‘Safety netting’ refers to a range of activities to manage clinical uncertainty during consultations. This can include uncertainty about diagnosis, how disease may progress or whether or not a treatment will work. It is frequently delivered along the lines of ‘come back and see me if this does not get better’, but can also include more specific advice and steps to monitor the patient through lists or codes in the patient record. The desired effect is for patients to seek medical attention again if needed and to be reassured if not. Ten years of safety netting research has established that safety netting is ubiquitous and inconsistently practised. It is invariably delivered with missing details about how quickly symptoms may resolve or reappear, and what the implication would be.1 Requests to ‘come back’ are often made without clear instructions for how and when to do so.1...

Dealing with uncertainty is an inherent part of scientific discovery. One of the ways in which scientists have tried to overcome uncertainty is through the concept of measurement—defined as the act or the process of finding the size, quantity or degree of something.1 Over centuries, standardised and consistent measurement systems have assumed fundamental importance in societies in parallel with the increasing dominance of the scientific paradigm. The 11th century proposal by Henry I of England to standardise the measurement of a yard as the distance from his nose to outstretched thumb2 may appear egotistical and humorous by modern standards. But the reader does not need to stray far outside of the scientific paradigm to appreciate the range of philosophical standpoints on the nature of measurement and the ongoing debate over the concept of measurable quantity.3 4 Examples of outcomes that cannot...

Our population is ageing and with increased age, comes more frequent presentation of people living with multiple long-term conditions (MLTCs), who are likely to experience polypharmacy and the risk that accompanies taking multiple medications. These risks disproportionately affect the most socioeconomically deprived people in our communities, including those from minority ethnic groups, among whom the incidence of MLTCs is higher1 and acquired at an earlier age.2 Polypharmacy increases the prevalence of errors relating to the medicines use process, defined as including prescribing, dispensing, administration and monitoring of medication. Estimates in England report around 237 million medication errors per annum with avoidable errors costing around £98 million each year resulting in 1700 deaths.3 In the USA, around 7 million people per year are affected by medication errors at a cost of around US$21 billion.4 While not all of these errors are clinically significant,...

Background

Safety-netting is intended to protect against harm from uncertainty in diagnosis/disease trajectory. Despite recommendations to communicate diagnostic uncertainty when safety-netting, this is not always done.

Aims

To explore how and why doctors safety-netted in response to several clinical scenarios, within the broader context of exploring how doctors communicate diagnostic uncertainty.

Methods

Doctors working in internal medical specialties (n=36) from five hospitals were given vignettes in a randomised order (all depicting different clinical scenarios involving diagnostic uncertainty). After reading each, they told an interviewer what they would tell a ‘typical patient’ in this situation. A follow-up semistructured interview explored reasons for their communication. Interviews were recorded, transcribed and coded. We examined how participants safety-netted using a content analysis approach, and why they safety-netting with thematic analysis of the semistructured follow-up interviews using thematic analysis.

Results

We observed n=78 instances of safety-netting (across 108 vignette encounters). We found significant variation in how participants safety-netted. Safety-netting was common (although not universal), but clinicians differed in the detail provided about symptoms to be alert for, and the action advised. Although many viewed safety-netting as an important tool for managing diagnostic uncertainty, diagnostic uncertainty was infrequently explicitly discussed; most advised patients to return if symptoms worsened or new ‘red flag’ symptoms developed, but they rarely linked this directly to the possibility of diagnostic error. Some participants expressed concerns that communicating diagnostic uncertainty when safety-netting may cause anxiety for patients or could drive inappropriate reattendance/over-investigation.

Conclusions

Participants safety-netted variously, even when presented with identical clinical information. Although safety-netting was seen as important in avoiding diagnostic error, concerns about worrying patients may have limited discussion about diagnostic uncertainty. Research is needed to determine whether communicating diagnostic uncertainty makes safety-netting more effective at preventing harm associated with diagnostic error, and whether it causes significant patient anxiety.

Background

Polypharmacy is frequently used as a quality indicator for older adults in Residential Aged Care Facilities (RACFs) and is measured using a range of definitions. The impact of data source choice on polypharmacy rates and the implications for monitoring and benchmarking remain unclear. We aimed to determine polypharmacy rates (≥9 concurrent medicines) by using prescribed and administered data under various scenarios, leveraging electronic data from 30 RACFs.

