Michelle Shardell
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Michelle Shardell has a total of 46 co-authors. Here are the top co-authors by number of shared publications:
- Gregory Evan Hicks
Department of Physical Therapy, University of Delaware, Newark, DE, USA
10 shared publications Ram R. Miller
Department of Epidemiology and Public Health, University of Maryland School of Medicine, Baltimore, MD
7 shared publicationsJay Magaziner
From the aDepartment of Epidemiology and Public Health, University of Maryland School of Medicine; bNational Institute on Aging; cDepartment of Physical Therapy, University of Delaware; and dUniversity of Maryland School of Nursing, Baltimore, MD
5 shared publicationsLuigi Ferrucci
National Institute on Aging, 3001 S Hanover Street, Baltimore, MD 21225, USA
5 shared publicationsDaniel O. Scharfstein
Johns Hopkins Bloomberg School of Public Health
3 shared publications- Eli N. Perencevich
Center for Access & Delivery Research and Evaluation (CADRE), Iowa City VA Health Care System , Iowa City, Iowa
3 shared publications
Latest Publications
- Open Access
- 4 July 2014
- Published by Oxford University Press (OUP) in Biostatistics
- Vol. 16 (1) , 155-168
Doubly robust estimation and causal inference in longitudinal studies with dropout and truncation by death
Motivated by aging research, we propose an estimator of the effect of a time-varying exposure on an outcome in longitudinal studies with dropout and truncation by death. We use an inverse-probability weighted (IPW) estimator to derive a doubly robust augmented inverse-probability weighted (AIPW) estimator. IPW estimation involves weights for the exposure mechanism, dropout, and mortality; AIPW estimation additionally involves estimating data-generating models via regression. We demonstrate that the estimators identify a causal contrast that is a function of principal strata effects under a set of assumptions. Simulations show that AIPW estimation is unbiased when weights or outcome regressions are correct, and that AIPW estimation is more efficient than IPW estimation when all models are correct. We apply the method to a study of vitamin D and gait speed among older adults.
- 4 July 2014
- Published by Oxford University Press (OUP) in Biostatistics
- Vol. 16 (1) , 155-168
Open AccessDoubly robust estimation and causal inference in longitudinal studies with dropout and truncation by death
Motivated by aging research, we propose an estimator of the effect of a time-varying exposure on an outcome in longitudinal studies with dropout and truncation by death. We use an inverse-probability weighted (IPW) estimator to derive a doubly robust augmented inverse-probability weighted (AIPW) estimator. IPW estimation involves weights for the exposure mechanism, dropout, and mortality; AIPW estimation additionally involves estimating data-generating models via regression. We demonstrate that the estimators identify a causal contrast that is a function of principal strata effects under a set of assumptions. Simulations show that AIPW estimation is unbiased when weights or outcome regressions are correct, and that AIPW estimation is more efficient than IPW estimation when all models are correct. We apply the method to a study of vitamin D and gait speed among older adults.
- 17 June 2014
- Published by Wiley in Statistics in Medicine
- Vol. 33 (25) , 4437-4452
Statistical analysis with missing exposure data measured by proxy respondents: a misclassification problem within a missing‐data problem
In studies of older adults, researchers often recruit proxy respondents, such as relatives or caregivers, when study participants cannot provide self‐reports (e.g., because of illness). Proxies are usually only sought to report on behalf of participants with missing self‐reports; thus, either a participant self‐report or proxy report, but not both, is available for each participant. Furthermore, the missing‐data mechanism for participant self‐reports is not identifiable and may be nonignorable. When exposures are binary and participant self‐reports are conceptualized as the gold standard, substituting error‐prone proxy reports for missing participant self‐reports may produce biased estimates of outcome means. Researchers can handle this data structure by treating the problem as one of misclassification within the stratum of participants with missing self‐reports. Most methods for addressing exposure misclassification require validation data, replicate data, or an assumption of nondifferential misclassification; other methods may result in an exposure misclassification model that is incompatible with the analysis model. We propose a model that makes none of the aforementioned requirements and still preserves model compatibility. Two user‐specified tuning parameters encode the exposure misclassification model. Two proposed approaches estimate outcome means standardized for (potentially) high‐dimensional covariates using multiple imputation followed by propensity score methods. The first method is parametric and uses maximum likelihood to estimate the exposure misclassification model (i.e., the imputation model) and the propensity score model (i.e., the analysis model); the second method is nonparametric and uses boosted classification and regression trees to estimate both models. We apply both methods to a study of elderly hip fracture patients. Copyright © 2014 John Wiley & Sons, Ltd.
