How well do international clinical guidelines on mental health and substance misuse address their coexistence? a systematic review of scope, quality and inclusivity


 Objective: It is estimated that up to 75% of patients with severe mental illness (SMI) also have substance use disorder (SUD). The aim of this systematic review is to explore the scope, quality and inclusivity of international clinical guidelines on mental health and/or substance misuse in relation to diagnosis and treatment of such co-existing disorders and considerations for wider social and contextual factors in treatment recommendations.Method: A protocol (PROSPERO CRD42020187094) driven systematic review was conducted. A systematic search was undertaken using six databases including MEDLINE, Cochrane Library, EMBASE, PsychInfo from 2010 till June 2020; and webpages of guideline bodies and professional societies. Guideline quality was assessed based on ‘Appraisal of Guidelines for Research & Evaluation II’ (AGREE II) tool. Data was extracted using a pre-piloted structured data extraction form and synthesized narratively. Reporting was based on PRISMA guideline.Result: A total of 12,644 records were identified. Of these, 21 guidelines were included in this review. Three of the included guidelines were related to coexisting disorders, 11 related to SMI, and 7 guidelines were related to SUD. Seven (out of 18) single disorder guidelines did not adequately recommend the importance of diagnosis or treatment of concurrent disorders despite their high co-prevalence. The majority of the guidelines (n = 15) lacked recommendations for medicines optimisation in accordance with concurrent disorders (SMI or SUD) such as in the context of drug interactions. Social cause and consequence of dual diagnoses such as homelessness and safeguarding and associated referral pathways were sparsely mentioned.Conclusion: Despite very high co-prevalence, clinical guidelines for SUD or SMI tend to have limited considerations for coexisting disorders in diagnosis, treatment and management. There is a need to improve the scope, quality and inclusivity of guidelines to offer person-centred and integrated care.

We considered the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) de nition of, 'substance use disorder' which is a single term combines both abuse and dependence (24). Such substances include legal drugs such as alcohol, illicit drugs such as heroin and cocaine, and prescription drugs such as oxycodone (25). The SMIs considered in this review were psychosis and other associated types of schizophrenia, as well as bipolar disorder. The terms coexisting disorder, co-occurring disorder, or dual diagnosis are frequently used to describe the existence of both conditions of SMI and SUD simultaneously.

Data extraction
After identi cation of eligible guidelines, data were extracted using a Microsoft Excel® spreadsheet. Data were extracted in relation to guideline characteristics, targeted patient population and health care providers, screening and management of co-existing disorders including recommendations for treatment adjustments and consideration of monitoring of physical health or drug interactions. Consideration of offending behavior, risks of homelessness, violence, and suicide were also extracted. Data extraction was done by two authors (RA and VP) in duplicate and independently and any disagreements were resolved by further discussion.

Quality assessment
The included guidelines are appraised by using the Appraisal of Guidelines for Research & Evaluation II (AGREE II) tool. The assessment of each guideline is carried out by following the users' instruction manual for AGREE II instruments (26). The assessment for the following domains: 'scope and purpose, stakeholder involvement, rigor of development, clarity of presentation, applicability, and editorial independence' (27). Each of the 23 items is scored 1 to 7 where 1 signals strong disagreement and 7 signals strong agreement and the nal score is rated from 0% to 100%. In addition, there are two overall assessments of each guideline. The rst one re ects the overall quality of each guideline. The second overall assessment allows assessment of whether or not the guideline is recommended for application in practice. Three distinct choices; namely, 'Yes', 'Yes with modi cation', or 'No' are utilized in relation to recommendation for use. Score sheet is demonstrated in appendix C. Two reviewers independently assessed the included guidelines.
In order to calculate domain rate, the following equation from AGREE II users' manual was used: The rate of each domain = (total score of all items within the domain − lowest score of all items within the domain) ⁄ (highest score of all items within the domain − lowest score of all items within the domain) x 100 A narrative synthesis was used to analyze the data. Comparisons between guidelines are pre-identi ed in accordance with the particular objectives of the review.

Results
The search and selection of guidelines In total, 12,644 records were identi ed through the searching of various databases. After the exclusion of data de-duplication and both title and abstract screening, 32 guidelines were screened for eligibility. Twenty-one guidelines were included in this study (Fig. 1).

General Characteristics Of The Included Guidelines
Of the 21 included, three guidelines related to coexisting disorders, seven guidelines related to SUD including alcohol use disorder and opioid disorder (Table   1). Eleven guidelines related to SMI (six of them were related to schizophrenia, and ve of them were related to bipolar disorder). The aim of each guideline is illustrated in Table 1.   (Table 1).