Method

A longitudinal cohort study of 5662 residents in New South Wales, Australia. Both prescribed and administered polypharmacy rates were calculated biweekly from January 2019 to September 2022, providing 156 assessment times. 12 different polypharmacy rates were computed separately using prescribing and administration data and incorporating different combinations of items: medicines and non-medicinal products, any medicines and regular medicines across four scenarios: no, 1-week, 2-week and 4-week look-back periods. Generalised estimating equation models were employed to identify predictors of discrepancies between prescribed and administered polypharmacy.

Results

Polypharmacy rates among residents ranged from 33.9% using data on administered regular medicines with no look-back period to 63.5% using prescribed medicines and non-medicinal products with a 4-week look-back period. At each assessment time, the differences between prescribed and administered polypharmacy rates were consistently more than 10.0%, 4.5%, 3.5% and 3.0%, respectively, with no, 1-week, 2-week and 4-week look-back periods. Diabetic residents faced over two times the likelihood of polypharmacy discrepancies compared with counterparts, while dementia residents consistently showed reduced likelihood across all analyses.

Conclusion

We found notable discrepancies between polypharmacy rates for prescribed and administered medicines. We recommend a review of the guidance for calculating and interpreting polypharmacy for national quality indicator programmes to ensure consistent measurement and meaningful reporting.

Background

The consultation process, where a clinician seeks an opinion from another clinician, is foundational in medicine. However, the effectiveness of group diagnosis has not been studied.

Objective

To compare individual diagnosis to group diagnosis on two dimensions: group size (n=3 or 6) and group process (interactive or artificial groups).

Methodology

Thirty-six internal or emergency medicine residents participated in the study. Initially, each resident worked through four written cases on their own, providing a primary diagnosis and a differential diagnosis. Next, participants formed into groups of three. Using a videoconferencing platform, they worked through four additional cases, collectively providing a single primary diagnosis and differential diagnosis. The process was repeated using a group of six with four new cases. Cases were all counterbalanced. Retrospectively, nominal (ie, artificial) groups were formed by aggregating individual participant data into subgroups of three and six and analytically computing scores. Presence of the correct diagnosis as primary diagnosis or included in the differential diagnosis, as well as the number of diagnoses mentioned, was calculated for all conditions. Means were compared using analysis of variance.

Results

For both authentic and nominal groups, the diagnostic accuracy of group diagnosis was superior to individual for both the primary diagnosis and differential diagnosis. However, there was no improvement in diagnostic accuracy when comparing a group of three to a group of six. Interactive and nominal groups were equivalent; however, this may be an artefact of the method used to combine data.

Conclusions

Group diagnosis improves diagnostic accuracy. However, a larger group is not necessarily superior to a smaller group. In this study, interactive group discussion does not result in improved diagnostic accuracy.

Background

Therapeutic relationships are a key domain in healthcare delivery. While well-understood in in-person interventions, how therapeutic relationships develop in more complex contexts is unclear. This study aimed to understand (1) how therapeutic relationships are developed during the telehealth delivery of a group-based, complex intervention and (2) the perceived impact of these relationships on intervention processes, such as intervention delivery and engagement, and patient outcomes, such as patient safety and satisfaction.

Methods

This qualitative study, nested within a randomised controlled trial, used an interpretivist approach to explore the perceptions of 25 participants (18 patients with shoulder pain and 7 clinicians) regarding developing therapeutic relationships in a group-based, complex intervention delivered via telehealth. Semi-structured interviews were conducted within 4 weeks of the telehealth intervention period and then analysed through in-depth, inductive thematic analysis.

Results

We identified six themes: (1) ‘Patients trust clinicians who demonstrate credibility, promoting the development of therapeutic relationships’; (2) ‘Simple features and approaches shape the therapeutic relationship’, including small talk, time spent together and social observation; (3) ‘A sense of belonging and support fosters connections’, facilitated by clinicians providing individualised attention within the group; (4) ‘Developing therapeutic relationships can impact the delivery of core intervention components’, reflecting challenges clinicians faced; (5) ‘Therapeutic relationships can facilitate intervention engagement’, through enhanced patient understanding and confidence and (6) ‘Therapeutic relationships can contribute to patient safety and satisfaction’, with patients feeling more comfortable reporting intervention-related issues.

