Seeing the Truth About Double Blinding.

Randomized clinical trials provide reassurances that confounding factors are balanced at baseline whereas blinding is essential to assure the balance of extraneous factors thereafter. This article provides a three-part taxonomy of pitfalls that can arise because of inadequate blinding in clinical trials. We introduce a cautionary framework for readers interpreting a blinded randomized trial for evidence-based medicine. Each pitfall is illustrated with a relevant example of a potential bias resulting from knowledge of group assignment. Several pitfalls occur during the conduct of the study including inadequate blinding of the intervention group, control group, or responsible clinicians. Additional pitfalls relate to data analysis including unsubstantiated assertions of blinding and subverted tests for blinding. Further pitfalls arise due to surrounding oversight including unblinding of research ethics boards and scientific reviewers. These caveats are sources of misunderstanding when observing the apparent connection between a clinical intervention and patient outcomes. An awareness of specific pitfalls might help advance the interpretation and application of blinded randomized clinical trials to inform evidence-based medical care.

Syncope While Driving and the Risk of a Subsequent Motor Vehicle Crash.

STUDY OBJECTIVE: Syncope that occurs while driving can result in a motor vehicle crash. Whether individuals with a prior syncope-related crash exhibit an exceptional risk of subsequent crash remains uncertain. METHODS: We performed a population-based retrospective observational study of patients diagnosed with 'syncope and collapse' at any of 6 emergency departments in British Columbia, Canada (2010 to 2015). Data were obtained from chart abstraction, administrative health records, insurance claims and police crash reports. We compared crash-free survival among individuals with crash-associated syncope (a crash and an emergency visit for syncope on the same date) to that among controls with syncope alone (no crash on date of emergency visit for syncope). RESULTS: In the year following their index emergency visit, 13 of 63 drivers with crash-associated syncope and 852 of 9,160 controls with syncope alone experienced a subsequent crash as a driver (crash risk 21% versus 9%). After accounting for censoring and potential confounders, crash-associated syncope was not associated with a significant increase in the risk of subsequent crash (adjusted hazard ratio [aHR] 1.38, 95% confidence interval [CI] 0.78 to 2.47). Individuals with crash-associated syncope were 31-fold more likely to have physician driving advice documented during their index visit (prevalence ratio 31.0, 95% CI, 21.3 to 45.1). In the subgroup without documented driving advice, crash-associated syncope was associated with a significant increase in subsequent crash risk (aHR 1.88, 95% CI 1.06 to 3.36). CONCLUSIONS: Crash risk after crash-associated syncope appears similar to crash risk after syncope alone.

On-call absences and academic recognition: A retrospective cohort analysis.

BACKGROUND: Resident call schedules require careful planning and are vulnerable to unanticipated absences from unpredictable factors. We tested whether unplanned absences from resident call schedules were associated with the likelihood of subsequent academic recognition. METHODS: We examined unplanned absences from call shifts for internal medicine residents at the University of Toronto from 2014 to 2022 (8 years). We identified institutional awards granted at the end of the academic year as an indicator of academic recognition. We defined the resident-year as the unit-of-analysis that started in July and ended in June of the subsequent year. Secondary analyses examined the association between unplanned absences and the likelihood of academic recognition in later years. RESULTS: We identified 1668 resident-years of training in internal medicine. In total, 579 (35%) had an unplanned absence, and the remaining 1089 (65%) had no unplanned absence. Baseline characteristics were similar between the two groups of residents. In total, 301 awards were received for academic recognition. The likelihood of receiving an award at the end of the year was 31% lower for residents who had any unplanned absence compared with those who had no absence (adjusted odds ratio = 0.69, 95% confidence interval 0.51-0.93, p = 0.015). The likelihood of receiving an award was further decreased for residents with multiple unplanned absences compared with those with none (odds ratio 0.54, 95% confidence interval 0.33-0.83, p = 0.008). An absence during the first year of residency was not significantly associated with the likelihood of academic recognition in later years of training (odds ratio 0.62, 95% confidence interval 0.36-1.04, p = 0.081). CONCLUSIONS: The results of this analysis suggest unplanned absences from scheduled call shifts may be associated with a decreased likelihood of academic recognition for internal medicine residents. This association could reflect countless confounders or the prevailing culture of medicine.

