Henry Krakauer

Gastroenterology Workforce Modeling

OBJECTIVE To examine the current supply and distribution of gastroenterologists and project future supply under various scenarios to provide a paradigm for workforce reform. DESIGN An analysis of current practices and distribution of gastroenterologists and a demographic model, using the 1992 gastroenterology workforce as a baseline. MAIN OUTCOME MEASURE Comparison of current supply, distribution, and practice profiles with past data and future projections, using analyses of data from the 1993 Area Resource File, 1992 Medicare Part B file, age- and sex-specific death and retirement rates from the Bureau of Health Professions, managed care staffing patterns, the National Survey of Internal Medicine Manpower, and the Bureau of the Census. RESULTS Rapid growth in the number of US gastroenterologists has resulted in a gastroenterologist-to-population ratio double that used on average by health maintenance organizations. In addition, the work profile of gastroenterologists is shared significantly by primary care physicians and other specialists, with the exception of a few specific and uncommon procedures. CONCLUSIONS Empirical evidence suggests that, even in the absence of detailed models to describe the desired supply/need balance for gastroenterology, the US health care system and clinicians may benefit from a reduction in gastroenterology training programs. The Gastroenterology Leadership Council endorsed a goal of 25% to 50% reduction in trainee numbers over 5 years, and recent National Resident Matching Program data indicate that a voluntary downsizing process is in full force. This study illustrates a paradigm for workforce planning that could be useful for other medical specialties.

The relationship of hospital ownership and teaching status to 30-and 180-day adjusted mortality rates

Hospital characteristics have been shown previously to be associated with variations in the probability of death within 30 days of admission. In the current study, the authors extend the examination of the relationship between hospital type to both short-term and long-term adjusted mortality. Observed and predicted 1988 hospital mortality rates were obtained from the Health Care Financing Administration (HCFA). A total of 3,782 acute care hospitals were divided into six mutually exclusive groups on the basis of their status as osteopathic, private for-profit, public teaching, public nonteaching, private teaching, and private nonteaching hospitals. After adjusting for the HCFA predicted mortality, Medicaid admissions, and emergency visits, 30-day and 30-to-180-day patient mortality rates were compared for these hospital types. Separate comparisons also were performed after stratifying hospitals into three groups defined by community size. The risk-adjusted 30-day mortality per 1,000 patients was 91.5, ranging from 85.4 for private teaching hospitals to 95.3 for nonteaching public hospitals, and 97.4 for osteopathic hospitals. The adjusted 30-to-180-day mortality was 84.7, ranging from 82.6 for nonteaching public hospitals to 87.4 and 88.2, respectively for public teaching and osteopathic hospitals. Differences among hospital types were minimal for small communities and increased with community size. In the large communities, the types of hospitals with high 30-day mortality also had higher mortality after 30 days. There was a strong association of hospital type with adjusted 30-day mortality, which should depend on the quality of hospital care, and a much weaker association with post-30-day mortality, which may be more dependent on patient risk. There was no evidence that types of hospitals with low 30-day mortality were postponing rather than preventing mortality.

Time‐to‐Event Modeling of Competing Risks With Intervening States in Transplantation

The criteria for the selection of who among the persons on the waiting is to receive an organ that has become available and who is to be placed on the list to begin with are the most contentious issues in organ transplantation. The decisions of whom to list and whom to transplant should take into account the net benefit to the individual patient and to the affected group as a whole. We present a method to compute the survival benefit by means of fully parametric modeling of the competing events (transplantation, death while awaiting the transplant, removal for other reasons), taking into account the transplant as an intervening state on the path to death post‐transplant, and apply it to decisions whether to list or not list and whether to transplant or to leave on the waiting list or to remove from the list. The data were obtained from the Organ Procurement and Transplantation Network. They describe the outcomes of listings in January, 1996 through June, 1999, with a follow up of at least 1 year possible for all cases. The models produce estimates of event probabilities that accord well with the observed probabilities and predictions of the survival benefit due to transplantation that range from small negative values to increases in survival probability of 20–40% points in liver and heart transplantation, with the larger benefits generally seen in the more severely ill transplant candidates. These estimates are stable under variations of case mix, as ascertained by bootstrap analysis. The survival benefit of alternative actions can be calculated for the complex circumstances encountered clinically – competing sequential events whose probability evolves over time. The range and stability of the estimates are sufficient to permit the use of this measure to rank candidates for listing and for transplantation.

