Implementation of an aortic screening program in clinical practice: Implications for the Screen for Abdominal Aortic Aneurysms Very Efficiently (SAAAVE) Act

Article Outline

• Abstract
• Methods
• Results
• Discussion
• Conclusions
• Author contributions
• Acknowledgment
• References
• Copyright

Objective

Screening for abdominal aortic aneurysms (AAA) significantly reduces aneurysm-related death. In January 2007, the Federal government enacted Medicare coverage guideline to screen persons at risk for the presence of an AAA, the Screen for Abdominal Aortic Aneurysms Very Efficiently (SAAAVE) Act. The purpose of this study is to evaluate the efficacy and costs of a large scale screening effort for identifying AAAs in patients in clinical practice.

Methods

A regional veterans affairs mandate for screening for AAA was implemented in February 2007. Data were extracted through the Northern California Veterans Affairs (VA) Service Network to identify veteran males 65-75 years of age who ever smoked at least 100 cigarettes during their lifetime. An AAA was defined as an aortic diameter 3.0 cm or greater. A Decision Support Systems software (LumiData, Minneapolis, Minn) package tracked true costs of conducting a large AAA screening protocol in the Northern California VA Health Care System.

Results

A total of 2918 patients (average age, 71 ± 6 years) were screened for AAA over a 1-year period from February 2007 to February 2008. An AAA was diagnosed in 5.1% (148/2918) of patients. Two hundred ninety patients out of the 2918 (9.9%) were inappropriately screened. The aneurysm distribution was as follows: 83% (123/148) of the aneurysms were 3.0-4.4 cm, 13% (19/148) were 4.5-5.5 cm, and 4.1% (6/148) were greater than 5.5 cm. Incidental findings of isolated iliac artery aneurysms were found in 0.1% (3/2918) of patients. The cost of AAA screening per patient is $53.

Conclusion

The results of a large AAA screening effort in clinical practice reflect the results reported in the major clinical trials at a reasonable cost. The identification of large iliac artery aneurysms in the screening has not been previously reported.

Abdominal aortic aneurysms (AAAs) account for over 15,000 deaths in the United States each year.¹ The aneurysm-related deaths underestimate the true number of deaths since patients with ruptured AAAs may never survive the trip to the hospital to receive medical care. Given the asymptomatic nature and potentially life-threatening risk, ultrasonography,² has been proposed as a routine screening modality. Its use has been studied in several randomized prospective studies.³, ⁴, ⁵, ⁶ Unanimously, the studies concluded that screening prevented aneurysm-related death. In 2005, a meta-analysis of these screening trials prompted the United States Preventive Services Task Force (USPSTF) to change its earlier neutral position on AAA screening to recommend the screening of all men between the ages of 65 to 75 who ever smoked 100 cigarettes or more.⁷

Importantly, the USPSTF cautioned that the limitation of their recommendations were that none of the studies were based on patients in the United States. The four well-performed and large-scale randomized clinical trials compared aneurysm-related death rates between a screened and non-screened patient population. The largest and best quality study was the UK's Multicentre Aneurysm Screening Study (MASS).³ In the United States, no randomized clinical trial data are available. However, based on sponsored screening trials, the Stroke and Aneurysm Vascular Evaluation (SAVE) program⁸ demonstrated an AAA prevalence of 2.1% and the dare to Carotid artery stenosis, Abdominal aortic aneurysms, Renal artery stenosis, and Extremity artery stenosis (CARE) trial⁹ revealed an AAA prevalence of 1.6%, significantly lower than the rates reported in the four clinical trials.

Despite the differing prevalence rates reported in the US and the USPSTF recommendations, several major medical societies and a bipartisan group of senators and congressional representatives introduced legislation to enable Medicare to cover the costs of AAA screening, the Screening Abdominal Aortic Aneurysms Very Efficiently (SAAAVE) act. The act became law in January 1, 2007. Medicare covers a one-time ultrasound scan screening of men aged 65-75 who ever smoked in their lifetime or men and women who have a family history of AAA disease as part of a “Welcome to Medicare” package. To allow for the covered ultrasound scan screening, the “Welcome to Medicare” physical examination must take place within 6 months that the beneficiary becomes eligible. In essence, the SAAAVE act excludes all participating Medicare beneficiaries before 2007 and potentially many future Medicare beneficiaries who fail to meet the strict Welcome to Medicare guidelines.

