Medication for opioid use disorder (MOUD) — methadone or buprenorphine — is recommended for people with opioid use disorder (OUD) during pregnancy. But there are gaps in what is known about the risks and benefits of each regimen, and whether particular medicines or prescribing patterns lead to better outcomes. And the problem grew: from 1999 to 2014, OUD among pregnant women in the United States quadrupled, from 1.5 to 6.5 per 1,000 delivery hospitalizations.
To start answering those questions, CDC built MAT-LINK — the Maternal and Infant Network to Understand Outcomes Associated with Medication for Opioid Use Disorder During Pregnancy. It is the first surveillance system to collect comprehensive, long-term, linked data on pregnant people and their infants across multiple clinical sites. Its first report describes how it works and who is in it.
The network
The seven MAT-LINK clinical sites. Credit: CDC, Morbidity and Mortality Weekly Report*.*
| Clinical sites | |
|---|---|
| 2019 (initial) | Boston Medical Center, Kaiser Permanente Northwest, The Ohio State University, University of Utah |
| 2021 (added) | University of New Mexico, University of Rochester, University of South Florida |
CDC runs the project with the Public Health Informatics Institute, funded through the Office of the Assistant Secretary for Planning and Evaluation's Patient-Centered Outcomes Research Trust Fund. The first sites were chosen for strong data systems, care protocols covering several MOUD regimens, and the ability to link mothers' and children's records and follow children to age 2; the 2021 expansion extended child follow-up to age 6, to study neurodevelopment and school-age outcomes.
What it collects
The network includes every known pregnancy outcome from January 1, 2014, through August 31, 2021, in a person with a diagnosis code for OUD during that pregnancy. Records follow the pregnant person from prenatal care to a year after birth (with diagnosis and procedure codes up to 6 years after), and the child from birth to age 6.
- During pregnancy: demographics, visits, medications, lab and ultrasound results, substance exposure, infections, and support such as behavioral therapy and peer programs.
- MOUD: when it started, how long it lasted and at what dose, plus medicines that interact with it, inpatient or residential stays, substance use and overdoses.
- Birth: pain management, delivery type, newborn care and measurements, neonatal abstinence or opioid withdrawal syndrome, readmissions, and every pregnancy outcome, including stillbirth, miscarriage and termination.
- Child and postpartum: growth, development, illnesses, vaccinations and referrals; the parent's anxiety, depression, contraception and substance use.
Data come from electronic health records, pharmacy and lab systems and public health records, either extracted automatically or abstracted by hand from charts. Sites check the data, CDC checks them again, and everything is protected under an Assurance of Confidentiality.
Who is in it
Of 5,541 pregnancies, 4,381 (79.1%) involved MOUD at some point and 1,160 (20.9%) did not.
| MOUD | No MOUD | All | |
|---|---|---|---|
| Mean age | 30.0 | 28.7 | 29.7 |
| White | 87.9% | 80.3% | 86.3% |
| Black or African American | 4.7% | 10.2% | 5.8% |
| Hispanic or Latino | 25.1% | 26.5% | 25.4% |
| Public insurance | 84.6% | 70.5% | 81.6% |
| Private insurance | 13.3% | 25.9% | 15.9% |
| Urban core | 84.3% | 85.2% | 84.4% |
| Rural | 7.1% | 6.2% | 6.9% |
People receiving MOUD were more likely to be older, White and publicly insured; those not receiving it tended to be younger and privately insured. There was no significant difference by ethnicity or by urban or rural residence. The authors call these differences important markers of possible differences in access to care and clinical care, and note that differences by race reflect systemic racism and implicit bias, not physiology.
How well the system works
CDC evaluated MAT-LINK against its guidelines for public health surveillance systems, drawing on documents and 16 interviews with clinicians, scientists, data specialists and the funder.
- Flexible: new sites, variables and longer follow-up can be added with little change.
- Representative, up to a point: the first four sites left out the South and Southwest and had low shares of Black (7.5%), Hispanic (6.6%) and uninsured (under 1.0%) patients; the 2021 sites improved that, but representativeness remains limited.
- Demanding: some sites underestimated the time needed to extract data by three times, so about 10% of variables were made optional.
- Blind spots: people without an OUD diagnosis — such as those treated with opioids for chronic pain — are not captured, nor are some who delivered at a site after prenatal care elsewhere, and non-live-birth outcomes appeared under-counted.
- Accurate where testable: checked against lab tests, a hepatitis C diagnosis code had a predictive value of 95%.
Limits and what comes next
The data are preliminary; outcome data for mothers and babies were not ready and will be analyzed from 2023. Race, ethnicity and insurance may be recorded inconsistently across sites, and education is too poorly documented to include. A new funding opportunity will build on MAT-LINK, adding sites and following pregnancies with MOUD exposure through 2027.
Sources
Based on Kathryn Miele, Shin Y. Kim, et al., "Medication for Opioid Use Disorder During Pregnancy — Maternal and Infant Network to Understand Outcomes Associated with Use of Medication for Opioid Use Disorder During Pregnancy (MAT-LINK), 2014–2021," MMWR Surveillance Summaries, volume 72, Centers for Disease Control and Prevention; a work of the United States government in the public domain.
Licence: CC0 1.0 (public domain) · Adapted from www.cdc.gov
1
0
0
0
Comments
