Advertisement

Screening for Abdominal Aortic Aneurysms in Outpatient Primary Care Clinics

Published:November 13, 2014DOI:https://doi.org/10.1016/j.amjmed.2014.10.036

      Abstract

      Background

      The US Preventive Services Task Force (USPSTF) guidelines recommend one-time abdominal aortic aneurysm ultrasound screening for men aged 65 to 75 years who ever smoked. Reported screening rates have been 13% to 26% but did not include computed tomography, magnetic resonance imaging, and nonaortic abdominal ultrasound, which provide adequate visualization of the aorta. The objective of this study was to evaluate rates of screening performed intentionally with ultrasound and incidentally with other abdominal imaging, determine rates of redundant screening, and evaluate patient and physician characteristics associated with screening.

      Methods

      Cross-sectional study of patient encounters in 2007 and 2012 to determine abdominal aortic aneurysm screening trends in primary care practices. Participants included all patients who were seen in a primary care office and were eligible for screening by USPSTF guidelines. The primary outcome was percentage of eligible patients screened for abdominal aortic aneurysm by ultrasound or other abdominal imaging.

      Results

      There were 15,120 patients eligible for screening in 2007, and 22,355 in 2012. Screening with ultrasounds increased from 3.6% in 2007 to 9.2% in 2012. Screening with any imaging that included the aorta increased from 31% in 2007 to 41% in 2012. Of 2595 screening ultrasounds performed in either cohort, 800 (31%) were performed on patients who had already undergone another imaging modality. Of 153 physicians who had >50 eligible patients, rates of abdominal aortic aneurysm screening ranged from 7.5% to 79% (median 39%, interquartile range 31%-47%), and rates of ultrasound screening ranged from 0% to 47% (median 6.3%, interquartile range 3.6%-11.4%). Physician characteristics positively associated with screened patients included female sex (odds ratio [OR] 1.32; 95% confidence interval [CI], 1.12-1.54), specialty (Internal Medicine vs Family Medicine: OR 1.32; 95% CI, 1.14-1.54), and location (academic medical center vs family health center: OR 1.30; 95% CI, 1.04-1.62).

      Conclusions

      Abdominal aortic aneurysm screening rates remain below 50%, but are improving over time. Screening by individual physicians varied widely, indicating substantial opportunity for educational interventions. Most abdominal aortic aneurysm screening is completed incidentally, and some patients later undergo unnecessary ultrasound screening. Before ordering screening, physicians and electronic health record-based reminder tools should ensure that the aorta has not been previously visualized.

      Keywords

      Clinical Significance
      • Seven years after publication of the US Preventive Services Task Force guidelines, abdominal aortic aneurysm screening rates remain low (9.2% in 2012), although they are higher when considering all studies that visualize the aorta (41.2%).
      • Rates are increasing with time. Thirty-one percent of patients who were screened had already had a study that would have visualized this finding.
      • Repeat screening may become a problem as organizations implement information technology solutions that prompt physicians to screen.
      Abdominal aortic aneurysm affects 1.4% of the US population aged 50-84 years, with 9000-14,000 deaths attributed annually.
      • Kent K.C.
      • Zwolak R.M.
      • Jaff M.R.
      • et al.
      Screening for abdominal aortic aneurysm: a consensus statement.
      • Kent K.C.
      • Zwolak R.M.
      • Egorova N.N.
      • et al.
      Analysis of risk factors for abdominal aortic aneurysm in a cohort of more than 3 million individuals.
      Once an abdominal aortic aneurysm ruptures, estimated mortality is 80%.
      • Kuivaniemi H.
      • Platsoucas C.D.
      • Tilson M.D.
      Aortic aneurysms: an immune disease with a strong genetic component.
      Therefore, management strategies focus on early detection through screening, which has been shown to reduce mortality from abdominal aortic aneurysm rupture.
      • Thompson S.G.
      • Ashton H.A.
      • Gao L.
      • Buxton M.J.
      • Scott R.A.
      Final follow-up of the Multicentre Aneurysm Screening Study (MASS) randomized trial of abdominal aortic aneurysm screening.
      In 2005, the US Preventive Services Task Force (USPSTF) recommended one-time abdominal aortic aneurysm ultrasound screening for all men aged 65 to 75 years who have ever smoked, made no recommendation for men aged 65 to 75 who have never smoked, and recommended against screening women for abdominal aortic aneurysm.
      U.S. Preventive Services Task Force
      Screening for abdominal aortic aneurysm: recommendation statement.
      The preferred screening test for abdominal aortic aneurysm is an abominal ultrasound, which is 95% sensitive and nearly 100% specific for detecting abdominal aortic aneurysm in asymptomatic patients.
      • Fleming C.
      • Whitlock E.P.
      • Beil T.L.
      • Lederle F.A.
      Screening for abdominal aortic aneurysm: a best-evidence systematic review for the U.S. Preventive Services Task Force.
      • Wilmink A.B.
      • Hubbard C.S.
      • Quick C.R.
      Quality of the measurement of the infrarenal aortic diameter by ultrasound.
      • Wilmink A.B.
      • Forshaw M.
      • Quick C.R.
      • Hubbard C.S.
      • Day N.E.
      Accuracy of serial screening for abdominal aortic aneurysms by ultrasound.
      • Barnett S.B.
      • Terhaar G.R.
      • Ziskin M.C.
      • Rott H.D.
      • Duck F.A.
      • Maeda K.
      International recommendations and guidelines for the safe use of diagnostic ultrasound in medicine.
      • Lederle F.A.
      • Wilson S.E.
      • Johnson G.R.
      • et al.
      Variability in measurement of abdominal aortic aneurysms. Abdominal Aortic Aneurysm Detection and Management Veterans Administration Cooperative Study Group.
      Despite insurance coverage for screening under the Screening Abdominal Aortic Aneurysms Very Efficiently (SAAAVE) Act of 2006, observed rates of ultrasound screening range from 8.2% to 12.9%.

