About

Almut G Winterstein is a distinguished professor in the Department of Pharmaceutical Outcomes and Policy at the University of Florida College of Pharmacy, where she also serves as the founding Director of the Center for Drug Evaluation and Safety. She holds a pharmacy degree from Friedrich Wilhelm University in Bonn, Germany, and a PhD in Pharmacoepidemiology from Humboldt University in Berlin, Germany. Since joining the University of Florida in 2000, she has established herself as a leading expert in drug safety, medication use, and pharmacoepidemiology, with a research focus on the post-marketing evaluation of drugs in pediatrics, perinatal care, infectious disease, and psychiatry, utilizing real-world data to improve public health outcomes. She has served as chair of the FDA’s Drug Safety and Risk Management Advisory Committee and is a fellow and current president of the International Society of Pharmacoepidemiology. In 2017, she was named the Dr. Robert and Barbara Crisafi Chair for Medication Safety in recognition of her contributions to drug safety research. Winterstein is also the director of the Consortium for Medical Marijuana Clinical Outcomes Research, a multi-university group funded by the State of Florida to study the risks and benefits of medical cannabis. Her scholarly work includes over 400 publications and more than 25 extramurally funded grants, emphasizing the development of evidence-based approaches to medication safety, policy evaluation, and pharmaceutical data analytics to support clinical decision-making and improve medication use in real-world populations.

Research topics

• Medicine
• Internal medicine
• Cardiology
• Machine Learning
• Artificial Intelligence
• Pediatrics
• Psychiatry
• Pharmacology
• Demography
• Computer Science
• Computer Security
• Obstetrics
• Family medicine
• Pathology
• Intensive care medicine
• Data science
• Engineering
• Environmental health
• Chemistry
• Biology
• Bioinformatics
• Radiology

Selected publications

• Joint Position Statement from the Society for Birth Defects Research and Prevention, the Organization of Teratology Information Specialists, and the Developmental Neurotoxicology Society: A Call for Implementation of Risk Evaluation and Mitigation Strategies to Reduce Prenatal Exposure to Valproate

