BackgroundAromatase inhibitors (AIs) are the preferred treatment for postmenopausal women with hormone receptor-positive (HR+) breast cancer, but their cardiovascular safety remains uncertain. This study compared cardiovascular risks associated with AIs versus tamoxifen in two distinct populations.Patients and methodsThis retrospective cohort study used data from the US SEER-Medicare database (2007–2020) and Taiwan’s National Health Insurance claims database and cancer registry (2010–2021). Postmenopausal women with HR+ stage I–III breast cancer were categorized by their initial endocrine therapy. Study outcomes included major adverse cardiovascular events (MACE), venous thromboembolism (VTE), and heart failure (HF). Covariate balance was achieved through inverse probability of treatment weighting, and cause-specific hazard models were used for risk assessment in each cohort.ResultsThe study included 36,950 US patients (89% AI users) and 22,676 Taiwan patients (71% AI users). In the US, AIs were associated with an HR of 1.13 (95% CI 0.83–1.52) for MACE and 1.17 (95% CI 0.93–1.48) for HF, compared to tamoxifen. In Taiwan, the HRs were 1.23 (95% CI 0.87–1.73) for MACE and 0.93 (95% CI 0.73–1.21) for HF. AI use was linked to a lower VTE risk in the US (HR: 0.35, 95% CI 0.27–0.45), while the Taiwan cohort showed a non-significant trend (HR: 0.72, 95% CI 0.42–1.25).ConclusionAIs were not associated with a definitive increase in MACE or HF risk compared with tamoxifen, although confidence intervals suggest some uncertainty in the estimates. The observed differences in VTE risk across cohorts highlight the need for population-specific considerations when selecting endocrine therapy.

PurposeTo examine the accuracy of the diagnoses of early-onset neonatal infection in the Danish National Patient Register.Patients and MethodsAll Danish neonates born after 35 weeks of gestation with an ICD10 diagnosis of sepsis or meningitis within the first week of life between 2010 and 2018 were identified. To examine the validity of the diagnoses, medical records were reviewed for 500 neonates diagnosed with sepsis and all 63 neonates diagnosed with meningitis. Proven infection was defined by bacterial pathogens identified from blood or cerebrospinal fluid. Probable sepsis was defined as at least 5 days of antibiotic treatment, two clinical signs of infection, and one abnormal blood biomarker. Probable meningitis was defined as at least 14 days of antibiotics, neurological affection, and cerebrospinal fluid analysis consistent with meningitis. Positive predictive values (PPVs) with 95% CI were estimated using the proven definition of infection as well as the criteria for either proven or probable infection.ResultsThe PPV for sepsis diagnoses using the proven definition was 0.04 (0.02, 0.06), while the PPV using both the proven and probable definition was 0.52 (0.45, 0.58). Sepsis diagnoses with a specified pathogen showed the highest PPV for proven sepsis at 0.36 (0.29, 0.43) and for proven or probable sepsis at 0.65 (0.58, 0.72). The PPV for meningitis diagnoses using the proven definition was 0.24 (0.13, 0.37), while the PPV using both the proven and probable definition was 0.38 (0.25, 0.52).ConclusionWe found a poor validity for the ICD10 diagnoses of early-onset neonatal infection in the Danish National Patient Register. This highlights that the diagnoses should be used cautiously in future research and surveillance and emphasises the need for supporting information on microbiological samples. Additionally, this highlights the importance of establishing a consensus definition of neonatal sepsis to improve diagnostic accuracy.

IntroductionStudying patterns of death, particularly premature deaths (<75 years), provides insights to address health inequities among those living. Multiple coding systems for cause of death (COD) exist. The Leading Causes of Death (LCD) scheme is designed for identifying priority COD for interventions in global populations. The extent to which such classification is effective for identifying priority causes of premature mortality among subpopulations with chronic health conditions, such as inflammatory bowel disease (IBD), is unknown.ObjectiveTo evaluate the usability of the LCD for characterizing premature mortality among those with IBD.MethodsWe conducted a population-based matched case control study of persons with IBD who died between 2010 and 2018 using linked health administrative data from Ontario, Canada. Individuals with IBD were matched with five decedents without IBD based on sex and years of birth and death. We compared COD for premature and overall mortality using two classification structures: the LCD scheme and the International Statistical Classification of Diseases and Related Health Problems, tenth revision (ICD-10) chapters.ResultsAmong 7,919 decedents with IBD (39,414 matched controls), 47% died prematurely. With the LCD framework, COD differences for premature mortality were not detectable as 29% were allocated to the residual category (Standardized differences [SD]: 18%). Most residual deaths were due to neoplasms (34%) or diseases of the gastrointestinal system (32%). Using ICD-10 chapters, premature deaths were more commonly due to diseases of the digestive system than for matched controls (13% vs 5%, SD: 31%).DiscussionThe LCD coding scheme provides more granular COD details compared to the ICD-10 chapters. However, a larger proportion of deaths among people with IBD were allocated to the residual category, limiting its utility for enabling healthcare systems to identify priority targets to reduce premature mortality. Further work to develop and validate a framework for premature COD classification in populations with IBD is needed.

