Accountable Care Organizations and Use of Surgery Among Patients With Alzheimer Disease and Related Dementias

ABSTRACT

Objective: To understand the effects of accountable care organizations (ACOs) on use of surgery in patients with Alzheimer disease and related dementias (ADRD).

Study Design: Retrospective national cohort study of all Medicare beneficiaries identified in a 20% sample between 2010 and 2017. The primary exposure was participation in ACOs. The primary outcome was use of 1 of 6 common surgical procedures (aortic valve replacement [AVR], abdominal aortic aneurysm [AAA] repair, colectomy, carotid artery repair, major joint repair, and prostatectomy).

Methods: Multivariable logistic regression models were fit using beneficiary-year as the unit of analysis to estimate the likelihood of undergoing each procedure among patients with ADRD and without ADRD, stratified by ACO participation. Additional models were fit to determine how the relationship between ACO participation and surgery was altered based on procedure urgency and the availability of minimally invasive technology.

Results: Adjusted odds for use of surgery were lower among patients with ADRD compared with patients without ADRD for all procedures. ACO participation had varying impact on patients with ADRD, with higher odds of AVR and major joint surgery and lower odds of carotid artery repair. Availability of minimally invasive technology increased the likelihood of AVR and AAA repair among patients with ADRD; however, ACO participation reduced these effects. The effect of ACO participation on the likelihood of undergoing surgery did not vary by urgency of the procedure.

Conclusions: The likelihood of undergoing surgery is overall lower among patients with ADRD and may vary by ACO participation for specific procedures.

Am J Manag Care. 2023;29(7):349-355. https://doi.org/10.37765/ajmc.2023.89395

Takeaway Points

• The likelihood of undergoing 6 common surgical procedures (aortic valve replacement [AVR], abdominal aortic aneurysm [AAA] repair, colectomy, carotid artery repair, major joint repair, and prostatectomy) is lower among patients with Alzheimer disease and related dementias (ADRD).
• Accountable care organization (ACO) participation attenuated the decreased likelihood of undergoing AVR and major joint surgery among patients with ADRD. Conversely, ACO participation further decreased the likelihood of undergoing carotid artery repair.
• Availability of minimally invasive technology tended to increase the likelihood of undergoing AVR and AAA repair among patients with ADRD. ACO participation attenuated these effects.

The prevalence of Alzheimer disease and related dementias (ADRD) is rapidly increasing and is expected to triple by 2050 to 13 million individuals in the United States.¹ Understanding how this growing population interacts with the health care system is becoming increasingly important as annual spending for managing these patients exceeds $200 billion nationally.² This spending in part stems from increased hospital utilization, which is further complicated by the presence of dementia, resulting in higher rates of adverse events and mortality.^3,4 Surgery poses additional challenges for this population due to operative risk, effects of anesthesia, and postoperative delirium.^5-7 Considering these consequences, the decision to offer surgery to patients with ADRD is not trivial. Providers must balance improvement in quality of life with subsequent surgical morbidity.

The shift in health care delivery toward value-based payment models may affect surgical decision-making among patients with ADRD. Accountable care organizations (ACOs)—which are held to spending and quality standards—would theoretically incentivize providers to be judicious when considering surgery in patients at higher risk of complications.⁵ However, it is unclear how accountability implied by ACO participation affects use of surgery in patients with ADRD. Physicians in ACOs could be more selective in several ways. First, to reduce the risk of morbidity, ACO providers may be more likely to offer surgery in this vulnerable population if less-invasive and better-tolerated options are available, such as minimally invasive procedures.^8,9 Second, physicians may be more selective in choosing to whom surgery is offered. ADRD varies in severity, with some patients requiring nursing home care whereas others reside in the community.¹⁰ To reduce morbidity and associated spending, ACO providers may selectively offer surgery to community-dwelling patients with ADRD. Third, incentives rewarding value may motivate physicians to avoid surgery altogether in this high-risk population. Finally, we expect that procedure indications (eg, emergent, palliative) would mitigate, at least in part, any differences in use of surgery between those with and without ADRD.

