator. For example, although having a shorter length of stay may initially appear to be a positive outcome, it was noted that it is ‘important to pick up inappropriate early discharges due to bed pressures’ among inpatient services. It was also noted that, for some forensic patients, an extended length of stay may be clinically appropriate and, therefore, not a ‘negative’ outcome, a theme echoed in the carer interviews. This ambivalence about length of stay as an outcome indicator may be reflected in its relatively low importance rating (indeed, it fell below the consensus threshold in both rounds).

• Ambivalence regarding closeness to home area: although it is emphasised as being desirable in a range of policy papers, interestingly, closeness to home area was not considered a priority by professionals. One respondent commented that ‘our regional and supra regional Gatekeeper role sometimes experiences the benefit of some “geographical distance”’. This implies that for some patients living near to their family or local authority area may not necessarily be a constructive outcome.

• Patient-rated outcomes: several respondents recommended using patients’ self-reported data to capture outcomes including patients’ assessment of how safe they perceive they are – often relates to ability to engage in therapeutic change and achieve treatment goals/reduce risk. Professionals also outlined ways that this could be achieved, for example ‘co-production of such [patient satisfaction] measures from within our patient advisors’. It is notable that ‘My Shared Pathway’80 was intended by the Department of Health to be an outcomes monitoring system; it was mandated for use as a CQUINN (Commissioning for Quality and Innovation) target for secure services in 2013 and includes patient ratings of progress on treatment goals. The almost total lack of mention of this tool by clinicians (one respondent) suggests that it is not being used consistently as an outcome measure or indicator by IDD services.

Finally, it is noteworthy that all subdomains were generally rated towards the upper end of the scale, indicating that respondents believed that all items had a degree of importance. Nevertheless, clear consensus around key priorities emerged.

Search strategy

1. exp intellectual development disorder/ (39,999)

2. Learning disorders/ (2170)

3. Developmental Disabilities/ (10,651)

4. ((intellectual$ or learning$ or development$) adj3 (impair$ or disab$ or disorder$ or difficult$ or subnormal$ or sub-normal$)).tw. (66,491)

5. ((mental$ or development$) adj3 (retard$ or disab$ or defici$ or handicap$ or impairment$ or subnormal$ or sub-normal$)).tw. (58,870)

6. (feeble-minded$ or moron$ or imbecil$ or cretin$).tw. (2851)

7. de lange$.tw. (112)

8. Cri-du-chat.tw. (87)

9. (Wagr adj3 syndrome).tw. (2)

10. (down$ adj3 syndrome$).tw. (6387)

11. (Rubinstein adj3 Taybi).tw. (39)

12. (prader adj3 willi).tw. (637)

13. (williams adj3 syndrome).tw. (1038)

14. fragile X.tw. (1897)

15. ((Martin Bell or FRAXE or FRAXA) adj3 syndrome$).tw. (19)

16. (les?h adj3 nyhan$).tw. (184)

17. Adrenoleukodystrophy.tw. (151)

18. (Coffin adj Lowry).tw. (32)

19. Menkes.tw. (60)

20. Glycogen Storage Disease.tw. (43)

21. Mucopolysaccharidosis.tw. (90)

22. Rett$.tw. (1073)

23. Pyruvate Dehydrogenase.tw. (80)

24. or/1-23 (118,478)

25. forensic.tw. (14,524)

26. (secure adj (unit$ or hospital$ or service$)).tw. (676)

27. Criminals/ (11,153)

28. offen$.tw. (36,629)

29. or/25-28 (51,835)

30. 24 and 29 (2011)

Search strategy

1. exp Intellectual Disability/ (82,875)

2. Mentally Disabled Persons/ (2227)

3. Learning disorders/ (12,856)

4. Developmental Disabilities/ (15,682)

5. ((intellectual$ or learning$ or development$) adj3 (impair$ or disab$ or disorder$ or difficult$ or subnormal$ or sub-normal$)).tw. (48,293)

6. ((mental$ or development$) adj3 (retard$ or disab$ or defici$ or handicap$ or impairment$ or subnormal$ or sub-normal$)).tw. (51,852)

7. (feeble-minded$ or moron$ or imbecil$ or cretin$).tw. (1645)

8. de lange$.tw. (793)

9. Cri-du-chat.tw. (368)

10. (Wagr adj3 syndrome).tw. (133)

11. (down$ adj3 syndrome$).tw. (17,498)

12. (Rubinstein adj3 Taybi).tw. (519)

