In this post, we’re offering summaries and comments on articles from BMJ Quality & Safety’s Top 10 Articles of 2021. To check out the full list of our 21 finalists, click here. The editors and the Editorial Board used data such as citation rates and social media engagement in addition to their own judgement to create these lists. Overall, they demonstrate the wonderful breadth of valuable articles published in BMJ Quality & Safety.
- Evaluating the safety of mental health-related prescribing in UK primary care: a cross-sectional study using the Clinical Practice Research Datalink (CPRD) by WY Khawagi (@khawagi) et al. Published Online First: 25 August 2021. Accompanying editorial available here.
By analysing prescriptions for more than 1.6 million at-risk patients, Khawagi et al. sought to determine the prevalence of, and patient and practice-level risk factors for, 18 mental health-related potentially hazardous prescribing indicators and four inadequate medication monitoring indicators in UK primary care. The authors applied a tailored suite of 22 prescribing safety indicators (PSIs), developed by mental health experts, across 361 practices. Applying such models to capture possible harm is particularly important, as up to 40% of general practice consultations may be related to mental health concerns and 90% of mental health diagnoses are managed entirely within primary care in the UK. Overall, 9.4% of patients had at least one potentially hazardous prescription, and 90.2% of over 42,000 patients who required medication monitoring had at least one potentially hazardous monitoring episode. Significant practice variation was seen for several indicators, including benzodiazepine prescribing and physical health monitoring for people taking antipsychotics. Ultimately, 11 of the 22 PSIs were considered to have adequate reliability to capture practice variability and facilitate benchmarking. Although these prescribing indicators may not fully capture the complexity of individual patients and the medications they require to maintain and improve their health, such work is an important step in engaging multidisciplinary primary care teams to optimise medication in a patient-centred manner.
- Impact of COVID-19 restrictions on diabetes health checks and prescribing for people with type 2 diabetes: a UK-wide cohort study involving 618,161 people in primary care by MJ Carr et al. Published Online First: 12 October 2021. Accompanying editorial available here.
In addition to the direct consequences of the COVID-19 pandemic on morbidity and mortality, many people have suffered indirect consequences related to ease and timing of access to healthcare. Using data from the UK’s Clinical Practice Research Datalink, Carr et al. examined the effects of the first national lockdown on six of the nine essential ‘health checks’ for high-quality diabetes care as defined by the National Institute for Health and Care Excellence, and on diabetes medication prescribing practices. The six chosen factors were those that could be verified within electronic health records, including measurements of HbA1c, urinary albumin excretion, and blood pressure; the medications studied were those most often prescribed to people with diabetes: antidiabetics, antihypertensives, lipid-lowering agents, and antiplatelets. Across more than 600,000 patients, health checks decreased approximately 80% for all six measures during April 2020 relative to historical trends. Although rates of performance subsequently increased following this precipitous drop, rates of most health checks remained ~30-50% below historical norms by the end of 2020. Similarly, prescription rates were also lower, with ~20% reductions in new prescriptions for both antidiabetic and antihypertensive medications. Furthermore, health checks were most likely to be reduced in older patients and populations with greater levels of deprivation. Extrapolated across the UK population, this would mean more than 2.3 million fewer blood pressure checks, 1.2 million fewer HbA1c tests, and 31,000 fewer new antidiabetic medication prescriptions performed in 2020, concentrated within the most vulnerable populations. As new models of care are developed and expanded in response to the COVID-19 pandemic, steps must be taken to ensure access and ongoing high-quality care for all those who received fewer necessary services during this period.
- Filling a gap in safety metrics: development of a patient-centred framework to identify and categorise patient-reported breakdowns related to the diagnostic process in ambulatory care by SK Bell et al. Published Online First: 16 October 2021. Accompanying editorial available here.
