Aim
The aim is to determine the effectiveness of selected clinical governance interventions in four public hospitals in South Africa’s Eastern Cape and Mpumalanga provinces.
Study design
This will be an applied multi-methods cluster randomised study characterised by co-design of the intervention with the participants. The design allows for the study to be kept iterative thereby allowing for modifications to be made throughout the study to incorporate new insights. Pre-intervention baseline assessment of clinical governance performance, documents review, interview of key informants and focus groups will produce findings that will form the fundamental pillars of the intervention, which will be co-designed with the involvement of hospital management, clinicians, frontline health workers, hospital boards and organised labour. This design process will draw from hospital experience, experience of the researcher and performance on a set of clinical governance indicators, guidance from the World Health Organization (WHO) and international literature.
Information will be sourced through a multi-method approach of six sub-studies comprising of three phases: a baseline, 12-months long intervention with an assessment at the end, and assessment at 36 months. The six sub-studies will be three quantitative studies and three qualitative studies. The quantitative studies will be conducted at baseline, 12- months, and 36-months, while the qualitative studies will only be conducted at baseline and 36 months.
The intervention entails the co-design of 8 carefully selected clinical governance protocols with two of the four hospitals for participants partaking in sub-studies 2 to 4. These protocols include complaints management protocol; patient safety incidents management protocol; infection prevention and control protocol; clinical auditing protocol; clinical guidelines protocol; clinical education and training protocol; and hospital governance protocol.
The research methods are summarised in Table1.
Study setting
The study is located in two rural provinces with a high degree of under-development and marginalisation, namely Eastern Cape and Mpumalanga provinces. A significant proportion of the population in these provinces live in rural areas (24). Most of the people rely on public health facilities for healthcare. Both provincial departments of health consist of a network of hospitals ranging from district hospitals to tertiary hospitals in Mpumalanga and up to a central hospital in the Eastern Cape province. Regional hospitals in these provinces are not developed enough to provide first-line general medical specialist services and do not have adequate capacity to provide reasonable access to specialist hospital care and thus do not protect tertiary and central hospitals from unnecessary referrals (25). Strong district hospitals and regional hospitals provide an important foundation for a sustainable referral network. Both provinces have developing tertiary hospital services, which provide tertiary care for a majority of the population in the public sector.
However, these hospitals have not yet reached their full potential. Based on anecdotal evidence, the Eastern Cape Province consistently demonstrates poor health outcomes despite significant interventions, strategies and investments made to strengthen the capacity of the health system to deliver quality health services. On the other hand, Mpumalanga Province demonstrates gradual improvement in health outcomes. In addition, an assessment of performance indicators amongst various district hospitals in the Eastern Cape and Mpumalanga provinces revealed differences in performance indicators between district hospitals of the provinces and within district hospitals of the same district. For example, in the OR Tambo Health District, performance assessments ranged from 4 to 11 days for average length of stay, 51 to 99 percent for bed utilisation rate, caesarean section rate ranging from 0.1 – 10 percent, a facility crude death rate ranging from 3-19 percent, a perinatal mortality rate ranging from 15-75 per 1000 live births, stillbirth rate ranging from 14-39 percent and costs per patient day equivalent ranging from a high of R3 398 to a low of R900 (26). The study setting is four hospitals: two in the Eastern Cape Province (a central and a regional hospital), and two in Mpumalanga Province (a tertiary and a regional hospital).
Population and Sampling
The purposive sampling technique was used to select the study hospitals. The hospitals were selected based on their levels of care, gazetted specialist package of care and concerns about hospital performance. The four hospitals have also been prioritised by respective provincial health management for improved service package, improved patient experience, reduction of user complaints and improved health outcomes. The central and tertiary hospitals will be placed in a separate pot from the two regional hospitals. One hospital will be drawn from each of the two pots to select the intervention hospitals. The hospitals that will remain undrawn from the pots will be non-intervention (control) sites.
A triangulation of approaches will be used to select study participants from the four hospitals.
Sub-study 1: Qualitative in-depth interviews (corporate management)
All hospital board chairpersons, CEOs, corporate services managers, and provincial representatives who oversee clinical governance activities will be recruited into the study to participate in individual semi-structured in-depth interviews. Sixteen individuals are therefore expected to participate in this part of the study.