- 17 June 2014
- Published by Wiley in Statistics in Medicine
- Vol. 33 (25) , 4437-4452
Statistical analysis with missing exposure data measured by proxy respondents: a misclassification problem within a missing‐data problem
In studies of older adults, researchers often recruit proxy respondents, such as relatives or caregivers, when study participants cannot provide self‐reports (e.g., because of illness). Proxies are usually only sought to report on behalf of participants with missing self‐reports; thus, either a participant self‐report or proxy report, but not both, is available for each participant. Furthermore, the missing‐data mechanism for participant self‐reports is not identifiable and may be nonignorable. When exposures are binary and participant self‐reports are conceptualized as the gold standard, substituting error‐prone proxy reports for missing participant self‐reports may produce biased estimates of outcome means. Researchers can handle this data structure by treating the problem as one of misclassification within the stratum of participants with missing self‐reports. Most methods for addressing exposure misclassification require validation data, replicate data, or an assumption of nondifferential misclassification; other methods may result in an exposure misclassification model that is incompatible with the analysis model. We propose a model that makes none of the aforementioned requirements and still preserves model compatibility. Two user‐specified tuning parameters encode the exposure misclassification model. Two proposed approaches estimate outcome means standardized for (potentially) high‐dimensional covariates using multiple imputation followed by propensity score methods. The first method is parametric and uses maximum likelihood to estimate the exposure misclassification model (i.e., the imputation model) and the propensity score model (i.e., the analysis model); the second method is nonparametric and uses boosted classification and regression trees to estimate both models. We apply both methods to a study of elderly hip fracture patients. Copyright © 2014 John Wiley & Sons, Ltd.
- Open Access
- 14 September 2013
- Published by Oxford University Press (OUP) in International Journal of Epidemiology
- Vol. 42 (5) , 1509-10
Commentary: Back to the future with Sir Bradford Hill: statistical analysis with hospital-acquired infections.
- 14 September 2013
- Published by Oxford University Press (OUP) in International Journal of Epidemiology
- Vol. 42 (5) , 1509-10
Open AccessCommentary: Back to the future with Sir Bradford Hill: statistical analysis with hospital-acquired infections.
- 1 March 2013
- Published by Wolters Kluwer Health in Epidemiology
- Vol. 24 (2) , 215-223
Sensitivity Analysis for Nonignorable Missingness and Outcome Misclassification from Proxy Reports
Researchers often recruit proxy respondents, such as relatives or caregivers, for epidemiologic studies of older adults when study participants are unable to provide self-reports (eg, because of illness or cognitive impairment). In most studies involving proxy-reported outcomes, proxies are recruited only to report on behalf of participants who have missing self-reported outcomes; thus, either a proxy report or participant self-report, but not both, is available for each participant. When outcomes are binary and investigators conceptualize participant self-reports as gold standard measures, substituting proxy reports in place of missing participant self-reports in statistical analysis can introduce misclassification error and lead to biased parameter estimates. However, excluding observations from participants with missing self-reported outcomes may also lead to bias. We propose a pattern-mixture model that uses error-prone proxy reports to reduce selection bias from missing outcomes, and we describe a sensitivity analysis to address bias from differential outcome misclassification. We perform model estimation with high-dimensional (eg, continuous) covariates using propensity-score stratification and multiple imputation. We apply the methods to the Second Cohort of the Baltimore Hip Studies, a study of elderly hip fracture patients, to assess the relation between type of surgical treatment and perceived physical recovery. Simulation studies show that the proposed methods perform well. We provide SAS programs in the eAppendix ( http://links.lww.com/EDE/A646 ) to enhance the methods’ accessibility.
