MCMary E. Charlson

Mary E. Charlson

*This is an algorithmically generated author record based on publicly available information.

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


Mary E. Charlson has a total of 124 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Charlson comorbidity health analytics: A population management strategy to identify risk of hospitalizations, repeated hospitalizations, and resultant high cost

    by Mary Charlson, Martin T. Wells, James P. Hollenberg

    Background: Building on prior development work, the objective of this study of health care utilization of all Weill Cornell Medicine health insurance beneficiaries over a six-year period was to demonstrate the validity of the Charlson Comorbidity Health Analytics (CCHA), a summed weighted measure of 38 chronic conditions in adults and children, that prospectively predict longitudinal risk of hospital admissions, repeated admissions and resultant high cost in populations. The objective of the Charlson Comorbidity Health Analytics (CCHA) is to provide a new foundational framework for population management strategies by identifying the highest risk patients who can then be the focus for interventions designed to reduce unplanned hospitalizations and resultant high costs. Methods: All 27,190 Weill Cornell Medicine beneficiaries in the years 2016–2021, that is, employees and their dependents, including spouses/partners and their children, were linked across the years in a de-identified way, and CCHA was calculated from claims data. In addition to basic demographics, data included all outpatient and inpatient claims, including payments for each service over each year, excluding pharmacy. While two pharmaceuticals are part of the CCHA (anticoagulants and anti-psychotics), no data about pharmaceuticals was available for this analysis. First, CCHA from each year 2016–2021 was evaluated cross-sectionally as a predictor of that year’s hospitalizations and costs. Second, the CCHA from 2016 beneficiaries who were followed for five years were used to predict longitudinal risk of hospitalizations, repeated hospitalizations, and costs in each of the next five years. Then the CCHA was compared to the CMS Chronic Conditions Warehouse 30 (CCW30) measure. Finally, the CCHA from any given year (2016–2021) was analyzed for its predictive ability over the remaining one to five years of follow-up to predict hospitalizations, repeated hospitalizations, and costs. Results: Of the total 27,190 beneficiaries over the six years, 55.8% were employees (66.2% women with an average age of 40.9 years), and 25.7% children (average age of 6.1 years). The Charlson Comorbidity Health Analytics (CCHA) score from an index year longitudinally predicts the risk of hospitalizations--including repeated hospitalizations--which drive healthcare costsover six years (p < .01), providing the foundation for interventions in the highest risk patients. Moreover, the 2016 CCHA was a more significant predictor of readmission in 2017–2021 than a 2016 admission. In addition, comorbidity from any index year can be used to predict subsequent admissions and costs; therefore, it works in dynamic populations, like employers and unions that have changes in beneficiaries over time. Conclusions and Relevance: The Charlson Comorbidity Health Analytics is a method for prospectively identifying the small percent of patients who are at high longitudinal risk for unplanned hospitalizations and high costs. Intervention efforts can then be focused on high-comorbidity patients at high risk [1], with the goal of preventing health deterioration leading to health crises. Comorbidity Health Analytics provides a new foundational framework for population management strategies and specifically for interventions designed to reduce unplanned hospitalizations and thereby reduce costs.

  • Open Access

    Statin Use Among Women and Men Following Coronary Artery Bypass Surgery

    by Sigrid Sandner, Alexandra Kaider, Julia Riebandt, Alissa Florian, Selma Rizvanovic, C. Noel Bairey Merz, Jennifer S. Lawton, Mary Charlson, Monika M. Safford, Thomas Bergmair, et al.

    Background: Limited data exist on sex differences in guideline‐recommended statin therapy for secondary prevention after coronary artery bypass surgery (CABG). We examined sex differences in statin use after CABG and the association between sex‐specific statin use and mortality. Methods and Results: Data from the Austrian national cardiac surgery registry and federal social insurance claims database for patients who underwent CABG between 2013 and 2021 were used. Multivariable logistic regression models were calculated to obtain women‐to‐men odds ratios for filling any statin and high‐intensity statin prescriptions. Cox proportional hazards models were used to evaluate the association between statin use and mortality. A total of 15 448 patients (19% women) were included. During the 5 years after CABG, statin use decreased from 95.7% to 85.9% in men and 95.2% to 84.3% in women (P for trend P int =0.48), high‐intensity statin use decreased from 69.4% to 57.2% in men and 67.8% to 54.3% in women (P for trend P int =0.59). The adjusted odds ratio for filling any statin prescription comparing women with men was 1.03 (95% CI, 0.92–1.16) and for filling a high‐intensity statin prescription was 1.12 (95% CI, 1.02–1.23). Statin use was associated with a significantly lower mortality risk in both sexes (any statin: hazard ratio [HR], 0.56 [95% CI, 0.46–0.68]; P P int =0.22; high‐intensity statin: HR, 0.52 [95% CI, 0.42–0.63]; P P int =0.48). Conclusions: Women were as likely as men to fill a statin prescription after CABG and more likely to fill a high‐intensity prescription. Statin use was associated with a similar mortality risk reduction among women and men.


