Association of the Affordable Care Act Medicaid Expansion with Access to and Quality of Care for Surgical Conditions
Andrew P. Loehrer, MD, MPH1; David C. Chang, PhD, MPH, MBA2; John W. Scott, MD, MPH3; Matthew M. Hutter, MD, MPH2; Virendra I. Patel, MD, MPH4; Jeffrey E. Lee, MD1; Benjamin D. Sommers, MD, PhD5
1. Department of Surgical Oncology, University of Texas MD Anderson Cancer Center, Houston, TX 2. Department of Surgery, Massachusetts General Hospital, Department of Surgery, Boston, MA 3. Department of Surgery, Brigham and Women’s Hospital, Boston, MA 4. Department of Surgery, Columbia University Medical Center, New York, NY 5. Department of Health Policy and Management, Harvard Chan School of Public Health and Department of
Medicine, Brigham and Women’s Hospital, Boston, MA
JAMA Surg. 2018;153(3):e175568. doi:10.1001/jamasurg.2017.5568 Published online January 24, 2018.

Corresponding Author Andrew P. Loehrer, MD, MPH
MD Anderson Cancer Center Department of Surgical Oncology
Unit 1484 P.O. Box 301402
Houston, TX 77230-1402 (o) 713-792-0032
(f) 713-745-5235 aploehrer@mdanderson.org

1

Affordable Care Act Medicaid Expansion and Access to Surgical Care KEY POINTS
Question: How did Medicaid expansion under the Affordable Care Act affect patient presentation with and management of common surgical conditions? Findings: Among patients with one of 5 common surgical conditions, the Medicaid expansion was associated with a 7.5-percentage point increase in insurance coverage at the time of hospital admission. The policy was also associated with patients obtaining care earlier in their disease course and with an increased probability of receiving optimal care for those conditions. Meaning: The ACA’s Medicaid expansion was associated with increased coverage of patients, earlier presentation with common diagnoses, and improved surgical care.
2

Affordable Care Act Medicaid Expansion and Access to Surgical Care
ABSTRACT
Importance: Lack of insurance coverage has been associated with delays in seeking care, more complicated diseases at the time of diagnosis, and decreased likelihood of receiving optimal surgical care. The Affordable Care Act’s (ACA) Medicaid expansion has increased coverage among millions of low-income Americans, but its impact on care for common surgical conditions remains unknown.
Objective: To evaluate the impact of the ACA’s Medicaid expansion on access to timely and recommended care for common and serious surgical conditions.
Design: Quasi-experimental difference-in-differences study design, using hospital administrative data to compare patient-level outcomes in expansion vs. non-expansion states, before (20102013) versus after (2014-2015) expansion.
Setting: Academic medical centers and affiliated hospitals in 27 Medicaid expansion states and 15 non-expansion states.
Participants: Patients aged 18 to 64 years admitted to a study hospital between January 2010 and September 2015 with appendicitis, cholecystitis, diverticulitis, peripheral artery disease (PAD) or aortic aneurysm (N=293,529).
Exposure(s): State adoption of Medicaid expansion
Main Outcome(s) and Measure(s): Presentation with early uncomplicated disease (diverticulitis without abscess, fistula, or sepsis; nonruptured aortic aneurysm at time of repair; and PAD without ulcerations or gangrene), and receipt of optimal management (cholecystectomy for acute cholecystitis; laparoscopic approach for cholecystectomy or appendectomy; limbsalvage for PAD).
Results: Medicaid expansion was associated with a 7.5 percentage-point decreased probability of patients being uninsured (95% CI -12.2 to -2.9; P=0.002) and an 8.6 percentage-point increased probability of having Medicaid (95% CI 6.1 to 11.1; P<0.001). Medicaid expansion was associated with a 1.8 percentage-point increase in the probability of early uncomplicated presentation (95% CI 0.7 to 2.9; P=0.001), and a 2.6 percentage-point increase in the probability of receiving optimal management (95% CI 0.8 to 4.4; P=0.006).
Conclusions and Relevance: The ACA’s Medicaid expansion was associated with increased insurance coverage and improved receipt of timely care for five common surgical conditions.
3

Affordable Care Act Medicaid Expansion and Access to Surgical Care
INTRODUCTION The Affordable Care Act (ACA) aimed to increase access to health care through expanded insurance coverage. Since enactment, more than 20 million individuals have gained health insurance through the ACA, disproportionately in states adopting Medicaid expansion.1,2 Surveybased studies have shown the ACA to be associated with self-reported access to primary care and medications, affordability, and overall health.3-5 However, the clinical impact of coverage expansion on patients with common and serious surgical conditions remains unclear.
Uninsured and underinsured patients are significantly more likely to delay care and present with more complicated disease for conditions including appendicitis, cholecystitis, diverticulitis, aortic aneurysms and lower extremity peripheral artery disease (PAD).6-9 Insurance status is also associated with decreased probability of receiving optimal care for such diagnoses, including a lower likelihood of minimally invasive surgery, receiving immediate cholecystectomy for acute cholecystitis, and higher likelihood of amputation for vascular disease.6,7,10-12 The acute onset or deterioration of these diagnoses provides an ideal setting for studying short-term changes in presentation and management following population wide insurance coverage expansion. As more individuals gain insurance coverage, one might also expect a greater probability of patients presenting earlier and with less severe disease at time of diagnosis and in turn receive more timely and less morbid management.
The 2006 Massachusetts health reform, a model for the ACA insurance expansions, was associated with an increased probability of presenting with less complicated surgical disease and a greater likelihood of receiving optimal surgical management after admission.13-15 However, it is
4

