Costs and Benefits of Bundled Community-Based Screening for Carotid Artery Stenosis, Peripheral Artery Disease, and Abdominal Aortic Aneurysm
Costs and Benefits of Bundled Community-Based Screening for Carotid Artery Stenosis, Peripheral Artery Disease, and Abdominal Aortic Aneurysm
Purpose: Perform an initial formal assessment of the costs and benefits of bundled cardiovascular screening. Primarily, determine the relative importance of data uncertainties to the integrity of modeled outcomes associated with bundled screening for carotid artery stenosis (CAS), peripheral artery disease (PAD), and abdominal aortic aneurysm (AAA); secondarily, establish parameters around potential costs and outcomes benefits of this screening bundle.
Design: A decision-analysis framework composed of four decision tree submodels with transition probabilities specific to four age- and gender-specific subgroups. Model transition probabilities for each of the four submodels are based on the prevalence of all possible combinations of the presence or absence of moderate CAS, significant CAS, PAD, and AAA, and the likelihood of appropriate or inappropriate medical follow-up.
Methodology: Evaluates a hypothetical self-funded employer with 10,000 beneficiaries and who is considering whether to provide bundled cardiovascular health screenings. Screenings would be performed in addition to any other medical screenings commonly performed by physicians to assess cardiovascular risk. Determine costs and catastrophic events (death, myocardial infarction, stroke, other cardiovascular-related events, amputation events) for these self-funded hypothetical beneficiaries aged ≥50 years and <65 years.
Results: The model predicts approximately $54 million in health care costs over 10 years for the control cohort and $51 million for the screening cohort, representing a 5.2% reduction in 10-year health care spending due to screening.
Conclusion: This initial analysis predicts robust cost and health benefits associated with the decision to provide bundled cardiovascular health screening to a self-funded employer’s beneficiaries aged ≥50 years and <65 years. Further analyses are necessary to better quantify the magnitude of the cost and health benefits.
Community-based screening for vascular disease is of growing popularity. Nevertheless, the cost-effectiveness of such consumer-directed screenings has been hotly debated (Rooke 2007). In this paper, we provide an analytical framework as well as a preliminary assessment of 10-year costs and health benefits associated with a bundled package of cardiovascular health screenings.
Cost-effectiveness analysis is used to examine the value, in terms of a combined measure of cost and clinical benefit (e.g., cost per correct diagnosis of disease) of one or more interventions designed to improve health (Gold 1996). These analyses are then used in the decision-making process for physicians, insurance companies, and many others (Gold 1996). Well-known examples of cost-effectiveness analyses have examined the value of screening for at-risk populations, such as the value of mammograms to detect breast cancer in women over 40 years old (vs. women over 50) and of colonoscopies or other screening methods to detect colorectal cancer in men and women over 50 (Eddy 1989, Pharoah 2013, Barouni 2012, Austin 2013). Newer analyses examine the benefits of other health screenings, such as lung cancer screenings (Goulart 2013), in populations whose risk is less defined. Because new screening programs must be in place for many years before impacts on patients’ lives and health care system costs can be measured, assumption-driven cost-effectiveness analyses are an appropriate step in determining screening interventions that have promise and should be investigated further.
Additionally, it is important to identify co-occurring diseases, as it has been demonstrated that the occurrence of multiple disease conditions is associated with health care cost escalation (Bhattacharjee 2013, Choi 2014, Joyce 2009). Earlier identification of co-occurring conditions to enable co-management of chronic conditions, especially those whose treatments may overlap, may reduce health care expenditures (Ajmera 2014, Brilleman 2013, DeLorenze 2014). In addition to detecting co-occurring disease, possible benefits of bundled medical screening include the potential to improve cost-effectiveness over screening for individual diseases (due to lower costs of bundled screening compared to individual screening). However, if multiple screening indications are rarely co-occurring, individual, targeted screening for at-risk populations may be more cost-effective than bundled screening.
To date, there appears to be a gap in understanding the potential benefits of bundled screening. Only one cost-effectiveness analysis was identified that has examined the possible benefits of simultaneous screening. This analysis examined the feasibility of mobile cancer screening vans that would screen for multiple types of cancer in urban settings (O’Malley 2002).
Community-based cardiovascular screenings have been conducted throughout the United States, with over 8 million of these screenings performed between January 2005 and August 2013, forming the largest database of its kind of health-related information. The characteristics and representativeness of this screening database to the U.S. population as a whole are the subject of a recent article by Weisman (2013).
Screenings performed include ultrasound assessment of peripheral artery disease (PAD) (determined by ankle-brachial index [ABI]), ultrasound assessment of carotid artery stenosis (CAS), and ultrasound assessment of the abdominal aorta for abdominal aortic aneurysm (AAA). PAD is defined as an ABI <0.9 (Rooke 2011); possibly clinically significant CAS is defined as an internal carotid artery peak systolic velocity of ≥140 cm/sec (moderate CAS defined as peak systolic velocity 110–139 cm/sec); and all patients with an outer diameter of the abdominal aorta ≥3 cm were considered to have an AAA.
Guidelines for conducting cardiovascular screening have been mixed. Screening for AAA has been generally accepted to be beneficial in a stratified population (recommended most strongly for men aged 65 to 75 years who have smoked while not recommended in women who have never smoked) (LeFevre 2014a, Hirsch 2006, Lim 2011). AAA screening has also been recommended in younger men (≥55 years) with a family history of AAA (Chaikof 2009).
ABI has been found to be a reasonable tool for screening for PAD (Greenland 2010) and has been recommended for adults ≥65 years of age (Rooke 2011). PAD screening has also been recommended in individuals aged ≥55 who have risk factors for PAD, including smoking and diabetes (Rooke 2011). However, most recently, the United States Preventive Services Task Force (USPSTF) has found insufficient evidence to recommend for or against the use of ABI for PAD screening (Moyer 2013).
