This article was produced in collaboration with The Marshall Project, a nonprofit news organization that covers the U.S. criminal justice system, Reveal and Mother Jones.

Amairani Salinas was 32 weeks pregnant with her fourth child in 2023 when medical staff in a Texas hospital found that her baby no longer had a heartbeat. While preparing her for an emergency cesarean section, the staff gave her midazolam, a benzodiazepine often used to keep patients calm. The next day, as she held her stillborn daughter, a social worker delivered terrible news: Salinas was being reported to child welfare because a drug test showed traces of the benzodiazepine — the same medication given to her by the hospital prior to surgery.

Similarly, Victoria Villanueva was pregnant with her first child when her baby tested positive for morphine. Villanueva arrived at an Indiana hospital at 41 weeks for labor induction. To alleviate her contractions, she was given narcotics. The following day, a social worker informed her that her newborn’s meconium — the first bowel movement — contained opiates. Instead of cherishing her time with her baby, Villanueva was consumed with anxiety over the possibility of having her child taken away. “I didn’t even know how to function,” she recounted.

Salinas’ and Villanueva’s cases are not unique. Nationwide, hospitals are giving medications to women in labor, then reporting them to child welfare when tests for drugs come back positive for those same substances, according to an investigation by The Marshall Project and Reveal.

These positive test results stem from medications routinely given to millions of women during childbirth in the United States each year. The list includes morphine or fentanyl for epidural anesthesia, anxiety-reducing drugs, and medications to control blood pressure during cesarean deliveries.

Amid heightened scrutiny and criminalization of expectant mothers since the reversal of Roe v. Wade, these hospital reports have led to police involvement, child welfare assessments, and even the removal of children from their families.

This investigation involved interviews with over two dozen patients and healthcare professionals, along with the examination of numerous medical and legal documents. Some individuals requested anonymity due to the sensitive nature of child custody cases.

In New York, a mother with no previous history of substance abuse had her toddler and newborn taken from her for five months after testing positive for fentanyl that was administered during her epidural. In Oklahoma, authorities removed her newborn and three other children from a mother after she tested positive for meth. They spent 11 days in foster care until a verification test confirmed that the positive result was due to a heartburn medication given at the hospital.

By the time Villanueva gave birth in 2017, medical professionals had long been aware that certain medications can quickly pass from mother to child, leading to positive drug test results. All three of Villanueva’s prenatal tests showed no drugs in her system, and the morphine she received for pain was documented in her medical files. Yet, the hospital still reported her to the state child welfare agency, as indicated in hospital records.

Marion General Hospital in Indiana did not provide comments when contacted. Brian Heinemann, a spokesperson for the Indiana Department of Child Services, refrained from discussing Villanueva’s situation but mentioned that guidelines have since been updated to ensure that positive drug screens are not the sole basis for abuse or neglect allegations.

Kimberly Walton, a representative for the Texas Health Resources system, did not comment on Salinas’ case but stated that drug testing is conducted when there’s a concern for patient safety, adding that health professionals are mandated to report suspected drug use that could jeopardize a child’s safety.

The practice of drug testing pregnant women in hospitals started in the 1980s and gained momentum during the opioid crisis. This was, in part, to identify infants who might experience withdrawal and require additional medical attention. Federal regulations mandate that hospitals notify child welfare services when such babies are born. However, an investigation by The Marshall Project and Reveal uncovered that the commonly used, low-cost urine tests are prone to false positives, errors, and misunderstandings. Many hospitals haven’t implemented necessary protections to prevent patients from being reported based on inaccurate results.

In some instances highlighted by The Marshall Project, neither doctors nor social workers considered a patient’s medication history as a possible reason for a positive test. In other cases, although a healthcare provider suspected a prescribed medication might be responsible, they still reported the patients to authorities.

“What level of harm are we willing to accept for birthing individuals? Our patients are suffering until we enhance our practices,” stated Dr. Davida Schiff. Her hospital system, Mass General Brigham in New England, implemented a new policy this year that instructs providers to only conduct drug tests on patients, with their consent, when medically necessary and has ceased the automatic reporting of positive results. She emphasized that hospitals have a duty to rectify these testing issues, saying, “The faults lie with the hospitals. The clinicians are to blame. Our policies need improvement.”

