Structural Cox Proportional Hazards Models Research Articles

Adjustment for duration of employment in occupational epidemiology

PurposeIn occupational epidemiology, the healthy worker survivor effect can manifest as a time-dependent confounder because healthier workers can accrue greater amounts of exposure over longer periods of employment. For example, in occupational studies of radiation exposure that focus on cumulative annualized radiation dose, workers can accrue greater amounts of cumulative radiation exposure over longer periods of employment, while workers with longer periods of employment can transition into jobs with a reduced potential for annualized radiation exposure. The extent to which confounding arising from the healthy worker survivor effect impacts radiation risk estimates is unknown. MethodsWe assessed the impact of the healthy worker survivor effect on estimates of radiation risk among nuclear workers in a Million Person Study cohort. In simulation studies, we contrasted the ability of marginal structural Cox models with inverse probability weighting and Cox proportional hazards models to account for time-dependent confounding arising from the healthy worker survivor effect. ResultsMarginal structural Cox models and Cox proportional hazards models with flexible functional forms for duration of employment provided reliable results. ConclusionsIt is crucial to flexibly adjust for duration of employment to account for confounding arising from the healthy worker survivor effect in occupational epidemiology.

Air pollution and serious bleeding events in high-risk older adults

ImportanceDespite biological plausibility, very few epidemiologic studies have investigated the risks of clinically significant bleeding events due to particulate air pollution. ObjectiveTo measure the independent and synergistic effects of PM2.5 exposure and anticoagulant use on serious bleeding events. DesignRetrospective cohort study (2008–2016). SettingNationwide Medicare population. ParticipantsA 50% random sample of Medicare Part D-eligible Fee-for-Service beneficiaries at high risk for cardiovascular and thromboembolic events. ExposuresFine particulate matter (PM2.5) and anticoagulant drugs (apixaban, dabigatran, edoxaban, rivaroxaban, or warfarin). Main outcomes and measuresThe outcomes were acute hospitalizations for gastrointestinal bleeding, intracranial bleeding, or epistaxis. Hazard ratios and 95% CIs for PM2.5 exposure were estimated by fitting inverse probability weighted marginal structural Cox proportional hazards models. The relative excess risk due to interaction was used to assess additive-scale interaction between PM2.5 exposure and anticoagulant use. ResultsThe study cohort included 1.86 million high-risk older adults (mean age 77, 60% male, 87% White, 8% Black, 30% anticoagulant users, mean PM2.5 exposure 8.81 μg/m3). A 10 μg/m3 increase in PM2.5 was associated with a 48% (95% CI: 45%–52%), 58% (95% CI: 49%–68%) and 55% (95% CI: 37%–76%) increased risk of gastrointestinal bleeding, intracranial bleeding, and epistaxis, respectively. Significant additive interaction between PM2.5 exposure and anticoagulant use was observed for gastrointestinal and intracranial bleeding. ConclusionsAmong older adults at high risk for cardiovascular and thromboembolic events, increasing PM2.5 exposure was significantly associated with increased risk of gastrointestinal bleeding, intracranial bleeding, and epistaxis. In addition, PM2.5 exposure and anticoagulant use may act together to increase risks of severe gastrointestinal and intracranial bleeding. Thus, clinicians may recommend that high-risk individuals limit their outdoor air pollution exposure during periods of increased PM2.5 concentrations. Our findings may inform environmental policies to protect the health of vulnerable populations.

CausalCmprsk: An R package for nonparametric and Cox-based estimation of average treatment effects in competing risks data

