scholarly journals Methodological comparison of marginal structural model, time-varying Cox regression, and propensity score methods: the example of antidepressant use and the risk of hip fracture

2016 ◽  
Vol 25 ◽  
pp. 114-121 ◽  
Author(s):  
M. Sanni Ali ◽  
Rolf H. H. Groenwold ◽  
Svetlana V. Belitser ◽  
Patrick C. Souverein ◽  
Elisa Martín ◽  
...  
Keyword(s):  
Hip Fracture ◽  
Propensity Score ◽  
Structural Model ◽  
Cox Regression ◽  
Time Varying ◽  
Marginal Structural Model ◽  
Propensity Score Methods ◽  
Antidepressant Use ◽  
Methodological Comparison

P4745Concomitant oral anticoagulant and antidepressant therapy in patients with atrial fibrillation and risk of stroke and bleeding: a population based cohort study

2019 ◽  
Vol 40 (Supplement_1) ◽  
Author(s):  
J J Komen ◽  
P Hjemdahl ◽  
A K Mantel - Teeuwisse ◽  
O H Klungel ◽  
B Wettermark ◽  
...  
Keyword(s):  
Atrial Fibrillation ◽  
Propensity Score ◽  
Oral Anticoagulant ◽  
Cox Regression ◽  
Oral Anticoagulants ◽  
Anticoagulant Treatment ◽  
Oral Anticoagulant Treatment ◽  
Increased Risk ◽  
Antidepressant Use

Abstract Background Anticoagulation treatment reduces the risk of stroke but increases the risk of bleeding in atrial fibrillation (AF) patients. Antidepressants use is associated with increased risk for stroke and bleeds. Objective To assess the association between antidepressant use in AF patients with oral anticoagulants and bleeding and stroke risk. Methods All AF patients newly prescribed with an oral anticoagulant in the Stockholm Healthcare database (n=2.3 million inhabitants) from July 2011 until 2016 were included and followed for one year or shorter if they stopped claiming oral anticoagulant treatment or had an outcome of interest. Outcomes were severe bleeds and strokes, requiring acute hospital care. During follow-up, patients were considered exposed to antidepressant after claiming a prescription for the duration of the prescription. With a time-varying Cox regression, we assessed the association between antidepressant use and strokes and bleeds, adjusting for confounders (i.e., age, sex, comorbidities, comedication, and year of inclusion). In addition, we performed a propensity score matched analysis to test the robustness of our findings. Results Of the 30,595 patients included after claiming a prescription for a NOAC (n=13,506) or warfarin (n=17,089), 4 303 claimed a prescription for an antidepressant during follow-up. A total of 712 severe bleeds and 551 strokes were recorded in the cohort. Concomitant oral anticoagulant and antidepressant use was associated with increased rates of severe bleeds (4.7 vs 2.7 per 100 person-years) compared to oral anticoagulant treatment without antidepressant use (aHR 1.42, 95% CI: 1.12–1.80), but not significantly associated with increased stroke rates (3.5 vs 2.1 per 100 person-years, aHR 1.23, 95% CI: 0.93–1.62). No significant differences were observed between different oral anticoagulant classes (i.e., warfarin or NOAC) or different antidepressant classes (i.e., SSRI, TCA, or other antidepressant). Additional propensity-score matched analyses yielded similar results but showed a significantly increased risk for stroke (HR: 1.47, 95% CI: 1.08–2.02). Incidence rates of strokes and bleeds Conclusion Concomitant use of an oral anticoagulant and an antidepressant, irrespective of type, is associated with an increased bleeding risk. Increased awareness and a critical consideration for the need of an antidepressant is recommended in this population. Acknowledgement/Funding Swedish Heart Lung Foundation

scholarly journals Ribavirin and Interferon Therapy for Critically Ill Patients With Middle East Respiratory Syndrome: A Multicenter Observational Study