      109th US Congress. Deficit Reduction Act of 2005. Available at: http://www.gpo.gov/fdsys/pkg/PLAW-109publ171/html/PLAW-109publ171.htm. Accessed June 11, 2014.

      • Eaton J.
      • Reed D.
      • Angstman K.B.
      • et al.
      Effect of visit length and a clinical decision support tool on abdominal aortic aneurysm screening rates in a primary care practice.
      • Federman D.G.
      • Carbone V.G.
      • Kravetz J.D.
      • Kancir S.
      • Kirsner R.S.
      • Bravata D.M.
      Are screening guidelines for abdominal aortic aneurysms being implemented within a large VA primary health care system?.
      Potential explanations for low screening rates include lack of familiarity with abdominal aortic aneurysm screening and the large number of other screening examinations recommended to this same cohort of patients. As a result, physicians or patients may prioritize other screening examinations (eg, colorectal cancer screening) over screening for abdominal aortic aneurysm.
      Although ultrasound is the recommended screening test, a number of other imaging modalities, including computed tomography (CT) and magnetic resonance imaging (MRI), can adequately assess the aorta for abdominal aortic aneurysm. If the definition of screening includes imaging of the abdomen performed for any indication, screening rates may be higher than previously reported. Moreover, some patients may be screened intentionally with ultrasound after having already had their aorta imaged previously.
      Knowledge of both physician and patient characteristics associated with screening may assist in targeted approaches to increase screening and decrease duplicate testing. The objective of this study was to determine trends in abdominal aortic aneurysm screening at a large integrated health care system and to identify patient and physician characteristics associated with abdominal aortic aneurysm screening.

      Methods

      We performed a retrospective chart review of all patients who had visited a primary care physician at the Cleveland Clinic main campus and family health centers in northeast Ohio (Internal Medicine or Family Medicine) in the calendar years of 2007 (Cohort 1) or 2012 (Cohort 2) and met screening criteria by the 2005 USPSTF guidelines. This included all men age 65-75 years that had ever smoked. At the time of this study, the Cleveland Clinic did not employ any electronic health record-based reminder system for abdominal aortic aneurysm screening. We queried the electronic health record for the following data to assess patient factors associated with screening: age, smoking status (current or former), number of physician visits during the calendar year, record of a periodic health examination in the calendar year, family history of abdominal aortic aneurysm, and other age-appropriate preventive care including lipid panel within 5 years of visit, colonoscopy within 10 years of visit, and pneumococcal vaccine. We also recorded the following physician-level data: years in practice, sex, practice setting (academic or community), and specialty (Internal or Family Medicine). All radiographic studies performed after the age of 50 years were noted. The primary outcome was completion of abdominal aortic aneurysm screening by the end of the calendar year in 2007 and 2012 in order to assess trends in screening over time. Patients were considered to have been screened if they underwent a CT scan of the abdomen or pelvis, MRI study of the abdomen or pelvis, or a renal or mesenteric vascular ultrasound, all of which visualize the aorta. The secondary outcome was completion of abdominal aortic aneurysm screening with a designated screening ultrasound. Redundant screening was defined as screening with ultrasound subsequent to another abdominal imaging test.

