What do we know about birth match?

Research has identified a number of characteristics within families that may correlate to higher risk of severe injury or fatality from child abuse and neglect. Although the vast majority of families with these characteristics will not harm their children, even one death from child abuse and neglect is too many. As a result, jurisdictions increasingly are investigating how data within public and private information systems might be harnessed, using what we know from research, to more effectively identify children at risk of harm before the unthinkable occurs.1 Predicting risk of future harm is not an exact science, however, and a number of key tensions and considerations must be addressed and resolved.

This brief describes birth match as one attempt to use data to predict risk and prevent future harm, offering examples of jurisdictional implementation, a snapshot of the evidence, and highlights of the national conversation regarding the ethical dilemmas inherent in this approach.

What is birth match?

Birth match refers to the use of data to proactively identify newborns in families that have some prior significant history with child protective services (CPS). This typically occurs through a formal data-sharing agreement between the jurisdiction’s CPS agency and the department of vital statistics or other source of birth data. Historically, birth match strategies have been implemented by law or agency policy following a tragic death or severe injury of an infant that might have been prevented had CPS been aware of the child’s birth. This strategy was recommended by the U.S. Commission to Eliminate Child Abuse and Neglect Fatalities in its final report, issued in March 2016,2 and is currently employed in at least four states — Maryland, Michigan, Minnesota, and Texas — and in New York City.

Our best predictions of harm occur when a child is highly vulnerable and the parent has clearly demonstrated inadequate or unsafe parenting…. It’s very important [to start] to think about ways to narrow down to populations that we could really intervene with high levels of efficiency.

– Dr. Richard Barth, Dean and Professor, University of Maryland

How have jurisdictions implemented birth match?

Jurisdictional policy regarding birth match programs exists on a continuum. Tensions and ethical considerations increase the further the program moves from linking birth records to families with an open/active case to families with no prior involvement with child welfare but considered high risk.

In New York City, any infant born to a family with a sibling in foster care in the city must receive his or her own child protective investigation soon after birth. Caseworkers are responsible for identifying pregnancy and births in their own active cases. The investigation includes a child safety conference among parents, CPS and foster care workers, and other service providers. In most cases, this is expected to lead to placement of the infant in out-of-home care; exceptions must be reviewed and approved at the assistant commissioner level.4

In Maryland, the Department of Health and Mental Hygiene conducts the birth match process. The agency receives periodic reports from the Department of Human Resources regarding parents who have had a termination of parental rights (TPR) within the past five years as a result of abuse or neglect. The Vital Statistics Division is responsible for matching that list against a list of parents with newborns. Notification of matches is sent weekly to the local department of social services, which is required to assess the child’s safety and offer preventive services to the parents.5

In Michigan, the birth match system automatically notifies the state’s centralized intake via daily emails when a new child is born to a parent who had a previous TPR, caused the death of a child due to abuse and/or neglect, or has been manually added to the match list for another reason. A perpetrator’s name must be manually added to the birth match list in serious child abuse/neglect cases when TPR will not be requested or ordered, including when a nonparent adult is the perpetrator. Once the match’s accuracy is verified, the family must be assigned for investigation unless a pending investigation or open case already exists. The information included in the birth match, including related history, must be used to evaluate child safety in the pending investigation or open case.6

In Minnesota, the birth match program seeks to identify infants born to parents who have had a previous involuntary TPR, involuntary transfer of custody, or “determination of egregious harm.” Past voluntary TPR or transfer of custody may be considered for reports but are not included in the birth match process. Birth match findings result in a mandatory report of substantial child endangerment to the local child welfare agency. The Department of Human Services makes the matches based on birth records received from the Department of Health. Birth match cases must be screened in and receive an investigation unless the family already is involved in an open case specific to the newborn child.7

What is the level of evidence for birth match?

Evidence of effectiveness for birth match suffers many of the same challenges that plague other preventive strategies. It is difficult to prove something didn’t happen as a result of an intervention, and this is even more challenging when the event to be prevented is relatively rare. Research is further complicated because the approach currently is employed in so few jurisdictions. Finally, birth match assumes a link between a prior report of child abuse or neglect and risk of future harm, which current research has not yet confirmed.

According to one study that examined the practice in three jurisdictions, “The empirical support for the birth match approach rests, primarily, in the findings that newborns are extremely vulnerable and families that are involved with (child welfare services) once are often involved again.”8 Along these lines, birth match may make sense because we know that infants are a particularly vulnerable population. Nearly half of all child abuse and neglect fatalities (49.4 percent in 2015) occur before the child’s first birthday, and infants under age 1 die from abuse and neglect at three times the rate of older children.9 This is due to their size and physiology, dependence on caregivers, and the fact that infants may have little or no exposure to community members who could report early warning signs, particularly in isolated families (where fatality and severe injuries are more likely to occur).