Conclusions

Therapeutic relationships were developed during group-based telehealth sessions through a set of factors that may require additional skills and effort compared with in-person interactions. While these relationships have a perceived positive impact on intervention engagement and patient outcomes, clinicians need to find a balance between building relationships and delivering the telehealth intervention with fidelity.

Trial registration number

ACTRN12621001650886.

Introduction

Pharmacist-led medication reviews are an established intervention to support patients prescribed multiple medicines or with complex medication regimes. For this systematic review, a medication review was defined as ‘a consultation between a pharmacist and a patient to review the patient’s total medicines use with a view to improve patient health outcomes and minimise medicines-related problems’. It is not known how varying approaches to medication reviews lead to different outcomes.

Aim

To explore the common themes associated with positive outcomes from pharmacist-led medication reviews.

Method

Randomised controlled trials of pharmacist-led medication reviews in adults aged 18 years and over were included. The search terms used in MEDLINE, EMBASE and Web of Science databases were "medication review", "pharmacist", "randomised controlled trial" and their synonyms, time filter 2015 to September 2023. Studies published before 2015 were identified from a previous systematic review. Risk of bias was assessed using the Cochrane risk of bias 2 tool. Descriptions of medication reviews’ components, implementation and outcomes were narratively synthesised to draw out common themes. Results are presented in tables.

Results

Sixty-eight papers describing 50 studies met the inclusion criteria. Common themes that emerged from synthesis include collaborative working which may help reduce medicines-related problems and the number of medicines prescribed; patient involvement in goal setting and action planning which may improve patients’ ability to take medicines as prescribed and help them achieve their treatment goals; additional support and follow-up, which may lead to improved blood pressure, diabetes control, quality of life and a reduction of medicines-related problems.

Conclusion

This systematic review identified common themes and components, for example, goal setting, action planning, additional support and follow-up, that may influence outcomes of pharmacist-led medication reviews. Researchers, health professionals and commissioners could use these for a comprehensive evaluation of medication review implementation.

PROSPERO registration number

CRD42020173907.

Diagnostic error: the problem

Failures in the diagnostic process are thought to affect at least 15% of patient encounters, cause 34% of adverse events in hospitals, are a leading cause in major malpractice claims and payouts and are recognised as a top priority in patient safety research.1–3 The National Academies of Science, Engineering and Medicine defines diagnostic error as a failure to establish an accurate and timely explanation of a patient’s medical problem and has been shown to contribute to the morbidity and mortality of an estimated 795 000 patients each year in the USA.1 Although diagnostic error has received significant research attention across multiple clinical settings over the last several decades, it continues to pose consequential challenges and requires improvement in systematic investigation and operational intervention.3 4 Additionally, few effective mitigation strategies have been designed for...

The widely known United Nations sustainable development goals indicate that good quality healthcare should be available to all those who need it.1 Unfortunately, the availability of good care tends to vary inversely with the actual need for it. This is not a new issue and had already been dubbed ‘the inverse care law’ as far back as 1971.2 In response, those working in healthcare strived to make changes, aiming to provide equitable access to high-quality healthcare to all patients. Despite this, differences between groups persist. Even in populations with good healthcare coverage, certain subgroups attend screening less,3 experience lower availability of care4 and have ongoing unmet needs.5 To address this, the concept of health equity was introduced, with multiple recent landmark articles stressing its importance.6 7 BMJ Quality and Safety has also embraced this concept,...

Background

Early intervention for unmet needs is essential to improve health. Clear inequalities in healthcare use and outcomes exist. The Children and Young People’s Health Partnership (CYPHP) model of care uses population health management methods to (1) identify and proactively reach children with asthma, eczema and constipation (tracer conditions); (2) engage these families, with CYPHP, by sending invitations to complete an online biopsychosocial Healthcheck Questionnaire; and (3) offer early intervention care to those children found to have unmet health needs. We aimed to understand this model’s effectiveness to improve equitable access to care.