  Birth Defects Research · 2026-04-01

  article1st authorCorresponding

  Prevention of prenatal exposure to teratogenic medications is a critical challenge in clinical and regulatory decision-making. Complex benefit–risk considerations need to balance health outcomes among women of childbearing age who need treatment access with potential adverse effects on pregnancy outcomes. In recent decades, regulatory agencies worldwide have implemented risk management programs to support safe use behaviors of patients and providers that can prevent adverse outcomes and help ensure that medications can be approved and used with a favorable benefit–risk profile. Risk management programs are used for medications with known teratogenic risk (e.g., isotretinoin, thalidomide, and mycophenolate) (Brown et al. 2023), with different types of programs in place in different countries. Valproate was initially approved in the United States for treatment of epilepsy, with later expansion of its FDA-approved indications to include acute mania in bipolar disorder and migraine prophylaxis. Several off-label uses have emerged to include treatment of alcohol withdrawal, panic disorders, and social phobias (Manoguerra et al. 2008; Romoli et al. 2019). Valproate has long been recognized as a teratogen, yet substantial numbers of prenatal exposures to valproate still occur. One role of our societies (the Society for Birth Defects Research and Prevention, the Organization of Teratology Information Specialists, and the Developmental Neurotoxicology Society) is to translate scientific evidence to inform clinical practice and public health policies to improve health outcomes. Here we review the scientific evidence for valproate as a teratogen, the efforts currently in place to reduce its utilization in pregnancy, and current levels of prenatal exposures in the United States. Based on review of this evidence, we end with a call for enhanced regulatory risk management measures to mitigate prenatal exposures to valproate in the United States. Valproate has been recognized as a teratogen for decades, with an association between maternal valproate use and an increased risk of spina bifida first recognized in the 1980s (Lammer et al. 1987). Later studies demonstrated an association between valproate and other major congenital malformations. In 2010, a study using European Surveillance of Congenital Anomalies (EUROCAT) data reported links between maternal use of valproate monotherapy and six different congenital malformations. In this study, adjusted odds ratios were 12.7 (95% confidence interval [CI] 7.7–20.7) for spina bifida; 2.5 (95% CI, 1.4–4.4) for atrial septal defect; 5.2 (95% CI, 2.8–9.9) for cleft palate; 4.8 (95% CI, 2.9–8.1) for hypospadias; 2.2 (95% CI, 1.0–4.5) for polydactyly; and 6.8 (95% CI, 1.8–18.8) for craniosynostosis (Jentink et al. 2010). Data from a recent Cochrane review of prospective cohort studies estimated a prevalence of major congenital malformations among children born to women taking valproate of 9.8% (95% CI, 8.1–11.9), compared to 2.1% (95% CI, 1.5–3.0) among children born to women without epilepsy and 3.0% (95% CI, 2.1–4.2) for women with untreated epilepsy. In comparison, the frequencies of congenital malformations among women taking low-risk antiepileptic medications (lamotrigine and levetiracetam) in that analysis were 2.7% (95% CI, 1.9–3.8) and 2.6% (95% CI, 1.6–4.4), respectively, not significantly higher than the baseline risk among women with epilepsy (Bromley, Adab, et al. 2023). In a recent analysis of the International Registry of Antiepileptic Drugs and Pregnancy (EURAP) data collected across 40 countries, major congenital malformations occurred in 9.9% (95% CI, 8.5%–11.5%) of pregnancies among women taking valproate compared to 3.1% (95% CI, 2.5%–3.7%) for lamotrigine and 2.5% (95% CI, 1.8%–3.5%) for levetiracetam (Battino et al. 2024). Similar findings were seen in a study using the North American Antiepileptic Drug Pregnancy Registry between 1997 and 2023 (Hernandez-Diaz et al. 2025). These analyses confirm that the risk of congenital malformations is substantially elevated with prenatal use of valproate, and that these findings are not solely due to maternal epilepsy. Risk of congenital malformations among children born to women taking valproate increases as the dose increases. A study by Tomson et al. demonstrated a frequency of major congenital malformations of 24% among women exposed to monotherapy with valproate at doses ≥ 1500 mg/day (Tomson et al. 2015). Polytherapy with other antiepileptic drugs is associated with a higher risk than with use of valproate alone, with an odds ratio of 1.46 (95% CI, 1.11–1.91) for monotherapy compared to 2.06 (95% CI, 1.32–3.20) for polytherapy (Kim et al. 2026). In addition to the evidence regarding congenital malformations, there is strong evidence that prenatal valproate exposure increases the risk of adverse neurodevelopmental outcomes in childhood (Ornoy et al. 2023). A Cochrane review of cohort studies in 2014 found that children exposed to valproate in utero had IQ scores approximately 9 points lower than unexposed children born to women without epilepsy and 8 points lower than children born to women with untreated epilepsy (Bromley et al. 2014). A 2024 systematic review in the journal Neurology updated these findings, confirming that valproate is consistently associated with 2- to 5-fold increased risk of intellectual disability, with clear dose-dependent effects (Honybun et al. 2024). Prenatal valproate exposure has also been associated with an increased risk of childhood neurodevelopmental conditions. A 2025 meta-analysis of 8 cohort studies found that gestational valproate exposure was associated with a modest but significant increase in attention deficit/hyperactivity disorder (ADHD) risk (adjusted hazard ratio [HR] 1.62, 95% CI, 1.30–2.01) and a three-fold increase in the risk of autism spectrum disorder (ASD) (adjusted HR 3.10, 95% CI, 2.24–4.28), with both risks showing dose-dependent relationships (Andrade et al. 2025). The most comprehensive examination comes from a 2023 Nordic registry study among children born to mothers with epilepsy, which assessed 13 psychiatric disorders. This study found that children with prenatal valproate exposure faced an absolute risk of 42.1% (95% CI, 38.2–45.8) of being diagnosed with any psychiatric disorder by age 18, compared to 31.3% (95% CI, 28.9–33.6) among children of untreated mothers and 30.8% (95% CI, 29.2–32.3) among children exposed to any antiseizure medication. Besides intellectual and other neurodevelopmental disorders, this increase was driven by elevated risks for ASD, ADHD, and attachment disorder (Dreier et al. 2023). Long-term impacts on developmental outcomes in adulthood among those exposed to valproate in utero have not been well examined. As with the risk for congenital malformations, a consistent and significant dose–response relationship is seen with higher doses of valproate linked to both higher risk for and severity of adverse neurodevelopmental outcomes (Andrade et al. 2025). Gestational timing of prenatal exposure is important. First-trimester exposure is a critical time for manifestation of congenital malformations during organogenesis (Fietz et al. 2024), while the timing for adverse effects on neurodevelopment spans early to late gestation (Sheehy et al. 2025). Hence, risk management measures need to focus on