Individual screening for CAS in asymptomatic populations is particularly controversial. The USPSTF has recommended against screening in the general population entirely (LeFevre 2014b). The American Heart Association and American College of Cardiology Foundation have come to similar recommendations for asymptomatic individuals (Goldstein 2011, Brott 2011). Despite the recommendations regarding cardiovascular screening, what is clear from the screening data utilized in this analysis is that disease still exists in individuals beneath the recommended screening thresholds. Therefore, an exploration of the cost-effectiveness of screening these individuals (cohort ages ≥55 to <65 years) is still warranted.
Previous cost-effectiveness analyses of CAS, PAD, and AAA screening (as individual assessments) have been conducted and have demonstrated mixed results. Some analyses have found benefits of these screening tests in some patient demographic groups, while others have found screening to not be cost-effective for any patient group (Cull 2011, Jacobowitz 2003, Lee 1997, Mark 2003, Meenan 2005, Silverstein 2005, Søgaard 2012).
Based on these previous studies and a lack of studies on the cost-effectiveness of bundled screening, the screening database utilized provides a unique opportunity for preliminary exploration of the costs and outcomes associated with the decision to provide bundled vascular screening and provides a necessary platform on which rigorous cost-effectiveness analyses can be performed.
The primary objective of this study is to determine, through sensitivity analyses, those variables for which small changes in value result in changes in the nature of the modeled result (i.e., that change the evaluation of bundled screening as the lower-cost scenario to the higher-cost scenario). Secondarily, the analysis establishes parameters around the potential cost and outcomes benefits of bundled screening from the perspective of an employer-funded health plan. Owing to the preliminary nature of this model, only direct clinical costs (procedures, hospitalizations, etc.) and outcomes (identification of disease, progression of disease, and catastrophic events associated with disease) are explored.
Unlike fully insured plans, self-funded employers may assume direct risk for payment of claims for benefits. As costs of health care escalate, self-funded employers, particularly those that offer continued medical benefits to retirees, are increasingly at risk, particularly for self-funded employers that offer the lowest deductible health plans. For this preliminary model, it is assumed that the self-funded employer bears the full risk of payment on behalf of their beneficiaries.
In 2014, median employee tenure was 4.6 years (U.S. Bureau of Labor 2014), a time frame that may be too short to experience long-term net cost savings resulting from many health screenings. However, workers aged ≥55 years had 10.4 median years of tenure in 2014 (U.S. Bureau of Labor 2014). Longer employee tenures may allow self-funded employers to benefit more from earlier identification of disease, particularly cardiovascular disease, through disease screening programs. For these reasons, the current analysis was designed to examine the costs and outcomes associated with the decision to provide or not to provide bundled screening for a self-funded employer’s employees, dependents, and retired beneficiaries aged ≥50 years and <65 years.
This analysis considers a single cohort of employees receiving (or not receiving) bundled screening at baseline and examines their expected costs and outcomes over either a 10-year period, or until the employee turns 65, whichever occurs first. For the purpose of this initial analysis we have assumed all eligible employees will choose to receive the bundled screening benefit available to them.
Ten decision tree submodels, each with the model structure shown in Figure 1 but with transition probabilities specific to each of four age- and gender-specific subgroups, compose the overall decision analysis. Two of the four subgroups correspond to men (and two to women) aged 50 to 54, and 55 to 64.
FIGURE 1 Subset of decision tree transition possibilities
Men and women progress through the submodels according to transition probabilities based on prevalence of all possible combinations of the presence (or absence) of AAA, moderate CAS, significant CAS, and PAD and on the likelihood that patients will appropriately seek medical treatment if AAA, moderate CAS, or possibly clinically significant CAS or PAD, or both, is identified by the bundled screening diagnostics, or will inappropriately seek medical intervention if all 3 diagnostics are negative. Table 1 provides the baseline values for transition probabilities in two of the four submodels (men aged 55 to 64 and women aged 55 to 64 years). Although the likelihood of either appropriately or inappropriately seeking medical treatment following knowledge of screening results is dependent on both age and gender, for this initial assessment we have assumed these values are consistent across all age and gender subgroups.
|Table 1 Subset of decision tree transition probabilities|
|Variable namea||Variable description||Base case value|
|P1||Likelihood that patient will not have AAA||0.9380d||0.9989b|
|1–P1||Likelihood that patient will have AAA||0.0620||0.0011|
|P2||Likelihood that patient without AAA will not have CAS or PAD||0.9599c||0.9551c|
|P3||Likelihood that patient without AAA will not have CAS but will have PAD||0.0134c||0.0210c|
|P4||Likelihood that patient without AAA will have moderate CAS but not PAD||0.0144c||0.0139c|
|P5||Likelihood that patient without AAA will have moderate CAS and PAD||0.0010c||0.0010c|
|P6||Likelihood that patient without AAA will have possibly significant CAS but not PAD||0.0096c||0.0073c|
|1–(P2+P3+P4+P5+P6)||Likelihood that patient without AAA will have possibly significant CAS and PAD||0.0017||0.0010|
|P7||Likelihood that patient with AAA will not have CAS or PAD||0.8375c||0.7430c|
|P8||Likelihood that patient with AAA will not have CAS but will have PAD||0.0722c||0.1020c|
|P9||Likelihood that patient with AAA will have moderate CAS but not PAD||0.0258c||0.0543c|
|P10||Likelihood that patient with AAA will have moderate CAS and PAD||0.0088c||0.0185c|
|P11||Likelihood that patient with AAA will have possibly significant CAS but not PAD||0.0391c||0.0517c|
|1–(P7+P8+P9+P10+P11)||Likelihood that patient with AAA will have possibly significant CAS and PAD||0.0170c||0.0300|
|P12||Likelihood that patient without CAS, without PAD, and without AAA will appropriately not follow up with MD||0.8000||0.8000|
|1–P12||Likelihood that patient without CAS, without PAD, and without AAA will inappropriately follow up with MD||0.2000||0.2000|
|P13||Likelihood that patient with AAA and/or with CAS and/or with PAD will appropriately follow up with MD||0.9000||0.9000|
|1–P13||Likelihood that patient with AAA and/or CAS and/or PAD will inappropriately NOT follow up with MD||0.1000||0.1000|
aSee the corresponding variable names in the decision tree structure shown in Figure 1;
bdata from DeRubertis 2007;
cLife Line Screening data 2005–2008;
ddata from Singh 2001.
