Long-term Effects of Carotid Screening on Patient Outcomes and Behaviors

There have been conflicting reports regarding the effects of carotid ultrasonographic screening (CUS) on health-related behaviors^1-3; however, long-term studies have not evaluated the effects of CUS on clinical outcomes.⁴ This study's objectives were to determine if patients with advanced subclinical atherosclerosis (AdvAthero) on CUS are more likely to achieve guideline-based risk factor goals and improve their long-term cardiovascular disease (CVD) health-related behaviors.

This study was approved by our institutional review board. Potential subjects were 40- to 70-year-old patients with 1 or more risk factor and no history of CVD, who were referred by their physician to the University of Wisconsin Vascular Health Screening Program for CUS from January 2002 through December 2006. We performed CUS using the carotid intima-media thickness (CIMT) protocol from the Atherosclerosis Risk in Communities Study.⁵ AdvAthero was defined as the presence of CIMT greater than 75th percentile for age, sex, and race, or carotid plaque presence.^4,5 All subjects had a primary care provider in the University of Wisconsin Health medical practice and insurance coverage for CUS. Qualifying patients were invited to participate by mail. To reduce selection bias, a waiver of consent for electronic medical record review was obtained for subjects who could not be contacted.

A patient questionnaire evaluated health-related behaviors and adherence to lifestyle and pharmacotherapy recommendations. The electronic medical record was reviewed for achievement of low-density lipoprotein cholesterol (LDL-C)⁶ and systolic blood pressure (SBP)⁷ goals. Multivariable logistic regression models evaluated predictors of correct recall of CUS results and recommendations and achievement of LDL-C and SBP goals. For each outcome, a multivariable model was created that included age, sex, and presence of AdvAthero. Baseline variables were added to each model, including LDL-C level, high-density lipoprotein cholesterol level, SBP, Framingham Risk Score, family history of premature CVD, body mass index, hypertension history, dyslipidemia history, educational level, and length of follow-up.

Of 1165 patients who had CUS examinations, 602 met inclusion criteria and were invited to participate; only 73 declined. Of the 529 subjects (Table), electronic medical records were abstracted for 100% and questionnaires were returned by 59.6%. AdvAthero was identified in 58.6%.

One year after CUS, LDL-C goal achievement increased from 62.1% to 84.1% (P < .001) in all subjects. Among subjects with AdvAthero, only 61.2% were at the LDL-C goal at baseline, but 86.7% were at goal 1 year after CUS (P < .001). Similarly, among subjects without AdvAthero, 63.2% were at the LDL-C goal at baseline, whereas 80.2% were at goal after CUS (P < .001). The presence of AdvAthero (odds ratio [OR], 2.15 [95% confidence interval {CI},1.38-3.34]; P < .001), but not baseline LDL-C level (P = .99), independently predicted prescription of lipid-lowering medication after CUS, but the interaction between prescription of lipid-lowering therapy and presence of AdvAthero on achieving the LDL-C goal was not significant (P value for interaction, .10). Thus, having AdvAthero on CUS did not influence the use of lipid-lowering therapy to achieve LDL-C goals.

One year after CUS, subjects with AdvAthero had a lower mean LDL-C level (Δ = 12 mg/dL; P < .001) compared with subjects with normal scan results; however, 32.7% of subjects that started a lipid-lowering medication after CUS did not have AdvAthero. There was no significant change in SBP. AdvAthero (P = .81) did not predict SBP goal achievement or prescription of antihypertensives (P = .20). AdvAthero predicted prescription of aspirin (OR, 2.25 [95% CI, 1.43-3.54]; P < .001).

Survey follow-up was a mean (SD) 55.6 (19.1) months after CUS. AdvAthero presence did not increase subjects' perceived risk of current (P = .15) or future (P = .21) CVD. Subjects accurately recalled 83.0% of prevention recommendations; however, the AdvAthero presence only modestly predicted correct recall of CUS results (OR, 3.20 [95% CI, 0.01-2.31]; P = .047). With longer follow-up after CUS, subjects were less likely to remember their results (P = .01), or post-CUS dietary (P = .005) and exercise (P = .008) recommendations. AdvAthero did not predict changes in diet, exercise frequency, or long-term health-related behaviors.

To our knowledge, this is the first study to evaluate long-term clinical outcomes after patients undergo CUS for measurement of CIMT and carotid plaque detection. Major limitations include retrospective electronic medical record abstraction, use of questionnaires that relied on recall and self-report, and referral bias. However, all subjects were part of an academic health network and had medical insurance coverage that paid for the test. AdvAthero did not predict LDL-C goal achievement. Nearly one-third of subjects prescribed lipid-lowering therapy after CUS did not have AdvAthero. This suggests that achievement of LDL-C goals after CUS was not primarily because of abnormal screening results. It may be that physicians who ordered the test were predisposed to prescribe lipid-lowering medication, regardless of the results, and LDL-C improvements were from the screening and management process and not the results. Over time, CVD risk perception among those with AdvAthero decreased. Despite intensive counseling about their CUS results and recommendations, our subjects had inappropriate risk perception, decay in recall over time, and poor adoption of lifestyle changes. These observations support the importance of recurring CVD risk education and counseling and emphasize the lack of long-term effectiveness of 1-time interventions, even if they are as powerful as arterial imaging.⁸

Correspondence: Dr Stein, Division of Cardiovascular Medicine, Atherosclerosis Imaging Research Program, University of Wisconsin School of Medicine and Public Health, 600 Highland Ave, G7/341 CSC (MC 3248), Madison, WI 53792 (jhs@medicine.wisc.edu).

Author Contributions:Study concept and design: Johnson and Stein. Acquisition of data: Einerson, Korcarz, and Aeschlimann. Analysis and interpretation of data: Johnson and Stein. Drafting of the manuscript: Johnson and Stein. Critical revision of the manuscript for important intellectual content: Einerson, Korcarz, Aeschlimann, and Stein. Obtained funding: Stein. Administrative, technical, and material support: Einerson, Korcarz, Aeschlimann, and Stein. Study supervision: Johnson and Stein.

Financial Disclosure: Dr Stein has served as a principal investigator on research grants to University of Wisconsin (Astra-Zeneca, PreMD, Sanofi-Aventis, Siemens Medical Solutions, and Sonosite), a consultant to PreMD, a member of data monitoring committees (Abbott, Lilly, and Takeda), and is an inventor of patent #US 67300235 “Ultrasonic Apparatus and Method for Providing Quantitative Indication of Risk of Coronary Heart Disease.” This patent has been assigned to the Wisconsin Alumni Research Foundation, from which Dr Stein receives royalties. Dr Korcarz previously served as a consultant to Siemens Medical Solutions.

Funding/Support: This study was funded in part by a grant from the University of Wisconsin Foundation from the Madison Community Foundation Tommy Bartlett Medical Research Field of Interest Fund. This study also was funded in part by Siemens Medical Solutions, Inc. Dr Johnson was supported by the Ruth L. Kirschstein National Research Service Award T32 HL07936 from the National Heart Lung and Blood Institute.

Role of the Sponsors: The sponsors had no role in the design, collection, or analysis of the data and did not contribute to the content of this manuscript.

Additional Contributions: We are grateful to the patients who participated in this study. Karen Moncher, MD, assisted with reading some of the Vascular Health Screening examinations.