exposure prevention throughout pregnancy. Early pregnancy, often before women even know they are pregnant, poses the greatest challenge for prevention. The benefits of periconceptional folic acid to decrease the risk of neural tube defects are well recognized and have led to recommendations from numerous professional groups for all women who are capable of becoming pregnant to take 400 mcg (0.4 mg) of folic acid daily to reduce the risk of neural tube defects (American College of Obstetricians and Gynecologists 2017; US Preventive Services Task Force et al. 2023). Valproate can lower folic acid levels, which has led some to recommend that women who take valproate during pregnancy, as well as women with epilepsy on other medications, take higher doses (1–5 mg daily) of folic acid before and during pregnancy to reduce valproate-associated risks (Ornoy et al. 2023; Reynolds and Green 2020). However, a recent systematic review did not demonstrate significant mitigation of the risk for valproate-associated anatomic abnormalities (e.g., neural tube defects) among women taking folic acid (Valentino et al. 2024). While some studies of women taking antiseizure medications suggest better cognitive development (Meador et al. 2020) and lower risk of autistic traits (Bjork et al. 2018) with periconceptional folic acid supplementation, other studies did not confirm these protective effects (Baker et al. 2015; Bromley, Bullen, et al. 2023; Valentino et al. 2024). Thus, it remains unclear whether higher doses of folic acid are beneficial. Consistent with this, the current joint practice guideline from the American Academy of Neurology (AAN), the American Epilepsy Society (AES), and the Society for Maternal-Fetal Medicine (SMFM) recommends “at least 0.4 mg of folic acid supplementation daily preconceptionally and during pregnancy” to any person with epilepsy of childbearing potential treated with an antiseizure medication to decrease the risk of neural tube defects and to possibly improve neurodevelopmental outcomes such as ASD and global IQ in the offspring (Practice Guideline from the AAN, AES, and SMFM, et al. 2025). Concerns regarding possible adverse reproductive effects of valproate are not confined to use by women. Several smaller studies suggest that men may experience a drop in fertility while using valproate, with fertility improving when use has stopped (Asghar et al. 2024; Markoula et al. 2020; Tallon et al. 2021). However, a 2025 published retrospective cohort study in more than 90,000 exposed men with bipolar disorder or epilepsy found no significant difference in lifetime risks of infertility, testicular hypofunction, testicular atrophy, or abnormal semen parameters compared to unexposed men (Mbizvo et al. 2025). Paternal use of valproate has not been found to result in an increased risk of major congenital malformations (Garey et al. 2024). However, although evidence is inconsistent, paternal use may increase the risk of neurodevelopmental disorders in offspring. A 2025 European Medicines Agency-mandated post-authorization safety study conducted in Denmark, Norway, and Sweden and published in 2025 found that paternal valproate exposure within 3 months prior to conception was associated with a 50% increased risk of neurodevelopmental disorders compared to paternal use of lamotrigine or levetiracetam (pooled adjusted HR 1.50, 95% CI, 1.09–2.07) (Colas et al. 2025; IQVIA 2023). These concerns prompted recommendations about precautionary measures for the treatment of male patients by European regulatory authorities (European Medicines Agency 2024). However, independent Danish analyses (Christensen et al. 2024) were unable to replicate the Danish risk estimates from the post-authorization safety study, and an earlier Swedish study found insignificant results, suggesting that adverse effects are likely attributable to epilepsy-related factors rather than valproate itself (Bjork et al. 2025; Olstad et al. 2025). The US Food and Drug Administration (FDA)-approved indications for valproate include treatment of seizure disorders, acute treatment of manic or mixed episodes associated with bipolar disorder, and migraine prophylaxis. FDA has approved various formulations, ranging from capsules to injections, made by multiple different manufacturers. However, successive label changes, including addition and then expansion of a boxed warning, have restricted its use in the United States, including a contraindication for prophylaxis of migraine headaches in pregnancy and in women of childbearing potential who are not using effective contraception (Figure 1). The current label also states that valproate should not be used to treat women with epilepsy or bipolar disorder who are pregnant or plan to become pregnant unless other medications have failed to provide adequate symptom control, and it should generally not be administered to a woman of childbearing potential unless other medications have failed and only if effective contraception is used. Besides the label, regulatory risk management to prevent prenatal exposure to valproate has been limited to several drug safety communications to health care providers (Figure 1), and the requirement for a medication guide, that is, standardized patient-directed information that must be dispensed when valproate prescriptions are filled. The 5-page medication guide, which covers a broad range of severe adverse effects, includes a half page summarizing the risk for birth defects and lower IQ, related precautions and contraindications consistent with the label, and a recommendation that in case of pregnancy, women should talk to their healthcare provider to decide whether they should continue treatment with valproate (DailyMed National Library of Medicine 2026). In addition to issuing drug safety communications and medication guides, regulatory agencies can also require the implementation of more stringent risk management programs. With the FDA Amendments Act signed into law in 2007, FDA has the authority to require Risk Evaluation and Mitigation Strategies (REMS). Prominent examples in prevention of teratogenic effects include the REMS for isotretinoin, thalidomide, and mycophenolate (Brown et al. 2023). Each of these REMS includes “elements to ensure safe use,” ranging from more restrictive components such as required pharmacy verification of negative pregnancy tests before dispensing and supplies capped at 30 days for isotretinoin and 28 days for thalidomide, to only mandatory provider training and patient counseling for mycophenolate. While the FDA has not implemented a REMS program for valproate, other countries have implemented restrictions similar to REMS to help mitigate exposure in pregnancy. In 2018, the European Medicines Agency (EMA) developed a valproate Pregnancy Prevention Programme (PPP) in which valproate is contraindicated in women of childbearing age unless the patient and prescriber complete a mandatory risk acknowledgement form, the patient uses effective contraception throughout treatment with negative pregnancy tests before and after treatment, and other restrictions (European Medicines Agency 2018). This PPP was endorsed by the European Commission, requiring implementation by all member countries. The UK's Medicines and Healthcare Products Regulatory Agency (MHRA) (Iacobucci 2018) and Health Canada have adopted similar PPPs (BGP Pharma ULC 2020). For years, US-based professional medical organizations