The dangers of administering medications to pregnant patients and subsequently testing for illegal substances are well recognized. A 2022 study from Massachusetts General Hospital revealed that 91% of women who received fentanyl through epidurals tested positive for the substance afterward. Other research has demonstrated that mothers can efficiently transfer these medications to their infants. According to Athena Petrides, the lead researcher on a study at Brigham and Women’s Hospital in Boston, “A positive drug test for a baby cannot and should not be used to accuse mothers of fentanyl abuse.”

Even with these cautions, many hospitals lack the regulations requiring a review of patient medication histories before reporting to authorities. Reporting laws shield doctors from liability for good-faith reports, regardless of their accuracy. Additionally, toxicologists and doctors note that many healthcare providers do not have enough time or specialized knowledge to accurately analyze drug test results.

“This isn’t a topic typically covered in medical school or during residency,” noted Dr. Tricia Wright, an OB-GYN professor at the University of California, San Francisco Medical Center, who focuses on substance use disorders during pregnancy. “It hinges on the individual’s interpretation.” Wright played a role in updating her hospital’s policy to state that doctors should avoid testing patients for drugs unless it is medically warranted.

While drug testing can assist pediatricians in deciding treatment for infants at risk of withdrawal, many OB-GYNs believe that positive drug tests generally do not impact the mother’s care and are therefore not a primary concern for them.

Often, it is the social workers at hospitals—who are tasked with communicating with child welfare agencies—who are more attentive to drug test outcomes. Some hospitals necessitate that social workers automatically report any positive test results, while others conduct evaluations to assess whether there is any risk to the newborn.

However, social workers frequently experience high workloads, as noted by Kylie Haines, who leads a program for pregnant women facing opioid use disorder at Vanderbilt University Medical Center, a premier teaching hospital for obstetrics. She observed that social workers typically receive even less training than doctors regarding drug testing and have limited power to question test outcomes, stating, “Investigating the reasons behind a positive test isn’t part of their responsibilities.”

“We aren’t medical practitioners,” stated Haines, a licensed social worker. “We don’t have the ability to interpret drug screenings.”

The timing of when tests are administered complicates interpretations further. Ideally, obtaining a urine sample for testing should happen when the patient arrives at the hospital, prior to any medication administration. However, it is common for hospitals to test urine from a birthing patient’s catheter bag or to request urine samples after medications have already been given, as noted by healthcare providers.

Additionally, drug testing newborns adds a layer of intricacy. While urine tests can reveal drug use over several days, testing the baby’s meconium can detect substances used months prior. Meconium tests are widely regarded as the most reliable method for newborn drug testing since they can indicate potential drug exposure that occurred earlier in pregnancy, signifying possible addiction.

However, it can be impossible to distinguish whether a newborn’s positive test result is due to medication administered at the hospital or a substance the mother might have used earlier in her pregnancy.

Additionally, drug tests do not provide information about the extent or frequency of substance use by an individual. For instance, a patient might have used drugs in the past but ceased when she learned she was pregnant. “You actually don’t know,” Schiff explained. “It’s quite chaotic.”

Drugs like morphine and fentanyl have caused some patients to be marked as opioid users. Medications like ephedrine and phenylephrine, often given to manage low blood pressure during C-sections, have incorrectly resulted in positive tests for methamphetamine. The Marshall Project also discovered cases where women faced inquiries about positive drug tests after being administered sedatives like benzodiazepines or barbiturates in the hospital.

When a woman receives a positive drug test result, other facets of her life may quickly come under scrutiny. Medical issues or childbirth complications linked to addiction—such as hypertension, placental abruption, or premature delivery—might be mistakenly interpreted as further proof of drug use. Even missing some prenatal appointments can be viewed through the lens of a positive drug test as potential evidence of a mother putting her baby’s health at risk. Women who had traumatic birthing experiences may suddenly feel threatened.

In 2023, after losing her baby due to stillbirth, Salinas found herself in deep sorrow when she was informed that the child welfare agencies would be notified about her situation. She was unaware that the hospital had administered benzodiazepine, which led to her positive test, and she denied taking any medications. She also tested positive for Delta-9, a legal product derived from hemp, which she claimed to have purchased at the grocery store. But soon, Salinas faced a prolonged investigation while battling debilitating depression, grieving, and caring for her other children.