To better understand implications of ACOs for patients with ADRD, we performed a national study using Medicare data to assess the effect of participation on rates of major surgery across common conditions. We hypothesized that ACOs would attenuate rates of surgery. We expected such attenuation to be muted by a greater local availability of minimally invasive surgery and by underlying indications (eg, emergent vs elective).

METHODS

Data and Study Population

We performed a retrospective national cohort study of all Medicare beneficiaries between 2010 and 2017 within a 20% sample who were 66 years or older to allow for 1-year look-back. Patients were included if they had continuous enrollment in both Medicare parts A and B throughout the study period. Patients participating in Medicare Advantage plans were excluded due to absence of complete claims data and ineligibility for participation in Shared Savings Plan ACOs. Beneficiary-years were excluded (4.3%) if the beneficiary did not receive any primary care in a given study year because they would be considered ineligible for ACO alignment. We identified beneficiaries diagnosed with ADRD using International Classification of Diseases, Ninth Revision and Tenth Revision codes (eAppendix Table [available at ajmc.com]).¹¹

Exposure

Our primary exposure was Medicare Shared Savings Program ACO participation. To determine ACO participation, we first assigned patients to a primary care physician based on the plurality of evaluation and management codes for primary care services during the study period.¹² Beneficiaries were attributed to an ACO if their primary care physician was aligned with a Medicare Shared Savings Program ACO based on the ACO alignment file. A time varying indicator was defined and set to 1 during a given year if the patient’s primary care physician was aligned with an ACO that year and 0 otherwise. For 2012, there were 2 ACO start dates, April 1 and July 1. For those who underwent surgery, ACO participation was defined relative to the date of surgery. For example, if surgery occurred prior to the ACO start date, the beneficiary was characterized as not participating in an ACO (indicator set to 0). For those who did not undergo surgery, random allocation to ACO participation was performed based on ACO start date to avoid bias. For example, for ACOs starting on April 1, 75% of patients who did not undergo surgery were randomly allocated as ACO participants and 25% were allocated as not participating in an ACO; these percentages represent the portion of the year after and prior to the start date. For ACOs beginning July 1, 50% of beneficiaries who did not undergo surgery were randomly allocated as ACO participants and 50% were allocated as not participating in an ACO. Our stratification resulted in 4 groups of patients for comparisons: ACO participants with ADRD, ACO participants without ADRD, those with ADRD not participating in ACOs, and those without ADRD not participating in ACOs.

Outcome

Our primary outcome was use of 1 of 6 surgical procedures, identified using Healthcare Common Procedure Coding System codes (eAppendix Table). The surgical procedures of interest were aortic valve replacement (AVR) (open or transcatheter AVR [TAVR]), abdominal aortic aneurysm (AAA) repair (open or endovascular repair), colectomy (open or laparoscopic/robotic), carotid artery repair (open or endovascular), major joint repair (total hip replacement or total knee replacement), and prostatectomy (open or laparoscopic/robotic). These procedures were chosen because they are relatively common, have varying degrees of invasiveness, encompass several surgical disciplines, and span a spectrum of potential benefits (eg, improving quality of life, lifesaving, palliative, purely elective). Thus, there were 6 separate cohorts of patients, 1 for each procedure. The denominator for each cohort included all beneficiaries identified within the 20% sample.

Analysis

For each cohort, we assessed for differences in patient characteristics between ACO participants and nonparticipants who underwent surgery using χ² test for categorical variables and Wilcoxon rank-sum test for continuous variables.

We examined the effect of ACO participation on use of surgery among beneficiaries with and without ADRD using multivariable logistic regression models. Beneficiary-year was our primary unit of analysis. We fit separate models for each procedure, adjusting for year, age, gender, race, socioeconomic status, nursing home status 90 days prior to procedure (June 30 was used as a referent date for those who did not undergo surgery), and Hierarchical Condition Category (HCC) risk score. HCC risk score was used as a measure of medical complexity,¹³ and socioeconomic status was measured at the zip code level using established methods.¹⁴ Given underlying incentives to reduce costs, particularly in at-risk patients, we expected that ACOs would attenuate use of surgery in each cohort in patients with ADRD.

