Health Care Fragmentation in Medicaid Managed Care vs. Fee for Service

Abstract

Managed care plans often attempt to control health care costs through strategies designed to decrease health care utilization. However, the extent to which the resulting patterns of utilization represent high-quality care (compared to fee-for-service products) remains controversial. The authors sought to compare patterns of ambulatory care (including how diffuse or fragmented the care patterns were) for Medicaid fee-for-service beneficiaries vs. Medicaid managed care beneficiaries. A serial cross-sectional study of adults (≥18 years old) was conducted using statewide Medicaid claims from New York State for calendar years 2010–2013. Beneficiaries were required to be continuously enrolled and have ≥4 ambulatory visits for each year they contributed data, yielding a sample of more than 1 million beneficiaries per year. Beneficiaries were characterized by age, sex, and case mix. For each year, ambulatory care patterns were compared across subgroups of beneficiaries using Poisson models (for numbers of visits and providers) and bounded Tobit models (for fragmentation scores). In 2010, among those who were not dual eligible, managed care beneficiaries had on average fewer visits (10.9 visits vs. 11.4 visits [P < 0.0001]) but more providers (3.8 providers vs. 3.3 providers [P < 0.0001]) and therefore more fragmentation (0.58 vs. 0.51 [P < 0.0001]) than fee-for-service beneficiaries, adjusting for age, sex, and case mix. These patterns persisted throughout the follow-up period and in sensitivity analyses. Less utilization is not necessarily more efficient care; a smaller number of visits spread across a larger number of providers creates more challenges for care coordination.

Introduction

States across the country (including California, Florida, Kansas, Kentucky, Louisiana, New Jersey, New York, Ohio, South Carolina, Texas, Washington, and Wisconsin) are attempting to decrease Medicaid costs in part through promoting or requiring enrollment in managed care rather than fee-for-service plans.¹ However, the effects of these initiatives are not yet known.

Managed care plans often require beneficiaries to have a designated primary care provider who is responsible for coordinating their care and preventing unnecessary utilization.² Because of these explicit goals of care coordination and utilization management, managed care beneficiaries would be expected to have fewer ambulatory visits and to have those visits concentrated among fewer providers (ie, to have less fragmented ambulatory care), compared to their fee-for-service counterparts, but this has not been measured.

This study sought to compare ambulatory visit patterns for adult Medicaid managed care vs. fee-for-service beneficiaries in New York State over the period 2010–2013. During this time, New York State expanded Medicaid under the Affordable Care Act, as did 32 other states.³ During this time, New York State also had an explicit goal of transitioning more Medicaid beneficiaries into managed care.⁴ Results of this study can inform ongoing Medicaid managed care initiatives in New York and in other states.

Methods

Overview

The research team conducted a serial cross-sectional analysis of ambulatory visit patterns, using claims for all Medicaid beneficiaries ages 18 years and older in New York State (2010–2013). The team compared patterns of care for Medicaid managed care vs. fee-for-service beneficiaries, stratified by dual eligible status. The Weill Cornell Medicine Institutional Review Board approved the study protocol.

Data

The research team used statewide Medicaid claims for the years noted, extracting the following claim-level variables: synthetic beneficiary ID, patient date of birth, patient sex, date of service, insurance product type (managed care vs. fee-for-service), rendering provider ID, rendering provider specialty, Current Procedural Terminology (CPT) codes, and International Classification of Diseases, Ninth Revision (ICD-9) codes. Claims had the same level of detail, regardless of insurance product type. Also extracted were monthly beneficiary-level enrollment data, including dual eligible status.