Affordable Care Act Medicaid Expansion and Access to Surgical Care
unclear whether the Massachusetts experience generalizes to the ACA’s Medicaid expansion. Given current debate on the ACA and reforms to the Medicaid program, evidence on the impacts of these policies are critical. The primary objective of this study was to analyze changes associated with the ACA’s Medicaid expansion in access to and receipt of optimal care for common and serious surgical conditions.
METHODS Study Design Our study uses a quasi-experimental difference-in-differences design to compare patient-level outcomes in Medicaid expansion vs. non-expansion states, before (2010-2013) versus after (2014-2015) expansion. We also adjusted for potential time-varying confounders that could bias our estimates, and tested the underlying assumption of the study design – namely, that in the absence of the Medicaid expansion, our study outcomes would have had similar trends over time between expansion and non-expansion states.
Data Source and Outcomes Our primary data source was the Vizient (formerly University HealthSystem Consortium) Clinical Database/Resource Manager (CD/RM). The CD/RM is an administrative database capturing all discharges from participating hospitals that include over 95% of academic medical centers and affiliated hospitals across the country. The database has been used in multiple previous analyses to study surgical care.16-19 Our analysis included admissions to 203 hospitals in 27 Medicaid-expansion states (including Washington D.C) and 69 hospitals in 15 non-expansion states (Appendix Table 1). Some hospitals enter and exit the sample during the study period; in
5

Affordable Care Act Medicaid Expansion and Access to Surgical Care
sensitivity analyses, we tested the effect of this compositional change. For confidentiality reasons, the public-use file does not include the specific state in which a hospital is located. For our analysis, we were provided with an indicator for whether a hospital was in a state that expanded Medicaid prior to September 2015.
Our sample contained patients admitted to any hospital in the dataset between January 2010 and September 2015 with appendicitis, cholecystitis, diverticulitis, aortic aneurysm, or peripheral arterial disease (PAD). Conditions were identified based on ICD-9 diagnosis and procedure codes (Appendix Table 2). We selected these particular surgical conditions given extensive documentation of disparities in patient presentation with and subsequent care received based on insurance coverage, in addition to data suggesting an association between insurance expansion and care delivery for these conditions.5-14 Surgery for appendicitis, cholecystitis, and diverticulitis represent the three most frequent and highest morbidity procedures in general surgery.20 Vascular diseases, including aortic aneurysm and PAD, are increasingly prevalent in the U.S., and timely appropriate management has been shown to directly impact both mortality and major morbidity.21,22
We limited our sample to patients whose age was between 19 and 64 years. Adults older than 64 years and those with Medicare coverage were excluded, as they were not eligible for the ACA’s Medicaid expansion. Although the majority of newly insured patients under the ACA have enrolled in Medicaid, our primary analysis also included those with private insurance, since Medicaid expansion may have affected transitions in coverage between private and public insurance, and possible enrollment in ACA Marketplace plans. Unable to specifically determine
6

Affordable Care Act Medicaid Expansion and Access to Surgical Care
which individuals gained Medicaid after expansion, we conducted a subgroup analysis, stratifying our sample into 1) those with private coverage and 2) those with Medicaid or no health insurance. If the ACA’s Medicaid expansion is causing the changes we observe, they should be concentrated among the latter group (Medicaid/uninsured).
Our study outcomes were composite measures in two broad domains – “early uncomplicated presentation,” and “receipt of optimal care.” These composite outcomes were created to include a variety of common, serious surgical diseases that share a common causal pathway associated with similar outcomes, and to increase power to detect significant effects across multiple conditions. For early, uncomplicated presentation, we defined the condition-specific outcomes as follows: uncomplicated diverticulitis as the absence of peritoneal or retroperitoneal abscess, enterocutaneous fistula, sepsis or septic shock; non-ruptured aortic aneurysm at the time of repair; and peripheral artery disease without ulcerations or gangrene of limb (see Appendix Table 3 for details of ICD-9 codes used in this definition).
We defined “optimal care” for select diagnoses based on prior studies showing improved morbidity, mortality, effect on quality of life, and costs.23-28 Condition-specific outcomes were as follows: receipt of cholecystectomy when admitted with acute cholecystitis; receipt of minimally-invasive appendectomy or cholecystectomy when undergoing surgery for acute appendicitis or cholecystitis; and avoidance of amputation when admitted with lower extremity peripheral artery disease. While alternative treatment may be appropriate for individual patients depending on specific presentation, we contend that receipt of optimal care as defined here represents the ideal management if patients present early and are healthy enough to receive it.
7

Affordable Care Act Medicaid Expansion and Access to Surgical Care
Statistical Analysis We used linear regression difference-in-difference models to evaluate changes in presentation and management of surgical conditions for patients in expansion states versus non-expansion states associated with implementation of the ACA’s Medicaid expansion. Linear models provide straightforward estimates of absolute changes, as is typical in difference-in-differences models.29 An “Expansion State” binary variable was created for discharges from hospitals in states that expanded Medicaid eligibility any time before September 2015. A separate “Post-Reform” indicator variable was created if the discharge occurred on or after January 1, 2014, the date expanded Medicaid eligibility in the majority of adopting states. An interaction term was created between Expansion State and Post-Reform, which provides the difference-in-differences estimate for the Medicaid expansion.
All models adjusted for patient age, sex, race/ethnicity, and comorbidity, in addition to a linear quarterly time trend. The CD/RM defines categorical age variables as “18-30 years”, “31-50 years”, or “51-64 years.” Patient race/ethnicity was defined in the dataset as “White”, “Black”, “Asian”, or “other.” Patient comorbidity was determined using the Charlson Comorbidity Index.30 Standard errors were clustered at the hospital level using cluster-correlated robust estimate of variance, as the public use file does not allow the direct identification of states.31
In sensitivity analyses, we assessed parallel trends in outcomes between expansion and nonexpansion states prior to 2014, which is the key to the validity of the difference-in-differences study design. These models included only admissions that occurred prior to January 1, 2014. An interaction term between the Expansion state variable and the continuous quarter variable was
8