have acknowledged the risks associated with valproate use in pregnancy and suggested mitigation strategies. As early as 2009, the AAN and the AES issued practice updates warning about the use of valproate in pregnant women and suggesting changing to another antiepileptic medication before pregnancy, if possible (Harden, Meador, Pennell, Allen Hauser, et al. 2009; Harden, Meador, Pennell, Hauser, et al. 2009). The current joint practice guideline from AAN, AES, and the SMFM regarding women with epilepsy of childbearing potential recommends avoiding the use of valproate if clinically feasible (Practice Guideline from the AAN, AES, and SMFM, et al. 2025). In 2023 the World Health Organization published an addendum to their Mental Health Gap Action Program (mhGAP) guidelines suggesting that valproate not be used in women and girls of reproductive potential, that those taking valproate be advised to use effective contraception and have their medication reviewed by a specialist on a periodic basis, and that every effort be made to switch to an alternative treatment prior to pregnancy (World Health Organization 2023). Finally, in 2025, several members of the American Psychiatric Association's Committee on Women's Mental Health met with FDA officials to request a REMS program for valproate products (Richmond 2025). This action, initiated by healthcare providers, is particularly noteworthy because certain components of REMS such as mandatory training might pose an additional burden on prescribers. It suggests that prescribers might value the reassurance provided by a REMS program when prescribing valproate to women of childbearing age, which is consistent with recent survey results about other REMS programs that include Elements To Assure Safe Use (Sarpatwari et al. 2023) and reports of clinician-initiated REMS-like processes such as patient informed consent when prescribing valproate (Suleiman et al. 2026). Several studies assessing valproate use over one or more decades have documented a decline, although it has remained the most widely used antiepileptic agent globally in 2022, with stronger growth observed in lower- and middle-income compared to higher-income countries (Chan et al. 2025). Studies describing decreases in use have typically focused on female populations of childbearing age and have attributed changes to the implementation of risk management programs or other regulatory action (Battino et al. 2024; Di Vito et al. 2023; Kim et al. 2019). However, while decreases in use have been more pronounced among female patients, use among males has also declined, likely attributable to the increasing availability of second-generation antiepileptic agents. Analyses in national private insurance claims data have estimated that about 3.6 per 10,000 women of childbearing age used valproate in 2020 in the US, compared to 6.4 of their male counterparts at a similar age range (Al-Bahou et al. 2022). Importantly, valproate remains one of the top ten known teratogenic medications used during pregnancy. A study examining prenatal exposure to a range of teratogenic medications between and found pregnancies with at least one during gestation per per in et al. 2024). A study from 2014 to assessing risk for prenatal exposure among teratogenic medication of childbearing potential reported the risk for prenatal exposure to valproate at (95% CI, per et al. 2026). during pregnancy from conception during treatment rather than of the medication during pregnancy and was as among women with public insurance compared to those in the risk for from valproate of the for benefit–risk analyses for use among women of childbearing potential during be more that is, the of the being valproate in this and whether effective or more effective are Hence, valproate use be to across for with use among women with epilepsy than with However, studies have a different women with epilepsy a to valproate of childbearing In a 2024 study of women of childbearing age in the United States with private insurance bipolar disorder for of valproate by migraine or while epilepsy only of treatment episodes et al. 2024). a 2025 analysis of data found that among women age valproate prescriptions between and 2022, the most was migraine or other by bipolar disorder with or epilepsy for only et al. 2026). decreases in valproate use among women of childbearing age have been with use for migraine showing no decrease over the (Al-Bahou et al. 2022). Finally, pregnancy during valproate use among women in private insurance in the United States with migraine or bipolar disorder were more than those of with epilepsy. Importantly, pregnancy have remained across the study from per in to in suggesting no in pregnancy prevention during use et al. 2024). In this study, than one of valproate use episodes with exposure to or suggesting that the teratogenic risk was likely or the need for effective contraception Studies on risk factors for exposure to teratogenic medications have that women at the of childbearing age, those in states in the United States with lower healthcare and those with public private insurance are at higher the of pregnancy prevention programs et al. et al. et al. 2022). With most exposure to teratogenic medications early in pregnancy, pregnancy and in prenatal care are other risk factors that place populations with limited healthcare access at risk et al. 2024). data suggest that approximately of pregnancies in the United States are with estimates to among certain groups National for Health 2026). For this about the potential effects of valproate use in pregnancy, and efforts to mitigate focus solely on pregnancy, but rather on women of reproductive age who can become In women of reproductive age healthcare providers and have limited access to et al. et al. 2024; and 2024). valproate risk to providers in a range of and to include and can help ensure that women of reproductive age taking valproate or who may be valproate can be at points of about associated and to medications when As with any medication before and during pregnancy, informed treatment should be to the the severity of their medication effective pregnancy maternal and is evidence that valproate is a teratogen that with for the offspring of exposed their and utilization of valproate has in both over time in the US, the risk for prenatal exposure among has suggesting that current risk management efforts have been in that women use effective contraception during use and that providers not valproate to women who may be the of evidence, frequency of prenatal and severity of it that valproate in a similar risk as isotretinoin or thalidomide, about currently risk management its significant public health implementation of pregnancy prevention programs in countries, and the that medical societies have such programs to ensure an safety for providers when valproate is the only treatment for a we call for implementation of a REMS for the of prenatal a valproate REMS the use of contraception or during treatment and require a negative pregnancy when prescribing valproate to a woman of childbearing In of or use during pregnancy, both provider and patient should that benefit–risk has been and found the of the we not it to a REMS that covers paternal use of to the development of this The have to reports from and and from and reports from by and to on pregnancy registry scientific The other no of Data availability is not to this as no data were or

