Language barriers between healthcare professionals and their patients remain a persistent challenge. Patients with limited proficiency in the primary language of the country where they receive care face higher risks of adverse events, misdiagnosis and unplanned readmissions.1 2 In linguistically diverse countries, services often fall short of meeting the needs of patients who speak minority languages. This leads to inequities in care across inpatient, outpatient and emergency settings. While concern for language discordance in healthcare is by no means a novel development, guidelines have rarely progressed beyond recommending implementation of professional interpreter services.3 In-person interpretation is generally considered the gold standard for addressing language barriers during direct care delivery4 5 and is in line with regulatory and ethical standards.6 Other modalities of interpretation, including telephone and video, are alternative options, although the feasibility of providing timely written...

Introduction

Machine translation of patient-specific information could mitigate language barriers if sufficiently accurate and non-harmful and may be particularly useful in healthcare encounters when professional translators are not readily available. We evaluated the translation accuracy and potential for harm of ChatGPT-4 and Google Translate in translating from English to Spanish, Chinese and Russian.

Methods

We used ChatGPT-4 and Google Translate to translate 50 sets (316 sentences) of deidentified, patient-specific, clinician free-text emergency department instructions into Spanish, Chinese and Russian. These were then back-translated into English by professional translators and double-coded by physicians for accuracy and potential for clinical harm.

Results

At the sentence level, we found that both tools were ≥90% accurate in translating English to Spanish (accuracy: GPT 97%, Google Translate 96%) and English to Chinese (accuracy: GPT 95%; Google Translate 90%); neither tool performed as well in translating English to Russian (accuracy: GPT 89%; Google Translate 80%). At the instruction set level, 16%, 24% and 56% of Spanish, Chinese and Russian GPT-translated instruction sets contained at least one inaccuracy. For Google Translate, 24%, 56% and 66% of Spanish, Chinese and Russian translations contained at least one inaccuracy. The potential for harm due to inaccurate translations was ≤1% for both tools in all languages at the sentence level and ≤6% at the instruction set level. GPT was significantly more accurate than Google Translate in Chinese and Russian at the sentence level; the potential for harm was similar.

Conclusion

These results support the potential of machine translation tools to mitigate gaps in translation services for low-stakes written communication from English to Spanish, while also strengthening the case for caution and for professional oversight in non-low-risk communication. Further research is needed to evaluate machine translation for other languages and more technical content.

Background

Healthcare patient safety investigations inappropriately focus on individual culpability and the target of recommendations is often on the behaviours of individuals, rather than addressing latent failures of the system. The aim of this study was to explore whether outcome bias might provide some explanation for this. Outcome bias occurs when the ultimate outcome of a past event is given excessive weight, in comparison to other information, when judging the preceding actions or decisions.

Methods

We conducted a survey in which participants were each presented with three incident scenarios, followed by the findings of an investigation. The scenarios remained the same, but the patient outcome was manipulated. Participants were recruited via social media and we examined three groups (general public, healthcare staff and experts) and those with previous incident involvement. Participants were asked about staff responsibility, avoidability, importance of investigating and to select up to five recommendations to prevent recurrence. Summary statistics and multilevel modelling were used to examine the association between patient outcome and the above measures.

Results

212 participants completed the online survey. Worsening patient outcome was associated with increased judgements of staff responsibility for causing the incident as well as greater motivation to investigate. More participants selected punitive recommendations when patient outcome was worse. While avoidability did not appear to be associated with patient outcome, ratings were high suggesting participants always considered incidents to be highly avoidable. Those with patient safety expertise demonstrated these associations but to a lesser extent, when compared with other participants. We discuss important comparisons between the participant groups as well as those with previous incident involvement, as victim or staff member.

Interpretation

Outcome bias has a significant impact on judgements following incidents and investigations and may contribute to the continued focus on individual culpability and individual focused recommendations observed following investigations.

Background

Natural hazards, such as earthquakes, pose a significant risk to both the public and healthcare professionals, jeopardising patient safety due to the disruption of healthcare systems and services. This study aimed to explore the lived experiences of healthcare professionals concerning patient safety during natural hazards, specifically earthquakes.

