New York State 1115 Demonstration Independent Evaluation

Interim Report

by Harry H. Liu, Andrew W. Dick, Nabeel Qureshi, Sangita M. Baxi, Katherine J. Roberts, J. Scott Ashwood, Laura A. Guerra, Teague Ruder, Regina A. Shih

This Article

RAND Health Quarterly, 2022; 9(3):5

Abstract

The broad goals of New York State's Medicaid Section 1115 Waiver are to enroll a majority of Medicaid beneficiaries into managed care, increase access and service quality, and expand coverage to more low-income New Yorkers. The RAND Corporation was competitively selected as the independent evaluator to assess two components under this 1115 Demonstration Waiver: the Managed Long-Term Care (MLTC) program and the 12-month continuous eligibility policy, which guarantees enrollees Medicaid coverage regardless of changes in income in the 12 months after eligibility determination and enrollment. This final interim evaluation examines whether these two components have helped achieve the program's goals. The RAND team's analyses show that the Demonstration has expanded access to managed care through mandatory MLTC enrollment and 12-month continuous eligibility. The team found no evidence of a significant change in patient safety or quality of care. The authors note that, although this means that there is no evidence the Demonstration achieved the goal of improving quality of care, increasing access without compromising quality of care is a success in its own right.

For more information, see RAND RR-A951-1 at https://www.rand.org/pubs/research_reports/RRA951-1.html

Full Text

Evaluation Objective

The broad goals of New York State's Medicaid Section 1115 Waiver are to enroll a majority of Medicaid beneficiaries into managed care, increase access and service quality, and expand coverage to more low-income New Yorkers. To meet the special terms and conditions specified by the Centers for Medicare & Medicaid Services under New York State's 1115 Medicaid Redesign Team Waiver, the RAND Corporation was competitively selected as the independent evaluator to assess two components under this 1115 Demonstration Waiver: the Managed Long-Term Care (MLTC) program and the 12-month continuous eligibility policy. Starting in September 2012, the State required individuals who are over 21, eligible for both Medicare and Medicaid, and in need of 120 days or more of long-term services and supports (LTSS) to enroll in MLTC plans, which are paid on a capitated basis. The 12-month continuous eligibility policy was based on the Modified Adjusted Gross Income guideline and was implemented in January 2014 for individuals eligible for Medicaid, including pregnant women; childless adults who are not pregnant, are younger than 65, and are not on Medicare; parents or caretaker relatives; and individuals eligible for the Family Planning Benefit Program. Individuals who qualified for 12-month continuous eligibility were guaranteed Medicaid coverage regardless of changes in income in the 12 months after eligibility determination and enrollment. This final interim evaluation examines whether these two programs have achieved the following:

  • expanding access to LTSS and improving patient safety, quality of care, and consumer satisfaction (in the case of MLTC [Domain 1])
  • reducing enrollment gaps and increasing Medicaid enrollment duration (in the case of 12-month continuous eligibility [Domain 2]).

Analytical Approach

To achieve the goals of this final interim evaluation, RAND researchers conducted a number of analyses applying primarily a quasi-experimental study design and using various data sources provided by the New York State Department of Health (NYS DOH), including the 2010–2018 MLTC monthly enrollment by county; 2007–2019 MLTC plan-level aggregate data on patient safety, quality of care, and consumer satisfaction (these data years vary across different outcome measures); and 2012–2018 Medicaid Data Warehouse data. The evaluation team described the trends in various outcomes over time and conducted statistical modeling and testing to answer the evaluation questions.

Findings and Conclusions

The results of our analyses showed that the MLTC mandate was associated with a large increase in MLTC enrollment during 2012–2018, with its effect having stabilized by month 19, i.e., by the time the most recent demonstration period started (December 2016 to March 2021); there is no evidence of a decline in patient safety, quality of care, or consumer satisfaction, except for a decrease in satisfaction with care managers (Table 1). Among those who transitioned from institutional settings to community settings, enrollment in MLTC increased during 2015–2018, but no statistically significant changes in patient safety and quality of care were observed except for an increase in receipt of dental exams.

The 12-month continuous eligibility policy was associated with a moderate increase in Medicaid enrollment duration among adults but a decline in monthly Medicaid cost, resulting in a small net increase in total Medicaid cost. The policy's impact was smaller among individuals enrolled through the Welfare Management System (WMS), administered by local departments of social services, than among those enrolled through New York State of Health (NYSoH), the State's health insurance exchange.