PurposeTo compare the effect estimates obtained from patients with different levels of electronic health record (EHR)-continuity in four empirical studies: risk of pneumonia among 1) new users of proton pump inhibitors (PPI) vs H2 receptor antagonists, or 2) new users of PPI vs non-PPI users; risk of major bleeding among 3) new users of warfarin vs direct-acting oral anticoagulants, or 4) new users of oral anti-coagulants (OAC) vs non-OAC users.Patients and MethodsPatients were identified in 2 US EHR systems (MA system, NC system) linked with Medicare claims data (2007/1/1 – 2014/12/31) separately. We calculated incidence rates (IR), incidence rate differences (IRD), and hazard ratios (HR) in the total linked study population and after excluding patients with the lowest 25%, 50%, or 75% of EHR-continuity scores. We quantified bias in IRD and propensity score (PS) decile-adjusted HR.ResultsIn the MA system, IRs based on EHR-only data underestimated true rates by 44.1% to 76.2%, reduced to 12.9% to 46.5%. After excluding the lowest 75% of EHR-continuity patients, underestimation was more pronounced in non-user comparator designs. Absolute IRD bias was small for PPI vs H2RA (0.4%) and warfarin vs DOAC (0.7%), but larger for PPI vs non-PPI (19.1%) and OAC vs non-OAC (7.8%). Relative HR bias was 13.3% (PPI vs H2RA), 18.9% (PPI vs non-PPI), 3.0% (warfarin vs DOAC), and 31.5% (OAC vs non-OAC). Excluding lower-continuity patients and PS adjustment reduced IRD bias, while restricting to higher-continuity patients modestly improved HR bias.ConclusionLimiting analyses to patients with higher EHR-continuity can reduce IR underestimation and bias in effect estimates, particularly on the IRD scale. While PS adjustment mitigates some bias, EHR-discontinuity remains a source of bias, especially in studies using non-user comparators. These findings underscore the importance of balancing EHR-continuity with sample size considerations in pharmacoepidemiologic research.

PurposeThyroid cancer (TC) survivors face an increased risk of second primary cancers (SPCs). While previous research on the risk of SPCs among TC patients has been conducted in other regions, comprehensive data from Northern Europe remain limited. Our register-based cohort study aimed to evaluate the risk of SPCs among Finnish TC survivors.Patients and MethodsWe obtained data on all patients diagnosed with TC in Finland in 1953–2022 from the Finnish Cancer Registry (FCR). Standardized incidence ratios (SIRs) of SPCs were calculated relative to the risk of cancer in the general population.ResultsWe identified 14,520 patients with TC (3,212 males, 22%; 11,308 females, 78%), with a median follow-up of 12.1 years. A metachronous SPC was diagnosed in 16.4% (n=527) of males and 15.5% (n=1,744) of females. The overall SPC risk was elevated for both males (SIR 1.23, 95% CI 1.12–1.34; EAR 2.39/1,000 PYs) and females (1.24, 1.19–1.30; EAR 1.92/1,000 PYs). Female patients with papillary TC (PTC) displayed increased SIRs for breast (SIR 1.26, 95% CI:1.15–1.37), urinary organ (1.50, 1.19–1.86), brain (1.85, 1.48–2.29), and hematolymphoid cancers (1.31, 1.10–1.54). Males with PTC showed an increased risk of urinary organ (SIR 1.39, 1.01–1.85), brain (2.73, 1.69–4.17), and hematolymphoid cancers (1.52, 1.12–2.01). The elevated SPC risk of the breast and brain persisted even after 20 years of follow-up. An increased incidence of SPCs of the urinary organs, brain, and hematolymphoid tissues was observed in females diagnosed with follicular TC (FTC) or medullary TC (MTC) in 1953–1985.ConclusionOur results indicate an excess risk of cancers of the breast, urinary organs, brain and hematolymphoid tissues for Finnish patients with a history of PTC. For FTC and MTC risk of cancers of the urinary organs, brain, and hematolymphoid tissues is specific for female patients diagnosed in 1953–1985.

