Can Patient Safety Incident Reports Be Used to Compare Hospital Safety? Results from a Quantitative Analysis of the English National Reporting and Learning System Data

BackgroundThere is an emerging interest in the inadvertent harm caused to patients by the provision of primary health-care services. To date (up to 2015), there has been limited research interest and few policy directives focused on patient safety in primary care. In 2003, a major investment was made in the National Reporting and Learning System to better understand patient safety incidents occurring in England and Wales. This is now the largest repository of patient safety incidents in the world. Over 40,000 safety incident reports have arisen from general practice. These have never been systematically analysed, and a key challenge to exploiting these data has been the largely unstructured, free-text data.AimsTo characterise the nature and range of incidents reported from general practice in England and Wales (2005–13) in order to identify the most frequent and most harmful patient safety incidents, and relevant contributory issues, to inform recommendations for improving the safety of primary care provision in key strategic areas.MethodsWe undertook a cross-sectional mixed-methods evaluation of general practice patient safety incident reports. We developed our own classification (coding) system using an iterative approach to describe the incident, contributory factors and incident outcomes. Exploratory data analysis methods with subsequent thematic analysis was undertaken to identify the most harmful and most frequent incident types, and the underlying contributory themes. The study team discussed quantitative and qualitative analyses, and vignette examples, to propose recommendations for practice.Main findingsWe have identified considerable variation in reporting culture across England and Wales between organisations. Two-thirds of all reports did not describe explicit reasons about why an incident occurred. Diagnosis- and assessment-related incidents described the highest proportion of harm to patients; over three-quarters of these reports (79%) described a harmful outcome, and half of the total reports described serious harm or death (n = 366, 50%). Nine hundred and ninety-six reports described serious harm or death of a patient. Four main contributory themes underpinned serious harm- and death-related incidents: (1) communication errors in the referral and discharge of patients; (2) physician decision-making; (3) unfamiliar symptom presentation and inadequate administration delaying cancer diagnoses; and (4) delayed management or mismanagement following failures to recognise signs of clinical (medical, surgical and mental health) deterioration.ConclusionsAlthough there are recognised limitations of safety-reporting system data, this study has generated hypotheses, through an inductive process, that now require development and testing through future research and improvement efforts in clinical practice. Cross-cutting priority recommendations include maximising opportunities to learn from patient safety incidents; building information technology infrastructure to enable details of all health-care encounters to be recorded in one system; developing and testing methods to identify and manage vulnerable patients at risk of deterioration, unscheduled hospital admission or readmission following discharge from hospital; and identifying ways patients, parents and carers can help prevent safety incidents. Further work must now involve a wider characterisation of reports contributed by the rest of the primary care disciplines (pharmacy, midwifery, health visiting, nursing and dentistry), include scoping reviews to identify interventions and improvement initiatives that address priority recommendations, and continue to advance the methods used to generate learning from safety reports.FundingThe National Institute for Health Research Health Services and Delivery Research programme.

Download Full-text

Learning from patient safety incidents involving acutely sick adults in hospital assessment units in England and Wales: a mixed methods analysis for quality improvement

Journal of the Royal Society of Medicine ◽

10.1177/01410768211032589 ◽

2021 ◽

pp. 014107682110325

Author(s):

Alexandra Urquhart ◽

Sarah Yardley ◽

Elin Thomas ◽

Liam Donaldson ◽

Andrew Carson-Stevens

Keyword(s):

Patient Safety ◽

Quality Improvement ◽

Mixed Methods ◽

Diagnostic Errors ◽

Learning System ◽

England And Wales ◽

Safety Incident ◽

Patient Safety Incident ◽

Incident Reports ◽

National Analysis

Objective Six per cent of hospital patients experience a patient safety incident, of which 12% result in severe/fatal outcomes. Acutely sick patients are at heightened risk. Our aim was to identify the most frequently reported incidents in acute medical units and their characteristics. Design Retrospective mixed methods methodology: (1) an a priori coding process, applying a multi-axial coding framework to incident reports; and, (2) a thematic interpretative analysis of reports. Setting Patient safety incident reports (10 years, 2005–2015) collected from the National Reporting and Learning System, which receives reports from hospitals and other care settings across England and Wales. Participants Reports describing severe harm/death in acute medical unit were identified. Main outcome measures Incident type, contributory factors, outcomes and level of harm were identified in the included reports. During thematic analysis, themes and metathemes were synthesised to inform priorities for quality improvement. Results A total of 377 reports of severe harm or death were confirmed. The most common incident types were diagnostic errors ( n = 79), medication-related errors ( n = 61), and failures monitoring patients ( n = 57). Incidents commonly stemmed from lack of active decision-making during patient admissions and communication failures between teams. Patients were at heightened risk of unsafe care during handovers and transfers of care. Metathemes included the necessity of patient self-advocacy and a lack of care coordination. Conclusion This 10-year national analysis of incident reports provides recommendations to improve patient safety including: introduction of electronic prescribing and monitoring systems; forcing checklists to reduce diagnostic errors; and increased senior presence overnight and at weekends.