Table 1. Summary of Evaluation Results

Domain Goal Outcome Result
Domain 1, Component 1: Managed Long-Term Care (MLTC) Goal 1: Expand access to MLTC for Medicaid enrollees in need of LTSS RQ1. Time for the MLTC mandate's effect on enrollment to stabilize

19 months, stabilizing at +0.6 percentage points per year; a 12-percentage point increase in enrollment rates during the 79 months post-mandate (p < 0.05)

Goal 2: Demonstrate stability or improvement in patient safety RQ1. Percentage of enrollees who had no emergency room visits

+0.8 percentage points (p > 0.05)

RQ2. Percentage of enrollees who had no falls that required medical intervention or resulted in major or minor injuries

−1.8 percentage points (p > 0.05)

Goal 3: Demonstrate stability or improvement in quality of care RQ1. Receipt of timely care

−0.8 percentage points (p > 0.05)

RQ2. Influenza vaccination

+0.2 percentage points (p > 0.05)

RQ2. Dental exam

−5.6 percentage points (p > 0.05)

Goal 4: Stabilize or reduce preventable acute hospital admissions RQ1. Potentially avoidable hospitalizations

−1.3 hospitalizations per 10,000 enrollee days (p > 0.05)

Goal 5: Demonstrate stability or improvement in consumer satisfaction RQ1. Satisfaction with MLTC plans

−1.8 percentage points (p > 0.05)

RQ2. Satisfaction with care managers

−3.1 percentage points (p < 0.05)

RQ3. Satisfaction with provider timeliness

−2.2 percentage points (p > 0.05)

RQ4. Satisfaction with service quality

−1.2 percentage points (p > 0.05)

Domain 1, Component 2: Individuals Moved from Institutional Settings to Community Settings for LTSS Goal 1: Improve access to MLTC for those who transitioned from an institutional setting to the community RQ1. Enrollment in MLTC within one year post-discharge from an institution

7% in 2015; 60% in 2018 (p < 0.05)

Goal 2: Demonstrate stability or improvement in patient safety RQ1. Percentage of enrollees who had no emergency room visits

50% in 2015; 85% in 2018 (p > 0.05)

RQ2. Percentage of enrollees who had no falls that required medical intervention or resulted in major or minor injuries

50% in 2015; 93% in 2018 (p > 0.05)

Goal 3: Demonstrate stability or improvement in quality of care RQ1. Percentage in community within one year post-discharge from an institution

85% in 2015; 81% in 2018 (p > 0.05)

RQ2. Influenza vaccination

50% in 2015; 73% in 2018 (p > 0.05)

RQ2. Dental exam

50% in 2015; 64% in 2018 (p < 0.05)

Domain 2: Mainstream Medicaid Managed Care and Temporary Assistance to Needy Families (TANF) Goal 1: Increase access to health insurance through Medicaid enrollment—Express Lane Eligibility RQ1. Medicaid enrollment, RQ2. demographic characteristics, and RQ3. percentage of ineligible enrollees Removed from the evaluation
Goal 2: Limit gaps in Medicaid eligibility due to fluctuations in recipient income—12-month continuous eligibility RQ1: Percentage with at least 12, 24, or 36 months of enrollment among the population affected by the continuous eligibility policy

≥12 months: 47% in 2012; 58% in 2017 (p < 0.01) for NYSoH and 47% in 2012; 58% in 2017 for WMS (p < 0.01)

≥24 months: 23% in 2012, 32% in 2016 (p < 0.01) for NYSoH and 23% in 2012, 34% in 2016 for WMS (p < 0.01)

36 months: 13% in 2012; 18% in 2015 (p < 0.01) for NYSoH and 13% in 2012; 29% in 2015 for WMS (p < 0.01)

RQ2: Difference in percentage with at least 12, 24, or 36 months of enrollment by enrollee characteristics

Demographics: Older members, White and Hispanic members, and members with a lower health status more likely to have longer enrollment duration for NYSoH and WMS populations (all p < 0.01)

Geographic area: Individuals in New York City had longer enrollment durations that those not in New York City for NYSoH and WMS populations (all p < 0. 01)

RQ3: Average number of continuous enrollment months

+0.8 and +1.9 months in a 12- and 24-month post-policy period for NYSoH populations, respectively (p < 0.05)

+0.4 and +1.2 months in a 12- and 24-month post-policy period for WMS populations, respectively (p < 0.05)

RQ4: Probability of being continuously enrolled for at least 12 months

+0.19 probability of being enrolled for the NYSoH population (p < 0.05)