Changes in Media Reporting Quality and Suicides Following National Media Engagement on Responsible Reporting of Suicide in Canada: Changements de la Qualité des reportages dans les médias sur les suicides suite à l'engagement des médias nationaux à la déclaration responsable du suicide au Canada.

OBJECTIVE: Responsible media reporting is an accepted strategy for preventing suicide. In 2015, suicide prevention experts launched a media engagement initiative aimed at improving suicide-related reporting in Canada; its impact on media reporting quality and suicide deaths is unknown. METHOD: This pre-post observational study examined changes in reporting characteristics in a random sample of suicide-related articles from major publications in the Greater Toronto Area (GTA) media market. Articles (n = 900) included 450 from the 6-year periods prior to and after the initiative began. We also examined changes in suicide counts in the GTA between these epochs. We used chi-square tests to analyse changes in reporting characteristics and time-series analyses to identify changes in suicide counts. Secondary outcomes focused on guidelines developed by media professionals in Canada and how they may have influenced media reporting quality as well as on the overarching narrative of media articles during the most recent years of available data. RESULTS: Across-the-board improvement was observed in suicide-related reporting with substantial reductions in many elements of putatively harmful content and substantial increases in all aspects of putatively protective content. However, overarching article narratives remained potentially harmful with 55.2% of articles telling the story of someone's death and 20.8% presenting an other negative message. Only 3.6% of articles told a story of survival. After controlling for potential confounders, a nonsignificant numeric decrease in suicide counts was identified after initiative implementation (ω = -5.41, SE = 3.43, t = 1.58, p = 0.12). CONCLUSIONS: We found evidence that a strategy to engage media in Canada changed the content of reporting, but there was only a nonsignificant trend towards fewer suicides. A more fundamental change in media narratives to focus on survival rather than death appears warranted.

A Dose of Reality About Dose-Response Relationships.

Observational research can be strengthened by examining potential dose-response relationships that correlate a clinical intervention with a patient outcome. Despite being a classic criterion for establishing causality, dose-response testing can be difficult to interpret in clinical medicine due to multiple diverse pitfalls. This review introduces a cautionary framework for investigators considering dose-response relationships in observational research to support evidence-based medicine. Each pitfall is illustrated with a specific example relevant when analyzing a dose-response relationship. Several pitfalls stem from faulty interpretation including confounding by indication and fallible range selection. Additional pitfalls relate to improper analysis including fitting a nonlinear model and misclassification error. Further pitfalls arise in special situations including subjective self-report and artifacts from survival bias. These caveats are common sources of misunderstanding in analyses that examine the link between varying exposures and the intensity of clinical outcomes. Awareness of specific pitfalls, we suggest, might help advance the conduct, application, and translation of dose-response relationships in observational research to inform evidence-based medical care.

COVID-19 deaths on weekends.

BACKGROUND: Mortality statistics about daily deaths might change on weekends due to delays in reporting, uneven staffing, a different mix of personnel, or decreased efficiency. We hypothesized that reported deaths for COVID-19 might increase on weekends compared to weekdays. METHODS: We collected data from the World Health Organization COVID-19 database. All deaths from March 7, 2020 to March 7, 2022 were included (two years). The primary analysis evaluated mean daily deaths on weekends compared to the preceding five workdays. Analyses were replicated in ten individual countries: United States, United Kingdom, France, Germany, Italy, Spain, Russia, India, Brazil, and Canada. RESULTS: The mean COVID-19 daily deaths was higher on weekends compared to weekdays (8,532 vs. 8,083 p < 0.001), equal to a 6% relative increase (95% confidence interval 3% to 8%). The highest absolute increase was in the United States (1,483 vs. 1,220 deaths, p < 0.001). The second highest absolute increase was in Brazil (1,061 vs. 823 deaths, p < 0.001). The increase in deaths on weekends remained significant during the earlier and later months of the pandemic, as well as during the greater and lesser weeks of the pandemic. CONCLUSIONS: The apparent increased COVID-19 deaths reported on weekends might potentially reflect patient care, confound community trends, and affect the public perception of risk.

Is the narrative the message? The relationship between suicide-related narratives in media reports and subsequent suicides.

