A marginal structural model for normal tissue complication probability.

The goal of radiation therapy for cancer is to deliver prescribed radiation dose to the tumor while minimizing dose to the surrounding healthy tissues. To evaluate treatment plans, the dose distribution to healthy organs is commonly summarized as dose-volume histograms (DVHs). Normal tissue complication probability (NTCP) modeling has centered around making patient-level risk predictions with features extracted from the DVHs, but few have considered adapting a causal framework to evaluate the safety of alternative treatment plans. We propose causal estimands for NTCP based on deterministic and stochastic interventions, as well as propose estimators based on marginal structural models that impose bivariable monotonicity between dose, volume, and toxicity risk. The properties of these estimators are studied through simulations, and their use is illustrated in the context of radiotherapy treatment of anal canal cancer patients.

Emulating a Target Trial Using Primary-Care Electronic Health Records: Sodium-Glucose Cotransporter 2 Inhibitor Medications and Hemoglobin A1c.

Substantial effort has been dedicated to conducting randomized controlled experiments to generate clinical evidence for diabetes treatment. Randomized controlled experiments are the gold standard for establishing cause and effect. However, due to their high cost and time commitment, large observational databases such as those comprised of electronic health record (EHR) data collected in routine primary care may provide an alternative source with which to address such causal objectives. We used a Canadian primary-care data repository housed at the University of Toronto (Toronto, Ontario, Canada) to emulate a randomized experiment. We estimated the effectiveness of sodium-glucose cotransporter 2 inhibitor (SGLT-2i) medications for patients with diabetes using hemoglobin A1c (HbA1c) as a primary outcome and marker for glycemic control from 2018 to 2021. We assumed an intention-to-treat analysis for prescribed treatment, with analyses based on the treatment assigned rather than the treatment eventually received. We defined the causal contrast of interest as the net change in HbA1c (percent) between the group receiving the standard of care versus the group receiving SGLT-2i medication. Using a counterfactual framework, marginal structural models demonstrated a reduction in mean HbA1c level with the initiation of SGLT-2i medications. These findings provided effect sizes similar to those from earlier clinical trials on assessing the effectiveness of SGLT-2i medications.

Estimation of marginal structural models under irregular visits and unmeasured confounder: calibrated inverse probability weights.

Clinical information collected in electronic health records (EHRs) is becoming an essential source to emulate randomized experiments. Since patients do not interact with the healthcare system at random, the longitudinal information in large observational databases must account for irregular visits. Moreover, we need to also account for subject-specific unmeasured confounders which may act as a common cause for treatment assignment mechanism (e.g. glucose-lowering medications) while also influencing the outcome (e.g. Hemoglobin A1c). We used the calibration of longitudinal weights to improve the finite sample properties and to account for subject-specific unmeasured confounders. A Monte Carlo simulation study is conducted to evaluate the performance of calibrated inverse probability estimators using time-dependent treatment assignment and irregular visits with subject-specific unmeasured confounders. The simulation study showed that the longitudinal weights with calibrated restrictions improved the finite sample bias when compared to the stabilized weights. The application of the calibrated weights is demonstrated using the exposure of glucose lowering medications and the longitudinal outcome of Hemoglobin A1c. Our results support the effectiveness of glucose lowering medications in reducing Hemoglobin A1c among type II diabetes patients with elevated glycemic index ([Formula: see text]) using stabilized and calibrated weights.

Textbook Outcome and Survival in Patients With Gastric Cancer: An Analysis of the Population Registry of Esophageal and Stomach Tumours in Ontario (PRESTO).