+0.14 probability of being enrolled for the WMS population (p < 0.05)

RQ5: Effect of the continuous eligibility policy on outpatient, inpatient, and emergency department visits and Medicaid cost of care

Utilization: −43 inpatient admissions, −295 outpatient visits, and −49 emergency room visits per 1,000 member-years for the NYSoH population (all p < 0.05)

−29 inpatient admissions (p < 0.05), +101 outpatient visits (p < 0.05), and +17 emergency room visits per 1,000 member-years for the WMS population (p > 0.05)

Medicaid cost: −$27 per member per month for the NYSoH population (p < 0.05), −$8 per member per month for the WMS population (p > 0.05)

RQ6: Increased number of enrollment months due to the continuous eligibility policy

+378k (p < 0.05), +1,030k (p < 0.05), +959k (p < 0.05), +1,046k (p < 0.05) enrollees for 2014–2017, respectively, for the NYSoH population

+530k (p < 0.05), +483k (p > 0.05) enrollees for 2016–2017, respectively, for the WMS population

RQ7: Percentage of individuals in fee for service (FFS) by calendar month

29% in January 2012; 23% in December 2018 (p < 0.01)

RQ8: Percentage in FFS for 1–2 months, among those with any MMC coverage in a year

All enrollees: 18% in 2012; 19% in 2018 (p < 0.01)

New enrollees: 25% in 2012; 36% in 2018 (p < 0.01)

NYSoH enrollees: 74% in 2014; 27% in 2018 (p < 0.01)

WMS enrollees: 8% in 2014; 6% in 2018 (p < 0.01)

RQ8: FFS enrollment months in the first enrollment year, among those with at least 6 months of MMC coverage in that year

−0.6 months of FFS enrollment during the first enrollment year (p < 0.01)

RQ9: Percentage of MMC enrollees remaining in the same MMC plan after the recertification, among those with at least 13 consecutive months of MMC coverage

All enrollees: 88% in 2013; 80% in 2018 (p < 0.01)

NYSoH enrollees: 70% in 2014; 77% in 2018 (p < 0.01)

WMS enrollees: 93% in 2014; 90% in 2018 (p < 0.01)

RQ10: Percentage of MMC enrollees who are auto-assigned to any health plan at the start of MMC enrollment

All enrollees: 6.6% in 2012; 4.4% in 2018 (p < 0.01)

NYSoH enrollees: ~0% in 2014; 2.7% in 2018 (p < 0.01)

WMS enrollees: 5.6% in 2014; 8.5% in 2018 (p < 0.01)

NOTE: RQ = research question. The color code: green represents favorable results, red unfavorable, and yellow neither. For Domain 1, Component 2, since no pre-MLTC mandate data were available, only the post-period trends are presented. Due to a large sample size of about 1 to 6 million individuals, the descriptive trend tests for Domain 2, Goal 2, RQs 7–10 result in small p values.

Domain 1, Component 1, Goal 1: MLTC Enrollment

The MLTC mandate increased enrollment rapidly and dramatically and then stabilized at a growth rate of about 0.05 percent per month, or 0.6 percent per year within 19 months of the mandate's implementation (Table 1). However, increases in enrollment and time for the MLTC mandate's effect on enrollment to stabilize differed across regions, suggesting that idiosyncratic factors may have affected implementation across the State. New York City, in which the mandate was implemented first, drove the results due to the size of its population compared to the rest of the State.

Domain 1, Component 1, Goals 2–5: Patient Safety, Quality of Care, and Consumer Satisfaction Among the MLTC Population

We found no evidence of changes in patient safety (percentage of enrollees who had no emergency room visits and percentage of enrollees who had no falls that required medical intervention or resulted in major or minor injuries) and quality of care (influenza vaccinations, dental exams, and potentially avoidable hospitalizations). Satisfaction measures remained high with MLTC, with no statistically significant evidence of decline occurring except for satisfaction with care managers. Thus, results indicate that MLTC plans were able to accommodate the large increases in enrollment without noticeably compromising patient safety, quality of care, or consumer satisfaction with care. These results are particularly important given the rapid and large increase in MLTC enrollment.

Domain 1, Component 2, Goals 1–3: Individuals Moved from Institutional Settings to Community Settings

Among those who transitioned from institutional to community settings, enrollment in MLTC increased, which is not surprising given that MLTC enrollment of new nursing home residents became mandatory starting in February 2015. We found no evidence of changes in patient safety measures (percentage of enrollees who had no emergency room visits and percentage of enrollees who had no falls that required medical intervention or resulted in major or minor injuries) among MLTC enrollees who transitioned from institutions to the community from 2015 through 2018. We also found that a substantial majority (66–85 percent) of the home- and community-based services (HCBS) expansion population remained in the community. Among the HCBS expansion population, the changes in influenza vaccination rates were not statistically significant. Receipt of dental exams increased, perhaps in response to a performance improvement project for MLTC enrollees during the period.