13. (prader adj3 willi).tw. (2673)

14. (williams adj3 syndrome).tw. (1673)

15. fragile X.tw. (4927)

16. ((Martin Bell or FRAXE or FRAXA) adj3 syndrome$).tw. (202)

17. (les?h adj3 nyhan$).tw. (969)

18. Adrenoleukodystrophy.tw. (1653)

19. (Coffin adj Lowry).tw. (187)

20. Menkes.tw. (1086)

21. Glycogen Storage Disease.tw. (2005)

22. Mucopolysaccharidosis.tw. (2613)

23. Rett$.tw. (3054)

24. Pyruvate Dehydrogenase.tw. (5100)

25. or/1-24 (17,7158)

26. forensic.tw. (25,673)

27. (secure adj (unit$ or hospital$ or service$)).tw. (315)

28. Criminals/ (1347)

29. offen$.tw. (15,415)

30. or/26-29 (40,461)

31. 25 and 30 (718)

Search strategy

1. exp mental deficiency/ (124,913)

2. intellectual impairment/ (12,230)

3. developmental disorder/ (27,419)

4. Learning disorders/ (21,725)

5. Developmental Disabilities/ (24,896)

6. ((intellectual$ or learning$ or development$) adj3 (impair$ or disab$ or disorder$ or difficult$ or subnormal$ or sub-normal$)).tw. (69,880)

7. ((mental$ or development$) adj3 (retard$ or disab$ or defici$ or handicap$ or impairment$ or subnormal$ or sub-normal$)).tw. (69,505)

8. (feeble-minded$ or moron$ or imbecil$ or cretin$).tw. (2271)

9. de lange$.tw. (1017)

10. Cri-du-chat.tw. (483)

11. (Wagr adj3 syndrome).tw. (157)

12. (down$ adj3 syndrome$).tw. (22,586)

13. (Rubinstein adj3 Taybi).tw. (652)

14. (prader adj3 willi).tw. (3535)

15. (williams adj3 syndrome).tw. (2497)

16. fragile X.tw. (6134)

17. ((Martin Bell or FRAXE or FRAXA) adj3 syndrome$).tw. (226)

18. (les?h adj3 nyhan$).tw. (1152)

19. Adrenoleukodystrophy.tw. (2064)

20. (Coffin adj Lowry).tw. (219)

21. Menkes.tw. (1331)

22. Glycogen Storage Disease.tw. (2559)

23. Mucopolysaccharidosis.tw. (3866)

24. Rett$.tw. (4102)

25. Pyruvate Dehydrogenase.tw. (6090)

26. or/1-25 (257,407)

27. forensic.tw. (40,822)

28. (secure adj (unit$ or hospital$ or service$)).tw. (577)

29. exp offender/ (8941)

30. offen$.tw. (23,256)

31. or/27-30 (65,604)

32. 26 and 31 (1413)

Search strategy

1. exp learning disability/ (3019)

2. exp mental retardation/ (2432)

3. exp Developmental disabilities/ (849)

4. ((intellectual$ or learning$ or development$) adj3 (impair$ or disab$ or disorder$ or difficult$ or subnormal$ or sub-normal$)).tw. (7938)

5. ((mental$ or development$) adj3 (retard$ or disab$ or defici$ or handicap$ or impairment$ or subnormal$ or sub-normal$)).tw. (4951)

6. (feeble-minded$ or moron$ or imbecil$ or cretin$).tw. (9)

7. de lange$.tw. (15)

8. Cri-du-chat.tw. (8)

9. (Wagr adj3 syndrome).tw. (0)

10. (down$ adj3 syndrome$).tw. (825)

11. (Rubinstein adj3 Taybi).tw. (5)

12. (prader adj3 willi).tw. (78)

13. (williams adj3 syndrome).tw. (74)

14. fragile X.tw. (81)

15. ((Martin Bell or FRAXE or FRAXA) adj3 syndrome$).tw. (1)

16. (les?h adj3 nyhan$).tw. (4)

17. Adrenoleukodystrophy.tw. (4)

18. (Coffin adj Lowry).tw. (0)

19. Menkes.tw. (0)

20. Glycogen Storage Disease.tw. (8)

21. Mucopolysaccharidosis.tw. (4)

22. Rett$.tw. (71)

23. Pyruvate Dehydrogenase.tw. (7)

24. or/1-23 (11,351)

25. forensic.tw. (226)

26. (secure adj (unit$ or hospital$ or service$)).tw. (22)

27. exp crime/ (312)