Methods

We used primary care and CYPHP service-linked records and applied the same methods as the CYPHP’s population health management process to identify children aged <16 years with a tracer condition between 1 April 2018 and 30 August 2020, those who engaged by completing a Healthcheck and those who received early intervention care. We applied multiple imputation with multilevel logistic regression, clustered by general practitioner (GP) practice, to investigate the association of deprivation and ethnicity, with children’s engagement and receiving care.

Results

Among 129 412 children, registered with 70 GP practices, 15% (19 773) had a tracer condition and 24% (4719) engaged with CYPHP’s population health management system. Children in the most deprived, compared with least deprived communities, had 26% lower odds of engagement (OR 0.74; 95% CI 0.62 to 0.87). Children of Asian or black ethnicity had 31% lower odds of engaging, compared with white children (0.69 (0.59 to 0.81) and 0.69 (0.62 to 0.76), respectively). However, once engaged with the population health management system, black children had 43% higher odds of receiving care, compared with white children (1.43 (1.15 to 1.78)), and children from the most compared with least deprived communities had 50% higher odds of receiving care (1.50 (1.01 to 2.22)).

Conclusion

Detection of unmet needs is possible using population health management methods and increases access to care for children from priority populations with the highest needs. Further health system strengthening is needed to improve engagement and enhance proportionate universalist access to healthcare.

Trial registration number

ClinicalTrials.gov Registry (NCT03461848).

Background

Large-scale improvement programmes are a frequent response to quality and safety problems in health systems globally, but have mixed impact. The extent to which they meet criteria for programme quality, particularly in relation to transparency of reporting and evaluation, is unclear.

Aim

To identify large-scale improvement programmes focused on intrapartum care implemented in English National Health Service maternity services in the period 2010–2023, and to conduct a structured quality assessment.

Methods

We drew on the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews guidance to inform the design and reporting of our study. We identified relevant programmes using multiple search strategies of grey literature, research databases and other sources. Programmes that met a prespecified definition of improvement programme, that focused on intrapartum care and that had a retrievable evaluation report were subject to structured assessment using selected features of programme quality.

Results

We identified 1434 records via databases and other sources. 14 major initiatives in English maternity services could not be quality assessed due to lack of a retrievable evaluation report. Quality assessment of the 15 improvement programmes meeting our criteria for assessment found highly variable quality and reporting. Programme specification was variable and mostly low quality. Only eight reported the evidence base for their interventions. Description of implementation support was poor and none reported customisation for challenged services. None reported reduction of inequalities as an explicit goal. Only seven made use of explicit patient and public involvement practices, and only six explicitly used published theories/models/frameworks to guide implementation. Programmes varied in their reporting of the planning, scope and design of evaluation, with weak designs evident.

Conclusions

Poor transparency of reporting and weak or absent evaluation undermine large-scale improvement programmes by limiting learning and accountability. This review indicates important targets for improving quality in large-scale programmes.

Background

Antiplatelet therapy (APT) can substantially reduce the risk of further vascular events in individuals with established atherosclerotic cardiovascular disease (ASCVD). However, knowledge regarding the extent and determinants of APT use is limited.

Objectives

Estimate the extent and identify patient groups at risk of suboptimal APT use at different stages of the treatment pathway.

Methods

Retrospective cohort study using linked NHS Scotland administrative data of all adults hospitalised for an acute ASCVD event (n=150 728) from 2009 to 2017. Proportions of patients initiating, adhering to, discontinuing and re-initiating APT were calculated overall and separately for myocardial infarction (MI), ischaemic stroke and peripheral arterial disease (PAD). Multivariable logistic regression and Cox proportional hazards models were used to assess the contribution of patient characteristics in initiating and discontinuing APT.

Results

Of patients hospitalised with ASCVD, 84% initiated APT: 94% following an MI, 83% following an ischaemic stroke and 68% following a PAD event. Characteristics associated with lower odds of initiation included female sex (22% less likely than men), age below 50 years or above 70 years (aged <50 years 26% less likely, and aged 70–79, 80–89 and ≥90 years 21%, 39% and 51% less likely, respectively, than those aged 60–69 years) and history of mental health-related hospitalisation (45% less likely). Of all APT-treated individuals, 22% discontinued treatment. Characteristics associated with discontinuation were similar to those related to non-initiation.