AAA=abdominal aortic aneurysm, CAS=carotid artery stenosis; PAD=peripheral artery disease.
As participants progress through the appropriate submodels, they accrue both costs and outcomes. Modeled results include those derived from a standard expected value decision analysis as well as from a series of one-way sensitivity analyses, examining the dependence of the modeled costs and outcomes on key variable uncertainties.
A hypothetical self-funded employer with 10,000 beneficiaries forms the basis for the modeled population. From the assumed beneficiary base, the modeled population consists of 1,351 men and 1,406 women, with age demographics drawn from 2011 U.S. Census Bureau data (U.S. Census Bureau 2012). This screening population consists only of individuals aged ≥50 years and <65 years.
Current smokers or tobacco users comprise 23.9% of all men and 17.9% of all women, with former smokers or tobacco users comprising another 8.0% and 8.3% of men and women, respectively (CDC 2013).
It is assumed for the initial analysis that all beneficiaries eligible for bundled screening are na&eunml;ve to CAS, PAD, and AAA screening and are not being treated for CAS, PAD, or AAA.
Main outcome measures
For the purposes of this analysis, costs are defined as treatment costs for the identified conditions (e.g., AAA repair, endarterectomy, peripheral bypass graft) and relevant postoperative and medical management costs up to Year 10 (or until a patient is 65 years old).
For each patient, costs are estimated over a 10-year window (or until that patient is 65 years old) to coincide with the time frame for the primary clinical outcomes measure described below. Base case values for cost variables utilized in the analysis are provided in Table 2. An annual distribution of treatment costs is assumed to be 5% for Years 1–4, 10% in Year 5, 20% in Years 6 and 7, and 10% in Years 8–10. These costs are exclusive of catastrophic events (stroke, myocardial infarction [MI], cardiovascular-related death, and amputation) and bundled screening cost. The clustering of costs in Years 6 and 7 is based on an assumption that major surgical procedures for AAA, CAS, and/or PAD would occur most frequently in this time frame.
|Table 2 Cost variablesa|
|Variable name||Variable description||Base case valueb||Justification|
|C1||Average health care cost of untreated no AAA, no CAS, no PAD||$21,940||Average health care costs for 10 yearsc|
|C2||Average cost of untreated no AAA, no CAS, PAD||$49,205||Assumes: cost of peripheral bypass graft;d average health care costs for 5 years;c costs of PAD treatment for 5 years;e 3% of patients require amputationf|
|C3||Average cost of untreated no AAA, moderate CAS, no PAD||$72,558||Assumes: cost of endarterectomy and extensive 2-year post-op medical management;g average health care costs for 5 years;c cost of CAS treatment for 3 years;e cost of medical management of CAs similar to cost of medical management of PAD|
|C4||Average cost of untreated no AAA, moderate CAS, PAD||$105,933||Assumes: cost of endarterectomy and extensive 2-year post-op medical management;g average health care costs for 5 years;c cost of peripheral bypass graft;d cost of CAS and PAD treatment for 3 years;e cost of medical management of CAS similar to cost of medical management of PAD; 3% of patients will require amputationf|
|C5||Average cost of untreated no AAA, possibly significant CAS, no PAD||$73,502||Assumes: cost of endarterectomy and extensive 2-year post-op medical management;g average health care costs for 1 year;c cost of CAS treatment for 7 years;e cost of medical management of CAS similar to cost of medical management of PAD|
|C6||Average cost of untreated no AAA, possibly significant CAS, PAD||$116,597||Assumes: cost of endarterectomy and extensive 2-year post-op medical management;g average health care costs for 1 year;c cost of peripheral bypass graft;d cost of CAS and PAD treatment for 7 years;e cost of medical management of CAS similar to cost of medical management of PAD; 3% of patients will require amputationf|
|C7||Average cost of untreated AAA, no CAS, no PAD||$70,770||Assumes: cost of emergency repair for ruptured AAA;h average health care costs for 10 yearsc|
|C8||Average cost of untreated AAA, no CAS, PAD||$98,035||Assumes: cost of emergency repair for ruptured AAA;h cost of peripheral bypass graft;d average health care costs for 5 years;c and cost of PAD treatment for 5 years;e 3% of patients will require amputationf|
|C9||Average cost of untreated AAA, moderate CAS, no PAD||$121,388||Assumes: cost of emergency repair for ruptured AAA;h cost of endarterectomy and extensive 2-year post-op medical management;g average health care costs for 5 years;c cost of CAS treatment for 3 years;e cost of medical management of CAS similar to cost of medical management of PAD|
|C10||Average cost of untreated AAA, moderate CAS, PAD||$154,763||Assumes: cost of emergency repair for ruptured AAA;h cost of endarterectomy and extensive 2-year post-op medical management;g peripheral bypass graft;d average health care costs for 5 years;c cost of CAS and PAD treatment for 3 years;e cost of medical management of CAS similar to cost of medical management of PAD; 3% of patients will require amputationf|
|C11||Average cost of untreated AAA, possibly significant CAS, no PAD||$122,332||Assumes: cost of emergency repair for ruptured AAA;h cost of endarterectomy and extensive 2-year post-op medical management;g average health care costs for 1 year;c cost of CAS treatment for 7 years;e cost of medical management of CAS similar to cost of medical management of PAD|
|C12||Average cost of untreated AAA, possibly significant CAS, PAD||$165,427||Assumes: cost of emergency repair for ruptured AAA;h cost of endarterectomy and extensive 2-year post-op medical management;g peripheral bypass graft;d average health care costs for 1 year;c cost of CAS and PAD treatment for 7 years;e cost of medical management of CAS similar to cost of medical management of PAD; 3% of patients will require amputationf|
|C13||Average cost of AAA diagnostic||$416||Cost of abdominal CT (with contrast)i|