  PublisherDOI

• Risk of Fetal Exposure to Teratogenic Medications: Development of Evidence for the Teratogenic Risk Impact and Mitigation (TRIM) Tool

  Drug Safety · 2026-04-17 · 1 citations

  articleSenior author
  PublisherDOI

• Neonatal Outcome Ascertainment in Mother‐Infant Paired Claims

  Pharmacoepidemiology and Drug Safety · 2026-01-18

  articleOpen accessSenior authorCorresponding

  PURPOSE: Claims data are a valuable source to study neonatal outcomes across a wide range of clinical questions. Infants' delayed enrollment in infant insurance poses challenges in capture of neonatal outcomes, which may be charged to the maternal health plan, posing misclassification risks. We evaluated outcome ascertainment across three infant enrollment scenarios. METHODS: We used Merative MarketScan databases (2012-2018) in the United States to construct a mother-infant linked cohort and assess the outcome ascertaiment precision with varying infant enrollment requirement. RESULT: We found that allowing delayed infant enrollment in their own insurance within the first 4 weeks of life retained sample size, nearly doubled case numbers and yielded outcome prevalences similar to those of cohorts with full enrollment since birth. Use of maternal claims in addition to infant claims in this cohort made minor contributions to case capture for neonatal-specific outcomes, while significantly decreasing specificity of more general outcomes. Longer delays in enrollment yielded lower outcome prevalences with higher contributions of maternal claims even for neonatal-specific outcomes. For small for gestational age (SGA), both maternal and infant claims contributed similar proportions of cases. CONCLUSION: These findings inform strategies for outcome ascertainment in claims-based perinatal research and emphasize outcome-specific case ascertainment strategies to balance sensitivity and specificity.