Affordable Care Act Medicaid Expansion and Access to Surgical Care used to identify differential trends in outcomes between expansion state and non-expansion states prior to 2014. Additionally, we broke down our composite outcomes to assess trends in presentation with and management of individual diagnoses, though with reduced power due to smaller sample sizes. Finally, additional sensitivity models limited analysis to 139 hospitals in expansion states and 44 hospitals in non-expansion states that were consistently represented in the dataset from 2010 through 2015.
This study was deemed exempt from IRB review by Harvard University, since we only examined de-identified data. Data were analyzed using STATA software, version 14 (STATA, College Station, TX). Statistical significance was tested using a two-sided threshold of p=0.05.
RESULTS Our final sample included 225,572 admissions in Medicaid expansion states and 67,957 admissions in non-expansion states. Patient demographic and clinical characteristics are shown in Table 1. Patients in expansion states were younger, more likely to be white, and a higher share had private insurance. Patient sex and comorbidity were comparable between expansion and non-expansion states.
Beginning in 2014, the share of uninsured patients dropped sharply in expansion states compared to non-expansion states (Figure 1). In adjusted models, the Medicaid expansion was associated with a 7.5 percentage-point decrease in the probability of being uninsured (95% CI -12.2 to -2.9; P=0.002) and an 8.6 percentage-point increase in the probability of Medicaid coverage (95% CI
9

Affordable Care Act Medicaid Expansion and Access to Surgical Care
6.1 to 11.1; P<0.001) in our sample of admissions for surgical conditions (Table 2). There was no significant change in the probability of admissions for patients with private insurance.
Figure 2 shows unadjusted trends in presentation and management throughout study period, and Table 3 presents adjusted difference-in-differences estimates of changes in patient presentation and management associated with Medicaid expansion. Medicaid expansion was associated with a 1.8 percentage-point increased probability (95% CI 0.7 to 2.9; P=0.001) of early uncomplicated presentation for all patients. Improvements in presentation were concentrated among the Medicaid/uninsured subgroup (2.9 percentage points, 95% CI 1.2 to 4.8; P=0.001). Medicaid expansion was associated with a 2.6 percentage-point increased probability of optimal care (95% CI 0.8 to 4.4; P=0.006) among all patients, and this estimate rose to 3.7 percentage points (95% CI 0.7 to 6.7; P=0.016) in the Medicaid/uninsured subgroup. Unadjusted trends in Figure 2 shows an apparent divergence in in timely presentation beginning in 2013. However, adjusted models comparing pre-expansion trends in expansion and non-expansion states did not show any significant differences for either of the composite surgical measures (Appendix Table 4), offering support for the difference-in-differences design.
Evaluating individual components of our composite measures produced similar directional estimates for the impact of the Medicaid expansion, but with variable statistical significance (Appendix Tables 5 and 6). Early presentation for diverticulitis remained significant (p=0.004), as did optimal management of acute cholecystitis (p=0.05), while the remaining outcomes had positive but non-significant coefficients. Results were also consistent when limiting analyses to patients admitted to hospitals present in sample throughout study period (Appendix Table 7).
10

Affordable Care Act Medicaid Expansion and Access to Surgical Care
Finally, given that our outcomes are calculated as a proportion of total admissions, we assessed whether admission rates for these conditions changed in association with Medicaid expansion (Appendix Table 8). In our full sample, there were no differential changes in admission rates for these conditions in expansion states compared to non-expansion states (Incident Rate Ratio [IRR] 0.99; 95% CI 0.97 to 1.00; P=0.14). Among the insurance subgroups, we detect significant but small changes – a slight increase in volume of admissions for Medicaid/uninsured (IRR 1.02, p<0.001) in expansion vs. non-expansion states, and a slight decrease for privately insured patients (IRR 0.99, P<0.001). This indicates that the changes in our composite outcomes (Figure 2 and Table 3) are primarily due to changes in the numerator (i.e. early presentation and optimal care received), rather than the denominator (i.e. number of admissions).
DISCUSSION In a study of over 290,000 hospital admissions in 42 states (including Washington DC), we find that the ACA’s Medicaid expansion was associated with significant increases in insurance coverage among patients presenting with serious surgical conditions. The Medicaid expansion was also associated with a significantly greater likelihood of patients’ presenting with earlier less complicated disease at the time of admission, and an increased likelihood of receiving optimal care after admission. As expected, these changes were concentrated among Medicaid and uninsured patients most likely to benefit from coverage expansion, rather than those with private insurance. To our knowledge, this is among the first non-survey based data indicating improvements in quality of care associated with the ACA’s Medicaid expansion. Our findings
11

Affordable Care Act Medicaid Expansion and Access to Surgical Care
supplement a large body of research demonstrating improved access, satisfaction with care, and self-reported health since enactment of the ACA.3,4,32,33
The ACA expanded insurance coverage to over 20 million Americans, largely through Medicaid, and one main goal of the law was to decrease barriers to timely receipt of high-quality care for serious medical conditions.2 Our study focuses on common surgical conditions for several reasons: they are typically urgent conditions requiring prompt evaluation; delays in care may respond quickly to changes in health insurance status; and they are highly costly and morbid conditions. Thus, our study captures outcomes with important implications for both public health and health care costs. Beyond financial protection for individual patients, Medicaid expansion has been shown to mitigate uncompensated care and improve finances for hospitals caring for vulnerable populations.34,35 These data also build on previous works from Massachusetts and the ACA’s dependent coverage provision showing an association between insurance expansion and improved clinical care of acute onset diseases, including appendicitis, cholecystitis and lower extremity vascular disease.13-15,36
Our findings are likely explained by several main factors. Newly insured individuals may be less likely to delay seeking care and thus present sooner when early symptoms develop. Such behavior is consistent with the Oregon health insurance experiment and numerous ACA-related studies that show Medicaid coverage to be associated with decreased delays in seeking care and increased use of emergency room, primary care, and outpatient clinics.4,5,37 Having presented earlier in the clinical progression or deterioration of their conditions, patients may be more likely to receive optimal management with decreased morbidity and improved outcomes. Additionally,
12