|C14||Average cost of CAS diagnostic||$405||Cost of carotid duplex scanj|
|C15||Average cost of PAD diagnostic||$234||Cost of arterial evaluation of extremityi|
|C15A||Cost of bundled AAA, CAS, and PAD diagnostic screening||$110||Cost of bundled cardiovascular screening (includes AAA, CAS, and PAD)k|
|C16||Average cost of treated no AAA, no CAS, no PAD||$21,940||Assumes average health care costs for 10 yearsc|
|C17||Average cost of treated no AAA, no CAS, PAD||$27,725||Assumes: cost of 10 years of average PAD treatment costs;h 10% of patients will require percutaneous intervention for PAD and 1% will require amputationf|
|C18||Average cost of treated no AAA, moderate CAS, no PAD||$24,300||Assumes: cost of 10 years of average CAS treatment costs;e,h medical management of CAS similar in cost to medical management of PAD|
|C19||Average cost of treated no AAA, moderate CAS, PAD||$52,025||Assumes: cost of 10 years of average PAD and CAS treatment costs;e medical management of CAS is similar in cost to medical management of PAD; 10% of patients will require percutaneous intervention for PAD and 1% will require amputationf|
|C20||Average cost of treated no AAA, possibly significant CAS, no PAD||$38,719||Assumes: cost of endarterectomy and extensive 2-year post-op medical management;g average health care costs for 8 yearsc|
|C21||Average cost of treated no AAA, possibly significant CAS, PAD||$41,079||Assumes: cost of endarterectomy and extensive 2-year post-op medical management;g annual costs associated with PAD medical management;c,e 10% of patients will require percutaneous intervention for PAD and 1% will require amputationf|
|C22||Average cost of treated AAA, no CAS, no PAD||$49,985||Assumes: cost of elective AAA surgery;h 10 years of average health care costsc|
|C23||Average cost of treated AAA, no CAS, PAD||$55,270||Assumes: cost of elective AAA surgery;h 10 years of average PAD treatment costs;e 10% of patients will require percutaneous intervention for PAD and 1% will require amputationf|
|C24||Average cost of treated AAA, moderate CAS, no PAD||$51,845||Assumes: cost of elective AAA surgery;h 10 years of average CAS treatment costs;e medical management of CAS similar in cost to medical management of PAD|
|C25||Average cost of treated AAA, moderate CAS, PAD||$79,570||Assumes: cost of elective AAA surgery;h 10 years of average PAD and CAS treatment costs;e 10% of patients will require percutaneous intervention for PAD and 1% will require amputation;f medical management of CAS similar in cost to medical management of PAD|
|C26||Average cost of treated AAA, possibly significant CAS, no PAD||$62,839||Assumes: costs of endarterectomy, and extensive 2-year post-op medical management;g elective AAA surgery;h average annual health care costs for 8 yearsc|
|C27||Average cost of treated AAA, possibly significant CAS, PAD||$68,624||Assumes: costs of endarterectomy, and extensive 2-year post-op medical management;g elective AAA surgery;h 10 years of average PAD treatment costs;e 10% of patients will require percutaneous intervention for PAD and 1% will require amputationf|
|C28||Average cost of general practitioner office visit||$78||American Academy of Family Practicel|
|C29||Average cost of specialist physician visit||$366||Harvard Pilgrim Healthcarem|
|C30||Cost of surgical intervention for PAD||$22,584||Cost of peripheral bypass graftc|
|C31||Cost of ischemic amputation||$116,700||Cost of ischemic amputation and associated 2-year costsn|
|C32||Average cost of stroke, myocardial infarction, or CV-related death||$47,178||Assumes average of: cost of stroke (hospitalization costo and 90-day post-hospitalization costp); cost of myocardial infarction (hospitalization costq and 180-day post-hospitalization costq); and cost of CV-related deathr|
aAll costs assume a 10-year time period;
ball costs are in 2012 U.S. dollars;
ddata from Mahoney 2008;
edata from Hirsch 2008;
fdata from Fintel 2008;
gdata from Ryan 2009;
hdata from Lee 2002;
idata from Healthcare Blue Book Web site, 2012;
jdata from Cull 2011;
kdata from Life Line Screening (Independence, Ohio);
ldata from American Academy of Family Physicians Web site, 2004;
mdata from Harvard Pilgrim Health Care Web site, 2012;
ndata from Mackenzie 2007;
odata from Lyden 1996;
pdata from Kramer 2006;
qdata from Kauf 2006;
qdata from Sloss 2004;
rdata from Naccarelli 2010.
AAA=abdominal aortic aneurysm, CAS=carotid artery stenosis, CT=computed tomography, CV=cardiovascular, PAD=peripheral artery disease.
Because precise cost data for each variable were not available in the literature, many base case costs are assumption-driven. This is certainly a limitation to the model that should be addressed in future claims database analyses. Table 2 includes details regarding data sources and assumptions for each cost variable.
Finally, although costs would certainly vary by age and by gender, for the initial analysis we have assumed that costs are not age- or gender-dependent.
Death, MI, stroke, or other cardiovascular-related events
An individual’s Reynolds Risk Score (RRS) is the likelihood that the individual will experience death, MI, stroke, or other cardiovascular-related events within 10 years. An RRS is calculated utilizing a subject’s age, blood pressure, cholesterol levels, smoking status, high-sensitivity C-reactive protein (hsCRP) testing (a marker of inflammation), and parental history of heart attack. The inclusion of hsCRP and parental history in the calculation of the RRS has created an improved predictive model of cardiovascular risk compared with other existing models that do not take these factors in to consideration (Ridker 2007, Ridker 2008). Screening data provided mean RRS for patients with each possible combination of the presence (or absence) of AAA, moderate CAS, significant CAS, PAD, and smoking history. To arrive at the values in Table 3, the expected RRS was calculated using the screening data, weighted according to the model’s assumptions regarding current and past smoking status of model participants. Like Table 1, Table 3 provides only the subset of variables corresponding to the 2 submodels of men and of women aged 55 to 64 years.