28. offen$.tw. (339)

29. or/25-28 (698)

30. 24 and 29 (193)

Search strategy

1. exp learning disabilities/ (6239)

2. ((intellectual$ or learning$ or development$) adj3 (impair$ or disab$ or disorder$ or difficult$ or subnormal$ or sub-normal$)).tw. (4440)

3. ((mental$ or development$) adj3 (retard$ or disab$ or defici$ or handicap$ or impairment$ or subnormal$ or sub-normal$)).tw. (3030)

4. (feeble-minded$ or moron$ or imbecil$ or cretin$).tw. (7)

5. de lange$.tw. (0)

6. Cri-du-chat.tw. (0)

7. (Wagr adj3 syndrome).tw. (0)

8. (down$ adj3 syndrome$).tw. (254)

9. (Rubinstein adj3 Taybi).tw. (0)

10. (prader adj3 willi).tw. (3)

11. (williams adj3 syndrome).tw. (0)

12. fragile X.tw. (12)

13. ((Martin Bell or FRAXE or FRAXA) adj3 syndrome$).tw. (0)

14. (les?h adj3 nyhan$).tw. (0)

15. Adrenoleukodystrophy.tw. (0)

16. (Coffin adj Lowry).tw. (2)

17. Menkes.tw. (0)

18. Glycogen Storage Disease.tw. (0)

19. Mucopolysaccharidosis.tw. (3)

20. Rett$.tw. (6)

21. Pyruvate Dehydrogenase.tw. (0)

22. or/1-21 (8443)

23. forensic.tw. (427)

24. exp secure mental health services/ (195)

25. exp psychiatric secure units/ (133)

26. (secure adj (unit$ or hospital$ or service$)).tw. (553)

27. offenders/ (1921)

28. offen$.tw. (1971)

29. or/23-28 (3691)

30. 22 and 29 (189)

Search strategy

1. BNI; (exp AND MENTAL AND RETARDATION).ti,ab (0)

2. BNI; ((mental* OR development*) adj3 (retard* OR disab* OR defici* OR handicap* OR impairment* OR subnormal* OR sub-normal*)).af (1904)

3. BNI; ((intellectual* OR learning* OR development*) adj3 (impair* OR disab* OR disorder* OR difficult* OR subnormal* OR sub-normal*)).af (7333)

4. BNI; (feeble-minded* OR moron* OR imbecil* OR cretin*).af (8)

5. BNI; Cri-du-chat.af (5)

6. BNI; (Wagr adj3 syndrome).af (0)

7. BNI; (down* adj3 syndrome*).af (453)

8. BNI; (Rubinstein adj3 Taybi).af (1)

9. BNI; (prader adj3 willi).af (27)

10. BNI; (williams adj3 syndrome).af (16)

11. BNI; “fragile X”.af (49)

12. BNI; ((Martin Bell OR FRAXE OR FRAXA)).af (2)

13. BNI; ((les?h adj3 nyhan*)).af (1)

14. BNI; Adrenoleukodystrophy.af (0)

15. BNI; (Coffin ADJ Lowry).af (0)

16. BNI; Menkes.af (3)

17. BNI; “Glycogen Storage Disease”.af (3)

18. BNI; Mucopolysaccharidosis.af (2)

19. BNI; Rett*.af (26)

20. BNI; “Pyruvate Dehydrogenase”.af (0)

21. BNI; “de lange*”.af (20)

22. BNI; 1 OR 2 OR 3 OR 4 OR 5 OR 6 OR 7 OR 8 OR 9 OR 10 OR 11 OR 12 OR 13 OR 14 OR 15 OR 16 OR 17 OR 18 OR 19 OR 20 OR 21 (7790)

23. BNI; forensic.af (904) results.

24. BNI; exp SECURE PSYCHIATRIC HOSPITALS/ OR exp MENTALLY DISORDERED OFFENDERS/ (770)

25. BNI; (secure ADJ (unit* OR hospital* OR service*)).af (527)

26. BNI; offen*.af (1270)

27. BNI; 27 OR 28 OR 29 OR 30 (2057)

28. BNI; 26 AND 31 (364)

Search strategy

1. S1. (MH “Intellectual Disability+”) (14,423)

2. S2. (MH “Mentally Disabled Persons”) (2211)

3. S3. (MH “Developmental Disabilities”) (4374)

4. S4. TX ((mental* OR development*) N3 (retard* OR disab* OR defici* OR handicap* OR impairment* OR subnormal* OR sub-normal*)) (40,779)

