Rebecca A Betensky
Rebecca Betensky
Scroll
Chair of the Department of Biostatistics
Professor of Biostatistics
-
Professional overview
-
Prior to NYU, Dr. Betensky was Professor of Biostatistics at the Harvard T.H. Chan School of Public Health. She was director of the Harvard Catalyst (Clinical and Translational Science Award) Biostatistics Program; director of the Data and Statistics Core for the Massachusetts Alzheimer’s Disease Research Center; and director of the Biostatistics Neurology Core at Massachusetts General Hospital. Previously, she was the Biostatistics Program Leader for the Dana-Farber/Harvard Cancer Center.
Dr. Betensky’s research focuses on methods for the analysis of censored and truncated outcomes and covariates, which frequently arise from the subsampling of cohort studies. She has a long-time interest in clinical trials, and has written on the evaluation of biomarkers and the use and interpretation of p-values. She has collaborated extensively in studies in neurologic diseases, and serves as statistical editor for Annals of Neurology.
Dr. Betensky was awarded, and directed for 15 years, an NIH T32 training program in neurostatistics and neuroepidemiology for pre- and post-doctoral students in biostatistics and epidemiology and for clinician-scientists. She previously directed Harvard’s Biostatistics programs to promote and support diversity at all levels in the field of quantitative public health. She was also a member of the BMRD Study Section for review of NIH statistical methodology grants; on committees for the Institute of Medicine; and a co-chair of the technical advisory committee for the scientific registry of transplant recipients.
Dr. Betensky an elected Fellow of the American Statistical Association and of the International Statistical Institute, and is a past recipient of the Spiegelman Award from the American Public Health Association. She currently serves as a member of the Board of Scientific Counselors for Clinical Science and Epidemiology at the National Cancer Institute.
-
Education
-
AB, Mathematics, Harvard University, Cambridge, MAPhD, Statistics, Stanford University, Stanford, CA
-
Areas of research and study
-
BiologyBiostatisticsNeuroepidemiologyNeurologyNeurostatisticsTranslational science
-
Publications
Publications
Causal inference in medical records and complementary systems pharmacology for metformin drug repurposing towards dementia
AbstractCharpignon, M. L., Vakulenko-Lagun, B., Zheng, B., Magdamo, C., Su, B., Evans, K., Rodriguez, S., Sokolov, A., Boswell, S., Sheu, Y. H., Somai, M., Middleton, L., Hyman, B. T., Betensky, R., Finkelstein, S. N., Welsch, R. E., Tzoulaki, I., Blacker, D., Das, S., & Albers, M. W. (n.d.).Publication year
2022Journal title
Nature communicationsVolume
13Issue
1AbstractMetformin, a diabetes drug with anti-aging cellular responses, has complex actions that may alter dementia onset. Mixed results are emerging from prior observational studies. To address this complexity, we deploy a causal inference approach accounting for the competing risk of death in emulated clinical trials using two distinct electronic health record systems. In intention-to-treat analyses, metformin use associates with lower hazard of all-cause mortality and lower cause-specific hazard of dementia onset, after accounting for prolonged survival, relative to sulfonylureas. In parallel systems pharmacology studies, the expression of two AD-related proteins, APOE and SPP1, was suppressed by pharmacologic concentrations of metformin in differentiated human neural cells, relative to a sulfonylurea. Together, our findings suggest that metformin might reduce the risk of dementia in diabetes patients through mechanisms beyond glycemic control, and that SPP1 is a candidate biomarker for metformin’s action in the brain.Cell-surface MuSK self-association : A crucial role for the putative signal sequence
AbstractBianchetta, M. J., Betensky, R., & Cohen, J. B. (n.d.).Publication year
2005Journal title
BiochemistryVolume
44Issue
49Page(s)
16229-16238AbstractThe receptor tyrosine kinase MuSK plays a crucial role-both as a signaling molecule and structurally-in the process of clustering nicotinic acetylcholine receptors at the neuromuscular junction. Immunofluorescence microscopy of transiently transfected fibroblasts has been used to visualize the cell-surface distribution of MuSK, which is found in discrete, punctate clusters. This distribution does not result from targeting of MuSK to identified plasma membrane subdomains, and MuSK's association with itself is specific, as MuSK clusters at the cell surface are segregated from clusters of other cotransfected receptor tyrosine kinases. A mutational analysis, using coexpressed pairs of MuSK mutants and chimeras, demonstrates that the putative signal peptide is both necessary and sufficient for association with MuSK. Removal of the intracellular domain or most of the extracellular domain, or replacement of the transmembrane domain, does not abolish MuSK self-association. The N-terminus of the MuSK protein, however, is sufficient to recruit another receptor tyrosine kinase to MuSK clusters. Quantitation and statistical analysis of the amount of colocalization between coexpressed MuSK mutants and chimeras confirm these results.Characteristics of Paid Malpractice Claims among Resident Physicians from 2001 to 2015 in the United States