Background and ObjectiveCompeting risks data arise in both observational and experimental clinical studies with time-to-event outcomes, when each patient might follow one of the multiple mutually exclusive competing paths. Ignoring competing risks in the analysis can result in biased conclusions. In addition, possible confounding bias of the treatment-outcome relationship has to be addressed, when estimating treatment effects from observational data. In order to provide tools for estimation of average treatment effects on time-to-event outcomes in the presence of competing risks, we developed the R package causalCmprsk. We illustrate the package functionality in the estimation of effects of a right heart catheterization procedure on discharge and in-hospital death from observational data. MethodsThe causalCmprsk package implements an inverse probability weighting estimation approach, aiming to emulate baseline randomization and alleviate possible treatment selection bias. The package allows for different types of weights, representing different target populations. causalCmprsk builds on existing methods from survival analysis and adapts them to the causal analysis in non-parametric and semi-parametric frameworks. ResultsThe causalCmprsk package has two main functions: fit.cox assumes a semiparametric structural Cox proportional hazards model for the counterfactual cause-specific hazards, while fit.nonpar does not impose any structural assumptions. In both frameworks, causalCmprsk implements estimators of (i) absolute risks for each treatment arm, e.g., cumulative hazards or cumulative incidence functions, and (ii) relative treatment effects, e.g., hazard ratios, or restricted mean time differences. The latter treatment effect measure translates the treatment effect from probability into more intuitive time domain and allows the user to quantify, for example, by how many days or months the treatment accelerates the recovery or postpones illness or death. ConclusionsThe causalCmprsk package provides a convenient and useful tool for causal analysis of competing risks data. It allows the user to distinguish between different causes of the end of follow-up and provides several time-varying measures of treatment effects. The package is accompanied by a vignette that contains more details, examples and code, making the package accessible even for non-expert users.

Comparison of Different Modeling Approaches for Prescription Opioid Use and Its Association With Adverse Events.

Previous research linking opioid prescribing to adverse drug events has failed to properly account for the time-varying nature of opioid exposure. This study aimed to explore how the risk of opioid-related emergency department visits, readmissions, or deaths (composite outcome) varies with opioid dose and duration, comparing different novel modeling techniques. A prospective cohort of 1,511 hospitalized patients discharged from 2 McGill-affiliated hospitals in Montreal, 2014-2016, was followed from the first postdischarge opioid dispensation until 1 year after discharge. Marginal structural Cox proportional hazards models and their flexible extensions were used to explore the association between time-varying opioid use and the composite outcome. Weighted cumulative exposure models assessed cumulative effects of past use and explored how its impact depends on the recency of exposure. The patient mean age was 69.6 (standard deviation= 14.9) years; 57.7% were male. In marginal structural model analyses, current opioid use was associated with a 71% increase in the hazard of opioid-related adverse events (adjusted hazard ratio = 1.71, 95% confidence interval: 1.21, 2.43). The weighted cumulative exposure results suggested that the risk cumulates over the previous 50 days of opioid consumption. Flexible modeling techniques helped assess how the risk of opioid-related adverse events may be associated with time-varying opioid exposures while accounting for nonlinear relationships and the recency of past use.

Association between contact with mental health and substance use services and reincarceration after release from prison.

ObjectivePeople released from prison who experience mental health and substance use problems are at high risk of reincarceration. This study aimed to examine the association between contact with mental health and substance use treatment services, and reincarceration, among adults released from prison.MethodsPre-release survey data from 1,115 adults released from prisons in Queensland, Australia were linked with administrative health and correctional records covering a median of 787 days post-release. We constructed marginal structural Cox proportional hazards models, adjusting for pre-release variables and time-varying indicators of emergent mental health and substance use problems, to examine the association between contact with mental health and substance use treatment services, and reincarceration.ResultsThe adjusted hazard ratio (AHR) for reincarceration associated with mental health service contact was 1.76 (95%CI 1.23,2.51). Among those not on parole following release, the AHR for reincarceration associated with substance use treatment service contact was 3.16 (95%CI 2.09,4.77); we found no evidence for an association among those who were released on parole (AHR = 1.07; 95%CI 0.80,1.43).ConclusionsAlthough we cannot eliminate the possibility of residual confounding, our findings suggest that infrequent or unsustained contact with community-based mental health and substance use treatment services is not protective against reincarceration, and may even be iatrogenic. Increased investment in high-quality and timely behavioural health services for people released from prison may simultaneously improve health outcomes, and reduce reincarceration.

The association of long-acting insulin analogue use versus neutral protamine Hagedorn insulin use and the risk of major adverse cardiovascular events among individuals with type 2 diabetes: A population-based cohort study.