  PublisherOA PDFDOI

• A new tool for pregnancy research: a unified definition for major congenital malformation across ICD eras

  American Journal of Epidemiology · 2026-03-18

  article

  Composite major congenital malformation (MCM) outcomes are commonly used to assess teratogenic effects of prenatal medication exposure, but this approach dilutes effect estimates when the risk is confined to a specific MCM. Tree-based scan statistics address this by screening outcomes using a hierarchical tree, enabling detection of specific risks without predefined hypotheses. To apply this method across ICD-9-CM and ICD-10-CM eras, we developed a unified hierarchical outcomes tree for MCM. We selected ICD-9-CM and ICD-10-CM codes classified as congenital anomalies, removing minor malformations, chromosomal anomalies, and single-gene conditions. A multi-level tree was built based on the Multilevel Clinical Classification Software, General Equivalence Mappings, and expert review. We validated the tree using birth cohorts from MarketScan and Medicaid databases (2011-2013; 2016-2018), assessing balance of MCM prevalences within one year of birth via standardized mean differences (SMDs). The final tree included 1023 codes, organized into 244 clinical MCM groups at the most granular level. We identified 572 107 (2011-2013) and 360 167 infants (2016-2018) in MarketScan and 362 820 and 3 500 589 infants in Medicaid. All SMDs were below 0.1, indicating consistency across coding eras. This hierarchical MCM tree bridges ICD-9-CM and ICD-10-CM, enabling consistent outcome definitions and enhancing detection of specific teratogenic risks.

  PublisherDOI

• Disparities in Initial Antihypertensive Intensity by Sex, Race and Ethnicity in Newly Treated Patients With Hypertension

  American Journal of Hypertension · 2025-04-22 · 1 citations

  articleOpen access

  BACKGROUND: Sex, race, and ethnicity disparities in hypertension (HTN) treatment intensity have been previously described. It remains unclear if these disparities occur at treatment onset and whether they can be explained by differences in clinical factors. METHODS: We conducted a retrospective cross-sectional study of adults with newly treated HTN using linked EHR + claims data from OneFlorida + Consortium. We included Florida Medicaid & Medicare-recipients diagnosed with HTN and prescribed ≥ 1 first-line antihypertensive during 2013-2020. We used generalized linear models to estimate differences in total therapeutic intensity score (TTIS)-a patient's total daily dose (TDD) divided by recommended maximum TDD for a drug, summed across entire regimen-by sex, race, and ethnicity. We then modeled the same, controlling for demographics, blood pressure, and relevant comorbidities. RESULTS: In total 4,094 patients (mean age 58 ± 16; female 57.6%; White 56.7%) were included. We observed variations in the initiation of antihypertensive classes by sex, race and ethnicity. In univariate analyses, men averaged 7.6% (95% CI: 3.9%-11.3%) greater TTIS versus women and Black individuals averaged 10.5% (95% CI: 6.6%-14.3%) greater TTIS versus White individuals, whereas no disparities were observed by ethnicity. After adjusting for clinical factors, these disparities persisted: men had 7.6% (95% CI: 3.9%-11.4%) greater TTIS versus women, and Black individuals had 17.9% (95% CI: 13.8%-21.9%) greater TTIS versus White individuals. CONCLUSIONS: We observed disparities in treatment intensity by sex and race that were not explained by differences in clinical factors. There was sex-based variation in practice patterns, and Black individuals received more intensive initial antihypertensive therapy than White individuals.

  PublisherOA PDFDOI

• Risk Management of Valproate and Other Teratogenic Anticonvulsants in the Era of Proliferating Use

  The Journal of Pediatric Pharmacology and Therapeutics · 2025-06-01

  articleOpen access1st authorCorresponding

  Valproic acid, carbamazepine and topiramate have well-known teratogenic risk and all 3 rank among the top 10 teratogenic medications with the highest prenatal exposure risk. Importantly, pregnancies exposed to valproic acid are not dominated by patients with epilepsy but rather with less serious conditions such as migraine. In the United States, only a weight loss combination product containing topiramate has a mandatory pregnancy prevention program, a so-called Risk Evaluation and Mitigation Strategy (REMS), while prevention of fetal exposure to all three single ingredient products relies on information in the product labeling and a medication guide provided at dispensing. REMS have been avoided for antinconvulsants because of concerns about reduced medication access for patients with serious conditions such as epilepsy, hence weighting maternal harm due to uncontrolled disease against adverse pregnancy or infant outcomes. However, the broad and growing spectrum of indications for all three medications, paired with increasingly strict abortion laws that may not allow pregnancy termination if accidental fetal exposure occurs, may require re-assessment of the benefit-risk of REMS. Here we argue that formal quantitative approaches are needed that allow assessments of maternal and infant risk, considering maternal disease, adverse pregnancy outcomes and teratogenic effects on infants, and the overall public health impact of REMS for anticonvulsants. For valproic acid, given its broad use, high risk of fetal exposure, and profound impact on child health, we predict the public health impact of a REMS will be favorable.

