Hostile aortic neck anatomy is associated with loss of proximal seal and increased late reinterventions. Although both EndoSuture aneurysm repair (ESAR) and fenestrated endovascular aortic repair (FEVAR) are commercially available options for treatment of short neck aneurysms, branch vessel patency is a potential tradeoff for improved seal with FEVAR owing to the incorporation of renovisceral vessels. This study compares the performance of ESAR vs FEVAR in hostile aortic necks.

Patients who underwent elective ESAR or FEVAR for hostile neck AAAs at a single center from 2012 to 2024 were reviewed retrospectively. Exclusion criteria included pararenal or thoracoabdominal aortic aneurysm, off-label modifications, and nonstandard FEVAR configurations. Propensity matching weights were generated based on age, year of operation, preoperative estimated glomerular filtration rate, neck length, neck diameter, and infrarenal angulation. Rates of survival, reintervention, dialysis, chronic kidney disease stage progression, type IA endoleak (EL), and sac regression (>5 mm) were assessed at latest follow-up.

Of 391 patients, 60 with ESAR and 207 with FEVAR were included. FEVAR patients were younger (74.4 years vs 79.8 years; P < .001) with larger neck diameters (25.0 mm vs 23.6 mm; P = .016), shorter neck length (5.0 mm vs 9.8 mm; P < .001), and decreased infrarenal angulation (20° vs 40°; P < .001). After propensity score-adjusted regression (58 ESAR, 169 FEVAR), FEVAR, compared with ESAR, was associated with decreased IA EL (hazard ratio, 0.341; 95% confidence interval [CI], 0.061-0.72; P = .031) and increased sac regression (hazard ratio, 3.92; 95% CI, 1.25-5.14; P = .02). Notably, FEVAR was associated with increased 1-year aneurysm-related reintervention (odds ratio, 4.33; 95% CI, 1.12-10.54; P = .046). On Kaplan-Meier analysis, FEVAR was associated with reduced freedom from reinterventions at 3 years (71.8% [95% CI, 0.63-0.78] vs 93.5% [95% CI, 0.80-0.97]; log-rank P = .019) but a trend toward improved survival at 3 years (79.15% [95% CI, 0.70-0.85] vs 61.5% [95% CI, 0.44-0.74]; log-rank P = .095). There was no significant difference in new-onset chronic dialysis between ESAR and FEVAR at 3 years (94.2% [95% CI, 0.82-0.98] vs 97.4% [95% CI, 0.93-0.99]; log-rank P = .124).

In the treatment of abdominal aortic aneurysms with hostile neck anatomy in this propensity-matched cohort, FEVAR was associated with fewer type IA ELs and greater sac regression compared with ESAR, with no detrimental impact on long-term renal function. There were more reinterventions, mostly branch related, in the FEVAR group. We await the results of the current randomized prospective trial comparing these strategies to further determine the impact of these clinical differences on aneurysm-related mortality.

To explore the ability of natural language processing (NLP) methods to identify confounder information beyond what can be identified using claims codes alone for pharmacoepidemiology.

We developed a retrospective cohort for high vs low dose proton pump inhibitors from linked Medicare claims (2008-2017) and electronic health record data for patients with a history of peptic ulcer disease or osteoarthritis. Clinical notes authored one year before first dispensing date were processed by off-the-shelf tools: bag-of-n-grams, latent Dirichlet allocation, a linguistics-focused tool, BERT sentence embeddings, BioBERT word embeddings, and GloVe word embeddings. Candidate features were ranked using Bross formula, a simple way to rank the marginal confounding impact of binary features on estimated causal effects.

The marginal confounding impact in the Bross rankings of NLP-derived features trended from 39 % in the top 100 to 77 % in the top 500 to 93 % in the top 5000 among patients with peptic ulcer disease. More specifically, the top 25 confounders are largely from factors identified by domain experts and structured fields, and the marginal impact of these confounders is stronger than others. Features 25 to 50 include features identified by a linguistics-focused tool and embeddings, whereas features 50 to 100 include more embeddings and bag-of-ngrams. After 100, the curve flattens, meaning that the marginal impact of those potential confounders gets smaller. Similarly, among patients with osteoarthritis, NLP-derived features trended from 66 % in the top 100 to 84 % in the top 500 to 95 % in the top 5000 when the outcome was gastrointestinal bleed and from 47 % in the top 100 to 81 % in the top 500 to 94 % in the top 5000 when the outcome was acute kidney injury. Similar trends were observed in the information gain data, though NLP-derived features had higher baselines.

NLP contributed to finding large numbers of features that can supplement claims data and prespecified variables to help provide additional confounder information. We found that unsupervised off-the-shelf NLP tools can scale to generate large numbers of features appropriate for high-dimensional proxy adjustment and pharmacoepidemiology use cases.

The Magistrates Early Referral into Treatment (MERIT) program is a voluntary, pre-plea diversion program for defendants appearing in the New South Wales (NSW), Australia, Local Court who have issues related to their alcohol and other drug (AOD) use.

Matched treatment and comparison groups were created using propensity score matching. The outcomes examined were AOD-related hospital admissions, AOD-related Emergency Department (ED) admissions, ED admissions (general), hospital admission (general), ambulance callouts, AOD related deaths, and deaths from any cause, as well as reoffending and imprisonment. Differences between outcomes were analysed using Cox regression (health outcomes), negative binomial regression (reoffending) and logistic regression (imprisonment).

Survival times for participants in the MERIT program were significantly shorter for all health outcomes except one (death). At the 12-month mark, MERIT participants offended 21 per cent less frequently than comparison group participants (IRR: 0.793. CI: 0.748-0.841). This gap increased to 27 per cent after 24 months (IRR: 0.870. CI: 0.829-0.912). At the conclusion of criminal proceedings participants in the MERIT program were significantly less likely to receive a prison sentence (OR: 0.728. CI: 0.674-0.787) or to die (OR: 0.674. CI: 0.502-0.904) CONCLUSION: The Magistrates' Early Referral Into Treatment Program appears to be an effective way of reducing the short-term risk of re-offending, imprisonment, and death.

While the impact of COVID-19 on bone metabolism has been extensively studied, the inverse relationship remains less understood. This study investigates whether impaired bone metabolism is associated with an increased risk of COVID-19 infection.

We conducted a nested case-control study within a population-based cohort, incorporating Kaplan-Meier analysis (KMA) to assess time to infection (TTI) differences. Propensity score matching (1:2) was performed and validated through standardized mean differences (<0.10), variance ratio (=1), and McFadden's pseudo-R2 (=0), ensuring balanced covariates. Bone status was evaluated using a composite index (AOMI), which included five components: P1NP and CTX (bone turnover markers), total hip bone mineral density (BMD-TH), trabecular bone score (TBS), and integral volumetric BMD (IvBMD). Inflammation and insulin resistance (IR) were assessed by albumin-to-globulin ratio (AGR <1.50) and the TG/HDL ratio (>2.50 in women and >2.80 in men), respectively.