No data exist to show that infants whose parents had a TPR with a previous child are more likely to experience a child maltreatment fatality or near fatality. While there is evidence that past reports to CPS on a specific child serve as an effective predictor of that child’s future maltreatment, this research demonstrates only that prior reports for the same child increase that same child’s fatality risk. Specifically, a study in California found that, after adjusting for other risk factors, a prior report to CPS was the single strongest predictor of a child’s potential risk for injury death (intentional or unintentional) before age 5. In particular, children with a prior CPS report had 5.8 times greater risk of death from intentional injuries.10 Research showing that prior reports for a sibling increase a newborn infant’s risk of serious harm would provide stronger support for current birth match policies.

Some data exists from states that have implemented birth match. For example, Michigan reports that 49 children were protected from “likely harm” in 2013, 74 children in 2012, and 82 in 2011.11 In Maryland, 30 percent of the matches (14 of 47) made between October 2010 and September 2011 that were previously unknown to CPS resulted in a case being opened for the child to address identified service needs.12 It is important to note, however, that the act of opening a case for an identified service need does not necessarily correlate to a finding of child abuse or neglect or threat of harm.


What are key considerations in operationalizing birth match?

Birth match in its broadest sense refers to the process of matching data from multiple sources to identify newborns who may be at elevated risk. There are multiple implementation and adaptation considerations, given that these types of programs have been conceptualized and operationalized differently from jurisdiction to jurisdiction.  Some key decision-making questions include:

  • Which children should be identified? In most states where birth match is practiced, the process identifies newborns whose mothers or fathers have had a prior TPR. (This is modified in some cases, for example, to include parents who lost a child to guardianship and/or exclude parents who voluntarily relinquished their rights.) Parents also may be flagged if they have been identified as “perpetrators” but are absent a TPR. These processes are limited to the identification of infants who are born in a hospital (or otherwise reported to a department of vital statistics) within the state. In New York City, on the other hand, birth match identifies only newborns with siblings who are in foster care at the time of their birth.
  • How far back should the birth match process look? In Maryland, records for TPR are reviewed only for the past five years. In Michigan, they extend back as far as computer records exist.
  • How will families be identified? Considerations in this area include which agency is responsible for doing the matching (vital records or child welfare) and whether the process is automated (as in Michigan, where emails are sent weekly to local offices) or relies on a caseworker. In New York City, newborns are identified only through family visits to mothers with open foster care cases.
  • What should happen when a family is identified? The actions required of a child welfare agency once a match is identified vary considerably. They may range from an offer of voluntary services to a mandatory investigation. In New York City, removal of the newborn is required unless a compelling reason exists to delay it.

What are the tensions or issues inherent in birth match?

Ethical tensions related to birth match programs center on the accuracy of predictions of future harm and the government’s duty, role, or rationale for intervening as a result. Proponents of birth match argue that few, if any legal or technological barriers prevent agencies from implementing this kind of strategy. Birth match could be implemented right now in every state. Proponents also argue that the federal Adoption and Safe Families Act establishes a precedent for responding to families differentially on the basis of previous TPRs.

Nevertheless, the practice does raise questions worthy of discussion and debate:

  • Are we taking a leap by linking any prior family history with CPS (not child-specific) to risk of future harm to a newborn? As noted earlier, the field currently lacks empirical evidence that prior family history places newborns at greater risk of serious harm. More research is needed in this area.
  • Does the practice violate family privacy? Although birth data are vital records collected under government authority, as opposed to private medical data, some family advocates express concern about this level of data sharing.
  • Do previous actions, such as issuance of a TPR, grant government the right to monitor an individual’s behavior for a lifetime, or are parents entitled to a “fresh start? Some programs address this question by limiting look-back timeframes, excluding parents who voluntarily relinquished a previous child for adoption, or allowing caseworkers to consider changed circumstances in the assessment of identified families.
  • Does birth match result in too many false positives? This can raise concerns about unnecessary intrusion into family lives, as well as concerns about the cost to the agency of pursuing assessments or investigations where no harm has been reported.
  • Does birth match identify enough families? Reliance on birth match at the expense of other strategies may fail to identify children who are born at home with no recorded birth certificate, out of state, or in homes with non-parental caregivers who may put the child at risk.
  • Is the act of offering preventive services an intrusion? Even when services offered are voluntary, families may view contact initiated by CPS as intrusive, possibly even traumatic. Without conclusive evidence to show that newborns identified through birth match are at risk of serious harm, it is reasonable to question the justification of supportive or preventive outreach to families.
  • Will the process perpetuate biases in data? Some may wonder whether birth match programs will exacerbate disproportionate treatment of minority families, for example.
  • Does birth match unnecessarily emphasize family risk factors at the expense of addressing systemic factors? Although a past CPS report appears to be a strong indicator of risk, more research is needed before concluding that birth match strategies are more effective at preventing future harm than strategies that focus on addressing issues such as poverty, mental health, substance use, housing, and domestic violence.
  • Would the resources of child protection agencies be better spent on more universal prevention strategies? Some may argue that child welfare agencies will see better results from an approach that is more universally supportive of families within their broader communities, rather than something that may be perceived as unfairly punitive to a small subset of families at risk.

