Patient safety has been on the research agenda since 2000, when unnecessary harm to patients in providers’ care came to light. In 2005, the improvements in patient safety fell short of expectations, and the patient safety research community recognized that the issues are more difficult to resolve than first thought. One of the tools to address this vexing problem has been voluntary incident reporting systems, although the literature has given incident reporting systems mixed reviews. This qualitative comparative case study comprises 85 semi-structured interviews in two separate divisions of a tertiary care hospital, General Internal Medicine (GIM) and Obstetrics and Neonatology (OBS/NEO). The main line of questioning concerned incident reporting; general views of patient safety were also sought. This is a thesis by publication. The thesis consists of a general introduction to patient safety, a literature review, a description of the methods and cases, followed by the manuscripts. The thesis concludes with a summarization of the findings, and implications of the study. Manuscript one focuses on the reporter end of incident reporting systems. It asks what frames underlie GIM nurse and physician self reporting and peer reporting practices. The findings showed that frames that inhibit reporting are shared by physicians and nurses, such as the fear of blame frame regarding self reporting, and the tattletale frame regarding peer reporting. These frames are underpinned by a focus on the individual, despite the organisational message of reporting for learning. A learning frame is an enabler to incident reporting. Viewing the objective of voluntary incident reporting as learning allows practitioners to depersonalize incident reporting. The focus becomes preventing recurrence and not the individual reporting or reported on. Manuscript two again focuses on the reporter end, and on one type of reportable incident – a problem that healthcare practitioners can fix themselves. The study asks: when faced with such a problem, do practitioners fix it in the moment and forget about it, or do they fix it in the moment and report it? We found that “fixing and forgetting” was the main choice that most GIM practitioners made in situations where they faced problems that they themselves could resolve. These situations included a) handling near misses, which were seen as unworthy of reporting since they did not result in actual harm to the patient, b) prioritizing solving individual patients’ safety problems, which were viewed as unique or one-time events, and c) encountering re-occurring safety problems, which were framed as inevitable, routine events. The paper argues that fixing and forgetting patient safety problems encountered may not serve patient safety as well as fixing and reporting. The latter approach aligns with a preventive view of patient safety. Manuscript three focuses on the practice of double checking, drawing from interviews conducted in both GIM and OBS/NEO. It asks what weaknesses are in the double checking process and what alternative views can help the double checking process enhance patient safety. The findings showed weaknesses in the double checking process, such as: a) double checking trusted as an independent process, b) double (or more) checking as a costly and time consuming procedure, and c) double checking as preventing reporting of near misses. It is proposed that there are alternative ways of viewing and practising double checking in order to enhance patient safety. These include: a) recognizing that double checking requires training, b) introducing automated double checking, and c) expanding double checking beyond error detection. The paper argues that practitioners need to be more aware of the caveats of double checking, and to view the double checking process through alternate lenses to help enhance its effectiveness. Manuscript four focuses on the reporting system more comprehensively, and attends to the reporting process in GIM and OBS/NEO. This is a comparative case study of the two divisions, and considers the different stages in the process and the factors that help shape the process. The findings showed that there were major differences between the two divisions in terms of: a) what comprised a typical report (outcome based vs communication and near-miss based); b) how the reports were investigated (individual manager vs interdisciplinary team); c) learning from reporting (interventions having ambiguous linkages to the reporting system vs interventions having clear linkages to reported incidents); and d) feedback (limited feedback vs multiple feedback). The differences between the two divisions can be explained in terms of: a) the influence of litigation on practice, b) the availability or lack of interprofessional training, and c) the introduction of the reporting system (top-down vs bottom-up approach). A model based on the findings portraying the influences on incident reporting and learning is provided. This thesis contributes to an in-depth understanding of front line perspectives on incident reporting systems and safety, and aims to provide insights into improving patient safety. Implications for practice and research will be addressed.