Methods

Employing a descriptive phenomenological approach, the study followed the Consolidated Criteria for Reporting Qualitative Research guidelines. 23 participants, including doctors, nurses and paramedics, were interviewed using purposive sampling. Data were gathered through semistructured interviews, which were audio recorded and transcribed. Ethical approval was obtained, and Colaizzi’s method was used for data analysis, with findings validated through researcher consensus and participant feedback.

Results

Nine overarching themes emerged, such as the emotional toll of communication breakdowns, struggles with patient identification, stress due to resource scarcity, operational chaos, ethical dilemmas and psychological impacts on both patients and staff. The study found that these factors collectively influenced patient safety during the earthquake.

Conclusion

The emotional strain caused by communication failures, patient identification issues and resource shortages compounded the challenges of providing safe care during the earthquake. Strengthening disaster preparedness through improved communication systems, resource management, psychological support, interagency coordination and regular realistic disaster drills is essential for safeguarding patient safety in future disasters.

Importance

Patients with poor glycaemic control have a high risk for major cardiovascular events. Improving glycaemic monitoring in patients with diabetes can improve morbidity and mortality.

Objective

To assess the effectiveness of a patient portal message in prompting patients with poorly controlled diabetes without a recent glycated haemoglobin (HbA1c) result to have their HbA1c repeated.

Design

A pragmatic, randomised clinical trial.

Setting

A large academic health system consisting of over 350 ambulatory practices.

Participants

Patients who had an HbA1c greater than 10% who had not had a repeat HbA1c in the prior 6 months.

Exposures

A single electronic health record (EHR)-based patient portal message to prompt patients to have a repeat HbA1c test versus usual care.

Main outcomes

The primary outcome was a follow-up HbA1c test result within 90 days of randomisation.

Results

The study included 2573 patients with a mean (SD) HbA1c of 11.2%. Among 1317 patients in the intervention group, 24.2% had follow-up HbA1c tests completed within 90 days, versus 21.1% of 1256 patients in the control group (p=0.07). Patients in the intervention group were more likely to log into the patient portal within 60 days as compared with the control group (61.2% vs 52.3%, p<0.001).

Conclusions

Among patients with poorly controlled diabetes and no recent HbA1c result, a brief patient portal message did not significantly increase follow-up testing but did increase patient engagement with the patient portal. Automated patient messages could be considered as a part of multipronged efforts to involve patients in their diabetes care.

Objective

Overconfidence is an important source of medical error. This review analyses experimental studies of confidence in medical diagnosis to identify factors affecting clinicians’ confidence in their diagnoses and how confidence impacts patient care.

Method

A scoping review of medical and psychological literature was conducted. Articles were categorised according to methodology and clinical specialty. Findings were analysed thematically. Our review methodology adheres to the JBI’s Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews checklist.

Data sources

We searched SCOPUS, MEDLINE, PsycINFO and Global Health. We then performed citation tracking within these papers’ references to identify additional articles.

Eligibility criteria

Papers were included if they reported quantitative results from an empirical study in which participants reported their confidence or certainty during a diagnostic decision. Studies comprised several medical subdisciplines.

Results

77 articles met the inclusion criteria. Across these articles, confidence was not found to be well-calibrated to true diagnostic accuracy regardless of clinician experience. We organised articles under two main themes: the determinants of confidence and the uses of confidence during the patient’s care pathway. Confidence is found to be affected by several factors, including case complexity, early diagnostic differentials and the healthcare environment. Factors that affect confidence, but not accuracy, demonstrate how the two can become decoupled, resulting in overconfidence/underconfidence. Confidence is found to affect patient testing, medication administration and referral rates, among other clinical actions.

Conclusions

Improving the calibration of confidence should be a priority for medical education and clinical practice (eg, via decision aids). We propose a theoretical model of factors that affect diagnostic confidence/certainty. Such a model can inform future work on how appropriate diagnostic confidence can be prompted and communicated among clinicians.