Next, we examined the effect of 2 potential mitigating factors in this relationship—local availability of technology (ie, less invasive approaches to a surgery; also called technology diffusion) and procedure indication (ie, emergent vs elective defined within Medicare claims). We hypothesized that greater local availability of technology and an emergent indication would attenuate any differences noted between those with and without ADRD, independent of ACO participation. We defined local availability of technology as use of a less invasive surgical approach for a given procedure (eg, TAVR for aortic valve replacement) in a hospital referral region (HRR). This variable was defined as the percent of cases of a given surgery of interest performed with new technology in an HRR for a given year. This was treated as a continuous variable and lagged 1 year so that local availability of technology in 1 year served as the exposure for the cohort in the following year.

Analyses were carried out using SAS version 9.4 (SAS Institute) statistical software. All tests were 2-sided and the probability of type I error was 0.05. The study was deemed exempt by the institutional review board of the University of Michigan.

RESULTS

We identified a total of 8,309,944 beneficiaries meeting eligibility criteria between 2010 and 2017. Of these, 740,334 underwent one of the procedures, with 121,011 (17%) participating in an ACO at the time of surgery. Table 1 [part A and part B] compares patient characteristics according to ACO participation and type of procedure. In general, those in ACOs were in higher socioeconomic strata and more often underwent a procedure using a minimally invasive approach, when applicable (P < .001). ACO participants were also less likely to undergo emergent surgery for all procedures (P < .001).

Use of surgery was lower among patients with ADRD compared with patients without ADRD for all procedures (Table 2). ACO participation modestly altered this effect for certain procedures. For example, relative to patients without ADRD not participating in ACOs, the odds of undergoing AVR for patients with ADRD decreased by 48% for those not participating in ACOs compared with 36% for those participating in ACOs. Similarly, odds of undergoing major joint surgery decreased by 67% for those not participating in ACOs and 62% for those participating in ACOs. Conversely, relative to patients without ADRD not participating in ACOs, the odds of undergoing carotid artery repair for patients with ADRD decreased by 20% for those not participating in ACOs and by 28% for those participating in ACOs. ACO participation did not alter the likelihood of patients with ADRD undergoing AAA repair, colectomy, or prostatectomy.

The impact of technology diffusion on utilization of surgery was mixed (Table 3). Technology diffusion increased the likelihood of receiving AVR and AAA repair but had no effect on use of colectomy or carotid artery repair for patients with ADRD. ACO participation reduced these effects among patients with ADRD undergoing AVR or AAA repair. For example, among patients with ADRD, greater diffusion of technology increased the odds of undergoing AVR by 103% for those not participating in ACOs vs 44% for those participating in ACOs. Similarly, technology diffusion increased the odds of AAA repair in patients with ADRD not participating in ACOs by 61% but had no significant effect among those participating in ACOs. Among patients without ADRD, technology diffusion only increased likelihood of prostatectomy.

We then assessed whether the likelihood of surgery among patients with ADRD differed between emergent and elective procedures. Compared with patients without ADRD, those with ADRD were less likely to undergo all procedures, regardless of urgency (Figure). The extent to which the odds decreased varied by procedure and urgency. For example, patients with ADRD were more likely to undergo AAA repair, colectomy, major joint surgery, or carotid artery repair if the procedure was emergent as opposed to elective. This difference was most pronounced for major joint surgery (69% decreased odds for elective vs 37% for emergent). Odds of undergoing AVR or prostatectomy did not significantly differ between emergent and elective procedures.

Although we found that the likelihood of receiving elective or emergent surgery was lower among patients with ADRD, we sought to understand how ACOs may moderate utilization in both settings. As shown in Table 4, ACO participation did not differentially affect likelihood of emergent and elective surgery among patients with ADRD (eg, direction of effects was in the same direction). Specifically, ACOs increased the likelihood of patients with ADRD undergoing both emergent and elective AVR and major joint surgery. Conversely, ACO participation decreased the likelihood of undergoing both emergent and elective carotid artery repair. ACOs did not significantly affect likelihood of patients with ADRD undergoing emergent or elective colectomy or AAA repair.