- 1 March 2013
- Published by Wolters Kluwer Health in Epidemiology
- Vol. 24 (2) , 215-223
Sensitivity Analysis for Nonignorable Missingness and Outcome Misclassification from Proxy Reports
Researchers often recruit proxy respondents, such as relatives or caregivers, for epidemiologic studies of older adults when study participants are unable to provide self-reports (eg, because of illness or cognitive impairment). In most studies involving proxy-reported outcomes, proxies are recruited only to report on behalf of participants who have missing self-reported outcomes; thus, either a proxy report or participant self-report, but not both, is available for each participant. When outcomes are binary and investigators conceptualize participant self-reports as gold standard measures, substituting proxy reports in place of missing participant self-reports in statistical analysis can introduce misclassification error and lead to biased parameter estimates. However, excluding observations from participants with missing self-reported outcomes may also lead to bias. We propose a pattern-mixture model that uses error-prone proxy reports to reduce selection bias from missing outcomes, and we describe a sensitivity analysis to address bias from differential outcome misclassification. We perform model estimation with high-dimensional (eg, continuous) covariates using propensity-score stratification and multiple imputation. We apply the methods to the Second Cohort of the Baltimore Hip Studies, a study of elderly hip fracture patients, to assess the relation between type of surgical treatment and perceived physical recovery. Simulation studies show that the proposed methods perform well. We provide SAS programs in the eAppendix ( http://links.lww.com/EDE/A646 ) to enhance the methods’ accessibility.
- Open Access
- 1 January 2013
- Published by Springer Nature in Antimicrobial Resistance & Infection Control
- Vol. 2 (1) , 28
Optimizing antimicrobial prescribing: Are clinicians following national trends in methicillin-resistant staphylococcus aureus (MRSA) infections rather than local data when treating MRSA wound infections
Clinicians often prescribe antimicrobials for outpatient wound infections before culture results are known. Local or national MRSA rates may be considered when prescribing antimicrobials. If clinicians prescribe in response to national rather than local MRSA trends, prescribing may be improved by making local data accessible. We aimed to assess the correlation between outpatient trends in antimicrobial prescribing and the prevalence of MRSA wound infections across local and national levels.
- 1 January 2013
- Published by Springer Nature in Antimicrobial Resistance & Infection Control
- Vol. 2 (1) , 28
Open AccessOptimizing antimicrobial prescribing: Are clinicians following national trends in methicillin-resistant staphylococcus aureus (MRSA) infections rather than local data when treating MRSA wound infections
Clinicians often prescribe antimicrobials for outpatient wound infections before culture results are known. Local or national MRSA rates may be considered when prescribing antimicrobials. If clinicians prescribe in response to national rather than local MRSA trends, prescribing may be improved by making local data accessible. We aimed to assess the correlation between outpatient trends in antimicrobial prescribing and the prevalence of MRSA wound infections across local and national levels.