  • Open Access

    Interval life events are an important determinant of heterogeneity in outcomes in a randomised trial: a novel, simple method of assessment

    by Mary Charlson, Martin Wells, Carol M Devine, Jerome Watts, Rosio Ramos, James Hollenberg, Ginger Winston, Erica Phillips, Elaine Wethington

    Objective Although life events are clearly important to health, most of the scientific focus has been on baseline life events that occur prior to a study. Life events that occur after enrolment, that is, interval life events, have had almost no attention. The aim of this analysis of data was to develop a method for measuring interval life events that could be used in clinical trials and other longitudinal studies. Design Small Changes and Lasting Effects (SCALE) was a 12-month weight-loss randomised controlled trial (RCT). This was an analysis of the SCALE follow-up data. Setting Healthcare networks, outpatient clinics and community churches in the South Bronx and Harlem areas of New York City. Participants Overweight black and Latino adults. This analysis focuses on the 330 of the 405 patients who had >4 weeks of follow-up with at least one perceived stress score (PSS). Intervention The SCALE RCT was published elsewhere and involved positive affect and self-affirmation to increase behaviour change. Outcome 5% weight loss. Follow-up Over 12 months, up to 27 follow-ups were conducted that evaluated interval life events, eating and physical activity behaviour, weight and perceived stress. During these follow-ups, participants were asked two open-ended questions to capture interval life events. The interval life events were qualitatively coded into categories. The interval life events categories were compared with interval monthly measures of perceived stress using the 4-item PSS scale. Results During the interval follow-ups for the RCT, 70.6% of the 330 patients reported at least one interval life event, which occurred during a median of 15 follow-ups (95% CI: 5 to 24). The median number of interval events was 2 (95% CI: 0 to 8): 30.6% reported their own illness; 22%, death or bereavement; 21.8%, illness in the family and 13.1%, family conflicts. The mean perceived stress score (PSS-4) assessed over the year of follow-up was 3.2±2.7. Mean perceived stress (PSS-4) increased, especially for interval financial events, major conflict with a partner and unemployment, but by less for deaths, family illness and family conflict. Participants with the most interval life events had the greatest increase in interval perceived stress (p20) nor baseline depression (Patient Health Questionnaire-9 >10) were associated with higher interval life events (p>0.05); but those with lower social support had more events. However, those with either depression or stress had higher interval stress responses. Most participants had neither baseline nor interval events, and the percentage with both was small so that baseline events did not predict subsequent perceived stress. Conclusions This method provides a straightforward method of assessing interval life events, by asking two open-ended questions, that can be coded in a simple categorical framework. Such events can affect outcomes in longitudinal studies and trials in part by increasing perceived stress. This framework moves beyond the events identified as important in the 1950s and recognises that specific life events may have significantly different life impacts in different individuals. Trial registeration number NCT01198990; Post-results.


  • Open Access

    Comorbidity: From a Confounder in Longitudinal Clinical Research to the Main Issue in Population Management