To compare the risk of cardiovascular outcomes associated with long-acting insulin analogues versus neutral protamine Hagedorn (NPH) insulin among patients with type 2 diabetes. We conducted a population-based retrospective cohort study using the UK Clinical Practice Research Datalink Aurum, linked with hospitalization and vital statistics data. Patients with type 2 diabetes who initiated basal insulin treatment between 2002 and 2018 were included in the study. Exposure was defined as current use of long-acting insulin analogues or NPH insulin, defined using a time-varying approach. The primary outcome was major adverse cardiovascular events (MACE; a composite endpoint of myocardial infarction, ischaemic stroke and cardiovascular death). We used a marginal structural Cox proportional hazards model to estimate the hazard ratio (HR) and 95% confidence interval (CI) for MACE with current use of long-acting insulin analogues versus NPH insulin, and in secondary analyses, by long-acting insulin molecule. Our cohort included 57 334 patients. A total of 3494 MACE occurred over a mean follow-up of 1.6 years (incidence rate 37.4, 95% CI 36.2 to 38.7 per 1000 person-years). Long-acting insulin analogues were associated with a decreased risk of MACE compared to NPH insulin (HR 0.89, 95% CI 0.83 to 0.96). Current use of long-acting insulin analogues is associated with a modestly reduced risk of MACE compared to current use of NPH insulin among patients with type 2 diabetes. This study could have important implications for drug plan managers and guideline-writing committees for recommendations of insulin treatment for type 2 diabetes.

Risk of Recurrent Acute Arterial Events Associated With Thiopurines and Anti-Tumor Necrosis Factor in Inflammatory Bowel Diseases

Patients with inflammatory bowel disease (IBD) are at increased risk of acute arterial events. Treatment with anti-tumor necrosis factor (anti-TNF) agents has been associated with a protective effect against the first occurrence of acute arterial events, but the impact of treatment with anti-TNF in patients with a previous history of acute arterial events remains unclear. We assessed the effect of anti-TNF and thiopurines on the risk of recurrent acute arterial events in patients with IBD in a nationwide cohort. Based on the French nationwide health insurance database, patients with IBD and a previous history of an acute arterial event were followed up from January 1, 2009, until December 31, 2018. The risk of acute arterial event recurrence associated with anti-TNF and thiopurine exposure was assessed using marginal structural Cox proportional hazard models adjusted for baseline and time-varying covariates. A total of 27,185 patients were included. During 121,822 person-years (median follow-up period, 4.0 y), 6865 recurrent acute arterial events occurred (incidence rate per 1000 person-years, 56.4; 95% CI, 55.0-57.7). Exposure to both anti-TNF and thiopurines were associated with a decreased risk of recurrent acute arterial events compared with the absence of exposure to either treatment (hazard ratio, 0.75; 95% CI, 0.63-0.90 and hazard ratio, 0.76; 95% CI, 0.66-0.88, respectively). In a nationwide cohort study of patients with IBD and a previous history of an acute arterial event, exposure to both anti-TNF and thiopurines were associated with a decreased risk of recurrent acute arterial events.

Impact of thiopurines and tumour necrosis factor antagonists on primary sclerosing cholangitis outcomes in patients with inflammatory bowel disease.

Patients with primary sclerosing cholangitis (PSC) and inflammatory bowel disease (IBD) are at risk of biliary tract cancer and liver damage (possibly leading to liver transplantation), and are often treated for IBD with thiopurines and/or tumour necrosis factor antagonists (anti-TNF) on a long-term basis. To assess the risk of biliary tract cancer and liver transplantation in patients exposed to thiopurines and/or anti TNF agents in a French nationwide cohort. We performed a population-based study of patients aged 18 years or older with PSC and IBD in the French national health insurance database. Patients were followed from 1 January 2009 to 31 December 2018. The risks of biliary tract cancer and liver transplantation associated with thiopurines and anti-TNF exposure were assessed with marginal structural Cox proportional hazard models, adjusting for baseline demographics and comorbidities, and time-varying medications and PSC activity. Among the 1929 patients with PSC and IBD included, 37 biliary tract cancers and 83 liver transplantations occurred. Compared with patients not exposed to thiopurines or anti-TNF agents, patients exposed to thiopurines (hazard ratio [HR], 1.05; 95% confidence interval [CI], 0.39-2.82) or anti-TNF agents (HR, 0.59; 95% CI, 0.13-2.80) had no excess risk of biliary tract cancer. Similarly, patients exposed to thiopurines (HR, 0.67; 95% CI, 0.30-1.48) or anti-TNF agents (HR, 0.68; CI, 0.22-2.09) had no excess risk of liver transplantation. Patients with PSC and IBD who are exposed to thiopurines or anti-TNF agents are not at excess risk of biliary tract cancer or liver transplantation.