- 1 May 2012
- Published by Elsevier in Nursing Outlook
- Vol. 60 (3) , 134-142.e1
Housing strain, mortgage foreclosure, and health
- 1 May 2012
- Published by Elsevier in Nursing Outlook
- Vol. 60 (3) , 134-142.e1
Housing strain, mortgage foreclosure, and health
- 27 January 2012
- Published by Wiley in Journal of the American Geriatrics Society
- Vol. 60 (2) , 256-264
Serum 25‐Hydroxyvitamin D , Transitions Between Frailty States, and Mortality in Older Adults: The Invecchiare in Chianti Study
Objectives: To assess whether serum 25‐hydroxyvitamin D (25( OH ) D ) concentrations relate to transitions between the states of robustness, prefrailty, and frailty and to mortality in older adults. Design: The Invecchiare in Chianti (In CHIANTI ) Study, a prospective cohort study. Setting: Tuscany, I taly. Participants: Adults aged 65 and older (N = 1,155). Measurements: Serum 25( OH ) D concentrations measured at baseline; frailty state (robust, prefrail, frail) assessed at baseline and 3 and 6 years after enrollment; and vital status determined 3 and 6 years after enrollment. Results: The median (interquartile range) 25( OH ) D concentration was 16.0 ng/mL (10.4–25.6 ng/mL; multiply by 2.496 to convert to nmol/L). Prefrail participants with 25( OH ) D levels less than 20 ng/mL were 8.9% (95% confidence interval ( CI ) = 2.5–15.2%) more likely to die, 3.0% (95% CI = −5.6–14.6%) more likely to become frail, and 7.7% (95% CI = −3.5–18.7%) less likely to become robust than prefrail participants with 25( OH ) D levels of 20 ng/mL or more. In prefrail participants, each 5‐ng/mL decrement of continuous 25( OH ) D was associated with 1.46 times higher odds of dying (95% CI = 1.18–2.07) and 1.13 higher odds of incident frailty (95% CI = 0.90–1.39) than with recovery of robustness. Transitions from robustness or frailty were not associated with 25( OH ) D levels. Conclusion: Results provide evidence that prefrailty is an “at risk” state from which older adults with high 25( OH ) D levels are more likely to recover than to decline, but high 25( OH ) D levels were not associated with recovery from frailty. Thus, 25( OH ) D levels should be investigated as a potential therapy to treat prefrailty and prevent further decline.
- 27 January 2012
- Published by Wiley in Journal of the American Geriatrics Society
- Vol. 60 (2) , 256-264
Serum 25‐Hydroxyvitamin D , Transitions Between Frailty States, and Mortality in Older Adults: The Invecchiare in Chianti Study
Objectives: To assess whether serum 25‐hydroxyvitamin D (25( OH ) D ) concentrations relate to transitions between the states of robustness, prefrailty, and frailty and to mortality in older adults. Design: The Invecchiare in Chianti (In CHIANTI ) Study, a prospective cohort study. Setting: Tuscany, I taly. Participants: Adults aged 65 and older (N = 1,155). Measurements: Serum 25( OH ) D concentrations measured at baseline; frailty state (robust, prefrail, frail) assessed at baseline and 3 and 6 years after enrollment; and vital status determined 3 and 6 years after enrollment. Results: The median (interquartile range) 25( OH ) D concentration was 16.0 ng/mL (10.4–25.6 ng/mL; multiply by 2.496 to convert to nmol/L). Prefrail participants with 25( OH ) D levels less than 20 ng/mL were 8.9% (95% confidence interval ( CI ) = 2.5–15.2%) more likely to die, 3.0% (95% CI = −5.6–14.6%) more likely to become frail, and 7.7% (95% CI = −3.5–18.7%) less likely to become robust than prefrail participants with 25( OH ) D levels of 20 ng/mL or more. In prefrail participants, each 5‐ng/mL decrement of continuous 25( OH ) D was associated with 1.46 times higher odds of dying (95% CI = 1.18–2.07) and 1.13 higher odds of incident frailty (95% CI = 0.90–1.39) than with recovery of robustness. Transitions from robustness or frailty were not associated with 25( OH ) D levels. Conclusion: Results provide evidence that prefrailty is an “at risk” state from which older adults with high 25( OH ) D levels are more likely to recover than to decline, but high 25( OH ) D levels were not associated with recovery from frailty. Thus, 25( OH ) D levels should be investigated as a potential therapy to treat prefrailty and prevent further decline.