Affordable Care Act Medicaid Expansion and Access to Surgical Care
hospitals may be more willing to provide care for newly-insured individuals who previously lacked coverage.38 As expected for the conditions studied, our analysis found no significant change in the overall number of individuals treated, but rather change the timeliness in which individuals receive care.
The study also has a number of limitations that should be considered. First, the CD/RM is an administrative dataset and as such is vulnerable to coding errors and limited clinical granularity captured through ICD-9 coding. However, these data have been used extensively for hospital quality benchmarking, including studies of the diagnoses used in our analyses, and it has been shown to be valid in comparison with clinical data and large administrative datasets.16-19 The CD/RM also only includes participating academic medical centers and affiliated hospitals that may differ in important ways from non-academic hospitals. Yet, academic centers play a key role in the health care safety net, making them highly relevant to evaluating the effects of the Medicaid expansion. Furthermore, our analysis included over 203 hospitals in 42 states (including Washington DC), capturing a fairly broad segment of inpatient care in the U.S. Recognizing these limitations, we feel the advantages of the CD/RM dataset outweigh many of these concerns and enable a robust, timely analysis of clinical outcomes with a validated data source.
While our analysis presents strong evidence that the expansion led to increases in insurance coverage, we recognize that the data on improved quality of care is not as clear. Our use of composite outcomes has specific limitations. Evaluation of individual components (Appendix Table 6 and 7) reveals varying patterns of change after expansion, some components with
13

Affordable Care Act Medicaid Expansion and Access to Surgical Care
marked improvement in expansion states after expansion and others with less of a decline after expansion. Despite this heterogeneity, the overall changes and pattern are consistent with improved care for each component in expansion states compared to changes in non-expansion states. Furthermore, Medicaid/uninsured patients saw the majority of improved outcomes in expansion states compared to privately insured patients (Table 3). Privately insured patients within expansion states may serve as an additional within-state control group to assess for regional differences independent of insurance expansion, although they are not truly a control group given the ACA’s numerous policies affecting this population.
Additionally, we defined post-reform period as any discharge after January 1, 2014. Although five states partially expanded Medicaid prior to 2014 and another seven states expanded after 2014, the insurance coverage trends seen in our dataset saw the most marked shift starting in 2014. Limitations on state identifiers in our dataset prevented additional sensitivity testing through including or excluding these states from analysis. However, the early expansion states may have contributed to a trend toward a decreasing uninsured rate in expansion states prior to 2014. Analysis of our sample revealed an increase in the percentage of surgical patients who were uninsured in nonexpansion states after 2014 (Table 2). This could partly explain the worsening of outcomes after 2014 while expansion states saw stabilization or improvement.
Our findings also may not be generalizable for other medical conditions. We selected these diagnoses and procedures based on well-established disparities in presentation and management based on insurance coverage.6-11 Underlying pathophysiology or comorbidities, rather than delays in care, may contribute to severity of disease at presentation and subsequent
14

Affordable Care Act Medicaid Expansion and Access to Surgical Care management.39,40 Compared to states that expanded Medicaid, nonexpansion states are known to have higher rates of comorbid conditions including obesity and cardiovascular disease which may limit generalizability.41-43 However, these differences between states should not bias our difference-in-differences analysis unless they were changing significantly during the study period; stable differences between states are filtered out with our study design.
Finally, additional changes in care delivery both associated with and independent of the ACA could have potentially influenced our findings. As these data do not enable the evaluation of hospital- or state-specific programs of redesigned care, ongoing assessment of alternative payment and delivery models is needed to ensure changes are equitable across populations.
Conclusions In this study of surgery patients in 42 states (including Washington DC), we found that the ACA’s Medicaid expansion was associated with higher coverage rates, earlier presentation, and improved probability of optimal care for common and serious surgical conditions. Our data reinforce that insurance coverage is an important contributor to earlier presentation with less severe disease at time of diagnosis. As policy makers weigh changes to or a potential repeal of the Affordable Care Act, these findings provide important new data on the early clinical impacts of the law’s coverage expansion.
15

Affordable Care Act Medicaid Expansion and Access to Surgical Care AKNOWLEDGMENTS We would like to recognize Vizient (formerly University HealthSystem Consortium) for collection and maintenance of data, and specifically Sam Hohmann for assistance in data acquisition and study-specific variable coding.
16

Affordable Care Act Medicaid Expansion and Access to Surgical Care
REFERENCES
1. Cohen RA, Martinez ME, Zammitti EP. Health Insurance Coverage: Early Release of Estimates From the National Health Interview Survey, 2015. Hyattsville, MD: National Center for Health Statistics; 2016.
2. Frean M, Gruber J, Sommers BD. Disentangling the ACA’s Coverage Effects – Lessons for policymakers. N Engl J Med. 2016; 375(17):1605-1607.
3. Sommers BD, Gunja MZ, Finegold K, Musco. Changes in self-reported insurance coverage, access to care, and health under the Affordable Care Act. JAMA. 2015;314(4):366-374.
4. Sommers BD, Blendon RJ, Orav J, Epstein AM. Changes in utilization and health among low-income adults after Medicaid expansion or expanded private insurance. JAMA Intern Med. 2016;176(10):1501-1509.
5. Wherry LR, Miller S. Early Coverage, Access, Utilization, and Health Effects Associated With the Affordable Care Act Medicaid Expansions: A Quasi-experimental Study. Ann Intern Med.2016;164:795-803.
6. Scott JW, Havens JM, Wolf LL, Zogg CK, Rose JA, Salim A, Haider AH. Insurance status is associated with complex presentation among emergency general surgery patients [published online October 4, 2016]. Surgery. http://dx.doi.org/10.1016/j.surg.2016.08.038
7. Greenstein AJ, Moskowitz A, Gelijns AC, Egorova NN. Payer status and treatment paradigm for acute cholecystitis. Arch Surg. 2012;147(5):453-458.
8. Giacovelli JK, Egorova N, Nowygrod R, Gelijns A, Kent KC, Morrissey NJ. Insurance status predicts access to care and outcomes of vascular disease. J Vasc Surg. 2008;48:905-911.
9. Andersen ND, Hanna JM, Ganapathi AM, Bhattacharya SD, Williams JB, Gaca JG, McCann RL, Hughes GC. Insurance status predicts acuity of thoracic aortic operations. J Thorac Cardiovasc Surg. 2014;148:2082-2086.
10. Varela JE, Nguyen NT. Disparities in access to basic laparoscopic surgery at US academic medical centers. Surg Endosc. 2011;25(4):1209-1214.
11. Guller U, Jain N, Curtis LH, Oertli D, Heberer M, Pietrobon R. Insurance status and race represent independent predictors of undergoing laparoscopic surgery for appendicitis: Secondary data analysis of 145,546 patients. J Am Coll Surg. 20014;199(4): 567-577.
12. Henry AJ, Hevelone ND, Belkin M, Nguyen LL. Socioeconomic and hospital-related predictors of amputation for critical limb ischemia. J Vasc Surg. 2011;53:330-339.
17