    by Mary E. Charlson, Martin T. Wells

    In his recent editorial “Forty Years of Clinimetrics,” Fava eloquently summarized the contribution of clinimetrics to moving clinical science from a focus on diagnosis and treatment of a single disease in isolation to the understanding that any illness occurs in a person whose life context, including their individual, family, community and environmental stresses and supports, immeasurably impacts their prognosis, overall physical and psychosocial well-being and outcomes [1]. Moreover, with any specific illness, an individual’s symptoms and severity, progression of illness, response to treatment and quality of life is complex and often not a linear function of a specific treatment [1].Alvan Feinstein coined the term “clinimetrics” to define a new set of methods and approaches to measure clinical phenomena [2, 3]. Under the broad umbrella of clinimetrics, he developed sets of principles relevant to many issues, among them assessment of prognosis [4]. In vintage form (taking no prisoners), Feinstein argued: “There never has been and never will be a natural history population. This belief may have been valid for agricultural plots, brewery vats and fruit flies, but the belief is not true for a free-living population. The first step for scientists who want to investigate prognosis, therefore is to recognize that we can study clinical course, but not natural history” [4]. At Yale, in his inimitable fashion, Feinstein went on to describe many new clinimetric methods and principles, among them the concept of comorbidity which he defined as “any distinct additional clinical entity that has existed or that may occur during the clinical course of a disease that is under study” and “that has the potential not only to impact a patient’s prognosis, but also to alter their therapeutic plans and outcomes” [5]. This includes conditions that can occur as lasting, but unintended consequence of treatment, that is, iatrogenic comorbidity [6]. Feinstein and Kaplan [7] subsequently developed the first method of assessing comorbidity.Building on the lessons learned at Yale, the Charlson comorbidity index was subsequently developed to provide an empirically based weighted measure of the prognostic impact of chronic conditions on long-term mortality [8]. Over the ensuing years, the Charlson comorbidity index has maintained its usefulness as a predictor of long-term mortality and has been employed in thousands of studies and has over 36,000 citations in Scopus. There are several likely reasons for its durability. It was developed in a population of hospitalized medical patients, controlling for age and the physician’s assessment of illness severity at the time of admission, to predict 1-year mortality. It was then validated in a completely separate population of breast cancer patients to predict 10-year mortality from diseases other than breast cancer [8]. The very different populations and the different follow-up intervals are likely key to its generalizability and its endurance over time. Before turning to the critical importance of comorbidity in the current era, the methodological issues in design of such measures of comorbidity and how they have been used (or not) will be reviewed.A number of studies that created new measures of comorbidity have developed them in one population and then validated it in a similar population – over a similar follow-up interval. Often, a given population has been divided into a split sample, with testing/development in one part of the sample (often selected at random) and validation in the other. In these circumstances, the predictions are likely to be largely concordant. Often, the validation of the measure will be good in the split sample but will perform considerably less well when applied to a different population [9]. In machine learning, this is called data drift – when an unexpected difference between the training and testing data violates the assumption that past predicts present [10].For those reasons, it’s preferable for such measures to be validated in a completely different population – over different time frames, in different geographic locations or in different clinical settings [11, 12]. This independent validation population is critical, because any validation population woven too tightly to the testing population is likely to drastically limit its generalizability [13]. Patients drawn from one population may differ from another population in ways that are not obvious at the outset. If the measure has enduring predictive validity, the specific outcome rates in each strata or rank of a predictive index may differ, but the gradients should be quite similar, that is, the index should discriminate between those with and without the outcome of interest in a graded fashion [14].Some efforts to assess comorbidity have included large numbers of conditions which may increase the prediction of the initial model; however, the use of a large number of conditions often radically reduces the predictive ability of the model when applied to another population [9]. Moreover, taking individual conditions, with or without weights, and testing them one by one for whether they should be included in a measure also tends to vastly overfit the models; again when the models are applied, their performance and generalizability drop dramatically [9].Some measures have been created to assess comorbidity relevant to one specific index disease. Yet the problem with assessing chronic conditions selected for their relevance to one disease is that it can result in many different indices relevant only to that disease (i.e., acute myelogenous leukemia, stem cell transplant, rheumatoid arthritis, etc.). If that is done for each chronic condition, there could be thousands of condition-specific comorbidity measures, which would not extrapolate across populations, resulting in confusion rather...


  • Reassembling the fragility index: a demonstration of statistical reasoning

    by Benjamin R. Baer, Mario Gaudino, Stephen E. Fremes, Mary Charlson, Martin T. Wells