AbstractGlover, M., McGee, G. W., Wilkinson, D. S., Singh, H., Bolick, A., Betensky, R., Harvey, H. B., Weinstein, D., & Schaffer, A. (n.d.).Publication year
2020Journal title
Academic MedicineVolume
95Issue
2Page(s)
255-262AbstractPurpose Limited information exists about medical malpractice claims against physicians-in-training. Data on residents' involvement in malpractice actions may inform perceptions about medicolegal liability and influence clinical decision-making at a formative stage. This study aimed to characterize rates and payment amounts of paid malpractice claims on behalf of resident physicians in the United States. Method Using data from the National Practitioner Data Bank, 1,248 paid malpractice claims against resident physicians (interns, residents, and fellows) from 2001 to 2015, representing 1,632,471 residents-years, were analyzed. Temporal trends in overall and specialty-specific paid claim rates, payment amounts, catastrophic (> $1 million) and small (< $100,000) payments, and other claim characteristics were assessed. Payment amounts were compared with attending physicians during the same time period. Results The overall paid malpractice claim rate was 0.76 per 1,000 resident-years from 2001 to 2015. Among 1,194 unique residents with paid claims, 95.7% had exactly 1 claim, while 4.3% had 2-4 claims during training. Specialty-specific paid claim rates ranged from 0.12 per 1,000 resident-years (pathology) to 2.96 (obstetrics and gynecology). Overall paid claim rates decreased by 52% from 2001-2005 to 2011-2015 (95% confidence interval [CI]: 0.45, 0.59). Median inflation-adjusted payment amount was $199,024 (2015 dollars), not significantly different from payments made on behalf of attending physicians during the same period. Proportions of catastrophic (11.2%) and small (33.1%) claims did not significantly change over the study period. Conclusions From 2001 to 2015, paid malpractice claim rates on behalf of resident physicians decreased by 52%, while median payment amounts were stable. Resident paid claim rates were lower than attending physicians, while payment amounts were similar.Characterization of amyloid deposition in the APPswe/PS1dE9 mouse model of Alzheimer disease
AbstractGarcia-Alloza, M., Robbins, E. M., Zhang-Nunes, S. X., Purcell, S. M., Betensky, R., Raju, S., Prada, C., Greenberg, S. M., Bacskai, B. J., & Frosch, M. P. (n.d.).Publication year
2006Journal title
Neurobiology of DiseaseVolume
24Issue
3Page(s)
516-524AbstractTransgenic mice carrying disease-linked forms of genes associated with Alzheimer disease often demonstrate deposition of the β-amyloid as senile plaques and cerebral amyloid angiopathy. We have characterized the natural history of β-amyloid deposition in APPswe/PS1dE9 mice, a particularly aggressive transgenic mouse model generated with mutant transgenes for APP (APPswe: KM594/5NL) and PS1 (dE9: deletion of exon 9). Ex vivo histochemistry showed Aβ deposition by 4 months with a progressive increase in plaque number up to 12 months and a similar increase of Aβ levels. In vivo multiphoton microscopy at weekly intervals showed increasing β-amyloid deposition as CAA and plaques. Although first appearing at an early age, CAA progressed at a significantly slower rate than in the Tg2576 mice. The consistent and early onset of β-amyloid accumulation in the APPswe/PS1dE9 model confirms its utility for studies of biochemical and pathological mechanisms underlying β-amyloid deposition, as well as exploring new therapeutic treatments.Clinical Manifestations
AbstractBriggs, A. Q., Betensky, R., Bernard, M. A., & Masurkar, A. V. (n.d.).Publication year
2024Journal title
Alzheimer's and DementiaVolume
20Page(s)
e091665AbstractBACKGROUND: Subjective cognitive decline (SCD) is a preclinical stage of Alzheimer's disease (AD), with correlations to cerebral amyloid and tau and accelerated cognitive decline. Studies have also revealed an association between sleep fragmentation and such AD biomarkers and cognitive decline, suggesting that cognitive function should be monitored in individuals experiencing excessive sleepiness. It is unclear if and how sleep dysfunction relates to SCD apart from AD biomarkers, as well as symptoms related to SCD and sleep dysfunction, such as depression. Furthermore, studies of SCD and sleep dysfunction have primarily focused on non-Hispanic Whites. Hence, a better understanding of how sleep impacts preclinical AD risk in SCD is needed in more diverse cohorts. We investigated the cross-sectional relationship between excessive sleepiness in cognitively normal older adults and SCD symptoms linked to cognitive domains in a cohort consisting of NHWs and