Bradley Efron’s research while affiliated with Stanford University and other places

Introduction: The evolution of pre- versus postoperative risk factors remains unknown in the development of persistent postoperative pain and opioid use. We identified preoperative versus comprehensive perioperative models of delayed pain and opioid cessation after total joint arthroplasty including time-varying postoperative changes in emotional distress. We hypothesized that time-varying longitudinal measures of postoperative psychological distress, as well as pre- and postoperative use of opioids would be the most significant risk factors for both outcomes. Methods: A prospective cohort of 188 patients undergoing total hip or knee arthroplasty at Stanford Hospital completed baseline pain, opioid use, and emotional distress assessments. After surgery, a modified Brief Pain Inventory was assessed daily for 3 months, weekly thereafter up to 6 months, and monthly thereafter up to 1 year. Emotional distress and pain catastrophizing were assessed weekly to 6 months, then monthly thereafter. Stepwise multivariate time-varying Cox regression modeled preoperative variables alone, followed by all perioperative variables (before and after surgery) with time to postoperative opioid and pain cessation. Results: The median time to opioid and pain cessation was 54 and 152 days, respectively. Preoperative total daily oral morphine equivalent use (hazard ratio-HR 0.97; 95% confidence interval-CI 0.96-0.98) was significantly associated with delayed postoperative opioid cessation in the perioperative model. In contrast, time-varying postoperative factors: elevated PROMIS (Patient-Reported Outcomes Measurement Information System) depression scores (HR 0.92; 95% CI 0.87-0.98), and higher Pain Catastrophizing Scale scores (HR 0.85; 95% CI 0.75-0.97) were independently associated with delayed postoperative pain resolution in the perioperative model. Conclusions: These findings highlight preoperative opioid use as a key determinant of delayed postoperative opioid cessation, while postoperative elevations in depressive symptoms and pain catastrophizing are associated with persistent pain after total joint arthroplasty providing the rationale for continued risk stratification before and after surgery to identify patients at highest risk for these distinct outcomes. Interventions targeting these perioperative risk factors may prevent prolonged postoperative pain and opioid use.

It is now 62 years since the publication of James and Stein’s seminal article on the estimation of a multivariate normal mean vector. The paper made a spectacular first impression on the statistical community through its demonstration of inadmissability of the maximum likelihood estimator. It continues to be influential, but not for the initial reasons. Empirical Bayes shrinkage estimation, now a major topic, found its early justification in the James–Stein formula. Less obvious downstream topics include Tweedie’s formula and Benjamini and Hochberg’s false discovery rate algorithm. This is a short and mainly non-technical review of the James–Stein rule and its effects on the machine learning era of statistical innovation.

Background: With advancements in allogeneic hematopoietic cell transplantation (alloHCT), the need for cytomegalovirus (CMV) surveillance persists. Methods: We present a retrospective analysis on the impact of CMV with preemptive therapy in 1,065 alloHCT patients with donor and/or recipient CMV seropositivity from 2009-2019. Results: 51% developed clinically significant CMV infection (CMV-CSI); 6.5% had CMV disease. In multivariate analysis stratified by serostatus and preparative regimen the use of ATG (HR 2.97, 95% CI 2.00 to 4.42, P < 0.001) was associated with development of CMV-CSI. Median length of stay for index hospitalization was longer in patients with CMV-CSI (27 d vs 25 d, respectively; P = .002), as were rates (32.9% vs 17.7%; P < .001) and duration (9 d vs 6 d; P < .001) of rehospitalization, and median total inpatient days (28 d vs 26 d; P < .001). Patients with CMV-CSI had higher rates of neutropenia (47% vs 20%; P < .001) and transfusion support (PRBC, median 5 vs 3; P < .001; platelets, median 3 vs 3; P < .001). Conclusion: Preemptive therapy does not negate the impact of CMV-CSI on peri-engraftment toxicity and healthcare utilization. This cohort represents a large single center study on the impact of CMV in the pre-letermovir era and serves as a real-world comparator for assessing the impact of future prophylaxis. This article is protected by copyright. All rights reserved.