The Department of Veterans Affairs (VA) adopted a more expansive view of the AAA screening program. Rather than exclude patients who are already in the VA system, as the SAAAVE act excludes all patients already in Medicare, a retroactive screening protocol was enacted. Approximately one million veteran patients meet criteria and will be screened. To institute the screening program to all eligible patients, the VA mandates that all practitioners remain judicious in working down the patient backlog. The purpose of this study is to evaluate 1-year results of the efficacy and associated costs in implementing a large AAA screening effort and to identify the prevalence of patients with AAAs in the context of a US clinical practice.

Back to Article Outline

Methods 

An initial evaluation was made of male veterans 65-75 who smoked or female patients older than 50 with a family history of AAA who had utilized the VA health care system from the previous fiscal year, 2006. This drew from a large database of patients in the Northern California VA Health Care System. Patient lists were generated and notices were sent to veterans by the radiology department as an invitation for AAA screening. Simultaneously, through the electronic medical records system, alerts were posted to the primary care physicians that an AAA screening was due for their patient. Implementation of the VA-mandated screening protocol began January 1, 2007, in response to the new federal law, the SAAAVE act. A 1-month period was required to develop and implement a unique CPT code for “AAA screening” in the VA system. This code was utilized beginning February 1, 2007, to track utilization and associated costs. All patients who underwent AAA screening from February 2007 to February 2008 with this CPT code were identified.

A retrospective review of all patients having been screened for an AAA during this time period was conducted under an Investigational Review Board (IRB) approved protocol at the Northern California VA Health Care System. Age, aortic diameter, and costs associated with the screening protocol were evaluated. Identification of an AAA was defined as an aortic diameter measuring 3.0 cm or larger. Prevalence and cost data were compared to a prior large-scale screening trial, the MASS.³ MASS was selected for comparison because of the “good quality” rating in the USPSTF recommendations as well as MASS had the narrowest confidence interval and contributed the most weight to the pooled odds ratio in the USPSTF recommendations.⁷

Cost data regarding the implementation of the AAA screening protocol was tracked prospectively using the VA Decision Support Systems.¹⁰, ¹¹ The Decision Support Systems is a computer software program where the budget allocation system combines clinical and fiscal information at a detailed level. This VA budget allocation system was piloted in 1986 and began full implementation at the behest of Congress in 1994.¹¹ The Northern California VA Health Care Systems has an independent department which tracks these costs so that budgetary planning can be made for successive years.

Since the VA does not generate a bill, direct costs such as the technician, administrative support, and radiologists' time are not calculated on a per patient basis. Rather, all costs in running the radiology department, for example, are estimated. The budget department then requests an estimate of the proportion effort devoted to the screening program. This value is then divided by the number of patients screened in that particular time period. Other large expense items such as the ultrasound machine, gurneys, probes, and software packages are not included in the cost estimate to implement the screening program since the VA already owns these items and no incremental costs were required.

Continuous variables are reported as a mean ± standard deviation. A two-tailed t test was employed to compare continuous variables, and a χ² test was utilized to compare proportions. A P value of less than .05 was considered statistically significant.

Back to Article Outline

Results 

Over the 1-year period, 2918 patients (average 71 ± 6 years) were successfully screened for AAA. A total of 84% (2918/3487) eligible veterans were screened. A total of 9.9% (290/2918) were inappropriately screened: 118 patients were under 65-years-old, 172 patients were over 75-years-old, and 7 women were screened without a clear family history of aneurysmal disease. All 7 women were also too old to be included in the screening. Only 0.3% (9/2918) of patients had a non-diagnostic ultrasound scan screening requiring a non-contrast abdominal/pelvic computed tomography (CT) scan for screening. The age distribution and the percent screened were not statistically different in this study when compared to the results obtained in the MASS trial. The age was 71 ± 6 in this VA study compared to the MASS trial of 69.2 ± 3 years (P = NS). The percent screened was 84% (2918/3487) in this VA study, whereas the MASS study screened 80% (27,147/33,839) (P = .11). A comparison of age and the percentage of patients screened are reported in Table I.

