An embedded mixed methods investigation of factors associated with the quality of mental health care in the Kingdom of Saudi Arabia (KSA): a structure, process and outcome evaluation

Background: Quality of care (QoC) is the essential determinant of service provision improvement for service users in mental health settings. However, there is a lack of research investigating the factors associated with the quality of mental health care and their relationship to service user outcomes, particularly in Saudi Arabia. This is despite the high priority given to QoC in mental health care in the Kingdom of Saudi Arabia’s (KSA) national health strategy. Aim: To investigate factors associated with the quality of mental health care and their relationship to service user outcomes (the so-called ‘5Ds’: death, disability, disease, discomfort and dissatisfaction), and to explore service users’ perceptions of the quality of mental health care that they receive. Methods: An embedded mixed methods design was applied in this study, whereby a qualitative strand was embedded in a quantitative study. In the quantitative strand, a survey was distributed to 176 staff nurses and 321 service users aged 18 years or over, diagnosed with mental disorders and admitted to 11 male and 3 female inpatient psychiatric wards. The nurses’ measures consisted of the staff characteristics sheet (SCS), used to elicit information about nurses’ age, gender, nationality, qualifications, years of experience in the hospital, attendance of psychiatric training courses and staff numbers; Attitudes Towards Acute Mental Health (ATAMH-33), used to assess nurses’ perceptions of the mentally ill/mental illness, and the Competency Assessment Tool-Mental Health (CAT-MH), used to assess nurses’ competency levels in the delivery of mental health care. Meanwhile, the service users’ measures comprised: the Service User Characteristics Sheet (SUCS), used to collect information about service users’ age, gender, marital status, employment status, educational level, area of residence and diagnosis; the Arabic version of the Social Functioning Scale (SFS), used to assess psychiatric service users’ capacity to engage in necessary social activities; the Arabic version of the Glasgow Antipsychotic Side-effect Scale (GASS), used to assess service users’ experience of the side-effects of psychiatric medication; the Arabic version of the Client Satisfaction Questionnaire (CSQ-8), used to assess service users’ satisfaction with the mental health care services received, and the Brief Psychiatric Rating Scale (BPRS), used to assess specific psychiatric service user symptoms, based on service users’ self-reports. In addition, participant observation was conducted of the interaction between staff and service users across six admission wards (4 male/2 female), using the Interaction-Observation Checklist (IOC). For the qualitative component, semi-structured interviews were conducted with 30 service users (24 male/6 female), using purposive sampling to explore their perceptions of the quality of their care provision. Findings: Multilevel modelling showed that service users’ characteristics, such as age, marital status, employment status, educational level, area of residence, and diagnosis, significantly predicted disability, whilst gender did not. In fact, with just one exception (unemployment), none of the service user characteristic variables was found to be a statistically significant predictor of disease, although educational level was found to be a significant predictor of discomfort and dissatisfaction. However, age and years of experience were found to significantly predict disability amongst nurses and years of experience was a significant predictor of their dissatisfaction. Moreover, nurses’ competence was found to be a statistically significant predictor of disability, while their attitudes toward the mentally ill/mental illness and staff-service user interaction did not significantly predict service user outcomes. An exploration of service users’ perceptions of important dimensions of perceived service quality subsequently included the themes: the ‘hospital is like a prison’, quality of interaction, service quality, staff qualities, and suggestions for achieving quality of care. Conclusion: Overall, the quantitative results revealed that the factors associated with quality in mental health care (i.e. structure and process measures) and/or measures of service user outcomes (i.e. SFS, BPRS, GASS and CSQ-8), used within Donabedian's Framework, indicate less relevance in the Saudi context, although the study used a considerable sample size and implemented valid and reliable measures. In contrast, the qualitative results identified factors of greater importance in the Saudi context. Therefore, future research should consider evaluating the conceptual model developed (a revised model); utilising Donabedian’s Framework for QoC, incorporated into the Andersen Behavioural Model of Health Service Use and based on the outcomes of the qualitative analysis. This can be tested quantitatively to determine the significant predictors. Thus, the originality of the current study comes from the first test of these proposed qualitative indicators, in a context that has not previously been studied. The implications of these findings for mental health care policy and clinical practice are subsequently discussed, with recommendations being made for future mental health service planning, in order to enhance QoC for patients with mental disorders in KSA