Use of survey and clinical data for screening and diagnosis.

BACKGROUND: U.S. prevalence of diabetes in 2005 was 20.8 million people (6.2 million undiagnosed) (Diabetes Statistics. 2006). Recognizable preclinical stage is between 7 and 12 years. Efficient screening and early diagnosis can help: 1. avoid or delay development of diabetes. 2. treat early and avoid co-morbidities. DESIGN: Retrospective cross-sectional study of 153 113 adults ages 24 to 83 from the Behavioural Risk Factor Surveillance Systems (BRFSS)-2003 of whom 4379 had diabetes at their current age or during the previous year and 2190 adults ages 40 to 74 from the National Health and Nutrition Examination Survey III, 211 of whom had glucose tolerance test result > or = 200 mg/dL. OBJECTIVES: To develop statistical models for screening and diagnosis. METHODS: Logistic and generalized linear and additive regression models, Akaike information criterion, area under the receiver operating characteristic (ROC) curve, Baye's rule. RESULTS: Area under the 'productivity' curve using BRFSS data is 0.65 indicating an average yield of 17.3 per cent. Survey data is useful also for diagnosis. Area under the ROC curve (AUC) using only survey data is 0.68. AUC for fasting plasma glucose (FPG) alone is 0.91. Stepwise drop one method of selecting co-variates for diagnosis included pre-test probability from BRFSS. When both FPG and pre-test information are included, AUC increases to 0.93. Reduction in residual deviance and 0.02 increase in AUC are statistically significant (p = 0.0012). Clinical significance of prior odds is shown by increase in accuracy and weighted average of sensitivity and specificity. Empirically estimated regression weights for pre-test and test information vary with age and race and are not equal, as required by Baye's theorem. Overfitting index was less than 1 per cent. CONCLUSIONS: Cross-section surveys are useful for increasing screening efficiency and diagnostic accuracy.

A cost-minimizing diagnostic methodology for discrimination between neoplastic and non-neoplastic brain lesions: utilizing a genetic algorithm.

RATIONAL AND OBJECTIVES: The purpose of this study was to make an improvement in the performance of a logistic regression model in predicting the presence of brain neoplasia with magnetic resonance spectroscopy data by using a new approach for logistic regression coefficient estimation. This new approach, termed cost minimizing (C-min), introduced by one of the authors (Chetty), uses the cost function for prediction outcomes to estimate model coefficients and the prediction decision rule. To do this requires use of a genetic algorithm. MATERIALS AND METHODS: Consecutive patients with suspected brain neoplasms or recurrent neoplasia referred for magnetic resonance spectroscopy were enrolled once a final diagnosis was established by histopathology or clinical course, laboratory data, and serial imaging. For the same magnetic resonance spectroscopy explanatory (input) variables, logistic regression models were constructed with conventional and C-min coefficient estimates, and sensitivity and specificity outcomes were compared at alternative probability threshold levels. RESULTS: The C-min approach dominated the conventional approach in 14 of 18 trials, in that C-min had either fewer of both false negatives and false positives, or it had the same number of one type, and less of the other type of diagnostic error. C-min was always less costly. CONCLUSION: The C-min approach to logistic or other regression model estimation may be a step forward in reducing the cost and, often, the errors of diagnostic (and treatment) processes. However, this new approach must be validated on larger and more varied datasets, and its statistical performance characteristics determined before it can be implemented as a practical clinical tool.

Medicare costs in urban areas and the supply of primary care physicians.

BACKGROUND: The supply of primary care physicians may be important determinants of health care costs. We examined the association between primary care physician supply and geographic location with respect to variation in Medicare Supplementary Medical Insurance (Part B) reimbursement. METHODS: We performed an analysis of data from all US metropolitan counties. Physician supply data were derived from the American Medical Association Masterfile. Medicare Part B reimbursements and enrollment data came from the Health Care Financing Administration. Physician supply was calculated for family practice, general internal medicine, and non-primary care specialties. Linear regression was used to test the association of physician supply and Medicare costs and to adjust for potential confounding variables. RESULTS: The average Medicare Part B reimbursement per enrollee was $1283. After adjusting for local price differences and county characteristics, a greater supply of family physicians and general internists was significantly associated with lower Medicare Part B reimbursements. The reduction in reimbursements between counties in the highest quintile of family physician supply and the lowest quintile was $261 per enrollee. In contrast, a greater supply of general practitioners and non-primary care physicians was associated with higher reimbursements per enrollee. CONCLUSIONS: These results add to the evidence than an increased supply of primary care physicians is associated with lower health care costs. If this association is causal, it supports the theory that increasing the number of primary care physicians may lower health care costs.

Primary care physicians in underserved areas. Family physicians dominate.

Using the definitions of "medically underserved areas" developed by the California Health Manpower Policy Commission and data on physician location derived from a survey of California physicians applying for licensure or relicensure between 1984 and 1986, we examined the extent to which different kinds of primary care physicians located in underserved areas. Among physicians completing postgraduate medical education after 1974, board-certified family physicians were 3 times more likely to locate in medically underserved rural communities than were other primary care physicians. Non-board-certified family and general physicians were 1.6 times more likely than other non-board-certified primary care physicians to locate in rural underserved areas. Family and general practice physicians also showed a slightly greater likelihood than other primary care physicians of being located in urban underserved areas.

Hubble space telescope observations of comet p/shoemaker-levy 9 (1993e).

The Hubble Space Telescope observed the fragmented comet P/Shoemaker-Levy 9 (1993e) (P indicates that it is a periodic comet) on 1 July 1993. Approximately 20 individual nuclei and their comae were observed in images taken with the Planetary Camera. After subtraction of the comae light, the 11 brightest nuclei have magnitudes between approximately 23.7 and 24.8. Assuming that the geometric albedo is 0.04, these magnitudes imply that the nuclear diameters are in the range approximately 2.5 to 4.3 kilometers. If the density of each nucleus is 1 gram per cubic centimeter, the total energy deposited by the impact of these 11 nuclei into Jupiter's atmosphere next July will be approximately 4 x 10(30) ergs ( approximately 10(8) megatons of TNT). This latter number should be regarded as an upper limit because the nuclear magnitudes probably contain a small residual coma contribution. The Faint Object Spectrograph was used to search for fluorescence from OH, which is usually an excellent indicator of cometary activity. No OH emission was detected, and this can be translated into an upper limit on the water production rate of approximately 2 x 10(27) molecules per second.

A study of Minnesota's high-risk health insurance pool.

This is a report of a study of Minnesota's high-risk health insurance pool for "medically uninsurable" persons. The study consisted of a survey of current and past enrollees carried out in the Spring of 1990 and an analysis of the claims and membership files for 1988 and 1989. The main policy conclusion we reached is that Minnesota's high-risk pool is an adequate approach to the problem raised by risk segmentation on the basis of health status, providing that enrollment remains a small fraction of the population. The recent high, enrollment growth rates the Minnesota risk pool has experienced raise the possibility that basic structural reforms of the nongroup and small-group health insurance markets are needed.

Asteroid families: observational evidence for common origins.

Colors of minor planets in the UBV system indicate compositions quite distinct from those of the field population in each of three Hirayama families. The Eos and Koronis families apparently originated from the collisional fragmentation of undifferentiated silicate bodies, and the Nysa group from a geochemically differentiated parent body.