2019 ◽  
Vol 70 (9) ◽  
pp. 1837-1844 ◽  
Author(s):  
Yaseen M Arabi ◽  
Sarah Shalhoub ◽  
Yasser Mandourah ◽  
Fahad Al-Hameed ◽  
Awad Al-Omari ◽  
...  
Keyword(s):  
Middle East ◽  
Observational Study ◽  
Critically Ill ◽  
Critically Ill Patients ◽  
Structural Model ◽  
Middle East Respiratory Syndrome ◽  
Time Varying ◽  
Marginal Structural Model ◽  
Recombinant Interferon ◽  
Multicenter Observational Study

Abstract Background The objective of this study was to evaluate the effect of ribavirin and recombinant interferon (RBV/rIFN) therapy on the outcomes of critically ill patients with Middle East respiratory syndrome (MERS), accounting for time-varying confounders. Methods This is a retrospective cohort study of critically ill patients with laboratory-confirmed MERS from 14 hospitals in Saudi Arabia diagnosed between September 2012 and January 2018. We evaluated the association of RBV/rIFN with 90-day mortality and MERS coronavirus (MERS-CoV) RNA clearance using marginal structural modeling to account for baseline and time-varying confounders. Results Of 349 MERS patients, 144 (41.3%) patients received RBV/rIFN (RBV and/or rIFN-α2a, rIFN-α2b, or rIFN-β1a; none received rIFN-β1b). RBV/rIFN was initiated at a median of 2 days (Q1, Q3: 1, 3 days) from intensive care unit admission. Crude 90-day mortality was higher in patients with RBV/rIFN compared to no RBV/rIFN (106/144 [73.6%] vs 126/205 [61.5%]; P = .02]. After adjusting for baseline and time-varying confounders using a marginal structural model, RBV/rIFN was not associated with changes in 90-day mortality (adjusted odds ratio, 1.03 [95% confidence interval {CI}, .73–1.44]; P = .87) or with more rapid MERS-CoV RNA clearance (adjusted hazard ratio, 0.65 [95% CI, .30–1.44]; P = .29). Conclusions In this observational study, RBV/rIFN (RBV and/or rIFN-α2a, rIFN-α2b, or rIFN-β1a) therapy was commonly used in critically ill MERS patients but was not associated with reduction in 90-day mortality or in faster MERS-CoV RNA clearance.

Regression-based Adjustment for Time-varying Confounders

2018 ◽  
Vol 49 (4) ◽  
pp. 906-946 ◽  
Author(s):  
Geoffrey T. Wodtke
Keyword(s):  
Structural Model ◽  
Panel Study ◽  
Time Varying ◽  
Marginal Structural Model ◽  
Income Dynamics ◽  
Social Scientists ◽  
Treatment Assignment ◽  
Conditional Mean ◽  
Marginal Effects ◽  
Conventional Regression

Social scientists are often interested in estimating the marginal effects of a time-varying treatment on an end-of-study continuous outcome. With observational data, estimating these effects is complicated by the presence of time-varying confounders affected by prior treatments, which may lead to bias in conventional regression and matching estimators. In this situation, inverse-probability-of-treatment-weighted (IPTW) estimation of a marginal structural model remains unbiased if treatment assignment is sequentially ignorable and the conditional probability of treatment is correctly modeled, but this method is not without limitations. In particular, it is difficult to use with continuous treatments, and it is relatively inefficient. This article explores using an alternative regression-based method—regression-with-residuals (RWR) estimation of a constrained structural nested mean model—that may overcome some of these limitations in practice. It is unbiased for the marginal effects of a time-varying treatment if treatment assignment is sequentially ignorable, the treatment effects of interest are invariant across levels of the confounders, and a model for the conditional mean of the outcome is correctly specified. The performance of RWR estimation relative to IPTW estimation is evaluated with a series of simulation experiments and with an empirical example based on longitudinal data from the Panel Study of Income Dynamics. Results indicate that it may outperform IPTW estimation in certain situations.

scholarly journals Adjusting Time-Varying Confounding by Marginal Structural Model in Observational Study