This article was prepared for the Special Issue on Resampling methods for statistical inference of the 2020s. Modern algorithms such as random forests and deep learning are automatic machines for producing prediction rules from training data. Resampling plans have been the key technology for evaluating a rule’s prediction accuracy. After a careful description of the measurement of prediction error the article discusses the advantages and disadvantages of the principal methods: cross-validation, the nonparametric bootstrap, covariance penalties (Mallows’ Cp and the Akaike Information Criterion), and conformal inference. The emphasis is on a broad overview of a large subject, featuring examples, simulations, and a minimum of technical detail.

The article The Bootstrap Method for Assessing Statistical Accuracy, written by Bradley Efron and Robert Tibshirani, was originally published Online First without Open Access.

The scientific needs and computational limitations of the twentieth century fashioned classical statistical methodology. Both the needs and limitations have changed in the twenty‐first, and so has the methodology. Large‐scale prediction algorithms—neural nets, deep learning, boosting, support vector machines, random forests—have achieved star status in the popular press. They are recognizable as heirs to the regression tradition, but ones carried out at enormous scale and on titanic datasets. How do these algorithms compare with standard regression techniques such as ordinary least squares or logistic regression? Several key discrepancies will be examined, centering on the differences between prediction and estimation or prediction and attribution (significance testing). Most of the discussion is carried out through small numerical examples.

Empirical Bayes inference assumes an unknown prior density g(θ) has yielded (unobservables) Θ1, Θ2, ..., ΘN, and each Θi produces an independent observation Xi from pi (Xi | Θi). The marginal density fi (Xi) is a convolution of the prior g and pi. The Bayes deconvolution problem is one of recovering g from the data. Although estimation of g - so called g-modeling - is difficult, the results are more encouraging if the prior g is restricted to lie within a parametric family of distributions. We present a deconvolution approach where g is restricted to be in a parametric exponential family, along with an R package deconvolveR designed for the purpose.

Despite effective therapies, CMV continues to have a significant impact on morbidity and mortality in hematopoietic cell transplant recipients. At particular risk are recipients of alternative grafts such as umbilical cord blood (UCB), haploidentical transplants (haplo), or patients conditioned with T-cell depleting regimens such as anti-thymocyte globulin (ATG). With the approval of letermovir, its impact on high risk patients is of particular interest. To evaluate the impact of letermovir prophylaxis at our center we performed a retrospective analysis of 114 high risk patients who received letermovir as prophylaxis (LET PPX) between January 2018 through December 2019, including 30 UCB and 22 haplo recipients, compared to 637 historical controls with comparable risk between January 2013 and December 2019. By D+100, letermovir prophylaxis significantly decreased the incidence of both CMV DNAemia compared to controls (45.37% vs 74.1%; P <.001) and clinically significant CMV infection (12.04% vs 48.82%; P <.001). The impact of LET PPX was even more profound on the incidence of clinically significant CMV infection (CSI) defined as the administration of antiviral therapy either as preemptive therapy for CMV DNAemia or treatment for CMV disease. CSI was significantly lower in haplo recipients on LET PPX compared to controls (13.64% vs 73.33%; P= .02) and UCB recipients on LET PPX compared to controls (3.45% vs 37.5%; P <.001). No patients on LET primary PPX developed CMV disease in any treatment group by D+100, compared to controls (0% vs 5.34%, respectively; P= .006). Patients on LET PPX had fewer hospitalizations involving initiation of anti-CMV therapy compared to controls (0.93% vs 15.23%, respectively). Our analysis of the largest cohort of patients at high risk for CMV reactivation published to date demonstrates that letermovir prophylaxis significantly reduces the number of patients who receive CMV-active antiviral therapy for either DNAemia or disease due to CMV.