OBJECTIVES: When journalists report on the details of a suicide, the way that they contextualize the meaning of the event (i.e. the 'narrative') can have significant consequences for readers. The 'Werther' and 'Papageno' narrative effects refer to increases and decreases in suicides across populations following media reports on suicidal acts or mastery of crises, respectively. The goal of this study was to investigate the impact of these different narrative constructs on subsequent suicides. METHODS: This study examined the change in suicide counts over time in Toronto, Canada. It used latent difference score analysis, examining suicide-related print media reports in the Toronto media market (2011-2014). Articles (N = 6367) were coded as having a potentially harmful narrative if they described suicide in a celebrity or described a suicide death in a non-celebrity and included the suicide method. Articles were coded as having potentially protective narratives if they included at least one element of protective content (e.g. alternatives to suicide) without including any information about suicidal behaviour (i.e. suicide attempts or death). RESULTS: Latent difference score longitudinal multigroup analyses identified a dose-response relationship in which the trajectory of suicides following harmful 'Werther' narrative reports increased over time, while protective 'Papageno' narrative reports declined. The latent difference score model demonstrated significant goodness of fit and parameter estimates, with each group demonstrating different trajectories of change in reported suicides over time: (χ2[6], N = 6367) = 13.16; χ2/df = 2.19; Akaike information criterion = 97.16, comparative fit index = 0.96, root mean square error of approximation = 0.03. CONCLUSION: Our findings support the notion that the 'narrative' matters when reporting on suicide. Specifically, 'Werther' narratives of suicides in celebrities and suicides in non-celebrities where the methods were described were associated with more subsequent suicides while 'Papageno' narratives of survival and crisis mastery without depictions of suicidal behaviours were associated with fewer subsequent suicides. These results may inform efforts to prevent imitation suicides.

Bans on Cellphone Use While Driving and Traffic Fatalities in the United States.

BACKGROUND: As of January 2020, 18 of 50 US states comprehensively banned almost all handheld cellphone use while driving, 3 states and the District of Columbia banned calling and texting, 27 states banned texting on a handheld cellphone, and 2 states had no general cellphone ban for all drivers. However, it remains unknown whether these bans were associated with fewer traffic deaths and whether comprehensive handheld bans are more effective than isolated calling or texting bans. We evaluated whether cellphone bans were associated with fewer driver, non-driver, and total fatalities nationally. METHODS: We conducted a longitudinal panel analysis of traffic fatality rates by state, year, and quarter. Population-based rate ratios and 95% CIs were estimated comparing state-quarters with and without cellphone bans. RESULTS: From 1999 through 2016, 616,289 persons including 344,003 drivers died in passenger vehicle crashes in the United States. Relative to no ban, comprehensive handheld bans were associated with lower driver fatality rates (adjusted rate ratio aRR = 0.93, 95% CI = 0.90, 0.97) but not for non-driver fatalities (aRR = 1.01, 95% CI = 0.95, 1.07) or total fatalities (aRR = 0.98, 95% CI = 0.94, 1.01). We found no differences in driver fatalities for calling-only bans (aRR = 1.00, 95% CI = 0.97, 1.03), texting-only bans (aRR = 1.02, 95% CI = 0.99, 1.05), texting plus phone-manipulating bans (aRR = 0.99, 95% CI = 0.93, 1.04), or calling and texting bans (aRR = 0.98, 95% CI = 0.88, 1.09). CONCLUSIONS: Comprehensive handheld bans were associated with fewer driver fatalities.

Understanding Patient Personality in Medical Care: Five-Factor Model.

Personality is the description of an individual's tendencies when acting or reacting to others. Clinicians spontaneously form impressions of a patient's apparent personality yet such unstructured impressions might lead to snap judgments or unhelpful labels. Here we review the evidence-based five-factor model from psychology science for understanding personalities (OCEAN taxonomy). Openness to experience is defined as the general appreciation for a variety of experiences. Conscientiousness is the tendency to exhibit self-discipline. Extraversion is the degree of engagement with the external world. Agreeableness is the general concern for social harmony. Neuroticism is the tendency to experience negative emotions. An awareness of these five dimensions might help clinicians avoid faulty judgments from casual contact. Expert assessment of personality requires extensive training and data, thereby suggesting that clinicians should take a humble view of their own unsophisticated impressions of a patient's personality.

Identifying Risk Factors for Diabetic Ketoacidosis Associated with SGLT2 Inhibitors: a Nationwide Cohort Study in the USA.

