Marjorie A. Rosenberg

Results of a gender-climate and work-environment survey at a midwestern academic health center.

Purpose To determine how facultys perceptions of medical school gender climate differ by gender, track, rank, and departmental affiliation. Method In 1997, a 115-item questionnaire was sent to all University of Wisconsin Medical School faculty to assess their perceptions of mentoring, networking, professional environment, obstacles to a successful academic career, and reasons for considering leaving academic medicine. Using Fishers exact two-tailed test, the authors assessed gender differences both overall and by track, rank, and departmental cluster. Results Of the 836 faculty on tenure, clinician-educator, and clinical tracks, 507 (61%) responded. Although equal proportions of men and women had mentors, 24% of the women (compared with 6% of men; p < .001) felt that informal networking excluded faculty based on gender. Womens and mens perceptions differed significantly (p < .001) on 12 of 16 professional environment items (p < .05 on two of these items) and on five of six items regarding obstacles to academic success. While similar percentages of women and men indicated having seriously considered leaving academic medicine, their reasons differed: women cited work-family conflicts (51%), while men cited uncompetitive salaries (59%). These gender differences generally persisted across tracks, ranks, and departmental clusters. The greatest gender differences occurred among clinician-educators, associate professors, and primary care faculty. Conclusions Women faculty perceived that gender climate created specific, serious obstacles to their professional development. Many of those obstacles (e.g., inconvenient meeting times and lack of child care) are remediable. These data suggest that medical schools can improve the climate and retain and promote women by more inclusive networking, attention to meeting times and child care, and improved professional interactions between men and women faculty.

Rapid Increase in Breast Magnetic Resonance Imaging Use: Trends From 2000 to 2011

IMPORTANCE Breast magnetic resonance imaging (MRI) is highly sensitive for detecting breast cancer. Low specificity, cost, and little evidence regarding mortality benefits, however, limit recommendations for its use to high-risk women. How breast MRI is actually used in community settings is unknown. OBJECTIVE To describe breast MRI trends and indications in a community setting. DESIGN, SETTING, AND PARTICIPANTS Retrospective cohort study at a not-for-profit health plan and multispecialty group medical practice in New England of 10,518 women aged 20 years and older enrolled in the health plan for at least 1 year who had at least 1 breast MRI between January 1, 2000, and December 31, 2011. MAIN OUTCOMES AND MEASURES Breast MRI counts were obtained from claims data. Clinical indication (screening, diagnostic evaluation, staging or treatment, or surveillance) was determined using a prediction model developed from electronic medical records on a subset of participants. Breast cancer risk status was assessed using claims data and, for the subset, also through electronic medical record review. RESULTS; Breast MRI use increased more than 20-fold from 6.5 per 10,000 women in 2000 to 130.7 per 10,000 in 2009. Use then declined and stabilized to 104.8 per 10,000 by 2011. Screening and surveillance, rare indications in 2000, together accounted for 57.6% of MRI use by 2011; 30.1% had a claims-documented personal history and 51.7% a family history of breast cancer, whereas 3.5% of women had a documented genetic mutation. In the subset of women with electronic medical records who received screening or surveillance MRIs, only 21.0% had evidence of meeting American Cancer Society (ACS) criteria for breast MRI. Conversely, only 48.4% of women with documented deleterious genetic mutations received breast MRI screening. CONCLUSIONS AND RELEVANCE Breast MRI use increased steeply over 10 years and then stabilized, especially for screening and surveillance among women with family or personal history of breast cancer; most women receiving screening and surveillance breast MRIs lacked documented evidence of meeting ACS criteria, and many women with mutations were not screened. Efforts are needed to ensure that breast MRI use and documentation are focused on those women who will benefit most.

Physicians vary in approaches to the clinical management of delirium

OBJECTIVES: To ascertain the variation in strategies for managing delirium of physicians with expertise in geriatrics.

An elementary introduction to Bayesian computing using WinBUGS.

Bayesian statistics provides effective techniques for analyzing data and translating the results to inform decision making. This paper provides an elementary tutorial overview of the WinBUGS software for performing Bayesian statistical analysis. Background information on the computational methods used by the software is provided. Two examples drawn from the field of medical decision making are presented to illustrate the features and functionality of the software.