Sub-study 2: Qualitative in-depth interviews using a Delphi technique (clinical management)
All clinical managers, quality assurance managers and nursing services managers will be recruited to participate as experts in this aspect of the study from each of the four hospitals. The number of participants will depend on the number of staff in each hospital as some hospitals have one and others three clinical managers and nursing managers.
Sub-study 3: Qualitative Focus Group Discussions (division heads)
All Heads of clinical Department (HODs), Operational Managers (OM) and Clinical Support and Allied health managers will be recruited to participate in focus group discussions. Identified participants will be randomly assigned to a focus group based on a plan to have a maximum of 5 focus groups per hospital. Each focus group will initially be assigned 4 participants and increased by one participant until saturation.
Sub-study 4: Quantitative cohort study (clinical staff)
Stratified random sampling of clinical staff will be undertaken to recruit medical doctors, nursing staff, dentists, pharmacists, and allied health professionals through a three-stage process.
First: a total combined sample size for all four hospitals will be calculated using the equation,
for a one-sided 95% confidence interval and a 5% significance level (z=1.96). Because the proportion (p) of clinical governance information available is not known, this (p) will be set at 50% and the desired precision (d) will be set at 4%. This thus yields a total minimum sample size of 600. To factor for data entry errors and loss-to-follow-up a further 20% (120) will be added to yield a desired sample size of 720 participants for all four sites.
Second: the clinical staff from the four hospitals as of 25 March 2021 were added together to provide the total population size (N = 2492), wherein hospital 1 = h1 (712); hospital 2 = h2 (480); hospital 3 = h3 (900); and hospital 4 = h4 (400). The weighted hospital sample size was calculated based on the equation,
where y is the value 1 to 4 depending on the hospital being calculated.
Third: clinical staff will be allocated into strata based on their profession. This will thus allow for a calculation of the strata specific sample per hospital. The calculation will be similar to that of calculating the hospital specific sample as above. Table 2 summarises the required samples for each of the hospitals.
Sub-study 5: Quantitative cross-sectional study (end- users)
Systematic random sampling of patients attending the outpatient’s department will be conducted by approaching every 5th patient in the queue until the sampling size has been reached. A total combined sample size for all four hospitals will be calculated using the equation,
for a one-sided 95% confidence interval and a 5% significance level (z=1.96). Because the proportion (p) of patients who are either satisfied or not satisfied with the quality of care received is not known, this (p) will be set at 50% and the desired precision (d) will be set at 4%. This thus yields a total minimum sample size of 600. To factor for data entry errors a further 10% (60) will be added to yield a desired sample size of 660 participants for all four sites. Participants will then be recruited proportionally to yield a sample size of 165 end-user participants per site.
Sub-study 6: Quantitative cross-sectional study (document review)
Information will be extracted from all documents with hospital plans, protocols, reports, and meeting minutes related to clinical governance for the most recent 12-month period. Five clinical records will be randomly sampled for each clinical department. Out of these clinical records, five will be randomly sampled to respond to the questions on the extraction tool. For discipline specific questions, the five clinical records from that clinical department will be used.
Measurements
Description of measurements
A multi-method approach to data collection will be adopted to get a comprehensive picture of the study before and after the intervention, to also compensate for the potential limitations of a single data collection method and to triangulate the data as a means of checking the consistency of the study findings.
Sub-study 1: Qualitative in-depth interviews (corporate management)
An in-depth interview guide (Appendix A) will be used for collecting data from the corporate management of the hospitals. The in-depth interviews will be conducted to solicit important perceptual information from the participants in their own words. The aim of the key informant interview is to probe understanding and beliefs about the intervention through the eyes of the hospital management and provide in-depth descriptive information to enhance our understanding of the ways in which respondents construct their understanding of the intervention. The use of an interview guide will allow for probing of topics that can be difficult to engage adequately in a structured questionnaire. This instrument asks questions on: Hospital challenges; Understanding of hospital performance; Indicators of hospital performance; Understanding of clinical governance; Presence of clinical quality improvement or Clinical governance plan; Implementation of the clinical quality improvement or Clinical governance plan; Systems for implementing key clinical governance activities/clinical quality improvement activities in the hospital; Existing structures for enforcing the implementation of clinical governance activities/clinical quality improvement activities; Barriers and facilitators to implementation of clinical governance/clinical quality improvement activities. With permission from the participant, an audio recorder will be used to capture the interview. In addition to the audio recorder, a notebook will be used to capture the discussions. In-depth interviews will only be conducted at baseline and at 36 months.
