DPH’s Process for Identifying and Notifying People Who May Be Eligible for WIC Benefits Is Not Effective.

Audit calls on DPH to revise its data-matching method to improve outreach to potential program participants.

Overview

The process DPH uses to identify and notify people who may be eligible for, but are not receiving, WIC benefits is not effective. First, DPH does not include recipients of the Department of Transitional Assistance’s (DTA’s) Transitional Aid to Families with Dependent Children (TAFDC) benefits—who are automatically income-eligible for WIC—in the data-matching process,¹⁰ even though DPH already has data-sharing agreements with DTA and the Office of Medicaid that enable it to identify potential WIC enrollees from the Supplemental Nutrition Assistance Program (SNAP) and MassHealth. Because TAFDC recipients are not included in that process, some of these WIC-eligible candidates are not identified for outreach efforts.

Second, DPH officials told us that based on the data matching they conduct, they send postcards to notify people who are identified as potentially eligible to receive WIC benefits. However, none of the 6,740 WIC participants who responded to DPH’s 2016 WIC participant satisfaction survey indicated that they first heard about WIC by receiving a postcard from DPH, and only four participants indicated that they first heard about WIC via unspecified mailings. Also, we visited 20 of the 120 WIC provider locations and asked each provider’s staff if their applicants ever cited the WIC postcards as a reason for applying for WIC benefits. None of the staff members to whom we spoke could recall an applicant mentioning a WIC postcard. Thus there is no indication that postcards are an effective method of notifying potential WIC enrollees that they might be eligible to participate.

Third, DPH’s data-matching process requires that four of six data fields match in order for DPH to identify a person as possibly eligible for WIC benefits and attempt to notify the person. This may be too many fields, resulting in unnecessary notifications being sent to people already receiving WIC benefits. During our fieldwork, a DPH employee found that matching on fewer data fields would probably eliminate many instances of matches not being identified because of issues like slight differences in addresses. In fact, the DPH employee who performed the data matching for us using the original method offered to perform a one-time data matching using the alternative method. Using that method, DPH demonstrated improved efficiency by reducing the number of unmatched women from the Medicaid Management Information System by 55.9% and the number of unmatched children by 40.2%. It also reduced the percentage of unmatched women from DTA’s Benefit Eligibility and Control Online Network information system by 80.0% and the percentage of unmatched children by 47.2%.

Authoritative Guidance

DPH has data-sharing agreements with the Office of Medicaid and DTA. The purpose of the data-sharing agreements, as stated therein, is to provide targeted outreach to families who receive MassHealth or SNAP benefits and who are eligible, but not enrolled, for WIC benefits. To this end, DPH should ensure that it uses the most efficient and effective data-matching process and outreach methods, which should include information on TAFDC recipients.

Reasons for Ineffectiveness and Inefficiency

DPH does not request the data for TAFDC recipients from DTA because, according to EOHHS, most SNAP recipients also receive TAFDC benefits and therefore the current data-matching process covers most WIC-eligible consumers. However, in OSA’s opinion, because the populations of SNAP recipients and TAFDC recipients are not the same, some recipients of TAFDC who are income eligible to receive WIC benefits may not be getting notified about their eligibility status.

In addition, DPH does not have a process for analyzing whether postcards are an effective method of contacting potential WIC participants compared with other possible options, such as emails or text messages.

Finally, the data-sharing agreement between DPH and the Office of Medicaid was executed in 2006, and the agreement between DPH and DTA was executed in 2008. The original data-matching processes defined in the data-sharing agreements established that the data matching should be done on six variables, and the agreements have never been updated to allow for a more effective process. Neither agreement has been reviewed since it was executed.

Recommendations

1. DPH should modify its data-sharing process to include TAFDC recipients.
2. DPH should implement a process to analyze whether postcards are an effective method of contacting potential WIC enrollees compared with other possible options, such as emails or text messages.
3. DPH should adopt the revised data-matching method it found to improve the results of data matching.

Auditee’s Response

DPH disagrees that WIC’s process for identifying and notifying people who may be eligible for WIC benefits is not effective. DPH is committed to ensuring all women and infants who are eligible receive services. While we acknowledge that there is always room for improvement, we are using effective methods including postcards and one on one outreach in locations frequented by low income women. DPH is currently engaged in a pilot project that will evaluate the effectiveness of conducting outreach to MassHealth members via text. Should the pilot yield positive results, this method will also be implemented. . . .

[The] assessment of the accuracy of the original match methodology isn’t valid. What DPH was being asked to do was to consume a file that was ten times larger than typically received and that file contained multiple duplicate records which caused the unmatched count to appear much higher than it was when using the original method. Prior to conducting the alternative record match, the file was de-duplicated. The de-duplicated file showed different results.

Auditor’s Reply

Although we do not dispute DPH’s assertion that it is committed to ensuring that all women and infants who are eligible receive services, our audit identified a number of issues with the process DPH uses to identify and notify people who may be eligible for WIC benefits, which in OSA’s opinion raises concerns about the effectiveness of this process. For example, not including TAFDC recipients in the identification process may result in WIC-eligible candidates not being identified and notified of their eligibility for those benefits. Further, DPH says it sends postcards to people it has determined might be eligible for WIC benefits but does not assess what effect, if any, these postcard notifications have on program enrollment. As noted above, our audit found that the postcard notification process may have a very limited effect on WIC enrollment, since none of the 6,740 WIC participants who responded to DPH’s 2016 WIC participant satisfaction survey indicated that they first heard about WIC by receiving a postcard from DPH, and only four participants indicated that they first heard about WIC via unspecified mailings. Also, during OSA’s visit to 20 WIC provider locations, none of the staff members to whom we spoke could recall an applicant mentioning a WIC postcard. Thus there is no indication that postcards are an effective method of notifying potential WIC enrollees that they might be eligible to participate.

To assess the effectiveness of EOHHS agencies’ data-matching process, we requested the data for all people who received MassHealth, SNAP, TAFDC, or WIC benefits during the audit period. However, this information was never provided (see the “Audit Constraints” section of this report). Instead of providing OSA with the requested data, EOHHS proposed and defined a data-matching method to be conducted by DPH and observed by OSA. The data-matching analysis presented is the result of this data-matching method agreed upon by EOHHS, DPH, and OSA.

Based on its comments, DPH is taking some measures to improve its WIC eligibility notification process, but we urge DPH to fully implement our recommendations.

10.    The records from each information system are compared using date of birth and last name first. The two records are then considered matched if any four of the following six data fields are the same: first name, middle initial, address line 1, address line 2, city, and ZIP code. The two records are considered unmatched if fewer than four of the six data fields are the same.

Date published:	September 9, 2019