Xiaojuan Li
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Xiaojuan Li has a total of 150 co-authors. Here are the top co-authors by number of shared publications:
- Sengwee Toh
Harvard Pilgrim Health Care Institute Department of Population Medicine Boston, MA 02215
7 shared publications M. Alan Brookhart
Department of Population Health Sciences Duke University Durham North Carolina USA
3 shared publicationsSengwee Toh
Department of Population Medicine, Harvard Medical School and Harvard Pilgrim Health Care Institute, Boston, US
3 shared publicationsDongdong Li
Department of Population Medicine, Harvard Pilgrim Health Care Institute, Boston, MA, USA
2 shared publications- Elizabeth A. Bayliss
Institute for Health Research, Kaiser Permanente Colorado, Aurora, CO, USA; Department of Family Medicine, University of Colorado School of Medicine, Aurora, CO, USA
2 shared publications - Jenna Wong
Department of Population Medicine, Harvard Medical School and Harvard Pilgrim Health Care Institute, Boston, MA, USA
2 shared publications
Latest Publications
- Open Access
- 18 October 2026
- Published by Wiley in Pharmacoepidemiology and Drug Safety
- Vol. 35 (11) , e70510
Selective Prescribing of First‐Line SGLT2 Inhibitor and GLP ‐1 RA in Adults With Severe Mental Illness and Type 2 Diabetes
Purpose: SGLT2 inhibitors (SGLT2i) and GLP‐1 receptor agonists (GLP‐1 RA) provide cardiovascular (CV) benefits, but their first‐line use among adults with severe mental illness (SMI) is not well characterized. We evaluated selective prescribing of SGLT2i, GLP‐1 RA, and metformin among adults with SMI and type 2 diabetes (T2D). Methods: Using 2017–2022 Medicare claims, we identified adults with SMI and T2D who initiated first‐line SGLT2i, GLP‐1 RA, or metformin during 1/1/2018–12/31/2021. By calendar year, standardized mean differences (SMD) assessed baseline characteristic differences, and propensity score overlap coefficients (OVL) assessed group comparability between SGLT2i or GLP‐1 RA initiators and metformin initiators. Results: We identified 1322 SGLT2i, 2306 GLP‐1 RA, and 21 144 metformin initiators. From 2018 to 2021, SGLT2i initiators shifted from younger to older than metformin initiators (SMD, −0.11 to 0.19), while differences in CVD (0.10 to 0.50) and CKD (0.18 to 0.62) increased. GLP‐1 RA initiators consistently had higher prevalence of obesity (SMD, 0.33 to 0.41) and CKD (0.42 to 0.28) than metformin initiators. Compared with metformin initiators, both SGLT2i and GLP‐1 RA initiators had lower prevalence of schizophrenia, antipsychotic use, and recent psychiatrist visits but higher prevalence of major depressive disorder; these differences generally persisted over time. OVL increased from 42.7% to 50.7% for SGLT2i and from 42.3% to 50.0% for GLP‐1 RA versus metformin. Conclusions: First‐line SGLT2i and GLP‐1 RA prescribing among adults with SMI was selective according to cardiometabolic and mental health‐related characteristics. Future comparative studies should account for these differences and temporal prescribing changes to mitigate confounding.