DISCUSSION

The likelihood of undergoing surgery for patients with ADRD was generally lower than that for those without ADRD. Participation in ACOs resulted in a mitigation of the decreased likelihood of undergoing AVR and major joint surgery while further decreasing the likelihood of undergoing carotid artery repair. These findings held true despite the degree of penetration of minimally invasive technology and procedure urgency.

Reducing low-value health care in patients with ADRD aligns with policies underlying ACOs that aim to reduce spending and improve quality (eg, by avoiding unnecessary care).¹⁵ With 1-year mortality approaching 20% among those with ADRD in Medicare, it is not surprising that their likelihood of receiving surgery was lower.¹⁰ However, the effect of ACO participation on use of surgery varied by procedure. We found the reduction in odds was less pronounced for ACO participants undergoing AVR and major joint surgery and more pronounced for those undergoing carotid artery repair, compared with patients with ADRD not participating in ACOs. Such mixed effects on utilization may reflect more selective use of surgery in patients with ADRD by ACO participants. Whereas some procedures are palliative and aimed at improving quality of life, others may lack sufficient benefit to outweigh the high perioperative risk of complications and mortality in the ADRD population.¹⁶ For instance, aortic stenosis can result in debilitating symptoms with a significant reduction in quality of life and result in multiple hospitalizations if left untreated.¹⁷ Similarly, hip fracture is 3-fold more common among those with dementia and results in a 32% increase in mortality at 6 months.^18,19 Thus, surgeries to address aortic stenosis and acute hip fracture have the potential to improve quality of life and mitigate hospitalization risk.

Although patients with ADRD are overall less likely to receive surgery and ACO participation partially mitigates this effect for some procedures, surgical technology has the potential to facilitate use of surgery in this high-risk population. Prior work has shown that rates of AVR have increased substantially in ADRD populations, largely driven by diffusion of the minimally invasive approach of TAVR.¹¹ Technology implied by less-invasive alternatives (eg, laparoscopy, robotics, endoscopic, or endovascular), relative to conventional, open approaches, is often associated with decreased requirement for hospitalization, shorter length of hospitalization if required, quicker recovery, and lower morbidity.²⁰ Such technology may allow for expanding indications to a population previously deemed unfit for conventional open surgery, such as those with high medical complexity.²¹ Therefore, diffusion of these technologies can affect decision-making, particularly in markets with high penetration. In such instances, it is conceivable that minimally invasive approaches may result in more liberal utilization of a procedure, even in clinical circumstances in which benefits are less clear. Interestingly, we found that likelihood of AVR and AAA repair in patients with ADRD is higher in areas with increased use of minimally invasive approaches for each procedure. Participation in an ACO reduced this effect for AVR and eliminated the effect for AAA repair. These findings build on prior work that has shown that rates of adoption of new surgical technologies for many procedures (including those in this study) were similar between ACO hospitals and non-ACO hospitals.^22,23 TAVR appears to be cost-effective and safe in patients with ADRD, providing justification for its adoption in this high-risk population.²⁴ Conversely, cost-effectiveness and benefits of endovascular aortic aneurysm repair in older patients are less clear.^25,26 In this instance, perioperative risk may be larger than the risk reduction in aneurysm rupture achieved by surgical correction.^27,28 Interestingly, areas with higher diffusion of robotic prostatectomy increased the likelihood of surgery for patients without ADRD but had no effect among patients with ADRD. This is reassuring because candidates for prostate cancer treatment are those with greater than 10 years of life expectancy, which many patients with ADRD may not have.²⁹

