Faculty Bibliography

BACKGROUND AND OBJECTIVE: For individuals with hypertension, diabetes, or hypercholesterolemia, the relative magnitude of cardiovascular risk factors and the effect of multiple risk factors remains controversial and both treatment practices and health care usage vary. We sought to determine the effect of hypertension, diabetes, hypercholesterolemia, and their combinations on health care utilization and health status through analysis of data from a large national survey. METHODS: We applied the Anderson model to a cross-sectional representative sample (n=15,107) of the U.S. civilian, noninstitutionalized population (the 1996 Medical Expenditure Panel Survey). RESULTS: For diabetes, additional risk factors did not increase the likelihood of emergency room (ER) visits or hospitalizations but were associated with increased outpatient visits and poorer health status. For hypertension, additional risk factors increased the likelihood of hospitalization (but not ER visits), the number of outpatient visits, and poorer health status. For hypercholesterolemia, additional risk factors were associated with increased likelihood of ER visits, hospitalizations, and poorer health status but not more outpatient visits. Diabetes had the largest effect on health care utilization and health status. CONCLUSION: These findings re-emphasize the magnitude of diabetes as a major risk factor associated with increased ER visits, hospitalizations, outpatients visits, and lower health status

We consider generalized linear models with a coarsened covariate. The term 'coarsened' is used here to refer to the case where the exact value of the covariate of interest is not fully observed. Instead, only some set or grouping that contains the exact value is observed. In particular, we propose a likelihood-based method for estimating regression parameters in a generalized linear model relating the mean of the outcome to covariates. We outline Newton-Raphson and EM algorithms for obtaining maximum likelihood estimates of the regression parameters. We also compare and contrast this likelihood-based approach with two somewhat ad hoc procedures: a complete-case analysis in which individuals with coarsened data are excluded and estimation is based on the remaining 'complete cases', and a coarsened data regression model in which the covariate values for all the complete cases are coarsened and then included in a regression model relating the mean to the coarsened covariate. The methodology that is presented is motivated by coarsened data on the racial-ethnicity categorization of patients in the US's National Ambulatory Medical Care Survey, a study to examine the medical care that is provided to a patient in a physician's office. In this study, the outcome of interest is the level of tests (none, non-invasive tests or invasive tests) ordered for the patient at the doctor's visit. One of the covariates of interest is the patient's four-level discrete covariate comprised of four racial-ethnicity categories: white-Hispanic, white-non-Hispanic, African-American-Hispanic and African-American-non-Hispanic. However, of the 19 095 patients in the sample, 14 955 (or 78%) have the exact category of race-ethnicity recorded and 4140 (or 22%) have race-ethnicity coarsened. For the latter group of 4140 individuals, the ethnicity is not recorded, but we know that 3683 are white and 457 are African-American

BACKGROUND: The sex-specific independent effect of diabetes mellitus and established coronary heart disease (CHD) on subsequent CHD mortality is not known. METHODS: This is an analysis of pooled data (n = 5243) from the Framingham Heart Study and the Framingham Offspring Study with follow-up of 20 years. At baseline (1971-1975), 134 men and 95 women had diabetes, while 222 men and 129 women had CHD. Risk for CHD death was analyzed by proportional hazards models, adjusting for age, hypertension, serum cholesterol levels, smoking, and body mass index. The comparative effect of established CHD vs diabetes on the risk of CHD mortality was tested by testing the difference in log hazards. RESULTS: The adjusted hazard ratios (HRs) with 95% confidence intervals (CIs) for death from CHD were 2.1 (95% CI, 1.3-3.3) in men with diabetes only, and 4.2 (95% CI, 3.2-5.6) in men with CHD only compared with men without diabetes or CHD. The HR for CHD death was 3.8 (95% CI, 2.2-6.6) in women with diabetes, and 1.9 (95% CI, 1.1-3.4) in women with CHD. The difference between the CHD and the diabetes log hazards was +0.73 (95% CI, 0.72-0.75) in men and -0.65 (95% CI, -0.68 to -0.63) in women. CONCLUSIONS: In men, established CHD signifies a higher risk for CHD mortality than diabetes. This is reversed in women, with diabetes being associated with greater risk for CHD mortality. Current treatment recommendations for women with diabetes may need to be more aggressive to match CHD mortality risk

Low-density lipoprotein (LDL)-based guidelines are currently used to initiate and monitor cholesterol-lowering therapy.Using stratified analyses, data from the Framingham Heart Study and the Coronary Primary Prevention Trial were evaluated to determine whether (1) cholesterol levels (total cholesterol [TC] or LDL [low-density lipoprotein]) better discriminated risk for coronary heart disease (CHD) than cholesterol ratios (LDL/HDL [high-density lipoprotein] or TC/HDL); and (2) whether changes in ratios better predicted risk reduction than changes in levels.Individuals with similar LDL/HDL ratios had similar risks for CHD regardless of whether they had high LDL levels or low LDL levels (23% vs 23% for the CPPT, 13.8% vs 14% for FHS men, and 8.6% vs 10.9% for FHS women). Among men with similar initial LDL/HDL ratios and similar changes in LDL/HDL ratios, risks for CHD did not differ (20.3% compared with 21.0%; p =0.96) between those with the largest and smallest reductions in LDL levels (21.3% compared with 6.5%). Among men with similar initial LDL levels and similar LDL reductions, a 20% reduction in risk for CHD was seen (19.5% compared with 24.5%; p =0.005) between those with the largest and smallest reductions in LDL/HDL ratios (23% compared with 4.6%). TC/HDL had predictive ability similar to LDL/HDL.Cholesterol levels do not provide incremental predictive value over cholesterol ratios in identifying people at risk for CHD. Changes in ratios are better predictors of successful CHD risk reduction than changes in levels. Future guidelines should consider incorporating ratios in initiating and monitoring successful lipid-lowering therapy