Adherence to a priori-Defined Diet Quality Indices Throughout the Early Disease Course Is Associated With Survival in Head and Neck Cancer Survivors: An Application Involving Marginal Structural Models.

No studies, to date, have scrutinized the role of a priori dietary patterns on prognosis following a head and neck squamous cell carcinoma (HNSCC) diagnosis. The purpose of this analysis was to evaluate the associations between adherence to six a priori defined diet quality indices (including AHEI-2010, aMED, DASH, and three low-carbohydrate indices) throughout the first 3 years of observation and all-cause and cancer-specific mortalities in 468 newly diagnosed HNSCC patients from the University of Michigan Head and Neck Specialized Program of Research Excellence (UM-SPORE). The dietary intake data were measured using a food frequency questionnaire administered at three annual time points commencing at study entry. Deaths and their causes were documented throughout the study using various data sources. Marginal structural Cox proportional hazards models were used to evaluate the role of diet quality, as a time-varying covariate, on mortality. There were 93 deaths from all causes and 74 cancer-related deaths adjudicated throughout the observation period. There was a strong inverse association between adherence to the AHEI-2010, all-cause mortality (HRQ5–Q1:0.07, 95% CI:0.01–0.43, ptrend:0.04), and cancer-specific mortality (HRQ5–Q1:0.15, 95% CI:0.02–1.07, ptrend:0.04). Other more modest associations were noted for the low-carbohydrate indices. In sum, higher adherence to the AHEI-2010 and a plant-based low-carbohydrate index throughout the first 3 years since diagnosis may bolster survival and prognosis in newly diagnosed patients with HNSCC.

Depressive symptoms are no longer a barrier to HCV treatment initiation in the HIV–HCV co-infected population in Canada

Background Psychiatric illness was a major barrier for HCV treatment during the Interferon (IFN) treatment era due to neuropsychiatric side effects. While direct acting antivirals (DAA) are better tolerated, patient-level barriers persist. We aimed to assess the effect of depressive symptoms on time to HCV treatment initiation among HIV–HCV co-infected persons during the IFN (2003–2011) and second-generation DAA (2013–2020) eras. Methods We used data from the Canadian Co-infection Cohort, a multicentre prospective cohort, and its associated sub-study on Food Security (FS). We predicted Center for Epidemiologic Studies Depression Scale-10 (CES-D-10) classes for depressive symptoms indicative of a depression risk using a random forest classifier and corrected for misclassification using predictive value-based record-level correction. We used marginal structural Cox proportional hazards models with inverse weighting for competing risks (death) to assess the effect of depressive symptoms on treatment initiation among HCV RNA-positive participants. Results We included 590 and 1127 participants in the IFN and DAA eras. The treatment initiation rate increased from 9 (95% confidence interval (CI): 7–10) to 21 (95% CI: 19–22) per 100 person-years from the IFN to DAA era. Treatment initiation was lower among those with depressive symptoms compared to those without in the IFN era (hazard ratio: 0.81 (95% CI: 0.69–0.95)) and was higher in the DAA era (1.19 (95% CI: 1.10–1.27)). Conclusion Depressive symptoms no longer appear to be a barrier to HCV treatment initiation in the co-infected population in the DAA era. The higher rate of treatment initiation in individuals with depressive symptoms suggests those previously unable to tolerate IFN are now accessing treatment.

Impact of Anti-tumour Necrosis Factor Agents on the Risk of Colorectal Cancer in Patients with Ulcerative Colitis: Nationwide French Cohort Study.