Black/African Americans (B/AA). METHOD: This retrospective cross-sectional study included cognitively normal older participants (n = 147, age> 60 years, 61 B/AA, 86 NHW) evaluated at the NYU Alzheimer's Disease Research Center. Participants underwent Uniform Data Set 3.0 psychometric testing to confirm normal cognition, Epworth Sleepiness Scale (ESS), Geriatric Depression Scale (GDS), and the Cognitive Change Index (CCI) as a measure of SCD. The CCI was further divided into subscores of non-amnestic (i.e. executive) and amnestic domains (i.e. memory). Participants underwent amyloid (florbetaben) and tau (PI-2620 or MK-6240) PET MRI, with composite SUVR calculated by averaging key cortical regions. Multivariable linear regression models tested the association of ESS and covariates (GDS, amyloid SUVR, tau SUVR, race) to the three CCI outcome measures (total, non-amnestic, amnestic). RESULT: ESS (excessive sleepiness) was significantly positively associated with total CCI (p = 0.01042) as well as non-amnestic (0.00771) and amnestic CCI subgroups (0.02583), after full covariate adjustment, as was GDS. Race did not show an independent statistically significant association. CONCLUSION: These results suggest that excessive sleepiness, independent of depression and amyloid and tau burdens, may be a significant contributor to SCD severity. Furthermore, these contributions impact both amnestic and non-amnestic domains. Future studies are needed to investigate the longitudinal impact of excessive sleepiness on SCD outcomes.Clinical pertinence metric enables hypothesis-independent genome-phenome analysis for neurologic diagnosis
AbstractSegal, M. M., Abdellateef, M., El-Hattab, A. W., Hilbush, B. S., De La Vega, F. M., Tromp, G., Williams, M. S., Betensky, R., & Gleeson, J. (n.d.).Publication year
2015Journal title
Journal of Child NeurologyVolume
30Issue
7Page(s)
881-888AbstractWe describe an "integrated genome-phenome analysis" that combines both genomic sequence data and clinical information for genomic diagnosis. It is novel in that it uses robust diagnostic decision support and combines the clinical differential diagnosis and the genomic variants using a "pertinence" metric. This allows the analysis to be hypothesis-independent, not requiring assumptions about mode of inheritance, number of genes involved, or which clinical findings are most relevant. Using 20 genomic trios with neurologic disease, we find that pertinence scores averaging 99.9% identify the causative variant under conditions in which a genomic trio is analyzed and family-aware variant calling is done. The analysis takes seconds, and pertinence scores can be improved by clinicians adding more findings. The core conclusion is that automated genome-phenome analysis can be accurate, rapid, and efficient. We also conclude that an automated process offers a methodology for quality improvement of many components of genomic analysis.Clinical trials using HIV-1 RNA-based primary endpoints : Statistical analysis and potential biases
AbstractMarschner, I. C., Betensky, R., DeGruttola, V., Hammer, S. M., & Kuritzkes, D. R. (n.d.).Publication year
1999Journal title
Journal of Acquired Immune Deficiency Syndromes and Human RetrovirologyVolume
20Issue
3Page(s)
220-227AbstractClinical trial endpoints based on magnitude of reduction in HIV-1 RNA levels provide an important complement to endpoints based on percentage of patients achieving complete virologic suppression. However, interpretation of magnitude of reduction can he biased by measurement limitations of virologic assays, particularly lower and upper limits of quantification. Using data from two AIDS Clinical Trials Group (ACTG) studies, widely used crude methods of analyzing HIV-1 RNA reductions were compared with methods that take into account censoring of HIV-1 RNA measurements. Such methods include Kaplan- Meier and censored regression analyses. It was found that standard crude methods of analysis consistently underestimated treatment effects. In some cases, the bias induced by crude methods masked statistically significant differences between treatment arms. Although statistically significant, adjustment for baseline HIV-1 RNA levels had little effect on estimated treatment differences. Furthermore, convenient parametric analyses performed as well as more complex nonparametric analyses. It is concluded that conveniently implemented censored data analyses should be conducted in preference to widely used crude analyses of magnitude of HIV-1 RNA reduction. To obtain complete information about virologic response to antiretroviral therapy, such analyses of magnitude of virologic response should be used to complement analyses of the percentage of patients having complete virologic suppression.Cognitive resilience in clinical and preclinical Alzheimer’s disease : the Association of Amyloid and Tau Burden on cognitive performance
AbstractRentz, D. M., Mormino, E. C., Papp, K. V., Betensky, R., Sperling, R. A., & Johnson, K. A. (n.d.).Publication year
2017Journal title