Quality Assessment Of Guidelines
The scores of each guideline against the criteria of the AGREE II tool are displayed in Table 2. In terms of 'scope and purpose', rst domain had the highest domain score. Only four guidelines scored below 80% (29,31,33,36) ( Table 2). In the second domain, 'stakeholder involvement', the guidelines that were developed by NICE and Scottish Intercollegiate Guidelines Network (SIGN) demonstrated the highest score; 84% and 83%, respectively (30,37-41) ( Table 2).The 'Rigour of development' domain scores were generally low (Fig. 2). Fifteen out of 21 included guidelines rated below 70% ( Table 2). Most of the guidelines scored higher in 'Clarity of presentation' domain ( Fig. 2). The guidelines that were developed by NICE and SIGN obtained the highest scores (30,37-41) ( Table 2). Figure 2 shows that the 'Applicability' domain has the lowest domain score. Fifteen guidelines were graded below 50% (Table 2).With regard to the 'Editorial independence' domain, the highest score was reported with the NICE guidelines, this being 83%. The rest of the included guidelines were graded below 80% (Table 2) (Fig. 2).   Three guidelines mentioned that patients with SMI with coexisting SUD who completed their SMI treatment course should not be discharged from a healthcare setting due to their substance misuse (28,35,39). Of the SMI guidelines, four guidelines highlighted the competency need of healthcare providers in each health care setting to consider for the co-existing disorders (30,32,36,43).Three out of seven SUD guidelines similarly covered competency aspects (31,34,37).
All coexisting disorder guidelines requested healthcare providers to gain training and expertise from other specialist staff (33,39,40) (Table 3).

Treatment Of Coexisting Disorders
All of the guidelines related to SMI or coexisting disorders described the importance of screening and/or treatment for both problems simultaneously (33,39,40). The SMI guidelines often stipulated SUD clinical guidelines and vice versa when recommending treatment of the other co-existing disorder ( Table 4). One SUD guideline (44) however, did not explicitly provide recommendation regarding treatment of both disorders.  Similarly, only three of the seven SUD guidelines mentioned recommendation about treatment adjustment (35,42,46) ( Table 4). Examples of treatment adjustments included recommendation for the use of long-acting injectable antipsychotic medication in cases where there was a history of non-adherence to medication in place of regular antipsychotic medication (39). In addition, two of the guidelines related to coexisting disorders (33,39), ve of SMI guidelines (28,38,43,47,48) and three of the SUD guidelines (34,37,49) considered potential drug interaction in patients with SMI and coexisting SUD (33,39). For example, the NICE (2011) guideline recommends that caution be exercised during the prescribing of medication for patients demonstrating substance misuse particularly that of alcohol, since alcohol will affect the metabolism of other medications and either diminish their e cacy or increase the risk of side effects (39) ( Table 4).
Importance of physical health monitoring were described by all guidelines related to coexisting disorders, nine SMI guidelines, and four of the seven SUD guidelines. These included monitoring and management of diabetes mellitus and hyperlipidemia (Table 4).

Care Pathway And Integrated Care Provision
All of the coexisting disorders guidelines, seven of the SMI guidelines, and three of SUD guidelines mentioned the importance of continuity of care. For example, the Australian government guideline advised that it is important to develop systems in order to facilitate the transition of patients with coexisting disorders by providing them with much-needed services and helping them to address their complex needs (33) ( Table 5).  Only one of the guidelines pertaining to coexisting disorders mentioned that healthcare providers in the emergency department should regularly ask patients about any potential drug misuse (39). Three of the guidelines related to SUD mentioned the role of the emergency department (31,34,37). Such consideration was missing from SMI guidelines. (Table 5).

Equity Consideration And Person-centered Care
Three guidelines pertaining to coexisting disorders, ten SMI guidelines, and six SUD guidelines described the essential role played by 'signi cant others' such as families and carers and encouraged their involvement along with any integrated care plans provided to patients ( Table 6). All of the three guidelines pertaining to coexisting disorders were explicit in reporting the need for assessment of any children cared for by patients with both disorders, according to safeguarding procedures. However, only three of the SMI guidelines and two of the SUD guidelines provided recommendations about children cared for by patients with both disorders (Table 6). All of the guidelines pertaining to coexisting disorders, ve of the SMI guidelines, and two of the SUD guidelines mentioned the importance of ensuring that healthcare providers who provide care to patients with coexisting disorders should engage with patients from different ethnicities and cultural backgrounds (Table 6). Only the NICE 2011 offered advice to healthcare providers to solve access to care issues in patients (39) ( Table 6).

Consideration Of Multiple Social Disadvantage
All of the guidelines pertaining to coexisting disorders, nine of the SMI guidelines, and ve of the SUD guidelines considered the assessment of risks of violence, suicide, and self-harm ( Table 7). Two of the guidelines pertaining to coexisting disorders highlighted the risk of certain getting involved with criminal justice system and the importance of prevention actions (39,40). Only the SMI guideline by Royal Australian and New Zealand College of Psychiatrists (RANZCP) (32) and three of the SUD guidelines (31,34,37) highlighted the risk of patients being registered in the criminal justice system (Table 7).
government mentioned the risk of homelessness in patients with coexisting disorders, but did not provide further recommendations about how such patients could receive support (33) ( Table 7). Assessment of the history of any kind of abuse suffered by the patient, including sexual abuse were only rarely considered (28,33,34,37,38,43) (Table 7).
Issue of stigma and discrimination from healthcare providers were covered well by guidelines for co-existing disorders but less so by either SMI or SUD guidelines ( Table 7).
Two of the guidelines pertaining to coexisting disorders, two of the SMI guidelines, and two of the SUD guidelines seemed to encourage seeking support from voluntary organizations (32,34,37,(39)(40)(41) (Table 7).