Patients with ulcerative colitis [UC] are at increased risk of colorectal cancer. Anti-tumour necrosis factor agents [anti-TNF] aim to reduce chronic colonic inflammation and may lower the risk of colorectal cancer [CRC], but the impact of anti-TNF exposure has not yet been assessed in population-based cohort studies. The aim of this nationwide study was to assess the risk of CRC in patients with UC exposed to anti-TNF. Based on the French health insurance database, patients aged 18 years or older with a diagnosis of UC, previously exposed to or initiating immunosuppressive treatment, were followed from January 1, 2009, until December 31, 2018. The risk of CRC associated with anti-TNF exposure was assessed using marginal structural Cox proportional hazard models adjusting for baseline and time-varying comorbidities including primary sclerosing cholangitis, UC disease activity, colonoscopic surveillance, and other medications. Among 32 403 patients with UC, 15 542 [48.0%] were exposed to anti-TNF. During a median follow-up of 6.1 years [198 249 person-years], 246 incident CRC occurred (incidence rate per 1000 person-years, 1.24; 95% confidence intrval [CI], 1.10-1.41). Whereas the risk of CRC associated with anti-TNF exposure was not decreased in the overall group of patients with UC (hazard ratio [HR], 0.85; 95% CI, 0.58-1.26), anti-TNF exposure was associated with a decreased risk of CRC in patients with long-standing colitis [disease duration ≥10 years] [HR, 0.41; 95% CI, 0.20-0.86]. In a nationwide cohort of patients with UC, anti-TNF exposure was associated with a decreased risk of CRC in patients with long-standing colitis.

Sunscreens With High Versus Low Sun Protection Factor and Cutaneous Squamous Cell Carcinoma Risk: A Population-Based Cohort Study.

Evidence on sunscreen use and cutaneous squamous cell carcinoma (cSCC) risk is limited. Most studies have not taken sun protection factor (SPF) into consideration and used nonusers of sunscreen as the reference group. Nonusers are likely a priori at lower cSCC risk than users. No study has investigated the effect of high- versus low-SPF sunscreens on cSCC, appropriately adjusting for time-varying confounding. Using data from the Norwegian Women and Cancer Study (1991–2016), we investigated whether use of SPF ≥15 versus SPF <15 sunscreens reduces cSCC risk. We used a marginal structural Cox proportional hazards model with inverse probability of treatment and censoring weights to estimate hazard ratios (HRs) and 95% confidence intervals (CIs). During follow-up of 148,781 women (mean follow-up, 14.3 years), 653 women were diagnosed with cSCC. The effect on cSCC risk of sunscreens with SPF ≥15 versus SPF <15 was close to the null when used at any latitudes (HR = 1.02, 95% CI: 0.82, 1.27) and when used in lower-latitude settings (HR = 1.05, 95% CI: 0.84, 1.32). In conclusion, we found no indication that sunscreens with SPF ≥15 reduced Norwegian women’s cSCC risk more than sunscreens with SPF <15, suggesting that either there is no difference in their effects long-term or the difference is diluted by incorrect application.

Immunosuppression does not prevent severe gastrointestinal tract involvement in systemic sclerosis.

We aimed to test the hypothesis that exposure to immunosuppression in early systemic sclerosis (SSc) could modify the risk of developing new onset severe gastrointestinal (GIT) involvement. A total of 762 subjects with <3 years of disease duration and without severe GIT disease at baseline study visit were identified from combined longitudinal cohort data from the Canadian Scleroderma Research Group (CSRG) and Australian Scleroderma Interest Group (ASIG). The primary exposure was ever use of methotrexate, cyclophosphamide, mycophenolate mofetil and/or azathioprine during the study period. Severe GIT disease was defined as: 1-malabsorption, 2-hyperalimentation, 3-pseudo-obstruction, and/or 4-≥10% weight loss in association with the use of antibiotics for bacterial overgrowth or oesophageal stricture. The change in the hazard of severe GIT disease due to exposure was estimated using a marginal structural Cox proportional hazards model fit by inverse probability of treatment weights (IPTW) to address potential confounding. Study subjects were 81.5% female, had a mean age of 53.7±13.0 years and mean disease duration at baseline of 1.4±0.8 years. During a mean follow-up of 4.0±2.6 years, severe GIT involvement developed in 11.6% of the 319 subjects exposed to immunosuppression and in 6.8% of the 443 unexposed subjects. In an IPTW-adjusted analysis, exposure to immunosuppression was not associated with severe GIT disease (weighted hazard ratio 0.91, 95% confidence interval 0.52-1.58). In this large inception SSc cohort, the risk of severe GIT involvement was not modified by exposure to immunosuppression.