Brain Imaging and BehaviorVolume
11Issue
2Page(s)
383-390AbstractWe explored the cross-sectional relationships between β-amyloid (Aβ) and inferior temporal tau deposition (IFT Tau) on cognitive performance and whether cognitive reserve (CR) modifies these associations. We studied 156 participants classified into groups of clinically normal (CN = 133), mild cognitive impairment (MCI = 17) and Alzheimer disease (AD = 6) dementia. AMNART IQ served as a proxy of CR and cognitive performance was assessed using the MMSE. In separate linear regression models predicting MMSE, we examined the interactions of CR x global Aβ and CR x IFT tau across all participants and within the CN group alone. In the whole sample, the interaction between CR and IFT tau was significant (p < 0.003), such that higher CR participants with elevated IFT tau had better MMSE scores compared with low CR participants with similar levels of IFT tau. The interaction between CR and Aβ status did not reach significance (p = 0.093). In CN only, no cross-sectional interactions among CR, Aβ, and IFT tau were observed on MMSE. These findings imply that CR may be protective against early AD processes and enable some individuals to remain cognitively stable despite elevated tau and Aβ burden.Combining gene expression profiles and clinical parameters for risk stratification in medulloblastomas
AbstractFernandez-Teijeiro, A., Betensky, R., Sturla, L. M., Kim, J. Y., Tamayo, P., & Pomeroy, S. L. (n.d.).Publication year
2004Journal title
Journal of Clinical OncologyVolume
22Issue
6Page(s)
994-998AbstractPurpose: Stratification of risk in patients with medulloblastoma remains a challenge. As clinical parameters have been proven insufficient for accurately defining disease risk, molecular markers have become the focus of interest. Outcome predictions on the basis of microarray gene expression profiles have been the most accurate to date. We ask in a multivariate model whether clinical parameters enhance survival predictions of gene expression profiles. Patients and Methods: In a cohort of 55 young patients (whose medulloblastoma samples have been analyzed previously for gene expression profile), associations between clinical and gene expression variables and survival were assessed using Cox proportional hazards models. Available clinical variables included age, stage (ie, the presence of disseminated disease at diagnosis), sex, histologic subtype, treatment, and status. Results: Univariate analysis demonstrated expression profiles to be the only significant clinical prognostic factor (P = .03). In multivariate analysis, gene expression profiles predicted outcome independent of other criteria. Clinical criteria did not significantly contribute additional information for outcome predictions, although an exploratory analysis noted a trend for decreased survival of patients with metastases at diagnosis but favorable gene expression profile. Conclusion: Gene expression profiling predicts medulloblastoma outcome independent of clinical variables. These results need to be validated in a larger prospective study.Comment on “Patient preference for cellulitis treatment : At-home care is preferred to hospital-based treatment”
AbstractShaw, K. S., Karagounis, T. K., Yin, L., Gibbon, G., Betensky, R., Lo Sicco, K. I., & Femia, A. N. (n.d.).Publication year
2021Journal title
Journal of the American Academy of DermatologyVolume
85Issue
3Page(s)
e157-e158Abstract~Commentary : Failure-rate functions for doubly-truncated random variables
AbstractBetensky, R., & Martin, E. C. (n.d.).Publication year
2003Journal title
IEEE Transactions on ReliabilityVolume
52Issue
1Page(s)
7-8AbstractNavarro and Ruiz [1] express the nonparametric maximum likelihood estimator (NPMLE) of the distribution of a failure-time random variable as a function of the NPMLE of generalized failure-rate functions. These generalized failure-rate functions are equal to the probability density functions of a doubly-truncated failure-time random variable at the endpoints of the truncating interval. Readers can infer from this paper that this simple estimator can be applied to a doubly-truncated sample of failure times. This commentary explains why that estimator does not apply to the general setting in which the observed failure times are doubly-truncated with subject-specific truncating intervals. A doubly-truncated sample of times to brain tumor progression illustrates the deviation of that estimator [1] from the NPMLE for these data. Definitions Quasar: unusually luminous objects found in remote areas of the universe; stellar object: nonquasar object; redshift: the shift in an object'S spectral lines toward the red end, indicating how fast the object is moving away from the observer.Comparison of clinical subgroup aCGH profiles through pseudolikelihood ratio tests
AbstractEngler, D., Shen, Y., Gusella, J., & Betensky, R. (n.d.).Publication year
2011Journal title
Statistical Applications in Genetics and Molecular BiologyVolume
10Issue
