Faculty Bibliography

The consistency of doubly robust estimators relies on the consistent estimation of at least one of two nuisance regression parameters. In moderate-to-large dimensions, the use of flexible data-adaptive regression estimators may aid in achieving this consistency. However, n^1/2 -consistency of doubly robust estimators is not guaranteed if one of the nuisance estimators is inconsistent. In this paper, we present a doubly robust estimator for survival analysis with the novel property that it converges to a Gaussian variable at an n^1/2 -rate for a large class of data-adaptive estimators of the nuisance parameters, under the only assumption that at least one of them is consistently estimated at an n^1/4 -rate. This result is achieved through the adaptation of recent ideas in semiparametric inference, which amount to (i) Gaussianizing (ie, making asymptotically linear) a drift term that arises in the asymptotic analysis of the doubly robust estimator and (ii) using cross-fitting to avoid entropy conditions on the nuisance estimators. We present the formula of the asymptotic variance of the estimator, which allows for the computation of doubly robust confidence intervals and p values. We illustrate the finite-sample properties of the estimator in simulation studies and demonstrate its use in a phase III clinical trial for estimating the effect of a novel therapy for the treatment of human epidermal growth factor receptor 2 (HER2)-positive breast cancer.

OBJECTIVES:To better understand indications for β-blocker (BB) prescriptions among older adults hospitalized with heart failure with preserved ejection fraction (HFpEF). DESIGN/SETTING:Retrospective observational study of hospitalizations derived from the geographically diverse Reasons for Geographic and Racial Differences in Stroke cohort. PARTICIPANTS:We examined Medicare beneficiaries aged 65 years or older with an expert-adjudicated hospitalization for HFpEF (left ventricular ejection fraction = 50% or greater). MEASUREMENTS:Discharge medications and indications for BBs were abstracted from medical records. RESULTS:Of 306 hospitalizations for HFpEF, BBs were prescribed at discharge in 68%. Among hospitalizations resulting in BB prescriptions, 60% had a compelling indication for BB-44% had arrhythmias, and 29% had myocardial infarction (MI) history. Among the 40% with neither indication, 57% had coronary artery disease (CAD) without MI and 38% had hypertension alone (without arrhythmia, MI, or CAD), both clinical scenarios with little supportive evidence of benefit of BBs. Among hospitalizations resulting in BB prescription at discharge, 69% had geriatric conditions (functional limitation, cognitive impairment, hypoalbuminemia, or history of falls). There were no significant differences in the prevalence of geriatric conditions between hospitalizations of individuals with compelling indications for BBs and hospitalizations of individuals with noncompelling indications. CONCLUSIONS:BBs are commonly prescribed following a hospitalization for HFpEF, even in the absence of compelling indications. This occurs even for hospitalizations of individuals with geriatric conditions, a subpopulation who may be at elevated risk for experiencing harm from BBs.

OBJECTIVE:We assessed whether providing detailed clinical information alongside computed tomography (CT) images improves their interpretation for acute stroke. METHODS:Using the prospective Cornell AcutE Stroke Academic Registry, we randomly selected 100 patients who underwent noncontrast head CT within 6 hours of transient ischemic attack or minor acute ischemic stroke and underwent magnetic resonance imaging (MRI) within 6 hours of the CT. Three radiologist investigators evaluated each of the 100 CT studies twice, once with and once without accompanying information on medical history, signs, and symptoms. In random sequence, each study was interpreted in one condition (ie, with or without detailed accompanying information) and then after a 4-week washout period, in the opposite condition. Using MRI diffusion-weighted imaging (DWI) as the reference standard, we classified CT interpretations as correct (true positives or negatives) or incorrect (false positives or negatives). We used logistic regression with sandwich estimators to compare the proportion of correct interpretations. RESULTS:In patients with DWI-defined infarcts, acute ischemia was called on 20% of CTs with detailed history and 18% without history. In patients without infarcts, the absence of ischemia was called on 77% of CTs with history and 77% without history. The proportion of correct interpretations of CTs accompanied by detailed clinical history (49%) did not differ significantly from those without history (47%; odds ratio: 1.1; 95% confidence interval: 0.8-1.4). CONCLUSIONS:Reported findings on head CT for evaluation of suspected acute ischemic stroke were similar regardless of whether detailed clinical history was provided.

We present a new estimator of the restricted mean survival time in randomized trials where there is right censoring that may depend on treatment and baseline variables. The proposed estimator leverages prognostic baseline variables to obtain equal or better asymptotic precision compared to traditional estimators. Under regularity conditions and random censoring within strata of treatment and baseline variables, the proposed estimator has the following features: (i) it is interpretable under violations of the proportional hazards assumption; (ii) it is consistent and at least as precise as the Kaplan-Meier and inverse probability weighted estimators, under identifiability conditions; (iii) it remains consistent under violations of independent censoring (unlike the Kaplan-Meier estimator) when either the censoring or survival distributions, conditional on covariates, are estimated consistently; and (iv) it achieves the nonparametric efficiency bound when both of these distributions are consistently estimated. We illustrate the performance of our method using simulations based on resampling data from a completed, phase 3 randomized clinical trial of a new surgical treatment for stroke; the proposed estimator achieves a 12% gain in relative efficiency compared to the Kaplan-Meier estimator. The proposed estimator has potential advantages over existing approaches for randomized trials with time-to-event outcomes, since existing methods either rely on model assumptions that are untenable in many applications, or lack some of the efficiency and consistency properties (i)-(iv). We focus on estimation of the restricted mean survival time, but our methods may be adapted to estimate any treatment effect measure defined as a smooth contrast between the survival curves for each study arm. We provide R code to implement the estimator.

BACKGROUND:In 2013, investigators from A Randomized Trial of Unruptured Brain Arteriovenous Malformations (AVM; ARUBA) reported that interventions to obliterate unruptured AVMs caused more morbidity and mortality than medical management. OBJECTIVE:We sought to determine whether interventions for unruptured AVM decreased after publication of ARUBA results. METHODS:We used the Nationwide Readmissions Database to assess trends in interventional AVM management in patients ≥18 years of age from 2010 through 2015. Unruptured brain AVMs were identified using International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) code 747.81 and excluding any patient with a diagnosis of intracranial hemorrhage. Our primary outcome was interventional AVM treatment, identified using ICD-9-CM procedure codes for surgical resection, endovascular therapy, and stereotactic radiosurgery. Join-point regression was used to assess trends in the incidence of interventional AVM management among adults from 2010 through 2015. RESULTS:There was no significant U.S. population level change in unruptured brain AVM intervention rates before versus after ARUBA (p = 0.59), with the incidence of AVM intervention ranging from 8.0 to 9.2 per 10 million U.S. residents before the trial publication to 7.7-8.3 per 10 million afterwards. CONCLUSIONS:In a nationally representative sample, we found no change in rates of interventional unruptured AVM management after publication of the ARUBA trial results.