       Statistical Analysis

      Screening rates were summarized as proportions along with 95% confidence intervals overall and for each cohort. Patient characteristics were summarized as frequencies and percentages by screening status, and the chi-squared test was used by different screening groups to compare these characteristics. For physician characteristics, analysis was limited to physicians with >50 eligible patients in the data set.
      To determine the relationship between screening and physician-level or patient-level characteristics, the generalized linear mixed-effects models were used. Each model contained one characteristic as a covariate and random intercept to account for potential correlation among data from patients seen by the same provider. All analyses were conducted in SAS 9.2 (SAS Institute Inc, Cary, NC) and statistical significance was established with a 2-sided P-value <.05.

      Results

      There were 15,120 patients who met at least one of the USPSTF criteria for screening in 2007, and 22,355 in 2012 (Table 1). Over this period, the proportion of eligible patients who had been screened increased from 30.7% in 2007 to 41.5% in 2012 (Figure 1). Most screening took place inadvertently via CT or MRI performed for other reasons. The proportion screened by ultrasound increased from 3.6% in 2007 to 9.2% in 2012. Of 2595 patients who had a screening ultrasound in 2007 and 2012, 800 (31%) had previously been screened inadvertently using some other modality.
      Table 1Patient Characteristics By 2007 and 2012 Cohort
      Patients Who Met USPSTF Criteria in 2007 (n = 15,120)

      n (%)
      Patients Who Met USPSTF Criteria in 2012 (n = 22,355)

      n (%)
      Current smoker2221 (14.7%)3660 (16.4%)
      Former smoker12,899 (85.3%)18,695 (83.6%)
      Family history of AAA32 (0.2%)134 (0.6%)
      0 PCP visit5947 (39.3%)6338 (28.4%)
      1-3 PCP visits in year4066 (26.9%)7621 (34.1%)
      4-5 PCP visits in year2105 (13.9%)3651 (16.3%)
      >6 PCP visits in year3002 (19.9%)4745 (21.2%)
      Periodic health examination in year2535 (16.8%)4608 (20.6%)
      Up to date on pneumonia vaccine7828 (51.8%)15,034 (67.3%)
      Up to date on lipid panel11,472 (75.9%)17,723 (79.3%)
      Up to date on colonoscopy4875 (32.3%)7978 (35.7%)
      Up to date on all 3 health maintenance3690 (24.4%)6409 (28.7%)
      AAA = abdominal aortic aneurysm; PCP = primary care physician; USPSTF = United States Preventive Services Task Force.
      Figure thumbnail gr1
      Figure 1Proportion of patients screened for abdominal aortic aneurysm by year (initial screening modality). CT = computed tomography; MRI = magnetic resonance imaging.
      Patient factors that were associated with screening appear in Table 2. Ultrasound screening was positively associated with having a family history of abdominal aortic aneurysm (odds ratio [OR] 2.46; 95% confidence interval [CI], 1.55-3.90), increasing age (OR per year 1.07; 95% CI, 1.06-1.09) and being up to date on the other health maintenance measures assessed (OR 1.93; 95% CI, 1.75-2.13). Having ≥4 primary care provider visits per year (OR 1.36; 95% CI, 1.23-1.50) and having a periodic health examination in the year (OR 1.49; 95% CI, 1.34-1.65) were associated positively with screening. Former smokers were less likely to be screened than current smokers (OR 0.88; 95% CI, 0.77–0.99).
      Table 2Patient Characteristics By Ultrasound Screening Status in 2012
      Screened