INTRODUCTION: Sodium glucose co-transporter-2 inhibitors (SGLT2) are commonly prescribed to patients with type 2 diabetes mellitus, but can increase the risk of diabetic ketoacidosis. Identifying patients prone to diabetic ketoacidosis may help mitigate this risk. METHODS: We conducted a population-based cohort study of adults initiating SGLT2 inhibitor use from 2013 through 2017. The primary objective was to identify potential predictors of diabetic ketoacidosis. Two machine-learning methods were applied to model high-dimensional pre-exposure data: gradient boosted trees and least absolute shrinkage and selection operator (LASSO) regression. We rank ordered the variables produced from LASSO by the size of their estimated coefficient (largest to smallest). With gradient boosted trees, a relative importance measure for each variable is provided rather than a coefficient. The "top variables" were identified after reviewing the distributions of the effect estimates from LASSO and gradient boosted trees to identify where there was a substantial decrease in variable importance. The identified predictors were then assessed in a logistic regression model and reported as odds ratios (ORs) with 95% confidence intervals (CIs). RESULTS: We identified 111,442 adults who started SGLT2 inhibitor use. The mean age was 57 years, 44% were female, the mean hemoglobin A1C was 8.7%, and the mean creatinine was 0.89 mg/dL. During a mean follow-up of 180 days, 192 patients (0.2%, i.e., 2 per 1000) were diagnosed and hospitalized with diabetic ketoacidosis (DKA) and 475 (0.4%, i.e., 4 per 1000) were diagnosed in either an inpatient or outpatient setting. Using gradient boosted trees, the strongest predictors were prior DKA, baseline hemoglobin A1C level, baseline creatinine level, use of medications for dementia, and baseline bicarbonate level. Using LASSO regression not including laboratory test results due to missing data, the strongest predictors were prior DKA, digoxin use, use of medications for dementia, and recent hypoglycemia. The logistic regression model incorporating the variables identified from gradient boosted trees and LASSO regression suggested the following pre-exposure characteristics had the strongest association with a hospitalization for DKA: use of dementia medications (OR = 7.76, 95% CI 2.60, 23.1), prior intracranial hemorrhage (OR = 11.5, 95% CI 1.46, 91.1), a prior diagnosis of hypoglycemia (OR = 5.41, 95% CI 1.92,15.3), prior DKA (OR = 2.45, 95% CI 0.33, 18.0), digoxin use (OR = 4.00, 95% CI 1.21, 13.2), a baseline hemoglobin A1C above 10% (OR = 3.14, 95% CI 1.95, 5.06), and baseline bicarbonate below 18 mmol/L (OR 5.09, 95% CI 1.58, 16.4). CONCLUSION: Diabetic ketoacidosis affected approximately 2 per 1000 patients starting to use an SGLT2 inhibitor. We identified both anticipated, e.g., low baseline serum bicarbonate, and unanticipated, e.g., digoxin, dementia medications, risk factors for SGLT2 inhibitor-induced DKA.

Sodium-glucose co-transporter-2 inhibitors versus dipeptidyl peptidase-4 inhibitors and the risk of heart failure: A nationwide cohort study of older adults with diabetes mellitus.

AIMS: To analyse the rate of heart failure hospitalization for older adults prescribed a sodium-glucose co-transporter-2 (SGLT2) inhibitor. MATERIALS AND METHODS: The study cohort included adults aged 66 years and older diagnosed with diabetes mellitus in Ontario, Canada, between July 2015 and March 2019, who received either an SGLT2 inhibitor or a dipeptidyl peptidase-4 (DPP-4) inhibitor. The primary outcome was a composite of heart failure hospitalization and all-cause mortality. Secondary outcomes included diabetic ketoacidosis and hypoglycaemia. RESULTS: A total of 29 916 adults prescribed an SGLT2 inhibitor were compared with 29 916 adults prescribed a DPP-4 inhibitor. The mean age was 72 years, 60% were men, the baseline glycated haemoglobin concentration was 8.2% and the baseline creatinine was 89 µmol/L. The incidence rate of the primary outcome was 19/1000 person-years for adults prescribed an SGLT2 inhibitor compared to 38/1000 person-years in those prescribed a DPP-4 inhibitor. This resulted in a hazard ratio (HR) of 0.49 (95% confidence interval [CI] 0.45, 0.54) and a rate difference (RD) of 19 fewer events per 1000 person-years (RD -19 [95% CI -22, -17]). Patients prescribed an SGLT2 inhibitor also had a lower rate of hypoglycaemia (HR 0.61 [95% CI 0.46, 0.81); RD -1.6 [95% CI -2.4, -0.8]), but a higher rate of diabetic ketoacidosis (HR 1.84 [95% CI 1.26, 2.70]; RD 1.0 [95% CI 0.4, 1.6]). CONCLUSIONS: Older adults prescribed an SGLT2 inhibitor had a lower rate of heart failure hospitalization or death, and a lower rate of hypoglycaemia, but an increased rate of diabetic ketoacidosis compared to older adults prescribed a DPP-4 inhibitor.