Table I. Comparison of age and % attendance between studies

	Age ± SD (years)	Age range (years)	% Attendance
VA study	70.5±6	65-75	84%
			2918/3487
MASS (2002)	69.2±3	65-74	80%
			27,147/33,839
Danish (2002)	67.5*	65-73	76%
			4843/6339
Australian (2004)	72.6±5	65-83	63%
			12,203/19,352

No significant differences identified in the ages between studies. Only the Australian study reported a lower attendance rate in the invitation for screening.

No standard deviation (SD) is reported for the Danish study. VA, Veterans Affairs; MASS, Multicentre Aneurysm Screening Study.

A total of 5.1% (148/2918) were diagnosed with an AAA 3.0 cm or larger, this result is similar to that reported in the MASS trial at 4.9% (1333/26,818) of patients invited (P = .82). The screened patients were mostly men 99.8% (2911/2918), 59.7% (1743/2918) White, 10.8% (314/2918) African American, 3.8% (112/2918) Asian/Pacific Islander, and 25.2% (736/2918) unknown. The distributions of the aneurysm diameter in the patients identified with an AAA were: 83% (123/148) for the 3.0-4.4 cm group, 13% (19/148) for the 4.5-5.4 cm, and 4.1% (6/148) for the 5.5 cm or greater group. The distribution of AAA diameters compare similarly to that of the MASS trial, except for the 5.5 cm or larger group, where the MASS trial reported a 12% rate (166/1333) (P = .006). The distribution of aneurysm prevalence and aneurysm distribution are summarized and compared with three other randomized controlled screening trials in Table II. The Chichester study was excluded from the table because of the differing diameter criteria reported and the use of 6.0 cm threshold for surgical intervention.

Table II. Comparison of AAAs detected and diameters between studies

Out of the total number of patients screened, there were 0.4% (12/2918) of patients who had an AAA 5.0 cm or larger. Isolated iliac artery aneurysms were found in 0.1% (3/2918) of patients. All 22 patients with a 3.0 cm iliac artery aneurysm or an AAA greater than 4.5 cm in size were appropriately referred to the vascular surgery service.¹²

An analysis of the cost data was based on the proportion of labor (administrative support, radiologists, and technicians) devoted to the screening in addition to supplies and other indirect costs. Weighted averages of costs estimate were calculated from the following Northern California VA Medical Centers: Martinez, Sacramento, Redding, and Oakland. In 2008, the estimated cost per screening was $53. To keep the dollar cost analysis in terms of 2008 dollars, the cost analysis was based on a data capture from January 1, 2008, to September 1, 2008, different than the time period for the evaluation of the screening protocol. Costs can differ based upon the number of radiologists on staff, the volume of studies performed, the cost of living in that geographical area, and the less tangible estimates of indirect costs, such as the amortized costs of the equipment and overhead. Based on cost data collected through Decisions Support Systems, the screening costs were $1050 to identify a patient with an AAA at least 3.0 cm in diameter, $6200 for an AAA at least 4.5 cm in diameter, and $25,900 for an AAA at least 5.5 cm in diameter. The escalating costs are due to the decreasing prevalence of larger aneurysms in the patient population. The costs reflect an estimate of what would be required to identify an aneurysm at each respective diameter.

Back to Article Outline

Discussion 

In January 1, 2007, the SAAAVE act became law. A payment will be made for a one-time screening of patients for identifying an AAA who meet the following criteria:¹³ (1) Complete an initial preventive physical examination (Welcome to Medicare Visit) within 6 months of Medicare eligibility, (2) Have not previously had an ultrasound scan screening examination under the Medicare program, and (3) Be a man aged 65-75 years of age who smoked at least 100 cigarettes or be a man or woman who has a family history of AAA.