Statistical process control (SPC) charts provide a natural approach to analysing time series data for healthcare quality improvement (QI) initiatives. A problem arising in practice is that having established baseline control limits, there is no accepted objective and transparent approach to deciding when to establish new control limits for a given chart. We present the Stable Shift Algorithm, a new algorithm to aid analysts by identifying when control limits should be re-established, partitioning a control chart of time series data into distinct time periods. The algorithm aims to achieve this while (1) using only the theory of SPC, familiar to many QI practitioners, (2) avoiding re-establishing limits prematurely and (3) remaining flexible to choice of basic parameters of typical control chart use in QI. This is achieved through the commonly used shift rule of control charts, applied to establish whether shifts warrant new control limits or not. We conducted a simulation study to evaluate the effectiveness of the algorithm in achieving its aims, and a case study demonstrating application of the algorithm to 557 time series of accident and emergency care measures for providers in England and Scotland. Simulation results show that the algorithm avoids premature re-establishment of limits more often than simply re-establishing at every shift rule break. Application of the algorithm to the accident and emergency measures demonstrates this is not achieved at the cost of excessive additional rule breaks that might indicate control limits do not represent the underlying process. The Stable Shift Algorithm offers a potentially highly valuable tool for QI practitioners and researchers undertaking SPC analyses, providing an automated, consistent and rigorous approach facilitating large-scale analyses.

Introduction

The role of quality and safety professionals and leaders in realising the potential and managing the risks of artificial intelligence (AI) tools has not been well defined. We suggest these leaders focus on three areas: using quality, safety and implementation sciences to increase the likelihood of beneficial AI adoption; using AI to enhance and support the methods of quality and safety management; and serving as experts and champions for AI tool use that promotes health and equity (figure 1).

Background and approach

We used 90-day research and development cycles1 to examine AI topics and the role of quality and safety leaders with respect to AI integration, including applying AI tools for quality improvement (QI) use cases,2 the broad application of AI in quality management systems, the implications of AI for patient safety3 and the use of AI...

Despite decades of progress in global child health, neonatal mortality remains high, accounting for nearly half of all under-five deaths worldwide.1 Most of these deaths occur in low- and middle-income countries and are preventable with timely, high-quality care for small and sick newborns.2 The WHO has called for every newborn to receive essential, high-impact interventions,3 yet the challenge lies not only in knowing what works, but in implementing those interventions at scale, with quality, and within real-world health systems. Quality improvement (QI) and implementation science (IS) offer complementary strategies to address this challenge. QI focuses on local, iterative problem solving to adapt and improve evidence-based or locally generated care processes,4 5 while IS provides structured, theory-driven methods to promote their uptake and sustainability.6 7 Yet too often, these fields operate independently rather than in a...

The number of complaints received by healthcare organisations from patients and families is on an upward trajectory.1 For example, in 2023–2024, the NHS in England received 241 922 complaints,2 an increase of 5% on the previous year and 37% since 2013–2014. Moreover, while relatively few NHS patient encounters result in a formal complaint (approximately 0.4%), just 9% of patients who report poor healthcare experiences actually submit one.3

Although the motivation for complainants can vary—for instance, some patients seek redress, and others want resolution of ongoing problems—they nearly always request organisational learning.4 Furthermore, while complaints can be incorrect or ill-intentioned, leading to concerns about their validity,5 the collective scale of the information they provide is hard to dismiss. They are, in effect, a massive rolling compendium of ethnographies from patients and families at the sharp end of treatment delivery, revealing perceived...

Background

Improving small and sick newborn care (SSNC) is crucial in resource-limited settings. Newborn Essential Solutions and Technologies (NEST360), a multicountry alliance, aims to reduce newborn mortality through evidence-based interventions. NEST360 developed a multipronged approach to improving quality. We use implementation research (IR) to describe this approach and report emerging implementation outcomes.

Methods

The implementation research logic model (IRLM) was applied to link contextual factors, implementation strategies, mechanisms and implementation outcomes, capturing successes and challenges of the improving quality approach. Data sources included programme data, peer-reviewed publications and team input. Contextual factors were organised by the NEST360-UNICEF SSNC implementation toolkit. Strategies were grouped by the Expert Recommendations for Implementation Change list, and implementation outcomes were measured using Proctor’s implementation outcomes.

Results

We developed an IRLM to describe the implementation of NEST360’s improving quality model. This IRLM included 33 contextual factors; 42% were barriers, 42% were facilitators, and 15% were both a barrier and facilitator. Additionally, we identified 10 implementation strategies that NEST360 used. The logic model also describes the connections between the contextual factors, the strategies that address them, and the preliminary implementation outcomes. Examples of the outcomes measured include Reach with 100% of units logging into the NEST360-Implementation Tracker (NEST-IT) at least once (October 2023 to March 2024), Adoption with 100% of units conducting a quality improvement (QI) project (April 2024 to June 2024), and Feasibility with 93% of units reporting NEST-IT data in their QI project documentation (April 2024 to June 2024). Finally, this study identified sustainability strategies as a critical need.