Comparative effectiveness and safety of sodium-glucose cotransporter-2 inhibitors versus metformin in patients with type 2 diabetes: An observational study using data from routine care.

AIM: To assess the effectiveness and safety of sodium-glucose cotransporter-2 (SGLT2) inhibitors in treatment-naïve patients compared with metformin. PARTICIPANTS AND METHODS: We conducted a cohort study of US adults with type 2 diabetes mellitus who had not filled a prescription for a diabetes medication in the preceding year. We then identified patients who newly filled a prescription for an SGLT2 inhibitor or metformin between 2013 and 2018. The primary outcome was a composite of heart failure, myocardial infarction or stroke. Safety outcomes included hypoglycaemia, diabetic ketoacidosis, genital infection, lactic acidosis and acute kidney injury. After 1:1 propensity-score (PS) matching, proportional hazards models were fit to estimate hazard ratios (HRs) with 95% confidence intervals (CIs). RESULTS: We identified 9964 individuals newly prescribed an SGLT2 inhibitor who were PS-matched to 9964 individuals newly prescribed metformin. The mean age was 54 years, 52% were women, and the duration of follow-up was 213 days for metformin and 147 days for SGLT2 inhibitors. The primary outcome occurred in 54 patients (7.2 events per 1000 person-years) who received an SGLT2 inhibitor, compared to 84 patients (8.5 per 1000 person-years) who received metformin (HR  0.82, 95% CI 0.58, 1.15). Similar results (HR 0.87, 95% CI 0.69, 1.09) were observed in an analysis with longer follow-up (ie, approximately 600 days). The rates of genital infection (HR 2.28, 95% CI 1.87, 2.78) and diabetic ketoacidosis (HR 1.58, 95% CI 0.92, 2.70) were higher for patients prescribed an SGLT2 inhibitor compared to metformin, while the rates of acute kidney injury (HR 0.94, 95% CI 0.60, 1.47) or hypoglycaemia (HR 0.83, 95% CI 0.48, 1.42) were not. CONCLUSIONS: We observed a numerically lower rate of short-/mid-term cardiovascular events for patients newly prescribed an SGLT2 inhibitor compared to metformin, albeit with wide CIs that include the possibility of a null effect. SGLT2 inhibitors were associated with a higher rate of genital infection and diabetic ketoacidosis. Larger cohort studies and long-term clinical trials powered to assess cardiovascular events are necessary to understand the risk-benefit profile of SGLT2 inhibitors as first-line therapy for adults with type 2 diabetes mellitus.

The association between Twitter content and suicide.

OBJECTIVE: A growing body of research has established that specific elements of suicide-related news reporting can be associated with increased or decreased subsequent suicide rates. This has not been systematically investigated for social media. The aim of this study was to identify associations between specific social media content and suicide deaths. METHODS: Suicide-related tweets (n = 787) geolocated to Toronto, Canada and originating from the highest level influencers over a 1-year period (July 2015 to June 2016) were coded for general, putatively harmful and putatively protective content. Multivariable logistic regression was used to examine whether tweet characteristics were associated with increases or decreases in suicide deaths in Toronto in the 7 days after posting, compared with a 7-day control window. RESULTS: Elements independently associated with increased subsequent suicide counts were tweets about the suicide of a local newspaper reporter (OR = 5.27, 95% CI = [1.27, 21.99]), 'other' social causes of suicide (e.g. cultural, relational, legal problems; OR = 2.39, 95% CI = [1.17, 4.86]), advocacy efforts (OR = 2.34, 95% CI = [1.48, 3.70]) and suicide death (OR = 1.52, 95% CI = [1.07, 2.15]). Elements most strongly independently associated with decreased subsequent suicides were tweets about murder suicides (OR = 0.02, 95% CI = [0.002, 0.17]) and suicide in first responders (OR = 0.17, 95% CI = [0.05, 0.52]). CONCLUSIONS: These findings largely comport with the theory of suicide contagion and associations observed with traditional news media. They specifically suggest that tweets describing suicide deaths and/or sensationalized news stories may be harmful while those that present suicide as undesirable, tragic and/or preventable may be helpful. These results suggest that social media is both an important exposure and potential avenue for intervention.