1 Barth, R.P., & Johnson-Reid, M. (2017). Better use data to protect children and support families. Grand challenges for social work: A policy action to stop family violence. Retrieved from http://aaswsw.org/wp-content/uploads/2017/03/PAS.3.1.pdf
Matthews, S. (2017). Protecting newborns from harm with data-sharing. Retrieved from https://datasmart.hks.harvard.edu/news/article/protecting-newborns-from-harm-with-data-sharing-1154 2 “Recommendation 7.2g, p. 101: Pass state legislation to establish policies for matching birth data to data on termination of parental rights and conducting preventive visits. These can be modeled after Michigan, Maryland, or New York City.” Commission to Eliminate Child Abuse and Neglect Fatalities. (2016). Within our reach: A national strategy to eliminate child abuse and neglect fatalities. Washington, DC: Government Printing Office. Retrieved from https://www.acf.hhs.gov/cb/report/within-our-reach-national-strategy-eliminate-child-abuse-and-neglect-fatalities
3 Barth, R. (2014). Testimony before the Commission to Eliminate Child Abuse and Neglect Fatalities. Retrieved from page 16 of https://cybercemetery.unt.edu/archive/cecanf/20160323230322/https:/eliminatechildabusefatalities.sites.usa.gov/files/2014/05/Transcript-Tampa-FINAL.pdf
4 Shaw, T. V., Barth, R. P., Mattingly, J., Ayer, D., & Berry, S. (2013). Child welfare birth match: Timely use of child welfare administrative data to protect newborns. Journal of Public Child Welfare, 7, 217–234.
5 Maryland Department of Human Resources, Social Services Administration. (2012). Birth match reports. Retrieved from http://dhs.maryland.gov/documents/SSA%20Policy%20Directives/Child%20Welfare/SSA%2019-05%20CW%20Birth-Match-Reports.pdf
6 Michigan Department of Health and Human Services. (2017). Children’s protective services manual – CPS intake: Special cases. Retrieved from page 16 of https://dhhs.michigan.gov/OLMWEB/EX/PS/Public/PSM/712-6.pdf
7 Minnesota Department of Human Services. (2015). Minnesota child maltreatment intake, screening and response path guidelines. https://www.lcc.leg.mn/tfcp/meetings/10132015/DHS_New_Screening_Guidelines.pdf, p. 31.
8 Shaw, T. V., Barth, R. P., Mattingly, J., Ayer, D., & Berry, S. (2013). Child welfare birth match: Timely use of child welfare administrative data to protect newborns. Journal of Public Child Welfare, 7, 217–234.
9 U.S. Department of Health & Human Services, Administration for Children and Families, Administration on Children, Youth and Families, Children’s Bureau. (2017). Child maltreatment 2015. Available from https://www.acf.hhs.gov/cb/report/child-maltreatment-2015
10 Putnam-Hornstein, E. (2010). An examination of injury mortality among children reported for maltreatment: Findings from California. Presentation at the 50th Annual National Association for Welfare Research and Statistics, September 26–29, 2010.
11 Bladen, S. (2014). Testimony before the Commission to Eliminate Child Abuse and Neglect Fatalities. Retrieved from page 100 of https://web.archive.org/web/20160603180724/https://eliminatechildabusefatalities.sites.usa.gov/files/2014/05/Michigan-Transcript-FINAL1.pdf
12 Shaw, T. V., Barth, R. P., Mattingly, J., Ayer, D., & Berry, S. (2013). Child welfare birth match: Timely use of child welfare administrative data to protect newborns. Journal of Public Child Welfare, 7, 217–234.
13 Barth, Mattingly & Parks. (2017). Child welfare reform: Birth match as an approach to early intervention. Child Advocacy Program – Art of Social Change: Child Welfare, Education, & Juvenile Justice: Reading packet for Session #4, February 16, 2017. Retrieved from https://cap.law.harvard.edu/art-of-social-change-selected-video-recordings-spring-2017/