- 18 October 2026
- Published by Wiley in Pharmacoepidemiology and Drug Safety
- Vol. 35 (11) , e70510
Open AccessSelective Prescribing of First‐Line SGLT2 Inhibitor and GLP ‐1 RA in Adults With Severe Mental Illness and Type 2 Diabetes
Purpose: SGLT2 inhibitors (SGLT2i) and GLP‐1 receptor agonists (GLP‐1 RA) provide cardiovascular (CV) benefits, but their first‐line use among adults with severe mental illness (SMI) is not well characterized. We evaluated selective prescribing of SGLT2i, GLP‐1 RA, and metformin among adults with SMI and type 2 diabetes (T2D). Methods: Using 2017–2022 Medicare claims, we identified adults with SMI and T2D who initiated first‐line SGLT2i, GLP‐1 RA, or metformin during 1/1/2018–12/31/2021. By calendar year, standardized mean differences (SMD) assessed baseline characteristic differences, and propensity score overlap coefficients (OVL) assessed group comparability between SGLT2i or GLP‐1 RA initiators and metformin initiators. Results: We identified 1322 SGLT2i, 2306 GLP‐1 RA, and 21 144 metformin initiators. From 2018 to 2021, SGLT2i initiators shifted from younger to older than metformin initiators (SMD, −0.11 to 0.19), while differences in CVD (0.10 to 0.50) and CKD (0.18 to 0.62) increased. GLP‐1 RA initiators consistently had higher prevalence of obesity (SMD, 0.33 to 0.41) and CKD (0.42 to 0.28) than metformin initiators. Compared with metformin initiators, both SGLT2i and GLP‐1 RA initiators had lower prevalence of schizophrenia, antipsychotic use, and recent psychiatrist visits but higher prevalence of major depressive disorder; these differences generally persisted over time. OVL increased from 42.7% to 50.7% for SGLT2i and from 42.3% to 50.0% for GLP‐1 RA versus metformin. Conclusions: First‐line SGLT2i and GLP‐1 RA prescribing among adults with SMI was selective according to cardiometabolic and mental health‐related characteristics. Future comparative studies should account for these differences and temporal prescribing changes to mitigate confounding.
- Open Access
- 6 December 2025
- Published by Springer Nature in BMC Geriatrics
- Vol. 26 (1)
Treatment patterns of symptomatic treatments for Alzheimer’s disease and related dementias
Limited real-world evidence exists on the patterns of symptomatic treatment use for Alzheimer’s disease and related dementias (ADRD). Using data from
- 6 December 2025
- Published by Springer Nature in BMC Geriatrics
- Vol. 26 (1)
Open AccessTreatment patterns of symptomatic treatments for Alzheimer’s disease and related dementias
Limited real-world evidence exists on the patterns of symptomatic treatment use for Alzheimer’s disease and related dementias (ADRD). Using data from
- Open Access
- 14 November 2025
- Published by Wiley in Pharmacoepidemiology and Drug Safety
- Vol. 34 (11) , e70255
The Reporting and Methodological Recommendations for Observational Studies Estimating the Effects of Deprescribing Medications (REMROSE‐D) ISPE‐Endorsed Guidance
Purpose: Pharmacoepidemiologic studies on deprescribing are challenging to implement, yet little guidance exists on methods to avoid bias and minimum reporting for replicability and appraisal. We developed consensus recommendations for the methods and reporting of observational studies that aim to examine the effects of deprescribing. Methods: We formed candidate recommendations based on our prior systematic review that methodologically appraised observational studies on deprescribing. We then conducted a two‐round modified Delphi process with researchers working in deprescribing pharmacoepidemiology to refine, select, and reach consensus on recommendations for a checklist based on > 70% agreement of their importance. We termed this list the REMROSE‐D (Reporting and Methodological Recommendations for Observational Studies estimating the Effects of Deprescribing medications) guidance. Results: Twenty‐three candidate recommendations were presented to the Delphi panel. The round 1 survey was completed by 55 participants, and 18 of the 23 candidate recommendations were selected for inclusion. Five candidate recommendations without consensus plus two additional items suggested by participants were included in a round 2 survey of 25 deprescribing researchers. Five of these seven items garnered consensus for inclusion, and two were excluded. The final REMROSE‐D guidance contains 23 recommendations for the methods and reporting of observational research on deprescribing. Conclusion: To ensure rigor and reproducibility in observational studies of the effects of deprescribing, the REMROSE‐D guidance provides recommendations for important reporting and methods considerations, including time zero, precise definitions of deprescribing, addressing confounding by indication, and careful consideration of follow‐up to avoid immortal time bias.