Download Full-text

The COVID‐19 outbreak in Saudi Arabia and the impact on patient safety incident reports: An empirical study among the medical facilities of Qassim health cluster

Journal of Healthcare Risk Management ◽

10.1002/jhrm.21464 ◽

2021 ◽

Author(s):

Sultan Al‐Shaya ◽

Ayed Al‐Reshidi ◽

Majeda Farajat ◽

Aliaa Elnefiely

Keyword(s):

Saudi Arabia ◽

Patient Safety ◽

Empirical Study ◽

Safety Incident ◽

Medical Facilities ◽

Patient Safety Incident ◽

Incident Reports ◽

The Impact

Download Full-text

Patient safety incidents in advance care planning for serious illness: a mixed-methods analysis

BMJ Supportive & Palliative Care ◽

10.1136/bmjspcare-2019-001824 ◽

2019 ◽

pp. bmjspcare-2019-001824

Author(s):

Toby Dinnen ◽

Huw Williams ◽

Sarah Yardley ◽

Simon Noble ◽

Adrian Edwards ◽

...

Keyword(s):

Patient Safety ◽

Mixed Methods ◽

Advance Care Planning ◽

Learning System ◽

Care Planning ◽

Safety Incident ◽

Patient Safety Incident ◽

Advance Care ◽

National Analysis ◽

Patient Safety Incidents

ObjectivesAdvance care planning (ACP) is essential for patient-centred care in the last phase of life. There is little evidence available on the safety of ACP. This study characterises and explores patient safety incidents arising from ACP processes in the last phase of life.MethodsThe National Reporting and Learning System collates patient safety incident reports across England and Wales. We performed a keyword search and manual review to identify relevant reports, April 2005–December 2015. Mixed-methods, combining structured data coding, exploratory and thematic analyses were undertaken to describe incidents, underlying causes and outcomes, and identify areas for improvement.ResultsWe identified 70 reports in which ACP caused a patient safety incident across three error categories: (1) ACP not completed despite being appropriate (23%, n=16). (2) ACP completed but not accessible or miscommunicated between professionals (40%, n=28). (3) ACP completed and accessible but not followed (37%, n=26). Themes included staff lacking the knowledge, confidence, competence or belief in trustworthiness of prior documentation to create or enact ACP. Adverse outcomes included cardiopulmonary resuscitation attempts contrary to ACP, other inappropriate treatment and/or transfer or admission.ConclusionThis national analysis identifies priority concerns and questions whether it is possible to develop strong system interventions to ensure safety and quality in ACP without significant improvement in human-dependent issues in social programmes such as ACP. Human-dependent issues (ie, varying patient, carer and professional understanding, and confidence in enacting prior ACP when required) should be explored in local contexts alongside systems development for ACP documentation.

Download Full-text

An Analysis of Patient Safety Incident Reports Associated with Electronic Health Record Interoperability

Applied Clinical Informatics ◽

10.4338/aci-2017-01-ra-0014 ◽

2017 ◽

Vol 08 (02) ◽

pp. 593-602 ◽

Cited By ~ 11

Author(s):

Katharine Adams ◽

Jessica Howe ◽

Allan Fong ◽

Joseph Puthumana ◽

Kathryn Kellogg ◽

...

Keyword(s):

Patient Safety ◽

Electronic Health Record ◽

Health Record ◽

Health It ◽

Specific Patient ◽

It Systems ◽

Safety Incident ◽

Patient Safety Incident ◽

Incident Reports ◽

Electronic Health

SummaryBackground: With the widespread use of electronic health records (EHRs) for many clinical tasks, interoperability with other health information technology (health IT) is critical for the effective delivery of care. While it is generally recognized that poor interoperability negatively impacts patient care, little is known about the specific patient safety implications. Understanding the patient safety implications will help prioritize interoperability efforts around architectures and standards.Objectives: Our objectives were to (1) identify patient safety incident reports that reflect EHR interoperability challenges with other health IT, and (2) perform a detailed analysis of these reports to understand the health IT systems involved, the clinical care processes impacted, whether the incident occurred within or between provider organizations, and the reported severity of the patient safety events.Methods: From a database of 1.735 million patient safety event (PSE) reports spanning multiple provider organizations, 2625 reports that were indicated as being health IT related by the event reporter were reviewed to identify EHR interoperability related reports. Through a rigorous coding process 209 EHR interoperability related events were identified and coded.Results: The majority of EHR interoperability PSE reports involved interfacing with pharmacy systems (i.e. medication related), followed by laboratory, and radiology. Most of the interoperability challenges in these clinical areas were associated with the EHR receiving information from other health IT systems as opposed to the EHR sending information to other systems. The majority of EHR interoperability challenges were within a provider organization and while many of the safety events reached the patient, only a few resulted in patient harm.Conclusions: Interoperability efforts should prioritize systems in pharmacy, laboratory, and radiology. Providers should recognize the need to improve EHRs interfacing with other health IT systems within their own organization.Citation: Adams KT, Howe JL, Fong A, Puthumana JS, Kellogg KM, Gaunt M, Ratwani RM. An analysis of patient safety incident reports associated with electronic health record interoperability. Appl Clin Inform 2017; 8: 593–602 https://doi.org/10.4338/ACI-2017-01-RA-0014