  PublisherOA PDFDOI

• Respiratory Syncytial Virus Incidence in Young Children in the United States: Impact of Methodologies and Patient Characteristics

  Influenza and Other Respiratory Viruses · 2025-04-01 · 2 citations

  articleOpen accessSenior authorCorresponding

  OBJECTIVES: This study aimed to estimate the incidence of respiratory syncytial virus (RSV) infections in US inpatient and outpatient settings. METHODS: We established national cohorts of privately insured children < 5 years (2011-2019) to estimate annual and seasonal incidences of lower respiratory tract infection (LRTI), RSV-LRTI, and RSV acute respiratory infection (RSV-ARI), stratified by age and high-risk conditions per American Academy of Pediatrics definitions. Sensitivity analyses varied episode definitions and assessed the impact of immunoprophylaxis and RSV under-ascertainment. RESULTS: Among 6,767,107 children, annual RSV-LRTI rates dropped with increasing age in both inpatient (7.9 for age < 1 year to 0.2 for age 4 per 1000 person-years) and outpatient settings (48.3 to 1.6). Most RSV-ARI (~80%-90%) was RSV-LRTI. RSV-LRTI accounted for > half of LRTI hospitalizations among infants (7.9 RSV-LRTI versus 14.7 LRTI) and for ~20% outpatient LRTI (48.3 versus 250.3), but this contribution declined with older age. Outpatient RSV-LRTI was > 5 times inpatient rates. Inpatient RSV-LRTI rates dropped consistently with increasing gestational age (GA) (35.6 for GA < 29 weeks versus 7.6 for term infants), while outpatient rates were similar across GA groups (54.0 versus 51.6). Infants with Down syndrome had the highest RSV-LRTI rates, and any high-risk group had rates >2 times higher than healthy term infants. Across all strata, seasonal rates were > 2 annual rates. Modeling suggested that claims data captured 42% of all RSV episodes. CONCLUSION: This study provides national, population-based estimates of medically attended RSV infections across age groups and high-risk strata. Results allow granular assessments of disease burden to guide recommendations for new RSV prevention strategies.

  PublisherOA PDFDOI

• Risk of Seizure Associated With Concomitant Use of Tramadol and Antidepressants in Older Nursing Home Residents

  Neurology · 2025-10-08

  articleOpen access

  BACKGROUND AND OBJECTIVES: Concomitant use of tramadol and antidepressants with potent inhibition of the cytochrome P450 2D6 (CYP2D6) enzyme is postulated to increase risk of seizures in older adults; yet, such an association has not been empirically tested in populations. We aimed to examine the association of concomitant tramadol and CYP2D6-inhibiting vs CYP2D6-neutral antidepressant use and the risk of seizures among older nursing home (NH) residents. METHODS: This population-based cohort study was conducted using a 100% Medicare NH sample from January 2010 to December 2021. We included long-term residents aged 65 years or older who initiated antidepressants on existing tramadol use (tramadol-antidepressant users) or initiated tramadol on existing antidepressant use (antidepressant-tramadol users). Patients were followed up until the end of 1 year, NH discharge, death, or study end. The key exposure was concomitant use of tramadol with CYP2D6-inhibiting vs CYP2D6-neutral antidepressants. The key outcome was incident rates of medical encounters with a diagnosis of seizure and analyzed using negative binomial or Poisson regression models adjusted for baseline covariates (e.g., pain status and depressive, physical, and cognitive function) through the inverse probability of treatment weighting. RESULTS: We identified 11,162 concomitant tramadol-antidepressant users (mean [SD] age, 86.2 [8.5] years; 9,077 [81.3%] female) and 58,994 concomitant antidepressant-tramadol users (mean [SD] age, 85.3 [8.4] years; 47,053 [79.8%] female). The incidence rate of seizures was 16.10 and 20.17 per 100 patient-years, respectively, for the tramadol-antidepressant and antidepressant-tramadol group. In both subgroups, co-use of tramadol with CYP2D6-inhibiting (vs with CYP2D6-neutral) antidepressants was associated with higher adjusted incidence rate ratios of seizures (1.09 [95% CI 1.02-1.18] and 1.06 [95% CI 1.03-1.10]). Findings were corroborated by a negative control exposure analysis in which co-use of hydrocodone with CYPD2D6-inhibiting (vs CYP2D6-neutral) antidepressants was not associated with risk of seizures. DISCUSSION: Concomitant use of tramadol with CYP2D6-inhibiting vs CYP2D6-neutral antidepressants was associated with increased risk of seizures. Findings are only generalizable to long-term NH populations and are subject to residual confounding. Clinicians should be mindful of seizure risk in older patients who use tramadol concomitantly with antidepressants, particularly CYP2D6-inhibiting antidepressants. CLASSIFICATION OF EVIDENCE: This study provides Class II evidence that the combination of tramadol and CYP2D6-inhibiting antidepressants is associated with a higher risk of seizures compared with the combination of tramadol and CYP2D6-neutral antidepressants.