Study sample and variables

In each year of data only those beneficiaries who were continuously enrolled that year and stayed with the same insurance product all year (eg, 12 months of fee-for-service, or 12 months of managed care, excluding those who temporarily lost coverage or who switched products midyear [Supplementary Table S1]) were included. Those who temporarily lost coverage or who switched products midyear are different from those who stayed with the same product all year and therefore are not included in this analysis.⁵

To define ambulatory visits, a modified version of the definition of ambulatory visits by the National Committee for Quality Assurance (NCQA) was used, which is a list of applicable CPT codes.⁶ The NCQA definition excludes emergency department visits. The modifications restricted the definition to evaluation-and-management visits for adults in an office setting, as has been done previously.⁷ The specialty of the rendering provider for each visit was determined using the specialty code on the claim or, if absent, the National Plan and Provider and Enumeration System (NPPES) crosswalk.⁸ (Of the ambulatory visits in the data set, approximately 95% of visits each year had specialty information in the claims; thus, the NPPES crosswalk was used to derive specialty in <5% of visits per year.)

Beneficiaries who had outlier observations (>99.9^th percentile) for number of visits or the number of unique providers were excluded, as those observations may have been erroneous. The sample in each year was then restricted to those with ≥4 ambulatory visits because describing patterns of care based on ≤3 visits can yield unstable estimates.⁹

The following beneficiary-level variables were calculated for each year: total number of ambulatory visits, total number of unique ambulatory providers, the proportion of visits with the most frequently seen provider, and a fragmentation score. The Bice-Boxerman Index,¹⁰ a previously validated measure,^7,9,11
–13 was used to calculate fragmentation scores. Scores on this index range from 0 to 1; the raw scores were transformed by reversing their direction (calculating 1 – raw score), so that higher scores reflect more fragmented care. Beneficiary-level Charlson-Deyo comorbidity indices were calculated using ICD-9 codes.^14,15

Statistical analysis

Descriptive statistics were used to characterize the distribution of age, sex, and case mix in the study sample for each year, stratified by dual eligible status and insurance product (managed care vs. fee-for-service). Dual eligible beneficiaries were separated out because they often are older and sicker than the typical Medicaid beneficiary.¹⁶ Chi-square tests were used to compare the characteristics of managed care vs. fee-for-service beneficiaries within year and within dual eligible category (not eligible or eligible).

Ambulatory utilization patterns were characterized, stratifying by year, insurance product, and dual eligible status. These patterns also were adjusted for age, sex, and case mix, using Poisson regression (for number of visits and providers) and bounded Tobit models (for fragmentation scores). Poisson regression is appropriate for count data, whereas Tobit regression is appropriate when dependent variables are observed in only a limited range of values (such as 0 to 1).¹⁷ These analyses were then replicated to stratify ambulatory care utilization by primary care vs. specialist provider.

To determine the robustness of the findings, a sensitivity analysis was conducted restricting the sample to only those in the top 20% of ambulatory visits.

P values <0.05 were considered statistically significant. SAS version 9.4 (SAS Institute Inc., Cary, NC) was used for all analyses.

Results

Study sample

The total number of Medicaid beneficiaries in the sample increased by .2 million from 2010 to 2013 (Table 1). The majority of Medicaid beneficiaries were only eligible for Medicaid (ie, not dual eligible) and had managed care insurance products (N = 669,550, or 59% in 2010, Table 1). As the overall size of the Medicaid population grew over time, the absolute number and proportion of non-dual managed care beneficiaries also grew (Table 1).

Table 1.

Sample Sizes of Adult Medicaid Beneficiaries by Year, Stratified by Dual Eligible Status and Insurance Product^*

Year	Total	Not dual eligible		Dual eligible
Year	Total	Fee-for-service	Managed care	Fee-for-service	Managed care
2010	1,125,768	87,784 (8%)	669,550 (59%)	289,157 (26%)	79,277 (7%)
2011	1,162,954	69,759 (6%)	743,745 (64%)	284,816 (24%)	64,634 (6%)
2012	1,205,152	48,497 (4%)	845,325 (70%)	259,827 (22%)	51,503 (4%)
2013	1,291,826	42,417 (3%)	964,007 (75%)	241,973 (19%)	43,429 (3%)

To be included, beneficiaries needed to have continuous enrollment and have ≥4 ambulatory visits in the calendar year for which they contributed data.

Sample characteristics

Dual eligible beneficiaries were older and sicker than beneficiaries who were not dual eligible (Table 2). Among those not dual eligible, managed care beneficiaries were slightly younger, more likely to be female, and healthier than their fee-for-service counterparts. These differences (by dual eligible status and by insurance product) were observed consistently in each year of the study (Table 2; 2011 and 2012 not shown).