  • Top Cited Papers
    Open Access

    Charlson Comorbidity Index: A Critical Review of Clinimetric Properties

    by Mary E. Charlson, Danilo Carrozzino, Jenny Guidi, Chiara Patierno

    The present critical review was conducted to evaluate the clinimetric properties of the Charlson Comorbidity Index (CCI), an assessment tool designed specifically to predict long-term mortality, with regard to its reliability, concurrent validity, sensitivity, incremental and predictive validity. The original version of the CCI has been adapted for use with different sources of data, ICD-9 and ICD-10 codes. The inter-rater reliability of the CCI was found to be excellent, with extremely high agreement between self-report and medical charts. The CCI has also been shown either to have concurrent validity with a number of other prognostic scales or to result in concordant predictions. Importantly, the clinimetric sensitivity of the CCI has been demonstrated in a variety of medical conditions, with stepwise increases in the CCI associated with stepwise increases in mortality. The CCI is also characterized by the clinimetric property of incremental validity, whereby adding the CCI to other measures increases the overall predictive accuracy. It has been shown to predict long-term mortality in different clinical populations, including medical, surgical, intensive care unit (ICU), trauma, and cancer patients. It may also predict in-hospital mortality, although in some instances, such as ICU or trauma patients, the CCI did not perform as well as other instruments designed specifically for that purpose. The CCI thus appears to be clinically useful not only to provide a valid assessment of the patient’s unique clinical situation, but also to demarcate major diagnostic and prognostic differences among subgroups of patients sharing the same medical diagnosis.


  • Open Access

    On clinical trial fragility due to patients lost to follow up

    by Benjamin R. Baer, Stephen E. Fremes, Mario Gaudino, Mary Charlson, Martin T. Wells

    Background: Clinical trials routinely have patients lost to follow up. We propose a methodology to understand their possible effect on the results of statistical tests by altering the concept of the fragility index to treat the outcomes of observed patients as fixed but incorporate the potential outcomes of patients lost to follow up as random and subject to modification.Methods: We reanalyse the statistical results of three clinical trials on coronary artery bypass grafting (CABG) to study the possible effect of patients lost to follow up on the treatment effect statistical significance. To do so, we introduce the LTFU-aware fragility indices as a measure of the robustness of a clinical trial’s statistical results with respect to patients lost to follow up.Results: The analyses illustrate that clinical trials can either be completely robust to the outcomes of patients lost to follow up, extremely sensitive to the outcomes of patients lost to follow up, or in an intermediate state. When a clinical trial is in an intermediate state, the LTFU-aware fragility indices provide an interpretable measure to quantify the degree of fragility or robustness.Conclusions: The LTFU-aware fragility indices allow researchers to rigorously explore the outcomes of patients who are lost to follow up, when their data is the appropriate kind. The LTFU-aware fragility indices are sensitivity measures in a way that the original fragility index is not.


  • The fragility index can be used for sample size calculations in clinical trials

    by Benjamin R. Baer, Mario Gaudino, Stephen E. Fremes, Mary Charlson, Martin T. Wells


  • Open Access

    Diabetes Intervention Accentuating Diet and Enhancing Metabolism (DIADEM-I): a randomised controlled trial to examine the impact of an intensive lifestyle intervention consisting of a low-energy diet and physical activity on body weight and metabolism in early type 2 diabetes mellitus: study protocol for a randomized controlled trial

    by Shahrad Taheri, Odette Chagoury, Hadeel Zaghloul, Sara Elhadad, Salma Hayder Ahmed, Omar Omar, Sherryl Payra, Salma Ahmed, Neda El Khatib, Rasha Abou Amona, et al.

    Type 2 diabetes mellitus (T2DM) and obesity are syndemic and will have a significant impact on affected individuals and healthcare services worldwide. Evidence shows that T2DM remission can be achieved with significant weight loss in those who are younger with early diabetes and requiring fewer medications for glycaemic control. DIADEM-I aims to examine the impact of an intensive lifestyle intervention (ILI) using a low-energy diet (LED) meal replacement approach combined with physical activity in younger individuals with early T2DM. The planned study is an ongoing, non-blinded, pragmatic, randomised controlled, parallel-group trial examining the impact of an LED-based ILI on body weight and diabetes remission in younger (18–50 years) T2DM individuals with early diabetes (≤ 3-year duration). The ILI will be compared to usual medical care (UMC). The primary outcome will be weight loss at 12 months. Other key outcomes of interest include diabetes remission, glycaemic control, diabetes complications, cardiovascular health, physical activity, mental health, and quality of life. It is planned for the study to include 138 subjects for assessment of the primary outcome. Safety will be assessed throughout. If DIADEM-I demonstrates a clinically significant effect for younger individuals with early T2DM, it will inform clinical guidelines and services of the future for management of T2DM. ISRCTN: ISRCTN20754766 (date assigned: 7 June 2017); ClinicalTrials.gov, ID: NCT03225339 Registered on 26 June 2017.


  • Environmental influences on small eating behavior change to promote weight loss among Black and Hispanic populations

    by Johanna D. Eldridge, Carol M. Devine, Elaine Wethington, Luz Aceves, Erica Phillips-Caesar, Brian Wansink, Mary E. Charlson