Fracture Risk After Initiation of Use of Canagliflozin: A Cohort Study.

Background: Sodium-glucose cotransporter-2 inhibitors promote glycosuria, resulting in possible effects on calcium, phosphate, and vitamin D homeostasis. Canagliflozin is associated with decreased bone mineral density and a potential increased risk for fracture. Objective: To estimate risk for nonvertebral fracture among new users of canagliflozin compared with a glucagon-like peptide-1 (GLP-1) agonist. Design: Population-based new-user cohort study. Setting: Two U.S. commercial health care databases providing data on more than 70 million patients from March 2013 to October 2015. Patients: Persons with type 2 diabetes who initiated use of canagliflozin were propensity score-matched in a 1:1 ratio to those initiating use of a GLP-1 agonist. Measurements: The primary outcome was a composite end point of humerus, forearm, pelvis, or hip fracture requiring intervention. Secondary outcomes included fractures at other sites. A fixed-effects meta-analysis that pooled results from the 2 databases provided an overall hazard ratio (HR). Results: 79 964 patients initiating use of canagliflozin were identified and matched to 79 964 patients initiating use of a GLP-1 agonist. Mean age was 55 years, 48% were female, average baseline hemoglobin A1c level was 8.7%, and 27% were prescribed insulin. The rate of the primary outcome was similar for canagliflozin (2.2 events per 1000 person-years) and GLP-1 agonists (2.3 events per 1000 person-years), with an overall HR of 0.98 (95% CI, 0.75 to 1.26). Risk for pelvic, hip, humerus, radius, ulna, carpal, metacarpal, metatarsal, or ankle fracture was also similar for canagliflozin (14.5 events per 1000 person-years) and GLP-1 agonists (16.1 events per 1000 person-years) (overall HR, 0.92 [CI, 0.83 to 1.02]). Limitation: Unmeasured confounding, measurement error, and low fracture rate. Conclusion: In this study of middle-aged patients with type 2 diabetes and relatively low fracture risk, canagliflozin was not associated with increased risk for fracture compared with GLP-1 agonists. Primary Funding Source: Brigham and Women's Hospital, Division of Pharmacoepidemiology and Pharmacoeconomics.

Challenges to Human Drivers in Increasingly Automated Vehicles.

OBJECTIVE: We examine the relationships between contemporary progress in on-road vehicle automation and its coherence with an envisioned "autopia" (automobile utopia) whereby the vehicle operation task is removed from all direct human control. BACKGROUND: The progressive automation of on-road vehicles toward a completely driverless state is determined by the integration of technological advances into the private automobile market; improvements in transportation infrastructure and systems efficiencies; and the vision of future driving as a crash-free enterprise. While there are many challenges to address with respect to automated vehicles concerning the remaining driver role, a considerable amount of technology is already present in vehicles and is advancing rapidly. METHODS: A multidisciplinary team of experts met to discuss the most critical challenges in the changing role of the driver, and associated safety issues, during the transitional phase of vehicle automation where human drivers continue to have an important but truncated role in monitoring and supervising vehicle operations. RESULTS: The group endorsed that vehicle automation is an important application of information technology, not only because of its impact on transportation efficiency, but also because road transport is a life critical system in which failures result in deaths and injuries. Five critical challenges were identified: driver independence and mobility, driver acceptance and trust, failure management, third-party testing, and political support. CONCLUSION: Vehicle automation is not technical innovation alone, but is a social as much as a technological revolution consisting of both attendant costs and concomitant benefits.

Practicing Medicine with Colleagues: Pitfalls from Social Psychology Science.