      (n = 2056, 9.2%)
      Not Screened

      (n = 20,299, 90.8%)
      P Value
      Smoking status
       Current338 (9.3%)3299 (90.7%).162
       Former1711 (9.2%)16,941 (90.8%)
       Passive6 (18.8%)26 (81.3%)
      Family history of AAA28 (20.9%)106 (79.1%)<.001
      Up to date on all 3 health maintenance markers1081 (16.9%)5328 (83.1%)<.001
      PHE in index year679 (14.7%)3929 (85.3%)<.001
      All characteristics are significantly different between screened and unscreened patients (P < .05).
      AAA = abdominal aortic aneurysm.
      For physicians with at least 50 eligible patients (n = 153), rates of screening by individual physicians using any modality ranged from 7.5%-79% (interquartile range 31%-47%, median = 39%). Rates of screening with ultrasound only ranged from 0%-47% (interquartile range 3.6%-11.4%, median = 6.3%) (Figure 2). Physician factors associated with screening by any modality included female sex (OR 1.31; 95% CI, 1.13-1.53), specialty (Internal Medicine vs Family Medicine, OR 1.32; 95% CI, 1.13-1.54), and practice location (main campus academic practice vs family health center community practice, OR 1.30; 95% CI, 1.04-1.63). For screening specifically by ultrasound, physician degree (MD vs DO; OR 2.70; 95% CI, 2.21-3.31), physician sex (female vs male; OR 1.33; 95% CI, 1.02-1.74), and physician specialty (Internal Medicine vs Family Medicine; OR 1.36; 95% CI, 1.03–1.79) were significantly associated with patient screening. Physicians who performed more health maintenance (screening for lipids and colon cancer and giving pneumococcal vaccination) also had higher rates of abdominal aortic aneurysm screening (Figure 3).
      Figure thumbnail gr2
      Figure 2Rates of ultrasound screening for abdominal aortic aneurysm by individual physicians. US = ultrasound.
      Figure thumbnail gr3
      Figure 3Physician abdominal aortic aneurysm screening rates in comparison with screening rates for other health maintenance markers.