Conclusions

Integrating IR and QI enhances SSNC in resource-limited settings. Addressing barriers, leveraging facilitators and using structured IR frameworks advanced QI efforts, thereby improving intervention reach, adoption and feasibility while building scalable systems for high-quality healthcare.

Background

Since 2015, the National Institute for Health and Care Excellence (NICE) guidelines have recommended antenatal magnesium sulfate (MgSO4) for mothers in preterm labour (<30 weeks’ gestation) to reduce the risk of cerebral palsy (CP) in the preterm baby. However, the implementation of this guideline in clinical practice was slow, and MgSO4 use varied between maternity units. In 2018, the PRrevention of Cerebral palsy in PreTerm labour (PReCePT) programme, an evidence-based quality improvement (QI) intervention to improve use of MgSO4, was rolled out across England. Earlier evaluation found this programme to be effective and cost-effective over the first 12 months. We extended the original evaluation to determine the programme’s longer-term impact over 4 years, its impact in later preterm births, the impact of the COVID-19 pandemic, and to compare MgSO4 use in England (where PReCePT was implemented) to Scotland and Wales (where it was not).

Methods

Quasi-experimental longitudinal study using data from the National Neonatal Research Database on babies born <30 weeks’ gestation and admitted to a National Health Service neonatal unit. Primary outcome was the percentage of eligible mothers receiving MgSO4, aggregated to the national level. Impact of PReCePT on MgSO4 use was estimated using multivariable linear regression. The net monetary benefit (NMB) of the programme was estimated.

Results

MgSO4 administration rose from 65.8% in 2017 to 85.5% in 2022 in England. PReCePT was associated with a 5.8 percentage points improvement in uptake (95% CI 2.69 to 8.86, p<0.001). Improvement was greater when including older preterm births (<34 weeks’ gestation, 8.67 percentage points, 95% CI 6.38 to 10.96, p<0.001). Most gains occurred in the first 2 years following implementation. PReCePT had a NMB of £597 000 with 89% probability of being cost-effective. Following implementation, English uptake appeared to accelerate compared with Scotland and Wales. There was some decline in use coinciding with the onset of the pandemic.

Conclusions

The PReCePT QI programme cost-effectively improved use of antenatal MgSO4, with anticipated benefits to the babies who have been protected from CP.

Background

The Child Opportunity Index (COI) is a multidimensional measure of US neighbourhood-level conditions needed for healthy development. COI is associated with healthcare delivery and outcomes. Formal quality improvement (QI) may influence the relationship between COI, quality of care and outcomes in children.

Objective

To assess the association between COI and paediatric sepsis care delivery and outcomes and determine if baseline disparities in care change over time among hospitals in the Improving Pediatric Sepsis Outcomes (IPSO) collaborative.

Methods

Retrospective cohort study of IPSO patients probabilistically linked to the Pediatric Health Information System database from 2017 to 2021. Primary exposure was COI. We estimated differences in the proportions of patients in each COI quintile identified via standardised sepsis recognition protocols (screening tool, huddle documentation and/or order set use) and who received a bundle of recommended care (standardised sepsis recognition, plus bolus <1 hour and antibiotic <3 hours). We further assessed the timeliness of each bundle component and mortality. We evaluated changes in standardised sepsis recognition over time using generalised linear models.

Results

31 260 sepsis cases from 24 hospitals were included. Cross-sectional analysis over the entire study period found patients in the Very High COI quintile were most likely to be identified via standardised recognition protocols and receive IPSO’s recommended care bundle (67.7% and 46%, respectively). Over time, standardised sepsis recognition improved for all; the greatest improvements were among inpatients in the Very Low COI quintile.

Conclusion

Disparities exist in paediatric sepsis care delivery by COI. Over the course of the IPSO collaborative, care improved most for children in the lowest COI quintile. QI collaboratives focused on standardisation and shared learning may reduce disparities.