Conclusions

APT use remains suboptimal for the secondary prevention of ASCVD, particularly among women and older patients, and following ischaemic stroke and PAD hospitalisations.

Objectives

To estimate the number and burden of medication errors associated with prescription information transfer within the National Health Service (NHS) in England and the impact of implementing an interoperable prescription information system (a single digital prescribing record shared across NHS settings) in reducing these errors.

Methods

We constructed a probabilistic mathematical model. We estimated the number of transition medication errors that would be undetected by standard medicines reconciliation, based on published literature, and scaled this up based on the annual number of hospital admissions. We used published literature to estimate the proportion of errors that lead to harm and applied this to the number of errors to estimate the associated burden (healthcare resource use and deaths). Finally, we used reported effect sizes for electronic prescription information sharing interventions to estimate the impact of implementing an interoperable prescription information system on number of errors and resulting harm.

Results

Annually, around 1.8 million (95% credibility interval (CrI) 1.3 to 2.6 million) medication errors were estimated to occur at hospital transitions in England, affecting approximately 380 000 (95% CrI 260 397 to 539 876) patient episodes. Harm from these errors affects around 31 500 (95% CrI 22 407 to 42 906) patients, with 36 500 (95% CrI 25 093 to 52 019) additional bed days of inpatient care (costing around £17.8 million (95% CrI £12.4 to £24.9 million)) and >40 (95% CrI 9 to 146) deaths. Assuming the implementation of an interoperable prescription information system could reduce errors by 10% and 50%, there could be 180 000–913 000 fewer errors, 3000–15 800 fewer people who experience harm and 4–22 lives saved annually.

Conclusions

An interoperable prescription information system could provide major benefits for patient safety. Likely additional benefits include healthcare professional time saved, improved patient experience and care quality, quicker discharge and enhanced cross-organisational medicines optimisation. Our findings provide vital safety and economic evidence for the case to adopt interoperable prescription information systems.

Objective

The objective of this review was to develop a taxonomy of pressures experienced by health services and an accompanying taxonomy of strategies for adapting in response to these pressures. The taxonomies were developed from a review of observational studies directly assessing care delivered in a variety of clinical environments.

Design

In the first phase, a scoping review of the relevant literature was conducted. In the second phase, pressures and strategies were systematically coded from the included papers, and categorised.

Data sources

Electronic databases (MEDLINE, Embase, CINAHL, PsycInfo and Scopus) and reference lists from recent reviews of the resilient healthcare literature.

Eligibility criteria

Studies were included from the resilient healthcare literature, which used descriptive methodologies to directly assess a clinical environment. The studies were required to contain strategies for managing under pressure.

Results

5402 potential articles were identified with 17 papers meeting the inclusion criteria. The principal source of pressure described in the studies was the demand for care exceeding capacity (ie, the resources available), which in turn led to difficult working conditions and problems with system functioning. Strategies for responding to pressures were categorised into anticipatory and on-the-day adaptations. Anticipatory strategies included strategies for increasing resources, controlling demand and plans for managing the workload (efficiency strategies, forward planning, monitoring and co-ordination strategies and staff support initiatives). On-the-day adaptations were categorised into: flexing the use of existing resources, prioritising demand and adapting ways of working (leadership, teamwork and communication strategies).

Conclusions

The review has culminated in an empirically based taxonomy of pressures and an accompanying taxonomy of strategies for adapting in response to these pressures. The taxonomies could help clinicians and managers to optimise how they respond to pressures and may be used as the basis for training programmes and future research evaluating the impact of different strategies.

The capabilities of artificial intelligence (AI) have accelerated over the past year, and they are beginning to impact healthcare in a significant way. Could this new technology help address issues that have been difficult and recalcitrant problems for quality and safety for decades? While we are early in the journey, it is clear that we are in the midst of a fundamental shift in AI capabilities. It is also clear these capabilities have direct applicability to healthcare and to improving quality and patient safety, even as they introduce new complexities and risks. Previously, AI focused on one task at a time: for example, telling whether a picture was of a cat or a dog, or whether a retinal photograph showed diabetic retinopathy or not. Foundation models (and their close relatives, generative AI and large language models) represent an important change: they are able to handle many different kinds of problems without additional datasets or training. This review serves as a primer on foundation models’ underpinnings, upsides, risks and unknowns—and how these new capabilities may help improve healthcare quality and patient safety.

