A Thesis submitted to the Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, in fulfilment of the requirements for the degree of Doctor of Philosophy December 2015 Johannesburg. / Introduction There has been little published in South Africa regarding quality improvement in health and in particular the involvement of patients in this intervention. There is evidence globally that both quality improvement efforts and particularly the engagement of the users adds value to health services. Three projects were conceived around this core concept as explained below. 1. Systematic review. Patients’ involvement in improvement initiatives: a qualitative systematic review After a search was done of databases, 5121 papers were found to be potentially relevant. After screening and critical appraisal for eligibility, it was found that 31 articles qualified for analysis. These were then assessed using JBI software and 5 categories and 2 metasynthesised findings were documented. In summary, there were enablers and barriers to involving patients. The five categories which lead to these 2 findings were the following: (1) although patient participation in QI is acknowledged and encouraged by many policies and documents globally, it is difficult to implement; (2) there are differing views between patients and providers as to the process; (3) on the positive side, different levels of involvement of patients in QI were demonstrated; 4) practical, appropriate and innovative results emerged; (5) individual or group support and incremental development through skills and enablement contributed towards success 2. The Integration of Non Communicable Chronic Diseases (NCDs) and HIV/Aids and mental health care through the involvement of chronically ill patients using Empowerment Evaluation (EE). At 9 primary care clinics, the process of EE was followed with chronically ill stable patients and appropriate healthcare workers. This was an additional intervention in an ongoing QI cycle on the integration of all chronic illnesses into one model, based on Lean principles. Steps followed were ‘’taking stock’’ ie assessing patients’ and HCWs’ impressions of the services at the clinic in a measured way, creating a vision and using this as a yardstick for the project and then problems and solutions being co-managed by the collaborative team. A total of 37 interventions were discussed and 23 implemented in the time frame. Innovative solutions were implemented and teams were empowered by the potential they experienced. 3. An exploration of childhood nutrition and wellness in a subdistrict by patient inclusivity in QI using experience based codesign (EBCD) with mothers/caregivers of malnourished children . Following the steps of EBCD, staff and patients exposed to health services regarding ill children, were interviewed, feedback was given of the findings separately and then in a combined meeting and co-design teams were created to work with the prioritised quality improvement interventions. Touch points in the system were examined through emotional mapping, video interviews and observations. Within the 10 month period of the project, 38 interventions were identified and 25 accepted and implemented at different levels. Conclusion The methodologies were chosen to fit with the qualitative aspect of the research. There were concrete appropriate improvement outcomes due to the engagement of service users in both the primary care clinics serving chronically ill patients and the paediatric system in one subdistrict eg the flow of patients improved, logistical improvements like direct admissions for very ill children, school and library opportunities for admitted ill children etc occurred. Subjective gains like the acknowledgement of their power role by patients and a flattening of the healthcare worker hierarchy were also experienced in the research. Other findings were that unexpected roleplayers were identified, the timeframe of such QI cycles needs to be considered especially regarding the resilience of patients and resources were not an importatn limitation. However some modifications would have to be considered to make these research approaches common practice. The particular research methodologies have not been published in a South African context before and have also not been used for paediatric or integrated chronic illness research and therefore contribute both content and process information to health systems research in South Africa. / MT2017

We assessed how estimates of need-expected inpatient hospital use differ depending on whether need-expected use was estimated for a population of all Canadians, Canadian health regions, or a subpopulation of higher income Canadians, who likely had minimal healthcare access problems. Data came from the 2009/2010 Canadian Community Health Survey, a national cross-sectional survey. Using zero-inflated negative binomial regression, we modeled inpatient hospital use separately based on the three aforementioned choices of population. We adjusted for demographic, health behaviour, health status, socioeconomic, and health care supply factors. We then estimated need-expected inpatient hospital use and compared the estimates across individuals and by income and province. The three choices of population that we used in this study had similar results. Our estimates of the average need-expected use by province or income group were not sensitive to the choice of population used to estimate need-expected use.