Background: Much emphasis has been placed on Quality Measurements or Key Performance Indicators in Emergency Medicine. Internationally, KPI's are used to measure and improve quality of care, with a major emphasis on waiting times, measured as time-based KPI's. These times are related to the various stages of a patient journey through the Emergency Center. In South Africa this has not been routinely done. The Western Cape has conducted audits in recent years to measure these. This study aims to provide a snapshot of waiting times (specifically time to triage, time to doctor, time to disposition decision and time to departure from the EC) within Cape Town public sector Emergency Centres. Methods: This is a retrospective descriptive study of waiting times for all patients presenting to Emergency Centres in the Western Cape in 2013-2014, as per six monthly waiting times audits conducted by the Western Cape Department of Health. A wide variety of emergency centers were audited, from 24 hour clinics to larger acute hospital based ECs. Results: The proportional acuity difference between hospitals and CHC for the first random 100 folders were statistically no different. Arrival to triage times were universally longer than internationally accepted as safe. The mean time for all-comers across all facilities was just under an hour, the higher acuity patients were triaged significantly faster (half an hour) than the lower acuity patients (hour or more). This difference was significant for hospitals, with a non-significant trend for CHCs. At hospital ECs, green patients were triaged significantly faster than yellow patients; this was not the case at CHCs. The mean time from triage to clinician consultation for all-comers across all facilities (over two hours) was significantly longer at hospitals as compared to clinics. Time from triage to clinician consultation, per triage category, were longer than the SATS guide times, although higher acuity cases were seen faster than lower acuity cases in a stepwise fashion. Red patients waited nearly an hour on average, with no significant difference between hospitals and CHCs. Orange patients had to wait one to two hours; this was significantly longer at hospitals. The mean time from assessment and management to a disposal decision for all-comers was significantly longer at hospitals as compared to CHCs across all priorities. Green patients took a lot longer at hospital compared to CHCs. A similar pattern was seen for the disposition decision to leaving time. The mean total time was significantly longer at hospitals as compared to clinics. Orange and yellow cases stayed significantly longer at hospitals as compared to CHCs; red and green cases also stayed longer at hospitals as compared to CHCs, though this was not significant. Red cases appeared to stay the longest at CHCs. Conclusions: Patients attending CHCs and hospitals are of similar illness acuity, despite policies dictating that sicker patients should be seen at hospitals not CHC level. CHCs have limited packages of care (decision making investigations, management options and expertise), and can only manage patients to a defined level. Thus, it takes longer for patients who are moderately or very ill to be seen and sorted in a CHC than a hospital, as at a CHC they are generally referred onwards to a hospital. Their journey through the EC will then begin again, so that for sicker patients the time spent in ECs in this study is underestimated. Models need to be explored so that patients receive care at point of contact as far as possible. Since CHC-based ECs see as many patients who are as ill as those in hospitals, these should have similar resources to hospitals, so that only those requiring definite admission need to be referred onwards. Point of care testing, bedside ultrasound, appropriate medications and EM skills should all be available at facilities closest to the patients with emergency conditions. Green patients, the lowest acuity, also take longer to be seen and sorted at hospitals versus CHCs, because investigations are available that are then done as an emergency versus outpatient basis. Efficient and timely outpatient appointments would help mitigate this.

Patient safety is founded on continuous learning because there is an urgent need to report and learn from errors, accidents, near misses, and adverse events. The traditional approach to patient safety, based on forming mortality committees and investigating accidents, will no longer be effective. Frameworks, surveys, and assessment tools have been developed over the last decade to assist organizations in measuring and understanding their culture. This a retrospective cross-sectional study included 67,010 respondents from Agency for Health care Research and Quality (AHRQ) 2018 comparative database was analyzed using partial least squares structural equation modeling (PLS-SEM). This research explored whether the dominant patient safety culture would impact the frequency of reported events and overall perceptions of patient safety. Furthermore, the study amid to examine whether respondents and hospital characteristics influence the perception of patient safety culture and the impact on healthcare staff. The results in this study showed that the perception of PSC positively influenced the overall perception of patient safety and frequency of event reporting. Moreover, the results revealed that hospital and respondents' characteristics (Staff Position, Teaching Status and Geographic Region) had varying influence on patient safety culture, overall perception of patient safety and frequency of event reporting.