|Table 3 Baseline Reynolds Risk Scores (RRS)a|
|Variable name||Variable description||Base case valuea|
55 to 64
55 to 64
|RR1||RRS for patient with untreated or treated no AAA, no CAS, no PAD||0.1092||0.0413|
|RR2||RRS for patient with untreated no AAA, no CAS, PAD||0.1551||0.0650|
|RR3||RRS for patient with untreated no AAA, moderate CAS, no PAD||0.1320||0.0622|
|RR4||RRS for patient with untreated no AAA, moderate CAS, PAD||0.1069||0.0585|
|RR5||RRS for patient with untreated no AAA, possibly significant CAS, no PAD||0.1541||0.0581|
|RR6||RRS for patient with untreated no AAA, possibly significant CAS, PAD||0.1850||0.1046|
|RR7||RRS for patient with untreated AAA, no CAS, no PAD||0.1376||0.0632|
|RR8||RRS for patient with untreated AAA, no CAS, PAD||0.0895||0.0177|
|RR9||RRS for patient with untreated AAA, moderate CAS, no PAD||0.2981||0.0622|
|RR10||RRS for patient with untreated AAA, moderate CAS, PAD||0.1069||0.0585|
|RR11||RRS for patient with untreated AAA, possibly significant CAS, no PAD||0.1403||0.0581|
|RR12||RRS for patient with untreated AAA, possibly significant CAS, PAD||0.2837||0.1046|
|RR Reduction||Percentage reduction in RRS due to treatment of AAA, CAS, and/or PAD||20%||20%|
|aLife Line Screening data, 2005–2008|
Individuals with PAD are at risk of amputation. For the current analysis, we utilized the following assumptions: individuals without PAD at baseline are not at risk of amputation in the 10-year model timeframe; individuals with unidentified PAD at baseline (i.e., patients with PAD who did not receive baseline screening) have a 3% likelihood of amputation within 10 years
(Fintel 2008; individuals with PAD identified by screening at baseline have a 1% likelihood of amputation within 10 years (Fintel 2008).
Distribution of catastrophic events
It is generally assumed that the number of catastrophic events, as defined previously, will increase from Year 1 through Year 10. The assumed annual distribution of catastrophic events is 5% for Years 1–3, 7.5% for Year 4, 10% for Year 5, 12.5% for Years 6–8, and 15% for Years 9 and 10.
Expected value analysis
The primary analysis utilizes average variable values (a subset of which are provided in Tables 1, 3, and 5) to determine: expected numbers of beneficiaries with AAA, moderate CAS, possibly significant CAS, and/or PAD who are either identified (bundled screening cohort) or whose disease remains latent (nonscreening or control cohort); expected number of disease occurrences receiving treatment through appropriate follow-up with the traditional health care system; and expected direct costs of the bundled screening and control cohorts.
Across all age and gender groups, the primary analysis shows an expected 2,580 beneficiaries (93.5%) without any of the diseases of interest, an expected 89 individuals (3.2%) with AAA, 54 individuals (2%) with CAS, and 52 cases (1.9%) of PAD.
The base case analysis depicts an opportunity afforded by bundled screening, namely, identification of latent disease, providing an opportunity for disease treatment. In the analysis, it is assumed that for the bundled screening cohort, 90% of patients with identified disease will seek appropriate medical treatment from the traditional health care system. Appropriate medical treatment is defined as follow-up with a primary care physician as an initial step to discuss screening results. This results in 80, 49, and 49 cases of AAA, CAS, and PAD, respectively, receiving treatment vs. no patients receiving treatment in the control cohort. Additionally, the model’s assumed RRS reduction of 20% due to treatment results in four fewer cardiovascular-related catastrophic health events over 10 years, a 2% reduction.
The modeled cost implications of early identification and treatment of AAA, CAS, and PAD are also determined from the base case analysis results. For patients in the control cohort, approximately $54.1 million in health care costs are expected over 10 years (or until they reach the age of 65). For the bundled screening cohort, health care costs of approximately $51.3 million are expected, representing a 5.2% reduction in health care spending.
Finally, Table 4 (below) provides an expected annual distribution of costs for the bundled screening and the control cohorts. These results take into account the appropriate longitudinal costs, as well as the assumed annual cost and event distributions. Table 4 also separates noncatastrophic event treatment costs, catastrophic event costs, and bundled screening costs (as applicable) to provide a more detailed breakdown of expected direct costs for the two modeled cohorts. From a budgetary impact perspective, the model predicts the hypothetical self-funded employer will recoup all screening costs and begin showing cost savings in the second year postscreening.
|Table 4 Bundled screening vs no screening: estimated annual costs for men and women ages 50–54 and ages 55–64|
|Year 1||Year 2||Year 3||Year 4||Year 5||Year 6||Year 7||Year 8||Year 9||Year 10|
|Bundled screening cohort|
|Noncatastrophic event costs||$3,222,214||$3,036,000||$2,849,785||$2,663,570||$4,954,711||$9,164,564||$8,419,706||$3,837,424||$3,464,994||$3,092,565|
|Catastrophic event costs||$474,190||$445,736||$417,283||$583,245||$720,752||$829,807||$758,673||$687,539||$739,686||$654,325|
|Bundled screening costs||$303,277||N/A||N/A||N/A||N/A||N/A||N/A||N/A||N/A||N/A|
|No screening cohort|
|Noncatastrophic event costs||$3,454,193||$3,251,942||$3,049,691||$2,847,440||$5,290,378||$9,771,754||$8,962,751||$4,076,873||$3,672,371||$3,267,870|
|Catastrophic event costs||$488,188||$458,788||$429,388||$599,982||$741,177||$852,973||$779,474||$705,974||$758,970||$670,771|
|Bundled screening costs||N/A||N/A||N/A||N/A||N/A||N/A||N/A||N/A||N/A||N/A|
|Savings due to bundled screening||–$57,301||$228,994||$212,011||$200,608||$356,092||$630,356||$563,846||$257,884||$226,661||$191,750|
To determine the robustness of the modeled cost and benefit findings, one-way sensitivity analyses were performed on three key model variables: (1) the likelihood of inappropriate health care follow-up for normal bundled screening findings (base value 0.2), (2) percent reduction in RRS for patients receiving appropriate disease treatment (base value 20%), and (3) the likelihood of appropriate health care follow-up for abnormal bundled screening results (base value 0.8).