5. S5. TX ((intellectual* OR learning* OR development*) N3 (impair* OR disab* OR disorder* OR difficult* OR subnormal* OR sub-normal*)) (61,814)

6. S6. TX (feeble-minded* OR moron* OR imbecil* OR cretin*) (1735)

7. S7. TX Cri-du-chat (159)

8. S8. TX (Wagr N3 syndrome) (23)

9. S9. TX (down* N3 syndrome*) (8682)

10. S10. TX (Rubinstein N3 Taybi) (58)

11. S11. TX (prader N3 willi) (806)

12. S12. TX (williams N3 syndrome) (796)

13. S13. TX “fragile X” (1781)

14. S14. TX ((Martin Bell OR FRAXE OR FRAXA) N3 syndrome*)  (25)

15. S15. TX (les?h N3 nyhan*) (150)

16. S16. TX Adrenoleukodystrophy (161)

17. S17. TX (Coffin N1 Lowry) (25)

18. S18. TX Menkes (277)

19. S19. TX “Glycogen Storage Disease” (220)

20. S20. TX Mucopolysaccharidosis (186)

21. S21. TX Rett* (3568)

22. S22. TX “Pyruvate Dehydrogenase” (196)

23. S23. TX “de lange*” (805)

24. S24. S1 OR S2 OR S3 OR S4 OR S5 OR S6 OR S7 OR S8 OR S9 OR S10 OR S11 OR S12 OR S13 OR S14 OR S15 OR S16 OR S17 OR S18 OR S19 OR S20 OR S21 OR S22 OR S23 (83,897)

25. S25. TI forensic or AB forensic (2740)

26. S26. TX (secure N1 (unit* OR hospital* OR service*)) (1797)

27. S27. (MH “Mentally Ill Offenders”) (1376)

28. S28. TI offen* or AB offen* (3951)

29. S29. S25 OR S26 OR S27 OR S28 (8721)

30. S30. S24 AND S29 (1195)

What we wanted to do

• We are a group of researchers who work in different secure services.

• This is a report on some work we did in 2015.

• We want to know what happens when patients with IDD are in secure hospital and after hospital. We call this ‘outcomes’.

• We want to know things like: how do they change, where do they end up and what support do they get later on?

• This small piece of work was part of planning our next big research project, which will take 5 years.

• Because it will be a big project we want to do it properly.

• So we want to know what questions to ask, what things are important to measure and how to measure them.

• To help to decide this we looked at work by other researchers, and we talked to secure patients and carers, and their doctors, nurses and psychologists.

• We wanted to find out to find out what these people thought we should look at when we are doing our big research project.

• The work was paid for by the NIHR.

What we did

• We looked carefully at 60 studies from around the world.

• We talked to 11 patients and six carers and 15 clinicians.

• We put all this information together.

What we decided

The main three areas we decided we should look at in future research were:

1. Does secure hospital treatment make people better?

2. How safe are patients in hospital?

3. What is hospital like for patients and their carers? And what is their experience on discharge from hospital?

We will talk about each of these.

1. Does secure hospital treatment make people better?

   The conclusions were that in our next research the most important things to measure are:

   1. How peoples’ mental health symptoms change and how patients see that change.

   2. Whether people are reaching their goals for treatment and how patients see their progress.

   3. How peoples’ behaviour changes (e.g. do they harm others or themselves any less often?).

   4. Whether people do actually commit offences or harm others when they come out of hospital.

   5. How people move through the hospital system and how fast do they do that?

2. How safe are patients in hospital?

   The conclusions were that in our next piece of research we should look at:

   1. How often people are secluded or restrained?

   2. How many people feel they are on too much medication?

   3. How many people died or took their own life?

   4. How often other patients are harmed by other patients?

3. What is hospital like for patients and their carers and what happens on discharge?

   The conclusions were that in our next research project we should look at:

   1. How satisfied patients are with their care?

   2. How satisfied carers are with their care?

   3. How involved patients and carers feel in the care – do they have a say?

   When they leave hospital:

   1. Where do people go?

   2. How much support do people get when they leave hospital?

   3. How involved they are in the community?

   4. Whether patients have enough to do in hospital and after hospital.

   5. Do they go back to hospital or prison?

We have suggested ways to measure all these things.

We plan to apply for funding from NIHR to collect information on ALL people with LDs and autism who go through secure hospitals. We want to describe their characteristics in detail and to see what happens to them over 5 years. We want to know what happens both while they are in hospital and after hospital.