Background

Regulation plays a central role in health and social care systems, particularly in ensuring quality, safety and accountability. However, there is limited understanding of how organisations effectively implement and adhere to these regulatory requirements. In particular, little is known about how providers of residential care facilities for people with disabilities (RCF-D) navigate and apply statutory care regulations.

Methods

We conducted semistructured interviews with managers of RCF-D. Participant recruitment followed a purposive maximum variation sampling approach. 19 participants were interviewed, representing 22 RCF-D and 16 provider organisations. Interview data were analysed using a mixed deductive–inductive approach.

Results

Most managers were supportive of regulatory goals, creating a more favourable environment for successful implementation. By making sense of regulatory requirements and sharing insights across their organisations, managers facilitated smoother implementation. Crucially, building strong internal and external networks played a pivotal role in driving success. Collaborative relationships with inspectors, centred on a shared commitment to improving residents’ lives, further strengthened the implementation process.

Conclusion

Managers of RCF-D devised a range of strategies to manage compliance, balancing regulatory demands with problem-solving and relationship-building. These efforts were supported by a collaborative approach to working with inspectors, which fostered a shared commitment to improving residents’ lives. Our findings offer practical guidance for organisations seeking to improve regulatory compliance through effective relationship management and resource alignment. Future research could investigate how framing regulation as an adaptive intervention could further enhance implementation and sustain compliance.

Background

There is increasing interest in how patient complaint material can be used to highlight areas requiring quality improvement (QI) in healthcare. However, knowledge of using complaint material to initiate or monitor QI is limited.

Objectives

This review explored the use of complaint material in QI by identifying problems related to substandard care that were addressed by QI initiatives, exploring how complaint material was used before or after a QI initiative, and mapping changes in complaint material after QI initiatives.

Methods

This scoping review followed the Joanna Briggs Institute methodology and adhered to the Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews reporting guideline.

Eligibility criteria

Studies were included if a QI initiative was initiated or monitored using complaint material. Eligible designs included observational studies, QI projects, pre-intervention and post-intervention studies and randomised controlled trials. Audio, online and symptom-based complaints were excluded.

Information sources

A systematic search was conducted on 10 December 2024 in Embase, Medline, CINAHL and Web of Science, and additional sources, with no language or date limitations.

Synthesis of results

Substandard problems targeted by QI initiatives were categorised using the Healthcare Complaints Analysis Tool by two independent coders. Findings were synthesised narratively and summarised using frequency analyses where applicable.

Results

We identified 58 QI initiatives, most frequently targeting safety (n=39). Before QI, complaint material was usually analysed through review (n=19), counts (n=17), content categorisation (n=9) or root cause analysis (n=2). After QI, analyses included counts (n=34), rates (n=20), content categorisation (n=7) and review (n=4). Reviewing or categorisation methods were often unspecified. Among studies using complaints as an outcome, most reported complaint reductions (n=43), while a few reported increases (n=2) or mixed results (n=4).

Discussion

The QI initiatives primarily targeted patient safety and applied simple quantitative analyses. Some studies relied on reviews or categorisations without reporting the validation or reliability of the used tools. Improved reporting standards are needed to strengthen learning. Furthermore, while QI initiatives appear to have the potential to change complaint patterns, this finding should be interpreted with caution, as this is based on a scoping review.

Other

Preregistered protocol: https://osf.io/6g4qw.

Introduction

Quality improvement (QI) and implementation science (IS) are distinct, yet related fields that aspire to improve the quality of healthcare for all people. QI is a systematic, data-driven approach to continuous problem solving in healthcare, with a focus on enhancing the efficiency, quality and safety of care delivery.1 Emerging in the mid-to-late twentieth century, QI identifies and analyses local problems, develops and implements targeted structural and process-focused solutions, and evaluates their outcomes. Separately, IS emerged in the late twentieth century due to growing awareness of the need to bridge the gap between research findings and their application in real-world settings. IS, defined as the study of methods to promote the systematic uptake of research findings and other evidence-based practices into routine healthcare and public health settings, emphasises improving healthcare processes and outcomes.2 While the disciplines share common goals, the fields diverge with respect...

More than 20 years since a landmark study1 documented hospitalised patients were more likely to die when their nurse cared for too many patients at a time, hundreds of rigorously conducted studies in over 30 countries have documented a relationship between nurse understaffing and poorer outcomes of all kinds, including preventable patient deaths and avoidable burnout of nurses.2–4 Despite the empirical evidence, chronic hospital nurse understaffing persists.