OBJECTIVE: To examine the association between Textbook Outcome (TO)-a new composite quality measurement-and long-term survival in gastric cancer surgery. BACKGROUND: Single-quality indicators do not sufficiently reflect the complex and multifaceted nature of perioperative care in patients with gastric adenocarcinoma. METHODS: All patients undergoing gastrectomy for nonmetastatic gastric adenocarcinoma registered in the Population Registry of Esophageal and Stomach Tumours of Ontario (PRESTO) between 2004 and 2015 were included. TO was defined according to negative margins; >15 lymph nodes sampled; no severe complications; no re-interventions; no unplanned ICU admission; length of stay ≤21 days; no 30-day readmission; and no 30-day mortality. Three-year survival was estimated using the Kaplan-Meier method. A marginal multivariable Cox proportional-hazards model was used to estimate the association between achieving TO metrics and long-term survival. E-value methodology was used to assess for risk of residual confounding. RESULTS: Of the 1836 patients included in this study, 402 (22%) achieved all TO metrics. TO patients had a higher 3-year survival rate compared to non-TO patients (75% vs 55%, log-rank P < 0.001). After adjustments for covariates and clustering within hospitals, TO was associated with a 41% reduction in mortality (adjusted hazards ratio 0.59, 95% confidence interval 0.48, 0.72, P < 0.001). These results were robust to potential residual confounding. CONCLUSIONS: Achieving TO is strongly associated with improved long-term survival in gastric cancer patients and merits further focus in surgical quality improvement efforts.

Influence of previous experience with and beliefs regarding anal cancer screening on willingness to be screened among men living with HIV.

BACKGROUND: Implementation of anal cancer screening requires the procedure to be acceptable to the target population. Our objective was to assess the beliefs of men living with HIV regarding anal cancer screening and identify factors associated with their willingness to participate in screening. METHODS: We developed a cross-sectional questionnaire using the Theory of Planned Behavior to examine beliefs regarding prevention of human papillomavirus (HPV)-related diseases, administered to men living with HIV in 2016-2017 in a multi-site HIV clinical cohort. Correspondence analysis was used to examine the interrelationships between men's beliefs and willingness to undergo anal cancer screening. We used multivariable proportional odds models to identify factors associated with increasing willingness. Results were reported as adjusted odds ratios (aOR) with 95% confidence intervals (CI). RESULTS: Among 1677 male participants, the vast majority (90%) would be willing to undergo screening by "anal Pap test"; willingness clustered with positive beliefs (e.g. confident they can get screened; disagree that they will feel pain) in the correspondence analysis. Higher self-perceived risk for anal cancer and positive beliefs regarding screening were associated with higher willingness to be screened. Gay, bisexual and other men who have sex with men had higher willingness (aOR = 1.62; 95% CI: 1.15, 2.29) than heterosexual men. Racialized men reported lower willingness (aOR = 0.68; 95% CI: 0.54, 0.89) than white men. CONCLUSIONS: Men generally had positive beliefs and were willing to undergo screening, though there were differences by sexual orientation and racial identity. Tailored community-led initiatives could focus on men's understanding of their risk and expectations of anal cancer screening to facilitate participation.

Normalized Augmented Inverse Probability Weighting with Neural Network Predictions.

The estimation of average treatment effect (ATE) as a causal parameter is carried out in two steps, where in the first step, the treatment and outcome are modeled to incorporate the potential confounders, and in the second step, the predictions are inserted into the ATE estimators such as the augmented inverse probability weighting (AIPW) estimator. Due to the concerns regarding the non-linear or unknown relationships between confounders and the treatment and outcome, there has been interest in applying non-parametric methods such as machine learning (ML) algorithms instead. Some of the literature proposes to use two separate neural networks (NNs) where there is no regularization on the network's parameters except the stochastic gradient descent (SGD) in the NN's optimization. Our simulations indicate that the AIPW estimator suffers extensively if no regularization is utilized. We propose the normalization of AIPW (referred to as nAIPW) which can be helpful in some scenarios. nAIPW, provably, has the same properties as AIPW, that is, the double-robustness and orthogonality properties. Further, if the first-step algorithms converge fast enough, under regulatory conditions, nAIPW will be asymptotically normal. We also compare the performance of AIPW and nAIPW in terms of the bias and variance when small to moderate L1 regularization is imposed on the NNs.