We analysed 294 COVID-19 cases and 528 controls. AOMI+ individuals had a higher prevalence of COVID-19 (41.5% vs. 33.2%; p = 0.031), an adverse lipid pattern ("A" profile: high ApoB, LDL and TC) and pronounced bone changes (higher P1NP and CTX, lower BMD-TH, TBS, and IvBMD). AOMI - individuals were more likely to have metabolic syndrome, displayed a different lipid profile ("B" profile: elevated TG, AIP, and TG/HDL ratio), fewer bone alterations, and lower COVID-19 prevalence. TG/HDL ratio was 1.66 ± 1 in "A" profile, while it was 2.85 ± 1.4 in "B" profile individuals (p = 0.0001). Age acted as an effect modifier, and lowest tercile significantly increased COVID-19 risk associated with AOMI+ [Mantel-Haenszel OR = 1.42 (95%CI: 1.08-1.9); p = 0.022]. KMA identified AOMI+ men and individuals of both sexes in lowest age tercile, as groups with shorter TTI: These younger individuals had high CTX (women), low TBS (men), and high ApoB (both). In multivariable analyses, plasma CTX levels negatively correlated with TTI (adjusted β= -0.325; p = 0.0001). AOMI+ status was associated with increased COVID-19 risk after controlling for confounders, including IR (adjusted OR = 1.51; 95%CI: 1.04-2.10; p = 0.027), although this association weakened when adjusting for AGR (95%CI: 0.99-2.28; p = 0.055). ANCOVA-estimated adjusted TBS means were lower in COVID-19 cases compared to controls (1.259 vs. 1.294; p = 0.013).

Impaired bone metabolism was found to be associated with increased COVID-19 risk, in a relationship potentially mediated by underlying inflammation. Elevated osteoclastic activity and a defined lipid profile with high ApoB, TC, LDL levels, played a crucial role in the results. Bone quality parameters more accurately captured COVID-19-related bone changes than BMD.

Pseudomonas aeruginosa bloodstream infections (Pa-BSIs) are still a major cause of mortality in ICUs, posing many treatment uncertainties.

This multicentre, retrospective study analysed data from 14 Italian hospitals, including all consecutive adults developing Pa-BSI in ICU during 2021-22 and treated with antibiotics for at least 48 h. The primary aim was to identify predictors of 30 day mortality using Cox regression. Results were adjusted with inverse probability of treatment weighting (IPTW) and for immortal time bias.

Overall, 170 patients were included. High-risk BSI (source: lung, intra-abdominal, CNS) occurred in 118 (69%) patients, and 54 (32%) had septic shock. In 37 (22%), 73 (43%), 12 (7%) and 48 (28%) the definitive backbone therapy was piperacillin/tazobactam, carbapenems, colistin or new antipseudomonal cephalosporins (ceftolozane/tazobactam, n = 20; ceftazidime/avibactam, n = 22; cefiderocol, n = 6), respectively. Moreover, 58 (34%) received a second drug as combination therapy. The incidence of 30 day all-cause mortality was 27.6% (47 patients). By Cox regression, Charlson comorbidity index, neutropenia, septic shock and high-risk BSI were independent predictors of 30 day mortality, while previous colonization by P. aeruginosa, use of antipseudomonal cephalosporins as definitive treatment, and combination therapy were shown to be protective. However, after IPTW adjustment, only the protective effect of antipseudomonal cephalosporins was confirmed (adjusted HR = 0.27, 95% CI = 0.10-0.69), but not for combination therapy. Hence, the treatment effect was calculated: antipseudomonal cephalosporins significantly reduced mortality risk [-17% (95% CI = -4% to -30%)], while combination therapy was beneficial only in the case of septic shock [-66% (95% CI = -44% to -88%].

In ICU, antipseudomonal cephalosporins may be the preferred target therapy for the treatment of Pa-BSI; in addition, initial combination therapy may be protective in the case of septic shock.

Peripartum cardiomyopathy (PPCM) remains a serious threat to maternal health around the world. While bromocriptine, in addition to standard treatment for heart failure, presents a promising pathophysiology-based disease-specific treatment option in PPCM, the evidence regarding its efficacy remains limited. This study aimed to determine whether bromocriptine treatment is associated with improved maternal outcomes in PPCM.

Peripartum cardiomyopathy patients from the EORP PPCM registry with available follow-up were included. The main exposure of this exploratory non-randomized analysis was bromocriptine treatment, and the main outcome was a composite endpoint of maternal outcome [death or hospital readmission within the first 6 months after diagnosis, or persistent severe left ventricular dysfunction (left ventricular ejection fraction < 35%) at 6-month follow-up]. Inverse probability weighting was used to minimize the effects of confounding by indication. Multiple imputation was used to account for the missing data.

Among the 552 patients with PPCM, 85 were treated with bromocriptine (15%). The primary endpoint was available in 491 patients (89%) and occurred in 18 out of 82 patients treated with bromocriptine in addition to standard of care (22%) and in 136 out of 409 patients treated with standard of care (33%) (P = .044). In complete case analysis, bromocriptine treatment was associated with reduced adverse maternal outcome [odds ratio (OR) 0.29, 95% confidence interval (CI) 0.10-0.83, P = .021]. This association remained after applying multiple imputation and methods to correct for confounding by indication (inverse probability weighted model on imputed data: OR 0.47, 95% CI 0.31-0.70, P < 0.001). Thromboembolic events were observed in 6.0% of the patients in the bromocriptine group vs. 5.6% in the standard of care group (P = .900).

Among women with PPCM, bromocriptine treatment in addition to standard of care was associated with better maternal outcomes after 6 months.

Rheumatoid arthritis (RA) and psoriatic arthritis (PsA) are currently treated similarly. However, it is unclear which patient-reported outcome (PRO) domains need specific attention in the management of RA and PsA. Therefore, we aimed to determine the difference in disease impact between matched RA and PsA patients at diagnosis and after 1 year in two different regions.

RA patients from the treatment in the Rotterdam Early Arthritis CoHort trial (tREACH), PsA patients from the Dutch southwest Early PsA cohoRt (DEPAR) and RA and PsA patients from the Leiden Early Arthritis Clinic (EAC) were included. The difference in disease impact between RA and PsA was measured with the following PROs: pain (Visual Analogue Scale (VAS), 0-100), fatigue (VAS), activity limitation (Health Assessment Questionnaire-Disability Index) and health impact (general health (VAS) and 36-item Short-Form Health Survey (SF-36)). Propensity scores were used to match RA and PsA patients, after which inverse probability weights (IPWs) were calculated. IPW-weighted linear regression models were used to measure PRO differences.

391 RA patients from tREACH, 416 PsA patients from DEPAR, 702 RA and 99 PsA patients from the EAC were included. At diagnosis, PsA-DEPAR patients scored 5.04 units worse (95% CI 2.21 to 7.87) on SF-36 mental health compared with RA-tREACH patients. This difference still existed after 1 year of treatment (3.88 (95% CI 1.90 to 5.86)). PsA-EAC patients had more activity limitations after 1 year of treatment compared with RA-EAC patients (-0.30 (95% CI -0.50 to -0.10)). No significant differences were present in the other PRO domains.

The disease impact of early RA patients is similar to matched early PsA patients on most PRO domains, except for mental health and functional limitations, which were worse in PsA after 1 year of treatment.

To compare rates of clinical response in children with Clostridioides difficile infection (CDI) treated with metronidazole vs vancomycin.