Affordable Care Act Medicaid Expansion and Access to Surgical Care
13. Loehrer AP, Hawkins AT, Auchincloss HG, Song Z, Hutter MM, Patel VI. Impact of expanded insurance coverage on racial disparities in vascular disease: Insights from Massachusetts. Ann Surg. 2016;263:705-711.
14. Loehrer AP, Song Z, Auchincloss HG, Hutter MM. Massachusetts health care reform and reduced racial disparities in minimally invasive surgery. JAMA Surg. 2013;148(02):11161122.
15. Loehrer AP, Song Z, Auchincloss HG, Hutter MM. Influence of health insurance expansion on disparities in the treatment of acute cholecystitis. Ann Surg. 2015;262:139-145.
16. Varela JE, Nguyen NT. Disparities in access to basic laparoscopic surgery at U.S. Academic Medical Centers. Surg Endosc. 2011;25:1209-1214.
17. Nguyen NT, Hinojosa MW, Fayad C, Varela E, Konyalian V, Stamos MJ, Wilson SE. Laparoscopic surgery is associated with a lower incidence of venous thromboembolism compared with open surgery. Ann Surg. 2007;246(6):1021-1027.
18. Simorov A, Shaligram A, Shostrom V, Boilese E, Thompson J, Oleynikov D. Laparoscopic colon resection trends in utilization and rate of coversion to open procedure: A national database review of academic medical centers. Ann Surg. 2012;256(3):462-468.
19. Sutton JM, Hayes AJ, Wilson GC, Quillin RC, Wima K, Hohmann S, Paquette IM, Sussman JJ, Ahmad SA, Haha SA, Abbott. Validation of the University HealthSystem Consortium administrative dataset: Concordance and discordance with patient-level institutional data. J Surg Res. 2014;190:484-490.
20. Scott JW, Olufajo OA, Brat GA, Rose JA, Zogg CK, Haider AH, Salim A, Havens JM. Use of national burden to define operative emergency general surgery [published online April 27, 2016]. JAMA Surg. http://jamanetwork.com/article.aspx?doi=10.1001/jamasurg.2016.0480.
21. Schermerhorn ML, Bensley RP, Giles KA, Hurks R, O’Malley AJ, Cotterill P, Chaikof E, Landon BE. Changes in abdominal aortic aneurysm rupture and short term mortality 19952008. Ann Surg. 2012;256(4):651-658.
22. Selvin E, Erlinger TP. Prevalence of and risk factors for peripheral arterial disease in the United States: Results from the National Health and Nutrition Examination Survey, 19992000. Circulation. 2004;110:738-743.
23. Lo CM, Liu CL, Fan ST, Lai EC, Wong J. Prospective randomized study of early versus delayed laparoscopic cholecystectomy for acute cholecystitis. Ann Surg. 1998;227:461-467.
18

Affordable Care Act Medicaid Expansion and Access to Surgical Care
24. Gurusamy K, Samraj K, Gluud C, Wilson E, Davidson BR. Meta-analysis of randomized controlled trials on the safety and effectiveness of early versus delayed laparoscopic cholecystectomy for acute cholecystitis. Br J Surg. 2010;97:141-150
25. Tiwari MM, Reynoso JF, Tsang AW, Oleynikov D. Comparison of outcomes of laparoscopic and open appendectomy in management of uncomplicated and complicated appendicitis. Ann Surg. 2011;254:927-932.
26. Yau KK, Siu WT, Tang CH, Yang GPC, Li MKW. Laparoscopic versus open appendectomy for complicated appendicitis. J Am Coll Surg. 2007;205:60-65.
27. Lujan JA, Parrilla P, Robles R, Marin P, Torralba JA, Garcia-Ayllon J. Laparoscopic cholecystectomy vs open cholecystectomy in the treatment of acute cholecystitis: a prospective study. Arch Surg. 1998;133(2):173-175.
28. Egorova NN, Guillerme S, Gelijns A, Morrissey N, Dayal R, McKinsey JF, Nowygrod R. An analysis of outcomes of a decade of experience with lower extremity revascularization including limb salvage, lengths of stay, and safety. J Vasc Surg. 2010;51:878-885.
29. Karaca-Mandic P, Norton EC, Dowd B. Interaction terms in nonlinear models. Health Serv Res 2012;47:255-74.
30. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method classifying prognostic comorbidity in longitudinal studies: Development and validation. J Chron Dis. 1987;40:373383.
31. Rogers WH. Regression standard errors in clustered samples. Stata Technical Bulletin. 1993;13:19-23. Stata Technical Bulletin Reprints, vol. 3, 88-94.
32. Collins SR, Gunja M, Doty MM, Beutel S. Americans’ Experiences with ACA Marketplace and Medicaid Coverage: Access to Care and Satisfaction. New York, NY: The Commonwealth Fund; 2016.
33. Shartzer A, Long SK, Anderson N. Access To Care And Affordability Have Improved Following Affordable Care Act Implementation; Problems Remain. Health Aff (Millwood) 2016;35:161-8.
34. Dranove D, Garthwaite C, Ody. Uncompensated care decreased at hospitals in Medicaid expansion states but not at hospitals in nonexpansion states. Health Aff (Millwood). 2016;35(8):1471-1479.
19