Comparative Risk of Serious Infections With Tumor Necrosis Factor α Antagonists vs Vedolizumab in Patients With Inflammatory Bowel Diseases

We conducted a retrospective cohort study comparing the risk of serious infections between patients treated with tumor necrosis factor-a (TNFa) antagonists vs. vedolizumab in patients with inflammatory bowel diseases (IBD). Using an administrative claims database, we identified patients with IBD who were new-users of either TNFa antagonists or vedolizumab between 2014-2018 and had insurance coverage for at least 1y before and after treatment initiation. We compared the risk of serious infections (infections requiring hospitalization) between patients treated with vedolizumab or TNFa antagonists using marginal structural Cox proportional hazard models adjusted for baseline disease characteristics, healthcare utilization, comorbidities, and time-varying use of corticosteroids, immunomodulators and opiates. We included 4881 patients treated with TNFa antagonists (age, 41 ± 15y, 60% with Crohn's disease [CD]) of whom 434 developed serious infections over 5786 person-year [PY] follow-up, and 1106 patients treated with vedolizumab (age, 44 ± 16y, 39% with CD) of whom 86 developed serious infections over 1040-PY follow-up. Vedolizumab was associated with 46% lower risk of serious infections as compared with TNFa antagonists in patients with ulcerative colitis (HR,0.54 [95% CI,0.35-0.83), but no significant differences were observed in patients with CD (HR,1.30 [0.80-2.11]). Vedolizumab was associated with lower risk of extra-intestinal serious infections in patients with UC, but higher risk of gastrointestinal serious infections in patients with CD. In an observational study of patients with IBD, vedolizumab was associated with lower risk of serious infections as compared with TNFa antagonists, in patients with UC, but not in patients with CD.

Survival on four compared with three times per week haemodialysis in high ultrafiltration patients: an observational study.

BackgroundThe harm caused by the long interdialytic interval in three-times-per-week haemodialysis regimens (3×WHD) may relate to fluid accumulation and associated high ultrafiltration rate (UFR). Four-times-per-week haemodialysis (4×WHD) may offer a solution, but its impact on mortality, hospitalization and vascular access complications is unknown.MethodsFrom the AROii cohort of incident in-centre haemodialysis patients, 3×WHD patients with a UFR >10 mL/kg/h were identified. The hazard for the outcomes of mortality, hospitalization and vascular access complications in those who switched to 4×WHD compared with staying on 3×WHD was estimated using a marginal structural Cox proportional hazards model. Adjustment included baseline patient and treatment characteristics with inverse probability weighting used to adjust for time-varying UFR and cardiovascular comorbidities.ResultsFrom 10 637 European 3×WHD patients, 3842 (36%) exceeded a UFR >10 mL/kg/h. Of these, 288 (7.5%) started 4×WHD and at baseline were more comorbid. Event rates while receiving 4×WHD compared with 3×WHD were 12.6 compared with 10.8 per 100 patient years for mortality, 0.96 compared with 0.65 per year for hospitalization and 14.7 compared with 8.0 per 100 patient years for vascular access complications. Compared with 3×WHD, the unadjusted hazard ratio (HR) for mortality on 4×WHD was 1.05 [95% confidence interval (CI) 0.78–1.42]. Following adjustment for baseline demographics, time-varying treatment probability and censoring risks, this HR was 0.73 (95% CI 0.50–1.05; P = 0.095). Despite these adjustments on 4×WHD, the HR for hospitalization remained elevated and vascular access complications were similar to 3×WHD.ConclusionsThis observational study was not able to demonstrate a mortality benefit in patients switched to 4×WHD. To demonstrate the true benefits of 4×WHD requires a large, well-designed clinical trial. Our data may help in the design of such a study.