Retrospective cohort study was performed as a secondary analysis of a previously established prospective cohort of hospitalized children with CDI. For 187 participants 2-17 years of age who were treated with metronidazole and/or vancomycin, the primary outcome of clinical response (defined as resolution of diarrhea within 5 days of treatment initiation) was identified retrospectively. Baseline variables associated with the primary outcome were included in a logistic regression propensity score model estimating the likelihood of receiving metronidazole vs vancomycin. Logistic regression using inverse probability of treatment weighting (IPTW) was used to estimate the effect of treatment on clinical response.

One hundred seven subjects received metronidazole and 80 subjects received vancomycin as primary treatment. There was no univariable association between treatment group and clinical response; 78.30% (N = 83) of the metronidazole treatment group and 78.75% (N = 63) of the vancomycin group achieved clinical response (P = 0.941). After adjustment using propensity scores with IPTW, the odds of a clinical response for participants who received metronidazole was 0.554 (95% CI: 0.272, 1.131) times the odds of those who received vancomycin (P = 0.105).

In this observational cohort study of pediatric inpatients with CDI, the rate of resolution of diarrhea after 5 days of treatment did not differ among children who received metronidazole vs vancomycin.

Respiratory syncytial virus (RSV) is a leading cause of infant morbidity. France has implemented a national campaign using nirsevimab to prevent RSV-related infections in infants. This study assessed its effectiveness in preventing hospitalization due to bronchiolitis in emergency department (ED). This retrospective study was conducted among six pediatric EDs in the Greater Paris area, France, and included infants aged < 3 months with a clinical diagnosis of bronchiolitis during the 2023-2024 RSV epidemic season. The primary outcome was hospitalization after the ED visits. The association with nirsevimab immunization was assessed using a multiple logistic model adjusted for potential confounding factors, with missing data handled using random forest imputation. Secondary analyses examined the risk of admission to the pediatric intensive care unit (PICU), RSV positivity, and subgroup analyses of prematurity, neonates, and deprivation using the FDep index (area-based measure of social deprivation in France). Between October 2 and December 31, 2023, 739 infants were included in the study. A total of 531 (72%) patients had a documented nirsevimab immunization status, and 402 (54%) were hospitalized following a bronchiolitis diagnosis. Nirsevimab showed 53.5% adjusted effectiveness in reducing hospitalizations (95% CI 34.1-67.3). Sensitivity analyses of complete-case data and propensity score matching yielded similar results. Nirsevimab also resulted in 51.1% reduction in PICU admissions (95% CI 10.7-74.3) and 79.6% reduction in RSV positivity (95% CI 68.0-87.1). The protective effect of immunization was consistent for preterm infants, neonates, and deprived groups, though the results were not statistically significant in these smaller subgroups.

Immunization with nirsevimab reduced hospitalization following an ED visit for bronchiolitis among infants aged < 3 months.

• Nirsevimab reduces the risk of bronchiolitis-related hospitalizations in clinical trials. • Real-world data from the immunization campaign in France remain limited.

• Nirsevimab showed 53.5% (95% CI 34.1-67.3) adjusted effectiveness in reducing hospitalizations for all-cause bronchiolitis in infants aged < 3 months in emergency departments. • Analyses included social deprivation and highlighted potential disparities in immunization access.

Abiraterone and enzalutamide are both approved in the United States for the treatment of metastatic castration-resistant prostate cancer (mCRPC). The objective of this study was to compare the real-world effectiveness and cardiovascular safety of these agents, drawing from a cohort of older adult patients diagnosed with mCRPC.

The Surveillance, Epidemiology, and End Results-Medicare database was used to conduct an observational study comparing three-year overall survival and one-year risk of major adverse cardiovascular events (MACE) between initiators of abiraterone or enzalutamide between September 2012 and June 2017. Inverse-probability-of-treatment weighting was used to balance measured confounders. MACE was defined as a hospitalization for myocardial infarction, heart failure, or ischemic event (stroke or transient attack). Results were additionally stratified by levels of a claims frailty index (robust, prefrail, frail) and the presence of baseline cardiovascular comorbidities.

The study population consisted of 4622 male adults 66 years of age and older diagnosed with mCRPC, of which 2430 initiated abiraterone and 2192 enzalutamide. The adjusted three-year overall survival was lower in patients initiating abiraterone (27.9 %) than enzalutamide (31.5 %) (adjusted survival difference [aSD] = -3.6 %, 95 % CI: -6.2 %, -0.9 %). In frailty-stratified analysis, no survival difference was found for the robust (aSD = 0.6 %, 95 % CI: -5.0 %, 6.3 %) or frail (aSD = -1.2 %, 95 % CI: -6.1 %, 3.7 %) subgroups, but there was lower survival with abiraterone for the prefrail group (aSD = -5.9 %, 95 % CI: -9.6, -2.3). The adjusted one-year risk of MACE was higher in abiraterone initiators (5.5 %) than enzalutamide initiators (3.6 %) (adjusted risk difference [aRD] = 1.8 %, 95 % CI: 0.6 %, 3.1 %); the increase was significant in the frail (aRD = 4.8 %, 95 % CI = 1.4 %, 8.3 %) and pre-frail subgroups (aRD =1.9 %, 95 % CI: 0.1 %, 3.6 %) but not the robust subgroup (aRD = -0.3 %, 95 % CI: -1.8 %, 1.2 %).

The three-year survival of abiraterone initiators was slightly lower than that of enzalutamide initiators, though the agents showed similar survival for patients with robust fitness. A one-year increase in MACE risk was observed in abiraterone initiators, especially amongst frail individuals, highlighting the importance of assessing frailty during therapy selection.

We performed a retrospective comparative study to clarify the optimal gallbladder drainage method prior to elective cholecystectomy.

We collected data from the Diagnosis Procedure Combination database about cholecystitis patients who underwent gallbladder drainage prior to cholecystectomy in a subsequent hospitalization between April 2014 and March 2020. We divided the study population into two groups: an endoscopic gallbladder stenting (EGBS) group and a percutaneous transhepatic gallbladder drainage (PTGBD) group. We performed propensity score matching and compared surgical outcomes related to cholecystectomy.

We collected 6306 cases (PTGBD: 6112 cases; EGBS: 194 cases). In propensity score matching, we obtained 193 matched pairs from the study population. Long-term postoperative antibiotics therapy (2 vs. 9; risk ratio 4.7 [95% CI: 1.1-30.9]) was more frequent in the EGBS group than the PTGBD group. There were no significant differences between the two groups for laparotomic cholecystectomy and postoperative bile duct drainage. For reoperation, postoperative abdominal drainage and postoperative blood transfusion, there were few outcome occurrences and effect measures were not obtained.

Gallbladder drainage by EGBS may have more risk of surgical complications related to elective cholecystectomy than PTGBD. There are a few reports on this topic, so further research should be conducted.