  PublisherOA PDFDOI

• Validation of diagnosis codes for low birth weight and small for gestational age in the Medicaid Analytic eXtract database

  American Journal of Epidemiology · 2025-01-14 · 1 citations

  articleSenior author

  The accuracy of low birth weight (LBW) and small for gestational age (SGA) in administrative health care records is crucial for perinatal studies but there are few published validity studies. Using 1999-2010 Medicaid Analytic eXtract (MAX) data linked to birth certificates (BCs), we identified mother-infant dyads (≥30 days enrollment after delivery, with valid gestational age [GA] and birth weight [BW] data). We identified LBW and SGA according to International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes. Infants with BW < 10% of the US reference were flagged as SGA. For LBW group diagnoses, we imputed BW using median, mean BW from BCs, and ICD code boundaries of infants in the same LBW group. We calculated the sensitivity, specificity, and positive and negative predictive values to assess performance. We identified 1 536 272 live births. All LBW groups had low Ses and high SPs and NPVs, whereas PPVs varied. Among infants with SGA diagnoses based on GA/BW from the BC, SE of the SGA codes was 13.36%, SP was 99.01%, and PPV was 67.37%. Combining imputation with LBW codes increased SE up to 22.09% (lower boundary) but decreased PPV to 41.53% (lower boundary). The ICD-9-CM codes from administrative health care records had low SE but high SP. Imputation based on GA and BW did not add much value to SGA identification.

  PublisherDOI

• Unrelieved pain and risk of opioid use disorder or overdose in older adults prescribed opioids

  Pain · 2025-03-17 · 3 citations

  articleOpen accessSenior author

  ABSTRACT: It is unclear to what extent unrelieved pain, the most common motive for prescription opioid misuse, is associated with risks of opioid use disorder (OUD) and opioid overdose (OD) among older adults with prescribed opioids. This retrospective cohort study was conducted among Health and Retirement Study (HRS) participants with linked Medicare claims data between 2006 and 2021. Participants aged 65 years or older with chronic pain who had received at least 1 opioid prescription entered the cohort in an HRS-assessed pain assessment (index) between 2008 and 2020. We included 2 time-varying measures of HRS-assessed pain exposure: uncontrolled pain, defined as having moderate or severe pain, and high-impact pain, defined as having moderate to severe pain that impacted daily activities. Primary outcomes of incident OUD or OD diagnosis were analyzed using separate Cox regression models with marginal structural modeling. Of 3104 eligible participants identified, 1359 (43.8%) had uncontrolled pain and 1044 (33.6%) experienced high-impact pain in the index wave. In the marginal structural modeling-adjusted Cox regression model, patients with uncontrolled (vs controlled) pain had higher risks of OUD (adjusted hazard ratio [AHR] 9.70; 95% confidence interval [CI], 4.56-20.63) and OD (AHR 2.46; 95% CI 1.30-4.66). The AHR for OUD was 6.74 (95% CI 3.76-12.08) and for OD was 1.96 (95% CI 1.07-3.60) times higher for patients with vs without high-impact pain. Our findings underscore the importance of regular assessment and modification of pain management for older patients whose pain remains unrelieved after opioid treatment, to lower the risk of OUD and OD.

  PublisherDOI

Recent grants

• NIH Grant R01HS018506

  NIH · $482k · 2012

Frequent coauthors

• Yu‐Jung Jenny Wei

  The Ohio State University

  134 shared

• Thuy Nhu Thai

  University of Florida

  118 shared

• Joshua D. Brown

  Center for Drug Evaluation and Research

  99 shared

• Amie Goodin

  63 shared

• Scott Martin Vouri

  Center for Drug Evaluation and Research

  63 shared

• Amir Sarayani

  Johnson & Johnson (United States)

  57 shared

• Laurence M. Solberg

  University of Florida

  51 shared

• Christian Hampp

  Regeneron (United States)

  50 shared

Labs

• Translational Drug Development CorePI

Education

• Ph.D., Pharmacoepidemiology

  Charité Humboldt University

• Other

  Friedrich Wilhelm University

Awards & honors

• Dr. Robert and Barbara Crisafi Chair for Medication Safety (…
• Fellow of the International Society of Pharmacoepidemiology…
• Inducted in the Academy of Science, Engineering and Medicine…