Sensitivity analyses on the likelihood of inappropriate health care follow-up for normal bundled screening findings showed that modeled cost savings are robust to any level of inappropriate health care follow-up for normal screening results. Even if all bundled screening patients receiving normal results have an additional standard health care visit, the model still predicts over $2.5 million in cost savings over 10 years.
The modeled results are similarly robust to changes in the assumed percent reduction in RRS for patients receiving appropriate disease treatment. Even if no reduction in RRS is assumed, the model predicts 10-year cost savings of approximately $2.8 million for the bundled screening cohort.
Finally, even if only approximately 5% of patients receiving abnormal bundled screening results visit their physicians for repeat diagnostic testing to confirm abnormal findings and, as relevant, to initiate appropriate disease treatment, the modeled analysis still shows 10-year cost savings (or cost parity) for the bundled screening cohort. Modeled costs are identical for the bundled screening and the control cohort if only 4.8% of patients with latent disease receive appropriate medical treatment.
This cost-benefit analysis predicts significant, robust cost and health benefits associated with the decision to provide bundled cardiovascular disease screening to a cohort of self-funded employer’s beneficiaries aged ≥50 years and <65 years. As demonstrated, the modeled cost benefits are due to the avoidance of very costly emergency and/or major medical procedures once latent disease becomes symptomatic. While it is established that current, published cardiovascular screening recommendations do not all recommend AAA, CAS, and PAD screening in the cohort examined in this analysis, it is clear from the screening data that form the basis of the model that disease does exist in this cohort, and is identified by screening. Presence of disease implies not only direct cost benefits but also individual and public health benefits.
The precise magnitude of this benefit remains unknown, however, due to limitations of the initial analysis. First, while the bundled screening strategy resulted in cost savings even if very few patients with disease received appropriate disease treatment, the cost savings varied from approximately $2.8 million (if 90% of patients seek treatment) to approximately $6,000 (if only about 5% of patients seek treatment). For this reason, it is important for providers of bundled screening outside the traditional health care environment (e.g., community-based screening where participants self-select to screen) to perform follow-up surveys to gauge likelihood of physician follow-up following a positive finding of AAA, CAS, and/or PAD as well as likelihood of physician follow-up following negative findings.
To this end, the screening provider has conducted preliminary follow-up surveys to investigate postscreening interaction with health care providers, details of which are provided in the previously mentioned article by Weisman (2013). Seventy-five percent of survey respondents shared their screening results with a physician or health care provider (unweighted, n=1,477). Respondents with abnormal test results (defined as a positive test result for PAD, CAS, AAA, atrial fibrillation, or osteoporosis) (unweighted, n=524) were more likely to share results than those respondents with normal test results (82% vs 74%). If we utilize these preliminary survey results in our model, the predicted 10-year cost savings of approximately $2.55 million reinforces the cost savings potential of bundled cardiovascular screening for a cohort group aged ≥50 years and <65 years.
Second, while all health care costs were drawn from published sources, the timing and selection of treatment procedures across the patient groups are assumption-driven. Therefore, collection of post-screening medical resource utilization such as additional diagnostics, medications, office visits, emergency department visits, and hospitalizations also will be important. Similarly, post-screening behavioral changes such as modified diet, increased exercise, etc., should be collected to inform future cost-benefit analyses.
Third, the current analysis provides only initial assessments of the cost and health benefits of bundled disease screening to a single cohort of beneficiaries of a self-funded employer who assumes complete risk for its beneficiaries. However, the initial analysis serves an important purpose: to highlight the cost-saving potential of bundled screening for cardiovascular disease within an appropriately defined population. More rigorous cost-benefit analyses should be conducted from a number of perspectives (e.g., self-funded employer [including varying models of benefits such as high-deductible plans], self-pay population, private payer, CMS, societal).
Fourth, future cost-benefit analyses should compare the costs and benefits of employer-provided bundled CAS, AAA, and PAD screening to the costs and benefits of more traditional screening for cardiovascular disease and diabetes.
Additionally, future cost-benefit assessments should be built as Markov models, rather than as simple decision analyses, and should incorporate elements such as:
- Rolling patient populations, with bundled screening assessments provided for new individuals entering/leaving the model on an annual basis
- Discounting of future cost and health benefits
- Implications of repeated bundled screening at predefined intervals (e.g., 5 years or 10 years)
- Patients choosing to opt out of the bundled screening benefit
- Lifetime health care costs
- All-cause mortality
- Expected survival and/or quality-adjusted survival benefits for patients receiving bundled screening
- Calculation of either incremental cost per year of survival or incremental cost per quality-adjusted life years (QALYs) (or assessment of dominance if the more effective strategy also results in lower costs as is suggested by the current analysis)
- The impact of reductions in early compliance with necessary medication and/or lifestyle changes on the incremental benefits of bundled screening
- The impact of additional comorbid conditions (e.g., diabetes) on the occurrence of catastrophic events and on RRS reductions achievable through CAS, PAD, and/or AAA treatments only.
While a preliminary survey of post-screening health care follow-up has been conducted, in advance of further prospective survey data regarding postscreening medical resource utilization, analyses could utilize social science research to refine the likelihood of appropriate or inappropriate follow-up across the different age and gender subgroups. Claims-based data analyses could also be used to determine actual reduction in costs due to early treatment of latent disease as compared to later treatment of symptomatic disease.
Finally, while additional research is necessary to fully characterize the cost-benefit profile of community-based bundled cardiovascular screening, this initial analysis suggests that future, definitive cost-effectiveness analyses will demonstrate clear value in the provision of bundled screening to appropriately selected populations.
American Academy of Family Physicians. Median expenses per office-based physician visit, per office type. http://www.aafp.org/dam/AAFP/documents/media_center/charts-graphs/MedianExpense2004.pdf. Accessed Dec. 23, 2014.
Ajmera M, Raval AD, Shen C, Sambamoorthi U. Explaining the increased health care expenditures associated with gastroesophageal reflux disease among elderly Medicare beneficiaries with chronic obstructive pulmonary disease: a cost-decomposition analysis. Int J Chron Obstruct Pulmon Dis. 2014;8:339–348.
Austin GL, Fennimore B, Ahnen DJ. Can colonoscopy remain cost-effective for colorectal cancer screening? The impact of practice patterns and the Will Rogers phenomenon on costs. Am J Gastroenterol. 2013;108:296–301.