Observational data, such as electronic clinical records and claims data, can prove invaluable for evaluating the Average Treatment Effect (ATE) and supporting decision-making, provided they are employed correctly. The Inverse Probability of Treatment Weighting (IPTW) method, based on propensity scores, has demonstrated remarkable efficacy in estimating ATE, assuming that the assumptions of exchangeability, consistency, and positivity are met. Directed Acyclic Graphs (DAGs) offer a practical approach to assess the exchangeability assumption, which asserts that treatment assignment and potential outcomes are independent given a set of confounding variables that block all backdoor paths from treatment assignment to potential outcomes. To ensure a consistent ATE estimator, one can adjust for a minimally sufficient adjustment set of confounding variables that block all backdoor paths from treatment assignment to the outcome. To enhance the efficiency of ATE estimators, our proposal involves incorporating both the minimally sufficient adjustment set of confounding variables and predictors into the propensity score model. Extensive simulations were conducted to evaluate the performance of propensity score-based IPTW methods in estimating ATE when different sets of covariates were included in the propensity score models. The simulation results underscored the significance of including the minimally sufficient adjustment set of confounding variables along with predictors in the propensity score models to obtain a consistent and efficient ATE estimator. We applied this proposed method to investigate whether tracheostomy was causally associated with in-hospital infant mortality, utilizing the 2016 Healthcare Cost and Utilization Project Kids' Inpatient Database. The estimated ATE was found to be approximately 2.30%-2.46% with p-value >0.05.

Observational studies assessing causal effects of interventions are subject to indication (selection) bias, which may be difficult to eliminate using traditional multivariable techniques. When properly specified, propensity score-adjusted analysis may offer an advantage over traditional regression by ensuring that investigators explicitly assess comparability of baseline prognostic factors between the treated and untreated. However, it is important to note that the effectiveness of a propensity score-adjusted analysis depends on the variables selected for the model and the analytic approach. Noninclusion of important prognostic factors and model misspecification among other errors may in fact increase bias; thus, in performing propensity score analysis, these errors must be minimized as much as possible or assessed using sensitivity analysis to ensure validity.

To compare pregnancy outcomes in a midwifery continuity of care (MCoC) model to standard midwifery care in Sweden.

Matched cohort study.

Public healthcare during pregnancy and childbirth, Stockholm, Sweden.

Women giving birth at Karolinska University Hospital site Huddinge in Stockholm between January 1, 2019, and August 31, 2021.

Data on all births including MCoC and standard care, during the time period, were retrieved from the national Swedish Pregnancy Register. Propensity score matching was applied to obtain a matched set from the standard care group for every woman in the MCoC model. Based on the matched cohort, we estimated risk ratios (RR) for binary outcomes with 95% confidence intervals (CI).

Interventions during labor, mode of birth, and preterm birth (< 37 gestational weeks).

Compared with standard care, women in the MCoC model were more likely to give birth spontaneously (RR 1.06 95% CI 1.02-1.10) and less likely to have an elective cesarean on maternal request (RR 0.24 95% CI 0.11-0.51). The risk of preterm birth was also reduced in the MCoC group (RR 0.51 95% CI 0.32-0.82).

The MCoC model was associated with fewer medical interventions and improved pregnancy outcomes.

Electroconvulsive therapy (ECT) and ketamine are effective treatments for depression; however, evidence-based guidelines are needed to inform individual treatment selection. We adapted the Personalized Advantage Index (PAI) using machine learning to predict optimal treatment assignment to ECT or ketamine using EHR data on 2506 ECT and 196 ketamine patients. Depressive symptoms were evaluated using the Quick Inventory of Depressive Symptomatology (QIDS) before and during acute treatment. Propensity score matching across treatments was used to address confounding by indication, yielding a sample of 392 patients (n = 196 per treatment). Models predicted differential minimum QIDS scores (min-QIDS) over acute treatment using pretreatment EHR measures and SHAP values identified prescriptive predictors. Patients with large PAI scores who received a predicted optimal had significantly lower min-QIDS compared to the non-optimal treatment group (mean difference = 1.19 [95% CI: 0.32, ∞], t = 2.25, q < 0.05, d = 0.26). Our model identified candidate pretreatment factors to provide actionable, effective antidepressant treatment selection guidelines.

Propensity scores are commonly used to reduce the confounding bias in non-randomized observational studies for estimating the average treatment effect. An important assumption underlying this approach is that all confounders that are associated with both the treatment and the outcome of interest are measured and included in the propensity score model. In the absence of strong prior knowledge about potential confounders, researchers may agnostically want to adjust for a high-dimensional set of pre-treatment variables. As such, variable selection procedure is needed for propensity score estimation. In addition, studies show that including variables related to treatment only in the propensity score model may inflate the variance of the treatment effect estimators, while including variables that are predictive of only the outcome can improve efficiency. In this article, we propose to incorporate outcome-covariate relationship in the propensity score model by including the predicted binary outcome probability as a covariate. Our approach can be easily adapted to an ensemble of variable selection methods, including regularization methods and modern machine-learning tools based on classification and regression trees. We evaluate our method to estimate the treatment effects on a binary outcome, which is possibly censored, across multiple treatment groups. Simulation studies indicate that incorporating outcome probability for estimating the propensity scores can improve statistical efficiency and protect against model misspecification. The proposed methods are applied to a cohort of advanced-stage prostate cancer patients identified from a private insurance claims database for comparing the adverse effects of four commonly used drugs for treating castration-resistant prostate cancer.

Generalized propensity score (GPS) methods have become popular for estimating causal relationships between a continuous treatment and an outcome in observational studies with rich covariate information. The presence of rich covariates enhances the plausibility of the unconfoundedness assumption. Nonetheless, it is also crucial to ensure the correct specification of both marginal and conditional treatment distributions, beyond the assumption of unconfoundedness.

We address limitations in existing GPS methods by extending balance-based approaches to high dimensions and introducing the Generalized Outcome-Adaptive LASSO and Doubly Robust Estimate (GOALDeR). This novel approach integrates a balance-based method that is robust to the misspecification of distributions required for GPS methods, a doubly robust estimator that is robust to the misspecification of models, and a variable selection technique for causal inference that ensures an unbiased and statistically efficient estimation.

Simulation studies showed that GOALDeR was able to generate nearly unbiased estimates when either the GPS model or the outcome model was correctly specified. Notably, GOALDeR demonstrated greater precision and accuracy compared to existing methods and was slightly affected by the covariate correlation structure and ratio of sample size to covariate dimension. Real data analysis revealed no statistically significant dose-response relationship between epigenetic age acceleration and Alzheimer's disease.

In this study, we proposed GOALDeR as an advanced GPS method for causal inference in high dimensions, and empirically demonstrated that GOALDeR is doubly robust, with improved accuracy and precision compared to existing methods. The R package is available at https://github.com/QianGao-SXMU/GOALDeR .

Direct oral anticoagulants (DOACs) are commonly co-prescribed with digoxin, but whether there is a drug interaction between them is unclear. We aimed to investigate potential drug interactions between DOACs and digoxin.

We identified DOAC users during January 1, 2011-December 31, 2019 using data from Clinical Practice Research Datalink Aurum in cohort design with propensity score to compare the hazards of effectiveness cardiovascular and mortality outcomes and safety bleeding outcomes, respectively, in DOAC + digoxin users versus DOAC + beta-blocker users. A case-crossover design was conducted to compare odds of exposure to different drug initiation patterns in hazard period versus referent period.

Of 397,459 DOAC users, we identified 25,251 co-prescribed digoxin and 109,779 co-prescribed beta-blockers in cohort study. A lower proportion of DOAC + digoxin users were men (46%) in contrast with that of DOAC + beta-blocker users (53%). Mean age of DOAC + digoxin users (77.1 years) were higher than DOAC + beta-blocker users (74.5 years). No increased risk of pharmacologically predictable DOAC safety outcomes or specific effectiveness outcomes was seen with DOAC + digoxin. A higher risk of all-cause mortality (hazard ratio: 1.35; 99% confidence interval [CI]: 1.14-1.61) was observed with DOAC + digoxin versus DOAC + beta-blockers. In the case-crossover study, a 24% higher odds of all-cause mortality was seen with initiating digoxin while taking DOAC (odds ratio: 1.24; 99% CI: 1.06-1.45); and a 63% higher odds was also seen with initiating DOAC while taking digoxin (odds ratio: 1.63; 99% CI: 1.41-1.88).