Affordable Care Act Medicaid Expansion and Access to Surgical Care
35. Blavin F. Association between the 2014 Medicaid expansion and US hospital finances. JAMA. 2016;316(14):1475-1483.
36. Scott JW, Rose JA, Tsai TC, Zogg CK, Shrime MG, Sommers BD, Salim AL, Haider AH. Impact of ACA insurance coverage expansion on perforated appendix rates among young adults. Med Care. 2016;54(9):818-826.
37. Finkelstein AN, Taubman SL, Allen HL, Wright BJ, Baicker K. Effect of Medicaid coverage on ED use – Further evidence from Oregon’s experiment. N Engl J Med. 2016;375:15051507.
38. Doyle JJ. Health insurance, treatment and outcomes: using auto accidents as health shocks. Review of Economics and Statistics 2005;87:256-70.
39. Livingston EH, Woodward WA, Sarosi GA, Haley RW. Disconnect between incidence of nonperforated and perforated appendicitis: Implications for pathophysiology and management. Ann Surg. 2007;245:886-892.
40. Livingston EH, Fomby TB, Woodward WA, Haley RW. Epidemiological similarities between appendicitis and diverticulitis suggesting a common underlying pathogenesis. Arch Surg. 2011;146(3):308-314.
41. Slack T, Myers CA, Martin CK, Heymsfield SB. The geographic concentration of US adult obesity prevalence and associated social, economic, and environmental factors. Obesity. 2014;22:868-874.
42. Cushman M, Cantrell RA, McClure LA, Howard G, Prineas RJ, Moy CS, Temple EM, Howard VJ. Ann Neurol. 2008;64:507-513.
43. Ezzati M, Oza S, Goodarz D, Murray CJL. Trends and cardiovascular mortality effects of state-level blood pressure and uncontrolled hypertension in the United States. Circulation. 2008;117:905-914
20

Affordable Care Act Medicaid Expansion and Access to Surgical Care

Table 1. Demographic and clinical characteristics of patients admitted in Medicaid expansion and non-expansion states.

Variable
Total Admissions
Age n (%) 18-30 years 31-50 years 51-65 years
Female n (%)
Race n (%) White Black Asian Other
Charlson Comorbidity Index mean (std)
Distribution of Diagnoses n (%) Appendicitis Cholecystitis Diverticulitis Aortic Aneurysm Peripheral Arterial Disease

Expansion States Non-Expansion States P-Value

225,529

67,947

32,889 (14.6) 72,000 (31.9) 120,683 (53.5)
98,710 (43.8)

7,636 (11.2) 21,434 (31.5) 38,887 (57.2)
29,682 (43.7)

P<0.001 P=0.705

151,654 (67.2) 32,148 (14.3)
5,631 (2.5) 36,139 (16.0)
0.97 (1.66)

42,919 (63.2) 14,685 (21.6)
937 (1.4) 9,416 (13.9)
1.16 (1.79)

P<0.001 P<0.001

70,979 (31.5) 13,175 (5.8)
93,150 (41.3) 14,089 (6.3)
34,136 (15.1)

16,616 (24.5) 4,077 (6.0)
29,257 (43.1) 5,327 (7.8)
12,670 (18.7)

P<0.001

	 	 	 	 	 	 	
21

Affordable Care Act Medicaid Expansion and Access to Surgical Care Figure 1. Unadjusted trends in health insurance status among admissions (2010-2015), by Medicaid expansion status.
	
	
22

Affordable Care Act Medicaid Expansion and Access to Surgical Care

Table 2. Changes in insurance coverage among admissions for surgical conditions after the

Medicaid expansion.

Probability of Coverage Expansion States

Non-Expansion States Difference-in-Differences

Pre Post Diff Pre Post Diff Unadjusted Adjusted*

Uninsured

14.1 6.8

-7.3

21.2

21.9

+0.6

-7.9 (P<0.001)

-7.5
CI -12.2 to -2.9 P=0.002

Medicaid

22.1 30.5 +8.4

18.0

18.0

0.0

+ 8.4 (P<0.001)

+8.6 CI 6.1 to 11.1
P<0.001

Private

63.9 62.7 -1.2

60.8

60.1 -0.7

-0.5 (P=0.26)

-1.0
CI -5.1 to 3.0 P=0.62

* Adjusted for patient age, sex, race, comorbidity, secular trends and clustering at hospital level

CI – 95% Confidence Interval

	

23

Affordable Care Act Medicaid Expansion and Access to Surgical Care Figure 2. Unadjusted trends in a) early uncomplicated presentation and b) receipt of optimal care among surgery admissions, by state Medicaid expansion status and insurance type.
24

Affordable Care Act Medicaid Expansion and Access to Surgical Care

Table 3. Changes in rates of early uncomplicated presentation and optimal care for select

surgical conditions after Medicaid expansion.

