In this post, we’re offering the second set of summaries and comments on articles from BMJ Quality & Safety’s Top 10 Articles of 2023. For the full list of our 23 finalists, click here, and to read more about the annual Top Article selection process, please click here. To briefly summarize, the editors used data such as citation rates and social media engagement to create the finalist article list, which was then scored by the Editorial team on novelty, generalizability/addition to understanding, rigor of methods, and importance to clinical practice to arrive at this Top 10. These articles demonstrate the wonderful breadth of valuable articles published in BMJ Quality & Safety every year.
- Initial opioid prescription characteristics and risk of opioid misuse, poisoning and dependence: retrospective cohort study by A García-Sempere et al. BMJ Quality & Safety 2024;33:13-23. Accompanying editorial available here.
Across high-income countries, both the number of opioid prescriptions and the doses of the average prescription have increased over the past few decades. While patient factors associated with increased likelihood of a serious opioid-related outcome have been identified, factors related to prescriptions have been less well studied. García-Sempere and colleagues therefore retrospectively studied initial opioid prescriptions and outcomes in five million patients from 2012-2018 in a Spanish regional health system. Over this period, they found more than 950,000 patients were initiated on prescription opioids for non-cancer diagnoses, and 1,289 of these (0.13%) experienced a serious opioid-related outcome (ORO). Three-quarters of patients were initiated on therapy with tramadol, with a median duration for any initiated medication of 9 days. Initiation with any non-codeine medication other than tramadol had increased hazard ratios for a serious ORO, although very few patients were initiated on the highest-risk medications (ultrafast and short-acting opioids – 0.1% and 0.2% of patients, respectively). However, when patients were initiated on higher-risk formulations, risks of OROs were significantly elevated, with oxycodone having a nearly fivefold increased risk compared with tramadol. Similarly, prescription duration > 3 days had higher risk than durations of 1-3 days, with the hazard ratios increasing across each studied duration. Finally, concurrent use of benzodiazepines, gabapentinoids, and/or antipsychotics increased risk of OROs, as did initiation of daily total doses of >120 morphine milligram equivalents (MMEs) versus those < 50 MMEs. Identification of these prescription factors may help individual clinicians and health systems optimize pain management and patient safety.
- Health services under pressure: a scoping review and development of a taxonomy of adaptive strategies by B Page (@bethanpage21) et al. BMJ Quality & Safety Published Online First: 29 November 2023.
Resources and capacity of health systems are finite and at times may be mismatched with the demands for services and health needs of the populations served. Systems adapt as best they are able during such times of increased pressure, but they often struggle to maintain standards of quality and safety. Page and colleagues therefore developed taxonomies for both pressures and short-term and medium-term adaptive strategies by performing a scoping review of such system responses. Across the 17 included studies, the authors noted 166 pressures and almost 350 adaptations. Their inductive thematic analysis yielded four broad system pressures that may be interrelated: contextual pressures (those external to the clinical teams, such as COVID-19 or funding); demand exceeding capacity (such as space limitations or inadequate staffing); difficult working conditions (including increased workload and time pressures); and problems with system functioning (including patient flow issues and delays in access). Adaptations were characterised based on timing, with some being anticipatory and some ‘on-the-day’. Anticipatory strategies included increasing resources (such as training staff to improve skill mix or providing overtime payments), controlling demand (including restricting services or transferring patients to less pressured services), and planning for workload management (for example, by shifting to telehealth to preserve physical space or developing contingency plans). On-the-day strategies included flexing resources (such as task shifting, transferring patients, and changing staffing ratios), prioritising demand (for example, by transitioning certain care areas into acute management spaces), and changing work tasks (including having ad hoc meetings for information-sharing or with increased interprofessional collaboration). The authors also note that many teams will enact multiple adaptive strategies. Overall, such a taxonomy is an early step in enabling organisations to better train personnel and coordinate team- and system-level responses to system pressures.
- Why is safety in intrapartum electronic fetal monitoring so hard? A qualitative study combining human factors/ergonomics and social science analysis by G Lamé et al. BMJ Quality & Safety Published Online First: 09 November 2023. Accompanying editorial available here.