      Discussion

      In this large retrospective cohort study conducted in an integrated health care system, we observed that 7 years after the USPSTF recommended ultrasound screening for abdominal aortic aneurysm and 6 years after an act of Congress mandated payment for it, the rate of abdominal aortic aneurysm screening with ultrasound was only 9%. This was similar to rates reported by others, and more than double the rate 5 years earlier. However, when including other imaging studies that visualize the aorta, we found that 41.5% of eligible patients had been screened adequately. Ultrasound screening rates varied widely among physicians, with some screening none of their patients and others screening close to 50%. Physician characteristics associated with ultrasound screening included MD degree, female sex, specialization in Internal Medicine, and ordering more preventive health measures. Patient characteristics associated with screening for abdominal aortic aneurysm included older age, status as a current smoker, family history of abdominal aortic aneurysm, having a periodic health examination, and completion of other health maintenance measures.
      Screening rates by ultrasound in previous studies have also been low, ranging from 8.2% in a Veteran Affairs primary care clinic to 12.9% in a health system with automated reminders.
      • Eaton J.
      • Reed D.
      • Angstman K.B.
      • et al.
      Effect of visit length and a clinical decision support tool on abdominal aortic aneurysm screening rates in a primary care practice.
      • Federman D.G.
      • Carbone V.G.
      • Kravetz J.D.
      • Kancir S.
      • Kirsner R.S.
      • Bravata D.M.
      Are screening guidelines for abdominal aortic aneurysms being implemented within a large VA primary health care system?.
      Low rates were attributed to physician lack of familiarity with the guidelines, competition with other screening examinations for priority, and provider perception of a lack of patient knowledge about screening.
      • Eaton J.
      • Reed D.
      • Angstman K.B.
      • et al.
      Effect of visit length and a clinical decision support tool on abdominal aortic aneurysm screening rates in a primary care practice.
      Lack of familiarity appeared to play a role in our institution, as rates increased over time and academic physicians had higher rates than others. Abdominal aortic aneurysm screening did not appear to compete with other health maintenance measures, as higher rates of screening were observed among those who had the most preventive care in other domains. In fact, physicians' rates of ultrasound screening increased exponentially in relation to completion of other health maintenance measures, indicating that ultrasound screening for abdominal aortic aneurysm was prioritized below other health maintenance practices.
      Clearly there is room for improvement. Reminder alerts built into the electronic health record might be expected to enhance compliance with evidence-based recommendations. However, one recent study evaluating the implementation of a new electronic health record best practice alert system revealed that 48% of otherwise eligible patients had already received a radiographic examination of the abdomen.
      • Hye R.J.
      • Smith A.E.
      • Wong G.H.
      • Vansomphone S.S.
      • Scott R.D.
      • Kanter M.H.
      Leveraging the electronic medical record to implement an abdominal aortic aneurysm screening program.
      In our health system, we found that 31% of screening ultrasounds performed were unnecessary because the patients had already undergone abdominal imaging, demonstrating the importance of incorporating prior imaging into any reminder rules. Every physician in our data set had at least one redundant ultrasound order, which indicates that physicians do not monitor routinely for previous imaging before ordering an ultrasound.
      Family history of abdominal aortic aneurysm has been noted to be among the strongest predictors of abdominal aortic aneurysm risk.
      • Kent K.C.
      • Zwolak R.M.
      • Egorova N.N.
      • et al.
      Analysis of risk factors for abdominal aortic aneurysm in a cohort of more than 3 million individuals.
      In the US, the prevalence of abdominal aortic aneurysm is 1.4%; however, only 1.2% of patients in our sample had a family history of abdominal aortic aneurysm documented in the electronic health record, and only half of these had undergone screening. This highlights the need to improve evaluation of family history as well as screening practices.
      This study has a number of limitations. First, we examined one large health system and our results may not be representative of national patterns. However, our rates were similar to those seen in other studies and low screening rates likely exist in other systems nationwide. We have no information about rates of patient refusal of offered screening or patient rationale for missing ordered screening. This would not, however, change the rate of patients who were screened. We included only tests done at Cleveland Clinic, which may underestimate screening rates if patients had tests conducted elsewhere. Finally, we did not evaluate results of the screenings or physician rational for ordering or not ordering a screening, so we cannot comment on the diagnostic yield or patient outcomes.
      The USPSTF recently updated their screening guidelines to include selective screening among men aged 65-75 years who have never smoked.
      • LeFevre M.L.
      U.S. Preventive Services Task Force
      Screening for abdominal aortic aneurysm: U.S. Preventive Services Task Force recommendation statement.
      This will increase the pool of those eligible for screening as well as the complexity of the screening recommendation. Both of these may contribute to declining rates of screening among those eligible. Future efforts to improve these rates and decrease variation among providers will likely involve system-based quality-improvement initiatives, such as reminder systems and risk calculators built into the electronic health record. These have been shown to increase screening rates for abdominal aortic aneurysm and for other diseases by 12%-14%.
      • Dexheimer J.W.
      • Talbot T.R.
      • Sanders D.L.
      • Rosenbloom S.T.
      • Aronsky D.
      Prompting clinicians about preventive care measures: a systematic review of randomized controlled trials.
      However, given the fact that nearly half of screening tests were performed on patients who had already been screened, it will be important that such reminder systems check for other abdominal imaging before suggesting abdominal aortic aneurysm ultrasound. In addition, although CT has been considered in many studies to be equivalent to or better than ultrasound for imaging abdominal aortic aneurysm, the use of CT for inadvertent screening has not been studied.
      • Federman D.G.
      • Carbone V.G.
      • Kravetz J.D.
      • Kancir S.
      • Kirsner R.S.
      • Bravata D.M.
      Are screening guidelines for abdominal aortic aneurysms being implemented within a large VA primary health care system?.
      • Hye R.J.
      • Smith A.E.
      • Wong G.H.
      • Vansomphone S.S.
      • Scott R.D.
      • Kanter M.H.
      Leveraging the electronic medical record to implement an abdominal aortic aneurysm screening program.
      It is not known whether radiologists routinely report the aortic diameter or whether such reports are followed up. If future screening strategies are to incorporate inadvertent screening, it will be important to ensure that radiologists address aortic diameter routinely or that primary care providers review the films when the issue of abdominal aortic aneurysm screening arises. Lastly, abdominal aortic aneurysm screening is currently recommended, but is not considered a quality measure for public reporting or pay-for-performance programs. Incorporation of abdominal aortic aneurysm screening measures into such programs would likely boost adherence.

      Conclusion

      Seven years after publication of the USPSTF guidelines, screening rates for abdominal aortic aneurysm remain low, although rates are substantially higher when considering all studies that visualize the aorta. Repeat screening with ultrasound in patients who previously had abdominal imaging is an emerging problem. These findings draw attention to a need for increased education around abdominal aortic aneurysm screening and information technology solutions that prompt physicians to screen while avoiding duplicate imaging.