2018 ◽  
Vol 23 (Supplement) ◽  
pp. s116-s117
Author(s):  
Shih-Wei Chiu ◽  
Takuhiro Yamaguchi
Keyword(s):  
Observational Study ◽  
Structural Model ◽  
Time Varying ◽  
Marginal Structural Model

scholarly journals Hip fracture in patients with non-dialysis chronic kidney disease stage 5

2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Chao-Hsiun Tang ◽  
Che-Yi Chou
Keyword(s):  
Chronic Kidney Disease ◽  
Hip Fracture ◽  
Kidney Disease ◽  
Propensity Score ◽  
Competing Risks ◽  
Hip Fractures ◽  
Cox Regression ◽  
Female Gender ◽  
Ckd Stage ◽  
History Of

AbstractHip fracture is a significant health problem and is associated with increased mortality. Patients with chronic kidney disease (CKD) are more at risk of hip fracture than the general population, but the hip fracture risk is not evident among non-dialysis CKD stage 5 patients. This study aims to assess the risk of hip fracture in patients with non-dialysis CKD stage 5 comparing to those with CKD stages 1–4. Patients with non-dialysis CKD stage 5 and CKD stages 1–4 were retrieved from Taiwan longitudinal health insurance database 2011–2014. All patients were followed to the end of 2018 for the development of hip fractures. We analyze the risk of hip fracture of propensity score-matched patients with CKD stage 5 compared to patients with CKD stages 1–4 using stepwise Cox regression and competing risks regression. We analyzed 5649 propensity score-matched non-dialysis CKD 1–4 patients and non-dialysis CKD 5 patients between 2011 and 2014. All patients were followed to the end of 2018, 229 (4.1%) of CKD 1–4 patients in 21,899 patient-year, and 290 (5.1%) of CKD 5 patients had hip fractures in 18,137 patient-year. CKD 5 patients had a higher risk of hip fracture than patients with CKD stages 1–4. The adjusted HR was 1.53 (95% CI 1.08–1.54) in the Cox regression with adjustments for age, gender, comorbidity, and history of fracture. In the competing risks regression, the subdistribution hazard ratio was 1.29 (95% CI 1.08–1.54). Female gender, age, history of fractures, and Charlson–Deyo comorbidity index were independently associated with increased hip fracture risks. Non-dialysis CKD 5 patients had a higher risk of hip fracture than patients with CKD stages 1–4. This association is independent of patients’ age, female gender, history of fractures, and comorbidities.

A marginal structural model approach to analyse work-related injuries: an example using data from the health and retirement study

2019 ◽  
Vol 26 (3) ◽  
pp. 248-253
Author(s):  
Navneet Kaur Baidwan ◽  
Susan Goodwin Gerberich ◽  
Hyun Kim ◽  
Andrew D Ryan ◽  
Timothy Church ◽  
...  
Keyword(s):  
Structural Model ◽  
Research Effort ◽  
Physical Work ◽  
Health And Retirement Study ◽  
Effect Estimate ◽  
Study Cohort ◽  
Time Varying ◽  
Marginal Structural Model ◽  
Work Related ◽  
Retirement Study

BackgroundBiases may exist in the limited longitudinal data focusing on work-related injuries among the ageing workforce. Standard statistical techniques may not provide valid estimates when the data are time-varying and when prior exposures and outcomes may influence future outcomes. This research effort uses marginal structural models (MSMs), a class of causal models rarely applied for injury epidemiology research to analyse work-related injuries.Methods7212 working US adults aged ≥50 years, obtained from the Health and Retirement Study sample in the year 2004 formed the study cohort that was followed until 2014. The analyses compared estimates measuring the associations between physical work requirements and work-related injuries using MSMs and a traditional regression model. The weights used in the MSMs, besides accounting for time-varying exposures, also accounted for the recurrent nature of injuries.ResultsThe results were consistent with regard to directionality between the two models. However, the effect estimate was greater when the same data were analysed using MSMs, built without the restriction for complete case analyses.ConclusionsMSMs can be particularly useful for observational data, especially with the inclusion of recurrent outcomes as these can be incorporated in the weights themselves.