Table 2.

Characteristics of Medicaid Beneficiaries Over Time, Stratified by Dual Eligible Status and Insurance Product^*

	2010				2013
	Not dual eligible		Dual eligible		Not dual eligible		Dual eligible
	FFS	MC	FFS	MC	FFS	MC	FFS	MC
Demographics
Age category, years
18–30	23%	22%	2%	3%	37%	26%	3%	1%
31–40	17%	20%	5%	5%	16%	20%	7%	2%
41–50	21%	23%	9%	8%	16%	22%	13%	3%
51–60	25%	23%	14%	12%	18%	22%	18%	7%
61+	13%	12%	69%	72%	14%	11%	59%	88%
Sex, % female	58%	70%	62%	66%	53%	68%	59%	69%
Charlson-Deyo score category
0	43%	59%	27%	29%	50%	55%	32%	15%
1–2	33%	34%	42%	47%	33%	34%	42%	42%
3–4	7%	5%	18%	16%	7%	5%	15%	25%
5+	18%	3%	13%	9%	10%	6%	12%	18%

All pairwise comparisons (FFS vs. MC) within year and within dual category (not dual eligible or dual eligible) were statistically significant, using chi-square tests (P < 0.0001). Results for 2011 and 2012 were similar (not shown).

FFS, fee-for-service; MC, managed care.

Ambulatory care patterns

Among those who were not dual eligible, managed care beneficiaries in 2010 had on average fewer visits (P < 0.0001) but more providers (P < 0.0001) than fee-for-service beneficiaries, adjusting for age, sex, and case mix (Table 3). On average, these managed care beneficiaries had a smaller proportion of their visits with the most frequently seen provider (0.58 vs. 0.65 [adjusted P < 0.0001]) and therefore more fragmentation (0.58 vs. 0.51 [adjusted P < 0.0001]) than fee-for-service beneficiaries. These patterns persisted over time (Table 3). For example, in 2013, managed care beneficiaries in adjusted models had an average of 11.5 visits and 4.1 providers (compared to 11.9 visits and 3.7 providers for fee-for-service beneficiaries, P < 0.0001), leading managed care beneficiaries to have fewer visits with the most frequently seen provider (P < 0.0001) and more fragmented care (P < 0.0001) (Table 3).

Table 3.

Ambulatory Care Patterns of Medicaid Beneficiaries Over Time, Stratified by Dual Eligible Status and Insurance Product^*

	2010				2013
	Not dual eligible		Dual eligible		Not dual eligible		Dual eligible
	FFS	MC	FFS	MC	FFS	MC	FFS	MC
UNADJUSTED
Total number of ambulatory visits, mean	12.3	10.4	13.9	13.9	12.1	11.5	13.0	13.9
Total number of unique ambulatory providers, mean	3.6	3.8	4.5	4.4	3.8	4.1	4.5	4.4
Proportion of total visits with the most frequently seen provider, mean	0.64	0.58	0.55	0.56	0.56	0.55	0.54	0.56
Fragmentation score, mean	0.53	0.60	0.62	0.61	0.56	0.63	0.64	0.61
ADJUSTED^†
Total number of ambulatory visits, mean	11.4	10.9	11.8	12.4	11.9	11.5	11.4	10.9
Number of primary care visits, mean^‡	8.5	6.0	7.0	6.6	7.6	6.5	6.5	6.2
Number of specialist visits, mean^‡	4.8	5.9	6.3	7.0	6.1	6.2	6.5	6.7
Total number of ambulatory providers, mean	3.3	3.8	4.2	4.2	3.7	4.1	4.2	3.8
Number of primary care providers, mean^‡	1.9	1.7	1.9	1.8	1.9	2.1	1.8	1.8
Number of specialist providers, mean^‡	2.2	2.7	2.9	3.0	2.5	2.7	2.9	2.7
Proportion of total visits with the most frequently seen provider, mean	0.65	0.58	0.56	0.56	0.61	0.56	0.55	0.58
Fragmentation score, mean	0.51	0.58	0.60	0.59	0.54	0.61	0.62	0.59

This study includes only those beneficiaries with ≥4 ambulatory visits. Fragmention scores are reversed Bice-Boxerman scores; higher scores reflect more fragmented care.