Permanent discontinuation of different anticoagulants in patients with atrial fibrillation and the impact on clinical outcome: data from the GARFIELD-AF registry

Abstract Background Non-Vitamin K Antagonists (NOAC) are replacing vitamin K Antagonists (VKA) as first line oral anticoagulant therapy (OAC) in patients with non-valvular atrial fibrillation (NVAF). Discontinuation of OAC might put patients at increased risk. It was anticipated that patients who were on NOAC would discontinue OAC less. Purpose We compare the rates and impact on outcome of the discontinuation of NOAC and VKA using data from the GARFIELD-AF registry. Methods Patients included in GARFIELD-AF, had a new diagnosis of NVAF and at least 1 stroke risk factor. In this analysis 26,299 patients (VKA: 13,012; NOAC: 13,287) that received OAC were included. Permanent discontinuation was defined as stopping OAC for at least 7 consecutive days (whether or not restarted during follow-up). Marginal structural Cox proportional hazards models estimated the effect of discontinuation on death, cardiovascular (CV) death, non-haemorrhagic stroke + systemic embolism (NHS+SE), myocardial infarction (MI), or combined endpoints. Adjustments were made for both baseline factors and time dependent variables. Results Of all patients, 15.6% discontinued OAC (VKA: 15.4%; NOAC: 15.8%) over a median follow-up of 181 days (IQR: 359). Most discontinued early (67.0% of patients on VKA and 47.1% of patients on NOAC ≤4 months). Significantly higher discontinuation risk was seen with worsening kidney function, coronary artery disease, history of bleeding (baseline factors), as well as with all types of bleeding (time dependent factors). Lower discontinuation rates were seen with history of stroke/TIA, hypertension, increasing age, permanent AF (all p&amp;lt;0.01). Mean CHA2DS2-VASc score was 3 in all groups. Patients in both treatment arms who discontinued were at increased risk for death, NHS+SE, MI as well as combined endpoints of death/NHS+SE/MI, death/NHS+SE and a trend towards higher CV death (Figure 1). All interaction tests for the interaction of treatment and discontinuation had a p value &amp;gt;0.4. The association between discontinuation and outcomes did not change when a 30 day discontinuation window was used. Conclusion The rate of discontinuation in this study was 15.8% and comparable for VKA and NOAC over a 2-year follow-up. Discontinuation rates were the highest soon after the initiation of treatment. When VKA or NOAC was stopped for ≥7 consecutive days, the risk of NHS+SE, death, MI or any combined endpoints were significantly worse in both treatment arms. These data suggest that discontinuation of anticoagulant treatment with VKA or NOAC should be discouraged. HR of patients who discontinued OAC Funding Acknowledgement Type of funding source: Private grant(s) and/or Sponsorship. Main funding source(s): The GARFIELD-AF registry is funded by an unrestricted research grant from Bayer AG.

S0861 Comparative Risk of Serious Infections With Tumor Necrosis Factor-α Antagonists vs Vedolizumab in Patients With Inflammatory Bowel Diseases

INTRODUCTION: Comparative safety of therapies is an important attribute in choosing optimal therapy in patients with inflammatory bowel diseases (IBD). We conducted a retrospective cohort study comparing the risk of serious infections between patients treated with tumor necrosis factor-α (TNFα) antagonists and those treated with vedolizumab. METHODS: Using an administrative claims database (OptumLabs Data Warehouse) we identified patients with IBD who were new-users of either TNFα antagonists or vedolizumab between 2014 and 2019 and had insurance coverage for at least 1y before and 1y after treatment initiation. We compared the risk of serious infections between patients treated with TNFα antagonists and those treated with vedolizumab using marginal structural Cox proportional hazard models adjusted for ulcerative colitis (UC) versus Crohn’s disease (CD), baseline demographics, healthcare utilization, comorbidities and time-varying use of IBD-related medications including corticosteroids and immunomodulators. RESULTS: We included 4896 patients treated with TNFα antagonists (age, 41 ± 15 y, 51% male, 60% with CD) of whom 434 developed serious infections over a 6009 person-year [py] follow-up, and 1106 patients treated with vedolizumab (age, 44 ± 16 y, 49% male, 39% with CD) of whom 86 developed serious infections over a 1077 py follow-up. On adjusted analysis, TNFα antagonists were associated with 2.4-fold higher risk of serious infections (hazard ratio [HR], 2.38 [95% CI, 1.38–4.13]) when compared with vedolizumab in all patients with IBD, and in subgroup analyses of patients with CD (HR, 2.28 [1.13–4.62]) and UC (HR, 1.62 [1.03–2.5]). This was primarily driven by increased risk of extra-intestinal serious infections (HR, 3.13 [1.48–6.63]), without any increase in risk of serious gastrointestinal infections (HR, 0.60 [0.33–1.08]). CONCLUSION: In a large claims-based study, TNFα antagonists were associated with increased risk of serious infections as compared with vedolizumab, in patients with CD and UC, particularly risk of extra-intestinal infections.