Discussion
This study provides an up-to-date assessment of the scope, quality and inclusivity of international clinical guidelines on mental health and/or substance misuse in relation to diagnosis and treatment of such co-existing disorders and consideration of wider social and contextual issues in treatment recommendations. The overall quality of the included guidelines rated from a high to moderate quality. The 'scope and purpose' and 'clarity of presentation' domains were well addressed by the included guidelines. Previous systematic reviews have also demonstrated that clinical guidelines often score high in these domains (50)(51)(52). For the 'Stakeholder involvement', it was noticed that there was a lack of incorporation of patient or public preferences in the guidelines development process. The 'applicability' domain was rated low amongst all the guidelines.
This review has demonstrated that there is a shortage in the number of guidelines that are produced to address concurrent problems. More importantly any existing single disorder guidelines should incorporate coexisting disorders in diagnosis and treatment recommendations. These guidelines need to be consistent with current evidence, and in providing care for patients' multiple disadvantages including wider social and contextual factors such as homelessness, involvement with criminal justice system.

Implication Of Practice And Research
Until recently, most of the guidelines and recommendations addressed a single disorder; namely, either SMI or SUD. The result of this review suggests that a greater number of guidelines are required in order to cover dual diagnoses given the high overlap of the concurrent disorders.
Most single disorder guidelines included in this review did emphasize the importance of assessment of dual diagnosis. However, treatment adjustment for dual diagnoses was rarely described. Barriers of access to medicines, adherence issues requiring long acting depot injections, and drug interactions (including interactions with drug and substance of abuse) are key issues that require further considerations in single disorder guidelines.
There needs to be better emphasis on the integrated and inclusive care to be offered to the patients with dual diagnosis. The need for liaison with emergency department, primary care, drug and alcohol services and hospital and specialist treatment centers also require further emphases. There is also scope to enhance cultural and ethnic speci c issues in treatment recommendations.
It is well documented in the evidence that the treatment of coexisting disorders multifaceted and requires the continued assessment of many social and contextual issues of a patient. Social and contextual factors were not however uniformly addressed in the included guidelines. While risk of homelessness in patients with SMI, SUD or dual diagnosis was commonly described, further information to health providers to support prevention actions were often missing. It is imperative to signpost patients to housing assistance, volunteer sectors and social bene ts system in order to prevent homelessness including repeat cycle of homelessness. Adequate evidence exist on the overlap between homelessness, SUD, SMI and dual diagnosis (53). Persons who are homeless or risk facing homelessness often nd accessing services di cult and future guidelines should consider addressing access issues better (54)(55)(56). These include perceived stigma and discrimination in healthcare setting. Some guidelines described risks of homelessness with dual diagnoses. There are various barriers which patients experiencing homelessness and SUD must overcome in order to obtain housing due to their criminal record and economic status, all of which make them more susceptible to being submerged in their current negative environment and seem to increase the risk of relapse (57,58).
Only a limited number of guidelines considered the continuity of care of offenders in community settings. Accordingly, this is an important factor to be considered and needs to be addressed during the production of any new guidelines.
There needs to be better emphases on the integrated and inclusive care to be offered to the patients with dual diagnosis. Liaison with emergency department, primary care, drug and alcohol services and hospital and specialist treatment centers require further emphases. There is also scope to enhance cultural and ethnic speci c issues in treatment recommendations. Roles of community based services such as community pharmacy and voluntary sectors should be better stipulated in the guidelines (59)(60)(61).
Future research is need needed to cover healthcare professional, patient, carer and payer's perspectives to identify ways to strengthen the guidelines and limitations and improve patient experiences of care and outcomes. It is also imperative to compare practices against the guideline recommendations. For example, research suggest that patients prescribed antipsychotic medicines are often poorly followed up for their cardiovascular and metabolic health in contrary to the recommendations from the guidelines (62). Guideline development procedures should learn and share best practices being adopted in other countries.

Study Strengths And Limitations
This is the rst systematic review to discuss coexisting disorders and aspects of their different complex needs. A comprehensive search was undertaken using databases and professional body web pages. Validated appraisal tool (AGREE II) was used for quality assessment. However, our search was restricted to English language guidelines only. In addition, we did not assess any supplementary patient screening, risk assessment and patient placement criteria that were not included or appended within the published guidelines.

Conclusion
Treatment guidelines for management of either SUD or SMI have tend to have limited considerations of coexisting disorders. There is a need for the guidelines to be more inclusive in order to enable better diagnosis and treatment and cover social cause and consequences of concurrent disorders.