Targeted L1-Regularization and Joint Modeling of Neural Networks for Causal Inference.

The calculation of the Augmented Inverse Probability Weighting (AIPW) estimator of the Average Treatment Effect (ATE) is carried out in two steps, where in the first step, the treatment and outcome are modeled, and in the second step, the predictions are inserted into the AIPW estimator. The model misspecification in the first step has led researchers to utilize Machine Learning algorithms instead of parametric algorithms. However, the existence of strong confounders and/or Instrumental Variables (IVs) can lead the complex ML algorithms to provide perfect predictions for the treatment model which can violate the positivity assumption and elevate the variance of AIPW estimators. Thus the complexity of ML algorithms must be controlled to avoid perfect predictions for the treatment model while still learning the relationship between the confounders and the treatment and outcome. We use two NN architectures with an L1-regularization on specific NN parameters and investigate how their certain hyperparameters should be tuned in the presence of confounders and IVs to achieve a low bias-variance tradeoff for ATE estimators such as AIPW estimator. Through simulation results, we will provide recommendations as to how NNs can be employed for ATE estimation.

Achieving the "trifecta" with open versus minimally invasive partial nephrectomy.

INTRODUCTION: The "trifecta" is a summary measure of outcome after partial nephrectomy (PN) that encompasses three parameters: negative surgical margin, ≤ 10% decrease in post-operative estimated glomerular filtration rate (eGFR) and absence of urological complications. We assessed trifecta rates in patients undergoing open (OPN), laparoscopic (LPN), and robotic PN (RPN) for a clinical T1 renal mass (≤ 7 cm). METHODS: Clinical and pathologic parameters were extracted from the prospectively maintained Canadian Kidney Cancer Information System for patients treated between January 2011 and October 2018. Comparisons between groups were made using Kruskal-Wallis test for continuous variables and Chi-squared independence test for categorical variables. Multivariable analysis was performed to identify predictors of each component of the trifecta and the trifecta itself. RESULTS: Of 1708 total patients, 746 underwent OPN, 678 LPN, and 284 RPN for a T1 renal mass. A 'trifecta' was achieved in 53% OPN, 52% LPN and 47% RPN (p = 0.194). On multivariable analysis, OPN and LPN were associated with less frequent post-operative decline in eGFR and more frequent trifecta when compared to RPN, but there was no difference between OPN and LPN. OPN also predicted a higher rate of negative margins compared to RPN but not LPN. CONCLUSION: After correction for confounding variables, OPN and LPN were more likely than RPN to achieve the trifecta, which appeared to be due primarily to loss of renal function. No difference was observed between OPN and LPN. Analyses were limited by the lack of nephrometry score.

Instrumental variable estimation of early treatment effect in randomized screening trials.

The primary analysis of randomized screening trials for cancer typically adheres to the intention-to-screen principle, measuring cancer-specific mortality reductions between screening and control arms. These mortality reductions result from a combination of the screening regimen, screening technology and the effect of the early, screening-induced, treatment. This motivates addressing these different aspects separately. Here we are interested in the causal effect of early versus delayed treatments on cancer mortality among the screening-detectable subgroup, which under certain assumptions is estimable from conventional randomized screening trial using instrumental variable type methods. To define the causal effect of interest, we formulate a simplified structural multi-state model for screening trials, based on a hypothetical intervention trial where screening detected individuals would be randomized into early versus delayed treatments. The cancer-specific mortality reductions after screening detection are quantified by a cause-specific hazard ratio. For this, we propose two estimators, based on an estimating equation and a likelihood expression. The methods extend existing instrumental variable methods for time-to-event and competing risks outcomes to time-dependent intermediate variables. Using the multi-state model as the basis of a data generating mechanism, we investigate the performance of the new estimators through simulation studies. In addition, we illustrate the proposed method in the context of CT screening for lung cancer using the US National Lung Screening Trial data.

Constructing inverse probability weights for institutional comparisons in healthcare.