Barouni M, Larizadeh MH, Sabermahani A, Ghaderi H. Markov’s modeling for screening strategies for colorectal cancer. Asian Pac J Cancer Prev. 2012;13:5125–5129.
Bhattacharjee S, Sambamoorthi U. Co-occurring chronic conditions and healthcare expenditures associated with Parkinson’s disease: a propensity score matched analysis. Parkinsonism Relat Disord. 2013; 19:746–750.
Brilleman SL, Purdy S, Salisbury C, et al. Implications of comorbidity for primary care costs in the UK: a retrospective observational study. Br J Gen Pract. 2013;63(609):e274–282.
Brott TG, Halperin JL, Abbara S, et al. 2011 ASA/ACCF/AHA/AANN/AANS/ACR/ASNR/CNS/SAIP/SCAI/SIR/SNIS/SVM/SVS guideline on the management of patients with extracranial carotid and vertebral artery disease. Stroke. 2011;42:e464-e540.
CDC (U.S. Centers for Disease Control and Prevention). National Center for Health Statistics. Health Behaviors of Adults: United States, 2008–2010. http://www.cdc.gov/nchs/data/series/sr_10/sr10_257.pdf. Aug. 14, 2013. Accessed Dec. 23, 2014.
Chaikof EL, Brewster DC, Dalman RL, et al; Society for Vascular Surgery. The care of patients with an abdominal aortic aneurysm: the Society for Vascular Surgery practice guidelines. J Vasc Surg. 2009;50(4 Suppl):S2–S49.
Choi S, Lee S, Matejkowski J, Baek YM. The relationships among depression, physical health conditions and healthcare expenditures for younger and older Americans. J Ment Health. 2014;23:140–145.
Cull DL, Cole T, Miller B, et al. The value of a carotid duplex surveillance program for stroke prevention. Ann Vasc Surg. 2011;25:887–894.
DeLorenze GN, Tsai AL, Horberg MA, Quesenberry CP. Cost of care for HIV-infected patients with co-occurring substance use disorder or psychiatric disease: Report from a large, integrated health plan. AIDS Res Treat. 2014; 2014: 570546. doi: 10.1155/2014/570546. Epub June 22, 2014.
DeRubertis BG, Trocciola SM, Ryer EJ, et al. Abdominal aortic aneurysm in women: prevalence, risk factors, and implications for screening. J Vasc Surg. 2007;46:630–635.
Eddy DM. Screening for breast cancer. Ann Intern Med. 1989;111:389–399.
Fintel DJ. The burden of PAD: diagnosis and contemporary medical management through risk-factor modification. Internet J Cardiol. 2008;6(2). DOI: 10.5580/1faa.
Gold MR, ed. Cost-effectiveness in Health and Medicine. New York: Oxford University Press; 1996.
Goldstein LB, Bushnell CD, Adams RJ, et al. Guidelines for the primary prevention of stroke: a guideline for healthcare professionals from the American Heart Association/American Stroke Association. Stroke. 2011;42:517–584.
Goulart BH, Ramsey SD. Moving beyond the national lung screening trial: discussing strategies for implementation of lung cancer screening programs. Oncologist. 2013;18:941–946.
Greenland P, Alpert JS, Beller GA, et al; American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. 2010 ACCF/AHA guideline for assessment of cardiovascular risk in asymptomatic adults: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. Circulation. 2010;122:e584–e636.
Harvard Pilgrim Health Care Web site. https://www.harvardpilgrim.org. Accessed May, 2014.
Healthcare Bluebook. Abdominal CT (with contrast). https://www.healthcarebluebook.com/page_Results.aspx?id=157&dataset=MD&g=Abdominal+CT+(with+contrast). Accessed Dec. 23, 2014.
Healthcare Bluebook. Arterial Evaluation of Extremity. https://www.healthcarebluebook.com/page_Results.aspx?id=200&dataset=MD&g=Arterial+Evaluation+of+Extremity. Accessed Dec. 23, 2014.
Hirsch AT, Hartman L, Town RJ, Virnig BA. National health care costs of peripheral arterial disease in the Medicare population. Vasc Med. 2008;13:209–215.
Hirsch AT, Haskal ZJ, Hertzer NR, et al. ACC/AHA 2005 Practice Guidelines for the management of patients with peripheral arterial disease (lower extremity, renal, mesenteric, and abdominal aortic). Circulation. 2006;113:e463–e654.
Jacobowitz GR, Rockman CB, Gagne PJ, et al. A model for predicting occult carotid artery stenosis: Screening is justified in a selected population. J Vasc Surg. 2003;38:705–709.
Joyce AT, Smith P, Khandker R, et al. Hidden cost of rheumatoid arthritis (RA): estimating cost of comorbid cardiovascular disease and depression among patients with RA. J Rheumatol. 2009;36:743–752.
Kauf TL, Velazquez EJ, Crosslin DR, et al. The cost of acute myocardial infarction in the new millennium: evidence from a multinational registry. Am Heart J. 2006;151:206–212.
Kramer A, Holthaus D, Goodrish, G, Epstein A. U.S. Department of Health and Human Services. Office of Disability, Aging and Long-Term Care Policy. A Study of Stroke Post-Acute Care Costs and Outcomes: Final Report. December 2006. http://aspe.hhs.gov/daltcp/reports/2006/strokePAC.pdf. Accessed Dec. 23, 2014.
Lee TT, Solomon NA, Heidenreich PA, et al. Cost-effectiveness of screening for carotid stenosis in asymptomatic persons. Ann Intern Med. 1997;126:337–346.
Lee TY, Korn P, Heller JA, et al. The cost-effectiveness of a “quick-screen” program for abdominal aortic aneurysms. Surgery. 2002;132:399–407.
LeFevre ML; U.S. Preventive Services Task Force. Screening for abdominal aortic aneurysm: U.S. Preventive Services Task Force recommendation statement. Ann Intern Med. 2014a;161:281–290.