Another factor that could affect use of surgery in patients with ADRD is the imperative nature of the surgery (ie, whether it is emergent or elective). Regardless of procedure urgency, patients with ADRD were less likely to receive surgery compared with patients without ADRD. ACO participation did not differentially affect the odds of undergoing emergent or elective surgery. Reduction in odds of surgery for patients with ADRD compared with those without ADRD was more pronounced for elective procedures compared with emergent procedures. Major joint surgery had the largest difference in odds between emergent and elective surgery (69% reduced odds for elective vs 37% reduced odds for emergent). This is likely due to reasons mentioned earlier; emergent joint repair may reflect a means to provide immediate palliation of symptoms whereas elective repair may not be worth exposure to perioperative risks.⁶ There are several reasons why even emergency surgery may be reduced among patients with ADRD. Typically, emergent procedures are performed when there is an imminent threat to life. However, in patients with dementia, retrospective series have demonstrated significantly high postoperative mortality rates, almost double that for healthy patients, following emergent surgery.^30,31 Given the reduced life expectancy of patients with ADRD, emergent surgery may not prolong life as it may in a healthier population.³¹ Additionally, patients with dementia often present with atypical symptoms that can delay diagnosis to a point at which even emergent surgery would be futile.³² Furthermore, patients and caregivers of those with dementia may choose to avoid intensive care. Qualitative and observational analyses have shown that most families choose comfort as the primary goal among nursing home residents with dementia.³³ Therefore, even emergent surgery may not be consistent with goals of care for these patients.

Limitations

Our findings must be interpreted in the context of several limitations. First, the underlying pathology leading to use of the surgical procedures can have varying severity, which we are unable to account for using claims. We mitigate this issue by stratifying our analysis by emergent and elective procedures, which provides some context for the nature of the disease. Similarly, patients with ADRD have varying degrees of ADRD severity, which we cannot directly capture. Our models do adjust for several markers of comorbidity, including nursing home stay within 90 days of surgery and HCC risk. Second, we define ACO participation by patients’ primary care physician, and the surgeon or hospital in which surgery was performed may not necessarily be affiliated with the ACO. However, our goal was to understand differential decision-making among patients with ADRD participating in ACOs and those not participating in ACOs. The initial decision point starts with primary care providers within ACOs who guide care and make referrals to specialists for consideration of surgery. Third, ACO participants and nonparticipants vary by demographic characteristics, and although we attempt to adjust for these factors within our models, there remains potential for unmeasured regional and socioeconomic variation.

CONCLUSIONS

This study finds that utilization of surgery is overall lower among patients with ADRD. Utilization appears to be influenced by ACO participation for some procedures, regardless of the degree of technology diffusion or the urgency of the procedure, in this high-risk group.

Author Affiliations: Division of Health Services Research, Department of Urology (AM, RLD, BKH), and Division of Geriatric Medicine, Department of Medicine (JPWB), University of Michigan, Ann Arbor, MI.

Source of Funding: This work was supported by the National Institute on Aging (NIA) grant R01 AG048071-04S1.

Author Disclosures: Drs Bynum and Hollenbeck report receiving a grant (NIA R01 AG048071-04S1) to support this work. The remaining authors report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.

Authorship Information: Concept and design (AM, RD, JPWB, BKH); acquisition of data (RD, JPWB, BKH); analysis and interpretation of data (AM, RD, BKH); drafting of the manuscript (AM); critical revision of the manuscript for important intellectual content (AM, RD, JPWB); statistical analysis (RD); obtaining funding (JPWB); and supervision (JPWB).

Address Correspondence to: Avinash Maganty, MD, MS, Division of Health Services Research, Department of Urology, University of Michigan, 2800 Plymouth Rd, Bldg 16, Ann Arbor, MI 48109-2800. Email: avmagant@med.umich.edu.