To date, most safety and quality improvement efforts to mitigate harm have focused on the single diagnosis for which the patient was admitted to the hospital. Most often, the objective has been to ensure patients receive the appropriate evidence-based therapies for their diagnosis using guidelines, checklists, learning from defect tools1 or other interventions. However, people often have multiple morbidities and the interactions between them may increase their risk of harm when hospitalised.

Approximately half of all Americans have a chronic disease.2 In addition, an estimated 100 million disability-adjusted life years were added between 2000 and 2019 from a global rise in diabetes, ischaemic heart disease and several other non-communicable diseases.3 However, healthcare has paid less attention to mitigating significant risks of harm from the chronic diseases or disabilities patients have when admitted for another health reason. For example, 63% of hospitalised patients with Parkinson’s...

Medications continue to represent a major cause of harm, both in inpatients and outpatients and in adults and children. In a recent large study, medications were the leading cause of harm in inpatient adults, and the same was true for adult outpatients.1 2 Medications were also the most frequent cause of harm in a large paediatric inpatient study.3 Despite these and other data, the magnitude of this issue has been systematically underestimated by healthcare organisations.

The main reason for this is that operational estimates of both medication harm and error rates have relied on spontaneous reporting, which is ineffective in this instance. We found in a study published 30 years ago that spontaneous reporting only finds about 1 in 20 harmful drug events.4 It almost certainly finds an even lower proportion of the total number of medication errors.5 But...

In this issue of the journal, Wegwarth et al report on a study that sought to identify general practitioner (GP) characteristics that predicted prescribing of potentially hazardous medications or, as the authors put it, ‘too much medicine’.1 An online survey of 304 English GPs measured their risk literacy, conflicts of interest, and perceived benefit-to-harm ratio in low-value prescribing scenarios. National Health Service record data were used to derive prescribing volumes for the participating GPs for antibiotics, opioids, gabapentin and benzodiazepines. The range of risk literacy scores was dichotomised and those GPs with low risk literacy were found to prescribe more opioids, gabapentin and benzodiazepines than GPs with high risk literacy—although no difference was found for antibiotics. The other two independent variables—conflicts of interest and benefit/harm perceptions—were not associated with prescribing volumes.

The risk literacy questions in the survey gauged GPs’ ability to interpret clinical trial results with...

Objectives

To compare medication errors identified at audit and via direct observation with medication errors reported to an incident reporting system at paediatric hospitals and to investigate differences in types and severity of errors detected and reported by staff.

Methods

This is a comparison study at two tertiary referral paediatric hospitals between 2016 and 2020 in Australia. Prescribing errors were identified from a medication chart audit of 7785 patient records. Medication administration errors were identified from a prospective direct observational study of 5137 medication administration doses to 1530 patients. Medication errors reported to the hospitals’ incident reporting system were identified and matched with errors identified at audit and observation.

Results

Of 11 302 clinical prescribing errors identified at audit, 3.2 per 1000 errors (95% CI 2.3 to 4.4, n=36) had an incident report. Of 2224 potentially serious prescribing errors from audit, 26.1% (95% CI 24.3 to 27.9, n=580) were detected by staff and 11.2 per 1000 errors (95% CI 7.6 to 16.5, n=25) were reported to the incident system. Although the prescribing error detection rates varied between the two hospitals, there was no difference in incident reporting rates regardless of error severity. Of 40 errors associated with actual patient harm, only 7 (17.5%; 95% CI 8.7% to 31.9%) were detected by staff and 4 (10.0%; 95% CI 4.0% to 23.1%) had an incident report. None of the 2883 clinical medication administration errors observed, including 903 potentially serious errors and 144 errors associated with actual patient harm, had incident reports.

Conclusion

Incident reporting data do not provide an accurate reflection of medication errors and related harm to children in hospitals. Failure to detect medication errors is likely to be a significant contributor to low error reporting rates. In an era of electronic health records, new automated approaches to monitor medication safety should be pursued to provide real-time monitoring.