1AbstractArray-based Comparative Genomic Hybridization (aCGH) is a microarray-based technology that assists in identification of DNA sequence copy number changes across the genome. Examination of differences in instability phenotype, or pattern of copy number alterations, between cancer subtypes can aid in classification of cancers and lead to better understanding of the underlying cytogenic mechanism. Instability phenotypes are composed of a variety of copy number alteration features including height or magnitude of copy number alteration level, frequency of transition between copy number states such as gain and loss, and total number of altered clones or probes. That is, instability phenotype is multivariate in nature. Current methods of instability phenotype assessment, however, are limited to univariate measures and are therefore limited in both sensitivity and interpretability. In this paper, a novel method of instability assessment is presented that is based on the Engler et al. (2006) pseudolikelhood approach for aCGH data analysis. Through use of a pseudolikelihood ratio test (PLRT), more sensitive assessment of instability phenotype differences between cancer subtypes is possible. Evaluation of the PLRT method is conducted through analysis of a meningioma data set and through simulation studies. Results are shown to be more accurate and more easily interpretable than current measures of instability assessment.Comparison of comorbidities and adverse events in dermatology and rheumatology patients prescribed tofacitinib: A retrospective analysis
AbstractBetensky, R., Needle, C. D., Klein, E. J., Gjonaj, J., Nohria, A., Karim, M., Liu, L., Shah, J., Betensky, R. A., Garshick, M., Lo Sicco, K., & Karagounis, T. K. (n.d.).Publication year
2024Journal title
Journal of the American Academy of DermatologyVolume
90Issue
3Page(s)
659-662Abstract~Comparison of serum neurodegenerative biomarkers among hospitalized COVID-19 patients versus non-COVID subjects with normal cognition, mild cognitive impairment, or Alzheimer's dementia
AbstractBetensky, R., Frontera, J. A., Boutajangout, A., Masurkar, A. V., Betensky, R. A., Ge, Y., Vedvyas, A., Debure, L., Moreira, A., Lewis, A., Huang, J., Thawani, S., Balcer, L., Galetta, S., & Wisniewski, T. (n.d.).Publication year
2022Journal title
Alzheimer's & dementia : the journal of the Alzheimer's AssociationVolume
18Issue
5Page(s)
899-910AbstractNeurological complications among hospitalized COVID-19 patients may be associated with elevated neurodegenerative biomarkers.Comparison of urine output among patients treated with more intensive versus less intensive RRT : Results from the acute renal failure trial network study
AbstractMcCausland, F. R., Asafu-Adjei, J., Betensky, R., Palevsky, P. M., & Waikar, S. S. (n.d.).Publication year
2016Journal title
Clinical Journal of the American Society of NephrologyVolume
11Issue
8Page(s)
1335-1342AbstractBackground and objectives Intensive RRT may have adverse effects that account for the absence of benefit observed in randomized trials of more intensive versus less intensive RRT. We wished to determine the association of more intensive RRT with changes in urine output as a marker of worsening residual renal function in critically ill patients with severe AKI. Design, setting, participants, & measurements The Acute Renal Failure Trial Network Study (n=1124) was a multicenter trial that randomized critically ill patients requiring initiation of RRT tomore intensive (hemodialysis or sustained low-efficiency dialysis six times per week or continuous venovenous hemodiafiltration at 35 ml/kg per hour) versus less intensive (hemodialysis or sustained low-efficiency dialysis three times per week or continuous venovenous hemodiafiltration at 20 ml/kg per hour) RRT. Mixed linear regression models were fit to estimate the association of RRT intensity with change in daily urine output in survivors through day 7 (n=871); Cox regression models were fit to determine the association of RRT intensity with time to ≥50% decline in urine output in all patients through day 28. Results Mean age of participants was 60615 years old, 72% were men, and 30% were diabetic. In unadjusted models, among patients who survived ≥7 days, mean urine output was, on average, 31.7 ml/d higher (95% confidence interval, 8.2 to 55.2 ml/d) for the less intensive group compared with the more intensive group (P=0.01). More intensive RRT was associated with 29% greater unadjusted risk of decline in urine output of ≥50% (hazard ratio, 1.29; 95% confidence interval, 1.10 to 1.51). Conclusions More intensive versus less intensive RRT is associated with a greater reduction in urine output during the first 7 days of therapy and a greater risk of developing a decline in urine output of ≥50%in critically ill patients with severe AKI.Computationally simple accelerated failure time regression for interval censored data
AbstractBetensky, R., Rabinowitz, D., & Tsiatis, A. A. (n.d.).Publication year
2001Journal title
BiometrikaVolume
88Issue
3Page(s)
