Publications (5)15.9 Total impact
-
Article: Causal inference in epidemiological studies with strong confounding.
[show abstract] [hide abstract]
ABSTRACT: One of the identifiability assumptions of causal effects defined by marginal structural model (MSM) parameters is the experimental treatment assignment (ETA) assumption. Practical violations of this assumption frequently occur in data analysis when certain exposures are rarely observed within some strata of the population. The inverse probability of treatment weighted (IPTW) estimator is particularly sensitive to violations of this assumption; however, we demonstrate that this is a problem for all estimators of causal effects. This is due to the fact that the ETA assumption is about information (or lack thereof) in the data. A new class of causal models, causal models for realistic individualized exposure rules (CMRIER), is based on dynamic interventions. CMRIER generalize MSM, and their parameters remain fully identifiable from the observed data, even when the ETA assumption is violated, if the dynamic interventions are set to be realistic. Examples of such realistic interventions are provided. We argue that causal effects defined by CMRIER may be more appropriate in many situations, particularly those with policy considerations. Through simulation studies, we examine the performance of the IPTW estimator of the CMRIER parameters in contrast to that of the MSM parameters. We also apply the methodology to a real data analysis in air pollution epidemiology to illustrate the interpretation of the causal effects defined by CMRIER.Statistics in Medicine 02/2012; 31(13):1380-404. · 1.88 Impact Factor -
Article: Robust extraction of covariate information to improve estimation efficiency in randomized trials.
[show abstract] [hide abstract]
ABSTRACT: In randomized trials, investigators typically rely upon an unadjusted estimate of the mean outcome within each treatment arm to draw causal inferences. Statisticians have underscored the gain in efficiency that can be achieved from covariate adjustment in randomized trials with a focus on problems involving linear models. Despite recent theoretical advances, there has been a reluctance to adjust for covariates based on two primary reasons: (i) covariate-adjusted estimates based on conditional logistic regression models have been shown to be less precise and (ii) concern over the opportunity to manipulate the model selection process for covariate adjustments to obtain favorable results. In this paper, we address these two issues and summarize recent theoretical results on which is based a proposed general methodology for covariate adjustment under the framework of targeted maximum likelihood estimation in trials with two arms where the probability of treatment is 50%. The proposed methodology provides an estimate of the true causal parameter of interest representing the population-level treatment effect. It is compared with the estimates based on conditional logistic modeling, which only provide estimates of subgroup-level treatment effects rather than marginal (unconditional) treatment effects. We provide a clear criterion for determining whether a gain in efficiency can be achieved with covariate adjustment over the unadjusted method. We illustrate our strategy using a resampled clinical trial dataset from a placebo controlled phase 4 study. Results demonstrate that gains in efficiency can be achieved even with binary outcomes through covariate adjustment leading to increased statistical power.Statistics in Medicine 08/2011; 30(19):2389-408. · 1.88 Impact Factor -
Article: Response to emerging infection leading to outbreak of linezolid-resistant enterococci.
[show abstract] [hide abstract]
ABSTRACT: Linezolid was approved in 2000 for treatment of gram-positive coccal infections. We performed a case-control study during a hospital outbreak of linezolid-resistant enterococci (LRE) infections, comparing cases of LRE infection (cases) with linezolid-sensitive enterococci infections (controls). Nasal and perirectal swab samples were obtained from all patients in a 1-day point-prevalence survey. We examined antimicrobial drug use and calculated the defined daily dose of linezolid per 1,000 patient-days. Fifteen LRE cases were identified (13 Enterococcus faecalis and 2 E. faecium); 7 were vancomycin-resistant. Compared with controls, case-patients had increased in-hospital mortality rates and lengths of stay. Multivariate analysis identified independent predictors of LRE infection: prior cultures positive for methicillin-resistant Staphylococcus aureus (adjusted odds ratio [AOR] 27), hospitalization duration before index culture (AOR 1.1 per day), and duration of preceding linezolid therapy (AOR 1.1 per day). Linezolid exposure and patient-to-patient transmission appear to be responsible for LRE infections, an important emeraina hospital problem.Emerging infectious diseases 08/2007; 13(7):1024-30. · 6.17 Impact Factor -
Article: Neonatal sepsis in Egypt associated with bacterial contamination of glucose-containing intravenous fluids.
[show abstract] [hide abstract]
ABSTRACT: Rates of sepsis exceeding 50% in a neonatal intensive care unit (NICU) in Cairo, Egypt, were not controlled by routine antimicrobial therapy. We investigated these conditions in September 2001. Case series and retrospective cohort studies were conducted on 2 groups of NICU infants admitted to an academic medical center between February 12 and July 31, 2001. Observation of clinical practices led us to culture in-use intravenous (i.v.) fluids and medications. We monitored rates of i.v. fluid contamination, clinical sepsis and mortality after interventions to establish new procedures for handling and disposal of i.v. fluids, infection control training and improved clinical laboratory capacity. Among infants in the retrospective cohort group, 88 (77%) of 115 had clinical sepsis, and 59 (51%) died. In the case series group, we documented the time of initial positive blood culture; 21 (64%) of 33 were septic <24 hours after birth. Klebsiella pneumoniae accounted for 24 (73%) of 33 isolates; 14 (58%) of 24 were extended spectrum beta-lactamase-producing and aminoglycoside-resistant. On admission, all neonates received glucose-containing i.v. fluids; i.v. bottles (500 mL) were divided among multiple infants. The i.v. fluids were prepared at the bedside; poor hand hygiene and poor adherence to aseptic techniques were observed. K. pneumoniae was isolated from 13 (65%) of 20 in-use glucose-containing i.v. fluids. Fluid contamination, sepsis and mortality rates declined significantly after intervention. Extrinsically contaminated i.v. fluids resulted in sepsis and deaths. Standard infection control precautions significantly improve mortality and sepsis rates and are prerequisites for safe NICU care.The Pediatric Infectious Disease Journal 07/2005; 24(7):590-4. · 3.58 Impact Factor -
Article: Anthrax and the mail: the making of an educational video for mail workers.
[show abstract] [hide abstract]
ABSTRACT: The anthrax bioterrorist attacks in 2001 affected millions of people who process, sort, and deliver mail. To more effectively communicate information intended to protect the health of these workers, the Centers for Disease Control and Prevention produced a short-format educational video in December 2001 that targets this diverse group. This report illustrates how an educational video can be rapidly produced to translate and disseminate public health recommendations as part of a public health emergency response.American Journal of Infection Control 06/2003; 31(3):178-80. · 2.40 Impact Factor
Top co-authors
Top Journals
Institutions
-
2011–2012
-
University of California, Berkeley
- Division of Biostatistics
Berkeley, CA, USA
-
-
2003
-
Centers for Disease Control and Prevention
- Division of Healthcare Quality Promotion
Druid Hills, GA, USA
-