†

Adjusted for age, sex, and Charlson-Deyo score, using Poisson models (for numbers of visits and providers), gamma models (for proportion with the most frequently seen provider), and bounded Tobit models (for fragmentation scores). All unadjusted and adjusted pairwise comparisons (FFS vs. MC) within year and within dual eligible category (dual eligible or not dual eligible) were statistically significant (P < 0.0001). Results for 2011 and 2012 were similar (not shown).

‡

The sum of (the average primary care utilization + the average specialist utilization) may be greater than the total observed, in part because the reported numbers are adjusted using different models.

FFS, fee-for-service; MC, managed care.

Also, among those who were not dual eligible, when visits and providers were stratified by primary care vs. specialist categories it was found that, in 2010, managed care beneficiaries on average had fewer primary care visits, fewer primary care providers, more specialist visits, and more specialist providers than their fee-for-service counterparts (adjusted P < 0.0001 for each comparison, Table 3). These patterns persisted over time.

Among dual eligible beneficiaries, the differences between managed care and fee-for-service beneficiaries were smaller (than among non-dual eligibles) and, although some comparisons may be statistically significant, they may not be clinically significant (such as the difference in adjusted average fragmented scores in 2010: 0.59 for managed care dual eligible beneficiaries and 0.60 for fee-for-service dual eligible beneficiaries (Table 3).

The main findings persisted when the sample was restricted to only those in the top 20% of ambulatory visits (≥13 visits in a year). Among those who were not dual eligible, in 2010, managed care beneficiaries had on average fewer visits (20.5 vs. 21.6), more providers (5.7 vs. 4.9), and more fragmented ambulatory care (0.65 vs. 0.55) than their fee-for-service counterparts (adjusted P < 0.0001 for each comparison, Table 4). The same patterns were observed in 2013 (Table 4). In this top 20% of beneficiaries, trends in primary and specialty visits are similar to the trends in population overall, with fewer primary care visits and more specialist visits each year (Table 4).

Table 4.

Sensitivity Analysis of Ambulatory Care Patterns Among Medicaid Beneficiaries in the Top 20% of Ambulatory Visit Utilization (≥13 Ambulatory Visits Each Year)^*

	2010		2013
	Not dual eligible		Not dual eligible
	FFS	MC	FFS	MC
	(N = 30,340)	(N = 175,507)	(N = 14,635)	(N = 305,181)
ADJUSTED^†
Total number of ambulatory visits, mean	21.6	20.5	21.5	20.8
Number of primary care visits, mean^‡	15.4	10.1	12.8	10.8
Number of specialist visits, mean^‡	7.7	10.9	9.7	10.8
Total number of ambulatory providers, mean	4.9	5.7	5.4	6.0
Number of primary care providers, mean^‡	2.5	2.1	2.4	2.7
Number of specialist providers, mean^‡	2.9	4.1	3.4	3.9
Proportion of total visits with the most frequently seen provider, mean	0.60	0.50	0.55	0.48
Fragmentation score, mean	0.55	0.65	0.60	0.68

This study includes only those beneficiaries with ≥4 ambulatory visits. Fragmention scores are reversed Bice-Boxerman scores; higher scores reflect more fragmented care.

†

Adjusted for age, sex, and Charlson-Deyo score, using Poisson models (for numbers of visits and providers), gamma models (for proportion with the most frequently seen provider), and bounded Tobit models (for fragmentation scores). All pairwise comparisons (FFS vs. MC) within year and within dual eligible category (dual eligible or not dual eligible) were statistically significant (P < 0.0001).

‡

FFS, fee-for-service; MC, managed care.