Combination therapy with rifampicin or fosfomycin in patients with Staphylococcus aureus bloodstream infection at high risk for complications or relapse: results of a large prospective observational cohort

To investigate whether Staphylococcus aureus bloodstream infection (SAB) patients at high risk for complications or relapse benefit from combination therapy with adjunctive rifampicin or fosfomycin. In this post hoc analysis, SAB patients with native valve infective endocarditis, osteoarticular infections or implanted foreign devices were included. The co-primary endpoints were all-cause 90 day mortality and death or SAB-related late complications within 180 days. To overcome treatment selection bias and account for its time dependence, inverse probability of treatment weights were calculated and included in marginal structural Cox proportional hazard models (MSCMs). A total of 578 patients were included in the analysis, of which 313 (54%) received combination therapy with either rifampicin (n = 242) or fosfomycin (n = 58). In the multivariable MSCM, combination therapy was associated with a better outcome, that is, a lower rate of death or SAB-related late complications within 180 days (HR 0.65, 95% CI 0.46-0.92). This beneficial effect was primarily seen in patients with implanted foreign devices, in which combination therapy was associated with a lower rate of death or SAB-related late complications within 180 days (HR 0.53, 95% CI 0.35-0.79) and a lower 90 day mortality (HR 0.57, 95% CI 0.36-0.91). Upon agent-specific stratification, we found no significant differences in outcomes between combination therapy containing rifampicin and fosfomycin; however, the number of patients in most subgroups was not large enough to draw firm conclusions. In patients with implanted foreign devices, combination therapy was associated with a better long-term outcome. Larger prospective studies are needed to validate these findings.

Risk of acute arterial events associated with treatment of inflammatory bowel diseases: nationwide French cohort study

ObjectivePatients with IBD are at increased risk of acute arterial events. Antitumour necrosis factor (TNF) agents and thiopurines may, via their anti-inflammatory properties, lower the risk of acute arterial events....

Smoking status and cause-specific discontinuation of tumour necrosis factor inhibitors in axial spondyloarthritis

BackgroundThe impact of smoking on TNF inhibition (TNFi) therapy is unclear. We examined the effect of smoking on all-cause and cause-specific TNFi discontinuation in axial spondyloarthritis (axSpA).MethodsWe used longitudinal data from the British Society for Rheumatology Biologics Register for Ankylosing Spondylitis (BSRBR-AS). Patients fulfilling the ASAS criteria for axSpA, who started their first TNFi, were eligible for analysis. Inverse-probability weights were used to balance differences in baseline disease severity and other confounders. We used marginal structural Cox proportional hazard models to estimate hazard ratios (HR) for TNFi discontinuation according to smoking status. In analyses of cause-specific discontinuation, competing risk events were considered as censoring, using inverse-probability weights.ResultsA total of 758 participants were included in the analysis (66% male, mean age 45 years), providing 954 patient-years of follow-up. TNFi was discontinued in 174 (23%) patients, among whom 26% stopped due to infections, 20% due to other adverse events and 44% due to inefficacy or other reasons. Thirty-four percent were current smokers and 30% ex-smokers. Compared to never smokers, current smokers’ risk of TNFi discontinuation was HR 0.79 (95%CI 0.53 to 1.20) and ex-smokers HR 0.68 (95%CI 0.45 to 1.04). Our data did not show evidence that current smoking influenced discontinuation due to infections (HR 0.79, 95%CI 0.40 to 1.54), other adverse events (HR 0.86, 95%CI 0.41 to 1.78) or inefficacy/other causes (HR 1.44, 95%CI 0.86 to 2.41).ConclusionBaseline smoking status did not impact TNFi discontinuation in this UK cohort of axSpA participants.