For years, Dr. Craig Towers's research convinced doctors at clinics in Tennessee and Daytona Beach, Florida, that pregnant women addicted to opioids could safely stop taking medication using a drug called naltrexone. His study reshaped how those clinics treat some of the most vulnerable patients of the opioid crisis: pregnant women trying to protect their babies.

But a University of Tennessee audit found the study never got required ethics board approval. The audit also found Towers supplied "inaccurate and/or unverifiable" source documentation, according to the university's audit, as reported by ProPublica, and it could not confirm that patients' private health information had been protected.

Addiction and pregnancy specialists who reviewed the study raised even more troubling questions. They noted it never mentions a single patient declining treatment or leaving before delivery, uncommon in research involving drug use, and that it reported almost no serious medical complications among the mothers. "That made my hair on the back of my neck stand up," said Hendrée Jones, a University of North Carolina researcher who signed a letter of concern to the university, according to ProPublica.

One former patient, who hadn't used drugs during her two previous pregnancies and had completed treatment two years earlier, told ProPublica she felt pressured into taking naltrexone.

The University of Tennessee asked the journal to retract the study. Instead, more than three years later, it published a 78-word correction and left the study standing, even as some clinics still use the detox approach it inspired.

Pregnant women fighting addiction deserve care backed by real evidence, not an unapproved study built on data no one can verify. Sign the petition telling the American Journal of Obstetrics and Gynecology to fully retract this study, before more clinics rely on results that were never proven safe.