We found no increased risk of bleeding when DOACs are used with digoxin, suggesting combined use does not lead to drug-drug interaction. Future work is recommended to investigate the underlying mechanism of association with all-cause mortality.

This study aimed to examine potential drug interactions between direct oral anticoagulants (DOACs) (a drug class to prevent blood clots) and digoxin (treatment of abnormal heart rhythms). We compared a range of clinical outcomes in people prescribed DOAC and digoxin with people prescribed DOAC and beta-blockers (a treatment alternative to digoxin). We also used a new study design (case-crossover design) to compare the risk of clinical outcomes between different periods within a person as a validation. In both study designs, we found no increased risk of bleeding when DOACs are used with digoxin, suggesting combined use does not lead to drug-drug interaction. However, we found an increased risk of all-cause death associated with digoxin in DOAC users which requires further investigation.

Immune checkpoint inhibitors (ICIs) are associated with an increased risk of major adverse cardiovascular events (MACE). Glucagon-like peptide-1 agonists (GLP1a), initially developed for type 2 diabetes mellitus (T2DM), have shown promising results in reducing cardiovascular events. We aimed to investigate the effect of GLP1a on cardiovascular events in patients receiving ICIs.

We conducted a retrospective, propensity score-matched cohort study using the TriNetX database. We identified adults with cancer and T2DM who received ICIs between April 2013 and May 2023. The primary efficacy outcome was incident MACE, defined as a composite of myocardial infarction, need for coronary revascularization, heart failure, ischemic stroke, and cardiac arrest. The secondary efficacy outcomes were the individual components of MACE as well as myocarditis and pericarditis. Safety outcomes included the occurrence of immune-related adverse events, serious adverse events related to GLP1a use, and all-cause mortality.

We identified 7651 patients eligible for inclusion, among which 479 received GLP1a and 7172 received non-GLP1a diabetes medications. After matching (469 patients each), baseline characteristics were well-balanced. Over a median 12-month follow-up, the GLP1a cohort had a significantly lower MACE incidence than the non-GLP1a cohort (9.0 vs. 17.1 events per 100 patient-years) with a 54 % lower risk of MACE (Hazard ratio (HR),0.46 [95 % CI: 0.32-0.67]). There were reductions in myocardial infarction or need for coronary revascularization, heart failure, and all-cause mortality, with no differences in other cardiovascular events. GLP1a use did not increase risk of adverse events, including pancreatitis, biliary disease, bowel obstruction, gastroparesis, and immune-related adverse events.

GLP1a use in cancer patients with T2DM receiving ICIs was associated with reduced MACE and all-cause mortality without an increased risk in serious adverse events.

Postpartum hemorrhage is the most common complication in obstetrics and is a leading cause of serious maternal morbidity. Women with one previous cesarean delivery are at risk for postpartum hemorrhage.

The aim of this study was to compare the risk of postpartum hemorrhage according to the planned mode of delivery among women with one previous cesarean delivery.

This study was a secondary analysis of the Lower Uterine Segment Trial. The primary outcome was postpartum hemorrhage, and the secondary outcome was a composite of severe postpartum hemorrhage including the need for blood transfusion, second-line therapy to stop bleeding, or hysterectomy. The exposure was the planned mode of delivery, which was decided at around 36 weeks of gestation. Multivariate logistic regression analysis with a random intercept of the maternity unit was used to assess the relationship between the planned mode of delivery and postpartum hemorrhage. A secondary analysis using inverse probability weighting was performed to limit indication bias.

Of the 2948 women included in the Lower Uterine Segment Trial, 2889 women were included in this secondary analysis; 2689 (93.0%) chose planned vaginal delivery, and 200 (7.0%) planned cesarean delivery. The rate of postpartum hemorrhage was lower in the planned cesarean delivery group than in the planned vaginal delivery group (2.5% vs 7.5%, adjusted odds ratio 0.28 [0.11-0.69] P<.01). The rate of the composite maternal outcome of severe postpartum hemorrhage was similar between the 2 groups (2.9% vs 2.0%, adjusted odds ratio 0.66 [0.24-1.83] P=.42). These results were consistent in the secondary analysis involving inverse probability weighting (odds ratio 0.35 [0.24-0.51] P<.01).

Planned cesarean delivery was associated with a lower percentage of women with postpartum hemorrhage compared with planned vaginal delivery after one previous cesarean delivery. However, the use of transfusion, second-line therapies, or hysterectomy was rare, and the rates were similar between the 2 groups.

To demonstrate the efficacy of machine learning models in predicting mortality in melanoma cancer, we developed an interpretability model for better understanding the survival prediction of cancer. To this end, the optimal features were identified, ten different machine learning models were utilized to predict mortality across various datasets. Then we have utilized the important features identified by those machines learning methods to construct a new model named NKECLR to forecast mortality of patient with cancer. To explicitly clarify the model's decision-making process and uncover novel findings, an interpretable technique incorporating machine learning and SHapley Additive exPlanations (SHAP), as well as LIME, has been employed, and four genes EPGN, PHF11, RBM34, and ZFP36 were identified from those machine learning(ML). The experimental analysis conducted on training and validation datasets demonstrated that the proposed model has a good performance com- pared to existing methods with AUC value 81.8%, and 79.3%, respectively. Moreover, when combined our NKECLR with PD-L1, PD-1, and CTLA-4 the AUC value was 83%0. Finally, these findings have been applied to comprehend the response of drugs and immunotherapy. Our research introduced an innovative predictive NKECLR model utilizing natural killer(NK) cell marker genes for cohorts with melanoma cancer. The NKECLR model can effectively predict the survival of melanoma cancer cohorts and treatment results, revealing distinct immune cell infiltration in the high-risk group.

This study investigates the causal relationships between hormone levels and growth and development of children, focusing specifically on height disparities in cases of dwarfism. Besides utilizing double-debiased machine learning approach, the study integrates three alternative causal inference methods: partialing-out lasso linear regression, cross-fit partialing-out lasso linear regression, and post-double selection LASSO. These machine learning techniques are pivotal in identifying causal effects within observational data. The findings reveal a positive correlation between luteinizing hormone (LH) levels and adolescent height, while follicle-stimulating hormone (FSH) and the LH/FSH ratio show inverse correlations. The study underscores the significant role of hormone levels, particularly LH, in determining height, offering valuable insights that could guide future interventions or treatments for children and adolescents with dwarfism.

The evidence informing the harms of gabapentin use are at risk of bias from comparing users with nonusers.

To describe the risk for fall-related outcomes in older adults starting treatment with gabapentin versus duloxetine.

New user, active comparator study using a target trial emulation framework.

MarketScan (IBM) commercial claims between January 2014 and December 2021.

Adults aged 65 years or older with diabetic neuropathy, postherpetic neuralgia, or fibromyalgia and without depression, anxiety, seizures, or cancer in the 365 days before cohort entry.