Early/Uncomplicated Presentation Expansion States
Pre Post Diff

Non-Expansion States Pre Post Diff

Full Sample (n=188,629)

85.8 84.4 -1.4 85.4 82.2 -3.2

88.3 86.7 -1.6 88.8 85.7 -3.1

Medicaid/Uninsured (n=66,904)

80.9

80.3

-0.6

80.1

76.4 -3.7

Difference-in-Differences

Unadjusted Adjusted*

+1.9 (P<0.001)
+1.5 (P=0.001)
+3.1 (P<0.001)

+1.8 CI 0.7 to 2.9
P=0.001
+1.4 CI 0.3 to 2.5
P=0.02
+2.9 CI 1.2 to 4.8
P=0.001

Optimal Management Expansion States Non-Expansion States Difference-in-Differences

Pre Post Diff Pre Post Diff Unadjusted Adjusted*

Full Sample (n=147,439)

81.0 82.9 +1.9 81.3 80.9 -0.4

+2.4 (P<0.001)

+2.6
CI 0.8 to 4.4 P=0.006

Private Insurance (n=84,689)

+82.9 84.4

+1.5

84.0

84.0 +0.0

+1.4 (P=0.028)

+1.7 CI -0.2 to 3.5
P=0.07

Medicaid/Uninsured (n=62,750)

78.1

81.0 +2.9

78.1

77.4

-0.7

+3.7 (P<0.001)

+3.7
CI 0.7 to 6.7 P=0.02

Full Sample – all patients included in study (private insurance, Medicaid, and uninsured)

* Adjusted for patient age, sex, race, comorbidity, secular trends and clustering at hospital level

CI – 95% Confidence Interval

	

	

	

	

25

Affordable Care Act Medicaid Expansion and Access to Surgical Care

Appendix Table 1. List of Medicaid expansion and non-expansion states included in study

Expansion States Arkansas
Arizona

Non-Expansion States Alabama
Florida

California * Colorado Connecticut * Washington, DC Delaware Iowa Illinois Indiana Massachusetts Maryland Michigan Minnesota * New Hampshire

Georgia Kansas Louisiana† Maine Missouri Mississippi North Carolina South Carolina Tennessee Texas Utah Wisconsin Wyoming

New Jersey * New Mexico

New York Ohio

Oregon Pennsylvania Rhode Island Virginia Washington * West Virginia

* Early adopter of Medicaid expansion † Adopted Medicaid expansion after study period (July 1, 2016)

26

Affordable Care Act Medicaid Expansion and Access to Surgical Care

Appendix Table 2. ICD-9 codes used for inclusion in study.

Diagnosis Acute Appendicitis

ICD-9 Diagnosis Codes 540.0, 540.1, 540.9, 541, 542

Acute Cholecystitis
Diverticular Disease Peripheral Artery Disease Aortic Aneurysm

574.00, 574.01, 574.10, 574.11, 574.30, 574.31, 574.40, 574.41, 574.60, 574.61, 574.70, 574.71, 574.80, 575.0, 575.12
562.10, 562.11, 562.12, 562.13
440.20, 440.21, 440.22, 440.23, 440.24, 440.29
441.1, 441.2, 441.3, 441.4, 441.5, 441.6, 441.7, 441.9

ICD-9 International Classification of Diseases, 9th Edition

ICD-9 Procedure Codes 47.01, 47.09, 17.32, 17.39, 17.42 N/A
N/A N/A
38.04, 38.05, 38.14, 38.15, 38.34, 38.35, 38.36, 38.44, 38.45, 38.46, 39.25, 39.71, 39.73, 39.78, 39.79, 39.90

27

Affordable Care Act Medicaid Expansion and Access to Surgical Care

Appendix Table 3. Coding for “Early, uncomplicated care” and “Optimal care”.

Component of “Optimal Care”

ICD-9 Diagnosis or Procedure Codes

Early, Uncomplicated Presentation Absence of Below Diagnosis Codes

Diverticulitis

567.31, 567.22, 567.38, 567.39, 569.81, 995.91, 995.92

Aortic Aneurysm

441.1, 441.3, 441.5, 441.6

Peripheral Artery Disease

440.23, 440.24

Optimal Surgical Care

Procedure Codes

Cholecystectomy

51.21, 51.22, 51.23, 51.24

Laparoscopic Appendectomy

47.01, 17.31, 17.39, 17.42

Laparoscopic Cholecystectomy

51.23, 51.24

Absence of Amputation

84.10, 84.11, 84.12, 84.13, 84.14, 84.15, 84.16, 84.17, 84.18

ICD-9 International Classification of Diseases, 9th Edition

	 	 	 	 	 	 	 	 	 	 	

28

Affordable Care Act Medicaid Expansion and Access to Surgical Care

Appendix Table 4. Trends in outcomes prior to the full Medicaid expansion.

Outcome

Pre-Reform Time Trend* 95% CI (P-Value)

Uninsured

-0.6 percentage-points

-1.1 to 0.0 (P=0.057)

Early, Uncomplicated Presentation -0.0 percentage-points

-0.2 to 0.1 (P=0.907)

Optimal Care

+0.0 percentage-points

-0.3 to 0.4 (P=0.828)

* Interaction between dichotomous Expansion State term and continuous period term,

representing quarterly trend in outcomes for expansion states compared to non-expansion states

prior to 2014, controlling for patient age, sex, race, comorbidity, secular trends and clustering at

hospital level

CI –Confidence Interval

	

	

	

	

	

	

	

	

	

	

	

	

29

Affordable Care Act Medicaid Expansion and Access to Surgical Care

Appendix Table 5. Changes in rates of early uncomplicated presentation and optimal care for select surgical conditions after Medicaid expansion, by condition and for full sample.