- 14 November 2025
- Published by Wiley in Pharmacoepidemiology and Drug Safety
- Vol. 34 (11) , e70255
Open AccessThe Reporting and Methodological Recommendations for Observational Studies Estimating the Effects of Deprescribing Medications (REMROSE‐D) ISPE‐Endorsed Guidance
Purpose: Pharmacoepidemiologic studies on deprescribing are challenging to implement, yet little guidance exists on methods to avoid bias and minimum reporting for replicability and appraisal. We developed consensus recommendations for the methods and reporting of observational studies that aim to examine the effects of deprescribing. Methods: We formed candidate recommendations based on our prior systematic review that methodologically appraised observational studies on deprescribing. We then conducted a two‐round modified Delphi process with researchers working in deprescribing pharmacoepidemiology to refine, select, and reach consensus on recommendations for a checklist based on > 70% agreement of their importance. We termed this list the REMROSE‐D (Reporting and Methodological Recommendations for Observational Studies estimating the Effects of Deprescribing medications) guidance. Results: Twenty‐three candidate recommendations were presented to the Delphi panel. The round 1 survey was completed by 55 participants, and 18 of the 23 candidate recommendations were selected for inclusion. Five candidate recommendations without consensus plus two additional items suggested by participants were included in a round 2 survey of 25 deprescribing researchers. Five of these seven items garnered consensus for inclusion, and two were excluded. The final REMROSE‐D guidance contains 23 recommendations for the methods and reporting of observational research on deprescribing. Conclusion: To ensure rigor and reproducibility in observational studies of the effects of deprescribing, the REMROSE‐D guidance provides recommendations for important reporting and methods considerations, including time zero, precise definitions of deprescribing, addressing confounding by indication, and careful consideration of follow‐up to avoid immortal time bias.
- Open Access
- 4 August 2025
- Published by Wiley in Clinical Pharmacology & Therapeutics
- Vol. 118 (5) , 1170-1179
Effectiveness and Safety of Reduced‐Dose Prasugrel in an East Asian Population: PRASFIT ‐ ACS Emulation Using National Health Insurance Claims Data
Prasugrel is recommended for acute coronary syndrome (ACS) patients undergoing percutaneous coronary intervention (PCI). However, East Asians exhibit a stronger antiplatelet response, raising concerns about bleeding risks with standard dosing. This retrospective cohort study evaluated the real‐world effectiveness and safety of reduced‐dose prasugrel in an East Asian population by emulating the pivotal PRASFIT‐ACS trial using Taiwan’s National Health Insurance claims data. In this study, ACS patients were classified into reduced‐dose prasugrel or clopidogrel groups post‐PCI. We estimated the intention‐to‐treat effect of initiating reduced‐dose prasugrel vs. clopidogrel on the risk of major adverse cardiovascular events (MACE) and the per‐protocol effect of treating with reduced‐dose prasugrel on major bleeding risk using Cox proportional hazard models. A total of 4833 ACS patients were analyzed, including 977 (20.2%) who received reduced‐dose prasugrel. Reduced‐dose prasugrel showed comparable effectiveness to clopidogrel in preventing MACE (HR: 0.96, 95% CI: 0.72–1.29) and did not increase bleeding risk (HR: 0.82, 95% CI: 0.36–1.90), consistent with PRASFIT‐ACS results (HR MACE : 0.85, 95% CI: 0.62–1.16; HR Bleeding : 0.82, 95% CI: 0.39–1.73). A significant reduction in cardiovascular mortality was observed (HR: 0.51, 95% CI: 0.29–0.89), differing from PRASFIT‐ACS findings (HR: 1.21, 95% CI: 0.48–3.06). Our results suggest that reduced‐dose prasugrel was effective in preventing thrombotic events in the Taiwanese population without significantly increasing bleeding risk. These findings align with pivotal trial outcomes while highlighting the importance of race‐specific dose adjustments to optimize ACS treatment strategies. Further research is needed to evaluate whether these results are applicable to other East Asian populations.