REFERENCES

1. Hebert LE, Scherr PA, Bienias JL, Bennett DA, Evans DA. Alzheimer disease in the US population: prevalence estimates using the 2000 Census. Arch Neurol. 2003;60(8):1119-1122. doi:10.1001/archneur.60.8.1119

2. Hurd MD, Martorell P, Delavande A, Mullen KJ, Langa KM. Monetary costs of dementia in the United States. N Engl J Med. 2013;368(14):1326-1334. doi:10.1056/NEJMsa1204629

3. Reynish EL, Hapca SM, De Souza N, Cvoro V, Donnan PT, Guthrie B. Epidemiology and outcomes of people with dementia, delirium, and unspecified cognitive impairment in the general hospital: prospective cohort study of 10,014 admissions. BMC Med. 2017;15(1):140. doi:10.1186/s12916-017-0899-0

4. Rao A, Suliman A, Vuik S, Aylin P, Darzi A. Outcomes of dementia: systematic review and meta-analysis of hospital administrative database studies. Arch Gerontol Geriatr. 2016;66:198-204. doi:10.1016/j.archger.2016.06.008

5. Hu CJ, Liao CC, Chang CC, Wu CH, Chen TL. Postoperative adverse outcomes in surgical patients with dementia: a retrospective cohort study. World J Surg. 2012;36(9):2051-2058. doi:10.1007/s00268-012-1609-x

6. Jämsen E, Peltola M, Puolakka T, Eskelinen A, Lehto MUK. Surgical outcomes of hip and knee arthroplasties for primary osteoarthritis in patients with Alzheimer’s disease: a nationwide registry-based case-controlled study. Bone Joint J. 2015;97-B(5):654-661. doi:10.1302/0301-620X.97B5.34382

7. White EM, Smith JG, Trotta RL, McHugh MD. Lower postsurgical mortality for individuals with dementia with better-educated hospital workforce. J Am Geriatr Soc. 2018;66(6):1137-1143. doi:10.1111/jgs.15355

8. Jackson RS, Chang DC, Freischlag JA. Comparison of long-term survival after open vs endovascular repair of intact abdominal aortic aneurysm among Medicare beneficiaries. JAMA. 2012;307(15):1621-1628. doi:10.1001/jama.2012.453

9. Fleshman J, Sargent DJ, Green E, et al; Clinical Outcomes of Surgical Therapy Study Group. Laparoscopic colectomy for cancer is not inferior to open surgery based on 5-year data from the COST Study Group trial. Ann Surg. 2007;246(4):655-662; discussion 662-664. doi:10.1097/SLA.0b013e318155a762

10. Koller D, Hua T, Bynum JPW. Treatment patterns with antidementia drugs in the United States: Medicare cohort study. J Am Geriatr Soc. 2016;64(8):1540-1548. doi:10.1111/jgs.14226

11. Hollenbeck BK, Dunn RL, Sukul D, et al. Aortic valve replacement among patients with Alzheimer’s disease and related dementias. J Am Geriatr Soc. 2021;69(12):3468-3475. doi:10.1111/jgs.17432

12. Borza T, Oerline MK, Skolarus TA, et al. Association between hospital participation in Medicare Shared Savings Program accountable care organizations and readmission following major surgery. Ann Surg. 2019;269(5):873-878. doi:10.1097/SLA.0000000000002737

13. Pope GC, Kautter J, Ellis RP, et al. Risk adjustment of Medicare capitation payments using the CMS-HCC model. Health Care Financ Rev. 2004;25(4):119-141.

14. Jacobs BL, Zhang Y, Schroeck FR, et al. Use of advanced treatment technologies among men at low risk of dying from prostate cancer. JAMA. 2013;309(24):2587-2595. doi:10.1001/jama.2013.6882

15. Schwartz AL, Chernew ME, Landon BE, McWilliams JM. Changes in low-value services in year 1 of the Medicare Pioneer accountable care organization program. JAMA Intern Med. 2015;175(11):1815-1825. doi:10.1001/jamainternmed.2015.4525

16. Modi PK, Kaufman SR, Borza T, et al. Medicare accountable care organizations and use of potentially low value procedures. Surg Innov. 2019;26(2):227-233. doi:10.1177/1553350618816594

17. Lauck SB, Gibson JA, Baumbusch J, et al. Transition to palliative care when transcatheter aortic valve implantation is not an option: opportunities and recommendations. Curr Opin Support Palliat Care. 2016;10(1):18-23. doi:10.1097/SPC.0000000000000180

18. Liu Y, Wang Z, Xiao W. Risk factors for mortality in elderly patients with hip fractures: a meta-analysis of 18 studies. Aging Clin Exp Res. 2018;30(4):323-330. doi:10.1007/s40520-017-0789-5