New initiation of treatment with gabapentin or duloxetine (comparator).

The primary outcome was the hazard of experiencing any fall-related visit in the 6 months after initiating gabapentin or duloxetine until discontinuation of treatment. Secondary outcomes were hazard of severe fall-related events defined as a fall associated with hip fracture or emergency department visit or hospitalization associated with a fall. Stabilized inverse probability of treatment weighting was used to adjust for baseline characteristics.

Our analytic cohort included 57 086 older adults with a diagnosis of interest initiating treatment with gabapentin (n = 52 152) or duloxetine (n = 4934). Overall median follow-up duration was 30 days (IQR, 30 to 90 days). Weighted cumulative incidence of a fall-related visit per 1000 person-years at 30, 90, and 180 days was 103.60, 90.44, and 84.44 for gabapentin users and 203.43, 177.73, and 158.21 for duloxetine users, respectively. At 6-month follow-up, incident gabapentin users had lower hazard of falls (hazard ratio, 0.52 [95% CI, 0.43 to 0.64]), but there was no difference in the hazards of experiencing severe falls. Results were similar across sensitivity and subgroup analyses.

Claims may contain fewer frail adults and undercount falls.

Compared with incident use of duloxetine, incident use of gabapentin was not associated with increased fall-related visits.

None.

Conversion to neovascular disease in patients with non-neovascular age-related macular degeneration (AMD) initiated on direct oral anticoagulants (DOACs) compared with matched patients treated with warfarin.

Retrospective cohort study.

The study included 20 300 patients and 13 387 patients with non-neovascular AMD initiated on DOACs or warfarin, respectively, before propensity score matching (PSM).

TriNetX was used to identify patients diagnosed with non-neovascular AMD stratified by treatment with DOACs or warfarin with at least 6 months of follow-up. Propensity score matching was performed to control for baseline demographics and medical comorbidities.

Relative risk (RR) of developing neovascular AMD, macular hemorrhage (MH), vitreous hemorrhage (VH), and requiring an ocular intervention (intravitreal anti-VEGF therapy or pars plana vitrectomy [PPV]) within 6 months and 1 year. Patients with chronic atrial fibrillation (AF) on anticoagulation were separately evaluated for the same measures within 5 years after initiating therapy.

Treatment with warfarin was associated with a higher risk of developing neovascular AMD at 6 months (RR, 1.24, 95% confidence interval [CI], 1.12-1.39; P < 0.001) and 1 year (RR, 1.26, 95% CI, 1.14-1.40; P < 0.001) when compared with matched patients treated with DOACs. There was an increased risk of requiring intravitreal anti-VEGF therapy (6 months: RR, 1.30; 95% CI, 1.13-1.49; P < 0.001; 1 year: RR, 1.31, 95% CI, 0.72-2.05; P < 0.001) and PPV (6 months: RR, 2.13; 95% CI, 1.16-3.94; P = 0.01; 1 year: RR, 2.29, 95% CI, 1.30-4.05; P = 0.003). Among patients with AMD and AF treated with warfarin, there was an increased risk of ocular complications (neovascular AMD: RR, 1.25; 95% CI, 1.14-1.38; P < 0.001; MH: RR, 1.86; 95% CI, 1.47-2.35; P < 0.001; VH: RR, 2.22; 95% CI, 1.51-3.26; P < 0.001) and need for intravitreal anti-VEGF therapy (RR, 1.34; 95% CI, 1.18-1.52; P < 0.001) over an extended 5-year period. There was no significant difference in the development of major systemic hemorrhagic events between the 2 cohorts over 5 years.

Patients with non-neovascular AMD treated with warfarin were more likely to develop neovascular disease and require ocular intervention for hemorrhagic complications when compared with matched patients initiated on DOACs.

Proprietary or commercial disclosure may be found after the references.

Potential benefits of enhancing nursing education using augmented reality (AR) invite further investigation of its applications. This study evaluated an AR system augmenting classes on complex cardiac problems in an undergraduate didactic nursing course.

This study used a mixed-methods, quasiexperimental, and non-concurrent controlled design. Pre- and posttest scores for self-assessment quizzes and two examinations were compared for control group participants who received usual teaching (n = 51) and intervention group participants who received AR-enhanced teaching (n = 39).

Usual teaching and the AR system were rated positively for realism, effectiveness, and usefulness by approximately 40% of the participants in both groups. The AR system had positive ratings for innovation, capturing interest, and engagement. Quiz and examination scores increased in both groups but were not significantly different between groups.

AR is an innovative approach promoting student engagement and interest in learning. [J Nurs Educ. 2025;64(2):81-89.].

Whether prescription opioid exposure, duration, and dose are associated with cognitive function remains inconclusive.

A longitudinal cohort among 3097 older adults with chronic pain and without dementia was conducted using Health and Retirement Study (HRS) linked to Medicare data from 2006 to 2020. Prescription opioid exposure, cumulative use for ≥ 90 days, and high-dose use (≥ 90 morphine milligram equivalents [MME] daily) were assessed biennially. Memory score and dementia probability were derived from HRS cognitive measures and analyzed using linear mixed-effects models.

Adjusted memory decline and dementia probability were not statistically different between patients with (vs. without) opioid exposure and between patients with cumulative use for ≥ 90 days (vs. < 90 days) but were higher between participants with high-dose opioid use (vs. low-dose) at the end of the follow-up.

Prescription opioid exposure and duration were not associated, but high-dose opioid use was associated with greater memory decline and dementia probability.

Opioid use versus no use was not related to memory decline and dementia probability. Long-term opioid use was not related to memory decline and dementia probability. High-dose opioid use was related to greater memory decline and dementia probability.

We are lacking data with a high level of evidence on the use of episiotomy during instrumental delivery to prevent anal sphincter injury, which nonetheless presents the highest risk.

Our main objective was to assess the protective effect of episiotomy against obstetric anal sphincter injury in nulliparous women during instrumental delivery according to type of instrument. We also investigated its impact on immediate maternal and neonatal morbidity.

We conducted a prospective comparative cohort study for clinical trial emulation by means of propensity score weighting. The study was especially designed for consideration of possible confounders. This was a nationwide observational multicenter study including 111 French public and private maternity units between April 2021 and March 2022. We included nulliparous women, with singleton cephalic fetus, at more than 34 weeks of gestation. We considered vacuum, forceps, and spatula deliveries. We proceeded to a comparative analysis between women with and without episiotomy. The main outcome was obstetric anal sphincter injury occurrence. We used composite criteria for both maternal and neonatal immediate morbidity.

The analyses pertained to 11,013 women. Overall prevalence of episiotomy was 23%: 17% for vacuum (N=7007), 37% for forceps (N=2378), and 29% in case of spatula-assisted (N=1628) delivery. Episiotomy was not associated with significantly decreased obstetric anal sphincter injury occurrence in vacuum delivery (from 5.2% without episiotomy to 3.8%, odds ratio=0.73 [0.48-1.03]) or forceps delivery (from 10.9% without episiotomy to 8.8%, odds ratio=0.81 [0.56-1.14]). In contrast, we observed significantly decreased obstetric anal sphincter injury occurrence (from 9.4% without episiotomy to 5.6%) in spatula delivery (odds ratio=0.60 [0.37-0.87]). Episiotomy was associated with increased maternal morbidity using forceps (from 13.6%-18.3%, odds ratio=1.35 [1.01-1.73]) and spatulas (from 9.0%-13.4%, odds ratio=1.51 [1.11-2.00]). We also observed increased neonatal morbidity in vacuum delivery associated with episiotomy (from 9.1%-13.6%, odds ratio=1.49 [1.21-1.79]), but a decrease in case of forceps delivery with episiotomy (from 12.6%-9.2%, odds ratio=0.74 [0.55-0.95]).