      Acknowledgment

      The authors would like to thank Jennifer DiPiero, Education Coordinator in Personalized Health for the Cleveland Clinic Medicine Institute, for her assistance in preparing the manuscript for publication.

      References

        • Kent K.C.
        • Zwolak R.M.
        • Jaff M.R.
        • et al.
        Screening for abdominal aortic aneurysm: a consensus statement.
        J Vasc Surg. 2004; 39: 267-269
        • Kent K.C.
        • Zwolak R.M.
        • Egorova N.N.
        • et al.
        Analysis of risk factors for abdominal aortic aneurysm in a cohort of more than 3 million individuals.
        J Vasc Surg. 2010; 52: 539-548
        • Kuivaniemi H.
        • Platsoucas C.D.
        • Tilson M.D.
        Aortic aneurysms: an immune disease with a strong genetic component.
        Circulation. 2008; 117: 242-252
        • Thompson S.G.
        • Ashton H.A.
        • Gao L.
        • Buxton M.J.
        • Scott R.A.
        Final follow-up of the Multicentre Aneurysm Screening Study (MASS) randomized trial of abdominal aortic aneurysm screening.
        Br J Surg. 2012; 99: 1649-1656
        • U.S. Preventive Services Task Force
        Screening for abdominal aortic aneurysm: recommendation statement.
        Ann Intern Med. 2005; 142: 198-202
        • Fleming C.
        • Whitlock E.P.
        • Beil T.L.
        • Lederle F.A.
        Screening for abdominal aortic aneurysm: a best-evidence systematic review for the U.S. Preventive Services Task Force.
        Ann Intern Med. 2005; 142: 203-211
        • Wilmink A.B.
        • Hubbard C.S.
        • Quick C.R.
        Quality of the measurement of the infrarenal aortic diameter by ultrasound.
        J Med Screen. 1997; 4: 49-53
        • Wilmink A.B.
        • Forshaw M.
        • Quick C.R.
        • Hubbard C.S.
        • Day N.E.
        Accuracy of serial screening for abdominal aortic aneurysms by ultrasound.
        J Med Screen. 2002; 9: 125-127
        • Barnett S.B.
        • Terhaar G.R.
        • Ziskin M.C.
        • Rott H.D.
        • Duck F.A.
        • Maeda K.
        International recommendations and guidelines for the safe use of diagnostic ultrasound in medicine.
        Ultrasound Med Biol. 2000; 26: 355-366
        • Lederle F.A.
        • Wilson S.E.
        • Johnson G.R.
        • et al.
        Variability in measurement of abdominal aortic aneurysms. Abdominal Aortic Aneurysm Detection and Management Veterans Administration Cooperative Study Group.
        J Vasc Surg. 1995; 21: 945-952
      1. 109th US Congress. Deficit Reduction Act of 2005. Available at: http://www.gpo.gov/fdsys/pkg/PLAW-109publ171/html/PLAW-109publ171.htm. Accessed June 11, 2014.

        • Eaton J.
        • Reed D.
        • Angstman K.B.
        • et al.
        Effect of visit length and a clinical decision support tool on abdominal aortic aneurysm screening rates in a primary care practice.
        J Eval Clin Pract. 2012; 18: 593-598
        • Federman D.G.
        • Carbone V.G.
        • Kravetz J.D.
        • Kancir S.
        • Kirsner R.S.
        • Bravata D.M.
        Are screening guidelines for abdominal aortic aneurysms being implemented within a large VA primary health care system?.
        Postgrad Med. 2009; 121: 132-135
        • Hye R.J.
        • Smith A.E.
        • Wong G.H.
        • Vansomphone S.S.
        • Scott R.D.
        • Kanter M.H.
        Leveraging the electronic medical record to implement an abdominal aortic aneurysm screening program.
        J Vasc Surg. 2014; 59: 1535-1543
        • LeFevre M.L.
        • U.S. Preventive Services Task Force
        Screening for abdominal aortic aneurysm: U.S. Preventive Services Task Force recommendation statement.
        Ann Intern Med. 2014; 161: 281-290
        • Dexheimer J.W.
        • Talbot T.R.
        • Sanders D.L.
        • Rosenbloom S.T.
        • Aronsky D.
        Prompting clinicians about preventive care measures: a systematic review of randomized controlled trials.
        J Am Med Inform Assoc. 2008; 15: 311-320