Why has the research evidence not substantively transformed hospital staffing practices? One possible explanation is that the benefits of eliminating nurse understaffing accrue to patients and nurses, while the costs of staffing more nurses accrue to hospitals.

Hospitals are the most expensive healthcare setting, largely because of the intensive nursing care that hospitalised patients require. Indeed, if patients can have procedures and treatments administered outside of hospitals (eg, outpatient offices, home care), they do,...

BMJ is concerned about the consent obtained from veterans to use their personal information in this research paper1 and broader Veterans' Medicines Advice and Therapeutics Education Services (MATES) program. BMJ was contacted by veterans who asked the journal to retract the content on this basis.

 The MATES program was operated by the Department for Veterans Affairs (DVA), and the research was conducted by the University of South Australia (UniSA).2 DVA was responsible for obtaining consent from the veterans to use their personal information in the MATES program; they were also responsible for managing opt-out requests.

 In 2018, a veteran lodged a complaint with the Australian Information and Privacy Commissioner about the validity of consent for DVA’s collection and use of their personal information in the MATES programme. In 2023, the Privacy Commissioner determined that DVA had breached an Australian privacy principle by using and disclosing...

Since the 18th century, bedside rounds have been a fundamental component of clinical care, serving as a setting where clinical information is gathered, processed and shared.1 This tradition highlights the importance of maintaining a high level of structure during clinical encounters. Over time, structured tools to guide care have been widely adopted across multiple specialties.2–6 Systematic checklists, in particular, have become the most used form of structured intervention in bedside wards to enhance patient care and safety.7–9 This intervention has been associated with improved non-clinical outcomes, such as communication and adherence to standard protocols.7–9 However, their impact on clinical outcomes remains a matter of debate.10

A 2014 systematic review found that safety checklists improved team communication, improved adherence to standards and reduced adverse...

Background

Understaffing by nursing staff in hospitals is linked to patients coming to harm and dying unnecessarily. There is a vicious cycle whereby poor work conditions, including understaffing, can lead to nursing vacancies, which in turn leads to further understaffing. Is hospital investment in nursing staff, to eliminate understaffing on wards, cost-effective?

Methods

This longitudinal observational study analysed data on 185 adult acute units in four hospital Trusts in England over a 5-year period. We modelled the association between a patient’s exposure to ward nurse understaffing (days where staffing was below the ward mean) over the first 5 days of stay and risk of death, risk of readmission and length of stay, using survival analysis and linear mixed models. We estimated the incremental cost-effectiveness of eliminating understaffing by registered nurses (RN) and nursing support (NS) staff, estimating net costs per quality-adjusted life year (QALY). We took a hospital cost perspective.

Findings

Exposure to RN understaffing is associated with increased hazard of death (adjusted HR (aHR) 1.079, 95% CI 1.070 to 1.089), increased chance of readmission (aHR 1.010, 95% CI 1.005 to 1.016) and increased length of stay (ratio 1.687, 95% CI 1.666 to 1.707), while exposure to NS understaffing is associated with smaller increases in hazard of death (aHR 1.072, 95% CI 1.062 to 1.081) and length of stay (ratio 1.608, 95% CI 1.589 to 1.627) but reduced readmissions (aHR 0.994, 95% CI 0.988 to 0.999). Eliminating both RN and NS understaffing is estimated to cost £2778 per QALY (staff costs only), £2685 (including benefits of reduced staff sickness and readmissions) or save £4728 (including benefits of reduced lengths of stay). Using agency staff to eliminate understaffing is less cost-effective and would save fewer lives than using permanent members of staff. Targeting specific patient groups with improved staffing would save fewer lives and, in the scenarios tested, cost more per QALY than eliminating all understaffing.

Interpretation

Rectifying understaffing on inpatient wards is crucial to reduce length of stay, readmissions and deaths. According to the National Institute for Health and Care Excellence £10 000 per QALY threshold, it is cost-effective to eliminate understaffing by nursing staff. This research points towards investing in RNs over NS staff and permanent over temporary workers. Targeting particular patient groups would benefit fewer patients and is less cost-effective.