scholarly journals Associations between Community-Acquired Pneumonia and Proton Pump Inhibitors in the Laryngeal/Voice-Disordered Population

2018 ◽  
Vol 160 (3) ◽  
pp. 519-525 ◽  
Author(s):  
Seth M. Cohen ◽  
Hui-Jie Lee ◽  
David A. Leiman ◽  
Nelson Roy ◽  
Stephanie Misono
Keyword(s):  
Proton Pump ◽  
Index Date ◽  
Structural Model ◽  
Cox Regression ◽  
Absolute Risk ◽  
Cohort Analysis ◽  
International Classification Of Diseases ◽  
Community Acquired Pneumonia ◽  
Voice Disorder ◽  
Marginal Structural Model

Objectives To examine the relationship between community-acquired pneumonia (CAP) and proton pump inhibitor (PPI) treatment among patients with laryngeal/voice disorders. Study Design Retrospective cohort analysis. Setting Large national administrative US claims database. Subjects and Methods Patients were included if they were ≥18 years old; had outpatient treatment for a laryngeal/voice disorder from January 1, 2010, to December 31, 2014 (per International Classification of Diseases, Ninth Revision, Clinical Modification codes); had 12 months of continuous enrollment prior to the index date (ie, first diagnosis of laryngeal/voice disorder); had no preindex diagnosis of CAP; and had prescription claims captured from 1 year preindex to end of follow-up. Patient demographics, comorbid conditions, index laryngeal diagnosis, number of unique preindex patient encounters, and CAP diagnoses during the postindex 3 years were collected. Two models—a time-dependent Cox regression model and a propensity score–based approach with a marginal structural model—were separately performed for patients with and without pre–index date PPI prescriptions. Results A total of 392,355 unique patients met inclusion criteria; 188,128 (47.9%) had a PPI prescription. The 3-year absolute risk for CAP was 4.0% and 5.3% among patients without and with preindex PPI use, respectively. For patients without and with pre–index date PPI use, the CAP occurrence for a person who had already received a PPI is 30% to 50% higher, respectively, than for a person who had not yet had a PPI but may receive one later. Conclusions Patients without and with pre–index date PPI use experienced a roughly 30% to 50% increased likelihood of CAP, respectively, as compared with patients who had not had PPI prescriptions.

scholarly journals Early use of acute medication for preventing migraine attacks: Results from a diary-based cohort study

2020 ◽  
Vol 3 ◽  
pp. 251581632094492
Author(s):  
Susanna Di Termini ◽  
Christian Wöber ◽  
Werner Brannath
Keyword(s):  
Regression Analysis ◽  
Structural Model ◽  
Cox Regression ◽  
Episodic Migraine ◽  
Severe Headache ◽  
Marginal Structural Model ◽  
Cox Regression Analysis ◽  
Acute Medication ◽  
Early Use ◽  
Prospective Trials

Background: Treating migraine attacks early may improve outcome. The aim of this analysis was to investigate whether certain premonitory symptoms could be indicators for taking acute medication. Methods: We analyzed 3-month diary data recorded by 271 patients with episodic migraine and looked at all migraine-free intervals. For investigating the interaction between acute medication and neck discomfort associated with sensitivity to lights, noises, or odors, we used a marginal structural model and a Cox regression analysis adjusted for moderate or severe headache. Results: The patients (mean age 43 ± 15.4 years, 88% women) recorded a total of 20,219 diary days without migraine. In the marginal structural model analysis, the risk for occurrence of a migraine attack on the subsequent day was reduced when acute medication was used in the presence of neck discomfort associated with sensitivity to lights (hazard ratio 0.4; 95% confidence interval 0.2–0.7), noises (0.4; 0.3–0.7), or odors (0.2; 0.1–0.4). The marginal structural model showed lower risk of migraine attacks than the Cox regression analysis adjusted for moderate or severe headache in the majority of the cases. Conclusion: Migraine attacks may be prevented when acute medication is used in the presence of neck discomfort associated with sensitivity to lights, noises, or odors. The results of this study may stimulate further prospective trials.