A Decision-Tree Approach to Cost Comparison of Newborn Screening Strategies for Cystic Fibrosis

OBJECTIVE: Because cystic fibrosis can be difficult to diagnose and treat early, newborn screening programs have rapidly developed nationwide but methods vary widely. We therefore investigated the costs and consequences or specific outcomes of the 2 most commonly used methods. METHODS: With available data on screening and follow-up, we used a simulation approach with decision trees to compare immunoreactive trypsinogen (IRT) screening followed by a second IRT test against an IRT/DNA analysis. By using a Monte Carlo simulation program, variation in the model parameters for counts at various nodes of the decision trees, as well as for costs, are included and applied to fictional cohorts of 100 000 newborns. The outcome measures included the numbers of newborns given a diagnosis of cystic fibrosis and costs of screening strategy at each branch and cost per newborn. RESULTS: Simulations revealed a substantial number of potential missed diagnoses for the IRT/IRT system versus IRT/DNA. Although the IRT/IRT strategy with commonly used cutoff values offers an average overall cost savings of

Predictive Modeling with Longitudinal Data: A Case Study Of Wisconsin Nursing Homes

2.30 per newborn, a breakdown of costs by societal segments demonstrated higher out-of-pocket costs for families. Two potential system failures causing delayed diagnoses were identified relating to the screening protocols and the follow-up system. CONCLUSIONS: The IRT/IRT screening algorithm reduces the costs to laboratories and insurance companies but has more system failures. IRT/DNA offers other advantages, including fewer delayed diagnoses and lower out-of-pocket costs to families.

A bayesian approach to understanding time series data

Abstract The recent development and availability of sophisticated computer software has facilitated the use of predictive modeling by actuaries and other financial analysts. Predictive modeling has been used for several applications in both the health and property and casualty sectors. Often these applications employ extensions of industry-specific techniques and do not make full use of information contained in the data. In contrast, we employ fundamental statistical methods for predictive modeling that can be used in a variety of disciplines. As demonstrated in this article, this methodology permits a disciplined approach to model building, including model development and validation phases. This article is intended as a tutorial for the analyst interested in using predictive modeling by making the process more transparent. This article illustrates the predictive modeling process using State of Wisconsin nursing home cost reports. We examine utilization of approximately 400 nursing homes from 1989 to 2001. Because the data vary both in the cross section and over time, we employ longitudinal models. This article demonstrates many of the common difficulties that analysts face in analyzing longitudinal health care data, as well as techniques for addressing these difficulties. We find that longitudinal methods, which use historical trend information, significantly outperform regression models that do not take advantage of historical trends.

Computing Population-based Estimates of Health-adjusted Life Expectancy

Abstract This paper explores the use of Bayesian models to analyze time series data. The Bayesian approach produces output that can be readily understood by actuaries and included in their own experience studies. We illustrate this Bayesian approach by analyzing U.S. unemployment rates, a macroeconomic time series. Understanding time series of macroeconomic variables can help actuaries in pricing and reserving their products. For example, a change in the level and/or variance of the unemployment series is of interest to actuaries, because its movement can explain a changing pattern of lapse rates of incidence rates. Our Bayesian analysis, based on models developed by McCulloch and Tsay (1993, 1994), allows for shifts in the level and in the error variance of a process. We develop a measure of model fit, based on the Akaike Information Criterion, that can be used in choosing between alternative models. Posterior prediction intervals for the fitted values are also created to pictorially show the range of pa...

Forecasting Social Security Actuarial Assumptions

Observed health-adjusted life expectancy (HALE) is an indicator of population health. There are a number of ways to compute HALE for a community. The authors surveyed several methods and demonstrate resulting variation in the estimates of HALE. Quality of well-being (QWB) measures from 1,430 participants in the Beaver Dam Health Out comes Study are taken as weights. Actuarial life-table methods using community mor tality data, State of Wisconsin census data from two time frames, and U.S. census data are used with the QWB to estimate HALE. Measurement of community population health using HALE computations can be completed with national, regional, or local data. Community-level estimates may not be well approximated using large-scale mor tality experience. A Bayesian method is developed combining the local data with re gional data. The Bayesian method creates a smooth set of rates, retains the local flavor of the community, and gives a measure of variability of the estimated HALE. Key words: quality-of-life measures; Bayesian methodology; Gibbs sampling. (Med Decis Making 1999;19:90-97)

A Statistical Model to Detect DRG Upcoding

Abstract This paper presents a forecasting model of economic assumptions that are inputs to projections of the Social Security system. Social Security projections are made to help policy-makers understand the financial stability of the system. Because system income and expenditures are subject to changes in law, they are controllable and not readily amenable to forecasting techniques. Hence, we focus directly on the four major economic assumptions to the system: inflation rate, investment returns, wage rate, and unemployment rate. Population models, the other major input to Social Security projections, require special demographic techniques and are not addressed here. Our approach to developing a forecasting model emphasizes exploring characteristics of the data. That is, we use graphical techniques and diagnostic statistics to display patterns that are evident in the data. These patterns include (1) serial correlation, (2) conditional heteroscedasticity, (3) contemporaneous correlations, and (4) cross-co...