Enrollment Barriers, Procedural Denials, and Loss of Medicaid Coverage: A Randomized Trial to Identify Effective Outreach Solutions

This study will test new outreach methods to improve access to health insurance coverage among people who lost Medicaid for procedural reasons. Specifically, we propose to evaluate the impacts of outreach from Navigators on gaps in Medicaid coverage among people disenrolled from Medicaid for procedural reasons. The implementation partner is Covering Wisconsin, Wisconsin’s Navigator program, which has a staff of professional enrollment assisters who inform and advise consumers on their Medicaid eligibility as well as enrollment and renewal processes on a one-on-one basis. The experimental arms will provide two forms of outreach (a pre-recorded outbound call providing a hotline number vs. a live outbound call from a Navigator). The control arm includes those not exposed to outreach—an available group given the capacity limits of Navigators.

The interventions address potential barriers to successful renewal, including lack of awareness of eligibility for Medicaid or the need for renewal or enrollment paperwork, and uncertainty about how to respond to information requests related to income or asset verification. Navigators will conduct initial outreach in English, Spanish, or Hmong according to the preference indicated in the DHS data. Details are as follows:

Live outbound call arm: Households in this arm will receive a call from a Navigator who can provide personalized assistance with health insurance enrollment. If this call results in a phone conversation, the Navigator will explain the status of the consumer’s renewal application and answer questions about eligibility and the paperwork required to renew or re-enroll in Medicaid coverage or other available coverage. If the consumer does not answer, a pre-recorded message will be left identical to the pre-recorded call arm.

Pre-recorded call arm: Households in this arm will receive an outbound call message with a hotline number and information about the availability of Navigators to help with the Medicaid enrollment process and answer related questions. Message wording will be informed by the research team’s understanding of the most common administrative barriers and the reasons these barriers are especially challenging.

Both types of calls will be conducted in the 90 days following the member’s renewal deadline. Within this period members can attempt renewal and, if successful, Medicaid will provide up to 3 months’ retroactive payment for any covered services provided during the gap in coverage.

2025-02-03

2026-07-20

Reinstating Medicaid coverage within 3 months of their renewal deadline.

Under Wisconsin’s renewal policy mentioned above, those who renew in this timeframe can maintain continuous Medicaid coverage ex post, even after losing coverage for procedural reasons.

Secondary outcomes include Medicaid coverage during the 6 or 12 months after the renewal deadline (measured each month); the duration of any coverage gaps up to 6 or 12 months after the renewal deadline; whether the individual spoke with Covering Wisconsin either by calling the hotline or receiving a live outbound call; whether the individual called the Covering Wisconsin hotline; and whether someone used any health care covered by Medicaid in the 3 months, 6 months, or 12 months after the renewal deadline.

The experimental population will include people in Wisconsin who lost Medicaid coverage for procedural reasons during an 18 month period. Exclusion criteria include preferring a language other than English, Spanish, or Hmong, and lacking a working phone number.

Assignment to treatment arms will occur monthly. Randomization will be clustered by household. Randomization will be stratified by key characteristics that can affect eligibility and enrollment processes.

Households will be equally assigned across the three treatment arms starting in July 2025, the fifth month of the study; this reflects increasing capacity of navigators to make outbound calls in the early months of the study. In the first month, 750 live calls were placed. In the second month 1500 live calls were placed. In the third and fourth months, we were able to increase that number to 2,000 calls per month and 2,500 calls per month, respectively, as it became clear that there was capacity to place more live calls. (In a treatment on the treated analysis, we will account for the possibility that if the call list is particularly long in a given month, not everyone on the list will receive an outbound call.)

Randomization done in office by a computer

Household

Yes

Based on an analysis of microdata from Wisconsin and state-level and national data from KFF, we estimate around 130,000 households will meet all the inclusion criteria.

Based on the number of people who lost Medicaid coverage each month for procedural reasons during the first few months of trial implementation, we anticipate there will be 169,692 households (288,738 people) in the study sample once the full 18-month outreach period is complete.

Live outbound call from a Navigator ~47,000 households
Pre-recorded outbound call ~61,000 households
No-outreach control arm ~61,000 households

Based on this sample size and the projected division of households across treatment arms, the study will have 80% power to detect impacts of the live call on Medicaid coverage within 6 months as small as 0.76 percentage points, impacts of the pre-recorded call as small as 0.68 percentage points, and differences between the live and pre-recorded calls as small as 0.76 percentage points. These calculations account for clustering by household (i.e., one person is contacted in each household, and the outcomes of people within the household are related). The calculations do not factor in the efficiency gains from incorporating covariates in the model, and thus are likely conservative (i.e., the true minimum detectable effect size is likely smaller).

Pre-analysis plan v07012025.docx