“I still have three healthy children. They need to eat, go to school, and have their mother present,” Salinas remembered thinking during the questioning.

The case against her was closed about four months later as “unsubstantiated.” However, it wasn’t until a year later that Salinas reviewed her medical records with a reporter from the Marshall Project and found the source of her positive test.

Marissa Gonzales, a representative for the Texas Department of Family and Protective Services, declined to comment on Salinas’ situation due to privacy regulations.

Salinas expressed her frustration, questioning why child welfare investigators were choosing to zero in on her, a devoted mother who had done nothing wrong, rather than focusing on actual cases of child endangerment. “Why is your attention on someone like me instead of a child who might genuinely be in peril?”

Recent studies indicate that drug tests are not the most reliable method for identifying individuals with substance use disorders—a straightforward screening questionnaire can also effectively identify these issues. Prominent medical organizations recommend that hospitals utilize these questionnaires rather than conduct universal drug tests. Nevertheless, the practice of drug testing and reporting among labor and delivery patients remains widespread.

In at least 27 states, hospitals are mandated to inform child welfare services of positive drug tests or potential substance exposure to infants. However, no state mandates that hospitals verify test results prior to reporting them. Hospitals often notify authorities without ordering follow-up tests or waiting for confirmation of results.

While not all states demand the reporting of a positive drug test, many hospitals do so regardless. Federal data shows that in 2022, over 35,000 infants were reported to child welfare agencies as being exposed to substances, without assurances that the underlying test results were accurate.

“The system is set up to generate reports,” stated Dr. Mishka Terplan, an OB-GYN in Maryland and a prominent researcher in the area of substance use disorders during pregnancy. “To slow it down or halt it requires effort.”

The fallout from an erroneous drug test can be particularly serious for individuals with addiction histories, as they are often less trusted when claiming that a hospital medication caused the positive test. In this scenario, the patient’s testimony is frequently weighed against a single positive drug test, with individual medical personnel deciding whether to investigate further.

Lisa Grisham, a nurse in Arizona, shared the story of a patient recovering from opioid use disorder who tested positive for fentanyl. The woman maintained that another nurse had administered the medication during her labor, even though it was not documented in her medical history. Grisham, who oversees a hospital program for infants exposed to substances at Banner University Medical Center in Tucson, took it upon herself to look into the case and eventually identified the nurse who verified the patient’s claim.

Grisham expressed her frustration, saying, “It makes me sick to think of all the moms that have come through and said they don’t use fentanyl and we don’t believe them.”

Victoria Villanueva, only 18 years old and new to marriage, was working on her GED when she welcomed her first child. Raised to be honest with healthcare providers, she informed a nurse during her hospital admission that she had experimented with drugs like marijuana and acid at the age of 15 but hadn’t used them since.

Upon her admission, Villanueva underwent a drug test, which returned negative for any illegal substances, matching the results from her two prenatal drug screenings. Afterward, she was given medication to induce labor, prompting painful contractions. She felt relief when a nurse informed her that morphine was being administered.

The day after Villanueva’s delivery, an investigator from the Indiana Department of Child Services appeared. The hospital had alerted the agency to Villanueva’s “history of drug use,” as noted in her medical records. Villanueva was unaware that the hospital had also sent her daughter’s meconium for testing, which revealed traces of morphine.

Villanueva believed that her hospital records would confirm her usage of morphine, and she was correct. The drug test results clarified, “Drugs administered during labor and delivery may be detected in meconium.”

Despite the agency receiving the hospital records, Villanueva noted they insisted she undergo further drug tests and allow home inspections. After several weeks, the agency ultimately closed the case.

It wasn’t until the investigators left her life that Villanueva felt free to celebrate the birth of her child and embrace her happiness.

“I couldn’t even really enjoy properly my child being born,” she recalled. “Until after the fact, when they were gone.”

This article was produced in collaboration with The Marshall Project, a nonprofit organization dedicated to covering the U.S. criminal justice system, and Reveal and Mother Jones. Subscribe to The Marshall Project’s newsletters, and connect with them on Instagram, TikTok, Reddit, and Facebook.