PurposeSmall cell lung cancer (SCLC) is a highly aggressive malignancy with poor prognosis. Studies capturing the impact of recently approved immunotherapies are limited, highlighting a knowledge gap regarding their real-world use and effectiveness. Patients and MethodsWe examined data from 29949 patients with extensive-stage (ES)-SCLC across observational studies from the United States, United Kingdom, Spain, Taiwan, South Korea, and Japan to describe the clinical course of ES-SCLC. Data sources included electronic health record databases, registries, and claims data over time periods ranging between 2013 and 2023. Patient characteristics, recent treatment patterns, and real-world overall survival (rwOS) were assessed in each country.ResultsThe most common first-line (1L) treatment was platinum plus etoposide without anti-PD-L1 agents (59–89%), followed by platinum plus etoposide with anti-PD-L1 agents (9–38%). Second-line (2L) and third-line (3L) treatments varied widely across countries. Median rwOS ranged from 8.1–11.3 months following 1L initiation, 4.8–6.9 months following 2L, and 4.1–5.5 months following 3L. Patients receiving compared to those not receiving 1L anti-PD-L1 therapy showed numerically higher median rwOS following 1L initiation, with no meaningful difference in rwOS following 2L or 3L therapy.ConclusionIn our evaluation of real-world treatment patterns and outcomes among patients with ES-SCLC from six countries, we found that rwOS in 1L, 2L and 3L was consistently poor across countries, despite differences in patient characteristics and treatment patterns. These findings may support the generalizability of clinical evidence across geographies and highlight the need for further research to optimize treatment strategies to improve patient outcomes globally.

IntroductionMild traumatic brain injury (mTBI) is an important public health issue, but France does not have a dedicated epidemiological surveillance system for mTBI. Data on Emergency departments (ED) of the OSCOUR network could be useful for setting up a dedicated mTBI surveillance system. However, the performance of potential algorithms based on ICD-10 codes for identifying cases of mTBI in OSCOUR has not been assessed. The objective of this study is to measure the performance of various potential algorithms, based on ICD-10 codes, for identifying mTBI ED visits in the OSCOUR database.Material and MethodsWe performed a retrospective multicenter validation study of algorithms for identifying mTBI based on ICD-10 codes using the OSCOUR database. We calculated sensitivity, specificity, positive and negative predictive values of the various algorithms by using medical charts from ED visits as a reference source. Our study population consisted of a random sample of patients of all ages in France who visited one of the four ED from the OSCOUR network, which participated in our study in 2019.Results5,185 medical charts were reviewed. Algorithms performance varied according to population characteristics, and none of the algorithms tested for the identification of mTBI cases achieved the minimum performance requirements (sensitivity and PPV ≥ 80%) over all age or sex groups. However, sub-group analyses highlighted that one algorithm (BA31_OPT1) had acceptable performance for identifying mTBI according to our “broad” definition for people under 18 years old. Sensitivity, specificity, PPV and NPV for this algorithm were 85.2%, 99.4%, 95.2% and 98%, respectively.ConclusionMost mTBI case identification algorithms performed poorly in identifying mTBI cases of all ages in the OSCOUR database. Nevertheless, it was possible to identify cases defined according to a “broad” mTBI definition in the paediatric population (0–17 years).

BackgroundSeveral prognostic models have been developed to estimate operative mortality in patients undergoing surgery for infective endocarditis (IE). However, their external validity and performance remain uncertain, limiting clinical applicability. This study aimed to externally validate and compare the performance of contemporary IE-specific and generic cardiac surgery (EuroSCORE II) risk scores in a large single-center cohort.MethodsEighteen operative IE-specific risk scores, along with EuroSCORE II, were retrospectively applied to a database of 689 patients undergoing cardiac surgery for IE. Discrimination was evaluated using the area under the receiver operating characteristic curve (AUC), while calibration was assessed using the Hosmer–Lemeshow test, Brier score, and calibration slopes/intercepts. For each score, the type of validation reported in the original study was critically examined, noting that validation was not always performed. Additionally, the inclusion of IE-specific variables, such as pathogen type and valvular complications, was assessed to evaluate the reliability and clinical applicability of each score.ResultsAmong the 689 patients, 30% were female, with a median age of 69 years. The most frequent pathogens were Streptococcus (26%), Staphylococcus aureus (18%), coagulase-negative staphylococci (18%), and Enterococcus faecalis (16%). Operative mortality was 10.6% (n = 73). The RISK-E score showed the highest discrimination (AUC: 0.742), followed by APORTEI (0.734) and modified MELD-XI (0.730). All scores demonstrated good calibration, with scaled Brier scores above 0.8. Scores incorporating IE-specific variables generally performed better, while several widely used generic scores, including EuroSCORE II, overestimated operative risk. External validation revealed lower AUCs for many scores compared to original reports, highlighting the importance of rigorous evaluation.ConclusionThe RISK-E score demonstrated the highest discriminative ability and satisfactory calibration for predicting operative mortality in patients undergoing surgery for infective endocarditis. These results support the role of externally validated, IE-specific prognostic tools in guiding clinical assessment and selecting appropriate perioperative strategies.