Episiotomy was not associated with a decreased risk of obstetric anal sphincter injury in vacuum or forceps delivery, and a marginal reduction was achieved using spatulas. Our results do not favor extensive episiotomy during instrumental delivery.

ClinicalTrial NCT 04446780.

Prospective Observational Propensity Score.

Randomization may lead to bias when the treatment is unblinded and there is a strong patient preference for treatment arms (such as in spinal device trials). This report describes the rationale and methods utilized to develop a propensity score (PS) model for an investigational device exemption (IDE) trial (NCT03115983) to evaluate decompression and stabilization with an investigational dynamic sagittal tether (DST) vs decompression and Transforaminal Lumbar Interbody Fusion (TLIF) for patients with symptomatic grade I lumbar degenerative spondylolisthesis with spinal stenosis.

Twenty-five baseline covariates were selected for their expected relationship to patient outcomes or enrollment bias. Subclassification by PS quintiles was used to design a sample of investigational DST patients and TLIF controls with excellent covariate balance in which to estimate causal treatment effects. Additionally, balance in PS covariates was compared to available matching covariates from seven randomized spine IDE trials.

The PS subclassification design resulted in excellent balance across baseline covariates, as evidenced by small standardized mean differences and no significant between group differences after accounting for the PS design (all P ≥ 0.768). Differences in SMDs among covariates of randomized spine IDE trials were not significant (P = 0.396).

The PS subclassification design achieved excellent covariate balance between DST investigational and TLIF control participants. This PS designed sample shows covariate balance similar to that observed in published studies in which patients were randomized to investigational or control arms.Clinical trial registered with https://www.clinicaltrials.gov (NCT03115983).

During the first wave of the COVID-19 outbreak in China, the surge of COVID-19 cases was rapid and drastic. Emerging evidence suggests that beyond the acute phase, patients with COVID-19 may experience a wide range of postacute or long COVID sequelae. However, the mechanism and burden of COVID-19, especially long COVID, have not yet been comprehensively clarified. To fill this knowledge gap, this large prospective follow-up study aims to investigate the short-term and long-term effects of COVID-19, explore the underlying biological mechanism and identify predictive neuroimaging and haematological biomarkers associated with these effects.

This multicentre study will recruit patients infected during the first wave of COVID-19 in China and healthy controls (HCs) with no history of COVID-19 infection from nine participating hospitals. Confirmed patients with mild or moderate COVID-19 will complete the following programmes during the acute infection phase and at 3, 12 and 24 months after infection: (a) blood test at the local laboratory, (b) multimodal brain and spine MRI scan and (c) the neuropsychological scales and questionnaires. Similarly, the uninfected HCs will complete the same programmes as the infected group mentioned above at the time of inclusion. At the first time point, 501 participants (418 patients and 83 HCs) from nine recruiting hospitals have been observed. Ultimately, all of these results will be analysed to explore the short-term and long-term effects of COVID-19.

Ethics approval was granted by Ethics Committee of the First Affiliated Hospital of Xi'an Jiaotong University (XJTU1AF2023LSK-013). Findings will be presented at national and international conferences, as well as published in peer-reviewed scientific journals.

NCT05745805.

New respiratory syncytial virus (RSV) vaccines have been approved in the USA for the prevention of RSV-associated lower respiratory tract disease in adults aged 60 years and older. Information on the real-world effectiveness of these vaccines is needed.

We used electronic health records in the Veterans Health Administration to emulate a target trial comparing a single dose of a recombinant stabilised prefusion F protein RSV vaccine versus no vaccination among veterans aged 60 years and older. We matched eligible vaccine recipients with up to four unvaccinated individuals in four monthly nested sequential trials from Sept 1 to Dec 31, 2023. Outcomes were ascertained up to March 31, 2024. The primary outcome was any positive RSV test from day 14 following the matched index date. Secondary outcomes included hospitalisation and emergency department or urgent care encounter occurring within 1 day before or after a positive RSV test. We estimated vaccine effectiveness as 100 × (1 - risk ratio).

We included 146 852 vaccinated individuals matched to 582 936 unique control individuals, weighted equally to represent 146 852 individuals. Across the two groups, 276 039 (94·0%) of 293 704 veterans were male, 17 665 (6·0%) were female, and median age was 75·9 years (IQR 71·7-79·7). Over a median follow-up of 124 days (IQR 102-150), the incidence rate of documented RSV infection was 1·7 (95% CI 1·4-2·1) events per 1000 person-years (88 total events) in the vaccinated group and 7·3 (6·6-8·1) per 1000 person-years in the unvaccinated group (372 total events), and vaccine effectiveness was estimated as 78·1% (72·6-83·5). Among the secondary outcomes, vaccine effectiveness was estimated at 78·7% (72·2-84·8) against RSV-associated emergency department or urgent care encounters, and 80·3% (65·8-90·1) against RSV-associated hospitalisation.

RSV vaccination was effective in preventing RSV-related illness, including associated health-care use, in adults aged 60 years and older during the 2023-24 respiratory illness season, supporting current recommendations for vaccination in this population.

US Department of Veterans Affairs Cooperative Studies Program, US Department of Health and Human Services Biomedical Advanced Research and Development Authority, and US Food and Drug Administration.

Mental health status may be associated with residential neighborhoods' physical and social characteristics; however, longitudinal evidence is limited, and the findings are inconsistent.

To examine the longitudinal associations between neighborhood physical and social environments and mental health among adults after residential relocation.

We used national-representative panel data of 3000 adults between 2008 and 2013 in the Netherlands. We included movers relocating to neighborhood environments identified as health-supportive or adverse; non-movers served as the control group. Mental health was measured pre- and post-move using the Mental Health Inventory. Time-varying exposure to physical environmental factors (i.e., green space, blue space, air pollution, and population density) and social environmental factors (i.e., socioeconomic deprivation and social fragmentation) were assigned based on adults' residential neighborhood histories. We used propensity score matching to select non-movers from the control group and difference-in-difference regressions to estimate the associations between environmental changes and mental health.

Our results showed that decreases in fine particulate matter (β = -3.869, 95% CI: [-7.583, -0.155]), population density (β = -5.893, 95% CI: [-9.468, -2.319]), and socioeconomic deprivation (β = -4.756, 95% CI: [-7.800, -1.712]) were significantly associated with improvements in mental health. An increase in neighborhood social fragmentation was significantly associated with improvements in mental health (β = -3.520, 95% CI: [-6.742, -0.298]). We observed null associations between changes in mental health due to changes in green and blue space.

Following relocation, changed neighborhood environmental conditions appear to have longitudinal mental health associations. Moving to a neighborhood with less air pollution, lower population density, and reduced socioeconomic deprivation was mental health-supportively associated. However, the finding that greater social fragmentation contributes to mental health improvements requires further investigation.