Thus, the eligibility for Medicare covered AAA screening is rather limited.

The Northern California VA Health Care System has worked aggressively to apply the VA directive. An analysis of the past year's efforts have been presented here and compared to the benchmarks set by the USPSTF best evidence reviews.⁷ Within the Task Force's evaluation of the data, the MASS³ results were heavily weighted accounting for 60.2% of the patients evaluated. Further sub-analysis excluding the MASS results by the Task Force led to the same conclusions, that screening prevented aneurysm-related death. In our analysis, the MASS results were used as the benchmark with which to compare the efficacy of the Northern California VA screening results in actual practice (Table I). A limitation of the Task Force recommendations was the applicability of European and Australian publications to a US population. This retrospective study is the first to evaluate the clinical application of a screening protocol in a large US clinical practice. The results reported in this study are more closely aligned to the data reported in the other large clinical trials as opposed to the initiated screening trials performed in the United States.⁸, ⁹ The design of the US studies were to offer a free ultrasound scan screening to eligible patients typically in shopping malls. The prevalence may be lower in these studies as healthier subjects in the invited screening would be more likely to participate than as part of a screening protocol in clinical care. The prevalence rate in these studies ranged from 1.6% to 2.1%, significantly lower than the 5% reported in this study and the other randomized clinical trials.

There are some practical challenges in assessing the results of the screening efforts. Patients in clinical practice will move in and out of eligibility based on geographic, insurance, and age concerns. Depending upon the time the eligibility assessment is made, a patient may be age eligible at the time of data capture (ie, 75-years-old) but be 76-years-old at the time the ultrasound scan examination is completed. The capture rate of 84% screened vs eligible patients, compares favorably with the 80% screening rate obtained in the MASS study. The 9.9% rate of inappropriately screened patients is likely due to primary care physicians incorrectly ordering an AAA screening examination. In the VA system, meeting clinical reminders are rewarded, where as inappropriate ordering of studies are not necessarily punished. The improved yield rate will likely reflect a higher incidence of inappropriately screened patients in clinical practice.

An ultrasound scan was usually technically adequate for screening. Only 0.3% of examinations were unable to measure aortic diameters, compared to the 1.2% screening failure reported in the MASS trial. The reason for this is unclear. Whether the VA radiologists were willing to render an interpretation of aortic diameter based on limited views compared to the more extensive visualization demanded by the MASS study design is speculation. Another possibility is the hands-on approach of the VA radiologists working with the technologists to obtain an adequate study.

The identification of isolated iliac artery aneurysms requiring treatment occurred in 0.1% of patients, whereas no mention of incidental iliac artery aneurysms were made in the MASS trial³ or the USPSTF review.⁷ This observation supports the importance of vascular laboratory technicians to evaluate the iliac arteries.

The data reported here showed a significantly lower prevalence of large aneurysms of 5.5 cm or greater at 4.1% of patients compared to the 12% of patients reported in the MASS trial. This difference could be attributed to a couple of factors. The VA included patients who are utilizing their health care benefits and it is possible that their risk factors may be more effectively managed. The MASS trial subjects were culled from a population-based group of patients who were not necessarily utilizing their health care benefits. Secondly, the incident rate of active smokers in our VA patients at the time of screening in a post-hoc analysis was 25%. Whether this rate of active smokers compares favorably to the MASS trial is unknown. A smaller fraction of active smokers in the aneurysmal population could argue for the effect of aneurysm expansion rate and the prevalence of larger aneurysms.

This study evaluated costs, not charges or reimbursement. The VA does not bill Medicare for screening studies. The VA has a Decision Support System for analysis of expenditures and associated indirect costs for services. In the analysis of cost, the VA does not generate a bill as does the private sector. The Decision Support System methods for cost analysis have been published elsewhere.¹⁰, ¹¹ The average cost per patient is $53. Reported here are slightly lower than prior estimates generated by Wanhainen et al¹⁴ who estimated the costs to be $55 per screening study for patients over 65 years of age in 2003 dollars.

