did a jama psychiatry study analyze insurance claims from over 4 million pregnancies?
“The researchers analyzed insurance claims from over 4 million pregnancies.”
a host said on Psychopharmacology and Psychiatry Updates, 2026-10-18, at 01:03. hear it in context → the speaker cited: a large study published in JAMA Psychiatry.
the episode: Are Z-Drugs Safe in the First Trimester of Pregnancy? (at 01:03). we link the show's own page; we do not re-host audio.
first, the part that holds
the number holds up. the 2024 JAMA Psychiatry cohort study of prescription stimulants in pregnancy drew on Medicaid and MarketScan claims data, and it covered about 4.3 million pregnancies, comfortably over 4 million.
what we found when we went and looked
we read the study in JAMA Psychiatry (2024;81(5):477-488) on the full journal page and on PubMed Central. its analysis groups add up to 4,278,460 pregnancies: 2,502,250 in Medicaid (2000-2018) and 1,776,210 in the MarketScan Commercial Claims Database (2003-2020). the full cohorts in the PMC text add up to 4,317,502 (2,522,490 plus 1,795,012). the paper calls its source "health care utilization data". both of its databases are insurance claims databases, so "insurance claims" is a fair plain-language description. one thing a listener might miss: almost all of those 4 million pregnancies were the unexposed comparison group. only about 8,200 had a stimulant dispensed in the second half of pregnancy (7,065 amphetamine/dextroamphetamine and 1,123 methylphenidate).
assertion by assertion
| assertion | verdict | what the record says |
|---|---|---|
| the study was published in JAMA Psychiatry | confirmed | the journal page gives the citation as JAMA Psychiatry 2024;81(5):477-488, doi:10.1001/jamapsychiatry.2023.5073. the clip doesn't name the study, so we matched it on journal, databases and size. we did not find another JAMA Psychiatry pregnancy study that fits all three. [1] |
| researchers analyzed over 4 million pregnancies | confirmed | the analysis groups total 4,278,460 pregnancies: Medicaid 2,496,771 unexposed, 4,693 amphetamine/dextroamphetamine and 786 methylphenidate; MarketScan 1,773,501 unexposed, 2,372 amphetamine/dextroamphetamine and 337 methylphenidate. the full cohorts in the PMC text total 4,317,502. these are only live-birth pregnancies linked to a child, not all pregnancies. [1] |
| the data came from two databases, Medicaid and MarketScan | confirmed | the two sources were Medicaid data from 2000 to 2018 (public insurance) and the MarketScan Commercial Claims Database from 2003 to 2020 (commercial insurance), covering people aged 12 to 55. [1] |
| the data were insurance claims | confirmed | the paper's own term is "health care utilization data". one database is named the Commercial Claims Database and the other is Medicaid billing data, so "insurance claims" is accurate as a plain description. the limitation is that claims record a prescription being dispensed, not whether the medication was taken. [1] |
| context: how many of those pregnancies were exposed to a stimulant | confirmed | 7,065 pregnancies had amphetamine/dextroamphetamine and 1,123 had methylphenidate dispensed in the second half of pregnancy, about 0.2% of the total. the 4 million figure is mostly the comparison group. after adjustment the study found no meaningful link to autism, ADHD or any neurodevelopmental disorder. methylphenidate had a raised hazard ratio for ADHD (1.43) that did not hold up in sensitivity analyses. [1] |
both directions
for the speaker: every clause of the sentence matches the published paper: the journal, the two databases (Medicaid and MarketScan), insurance-claims data, and a pregnancy count over 4 million (4.28 million in the analysis groups, 4.32 million in the full cohorts). "a large study" is a fair description. by pregnancy count it is one of the largest US studies of stimulant exposure in pregnancy.
against: nothing in the record contradicts the number. the one framing risk is that "over 4 million pregnancies" can sound like 4 million people took the medication. the study's own key points describe 7,065 pregnancies exposed to amphetamine/dextroamphetamine and 1,123 exposed to methylphenidate. the medication findings rest on those exposed groups, especially the smaller methylphenidate group, not on the full 4 million. the paper also says "health care utilization data" rather than "insurance claims", but that is a difference of wording, not of fact.
what this cannot say
the clip doesn't name the study. we identified it from the journal, the two databases and the size, and that match is our inference. only live-birth pregnancies linked to a child with continuous insurance coverage were included. miscarriages, stillbirths and uninsured pregnancies are not in the 4 million. claims show that a prescription was dispensed, not that it was taken, and diagnoses in claims depend on how they were coded. we checked the count and the attribution, not whether the study's conclusions are right.
sources, each one fetched and read
- Prescription stimulant use during pregnancy and risk of neurodevelopmental disorders in children. JAMA Psychiatry 2024, full text on PubMed Central, PMC10809143 What we read there: the Medicaid cohort comprised 2,522,490 pregnancies and the MarketScan cohort 1,795,012. pregnant individuals aged 12 to 55 "with live-birth deliveries linked to children were required to have insurance coverage from 3 months before the estimated last menstrual period." 7,065 pregnancies were exposed to amphetamine/dextroamphetamine and 1,123 to methylphenidate in the second half of pregnancy. Fetched 2026-10-19 (HTTP 200). https://pmc.ncbi.nlm.nih.gov/articles/PMC10809143/
speakers are described by role, not named. one sentence is quoted under fair use, with attribution and a link to the original at the timestamp; we do not re-host audio or video. nothing on this page is a reason to start, stop or change a medication. if you have the file behind anything marked unverifiable, the research board is where people can answer. adam