Discussion

In this study of adult Medicaid beneficiaries in New York State, among those who were not dual eligible, managed care was successful in achieving lower rates of ambulatory visit utilization than fee-for-service. However, the smaller number of visits for managed care beneficiaries was distributed across a larger number of providers, leading managed care beneficiaries to have significantly more fragmented care than their fee-for-service counterparts, even after adjusting for age, sex, and case mix. These patterns persisted over time and also persisted in the subset of beneficiaries in the top 20% of ambulatory utilization.

These patterns were not observed in the subset of beneficiaries who were dual eligible. It is not clear why there was not a greater difference between managed care and fee-for-service for the dual eligible population. One possibility is that the time period of the study predated robust managed care interventions for dual eligibles, as that requires alignment with Medicare and was pursued through a specific initiative in New York State after this study period ended.¹⁸

Previous studies have compared rates of hospitalization among adult Medicaid managed care vs. fee-for-service beneficiaries, with mixed results.^19,20 Other previous studies comparing adult Medicaid managed care vs. fee-for-service have measured the relative performance on selected quality measures, beneficiaries' self-reported access to care, and/or satisfaction with care.^21

–26 However, this is the first study to the research team's knowledge to compare managed care vs. fee-for-service Medicaid beneficiaries on their patterns of ambulatory care, including the relative extent of health care fragmentation in these populations.

This study suggests that in managing health care utilization, payers, providers, and patients should consider not only how many visits a patient has but how many providers he or she sees and the proportion of care delivered by the most frequently seen provider. Providers do not consistently share clinical information with each other.²⁷ Thus, when care is provided in a fragmented pattern (with many providers and with no dominant provider), there is a substantial risk that relevant clinical information will be missing for some or all of those providers.^28
–30 Previous work has shown that more fragmented ambulatory care is associated with more testing, more procedures, more emergency department visits, and more hospitalizations than less fragmented ambulatory care.^7,9,11,13,31 Therefore, while patterns of ambulatory care are important in their own right, they also have implications for downstream health care utilization and costs.

Data from this study suggest that the greater fragmentation in the non-dual managed care beneficiaries was driven by specialist utilization rather than primary care utilization. That is, managed care beneficiaries saw more specialists than the fee-for-service beneficiaries, even after adjusting for case mix. This pattern was especially prominent among the top 20% of beneficiaries by number of visits. Having more specialist providers increases the scope of care coordination, while simultaneously decreasing the likelihood that care coordination across all involved providers will occur.^27,32,33 Having more specialist providers is also likely to trigger downstream health care utilization (eg, more tests and procedures) that increases the cost of care.¹³ Managing specialist utilization is important for providers attempting to succeed under value-based purchasing contracts,^34,35 which are increasingly common in Medicaid.³⁶ However, strategies for how to effectively manage specialist utilization are not yet well developed,³⁷ and the results of this study underscore the need for them.

There are several limitations of this study. It is observational; therefore, unmeasured confounding cannot be ruled out, and the research team cannot make inferences of causality. It also is cross-sectional and compares populations of beneficiaries with Medicaid managed care to those with fee-for-service; it does not follow over time specific individuals who may change from one insurance product to the other. The analysis cannot determine medical appropriateness and does not capture communication (or lack thereof) across providers or patient outcomes. The team did not analyze care for specific diagnoses or by specific specialists, such as care for diabetes or care from obstetrician–gynecologists, and future studies could address care patterns in greater detail. This analysis of dual eligible beneficiaries was able to include only those claims that were paid by Medicaid, not those paid by Medicare; thus, the observations of ambulatory utilization for dual eligibles (which were higher than the observations of ambulatory utilization for non-duals) still may be an underestimate of all of their utilization. This study took place in New York State in 2010–2013 and can serve as a baseline, prior to implementation of Medicaid redesign, which began in 2014 and is continuing through 2020.³⁸ Additional studies are needed in other states as well. Future studies could also measure the effect of fragmentation (as well as the effect of interventions to reduce fragmentation) on outcomes, such as quality and cost, for fee-for-service vs. managed care beneficiaries.