Although drug interactions between clarithromycin/erythromycin/fluconazole and direct oral anticoagulants (DOACs) are mechanistically plausible, it is uncertain whether they are clinically relevant.

This study aims to investigate the association among coprescribed DOACs and antimicrobials and bleeding, cardiovascular disease and mortality.

We identified DOAC users in the Clinical Practice Research Datalink Aurum from January 1, 2011 to March 29, 2021. We used a cohort design to estimate hazard ratios (HRs) for bleeding outcomes (intracranial bleeding, gastrointestinal bleeding, other bleeding), comparing DOACs + clarithromycin/erythromycin/fluconazole users with DOACs users not receiving these antimicrobials. Cardiovascular outcomes were ischaemic stroke, myocardial infarction, venous thromboembolism, cardiovascular mortality, and all-cause mortality. A 6-parameter case-crossover design comparing odds of exposure with different drug initiation patterns for all outcomes in hazard window vs referent window within an individual was also conducted.

Of 483,815 DOAC users, we identified 21,701 coprescribed clarithromycin, 4532 coprescribed erythromycin, and 4840 coprescribed fluconazole. We observed an increased risk of gastrointestinal bleeding over 7 days following coprescription of DOAC + erythromycin vs DOAC alone (HR 3.66; 99% confidence interval [CI] 1.27-10.51), with wide CIs in case-crossover analysis. No evidence of increased risk of bleeding outcomes was seen for DOAC + clarithromycin/fluconazole in cohort and case-crossover analyses. For cardiovascular outcomes, compared with DOAC alone, an increased risk of cardiovascular mortality with DOAC + clarithromycin (HR 3.36; 99% CI 1.73-6.52) and increased risk of all-cause mortality with DOAC + clarithromycin/erythromycin/fluconazole were observed in cohort analysis. However, similar risks were found when initiating erythromycin/fluconazole with and without DOACs.

We found no strong evidence of increased risks of bleeding and cardiovascular outcomes in DOACs + clarithromycin/fluconazole/erythromycin users except a possible short-term increased risk of gastrointestinal bleeding in DOACs + erythromycin users.

This study introduces Smart Imitator (SI), a 2-phase reinforcement learning (RL) solution enhancing personalized treatment policies in healthcare, addressing challenges from imperfect clinician data and complex environments.

Smart Imitator's first phase uses adversarial cooperative imitation learning with a novel sample selection schema to categorize clinician policies from optimal to nonoptimal. The second phase creates a parameterized reward function to guide the learning of superior treatment policies through RL. Smart Imitator's effectiveness was validated on 2 datasets: a sepsis dataset with 19 711 patient trajectories and a diabetes dataset with 7234 trajectories.

Extensive quantitative and qualitative experiments showed that SI significantly outperformed state-of-the-art baselines in both datasets. For sepsis, SI reduced estimated mortality rates by 19.6% compared to the best baseline. For diabetes, SI reduced HbA1c-High rates by 12.2%. The learned policies aligned closely with successful clinical decisions and deviated strategically when necessary. These deviations aligned with recent clinical findings, suggesting improved outcomes.

Smart Imitator advances RL applications by addressing challenges such as imperfect data and environmental complexities, demonstrating effectiveness within the tested conditions of sepsis and diabetes. Further validation across diverse conditions and exploration of additional RL algorithms are needed to enhance precision and generalizability.

This study shows potential in advancing personalized healthcare learning from clinician behaviors to improve treatment outcomes. Its methodology offers a robust approach for adaptive, personalized strategies in various complex and uncertain environments.

Effective use of health services by pregnant and postpartum woman (PPWs) is crucial to maternal and child health. Most maternal deaths are attributed to inadequate maternal health services, especially in rural areas. As a vulnerable group, rural PPWs can effectively prevent and reduce maternal and infant health risk factors through whole-process health management and ensure the health and safety of mothers and infants. Therefore, improving the utilization rate of rural maternal health services is a key issue that needs to be addressed urgently. This study aimed to explore the influence of peer interaction on the utilization of maternal health services in rural areas and the mediating effect of maternal health service knowledge.

Based on cross-sectional data of 821 PPWs in rural northwest China. This study used propensity score matching (PSM) to analyze the effects of peer interaction (PI) on maternal health service utilization, including maternal system management rate (Y1), prenatal screening rate (Y2), and postpartum visit rate (Y3). In addition, the mediating role of maternal health service knowledge between peer interaction and health service utilization was empirically tested.

The findings highlight the important role of peer interaction in improving the utilization rate of maternal health services in rural northwest China. The study found that peer interaction significantly improved the maternal system management rate, prenatal screening rate, and postpartum visit rate. At the same time, peer interaction enhances knowledge of maternal health services, which plays a key role in improving maternal health behaviors.

Through experience sharing and knowledge exchange among peers, understanding of health services can be enhanced, and positive health behaviors can be promoted. Policymakers and healthcare providers should integrate peer support programs into existing maternal health initiatives and fully use social media and community resources to create interactive platforms for maternal and infant knowledge that combine online and offline. By actively promoting peer interaction and experience sharing, these initiatives can maximize the positive role of peer support, increase the utilization rate of health services, and effectively ensure their safety.

Incidence of visual impairment (VI) and dyslipidemia is increasing with aging. Although good medication adherence (MA) is a crucial factor in achieving therapeutic goals for dyslipidemia, there is a paucity of studies measuring MA in the visually impaired with dyslipidemia. We investigated whether patients with VI had worse MA to dyslipidemia drugs than non-disabled people and determined the factors affecting MA among patients with VI. Data on dyslipidemia patients with VI were extracted in 2017 from the sample cohort database of the National Health Insurance Service. MA to dyslipidemia drugs was measured for two years based on the proportion of days covered (PDC). Conditional logistic regression analysis was performed to analyze the effect of VI on good MA (PDC ≥0.8). The VI group (0.860) had a larger PDC than the non-disabled group (0.850). The adjusted odds ratio (aOR) for good MA among VI vs. non-disabled individuals was statistically insignificant (1.137, 95% confidence interval:0.958-1.350). Significant factors for poor MA in the VI group were younger age (aOR for 20-39 vs. ≥75 years old: 0.124), lower income (aOR for 9-10th decile (rich) vs. 1-4th decile (poor): 1.771), shorter duration of dyslipidemia (aOR for 1-4 vs. 15 years: 0.416), having lower-level providers sas their main providers (aOR for clinics vs. general/tertiary-care hospitals: 0.545), and having mental diseases (aOR: 0.679). Patients with VI did not have worse MA than non-disabled patients taking dyslipidemia medication.

The ICH E9 addendum on estimands in clinical trials provides a framework for precisely defining the treatment effect that is to be estimated, but says little about estimation methods. Here, we report analyses of a clinical trial in type 2 diabetes, targeting the effects of randomized treatment, handling rescue treatment and discontinuation of randomized treatment using the so-called hypothetical strategy. We show how this can be estimated using mixed models for repeated measures, multiple imputation, inverse probability of treatment weighting, G-formula, and G-estimation. We describe their assumptions and practical details of their implementation using packages in R. We report the results of these analyses, broadly finding similar estimates and standard errors across the estimators. We discuss various considerations relevant when choosing an estimation approach, including computational time, how to handle missing data, whether to include post intercurrent event data in the analysis, whether and how to adjust for additional time-varying confounders, and whether and how to model different types of intercurrent event data separately.