Sub-study 2: Qualitative in-depth interviews (clinical management)
A Delphi technique will be used for collecting data from the clinical management of the hospitals. The Delphi technique has an inherent ability of being a forecasting process framework. In this instance, the researchers are interested in establishing clinical management views on how they foresee clinical governance in future. The Delphi technique will be employed to solicit expert opinion about key areas and factors to be considered to strengthen hospital’s capacity to perform. Additionally, this approach will allow for probing understanding and beliefs about the intervention through the eyes of the hospital management and provide in-depth descriptive information to enhance our understanding of the ways in which respondents construct their understanding of the intervention. The in-depth interview guide (Appendix B) questions will be used for this process. The use of an interview guide will allow for probing of topics that can be difficult to engage adequately in a structured questionnaire. This instrument asks questions on: Hospital challenges; Understanding of hospital performance; Indicators of hospital performance; Understanding of clinical governance; Presence of clinical quality improvement or clinical governance plan; Implementation of the clinical quality improvement or clinical governance plan; Systems for implementing key clinical governance activities/clinical quality improvement activities in the hospital; Existing structures for enforcing the implementation of clinical governance activities/clinical quality improvement activities; Barriers and facilitators to implementation of clinical governance/clinical quality improvement activities; and adaptation of clinical governance protocols during the covid-19 pandemic. With permission from the participant, an audio recorder will be used to capture the interview. In addition to the audio recorder, a notebook will be used to capture the discussions. The Delphi technique collection will only be conducted at baseline and at 36 months.
Sub-study 3: Qualitative Focus Group Discussions (division heads)
The main aim for conducting focus group discussions is to provide for group interactions that will reveal common experiences amongst clinical staff and frontline health workers about the intervention. This will help us get a perspective of what is happening on the ground as against what is ought to happen. A focus group interview guide (Appendix C) will be used to guide focus group discussions so that topics that can be difficult to engage adequately in a structured questionnaire can be probed. This instrument asks questions on: Hospital challenges; Understanding of hospital performance; Indicators of hospital performance; Understanding of clinical governance; Presence of clinical quality improvement or clinical governance plan; Implementation of the clinical quality improvement or clinical governance plan; Systems for implementing key clinical governance activities/clinical quality improvement activities in the hospital; Existing structures for enforcing the implementation of clinical governance activities/clinical quality improvement activities; Barriers and facilitators to implementation of clinical governance/clinical quality improvement activities. The clinical governance management team will comprise of head of departments/divisions/units. An audio recorder will be used to capture the interview if permission has been granted by the participant. In addition to the audio recorder, a notebook will be used to capture the discussions. Focus groups will only be conducted at baseline and at 36 months.
Sub-study 4: Quantitative cohort study (clinical staff)
The main aim for conducting this survey is to quantify the views of general clinical and support staff in relation to the implementation of clinical governance in their respective hospitals. A validated structured questionnaire (Appendix D) consisting of questions on demographic data; governance and management; patient safety and leadership; clinical effectiveness; adverse events management; organisational culture; complaints management; clinical information monitoring; infection prevention and control; and performance management systems adopted from a study on assessment of clinical governance implementation study (15)will be used. Data will be solicited from ward-based clinical staff including pharmacy, dental and allied health professionals. This survey will be conducted at baseline, at 12 months, and at 36 months.
Sub-study 5: Quantitative cross-sectional study (end-users)
The aim of this survey is to get the user perspective. The general patient experience survey (Appendix E) used by the South African department of health will be adopted and used for collecting data. This instrument asks questions on biographical data; access to care; availability and use of medicines; patient safety; cleanliness; values and attitudes; waiting times and clinical audits. This questionnaire will be translated into local languages such as isiXhosa, siSwati, and isiZulu to accommodate participants who might not be comfortable with English. This survey will be conducted at baseline, at 12 months, and at 36 months.
Sub-study 6: Quantitative cross-sectional study (document review)
Document review (Appendix F) will be conducted in addition to the key informant interviews and focus group discussions. The main aim of the document review is to provide access to hospital documents with historic information which pre-dates the study and provide important context about official statements, policies, priorities, and strategies that can be compared to what is observed through primary data collection. During the review, the focus will be on official policy documents, strategic plans, annual operational plans, hospital reports and recent reviews of the performance of the hospitals to add weight and granularity to other sources of data. Relevant information will be extracted to give insight to research objectives. In addition, information from clinical records will be used to validate responses given during the interview of clinical staff. The review will be conducted at baseline, at 12 months, and at 36 months.