703-711AbstractAn approach is presented for fitting the accelerated failure time model to interval censored data that does not involve computing the nonparametric maximum likelihood estimate of the distribution function at the residuals. The approach involves estimating equations computed with the examination times from the same individual treated as if they had actually been obtained from different individuals. The dependence between different measurements obtained from the same individual is then accounted for in the calculation of the standard error of the regression coefficients. The approach is applicable to interval censored data in settings in which examinations continue to occur regardless of whether the failure time has occurred. Simulations are presented to assess the behaviour of the approach, and the methodology is illustrated through an application to data from an AIDS clinical trial.Computationally simple analysis of matched, outcome-based studies of ordinal disease states
AbstractBetensky, R., Szymonifka, J., Lee, E. Q., Nutt, C. L., & Batchelor, T. T. (n.d.).Publication year
2015Journal title
Statistics in MedicineVolume
34Issue
17Page(s)
2514-2527AbstractOutcome-based sampling is an efficient study design for rare conditions, such as glioblastoma. It is often used in conjunction with matching, for increased efficiency and to potentially avoid bias due to confounding. A study was conducted at the Massachusetts General Hospital that involved retrospective sampling of glioblastoma patients with respect to multiple-ordered disease states, as defined by three categories of overall survival time. To analyze such studies, we posit an adjacent categories logit model and exploit its allowance for prospective analysis of a retrospectively sampled study and its advantageous removal of set and level specific nuisance parameters through conditioning on sufficient statistics. This framework allows for any sampling design and is not limited to one level of disease within each set, such as in previous publications. We describe how this ordinal conditional model can be fit using standard conditional logistic regression procedures. We consider an alternative pseudo-likelihood approach that potentially offers robustness under partial model misspecification at the expense of slight loss of efficiency under correct model specification for small sample sizes. We apply our methods to the Massachusetts General Hospital glioblastoma study.Computationally simple estimation and improved efficiency for special cases of double truncation
AbstractAustin, M. D., Simon, D. K., & Betensky, R. (n.d.).Publication year
2014Journal title
Lifetime Data AnalysisVolume
20Issue
3Page(s)
335-354AbstractDoubly truncated survival data arise when event times are observed only if they occur within subject specific intervals of times. Existing iterative estimation procedures for doubly truncated data are computationally intensive (Turnbull 38:290-295, 1976; Efron and Petrosian 94:824-825, 1999; Shen 62:835-853, 2010a). These procedures assume that the event time is independent of the truncation times, in the sample space that conforms to their requisite ordering. This type of independence is referred to as quasi-independence. In this paper we identify and consider two special cases of quasi-independence: complete quasi-independence and complete truncation dependence. For the case of complete quasi-independence, we derive the nonparametric maximum likelihood estimator in closed-form. For the case of complete truncation dependence, we derive a closed-form nonparametric estimator that requires some external information, and a semi-parametric maximum likelihood estimator that achieves improved efficiency relative to the standard nonparametric maximum likelihood estimator, in the absence of external information. We demonstrate the consistency and potentially improved efficiency of the estimators in simulation studies, and illustrate their use in application to studies of AIDS incubation and Parkinson's disease age of onset.Concordance measures and time-dependent ROC methods
AbstractPantoja-Galicia, N., Okereke, O. I., Blacker, D., & Betensky, R. (n.d.).Publication year
2021Journal title
Biostatistics and EpidemiologyVolume
5Issue
2Page(s)
232-249AbstractThe receiver operating characteristic (ROC) curve displays sensitivity versus 1-specificity over a set of thresholds. The area under the ROC curve (AUC) is a global scalar summary of this curve. In the context of time-dependent ROC methods, we are interested in global scalar measures that summarize sequences of time-dependent AUCs over time. The concordance probability is a candidate for such purposes. The concordance probability can provide a global assessment of the discrimination ability of a test for an event that occurs at random times and may be right censored. If the test adequately differentiates between subjects who survive longer times and those who survive shorter times, this will assist clinical decisions. In this context, the concordance probability may support the assessment of precision medicine tools based on prognostic biomarkers models for overall survival. Definitions of time-dependent sensitivity and specificity are reviewed. Some connections between such definitions and concordance measures are also reviewed and we establish new connections via new measures of global concordance. We explore the relationship between such measures and their corresponding time-dependent AUC. To illustrate these concepts, an application in the context of Alzheimer's disease is presented.Conditional power calculations for early acceptance of H(O) embedded in sequential tests