19. Petersen JD, Siersma VD, Wehberg S, Nielsen CT, Viberg B, Waldorff FB. Clinical management of hip fractures in elderly patients with dementia and postoperative 30-day mortality: a population-based cohort study. Brain Behav. 2020;10(11):e01823. doi:10.1002/brb3.1823

20. Touijer K. Marketing versus science: a fight between necessary evil and stern good over the adoption of new technology in medicine. Eur Urol. 2010;58(4):522-524. doi:10.1016/j.eururo.2010.06.004

21. Fried GM, Clas D, Meakins JL. Minimally invasive surgery in the elderly patient. Surg Clin North Am. 1994;74(2):375-387.

22. Modi PK, Kaufman SR, Caram ME, Ryan AM, Shahinian VB, Hollenbeck BK. Medicare accountable care organizations and the adoption of new surgical technology. J Am Coll Surg. 2021;232(2):138-145.e2. doi:10.1016/j.jamcollsurg.2020.10.016

23. Modi PK, Sukul DA, Oerline M, et al. Episode payments for transcatheter and surgical aortic valve replacement. Circ Cardiovasc Qual Outcomes. 2019;12(12):e005781. doi:10.1161/CIRCOUTCOMES.119.005781

24. Carino D, Sarac TP, Ziganshin BA, Elefteriades JA. Abdominal aortic aneurysm: evolving controversies and uncertainties. Int J Angiol. 2018;27(2):58-80. doi:10.1055/s-0038-1657771

25. Canning P, Tawfick W, Whelan N, Hynes N, Sultan S. Cost-effectiveness analysis of endovascular versus open repair of abdominal aortic aneurysm in a high-volume center. J Vasc Surg. 2019;70(2):485-496. doi:10.1016/j.jvs.2018.11.018

26. Baron SJ, Wang K, Arnold SV, et al; COAPT Investigators. Cost-effectiveness of transcatheter mitral valve repair versus medical therapy in patients with heart failure and secondary mitral regurgitation: results from the COAPT trial. Circulation. 2019;140(23):1881-1891. doi:10.1161/CIRCULATIONAHA.119.043275

27. Ulug P, Powell JT, Martinez MAM, Ballard DJ, Filardo G. Surgery for small asymptomatic abdominal aortic aneurysms. Cochrane Database Syst Rev. 2020;7(7):CD001835. doi:10.1002/14651858.CD001835.pub5

28. de Blic R, Alsac JM, Julia P, et al. Elective treatment of abdominal aortic aneurysm is reasonable in patients >85 years of age. Ann Vasc Surg. 2014;28(1):209-216. doi:10.1016/j.avsg.2013.01.022

29. Tom SE, Hubbard RA, Crane PK, et al. Characterization of dementia and Alzheimer’s disease in an older population: updated incidence and life expectancy with and without dementia. Am J Public Health. 2015;105(2):408-413. doi:10.2105/AJPH.2014.301935

30. Kassahun WT. The effects of pre-existing dementia on surgical outcomes in emergent and nonemergent general surgical procedures: assessing differences in surgical risk with dementia. BMC Geriatr. 2018;18(1):153. doi:10.1186/s12877-018-0844-x

31. Kassahun WT, Staab H, Gockel I, Mehdorn M. Factors associated with morbidity and in-hospital mortality after surgery beyond the age of 90: comparison with outcome results of younger patients matched for treatment. Am J Surg. 2018;216(6):1063-1069. doi:10.1016/j.amjsurg.2017.11.032

32. Torrance ADW, Powell SL, Griffiths EA. Emergency surgery in the elderly: challenges and solutions. Open Access Emerg Med. 2015;7:55-68. doi:10.2147/OAEM.S68324

33. Ernecoff NC, Zimmerman S, Mitchell SL, et al. Concordance between goals of care and treatment decisions for persons with dementia. J Palliat Med. 2018;21(10):1442-1447. doi:10.1089/jpm.2018.0103