Download Full-text

Using multiclass classification to automate the identification of patient safety incident reports by type and severity

BMC Medical Informatics and Decision Making ◽

10.1186/s12911-017-0483-8 ◽

2017 ◽

Vol 17 (1) ◽

Cited By ~ 19

Author(s):

Ying Wang ◽

Enrico Coiera ◽

William Runciman ◽

Farah Magrabi

Keyword(s):

Patient Safety ◽

Multiclass Classification ◽

Safety Incident ◽

Patient Safety Incident ◽

Incident Reports

Download Full-text

Analysis of paediatric long-term ventilation incidents in the community

Archives of Disease in Childhood ◽

10.1136/archdischild-2019-317965 ◽

2019 ◽

Vol 105 (5) ◽

pp. 446-451

Author(s):

Rasanat Fatima Nawaz ◽

Bethan Page ◽

Emily Harrop ◽

Charles A Vincent

Keyword(s):

Patient Safety ◽

Learning System ◽

Free Text ◽

Delivery Of Care ◽

Contributory Factors ◽

Safety Incident ◽

Provision Of Services ◽

Patient Safety Incident ◽

Long Term Ventilation

AimTo describe the nature and causes of reported patient safety incidents relating to care in the community for children dependent on long-term ventilation with the further aim of improving safety.MethodsWe undertook an analysis of patient safety incident data relating to long-term ventilation in the community using incident reports from England and Wales’ National Reporting and Learning System occurring between January 2013 and December 2017. Manual screening by two authors identified 220 incidents which met the inclusion criteria. The free text for each report was descriptively analysed to identify the problems in the delivery of care, the contributory factors and the patient outcome.ResultsCommon problems in the delivery of care included issues with faulty equipment and the availability of equipment, and concerns around staff competency. There was a clearly stated harm to the child in 89 incidents (40%). Contributory factors included staff shortages, out of hours care, and issues with packaging and instructions for equipment.ConclusionsThis study identifies a range of problems relating to long-term ventilation in the community, some of which raise serious safety concerns. The provision of services to support children on long-term ventilation and their families needs to improve. Priorities include training of staff, maintenance and availability of equipment, support for families and coordination of care.

Download Full-text

Using convolutional neural networks to identify patient safety incident reports by type and severity

Journal of the American Medical Informatics Association ◽

10.1093/jamia/ocz146 ◽

2019 ◽

Vol 26 (12) ◽

pp. 1600-1608 ◽

Cited By ~ 3

Author(s):

Ying Wang ◽

Enrico Coiera ◽

Farah Magrabi

Keyword(s):

Patient Safety ◽

Test Data ◽

Reporting System ◽

Word Embedding ◽

Support Vector ◽

Data Sets ◽

Safety Incident ◽

Patient Safety Incident ◽

Incident Reports ◽

Severity Levels

Abstract Objective To evaluate the feasibility of a convolutional neural network (CNN) with word embedding to identify the type and severity of patient safety incident reports. Materials and Methods A CNN with word embedding was applied to identify 10 incident types and 4 severity levels. Model training and validation used data sets (n_type = 2860, n_severity = 1160) collected from a statewide incident reporting system. Generalizability was evaluated using an independent hospital-level reporting system. CNN architectures were examined by varying layer size and hyperparameters. Performance was evaluated by F score, precision, recall, and compared to binary support vector machine (SVM) ensembles on 3 testing data sets (type/severity: n_benchmark = 286/116, n_original = 444/4837, n_independent = 6000/5950). Results A CNN with 6 layers was the most effective architecture, outperforming SVMs with better generalizability to identify incidents by type and severity. The CNN achieved high F scores (> 85%) across all test data sets when identifying common incident types including falls, medications, pressure injury, and aggression. When identifying common severity levels (medium/low), CNN outperformed SVMs, improving F scores by 11.9%–45.1% across all 3 test data sets. Discussion Automated identification of incident reports using machine learning is challenging because of a lack of large labelled training data sets and the unbalanced distribution of incident classes. The standard classification strategy is to build multiple binary classifiers and pool their predictions. CNNs can extract hierarchical features and assist in addressing class imbalance, which may explain their success in identifying incident report types. Conclusion A CNN with word embedding was effective in identifying incidents by type and severity, providing better generalizability than SVMs.

Download Full-text