Within current safety frameworks of diagnostic error, there often has been no avenue to elicit patient and family input. To create a novel classification system for ‘patient-reported diagnostic process-related breakdowns’ (PRDBs), Bell et al. integrated findings from a multi-stakeholder advisory group (MAG) with the results of two surveys identifying greater than 2,000 patient-reported ambulatory errors in more than 25,000 patients. The resulting framework for classifying PRDBs included three domains: 1) breakdown categories (of which there were seven, such as ‘medical history’, with a total of 40 subcategories, such as ‘wrong symptoms or main concern’); 2) contributing factors (three categories, such as ‘patient factors’, with 19 subcategories, such as ‘barriers to communication or speaking up’); and 3) impact on patients (two categories, including both ‘patient activation and mitigation’ and ‘negative patient impact’, with 11 subcategories, such as ‘negative impact on relationships’). Importantly, as noted by the MAG, the impact of a PRDB on the patient could lead to positive outcomes, such as when experiencing a delay leads to greater patient activation in staying vigilant or requesting/reviewing records. Additionally, the MAG noted that relationships with clinicians can be strengthened by breakdowns, that patient activation can coexist with negative effects, and that patients and clinicians may have goals that are misaligned during and after encounters. Overall, review of the two surveys identified PRDBs in roughly 6.5% of cases, with the category of ‘Communication and Respect’ present in about 30% of PRDBs. The authors note several opportunities to incorporate such information into existing structures, such as review of safety events and training of patient relations personnel. Eventually, however, the goal must be to create strategies using the PRDB framework to reduce their occurrence and proactively engage patients and families across the continuum of care.
- Effectiveness of a multifaceted intervention to improve emergency department care of low back pain: a stepped-wedge, cluster-randomised trial by DM Coombs et al. BMJ Quality & Safety 2021;30:825-835.
Patients presenting for emergency care for low back pain receive far more imaging, opioid medications, and hospital admissions than would be predicted based on the numbers of patients presenting with severe, urgent spinal pathology and clinical practice guidelines. Coombs et al. performed a stepped-wedge, cluster-randomised trial evaluating implementation of an evidence-based model of care for low back pain management in four Australian emergency departments (EDs), involving 269 clinicians. Each ED had a four-week intervention phase preceded by 13 months of usual care and followed by at least three months of post-intervention follow-up. The intervention included five steps: educational sessions, educational materials, easier access to non-opioid treatments in the ED, fast-track referrals for outpatient services such as physiotherapy, and audit-and-feedback. Of the healthcare utilisation outcomes for the 4,625 encounters, an absolute decrease was seen only for the secondary outcome of opioid medication use while patients were being seen in the ED (from 62.8% to 50.5% of low back pain presentations). There were no significant changes in receipt of lumbar imaging, strong opioids and non-opioid administration, ED length of stay, specialist consultations, admissions, and return to the ED within 48 hours. Even with reduced use of opioid medications during ED encounters, patient-reported outcomes in the post-intervention periods were non-inferior to the pre-intervention periods, including pain at one week post-presentation, physical functioning, satisfaction with the ED visit/clinicians, and overall quality of life. Although rates of baseline lumbar imaging in the studied EDs were lower than in some previous studies, factors such as patient expectations and clinicians’ fears of missing significant pathology may make care for such common problems difficult to modify in urgent and emergency care settings.
- Incidence, origins and avoidable harm of missed opportunities in diagnosis: longitudinal patient record review in 21 English general practices by S Cheraghi-Sohi (@Dr_Sudeh) et al. BMJ Quality & Safety 2021;30:977-985. Accompanying editorial available here.
Reviewing almost 2,100 consultations across 21 general practices over the course of one year in North West England, Cheraghi-Sohi et al. sought to clarify the current rate of ‘missed diagnostic opportunities’ (MDOs), defined as opportunities to make a correct or timely diagnosis based on the then available evidence. The authors selected patient records from each practice in four spaced intervals across one year, reviewing cases for all represented patients for at least three months prior to and nine months following the sampling date. New diagnoses from cases were each coded into six categories based on likelihood of MDO. Diagnoses with possible, likely, or certain MDOs were then characterized using a 5-point scale for degree of resultant harm. Ultimately, MDOs were implicated in roughly 4% of reviewed consultations, and just over 7% of diagnoses (as not every encounter had a new diagnosis). The top three issues leading to MDOs were 1) problems within the patient-practitioner encounter (a primary or secondary factor in 68%), 2) performance and interpretations of diagnostic testing (35%), and 3) follow-up and tracking of diagnostic information (48%), with approximately 75% of MDOs involving more than one issue. Roughly half of MDOs caused mild harm, but 37% were noted to lead to moderate or severe harm. Given the number of general practice encounters, millions of MDOs may occur within the NHS annually. As most MDOs are multifactorial, interventions addressing MDOs will likely need to target multiple steps throughout the diagnostic process in order to inform the creation of safer systems.