Estimating model-adjusted risks, risk differences, and risk ratios from complex survey data.
ABSTRACT There is increasing interest in estimating and drawing inferences about risk or prevalence ratios and differences instead of odds ratios in the regression setting. Recent publications have shown how the GENMOD procedure in SAS (SAS Institute Inc., Cary, North Carolina) can be used to estimate these parameters in non-population-based studies. In this paper, the authors show how model-adjusted risks, risk differences, and risk ratio estimates can be obtained directly from logistic regression models in the complex sample survey setting to yield population-based inferences. Complex sample survey designs typically involve some combination of weighting, stratification, multistage sampling, clustering, and perhaps finite population adjustments. Point estimates of model-adjusted risks, risk differences, and risk ratios are obtained from average marginal predictions in the fitted logistic regression model. The model can contain both continuous and categorical covariates, as well as interaction terms. The authors use the SUDAAN software package (Research Triangle Institute, Research Triangle Park, North Carolina) to obtain point estimates, standard errors (via linearization or a replication method), confidence intervals, and P values for the parameters and contrasts of interest. Data from the 2006 National Health Interview Survey are used to illustrate these concepts.
SourceAvailable from: Christine Mattson[Show abstract] [Hide abstract]
ABSTRACT: Background: The negative health effects of cigarette smoking and HIV infection are synergistic. Objective: To compare the prevalence of current cigarette smoking and smoking cessation between adults with HIV receiving medical care and adults in the general population. Design: Nationally representative cross-sectional surveys. Setting: United States. Patients: 4217 adults with HIV who participated in the Medical Monitoring Project and 27 731 U.S. adults who participated in the National Health Interview Survey in 2009. Measurements: The main exposure was cigarette smoking. The outcome measures were weighted prevalence of cigarette smoking and quit ratio (ratio of former smokers to the sum of former and current smokers). Results: Of the estimated 419 945 adults with HIV receiving medical care, 42.4% (95% CI, 39.7% to 45.1%) were current cigarette smokers, 20.3% (CI, 18.6% to 22.1%) were former smokers, and 37.3% (CI, 34.9% to 39.6%) had never smoked. Compared with the U.S. adult population, in which an estimated 20.6% of adults smoked cigarettes in 2009, adults with HIV were nearly twice as likely to smoke (adjusted prevalence difference, 17.0 percentage points [CI, 14.0 to 20.1 percentage points]) but were less likely to quit smoking (quit ratio, 32.4% vs. 51.7%). Among adults with HIV, factors independently associated with greater smoking prevalence were older age, non-Hispanic white or non-Hispanic black race, lower educational level, poverty, homelessness, incarceration, substance use, binge alcohol use, depression, and not achieving a suppressed HIV viral load. Limitation: Cross-sectional design with some generalizability limitations. Conclusion: Adults with HIV were more likely to smoke and less likely to quit smoking than the general adult population. Tobacco screening and cessation strategies are important considerations as part of routine HIV care. Primary Funding Source: Centers for Disease Control and Prevention.Annals of internal medicine 03/2015; Vol 162(5):335-344. DOI:10.7326/M14-0954 · 16.10 Impact Factor
[Show abstract] [Hide abstract]
ABSTRACT: Current guidelines recommend that adults with atherosclerotic cardiovascular disease take low-dose aspirin or other antiplatelet medications as secondary prevention of recurrent cardiovascular events. Yet, no national level assessment of low-dose aspirin use for secondary prevention of cardiovascular disease has been reported in a community-based population. Using data from the 2012 National Health Interview Survey, we assessed low-dose aspirin use in those with atherosclerotic cardiovascular disease. We estimated the prevalence ratios of low-dose aspirin use, adjusting for sociodemographic status, health insurance, and cardiovascular risk factors. In those with atherosclerotic cardiovascular disease (n = 3,068), 76% had been instructed to take aspirin and 88% of those were following this advice. Of those not advised, 11% took aspirin on their own. Overall, 70% were taking aspirin (including those who followed their health care provider's advice and those who were not advised but took aspirin on their own). Logistic regression models showed that women, non-Hispanic blacks and Hispanics, those aged 40 to 64 years, with a high school education or with some college, or with fewer cardiovascular disease risk factors were less likely to take aspirin than men, non-Hispanic whites, those aged ≥65 years, with a college education or higher, or with all 4 selected cardiovascular disease risk factors, respectively. Additional analyses conducted in those with coronary heart disease only (n = 2,007) showed similar patterns. In conclusion, use of low-dose aspirin for secondary prevention was 70%, with high reported adherence to health care providers' advice to take low-dose aspirin (88%) and significant variability within subgroups. Copyright © 2015 Elsevier Inc. All rights reserved.The American Journal of Cardiology 01/2015; 115(7). DOI:10.1016/j.amjcard.2015.01.014 · 3.43 Impact Factor
[Show abstract] [Hide abstract]
ABSTRACT: To estimate and evaluate the sensitivity and specificity of providers' diagnosis codes and medication lists to identify outpatient visits by patients with diabetes. We used data from the 2006 to 2010 National Ambulatory Medical Care Survey and National Hospital Ambulatory Medical Care Survey. We assessed the sensitivity and specificity of providers' diagnoses and medication lists to identify patients with diabetes, using the checkbox for diabetes as the gold standard. We then examined differences in sensitivity by patients' characteristics using multivariate logistic regression models. The checkbox identified 12,647 outpatient visits by adults with diabetes among the 70,352 visits used for this analysis. The sensitivity and specificity of providers' diagnoses or listed diabetes medications were 72.3% (95% CI: 70.8% to 73.8%) and 99.2% (99.1% to 99.4%), respectively. Diabetic patients ≥75years of age, women, non-Hispanics, and those with private insurance or Medicare were more likely to be missed by providers' diagnoses and medication lists. Diabetic patients who had more diagnosis codes and medications recorded, had glucose or hemoglobin A1c measured, or made office- rather than hospital-outpatient visits were less likely to be missed. Providers' diagnosis codes and medication lists fail to identify approximately one quarter of outpatient visits by patients with diabetes. Copyright © 2015. Published by Elsevier Inc.Journal of diabetes and its complications 04/2015; DOI:10.1016/j.jdiacomp.2015.03.019 · 1.93 Impact Factor