Did governments in different countries regulate common concerns about patient safety differently? If so how and why did they do this? This thesis undertakes a historical comparison of the regulation of patient safety in Britain and Canada between 1980 and 2005. These jurisdictions began the period with very similar regulatory frameworks, but by 2005 there were distinct differences in each jurisdiction‘s regulatory response to patient safety. Britain was very actively regulating all aspects of service provision within its health system in the name of patient safety, whereas Canada‘s regulatory direction showed adherence to the 1980s model with only scattered incremental developments. This thesis assesses the broader sociopolitical context and the structure of the health systems in each jurisdiction and concludes there are differences in the logics of these systems that established a foundation for future regulatory divergence. It is argued that between 1980 and 2005 there were two factors that influenced regulatory directionality in each jurisdiction: changing political norms associated with the development of neoliberalism and the New Public Management; and events or scandals associated with the provision of health services. The differing levels of penetration of both the changing political norms into governance cultures and of scandals into the public and political consciousness are critical to explaining regulatory differences between jurisdictions. The thesis concludes that what and how governments chose to regulate is a function of the perceived need for action and the dominant social and political norms within that society. Context is everything in the formulation of regulatory approaches to address pressing social problems.

Yes / Purpose: To summarize studies that have examined patient safety culture (PSC) in maternity units and describe the different purposes, study designs and tools reported in these studies, whilst highlighting gaps in the literature. Methodology: Peer-reviewed studies published in English during 1961-2016 across eight electronic databases were subjected to a narrative literature review. Findings: Among 100 articles considered, 28 met the inclusion criteria. The main purposes for studying PSC were: (a) assessing intervention effects on PSC (n= 17); and (b) assessing PSC level (n=7). Patient safety culture was mostly assessed quantitatively using validated questionnaires (n=23). The Safety Attitude Questionnaire was the most commonly used questionnaire (n=17). Intervention varied from a single action lasting five weeks to a more comprehensive package lasting more than four years. The time between the baseline and the follow-up assessment varied from six months up to 24 months. No study reported measurement or intervention costs, and none incorporated the patient’s voice in assessing PSC. Practical Implications: Assessing PSC in maternity units is feasible using validated questionnaires. Interventions to enhance PSC have not been rigorously evaluated. Future studies should report PSC measurement costs, adopt more rigorous evaluation designs, and find ways to incorporate the patient’s voice. Originality/Value: This review summarized studies examining PSC in a highly important area and highlighted main limitations that future studies should consider.

Yes / Background: A recent initiative in hospital settings is the patient safety huddle (PSH): a brief multidisciplinary meeting held to highlight patient safety issues and actions to mitigate identified risks. Aim: The authors studied eight ward teams that had sustained PSHs for over 2 years in order to identify key contributory factors. Methods: Unannounced observations of the PSH on eight acute wards in one UK hospital were undertaken. Interviews and focus groups were also conducted. These were recorded and transcribed for framework analysis. Findings: A range of factors contributes to the sustainability of the PSH including a high degree of belief and consensus in purpose, adaptability, determination, multidisciplinary team involvement, a non-judgemental space, committed leadership and consistent reward and celebration. Conclusion: The huddles studied have developed and been shaped over time through a process of trial and error, and persistence. Overall this study offers insights into the factors that contribute to this sustainability.