- 4 August 2025
- Published by Wiley in Clinical Pharmacology & Therapeutics
- Vol. 118 (5) , 1170-1179
Open AccessEffectiveness and Safety of Reduced‐Dose Prasugrel in an East Asian Population: PRASFIT ‐ ACS Emulation Using National Health Insurance Claims Data
Prasugrel is recommended for acute coronary syndrome (ACS) patients undergoing percutaneous coronary intervention (PCI). However, East Asians exhibit a stronger antiplatelet response, raising concerns about bleeding risks with standard dosing. This retrospective cohort study evaluated the real‐world effectiveness and safety of reduced‐dose prasugrel in an East Asian population by emulating the pivotal PRASFIT‐ACS trial using Taiwan’s National Health Insurance claims data. In this study, ACS patients were classified into reduced‐dose prasugrel or clopidogrel groups post‐PCI. We estimated the intention‐to‐treat effect of initiating reduced‐dose prasugrel vs. clopidogrel on the risk of major adverse cardiovascular events (MACE) and the per‐protocol effect of treating with reduced‐dose prasugrel on major bleeding risk using Cox proportional hazard models. A total of 4833 ACS patients were analyzed, including 977 (20.2%) who received reduced‐dose prasugrel. Reduced‐dose prasugrel showed comparable effectiveness to clopidogrel in preventing MACE (HR: 0.96, 95% CI: 0.72–1.29) and did not increase bleeding risk (HR: 0.82, 95% CI: 0.36–1.90), consistent with PRASFIT‐ACS results (HR MACE : 0.85, 95% CI: 0.62–1.16; HR Bleeding : 0.82, 95% CI: 0.39–1.73). A significant reduction in cardiovascular mortality was observed (HR: 0.51, 95% CI: 0.29–0.89), differing from PRASFIT‐ACS findings (HR: 1.21, 95% CI: 0.48–3.06). Our results suggest that reduced‐dose prasugrel was effective in preventing thrombotic events in the Taiwanese population without significantly increasing bleeding risk. These findings align with pivotal trial outcomes while highlighting the importance of race‐specific dose adjustments to optimize ACS treatment strategies. Further research is needed to evaluate whether these results are applicable to other East Asian populations.
- 17 July 2025
- Published by Wiley in Journal of the American Geriatrics Society
- Vol. 73 (10) , 3203-3207
Lecanemab Therapy Use Patterns for Alzheimer's Disease Among Early Initiators in a Large National Health Plan
Background: Lecanemab, an anti‐amyloid monoclonal antibody, has been approved for treatment of early Alzheimer's disease. However, real‐world data remain limited regarding use. Methods: Using data from a large national health plan with primarily Medicare Advantage enrollees, we characterized demographic and clinical characteristics in the 6 months preceding the initial infusion among individuals who initiated lecanemab between 1/1/2023 and 06/30/2024. We also characterized the patterns of infusions and the timing of MRIs for safety monitoring following the initial infusion among initiators who had ≥ 3 months of follow‐up. We further assessed trends in lecanemab use between 1/1/2023 and 12/31/2024. Results: Of the 195 lecanemab initiators, the average age was 74.6 (SD = 5.5) years, 62.1% were female, 87.7% were White, 4.1% were Black, 1.5% were Hispanic, and 98.5% were on a Medicare Advantage plan. Almost all initiators resided in nonrural areas (96.4%); 2.6% used anticoagulants, and 1.5% used antiplatelets. Among the 119 initiators who had ≥ 3 months of follow‐up, 40 (33.6%) received 7 total infusions, the expected number of infusions based on the recommended schedule of every 2 weeks. From the initial infusion, the average number of days to the first MRI was 47.1 (SD = 15.8), and to the second MRI scan was 73.4 (SD = 12.0) days, consistent with the recommended schedule. During 2023–2024, there were 526 patients who had ever received lecanemab. The increase in new initiators was modest over 2024, with only 43 more initiators during the final ( n = 137) versus the first quarter ( n = 94) of 2024. Conclusions: This study demonstrated slow uptake of lecanemab among Medicare Advantage beneficiaries. Adherence to the ideal treatment schedule was less than recommended. Early lecanemab users were not representative of the population likely eligible for treatment nationally. Further research is warranted to track longer‐term trends in utilization, as well as reasons for treatment interruption or discontinuation in real‐world populations.