In the analysis of recommending a screening program, the cost to administer the study is an important variable. Costs of screening will decrease and quality should improve with greater volumes. Based on this analysis, a more expansive application of the SAAAVE act for Medicare beneficiaries should be considered. The SAAAVE act excludes millions of Medicare beneficiaries who enrolled prior to 2007 and new beneficiaries who fail to meet fairly strict guidelines: complete a Welcome to Medicare physical within 6 months of enrolling and have an AAA ultrasound scan screening performed. This further detracts from the utility of the screening efforts since many primary care physicians will not routinely think AAA screening.

The important aspect of this screening is not limited to identifying patients with large aneurysms (greater than 5.5 cm) for repair, but rather the identification of patients with smaller aneurysms of 4.0-5.4 cm. Several aneurysm trials have shown the natural history of patients with small aneurysms to require repair 60% of the time in a 3-year follow-up period.¹⁵, ¹⁶ Given this strategy, further ultrasound scan studies will be required in this subset of patients for follow-up. Wanhainen et al have already included this possibility in their Markov analysis model. They conclude that a mass AAA screening protocol is still cost-effective.¹⁴ A careful analysis of costs and outcome of patients with a 3.0-4.5 AAA diameter will need to be followed prospectively.

Although no US randomized clinical trial was performed comparing AAA screening to a control population without screening, several salient points should be considered from these large randomized controlled trials. From the MASS study, screened patients with a 4.5 cm or larger AAA that were detected in a study population (<1%) had a significant 42% relative risk reduction in aneurysm related deaths. The US population may have a greater risk reduction since the spread between elective AAA repair mortality rates are 3.1%¹⁷ and emergent AAA repair mortality rates are 44%.¹⁸ In the MASS UK dataset, there was a 7% perioperative mortality for elective AAA repair and a 37% mortality for ruptured AAA repair. This is important since the non-screened group had significantly higher rates of ruptured aneurysms.

The overall mortality from all causes, the MASS study reports a reduction from 3% in the unscreened group to 2% in the screened group. Although the all-cause mortality was not significantly reduced, there was a significant reduction in ischemic heart disease deaths signifying a secondary benefit to screening. Furthermore, through a screening program, an additional 2.5 patients would have an AAA detected for every AAA detected through routine care. This estimate is extrapolated from the 92 large aneurysms (>5.5 cm) detected through routine care as opposed to 322 aneurysms detected through a screening program in the MASS study. Given that over 15,000 elective repairs were performed in US Medicare beneficiaries annually, a more expansive screening program could potentially generate an additional 37,500 Medicare patients that have an AAA >5.5 cm.

Back to Article Outline

Conclusions 

The clinical application of a large AAA screening program, compare well with the results of several randomized controlled screening studies. Isolated iliac artery aneurysms greater than 3.0 cm have been identified, demonstrating the importance of the ultrasonographer to screening the iliac arteries. Serious consideration should be given to expanding the SAAAVE act for Medicare beneficiaries to a retrospective expanded screening effort, similar to that enjoyed by our veteran patients.

Back to Article Outline

Author contributions 

Back to Article Outline

The authors would like to thank Lisa Han, Kourteney Johnson, and Lenny Eusebio for data collection.

