A self-described libertarian friend once described to me the feeling she had when it was time to leave the hospital with her newborn baby. She remembered looking at the nurse and thinking, “You’re just going to let me take this thing home? I have no idea what I’m doing.” Even those of us who are very skeptical of government intervention know instinctively that a lot can go wrong with an infant. They might not be eating enough. They might catch a virus. They might be injured by a well-meaning toddler.

The constant attention required of new parents is hard enough when you’re sober. Now imagine trying to do it when you’re high—or suffering withdrawal. Surely, if any parent needs a nurse or doctor to check up on them before taking a baby home, it is parents using drugs. But a new policy enacted at Mass General Brigham in Boston last week will discourage medical professionals from reporting mothers who test positive for illegal substances to the state’s child welfare agency.

Why would the hospital system adopt such a policy? You guessed it: to avoid perpetuating “systemic racism.” Representatives of the Mass General Brigham administration’s “United Against Racism” initiative found that “Black pregnant people are more likely to be drug tested and to be reported to child welfare systems than white pregnant people.” As a result, the hospital will “update policies that automatically trigger mandatory filings with child welfare agencies when a pregnant individual is engaged in treatment for substance use disorder, absent any other concerns for potential abuse or neglect.”

Let’s start from the top. Racial disparities are not prima facie evidence of racism. Black children are three times as likely to die from maltreatment as white children, so it is not unreasonable to assume that they are at higher risk. It’s also not unreasonable to assume that black mothers would get reported to child protective services more often than white mothers.

It is sensible for a doctor not to report a mother who discloses that she is in a treatment program, discloses the fact that she is using methadone to control her addiction, and displays no other problematic behaviors. And there’s no evidence that Mass General or any other hospital is reporting such women.

But here’s the kicker. The new policy says that reports “after delivery should be filed only if there is reasonable cause to believe that the infant is suffering or at imminent risk of suffering physical or emotional injury.” Evidence of injury or of imminent risk of injury is a very high bar; setting it so high will invite considerable trouble. What does “imminent risk” mean in this context? Does it mean that you can only report a mother who is high when she leaves the hospital and forgets to put the baby in a car seat, and who is so out of sorts that she won’t remember how or when to feed the baby?

Why are hospitals so afraid of getting child-welfare authorities involved? Even when a mother is found to be using drugs, or when the baby tests positive for in utero exposure, that doesn’t necessarily mean that authorities will remove the child from the home. It means that the agency will conduct an investigation to determine if the child is safe and if the parents need help with rehabilitation. Do black children and families deserve this intervention less than white families?

The alternative to an investigation is the child being sent home from the hospital without any follow-up. Such scenarios are already playing out in some states instituting voluntary Plans of Safe Care instead of reporting signs of drug abuse to child welfare. These plans consist of forms given to families that make note of services the hospital has offered. As a report from New Mexico noted, “data show that a high percentage of families are declining services when referred.” Indeed, a survey found that, of those receiving a plan, “41.8% of families did not know what a Plan of Safe Care was or had no one talk to them about it in the hospital. In addition, 57.1% of families completing the survey either were not contacted by a care coordinator or refused services.”

In recent years, an average of one infant who has tested positive for drugs at birth has died every month in New Mexico. And most of those who died did so because of later substance exposure (often when a toddler got a hold of fentanyl) or unsafe sleeping arrangements while a parent was intoxicated (typically, a parent would roll over onto a child and suffocate him or her). The assumption that infants born substance-exposed are fine to go home with their parents without follow-up is deeply misguided and ignores the growing evidence that doing so poses enormous risks.

Before Mass General’s recent announcement, a working group that I have convened at the American Enterprise Institute was preparing to release a statement about the need to take prenatal exposure more seriously. As the United States continues to move toward a less-punitive approach to substance use by adults, we cannot simply assume that kids will be fine. As the working group, which included former agency heads, judges, researchers, and nonprofit leaders across the political spectrum, noted:

Individuals with substance use disorders may take many years to be “ready” for treatment, but there is no pause button on infants’ development. . . . As states and the federal government myopically prioritize the avoidance of CPS as an end unto itself, [plans of safe care] have been applauded without any evidence—or even requirements to collect evidence—as to their actual efficacy in keeping children safe and ensuring that their basic developmental needs are met. This is not compassion but instead an abdication of responsibility.