Biases, validity, and reliability
Potential biases include power imbalances between researcher and patient, where a patient’s voice may be suppressed, interviewer bias, selection bias and measurement bias. All those who will be involved in the collection of data will undergo similar training to ensure standardisation. All data collection instruments will be standardised and validated to ensure reliability. Tools will be piloted in one regional hospital each in Eastern Cape and Mpumalanga provinces, thereafter all necessary adjustments to the tools will be made.
Data analysis
Qualitative data (key informant interviews and focus group discussions) will be analysed looking for the meaningful and symbolic content of data collected using NVIVO version 12. This will be achieved through a process of inductive analysis of qualitative data to allow research findings to emerge from the dominant themes derived from the raw data (27). Emerging findings will be summarised in descriptive words, phrases, themes, or patterns to assist in the understanding and interpretation of emerging themes. Analysed data will be interpreted in the context of existing literature to show how it corroborates existing knowledge or bring new insights to the existing body of knowledge.
Post-hoc analysis will be undertaken for information arising out of the document reviews using prevailing themes. Quantitative data analysis will include capturing survey data into the Redcap software and exporting the data into STATA version 17 (STATA Corp, College Station, Texas, USA) for analysis. Some descriptive and categorical data will be compared using frequencies, percentages, and graphs. Numerical data will be explored for normality using the Shapiro-Wilk test. If normally distributed the mean, range and standard deviation will be used. If not normally distributed, then the median, and interquartile range (IQR) will be used. The Chi-squared or Fisher’s exact tests will be used depending on the value of the expected frequencies. The level of statistical significance will be set at p-value ≤0.05. The 95% confidence interval will be used for the precision of estimates. The survey database will be used to perform descriptive analysis and examination of data for reliability and validity. The survey will be administered, in part, to triangulate and confirm the in-depth information gleaned from the interviews, document reviews and focus groups. The emerging findings from the qualitative data (key informant interviews and focus groups) and the quantitative findings from the survey data will be integrated to gain insight and deeper understanding of hospital performance and the intervention.
Ethical considerations and Dissemination
Ethical clearance with reference number: 040/21 has been granted by the Research Ethics Committee of the Faculty of Health Sciences at Walter Sisulu University. The study has also received approval to access the research sites from the Eastern Cape and Mpumalanga Provincial Health Research Committees with refence number: EC_202106_019 and MP_202106_009 respectively. Entry to the study sites will be further negotiated with the hospital CEOs before data collection. The study will abide by the four ethical principles of autonomy, beneficence, non-maleficence, and justice.
Before commencement of all data collection, participants will be asked to sign an informed consent form including consent for the qualitative study interviews to be recorded. The consent forms will be explained in detail, including the purpose of the study, the objectives of the study, what is expected of a research participant and any risks and benefits that accrue from participating in the study. A provision will be made in the event that a study participant does not wish to be recorded, wherein a second researcher will always be at hand to take notes by hand. Participants will be informed that their participation in this study is voluntary and that their confidentiality will be maintained throughout the study. All research staff will always sign a confidentiality agreement, agreeing to maintain and protect participant confidentiality. Research staff taking the informed consent will provide study participants with all the necessary information they require to make an informed decision to participate and will allow time for any questions to be addressed prior to signing the consent. Participants will also be assured that they are free to withdraw at any stage of the study and could opt out of questions that they are not comfortable with without any adverse consequences. Furthermore, participants will also be assured that there is no risk posed by their participation in the study. Even though findings from this study will benefit the broader public through improvement in quality of healthcare, there won’t be individual benefits or incentives for participating in this study. All identifying information will be removed including data on audiotape collected during one-on-one interviews/focus groups. Furthermore, audiotapes will be downloaded to a computer that will be locked with a password, after which the recorded information will be deleted from the tape recorder. Once the audiotape has been transcribed, the audio recording will be destroyed to ensure confidentiality. All electronic records will be accessed through a password encrypted database that only the principal investigator has access to.
There is a potential burden on clinical staff for taking time out from patient care to participate in a focus group discussion. Benefits include the interactions between the research team, hospital management, clinical staff, and patients, who will provide the opportunity for introspection, reflection on own practice, and space to reflect on their roles as individual health professionals and as members of a healthcare team.
Finally, research findings will be communicated through several dissemination mechanisms including workshops or seminars to engage with the stakeholders, policy briefs, technical reports, and publication in peer-reviewed journals.