Early/Uncomplicated Presentation Expansion States
Pre Post Diff

Diverticulitis (n=123,202)

89.7 88.1 -1.6

Aortic Aneurysm (n=20,127)
PAD (n=46,806)

93.8 94.1 +0.3 72.1 69.1 -3.0

Non-Expansion States Pre Post Diff 89.3 86.0 -3.3
93.7 92.7 -1.0
73.1 68.2 -4.9

Difference-in-Differences

Unadjusted Adjusted*

+1.7 (P<0.001)
+1.3 (P=0.091)
+1.9 (P=0.057)

+1.6 CI 0.5 to 2.7
P=0.004
+1.3 CI -1.1 to 3.8
P=0.288
+2.3 CI -0.2 to 4.9
P=0.072

Optimal Management Expansion States Non-Expansion States Difference-in-Differences

Pre Post Diff Pre Post Diff Unadjusted Adjusted*

Cholecystectomy
for Cholecystitis (n=19,093)

73.7 70.5 -3.2 73.8 67.5 -6.3

+3.1 (P=0.048)

+2.8
CI -0.0 to 5.7 P=0.051

MIS Appendectomy or Cholecystectomy 78.6 83.7 +5.1 79.2 82.7 +3.5
(n=95,500)

+1.5 (P=0.024)

+1.9 CI -0.7 to 4.5
P=0.154

Limb-sparing PAD (n=46,806)

84.7 82.7

-2.0

82.0 78.9 -3.2

+1.2 (P=0.128)

+1.5
CI 0.9 to 4.0 P=0.219

Full Sample – all patients included in study (private insurance, Medicaid, and uninsured)

* Adjusted for patient age, sex, race, comorbidity, secular trends and clustering at hospital level

CI – 95% Confidence Interval
	

	

	

30

Affordable Care Act Medicaid Expansion and Access to Surgical Care

Appendix Table 6. Changes in rates of early uncomplicated presentation and optimal care

for select surgical conditions after Medicaid expansion, by condition and for Medicaid and

uninsured patients.

Early/Uncomplicated Presentation Expansion States
Pre Post Diff

Diverticulitis (n=40,644)

88.1 86.8 -1.3

Aortic Aneurysm (n=4,201)

87.5 88.2 +0.7

PAD (n=22,559)

66.6 65.3 -1.3

Non-Expansion States Pre Post Diff 86.7 83.6 -3.1
87.3 86.1 -1.2
68.7 62.5 -6.2

Difference-in-Differences

Unadjusted Adjusted*

+1.7 (P=0.026)
+2.1 P=0.369
+4.8 (P=0.001)

+1.8 CI -0.3 to 3.8
P=0.086
+2.2 CI -2.3 to 6.6
P=0.338
+4.9 CI 1.7 to 8.1
P=0.003

Optimal Management Expansion States Non-Expansion States Difference-in-Differences

Pre Post Diff Pre Post Diff Unadjusted Adjusted*

Cholecystectomy
for Cholecystitis (n=8,308)

68.9 65.5 -3.4 70.0 62.9 -7.1

+3.6 (P=0.140)

+3.4
CI -1.1 to 8.0 P=0.140

MIS Appendectomy
or Cholecystectomy 76.9 84.0 +7.1 77.0 81.9 +4.9 (n=37,574)

+2.1 (P=0.050)

+2.3
CI -1.8 to 6.4 P=0.265

Limb-sparing PAD (n=22,559)

80.7 80.2

-0.5

78.5 73.8 -4.7

+4.2 (P=0.001)

+4.1 CI 0.5 to 7.7
P=0.027

* Adjusted for patient age, sex, race, comorbidity, secular trends and clustering at hospital level

CI – 95% Confidence Interval
	

	

	

31

Affordable Care Act Medicaid Expansion and Access to Surgical Care

Appendix Table 7. Changes in rates of early uncomplicated presentation and optimal care

for select surgical conditions after Medicaid expansion, for hospitals present in database

for all 6 years.

Early/Uncomplicated Presentation Expansion States
Pre Post Diff

Full Sample (n=160,475)

85.9 84.4 -1.5

Private Insurance (n=105,192)

88.3 86.8

-1.5

Medicaid/Uninsured (n=55,283)

81.1

80.0

-1.1

Non-Expansion States Pre Post Diff 85.5 82.3 -3.2
88.6 85.5 -3.1
80.3 76.1 -4.2

Difference-in-Differences

Unadjusted Adjusted*

+1.8 (P<0.001)
+1.6 (P=0.001)
+3.0 (P<0.001)

+1.8
CI 0.6 to 2.9 P=0.003
+1.6 CI 0.3 to 2.8
P=0.012
+3.0 CI 1.1 to 4.9
P=0.002

Optimal Management Expansion States
Pre Post Diff

Non-Expansion States Pre Post Diff

Difference-in-Differences Unadjusted Adjusted*

All Payers (n=124,006)
Private Insurance (n=72,751)

81.0 82.7 +1.7 83.3 84.5 +1.2

81.2 80.7 -0.5 84.0 83.7 -0.3

+2.2 (P<0.001)
+1.5 (P=0.035)

+2.3 CI 0.7 to 4.0
P=0.006
+1.7 CI -0.2 to 3.5
P=0.073

Medicaid/Uninsured (n=51,255)

77.4

80.3

+2.9

77.7 76.7 -1.0

+4.0 (P<0.001)

+3.8 CI 1.0 to 6.5
P=0.007

Full Sample – all patients included in study (private insurance, Medicaid, and uninsured)

* Adjusted for patient age, sex, race, comorbidity, secular trends and clustering at hospital level

CI – 95% Confidence Interval

32

Affordable Care Act Medicaid Expansion and Access to Surgical Care

Appendix Table 8. Changes in admission rates after Medicaid expansion.

Admission Rates*

Expansion States Pre Post Diff

Full Sample (n=293,529)

30.5 29.4 -1.1

Private Insurance (n=184,269)

36.2 34.6

-1.6

Medicaid/Uninsured (n=109,260)

20.4

20.6

-0.2

* Admissions per hospital per quarter

Non-Expansion States Pre Post Diff 27.8 27.0 -0.8
32.3 31.2 -0.9
20.8 20.7 -0.1

Difference-in-Differences
Unadjusted Rate Adjusted IRR
0.99 -0.3 CI 0.97 to 1.00
P=0.135 0.98
-0.7 CI 0.98 to 0.99 P<0.001 1.02
-0.1 CI 1.01 to 1.03 P<0.001

Full Sample – all patients included in study (private insurance, Medicaid, and uninsured)

IRR – admission incident rate ratio, adjusted for patient age, sex, race, comorbidity, secular

trends and clustering at hospital level

CI – 95% Confidence Interval

33