This perspective reviews three pitfalls from psychology science that can distort clinical assessments and contribute to interpersonal conflicts. One pitfall is the illusion that one's own subjective perceptions or judgments are objective observations or interpretations that reasonable colleagues would share. A second pitfall involves self-serving situational attributions rather than disposition attributions for explaining missteps after things go wrong. A third pitfall is confirmation bias that leads to a perseverance of erroneous beliefs, a tendency to mostly seek supportive colleagues, and a failure to check for dissenting viewpoints. An awareness of these three pitfalls may help clinicians improve patient care when practicing with colleagues.

The association of continuity of care and cancer centre affiliation with outcomes among patients with cancer who require emergency department care.

BACKGROUND: Patients with cancer have complex care requirements and frequently use the emergency department. The purpose of this study was to determine whether continuity of care, cancer expertise of an institution or both affect outcomes in patients with cancer in the emergency setting. METHODS: We conducted a retrospective cohort study using administrative databases from Ontario, Canada, involving records of patients aged 20 years and older who received chemotherapy or radiation in the 30 days before a cancer-related visit to the emergency department between 2006 and 2011. Patients seen in an emergency department at an alternative hospital (not the site where cancer treatment was given) were matched based on propensity score to patients who visited their original hospital (site where cancer treatment was given). Next, patients seen at an alternative emergency department that was in a general hospital (i.e., not a cancer centre) were matched to patients who visited their original hospital or a cancer centre. Outcomes were admission to hospital at the index visit to the emergency department, 30-day mortality, having imaging with computed tomography and return visits to the emergency department. RESULTS: We found 42 820 patients who were eligible for our study. Patients seen in the emergency departments at alternative hospitals were less likely to be admitted to hospital (odds ratio [OR] 0.78, 95% confidence interval [CI] 0.74-0.83) and had higher hazards of return visits to the emergency department than matched patients at original hospitals (hazard ratio [HR] 1.06, 95% CI 1.03-1.11). In comparison, patients at alternative general hospitals also had lower odds of admission to hospital (OR 0.83, 95% CI 0.79-0.88) and higher hazards of return visits to the emergency department (HR 1.07, 95% CI 1.03-1.11) compared with matched counterparts; however, these patients had higher 30-day mortality (OR 1.13, 95% CI 1.05-1.22) and lower odds of having CT imaging (OR 0.74, 95% CI 0.69-0.80). INTERPRETATION: Cancer expertise of an institution rather than continuity of care may be an important predictor of outcomes following emergency treatment of patients with cancer.

Traumatic brain injury as an independent risk factor for problem gambling: a matched case-control study.

PURPOSE: To assess whether traumatic brain injury (TBI) increases the risks of subsequent problem gambling. METHODS: We conducted a matched case-control analysis of adults in Ontario, Canada. The study included those who self-reported their gambling activities in the Canadian Community Health Survey 2007-2008. Using Problem Gambling Severity Index, we defined cases as those who were problem gamblers and controls who were recreational gamblers. Cases were matched to controls 1:2 using propensity scores based on demographics, prior mental health, and self-reported behaviours. The main predictor was prior TBI defined as requiring emergency care and identified using ICD-10 codes from administrative health databases. We estimated the likelihood of prior TBI in problem gamblers compared to controls using conditional logistic regression. RESULTS: Of 30,652 survey participants, 16,002 (53%) reported gambling activity of whom 14,910 (49%) were recreational gamblers and 4% (n = 1092) were problem gamblers. A total of 1469 respondents (5%) had a prior TBI. Propensity score matching yielded 2038 matched pairs with 1019 cases matched to 2037 controls. Case-control analysis showed a significant association between prior TBI and subsequent problem gambling (odds ratio 1.27, 95% confidence interval 1.07-1.51, P = 0.007). The increased risk was mostly apparent in men aged 35 to 64 years who reported alcohol use or smoking. The relative risk of problem gambling in those with two or more TBIs equated to an odds ratio of 2.04 (95% confidence interval 1.05-3.99). CONCLUSIONS: We found that a prior TBI was associated with an increased subsequent risk of problem gambling. Our findings support more awareness, screening, and treating problem gambling risks among TBI patients.

The Limits of Clinician Vigilance as an AI Safety Bulwark.

This Viewpoint examines the potential problems of clinician reliance on the use of artificial intelligence (AI) in health care and offers suggestions on how AI could be designed to promote clinician vigilance.