- 17 July 2025
- Published by Wiley in Journal of the American Geriatrics Society
- Vol. 73 (10) , 3203-3207
Lecanemab Therapy Use Patterns for Alzheimer's Disease Among Early Initiators in a Large National Health Plan
Background: Lecanemab, an anti‐amyloid monoclonal antibody, has been approved for treatment of early Alzheimer's disease. However, real‐world data remain limited regarding use. Methods: Using data from a large national health plan with primarily Medicare Advantage enrollees, we characterized demographic and clinical characteristics in the 6 months preceding the initial infusion among individuals who initiated lecanemab between 1/1/2023 and 06/30/2024. We also characterized the patterns of infusions and the timing of MRIs for safety monitoring following the initial infusion among initiators who had ≥ 3 months of follow‐up. We further assessed trends in lecanemab use between 1/1/2023 and 12/31/2024. Results: Of the 195 lecanemab initiators, the average age was 74.6 (SD = 5.5) years, 62.1% were female, 87.7% were White, 4.1% were Black, 1.5% were Hispanic, and 98.5% were on a Medicare Advantage plan. Almost all initiators resided in nonrural areas (96.4%); 2.6% used anticoagulants, and 1.5% used antiplatelets. Among the 119 initiators who had ≥ 3 months of follow‐up, 40 (33.6%) received 7 total infusions, the expected number of infusions based on the recommended schedule of every 2 weeks. From the initial infusion, the average number of days to the first MRI was 47.1 (SD = 15.8), and to the second MRI scan was 73.4 (SD = 12.0) days, consistent with the recommended schedule. During 2023–2024, there were 526 patients who had ever received lecanemab. The increase in new initiators was modest over 2024, with only 43 more initiators during the final ( n = 137) versus the first quarter ( n = 94) of 2024. Conclusions: This study demonstrated slow uptake of lecanemab among Medicare Advantage beneficiaries. Adherence to the ideal treatment schedule was less than recommended. Early lecanemab users were not representative of the population likely eligible for treatment nationally. Further research is warranted to track longer‐term trends in utilization, as well as reasons for treatment interruption or discontinuation in real‐world populations.
- Open Access
- 15 April 2025
- Published by Wiley in Alzheimer's & Dementia: Translational Research & Clinical Interventions
- Vol. 11 (2) , e70080
Generating real‐world evidence in early Alzheimer's disease: Considerations for applying the target trial emulation framework to study the safety of anti‐amyloid therapies
Anti‐amyloid beta monoclonal antibodies (anti‐Aβ mAbs) have received approval from the US Food and Drug Administration for the treatment of patients with mild cognitive impairment or mild dementia due to Alzheimer's disease (collectively known as early AD) based on evidence from clinical trials. However, whether findings from these trials are generalizable to the real world is uncertain. We need reliable evidence on the real‐world safety of these treatments to inform decision making for clinicians, patients, and caregivers. Using lecanemab as an exemplar, we outline the key considerations in designing and implementing an observational study on safety and utilization outcomes using established administrative healthcare claims data sources with the target trial emulation framework. The target trial emulation framework is a rigorous causal inference framework that minimizes common biases in observational studies. The approach proposed here can be applied to evaluation of additional mAbs as they become available. Highlights: Little is known about real‐world safety of anti‐amyloid beta monoclonal antibodies for early Alzheimer's disease. Existing real‐world data can support studies of their safety and utilization outcomes. Target trial emulation can guide the design of these studies while minimizing bias. We provide key design and analytical considerations for future studies.