In comparing quality of care between hospitals, disease-specific quality indicators measure structural, process, or outcome elements related to the care of a particular condition. Such comparisons can be framed in terms of causal contrasts, answering the question of whether a patient (or a population of patients on average) would receive different care if treated at the care level of a different hospital. Fair comparisons have to be adjusted for patient case-mix, which is equivalent to controlling for confounding by the patient-level factors, including demographic factors, comorbidities, and disease progression. The methodological choice for such comparisons is usually between direct and indirect standardization methods. In this article, we discuss the alternative of inverse probability weighting as a tool for standardization in hospital comparisons. This involves fitting multinomial logistic hospital assignment models and using these to construct the inverse probability weights. The challenge in the present context is the presence of large number of hospitals being compared, many of which have a small patient volume. We propose methods to include small categories in the weighted analysis, as well as metrics and visualizations for checking the positivity/overlap and covariate balance in constructing such weights. The methods are illustrated in a running example using linked administrative data on surgical treatment of kidney cancer patients in Ontario.

Low human papillomavirus (HPV) knowledge related to low risk perception among men living with HIV: Implications for HPV-associated disease prevention.

Human papillomavirus (HPV)-associated anal cancer is orders of magnitude higher among men living with HIV than the general male population. Our objective was to examine factors associated with HPV awareness and self-perceived risk for HPV-associated anal cancer among men living with HIV, which may influence uptake of cancer prevention strategies. A cross-sectional questionnaire on HPV was administered from 2016 to 2017 to 1677 men in a multisite, HIV clinical cohort in Ontario, Canada. We used logistic regression and proportional odds models to identify factors associated with being familiar with HPV and increasing self-perceived risk for anal cancer, respectively. We used correspondence analysis to examine associations of specific HPV-related knowledge with self-perceived risk. Only 52% were familiar with HPV, and 72% felt they had no or low risk for anal cancer. Familiarity with HPV was more common among men who have sex with men than heterosexual men (58% vs. 21%). Older men were less likely to be familiar with HPV (adjusted odds ratio [aOR] per 10 years = 0.77; 95% confidence interval [CI]: 0.69, 0.85). Familiarity with HPV was associated with increasing self-perceived risk (aOR = 2.39; 95% CI: 1.87, 3.04). After accounting for differences in HPV awareness and sexual orientation, racialized men had lower self-perceived risk (aOR = 0.68; 95% CI: 0.52, 0.88). In the correspondence analysis, risk-focused HPV-related knowledge (e.g., knowing smoking increases risk) was associated with highest risk perception. Efforts are needed to improve HPV-related health literacy in this population. Our findings suggest specific HPV-related knowledge may differentially influence self-perceived risk for anal cancer.

Gastrectomy case volume and textbook outcome: an analysis of the Population Registry of Esophageal and Stomach Tumours of Ontario (PRESTO).

OBJECTIVE: To determine the association between gastric cancer surgery case-volume and Textbook Outcome, a new composite quality measurement. BACKGROUND: Textbook Outcome included (a) negative resection margin, (b) greater than 15 lymph nodes sampled, (c) no severe complication, (d) no re-intervention, (e) no unplanned ICU admission, (f) length of stay of 21 days or less, (g) no 30-day readmission and (h) no 30-day mortality following surgery. METHODS: All patients undergoing gastrectomy for non-metastatic gastric adenocarcinoma registered in the Population Registry of Esophageal and Stomach Tumours of Ontario between 2004 and 2015 were included. We used multivariable generalized estimating equation (GEE) logistic regression modelling to estimate the association between gastrectomy volume (surgeon and hospital annual volumes) and Textbook Outcome. Volumes were considered as continuous variables and quintiles. RESULTS: Textbook Outcome was achieved in 378 of 1660 patients (22.8%). The quality metrics least often achieved were inadequate lymph node sampling and presence of severe complications, which occurred in 46.1% and 31.7% of patients, respectively. Accounting for covariates and clustering, neither surgeon volume nor hospital volume were significantly associated with Textbook Outcome. However, hospital volume was associated with adequate lymphadenectomy and fewer unplanned ICU admissions. CONCLUSIONS: Higher case volume can impact certain measures of quality of care but may not address all care structures necessary for ideal Textbook recovery. Future quality improvement strategies should consider using case-mix adjusted Textbook Outcome rates as a surgical quality metric.