scholarly journals Erythropoietin Dose and Mortality in Hemodialysis Patients: Marginal Structural Model to Examine Causality

2016 ◽  
Vol 2016 ◽  
pp. 1-8 ◽  
Author(s):  
Elani Streja ◽  
Jongha Park ◽  
Ting-Yan Chan ◽  
Janet Lee ◽  
Melissa Soohoo ◽  
...  
Keyword(s):  
Mortality Risk ◽  
Structural Model ◽  
Positive Association ◽  
Hemodialysis Patients ◽  
Adverse Outcomes ◽  
Time Varying ◽  
Marginal Structural Model ◽  
Time Varying Covariates ◽  
Dose Dependent

It has been previously reported that a higher erythropoiesis stimulating agent (ESA) dose in hemodialysis patients is associated with adverse outcomes including mortality; however the causal relationship between ESA and mortality is still hotly debated. We hypothesize ESA dose indeed exhibits a direct linear relationship with mortality in models of association implementing the use of a marginal structural model (MSM), which controls for time-varying confounding and examines causality in the ESA dose-mortality relationship. We conducted a retrospective cohort study of 128 598 adult hemodialysis patients over a 5-year follow-up period to evaluate the association between weekly ESA (epoetin-α) dose and mortality risk. A MSM was used to account for baseline and time-varying covariates especially laboratory measures including hemoglobin level and markers of malnutrition-inflammation status. There was a dose-dependent positive association between weekly epoetin-αdoses ≥18 000 U/week and mortality risk. Compared to ESA dose of <6 000 U/week, adjusted odds ratios (95% confidence interval) were 1.02 (0.94–1.10), 1.08 (1.00–1.18), 1.17 (1.06–1.28), 1.27 (1.15–1.41), and 1.52 (1.37–1.69) for ESA dose of 6 000 to <12 000, 12 000 to <18 000, 18 000 to <24 000, 24 000 to <30 000, and ≥30 000 U/week, respectively. High ESA dose may be causally associated with excessive mortality, which is supportive of guidelines which advocate for conservative management of ESA dosing regimen in hemodialysis patients.

scholarly journals Prednisone, Lupus Activity, and Permanent Organ Damage

2009 ◽  
Vol 36 (3) ◽  
pp. 560-564 ◽  
Author(s):  
MAE THAMER ◽  
MIGUEL A. HERNÁN ◽  
YI ZHANG ◽  
DENNIS COTTER ◽  
MICHELLE PETRI
Keyword(s):  
Lupus Erythematosus ◽  
Structural Model ◽  
Cox Regression ◽  
Laboratory Data ◽  
Organ Damage ◽  
Clinical Activity ◽  
Average Dose ◽  
Marginal Structural Model ◽  
Loss To Follow Up ◽  
Increased Risk

Objective.To estimate the effect of corticosteroids (prednisone dose) on permanent organ damage among persons with systemic lupus erythematosus (SLE).Methods.We identified 525 patients with incident SLE in the Hopkins Lupus Cohort. At each visit, clinical activity indices, laboratory data, and treatment were recorded. The study population was followed from the month after the first visit until June 29, 2006, or attainment of irreversible organ damage, death, loss to follow-up, or receipt of pulse methylprednisolone therapy. We estimated the effect of cumulative average dose of prednisone on organ damage using a marginal structural model to adjust for time-dependent confounding by indication due to SLE disease activity.Results.Compared with non-prednisone use, the hazard ratio of organ damage for prednisone was 1.16 (95% CI 0.54, 2.50) for cumulative average doses > 0–180 mg/month, 1.50 (95% CI 0.58, 3.88) for > 180–360 mg/month, 1.64 (95% CI 0.58, 4.69) for > 360–540 mg/month, and 2.51 (95% CI 0.87, 7.27) for > 540 mg/month. In contrast, standard Cox regression models estimated higher hazard ratios at all dose levels.Conclusion.Our results suggest that low doses of prednisone do not result in a substantially increased risk of irreversible organ damage.

Sign in / Sign up

Export Citation Format

Share Document