Back to Article Outline

References 

1. Lesperance K, Andersen C, Singh N, Starnes B, Martin MJ. Expanding use of emergency endovascular repair for ruptured abdominal aortic aneurysms: disparities in outcomes from a nationwide perspective. J Vasc Surg. 2008;47:1165–1170discussion 1170-1.
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (223 KB)
   • CrossRef
2. Wilmink AB, Hubbard CS, Quick CR. Quality of the measurement of the infrarenal aortic diameter by ultrasound. J Med Screen. 1997;4:49–53
   • View In Article
   • MEDLINE
3. Ashton HA, Buxton MJ, Day NE, Kim LG, Marteau TM, Scott RA, et al. The Multicentre Aneurysm Screening Study (MASS) into the effect of abdominal aortic aneurysm screening on mortality in men: a randomised controlled trial. Lancet. 2002;360:1531–1539
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (338 KB)
   • CrossRef
4. Lindholt JS, Juul S, Fasting H, Henneberg EW. Hospital costs and benefits of screening for abdominal aortic aneurysms (Results from a randomised population screening trial). Eur J Vasc Endovasc Surg. 2002;23:55–60
   • View In Article
   • Abstract
   • Full-Text PDF (97 KB)
   • CrossRef
5. Norman PE, Jamrozik K, Lawrence-Brown MM, Le MT, Spencer CA, Tuohy RJ, et al. Population based randomised controlled trial on impact of screening on mortality from abdominal aortic aneurysm. BMJ. 2004;329:1259
   • View In Article
   • CrossRef
6. Scott RA, Wilson NM, Ashton HA, Kay DN. Influence of screening on the incidence of ruptured abdominal aortic aneurysm: 5-year results of a randomized controlled study. Br J Surg. 1995;82:1066–1070
   • View In Article
   • MEDLINE
   • CrossRef
7. Fleming C, Whitlock EP, Beil TL, Lederle FA. 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
   • View In Article
8. Derubertis BG, Trocciola SM, Ryer EJ, Pieracci FM, McKinsey JF, Faries PL, et al. Abdominal aortic aneurysm in women: prevalence, risk factors, and implications for screening. J Vasc Surg. 2007;46:630–635
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (93 KB)
   • CrossRef
9. Hupp JA, Martin JD, Hansen LO. Results of a single center vascular screening and education program. J Vasc Surg. 2007;46:182–187discussion 188-9.
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (106 KB)
   • CrossRef
10. Barnett PG. Review of methods to determine VA health care costs. Med Care. 1999;37(4 Suppl Va):AS9–AS17
    • View In Article
    • MEDLINE
    • CrossRef
11. Hendricks AM, Lotchin TR, Hutterer J, Swanson J, Kenneally K Decision Support System Cost Evaluation Work Group. Evaluating VA patient-level expenditures: decision support system estimates and Medicare rates. Med Care. 2003;41(6 Suppl):II111–II117
    • View In Article
    • MEDLINE
12. Kent KC, Zwolak RM, Jaff MR, Hollenbeck ST, Thompson RW, Schermerhorn ML, et al. Screening for abdominal aortic aneurysm: a consensus statement. J Vasc Surg. 2004;39:267–269
    • View In Article
    • Full Text
    • Full-Text PDF (55 KB)
    • CrossRef
13. Abdominal Aortic Aneurysm Screening Provider Resources. http://www.cms.hhs.govAccessed September 2008.
    • View In Article
14. Wanhainen A, Lundkvist J, Bergqvist D, Björck M. Cost-effectiveness of different screening strategies for abdominal aortic aneurysm. J Vasc Surg. 2005;41:741–751discussion 751.
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (231 KB)
    • CrossRef
15. [No authors listed] Mortality results for randomised controlled trial of early elective surgery or ultrasonographic surveillance for small abdominal aortic aneurysms.The UK Small Aneurysm Trial Participants. Lancet. 1998;352:1649–1655
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (98 KB)
    • CrossRef
16. Lederle FA, Wilson SE, Johnson GR, Reinke DB, Littooy FN, Acher CW, et al. Immediate repair compared with surveillance of small abdominal aortic aneurysms. N Engl J Med. 2002;346:1437–1444
    • View In Article
    • CrossRef
17. Schermerhorn ML, O'Malley AJ, Jhaveri A, Cotterill P, Pomposelli F, Landon BE. Endovascular vs. open repair of abdominal aortic aneurysms in the Medicare population. N Engl J Med. 2008;358:464–474
    • View In Article
    • CrossRef
18. Egorova N, Giacovelli J, Greco G, Gelijns A, Kent CK, McKinsey JF. National outcomes for the treatment of ruptured abdominal aortic aneurysm: comparison of open versus endovascular repairs. J Vasc Surg. 2008;48:1092–11001100 e1-2.
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (1098 KB)
    • CrossRef