- 15 April 2025
- Published by Wiley in Alzheimer's & Dementia: Translational Research & Clinical Interventions
- Vol. 11 (2) , e70080
Open AccessGenerating real‐world evidence in early Alzheimer's disease: Considerations for applying the target trial emulation framework to study the safety of anti‐amyloid therapies
Anti‐amyloid beta monoclonal antibodies (anti‐Aβ mAbs) have received approval from the US Food and Drug Administration for the treatment of patients with mild cognitive impairment or mild dementia due to Alzheimer's disease (collectively known as early AD) based on evidence from clinical trials. However, whether findings from these trials are generalizable to the real world is uncertain. We need reliable evidence on the real‐world safety of these treatments to inform decision making for clinicians, patients, and caregivers. Using lecanemab as an exemplar, we outline the key considerations in designing and implementing an observational study on safety and utilization outcomes using established administrative healthcare claims data sources with the target trial emulation framework. The target trial emulation framework is a rigorous causal inference framework that minimizes common biases in observational studies. The approach proposed here can be applied to evaluation of additional mAbs as they become available. Highlights: Little is known about real‐world safety of anti‐amyloid beta monoclonal antibodies for early Alzheimer's disease. Existing real‐world data can support studies of their safety and utilization outcomes. Target trial emulation can guide the design of these studies while minimizing bias. We provide key design and analytical considerations for future studies.
- 24 October 2024
- Published by Wiley in Journal of the American Geriatrics Society
- Vol. 73 (3) , 697-706
Assessing causality in deprescribing studies: A focus on adverse drug events and adverse drug withdrawal events
Generating real‐world evidence about the effect of medication discontinuation or dose reduction on outcomes, such as reduction of adverse drug effects (ADE; intended benefit) and occurrence of adverse drug withdrawal events (ADWE; unintended harm), is crucial to informing deprescribing decisions. Determining the causal effects of deprescribing is difficult for many reasons, including lack of randomization in real‐world study designs and other design and measurement issues that pose threats to internal validity. The inherent challenge is how to identify the effects, both intended benefits and unintended harms, of a new medication stoppage or reduction when implemented in patients with many potential clinical and social risks that may influence the likelihood of deprescribing as well as outcomes. We discuss methodological issues of estimating the effect of medication discontinuation or reduction on risk of ADEs and ADWEs considering: (1) sampling study populations of sufficient size with the potential to demonstrate clinically meaningful and quantifiable outcomes, (2) accurate and appropriately timed measurement of covariates, outcomes, and discontinuation, and (3) statistical approaches to managing confounding and other biases inherent in long‐term medication use by individuals with multiple morbidities. Designing rigorous deprescribing studies that address internal validity threats will support evidence generation by improving the ability to assess benefits and harms when the exposure of interest is the absence of a medication. Iterative learnings about data quality, variable definition, variable measurement, and exposure‐outcome associations will inform strategies to improve the causal inferences possible in real‐world deprescribing studies.
- 24 October 2024
- Published by Wiley in Journal of the American Geriatrics Society
- Vol. 73 (3) , 697-706
Assessing causality in deprescribing studies: A focus on adverse drug events and adverse drug withdrawal events
Generating real‐world evidence about the effect of medication discontinuation or dose reduction on outcomes, such as reduction of adverse drug effects (ADE; intended benefit) and occurrence of adverse drug withdrawal events (ADWE; unintended harm), is crucial to informing deprescribing decisions. Determining the causal effects of deprescribing is difficult for many reasons, including lack of randomization in real‐world study designs and other design and measurement issues that pose threats to internal validity. The inherent challenge is how to identify the effects, both intended benefits and unintended harms, of a new medication stoppage or reduction when implemented in patients with many potential clinical and social risks that may influence the likelihood of deprescribing as well as outcomes. We discuss methodological issues of estimating the effect of medication discontinuation or reduction on risk of ADEs and ADWEs considering: (1) sampling study populations of sufficient size with the potential to demonstrate clinically meaningful and quantifiable outcomes, (2) accurate and appropriately timed measurement of covariates, outcomes, and discontinuation, and (3) statistical approaches to managing confounding and other biases inherent in long‐term medication use by individuals with multiple morbidities. Designing rigorous deprescribing studies that address internal validity threats will support evidence generation by improving the ability to assess benefits and harms when the exposure of interest is the absence of a medication. Iterative learnings about data quality, variable definition, variable measurement, and exposure‐outcome associations will inform strategies to improve the causal inferences possible in real‐world deprescribing studies.