BACKGROUND: Few studies have documented national trends in screening, awareness, and treatment of cardiovascular risk factors. We evaluated trends in screening, prevalence, and treatment of hypertension, hypercholesterolemia, and smoking. METHODS: Data were analyzed from the 1984-1998 Behavioral Risk Factor Surveillance System, a series of yearly cross-sectional population-based surveys of U.S. adults. Unadjusted and adjusted time trends (age-, gender-, ethnicity-, education-, and income-adjusted) in screening, prevalence, and treatment were evaluated. RESULTS: From 1984 to 1998, a larger proportion of U.S. adults were older, more educated, richer, and Hispanic. Hypertension screening was >97% (1988-1998), prevalence ranged from 21 to 24% (1984-1998), and approximately 58% (1984-1992) were prescribed blood-pressure-lowering medications. Hypercholesterolemia screening increased from 47 to 67% (1987-1998), prevalence from 18 to 31% (1987-1998), and cholesterol-lowering prescriptions from 22 to 25% (1988-1990). Smoking prevalence remained around 28% (1984-1998), while quit attempts declined from 63 to 47% (1990-1998). CONCLUSIONS: Although screening for hypertension and hypercholesterolemia has increased, a substantial proportion of cases were not being prescribed medications. While the prevalence of smoking remains constant, quit attempts have fallen. Continuing challenges for cardiovascular disease prevention include identification of individuals with hypercholesterolemia, appropriate prescription (initiation and/or maintenance) of antihypertensive and lipid-lowering medications, and intensifying smoking cessation efforts

OBJECTIVES: To examine the generalisability of multivariate risk functions from diverse populations in three contexts: ordering risk, magnitude of relative risks, and estimation of absolute risk. DESIGN: Meta-analysis of prospective cohort studies. PATIENTS: Participants from various epidemiological studies. MAIN OUTCOME MEASURE: Death from coronary heart disease (CHD). RESULTS: The analysis included 105 420 men and 56 535 women 35-74 years of age and free of CHD at baseline from 16 observational studies with a total of 27 analytical groups. The area under the receiver operating characteristic curve (AUC) was used to judge the ability of the multivariate risk function to order risk correctly. AUCs ranged from 0.60 to 0.80. The AUCs differed significantly between the studies (p < 0.01) but were very similar for different risk functions applied to the same population, indicating similar ability to rank risk for different models. The magnitudes of the relative risks associated with major risk factors (age, systolic blood pressure, serum total cholesterol, smoking, and diabetes) varied significantly across studies (p < 0.05 for homogeneity). The prediction of absolute risk was not very accurate in most of the cases when a model derived from one study was applied to a different study. CONCLUSIONS: When considered qualitatively, the major risk factors are associated with CHD mortality in a diverse set of populations. However, when considered quantitatively, there was significant heterogeneity in all three aspects: ordering risk, magnitude of relative risks, and estimation of absolute risk

A study of 74 patients in a clinic serving indigent primarily African American people found that most were in early stages of readiness to start exercising, and that HbA1c levels were lower in the later stages

BACKGROUND: Hypercholesterolemia is a major cardiovascular risk factor, and cholesterol awareness is important in both clinical practice and in public health. We evaluated the validity of self-reported hypercholesterolemia and identified determinants of validity. METHODS: The study design was a cross-sectional survey, from 1988 to 1994, of adult participants (N=8236) from the Third National Health and Nutrition Examination Survey for whom self-report of hypercholesterolemia and serum measurement were available. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) for self-reported hypercholesterolemia were calculated using total cholesterol > or =5.17 mmol/L (200 mg/dL) and/or taking cholesterol-lowering medication as the criterion standard. RESULTS: Overall test characteristics for self-report were sensitivity, 51%; specificity, 89%; PPV, 87%; and NPV, 55%. Sensitivity of self-report was higher among older subjects and non-Hispanic whites, specificity was higher among subjects with >12 years of education, PPV was higher in older subjects, and NPV was higher in younger subjects and in those with >12 years of education. Using higher cholesterol thresholds to define hypercholesterolemia led to higher sensitivity, lower specificity, lower PPV, and higher NPV. Sociodemographic and anthropometric predictors of validity were identified by logistic regression. CONCLUSIONS: Due to low sensitivity, self-reported hypercholesterolemia should be used with caution, both during the patient encounter and for surveillance of trends in hypercholesterolemia in the absence of measured cholesterol levels. Specificity is consistently much higher than sensitivity. The high PPV may be of use in certain clinical situations. Such validation studies should form the foundation for future research based on self-report