In conclusion, this study found that non-dual Medicaid managed care beneficiaries had fewer visits but more providers and therefore more ambulatory fragmentation than their fee-for-service counterparts, even after adjusting for age, sex, and case mix. These patterns persisted over all 4 years of the study, even as Medicaid expanded overall and even as the proportion of beneficiaries in managed care grew. This study suggests that managed care may be focusing on reducing visit utilization but may be missing opportunities to consolidate care in a smaller number of providers, which may decrease the risk of gaps in communication and, in doing so, may decrease the risk of harm.

Footnotes

Acknowledgments

The authors thank the New York State Department of Health for providing access to the data. This article does not necessarily reflect the views of the Commonwealth Fund or the New York State Department of Health.

Author Disclosure Statement

Dr. Pincus has received funding from the New York State Department of Health and the Commonwealth Fund for research/training programs, and has been a consultant or on an advisory committee for the National Committee for Quality Assurance and the Commonwealth Fund. The other authors declare that they have no conflicts of interest. This work was funded by the Commonwealth Fund (grant #20170769). The funder had no role in the study design, data collection, data analysis, interpretation, preparation or approval of the manuscript.

Supplementary Material

Supplementary Table S1

References

Iglehart

. Expanding eligibility, cutting costs—a Medicaid update. N Engl J Med, 2012; 366:105–107.

New York State Department of Health. Managed care. 2018. https://www.health.ny.gov/health_care/managed_care/ Accessed April 3, 2019 .

Henry

. Kaiser Family Foundation. Status of state action on the Medicaid expansion decision. 2018. https://www.kff.org/health-reform/state-indicator/state-activity-around-expanding-medicaid-under-the-affordable-care-act/ Accessed April 3, 2019 .

Centers for Medicare and Medicaid Services. Managed Care in New York. 2011. https://www.medicaid.gov/medicaid-chip-program-information/by-topics/delivery-systems/managed-care/downloads/new-york-mcp.pdf Accessed April 3, 2019 .

Saunders

, Alexander

. Turning and churning: loss of health insurance among adults in Medicaid. J Gen Intern Med, 2008; 24:133–124.

National Committee for Quality Assurance. HEDIS Volume 2: Technical Specifications. 2015. http://www.ncqa.org/HEDISQualityMeasurement/HEDISMeasures/HEDIS2015.aspx. Accessed April 3, 2019 .

Kern

, Seirup

, Casalino

, Safford

. Healthcare fragmentation and the frequency of radiology and other diagnostic tests: a cross-sectional study. J Gen Intern Med, 2017; 32:175–181.

Centers for Medicare and Medicaid Services. National Plan & Provider Enumeration System (NPPES). 2018. https://nppes.cms.hhs.gov/#/ Accessed April 3, 2019 .

Nyweide

, Anthony

, Bynum

, et al. Continuity of care and the risk of preventable hospitalization in older adults. JAMA Intern Med, 2013; 173:1879–1885.

10.

Bice

, Boxerman

. A quantitative measure of continuity of care. Med Care, 1977; 15:347–349.

11.

Hussey

, Schneider

, Rudin

, Fox

, Lai

, Pollack

. Continuity and the costs of care for chronic disease. JAMA Intern Med, 2014; 174:742–748.

12.

Pollack

, Hussey

, Rudin

, Fox

, Lai

, Schneider

. Measuring care continuity: a comparison of claims-based methods. Med Care, 2016; 54:e30–e34.

13.

Romano

, Segal

, Pollack

. The association between continuity of care and the overuse of medical procedures. JAMA Intern Med, 2015; 175:1148–1154.

14.

Charlson

, Pompei

, Ales

, MacKenzie

. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis, 1987; 40:373–383.

15.

Deyo

, Cherkin

, Ciol

. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol, 1992; 45:613–619.

16.

Henry

. Kaiser Family Foundation. Dual eligible. 2018. https://www.kff.org/tag/dual-eligible/ Accessed April 3, 2019 .

17.

SAS Institute, Inc. SAS/ETS(R) 13.2 User's Guide. The QLIM procedure. 2014. http://support.sas.com/documentation/cdl/en/etsug/67525/HTML/default/viewer.htm#etsug_qlim_overview.htm. Accessed April 3, 2019 .