A population-based study of fall-related traumatic brain injury identified in older adults in hospital emergency departments.

OBJECTIVE: The purpose of this study was to examine the population-based trends and factors associated with hospitalization of patients with traumatic brain injury (TBI) treated in the Emergency Department (ED) among those 65 years and older. The implications of these trends for neurosurgery and the broader society are discussed. METHOD: With a national, mandatory reporting system of ED visits, the authors used Poisson regression controlling for age and sex to analyze trends in fall-related TBI of those aged 65 years and older between 2002 and 2017. RESULTS: The overall rate of ED visits for TBI increased by 78%-from 689.51 per 100,000 (95% CI 676.5-702.8) to 1229 per 100,000 (95% CI 1215-1243) between 2002 and 2017. Females consistently experienced higher rates of fall-related TBI than did males. All age groups 65 years and older experienced significant increases in fall-related TBI rate over the study period; however, the highest rates occurred among the oldest individuals (90+ and 85-89 years). The hospital admission rate increased with age and Charlson Comorbidity Index. Males experienced both a higher admission rate and a greater percentage change in admission rate than females. CONCLUSIONS: Rates of ED visits for fall-related TBI, hospitalization, and in-ED mortality in those aged 65 years and older are increasing for both sexes. The increasing hospital admission rate is related to more advanced comorbidities, male sex, and increasing age. These findings have significant implications for neurosurgical resources; they emphasize that health professionals should work proactively with patients, families, and caregivers to clarify goals of care, and they also outline the need for more high-level and, preferably, randomized evidence to support outcomes-based decisions. Additionally, the findings highlight the urgent need for improved population-based measures for prevention in not only this age demographic but in younger ones, and the need for changes in the planning of health service delivery and long-term care.

Hospital Quality Metrics for Radical Cystectomy: Disease Specific and Correlated to Mortality Outcomes.

PURPOSE: Measuring quality is a high priority for health care systems globally. Despite the high perioperative morbidity, mortality, expenditures and performance variation of radical cystectomy there is a paucity of validated bladder cancer quality metrics. We aimed to create a hospital quality scoring system for radical cystectomy which is disease specific and associated with patient centered outcomes. MATERIALS AND METHODS: We used the National Cancer Database to identify hospitals where radical cystectomy was performed from 2004 to 2014. Mixed effects models were constructed to assess variation in hospital performance across 7 quality indicators. Indirect standardization was used to case mix adjust hospital performance. We assessed associations between quality indicators as well as the novel BC-QS (Bladder Cancer Quality Score) composite hospital quality metric with 30-day, 90-day and overall mortality using logistic and Cox regression, respectively. RESULTS: At 1,200 facilities radical cystectomy was performed in a total of 48,341 patients from 2004 to 2014. Mixed effects models demonstrated significant between hospital variation across all quality indicators after case mix adjustment. The composite BC-QS metric was composed of the hospital positive margin rate, the lymph node dissection rate and the neoadjuvant chemotherapy rate. Better BC-QS performance was associated with lower 30-day and 90-day mortality (adjusted OR 0.78, 95% CI 0.64-0.96, and OR 0.84, 95% CI 0.72-0.97, respectively) and overall mortality (HR 0.86, 95% CI 0.81-0.92). Hospitals with a higher BC-QS had higher volume and more were affiliated with an academic institution than hospitals with a lower BC-QS (p <0.0001). CONCLUSIONS: The BC-QS captures variations in the hospital performance of radical cystectomy and it shows an association of higher quality with lower patient mortality. Our validation of this quality metric provides support for its potential use by policy makers and payers in efforts to measure hospital quality for high cost surgeries.