LeFevre ML; U.S. Preventive Services Task Force. Screening for asymptomatic carotid artery stenosis: U.S. Preventive Services Task Force recommendation statement. Ann Intern Med. 2014b;161:356–362.
Lim LS, Haq N, Mahmood S, Hoeksema L; ACPM Prevention Practice Committee; American College of Preventive Medicine. Atherosclerotic cardiovascular disease screening in adults: American College of Preventive Medicine position statement on preventive practice. Am J Prev Med. 2011;40:381.
Lyden PD. Magnitude of the problem of stroke and the significance of acute intervention. In: Marler JR, Jones PW, Emr M, eds. Proceedings of a National Symposium on Rapid Identification and Treatment of Acute Stroke; 1996 Dec 12–13. Bethesda, Md.: National Institutes of Health; 1997.
MacKenzie EJ, Jones AS, Bosse MJ, et al. Health-care costs associated with amputation or reconstruction of a limb-threatening injury. J Bone Joint Surg Am. 2007;89:1685–1692.
Mahoney EM, Wang K, Cohen DJ, et al; REACH Registry Investigators. One-year costs in patients with a history of or at risk for atherothrombosis in the United States. Circ Cardiovasc Qual Outcomes. 2008;1:38–45.
Mark DB, Shaw LJ, Lauer MS, et al. 34th Bethesda Conference: Task force #5 — is atherosclerosis imaging cost effective? J Am Coll Cardiol. 2003;41:1906–1917.
Meenan RT, Fleming C, Whitlock EP, et al. Cost-Effectiveness Analyses of Population-Based Screening for Abdominal Aortic Aneurysm: Evidence Synthesis. Guide to Clinical Preventive Services, Third Edition: Periodic Updates. February 2005; AHRQ Pub. No. 05–0569-C.
Moyer VA, U.S. Preventive Services Task Force. Screening for peripheral artery disease and cardiovascular disease risk assessment with the ankle-brachial index in adults: U.S. Preventive Services Task Force recommendation statement. Ann Intern Med. 2013;159:342–348.
Naccarelli GV, Johnston SS, Lin J, et al. Cost burden of cardiovascular hospitalization and mortality in ATHENA-like patients with atrial fibrillation/atrial flutter in the United States. Clin Cardiol. 2010;33:270–279.
O’Malley AS, Lawrence W, Liang W, et al. Feasibility of mobile cancer screening and prevention. J Health Care Poor Underserved. 2002;13:298–319.
Pharoah PD, Sewell B, Fitzsimmons D, et al. Cost effectiveness of the NHS Breast Screening Programme: life table model. BMJ. 2013;346:f3822
Ridker PM, Buring JE, Rifai N, Cook NR. Development and validation of improved algorithms for the assessment of global cardiovascular risk in women: the Reynolds Risk Score. JAMA. 2007;297:611–619. Erratum in: JAMA. 2004;297:1433.
Ridker PM, Paynter NP, Rifai N, et al. C-reactive protein and parental history improve global cardiovascular risk prediction: the Reynolds risk score for men. Circulation. 2008;118:2243–2251, 4p following 2251.
Rooke TW. Controversies in vascular screening — art versus science. Vasc Med. 2007;12:235–242.
Rooke TW, Hirsch AT, Misra S, et al; American College of Cardiology Foundation; American Heart Association Task Force; Society for Cardiovascular Angiography and Interventions; Society of Interventional Radiology; Society for Vascular Medicine; Society for Vascular Surgery. 2011 ACCF/AHA focused update of the guideline for the management of patients with peripheral artery disease (updating the 2005 guideline). Vasc Med. 2011;16:452–476.
Ryan J, Linde-Zwirble W, Engelhart L, et al. Temporal changes in coronary revascularization procedures, outcomes, and costs in the bare-metal stent and drug-eluting stent eras: results from the US Medicare program. Circulation. 2009;119:952–961.
Silverstein MD, Pitts SR, Chaikof EL, Ballard DJ. Abdominal aortic aneurysm (AAA): cost-effectiveness of screening, surveillance of intermediate-sized AAA, and management of symptomatic AAA. Proc (Bayl Univ Med Cent). 2005;18:345–367.
Singh K, Bønaa KH, Jacobsen BK, et al. Prevalence of and risk factors for abdominal aortic aneurysms in a population-based study: the Tromsø Study. Am J Epidemiol. 2001;154:236–244.
Sloss EM, Wickstrom SL, McCaffrey DF, et al. Direct medical costs attributable to acute myocardial infarction and ischemic stroke in cohorts with atherosclerotic conditions. Cerebrovasc Dis. 2004;18:8–15.
Søgaard R, Laustsen J, Lindholt JS. Cost effectiveness of abdominal aortic aneurysm screening and rescreening in men in a modern context: evaluation of a hypothetical cohort using a decision analytical model. BMJ. 2012;345:e4276.
U.S. Bureau of Labor Statistics. Employee tenure in 2014. http://www.bls.gov/news.release/pdf/tenure.pdf. Sept 18, 2014. Accessed Dec. 23, 2013.
U.S. Census Bureau. Age and Sex Composition in the United States: 2011. http://www.census.gov/population/age/data/files/2011/2011gender_table1.xlsx. Updated November 28, 2012. Accessed Dec. 23, 2014.
Weisman SM, Manganaro AJ. Community-based screening: Identifying risk and motivating healthy lifestyle changes. Postgrad Med. 2013;125:18–27.
Funding source: Life Line Screening, LLC, provided the data sets used for this study, funded its research and writing, participated in study design, and had editorial input on the article’s content.
Conflict disclosure: Steven Weisman and Elizabeth Brooks are consultants to Life Line Screening and have received compensation for their efforts.
Steven M. Weisman, PhD
67 Park Place East
Morristown, NJ 07960
Phone: (973) 889-1600 x101
Fax: (973) 889-1619
More like this
- A Conversation With François de Brantes: Bundled Payment’s Many Challenges
- What's Fair In Bundled Payment Contracting?
- Clinical and Economic Outcomes of the Cincinnati Pharmacy Coaching Program for Diabetes and Hypertension
- Hospitals embrace bundled payment systems
- Is Bundled Payment an Idea Whose Time Has Come?