AbstractBetensky, R. (n.d.).Publication year
1997Journal title
Statistics in MedicineVolume
16Issue
4Page(s)
465-477AbstractFor ethical and efficiency concerns one often wishes to design a clinical trial to stop early if there is a strong treatment effect or if there is strong evidence of no treatment effect. There is a large literature to address the design of sequential trials for detecting treatment differences. There has been less attention paid to the design of trials for detecting lack of a treatment difference and most of the designs proposed have been ad hoc modifications of the traditional designs. In the context of fixed sample tests, various authors have proposed basing the decision to stop in favour of the null hypothesis, H(O), on conditional power calculations for the end of the trial given the current data. Here I extend this procedure to the popular sequential designs: the O'Brien-Fleming test and the repeated significance test. I derive explicit boundaries for monitoring the test statistic useful for visualizing the impact of the parameters on the operating characteristics of the tests and thus for the design of the tests. Also, they facilitate the use of boundary crossing methods for approximations of power. I derive appropriate boundaries retrospectively for two clinical trials: one that concluded with no treatment difference (AIDS Clinical Trials Group protocol 118) and one that stopped early for positive effect (Beta-Blocker Heart Attack Trial). Finally, I compare the procedures based on the different upper boundaries and assess the impact of allowing for early stopping in favour of H(O), in numerical examples.Construction of a continuous stopping boundary from an alpha spending function
AbstractBetensky, R. (n.d.).Publication year
1998Journal title
BiometricsVolume
54Issue
3Page(s)
1061-1071AbstractLan and DeMets (1983, Biometrika 70, 659-663) proposed a flexible method for monitoring accumulating data that does not require the number and times of analyses to be specified in advance yet maintains an overall Type I error, α. Their method amounts to discretizing a preselected continuous boundary by clumping the density of the boundary crossing time at discrete analysis times and calculating the resultant discrete-time boundary values. In this framework, the cumulative distribution function of the continuous-time stopping rule is used as an alpha spending function. A key assumption that underlies this method is that future analysis times are not chosen on the basis of the current value of the statistic. However, clinical trials may be monitored more frequently when they are close to crossing the boundary. In this situation, the corresponding continuous-time boundary should be used. Here we demonstrate how to construct a continuous stopping boundary from an alpha spending function. This capability is useful also in the design of clinical trials. We use the Beta-Blocker Heart Attack Trial (BHAT) and AIDS Clinical Trials Group protocol 021 for illustration.Corelease of dopamine and GABA by a retinal dopaminergic neuron
AbstractHirasawa, H., Betensky, R., & Raviola, E. (n.d.).Publication year
2012Journal title
Journal of NeuroscienceVolume
32Issue
38Page(s)
13281-13291AbstractNumerous neurons release two transmitters of low molecular mass, but it is controversial whether they are localized within the same synaptic vesicle, with the single exception of GABA and glycine because they are ferried into the vesicle by the same transporter. Retinal dopaminergic (DAergic) amacrine cells synthesize both dopamine (DA) and GABA. Both transmitters are released over the entire cell surface and act on neighboring and distant neurons by volume transmission, but, in addition, DAergic cells establish GABAergic synapses onto AII amacrine cells, the neurons that transfer rod signals to cone bipolars. By combining recordings of DA and GABA release from isolated, genetically identified perikarya of DAergic cells from the mouse retina, we observed that a proportion of the events of DA and GABA exocytosis were simultaneous, suggesting corelease. Furthermore, a proportion of the secretory organelles in the perikaryon and synaptic endings of DAergic cells contained both vesicular transporters for DA [vesicular monoamine transporter 2 (VMAT2)] and GABA [vesicular GABA transporter (VGAT)]. Because the majority of the DA release events concerned a single transmitter