This study examines the factors that influence how patients, families, and clinicians make decisions about risk-taking and safety in brain injury rehabilitation. Despite the importance of these decisions, particularly during transitions in care, there is scant literature to help guide these care partners in ethical and clinical decision-making related to risk-taking and safety. This study suggests that there are tensions between rehabilitation and patient safety efforts. Risk-taking lies at the core of brain injury rehabilitation; however, decisions about risk-taking are also influenced by conflicting values, system pressures, and patient abilities. A relational approach to autonomy that addresses patients’ decisional and functional abilities within their social contexts is more nuanced than a liberal individualist approach to autonomy, and provides a better framework for understanding decision-making. Relational autonomy may help clinicians make decisions that better balance risk-taking and safety, decisions that are committed to the principles of respecting autonomy and advancing safety.

This study examines the factors that influence how patients, families, and clinicians make decisions about risk-taking and safety in brain injury rehabilitation. Despite the importance of these decisions, particularly during transitions in care, there is scant literature to help guide these care partners in ethical and clinical decision-making related to risk-taking and safety. This study suggests that there are tensions between rehabilitation and patient safety efforts. Risk-taking lies at the core of brain injury rehabilitation; however, decisions about risk-taking are also influenced by conflicting values, system pressures, and patient abilities. A relational approach to autonomy that addresses patients’ decisional and functional abilities within their social contexts is more nuanced than a liberal individualist approach to autonomy, and provides a better framework for understanding decision-making. Relational autonomy may help clinicians make decisions that better balance risk-taking and safety, decisions that are committed to the principles of respecting autonomy and advancing safety.

Risk assessment is essential to ensure safety in hospitals. However, hospitals have paid little attention to risk assessment. Several problems have already been identified in the literature about current risk assessment practice, such as inadequate risk assessment guidance and bias in risk scoring. This research aimed to improve current risk assessment practice in hospitals in the National Health Service (NHS) in England. To address this aim, the research investigated current risk assessment practice and designed a new risk assessment approach by the use of mixed methods. One hundred hospitals’ risk assessment documents were reviewed to examine the current recommended risk assessment practice. Seventeen interviews and sixty-one questionnaires were conducted, a risk management system from a single hospital was reviewed, and strategic risks from thirty-four hospitals were reviewed, in order to examine how risks are assessed in actual practice. Following that, the proposed approach was designed by conducting requirements analysis and then evaluated by interviews and questionnaires with ten healthcare staff. The findings of this research reveal that hospitals conduct risk assessments in different ways (i.e. with a focus on individual patient-based, operational and strategic risks). There are also many problems involved in current risk assessment practice regarding both the foundations and use of risk assessment. For example, organisation-wide risk assessments predominantly rely on risk matrices which might lead to wrong risk prioritisation and resource allocation; and risks tend to reflect existing or past problems rather than being proactive. All these reveal a need to improve current risk assessment practice. This research makes an important contribution to the current understanding of risk assessment practice in hospitals by providing extensive evidence on both recommended and actual practice, and proposes a new risk assessment framework. The framework guides healthcare staff on how to conduct risk assessment in a more comprehensive way by encouraging its potential users to consider good risk assessment practice.

No / Patient safety is a major priority for health services. It is a multi-disciplinary problem and requires a multi-disciplinary solution; any education should therefore be a multi-disciplinary endeavour, from conception to implementation. The starting point should be at undergraduate level and medical education should not be an exception. It is apparent that current educational provision in patient safety lacks a systematic approach, is not linked to formal assessment and is detached from the reality of practice. If patient safety education is to be fit for purpose, it should link theory and the reality of practice; a human factors approach offers a framework to create this linkage. Learning outcomes should be competency based and generic content explicitly linked to specific patient safety content. Students should ultimately be able to demonstrate the impact of what they learn in improving their clinical performance. It is essential that the patient safety curriculum spans the entire undergraduate programme; we argue here for a spiral model incorporating innovative, multi-method assessment which examines knowledge, skills, attitudes and values. Students are increasingly learning from patient experiences, we advocate learning directly from patients wherever possible. Undergraduate provision should provide a platform for continuing education in patient safety, all of which should be subject to periodic evaluation with a particular emphasis on practice impact.