Domain 2, Goal 2: 12-Month Continuous Eligibility

There was an overall increasing trend in average Medicaid enrollment duration after the implementation of the 12-month continuous eligibility policy. Because of differences in operational processes, we analyzed the WMS and NYSoH populations separately. We found that the policy was associated with approximately 4- and 8-percent increases in enrollment duration among individuals enrolled in WMS and NYSoH, respectively. The policy impact in NYSoH could partially be attributed to the simplified and more convenient enrollment and renewal process under NYSoH versus WMS. The simultaneous implementation of the Medicaid expansion did not seem to affect the policy effect on enrollment because the estimates were similar after excluding the expansion population. In both NYSoH and WMS populations, we observed a statistically significant decline in annual patient admissions, as well as in average monthly Medicaid cost. Combining the increase in enrollment months and the decrease in monthly Medicaid cost, we estimated that the 12-month continuous eligibility policy has led to an increase in total Medicaid cost by about 3 percent. The State did make progress in reducing FFS enrollment and auto-assignment to a health plan at Medicaid managed care (MMC) enrollment start, although the proportion of MMC enrollees who stayed with the same plan after the 12-month recertification decreased during 2012–2018.

Limitations

We acknowledge that there are several major limitations to our evaluation. When examining MLTC enrollment rates, the number of dual eligible individuals was used as the denominator, but it is only a gross approximation of the actual eligible population. The definitions of some MLTC outcome measures changed over time, such as emergency room visits, falls, and perceived timely access to care, and such definitional changes made it difficult to evaluate changes in outcomes across years. Also, most of plan-level MLTC outcomes measures were risk-adjusted, and the adjustment methodologies changed over time. Because of the lack of individual-level data, we were not able to risk-adjust for the differences between voluntary enrollees before the MLTC mandate and new enrollees under the mandate, and these differences may affect the outcomes. The lack of individual-level data has also reduced the precision of our estimates of the impact of MLTC on outcomes.

In assessing the impact of 12-month continuous eligibility on Medicaid enrollment, our analysis is limited by the use of children as the control group. Children often have a broader income band, so that there is more room for income to fluctuate though they remain eligible for Medicaid. Furthermore, despite the difference-in-differences approach used in the analyses, we were not able to control for time-dependent changes that occurred simultaneously with the implementation of 12-month continuous eligibility policy and impacted the adult Medicaid population differently from the child population.

Conclusions

Based on the results of our analyses, the MLTC program under the 1115 Demonstration Waiver has achieved its goal of increasing access to LTSS through MLTC, as illustrated by the rapid expansion of MLTC across the State from 2012 through 2018. There is little evidence suggesting that the speed of the expansion has led to a significant change in patient safety or quality of care by the measures used in this evaluation.

We found that the 12-month continuous eligibility policy was associated with statistically significant increases in enrollment duration and outpatient visits, but decreases in inpatient admissions and per member per month Medicaid cost. When considering both increases in enrollment and decreases in per member per month Medicaid cost, the policy is associated with a net increase in total Medicaid cost. Finally, during 2012 through 2018, descriptive trends show that the State has been able to reduce the length of FFS enrollment among MMC enrollees.

The results for the most recent demonstration period (December 2016 to March 2021) covered by the data under this interim evaluation, i.e., December 2016 to December 2018 or 2019, showed similar trends or patterns to those from earlier post-policy years except that the MLTC mandate's impact on enrollment had stabilized by the end of 2016.

In brief, the State has achieved the Demonstration's first goal of expanding access to managed care through mandatory MLTC enrollment and 12-month continuous eligibility. Although we did not find evidence of improved quality, the second goal, increasing access without compromising quality of care, is a success in its own right. Questions remain about whether the MLTC mandate has generated efficiencies in spending––the third goal of the overall 1115 Demonstration––and the extent to which public reporting and quality assurance programs have affected quality of care. Future evaluations may be conducted to answer these questions to guide State policies.

This research was funded by the New York State Department of Health and carried out within the Payment, Cost, and Coverage Program in RAND Health Care.

RAND Health Quarterly is produced by the RAND Corporation. ISSN 2162-8254.