Severity of illness measures derived from the Uniform Clinical Data Set (UCDSS).

The Health Care Financing Administration (HCFA) plans to use the Uniform Clinical Data Set System (UCDSS) to collect data on hospitalized Medicare patients. This study examined the value of UCDSS data for creating severity of illness measures. UCDSS data were obtained from a study hospital and from a national data set for patients with pneumonia (n = 528) and stroke (n = 565). Models to predict length of stay or an adverse event were derived for each condition using HCFA claims data alone, UCDSS data alone, and UCDSS data supplemented with additional information also abstracted from charts. The models were derived from one set of patients and validated on another. The R2 for predicting length of stay in the validation data for the UCDSS model was 0.29 for pneumonia and 0.19 for stroke compared to R2 values from the claims model of 0.09 for stroke and 0.06 for pneumonia. UCDSS models also were better than claims models for predicting adverse events. The best UCDSS models included International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes and other information requiring clinical judgment, and were improved by adding more information on patient functional status. Some findings were more strongly associated with outcome for the study hospital than for the national data. These results suggest that UCDSS models will predict outcome much better than the claims based models currently used by HCFA for the analysis of hospitalization-related mortality; more functional status information should be added to UCDSS; and despite an extensive objective database, the most predictive UCDSS models require clinician-assigned diagnostic codes.

Beyond Survival: The Burden of Disease in Decision Making in Organ Transplantation

Organ transplantation has long been perceived as a life‐saving intervention. However, it is now recognized that the broader objectives include reducing the patients burden of disease. The components of the burden of disease include (i) mortality, (ii) morbidity, (iii) disability, (iv) psychological distress and (v) resource use. These components may be correlated either positively or negatively, reflecting common disease antecedents or the trade‐offs in clinical decisions. Our proposed approach to modeling includes measures of each outcome and, explicitly, their correlations. Its results are the predicted independent and joint probabilities that a patient will experience any specified range of each measure and of combinations of the measures. The methods were tested with data on mortality, morbidity and resource use from patients following kidney transplantation. The predicted probabilities of the measures are consistent with their observed values and distinguish among the prognoses of patients with distinctive risks, such as diabetics. We have shown, and we believe, that it is feasible to assess the interrelated components of the burden of disease in individual patients and hope that our approach may serve as a starting point in the development of a program for inclusion of alternative measures of outcome in decision making in organ transplantation.

A method for the quantitative analysis and standardization of interleukin-1 bioactivity

A method is presented for the reproducible quantitation of the biological activity of interleukin 1 (IL‐1). This method provides diagnostic tools which give insights into the qualitative aspects of the binding of IL‐1 and of the resulting activation of the responder thymocytes; for example, whether the lymphokine and/or the responder population is heterogeneous, or whether a threshold level exists. It establishes under what circumstances the assumptions on which it is based are reasonably adhered to and, consequently, quantitative estimation in the manner it prescribes is justified. It also gives a simple way to calculate both the maximal response attainable for each preparation in an assay and the dilution of a particular preparation that would produce a half‐maximal response, the accepted unit of activity of IL‐1. This empirical technique provides an improved means of comparing the activities of various preparations of IL‐1 in bioassays using various stocks of responder cells and reagents. It should also be applicable to the evaluation of the biological activity of lymphokines in general.

Labor epidurals improve outcomes for babies of mothers at high risk for unscheduled cesarean section.