18.

New York State Department of Health. Introducing FIDA: New York State's Fully Integrated Duals Advantage (FIDA) Program. 2015. https://www.health.ny.gov/health_care/medicaid/redesign/fida/fida_fact_sheet.htm. Accessed April 3, 2019 .

19.

Basu

, Friedman

, Burstin

. Managed care and preventable hospitalization among Medicaid adults. Health Serv Res. 2004; 39:489–510.

20.

Bindman

, Chattopadhyay

, Osmond

, Huen

, Bacchetti

. The impact of Medicaid managed care on hospitalizations for ambulatory care sensitive conditions. Health Serv Res, 2005; 40:19–38.

21.

Carey

, Weis

. Diagnostic testing and return visits for acute problems in prepaid, case-managed Medicaid plans compared with fee-for-service. Arch Intern Med, 1990; 150:2369–2372.

22.

Cook

. Effect of Medicaid managed care on racial disparities in health care access. Health Serv Res, 2007; 42:124–145.

23.

Coughlin

, Long

. Effects of Medicaid managed care on adults. Med Care, 2000; 38:433–446.

24.

Garrett

, Davidoff

, Yemane

. Effects of Medicaid managed care programs on health services access and use. Health Serv Res, 2003; 38:575–594.

25.

Seligman

, Chattopadhyay

, Vittinghoff

, Bindman

. Racial and ethnic differences in receipt of primary care services between medicaid fee-for-service and managed care plans. J Ambul Care Manage, 2007; 30:264–273.

26.

Sisk

, Gorman

, Reisinger

, Glied

, DuMouchel

, Hynes

. Evaluation of Medicaid managed care. Satisfaction, access, and use. JAMA, 1996; 276:50–55.

27.

O'Malley

, Reschovsky

. Referral and consultation communication between primary care and specialist physicians: finding common ground. Arch Intern Med, 2011; 171:56–65.

28.

Anderson

. Chronic care: making the case for ongoing care. 2010. http://www.rwjf.org/en/research-publications/find-rwjf-research/2010/01/chronic-care.html. Accessed April 3, 2019 .

29.

Berwick

, Hackbarth

. Eliminating waste in US health care. JAMA, 2012; 307:1513–1516.

30.

Smith

, Araya-Guerra

, Bublitz

, et al. Missing clinical information during primary care visits. JAMA, 2005; 293:565–571.

31.

Kern

, Seirup

, Rajan

, Jawahar

, Stuard

. Fragmented ambulatory care and subsequent emergency department visits and hospital admissions among Medicaid beneficiaries. Am J Manag Care, 2019; 25:107–112.

32.

Pham

, O'Malley

, Bach

, Saiontz-Martinez

, Schrag

. Primary care physicians' links to other physicians through Medicare patients: the scope of care coordination. Ann Intern Med, 2009; 150:236–242.

33.

Pham

, Schrag

, O'Malley

, Wu

, Bach

. Care patterns in Medicare and their implications for pay for performance. N Engl J Med, 2007; 356:1130–1139.

34.

Barnett

, McWilliams

. Changes in specialty care use and leakage in Medicare accountable care organizations. Am J Manag Care, 2018; 24:e141–e149.

35.

McWilliams

, Chernew

, Dalton

, Landon

. Outpatient care patterns and organizational accountability in Medicare. JAMA Intern Med, 2014; 174:938–945.

36.

Houston

. Approaching the tipping point? Medicaid ACO momentum continues to build. 2015. https://www.chcs.org/approaching-tipping-point-medicaid-aco-momentum-continues-build/ Accessed April 3, 2019 .

37.

Song

, Sequist

, Barnett

. Patient referrals: a linchpin for increasing the value of care. JAMA, 2014; 312:597–598.

38.

Bachrach

, Bernstein

, Augenstein

, Lipson

, Ellis

. Implementing New York's DSRIP program: implications for Medicaid payment and delivery system reform. 2016. www.commonwealthfund.org/publications/fund-reports/2016/apr/new-york-dsrip-medicaid Accessed April 3, 2019 .

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.02 MB