- 29 December 2011
- Published by SAGE Publications in Journal of Aging and Health
- Vol. 24 (3) , 367-383
Comparing Reports From Hip-Fracture Patients and Their Proxies
Objective: This study compared sex differences in disability and depressive symptoms using reports from hip fracture patients and their proxies. Method: Hip fracture patients (49 men, 183 women) aged ≥65 years and proxies were interviewed 1 year postfracture. Outcomes were Center for Epidemiological Studies Depression (CES-D) and number of dependencies in performing activities of daily living and instrumental activities of daily living. Results: Mean ADL sex differences (men minus women) were 0.40 (p = .37) using proxy reports and 0.70 (p = .08) using patient self-reports. Mean CES-D sex differences were −3.60 (p = .02) using proxy reports and −1.26 (p = .38) using patient self-reports. Discrepancies between patients and proxies were smallest for proxies who have lived with the patient ≥ 1 year. Discussion: Patients and proxies produced conflicting conclusions about sex differences. Results suggest that ideal proxies to recruit are those who have been cohabitating with the patients for an extended length of time.
- 29 December 2011
- Published by SAGE Publications in Journal of Aging and Health
- Vol. 24 (3) , 367-383
Comparing Reports From Hip-Fracture Patients and Their Proxies
Objective: This study compared sex differences in disability and depressive symptoms using reports from hip fracture patients and their proxies. Method: Hip fracture patients (49 men, 183 women) aged ≥65 years and proxies were interviewed 1 year postfracture. Outcomes were Center for Epidemiological Studies Depression (CES-D) and number of dependencies in performing activities of daily living and instrumental activities of daily living. Results: Mean ADL sex differences (men minus women) were 0.40 (p = .37) using proxy reports and 0.70 (p = .08) using patient self-reports. Mean CES-D sex differences were −3.60 (p = .02) using proxy reports and −1.26 (p = .38) using patient self-reports. Discrepancies between patients and proxies were smallest for proxies who have lived with the patient ≥ 1 year. Discussion: Patients and proxies produced conflicting conclusions about sex differences. Results suggest that ideal proxies to recruit are those who have been cohabitating with the patients for an extended length of time.
- Open Access
- 19 October 2011
- Published by Springer Nature in BMC Infectious Diseases
- Vol. 11 (1) , 279
Comparative effectiveness of nafcillin or cefazolin versus vancomycin in methicillin-susceptible Staphylococcus aureus bacteremia
The high prevalence of methicillin-resistant S. aureus (MRSA) has led clinicians to select antibiotics that have coverage against MRSA, usually vancomycin, for empiric therapy for suspected staphylococcal infections. Clinicians often continue vancomycin started empirically even when methicillin-susceptible S. aureus (MSSA) strains are identified by culture. However, vancomycin has been associated with poor outcomes such as nephrotoxicity, persistent bacteremia and treatment failure. The objective of this study was to compare the effectiveness of vancomycin versus the beta-lactam antibiotics nafcillin and cefazolin among patients with MSSA bacteremia. The outcome of interest for this study was 30-day in-hospital mortality.
- 19 October 2011
- Published by Springer Nature in BMC Infectious Diseases
- Vol. 11 (1) , 279
Open AccessComparative effectiveness of nafcillin or cefazolin versus vancomycin in methicillin-susceptible Staphylococcus aureus bacteremia
The high prevalence of methicillin-resistant S. aureus (MRSA) has led clinicians to select antibiotics that have coverage against MRSA, usually vancomycin, for empiric therapy for suspected staphylococcal infections. Clinicians often continue vancomycin started empirically even when methicillin-susceptible S. aureus (MSSA) strains are identified by culture. However, vancomycin has been associated with poor outcomes such as nephrotoxicity, persistent bacteremia and treatment failure. The objective of this study was to compare the effectiveness of vancomycin versus the beta-lactam antibiotics nafcillin and cefazolin among patients with MSSA bacteremia. The outcome of interest for this study was 30-day in-hospital mortality.
- 1 March 2011
- Published by Elsevier in Nutrition Research
- Vol. 31 (3) , 178-189
Low-serum carotenoid concentrations and carotenoid interactions predict mortality in US adults: the Third National Health and Nutrition Examination Survey
- 1 March 2011
- Published by Elsevier in Nutrition Research
- Vol. 31 (3) , 178-189
Low-serum carotenoid concentrations and carotenoid interactions predict mortality in US adults: the Third National Health and Nutrition Examination Survey