CONTEXT: Epidural placement for labor in the general population of laboring women is associated with increased incidence of operative deliveries, prolongation of labor, and may be associated with an increased cesarean section rate. The risks and benefits associated with epidural placement for labor in the subpopulation of mothers at high risk for cesarean section have not been studied.OBJECTIVE: To determine if a population of mothers and babies at high risk for cesarean section will have improved outcomes with labor epidural placement.DESIGN: A decision and cost analysis examining epidural placement for labor on a population of women who are at high risk for unscheduled cesarean section and may benefit from scheduled cesarean section as determined by threshold analysis was performed. Outcomes and probabilities were determined through analysis of the Department of Defenses 1996 National Quality Management Program (NQMP) Birth Product Line data set containing more than 7000 deliveries. Outcomes were defined using variables comprised of all documented conditions that occurred during the peripartum and neonatal hospitalizations. The 1997 NQMP data set was used to validate the results.SETTING: Military Treatment Facilities throughout the United States and abroad and civilian facilities in the United States providing care to military dependents.PATIENT POPULATION: Active duty and dependent pregnant women and babies.RESULTS: About 8% of mothers in this patient population were found to be at high risk for cesarean section. The decision and cost analyses showed that babies of the high risk mothers who received epidurals for labor had better clinical outcomes ( p<0.05) and the procedure was cost neutral ( p=0.23). The procedure did not increase the frequency of cesarean section, and there was no effect on maternal outcomes scores. These results were confirmed by the validation study.CONCLUSIONS: There is a sizable subpopulation of women at high risk for cesarean section whose babies may have better outcomes with epidural placement with no sacrifice in maternal outcomes or costs.

A METHOD FOR CORRECTING FOR THE VARIABILITY OF INHIBITORY EFFECTS OF SOLUBLE HUMAN INTERLEUKIN 1 RECEPTOR II MEASURED BY DIFFERENT ELISAS

Seven ELISAs were developed by using several combinations of anti-human IL-1beta antibodies for detecting interleukin 1beta (IL-1beta) in cell culture supernatants. These ELISAs have different sensitivities in detecting standard preparations of recombinant human IL-1beta (WHO reference standard) compared with conventional preparations of IL-1beta produced by stimulated human peripheral blood mononuclear cells. The observed differences were attributed to differences in epitope specificity of the various monoclonal antibodies used and the heterogeneity of IL-1beta secreted into culture supernatants. The presence of soluble IL-1 receptor type I did not alter the levels of IL-1beta detected by these ELISAs. However, soluble IL-1 receptor type II interfered with the detection of IL-1beta to different degrees in these ELISAs. A method involving standarization by means of separate measurement of the amount of receptor and its inhibitory effect in the IL-1beta ELISA, yields consistent estimates of the correct IL-1beta levels.

Analysis of the heterogeneity of the biological responses to native and mutant human interleukin-6.

The structure‐function relationships of the biological activities of mutant varieties of the pleiotropic cytokine interleukin‐6 (human) were measured by three assays: induction of immunoglobulin M (IgM) secretion from an Epstein‐Barr virus‐transformed human B cell line and induction of fibrinogen secretion from either a human hepatoma cell line or a rat hepatoma cell line. The biological effects of the cytokine were characterized by three parameters as determined by a novel analysis: effectiveness (the maximal response attainable), efficiency (the concentration yielding a half‐maximal response), and complexity (a measure of heterogeneity and feedback control). Substitution of serine for cysteine was associated with a reduction in the effectiveness of interleukin‐6 in both fibrinogen secretion assays. In the assay with human hepatoma cells, there was also a profound reduction in efficiency. Serine substitution in the human IgM synthesis assay appears mainly to reduce the efficiency. Deletion of amino acids 4 to 23 increased the efficiency in the rat hepatoma assay. The complexity parameter suggests the presence of multiple receptor classes or negative feedback in all three assays. Use of the proposed sequential approach to the analysis of dose‐response relations in bio‐ assays provides a more useful quantitative assessment of activities as well as more insight into the complexity of the reactions.

Competency in health care management: a training model in epidemiologic methods for assessing and improving the quality of clinical practice through evidence-based decision making.

This article describes a training model that focuses on health care management by applying epidemiologic methods to assess and improve the quality of clinical practice. The models uniqueness is its focus on integrating clinical evidence-based decision making with fundamental principles of resource management to achieve attainable, cost-effective, high-quality health outcomes. The target students are current and prospective clinical and administrative executives who must optimize decision making at the clinical and managerial levels of health care organizations.