and organelles were present that contained a single transporter, either VMAT2 or VGAT, we conclude that the secretory organelles of DAergic cells contain variable concentrations of the two transmitters, which are in turn determined by a variable mixture of the two transporter molecules in their limiting membrane. This variability can be explained if the relative numbers of transporter molecules is determined stochastically during the budding of the somatic organelles from the trans-Golgi network or the retrieval of the vesicular membrane from the plasmalemma after exocytosis.Correction : Association of anxiety with subcortical amyloidosis in cognitively normal older adults (Molecular Psychiatry, (2020), 25, 10, (2599-2607), 10.1038/s41380-018-0214-2)
AbstractHanseeuw, B. J., Jonas, V., Jackson, J., Betensky, R., Rentz, D. M., Johnson, K. A., Sperling, R. A., & Donovan, N. J. (n.d.).Publication year
2020Journal title
Molecular PsychiatryVolume
25Issue
10Page(s)
2644AbstractThis Article was originally published under Nature Research’s License to Publish, but has now been made available under a CC BY 4.0 license. The PDF and HTML versions of the Article have been modified accordingly.Correction to : Predictors for the use of systemic therapy in stage IB Mycosis fungoides (Archives of Dermatological Research, (2024), 316, 6, (337), 10.1007/s00403-024-03005-0)
AbstractRodriguez, E., Needle, C. D., Martinez, M. J., Nohria, A., Xing, Y., Song, C., Betensky, R., Latkowski, J. A., & Adotama, P. (n.d.).Publication year
2024Journal title
Archives of Dermatological ResearchVolume
316Issue
8AbstractThe abstract was incorrect in this article and should have read as follows. Cutaneous T Cell Lymphoma (CTCL) is an increasingly prevalent condition that impacts patients of all skin types. The most common subtype of CTCL is Mycosis Fungoides (MF), an indolent non-Hodgkin lymphoma of T cell origin that primarily develops in the skin. Most patients with Stage IB MF receive phototherapy as a first-line treatment; however, some require additional systemic therapy. We conducted an institutional review board (IRB)-approved single-center retrospective study of 172 patients diagnosed with Stage IB MF at NYU Langone Health from January 1990 to April 2021. Our study identified demographic and clinical parameters that may indicate which Stage IB patients will require systemic therapy in order to guide clinical decision making and ensure that these treatments can be offered earlier in the disease course. Of 172 total patients enrolled, 129 (75.0%) received phototherapy only, 33 (19.2%) received phototherapy and a systemic agent, and 10 (5.8%) were excluded for lack of documented treatment modality. Univariate statistical analysis demonstrated that advanced age at diagnosis (p < 0.005) and elevated LDH (p < 0.001) were statistically significant markers in the systemic therapy group. There were no significant differences in anatomical regions involved for patients on systemic agents. Other variables including ethnicity, Fitzpatrick skin type, and health insurance status were not statistically significant predictors of receiving systemic therapy. These results suggest that clinicians should routinely order LDH upon diagnosis for all patients with Stage IB MF. If elevated, patients should be more closely monitored for earlier initiation of systemic agents. Patients with Stage IB MF with an advanced age of diagnosis are also at higher risk of requiring systemic therapy. The original article has been corrected.Correlation among baseline variables yields non-uniformity of p-values
AbstractBetensky, R., & Chiou, S. H. (n.d.).Publication year
2017Journal title
PloS oneVolume
12Issue
9AbstractA recent paper in Neurology used statistical techniques to investigate the integrity of the randomization in 33 clinical trials conducted by a group of investigators. Without justification, the approach assumed that there would be no impact of correlation among baseline variables. We investigated the impact of correlation on the conclusions of the approach in several large-scale simulation studies that replicated the sample sizes and baseline variables of the clinical trials in question and utilized proper randomization. Additionally, we considered scenarios with larger numbers of baseline variables. We found that, with even moderate correlation, there can be substantial inflation of the type I error of statistical tests of randomization integrity. This is also the case under no correlation, in the presence of some discrete baseline variables, with a large number of variables. Thus, statistical techniques for assessing randomization integrity should be applied with extreme caution given that very low p-values, which are taken as evidence against valid randomization, can arise even in the case of valid randomization, in the presence of correlation. More generally, the use of tests of goodness of fit to uniformity for the purpose of testing a global null hypothesis is not advisable in the presence of correlation.