Intrapartum electronic fetal monitoring (EFM) is a necessary procedure during many deliveries but misinterpretation or failures to act on the results carry the risk of potentially reducible or preventable harm. Improvement efforts have often focused on interpretation of monitoring results, rather than the totality of EFM care. Lamé and a multidisciplinary team of colleagues conducted more than 300 hours of observations and 23 interviews using the Systems Engineering Initiative for Patient Safety (SEIPS) v. 2.0 framework in three maternity units across the United Kingdom, during clinical situations in which the decision to perform EFM had been made. In all three units, a similar procedure existed for midwives to monitor the tracing results, classify deviations, and escalate clinical decision-making after consultation with a second midwife on abnormal tracings. In their analysis, the researchers identified multiple concerns in these complex clinical microsystems around work systems, processes, and outcomes. For example, staffing levels were often noted to be below the desired levels, increasing workload and staff fatigue and cognitive burden, as well as reducing attendance at training or case review opportunities. Furthermore, suboptimal equipment design, location, and inadequate maintenance led to wasted time spent finding supplies or restarting machines and risked the loss of important clinical data. The studied units used different standards for interpretation and action, increasing the difficulty in incorporating updates and variability in rendering judgments, and confusing staff that rotate at different locations. Escalating concerns was complicated further by clinicians potentially biasing the secondary reviewer and the relatively high false-positive rate of EFM tracings that might increase the likelihood of unnecessary intervention. Given the interdependent elements characterised, multifaceted improvement efforts with a focus on design and rationalisation of guidance will likely be required to improve EFM systems.
- Equitable and accessible informed healthcare consent process for people with intellectual disability: a systematic literature review by M Dunn et al. BMJ Quality & Safety Published Online First: 09 December 2023. Accompanying editorial available here.
Informed consent processes need to be accessible to all individuals, including those with intellectual disability. In reality, persons with such disabilities often encounter significant access barriers to healthcare, including to the adjustments necessary to allow them to practice greater bodily autonomy around decisions for medical procedures or interventions. Dunn and colleagues, including two co-researchers with an intellectual disability, performed a systematic review of facilitators and barriers to informed consent in this population, finding 23 studies. Most studies were from the UK, and only one involved pediatric patients. Consent was studied in different scenarios, most commonly around contraception/sexual health or other medications. The authors identified six themes from these articles. Three of the themes were noted as barriers to informed consent: clinician attitudes and lack of education and training on disability, such as assuming patients had incapacity and using proxy consent inappropriately; inadequate access to health information for those with disability, such as inability to find information in formats that met an individual’s communication needs; and health system constraints, such as inadequate time and lack of longitudinal consent processes to facilitate ongoing conversations. Alternatively, two themes were identified as facilitators of informed consent: person-centred consent processes such as awareness of power imbalances and emphasizing available choices, and effective communication between patients and professionals such as use of multimodal communication aids. Finally, one theme served as both a potential barrier and a potential facilitator: the engagement of support personnel, who might either facilitate communication or impede individual autonomy. Ideally, coproduction of resources for individuals with intellectual disability and improved guidance and training for clinicians will improve informed consent for all patients.
- What works in medication reconciliation: an on-treatment and site analysis of the MARQUIS2 study by JL Schnipper (@drjschnip) et al. BMJ Quality & Safety 2023;32:457-469. Accompanying editorial available here.
Medication discrepancies occur frequently during hospitalizations and transitions of care back to community settings. Previous work from the MARQUIS2 study showed decreased discrepancies during an 18-month study across 18 hospitals using various combinations of their studied interventions. In this article, Schnipper and colleagues investigate the 17 system-level and seven individual-level interventions from the study toolkit and implementation framework to isolate each component’s association with medication discrepancies, which was made possible by their monthly surveys of site leads and study pharmacist sampling of patient records from each site. The 10 hospitals with significant decreases in discrepancies during the study had higher median numbers of implemented systems-level interventions than those hospitals that had no significant decrease in discrepancies, as well as a greater increase in the mean proportion of patients who received patient-level interventions during the study (resulting in a slightly higher overall mean proportion of patients receiving at least one patient-level intervention in those hospitals with significant decreases). Overall, 15 of the 17 system-level interventions had at least a small positive effect. At the system level, the implementation of systems to identify and correct discrepancies in real time and improvements in health information technology were associated with the largest reductions in discrepancy rates. Patient-level interventions had larger effects, with best possible medication history (BPMH) being taken in the ED by a trained clinician, and admission and discharge medication reconciliations by trained clinicians all substantially lowering the risk of medication discrepancies. Combining a BPMH in the ED and discharge medication reconciliation was the most effective overall strategy, with a rate of discrepancy of only 0.08 per medication. As noted by the authors, the systems-level interventions, while having a lower magnitude than patient-level interventions, may be important foundations for effective individual interventions and may have a broader impact on patients over time. Ultimately, such dedicated analysis of complex interventions offers opportunities to streamline future efforts and help other hospitals adopt the highest yield components.