Causal Mediation Analysis for Standardized Mortality Ratios.

Indirectly standardized mortality ratios (SMR) are often used to compare patient outcomes between health care providers as indicators of quality of care. Observed differences in the outcomes raise the question of whether these could be causally attributable to earlier processes or outcomes in the pathway of care that the patients received. Such pathways can be naturally addressed in a causal mediation analysis framework. Adopting causal mediation models allows the total provider effect on outcome to be decomposed into direct and indirect (mediated) effects. This in turn enables quantification of the improvement in patient outcomes due to a hypothetical intervention on the mediator. We formulate the effect decomposition for the indirectly standardized SMR when comparing to a health care system-wide average performance, propose novel model-based and semiparametric estimators for the decomposition, study the properties of these through simulations, and demonstrate their use through application to Ontario kidney cancer data.

Circulating Levels of Interleukin 1-Receptor Antagonist and Risk of Cardiovascular Disease: Meta-Analysis of Six Population-Based Cohorts.

OBJECTIVE: Interleukin (IL)-1ß represents a key cytokine in the development of cardiovascular disease (CVD). IL-1ß is counter-regulated by IL-1 receptor antagonist (IL-1RA), an endogenous inhibitor. This study aimed to identify population-based studies on circulating IL-1RA and incident CVD in a systematic review, estimate the association between IL-1RA and incident CVD in a meta-analysis, and to test whether the association between IL-1RA and incident CVD is explained by other inflammation-related biomarkers in the MONICA/KORA Augsburg case-cohort study (Multinational Monitoring of Trends and Determinants in Cardiovascular Disease/Cooperative Health Research in the Region of Augsburg). APPROACH AND RESULTS: We performed a systematic literature search and identified 5 cohort studies on IL-1RA and incident CVD in addition to the MONICA/KORA Augsburg case-cohort study for a meta-analysis based on a total of 1855 CVD cases and 18 745 noncases with follow-up times between 5 and 16 years. The pooled standardized hazard ratio (95% confidence interval) for incident CVD was 1.11 (1.06-1.17) after adjustment for age, sex, anthropometric, metabolic, and lifestyle factors (P<0.0001). There was no heterogeneity in effect sizes (I2=0%; P=0.88). More detailed analyses in the MONICA/KORA study showed that the excess risk for CVD was attenuated by ≥10% after additional separate adjustment for serum levels of high-sensitivity C-reactive protein, IL-6, myeloperoxidase, soluble E-selectin, or soluble intercellular adhesion molecule-1. CONCLUSIONS: Serum IL-1RA levels were positively associated with risk of CVD after adjustment for multiple confounders in a meta-analysis of 6 population-based cohorts. This association may at least partially reflect a response to triggers inducing subclinical inflammation, oxidative stress, and endothelial activation.

Association of Repeatedly Measured High-Sensitivity-Assayed Troponin I with Cardiovascular Disease Events in a General Population from the MORGAM/BiomarCaRE Study.

BACKGROUND: High-sensitivity troponin I (hs-cTnI) concentrations reflect myocardial stress. The role of hs-cTnI in predicting long-term changes in the risk of cardiovascular disease (CVD) in general populations is not clearly defined. METHODS: We investigated whether the change in 3 repeated measures of hs-cTnI collected 5 years apart in a prospective Danish study (3875 participants, initially aged 30-60 years, 51% female, disease free at baseline) improves 10-year prediction of incident CVD compared to using a single most recent hs-cTnI measurement. The change process was modelled using a joint (longitudinal and survival) model and compared to a Cox model using a single hs-cTnI measure adjusted for classic CVD risk factors, and evaluated using discrimination statistics. RESULTS: Median hs-cTnI concentrations changed from 2.6 ng/L to 3.4 ng/L over 10 years. The change in hs-cTnI predicts 10-year risk of CVD (581 events); the joint model gave a hazard ratio of 1.31 per interquartile difference in hs-cTnI (95% CI 1.15-1.48) after adjustment for CVD risk factors. However, the joint model performed only marginally better (c-index improvement 0.0041, P = 0.03) than using a single hs-cTnI measure (c-index improvement 0.0052, P = 0.04) for prediction of CVD, compared to a model incorporating CVD risk factors without hs-cTnI (c-index 0.744). CONCLUSIONS: The change in hs-cTnI in 5-year intervals better predicts risk of CVD in the general population, but the most recent measure of hs-cTnI, (at 10 years) is as effective in predicting CVD risk. This simplifies the use of hs-cTnI as a prognostic marker for primary prevention of CVD in the general population.

Safety, reliability and accuracy of small renal tumour biopsies: results from a multi-institution registry.

OBJECTIVE: To validate, in a multi-institution review, the safety, accuracy and reliability of renal tumour biopsy (RTB) and its role in decreasing unnecessary treatment. MATERIALS AND METHODS: We conducted a multi-institution retrospective study of patients who underwent RTB to characterize a small renal mass (SRM) between 2011 and May 2015. Patients were identified using the prospectively maintained Canadian Kidney Cancer information system. Diagnostic and concordance rates were presented using proportions, whereas factors associated with a diagnostic RTB were identified using a logistic regression model. RESULTS: Of the 373 biopsied SRMs, the initial biopsy was diagnostic in 87% of cases. Of the 47 non-diagnostic biopsies, 15 had a repeat biopsy of which, 80% were diagnostic. When both were combined, therefore, a diagnosis was obtained in 91% of SRMs. Of these, 18% were benign. Size was the only factor found to be associated with achieving a diagnostic biopsy. RTB histology and nuclear grade (high or low) were found to be highly concordant with surgical pathology (86 and 81%, respectively). Of the discordant tumours (n = 16), all were upgraded from low to high grade on surgical pathology. Adverse events were rare (<1% of cases). CONCLUSION: The present multi-institution study confirms that RTB of SRMs is safe, accurate and reliable across institutions, while decreasing unnecessary treatment. Given our findings, RTBs may be a helpful tool with which to triage SRMs and guide appropriate management.

A flexible parametric approach for estimating continuous-time inverse probability of treatment and censoring weights.

Marginal structural Cox models are used for quantifying marginal treatment effects on outcome event hazard function. Such models are estimated using inverse probability of treatment and censoring (IPTC) weighting, which properly accounts for the impact of time-dependent confounders, avoiding conditioning on factors on the causal pathway. To estimate the IPTC weights, the treatment assignment mechanism is conventionally modeled in discrete time. While this is natural in situations where treatment information is recorded at scheduled follow-up visits, in other contexts, the events specifying the treatment history can be modeled in continuous time using the tools of event history analysis. This is particularly the case for treatment procedures, such as surgeries. In this paper, we propose a novel approach for flexible parametric estimation of continuous-time IPTC weights and illustrate it in assessing the relationship between metastasectomy and mortality in metastatic renal cell carcinoma patients. Copyright © 2016 John Wiley & Sons, Ltd.

A case-base sampling method for estimating recurrent event intensities.

Case-base sampling provides an alternative to risk set sampling based methods to estimate hazard regression models, in particular when absolute hazards are also of interest in addition to hazard ratios. The case-base sampling approach results in a likelihood expression of the logistic regression form, but instead of categorized time, such an expression is obtained through sampling of a discrete set of person-time coordinates from all follow-up data. In this paper, in the context of a time-dependent exposure such as vaccination, and a potentially recurrent adverse event outcome, we show that the resulting partial likelihood for the outcome event intensity has the asymptotic properties of a likelihood. We contrast this approach to self-matched case-base sampling, which involves only within-individual comparisons. The efficiency of the case-base methods is compared to that of standard methods through simulations, suggesting that the information loss due to sampling is minimal.