- Open Access
- 11 September 2024
- Published by Oxford University Press (OUP) in American Journal of Epidemiology
- Vol. 194 (7) , 1999-2011
Tree-based scan statistics to generate drug repurposing hypotheses: a test case using sodium-glucose cotransporter-2 inhibitors
Most drug repurposing studies using real-world data focused on validating, instead of generating, hypotheses. We used tree-based scan statistics to generate repurposing hypotheses for sodium-glucose cotransporter-2 inhibitors (SGLT2i). We used an active-comparator, new-user design to create a 1:1 propensity-score matched cohort of SGLT2i and dipeptidyl peptidase-4 inhibitors (DPP4i) initiators in the MerativeTM MarketScan® Research Databases. Tree-based scan statistics were estimated across an ICD-10-CM-based hierarchical outcome tree using incident outcomes identified from hospital and outpatient diagnoses. We used an adjusted P≤0.01 as the threshold for statistical alert to prioritize associations for evaluation as repurposing signals. We varied the analyses by tree size, scanning level, and clinical settings for outcomes. There were 80,510 matched SGLT2i-DPP4i initiator pairs with 215,333 outcomes among SGLT2i initiators and 223,428 outcomes among DPP4i initiators. There were 18 prioritized associations, which included chronic kidney disease (P=0.0001), an expected signal, and anemia (P=0.0001). Heart failure (P=0.0167), another expected signal, was identified slightly beyond the statistical alert threshold. Narrowing the outcome tree, scanning at different tree levels, and including outcomes from different clinical settings influenced the scan statistics. We identified signals aligning with recently approved indications of SGLT2i, plus potential repurposing signals supported by existing evidence but requiring future validation.
- 11 September 2024
- Published by Oxford University Press (OUP) in American Journal of Epidemiology
- Vol. 194 (7) , 1999-2011
Open AccessTree-based scan statistics to generate drug repurposing hypotheses: a test case using sodium-glucose cotransporter-2 inhibitors
Most drug repurposing studies using real-world data focused on validating, instead of generating, hypotheses. We used tree-based scan statistics to generate repurposing hypotheses for sodium-glucose cotransporter-2 inhibitors (SGLT2i). We used an active-comparator, new-user design to create a 1:1 propensity-score matched cohort of SGLT2i and dipeptidyl peptidase-4 inhibitors (DPP4i) initiators in the MerativeTM MarketScan® Research Databases. Tree-based scan statistics were estimated across an ICD-10-CM-based hierarchical outcome tree using incident outcomes identified from hospital and outpatient diagnoses. We used an adjusted P≤0.01 as the threshold for statistical alert to prioritize associations for evaluation as repurposing signals. We varied the analyses by tree size, scanning level, and clinical settings for outcomes. There were 80,510 matched SGLT2i-DPP4i initiator pairs with 215,333 outcomes among SGLT2i initiators and 223,428 outcomes among DPP4i initiators. There were 18 prioritized associations, which included chronic kidney disease (P=0.0001), an expected signal, and anemia (P=0.0001). Heart failure (P=0.0167), another expected signal, was identified slightly beyond the statistical alert threshold. Narrowing the outcome tree, scanning at different tree levels, and including outcomes from different clinical settings influenced the scan statistics. We identified signals aligning with recently approved indications of SGLT2i, plus potential repurposing signals supported by existing evidence but requiring future validation.
- Open Access
- 7 May 2024
- Published by Springer Nature in Genome Biology
- Vol. 25 (1) , 1-46
Disruption of maternal vascular remodeling by a fetal endoretrovirus-derived gene in preeclampsia
Background: Preeclampsia, one of the most lethal pregnancy-related diseases, is associated with the disruption of uterine spiral artery remodeling during placentation. However, the early molecular events leading to preeclampsia remain unknown.Results: By analyzing placentas from preeclampsia, non-preeclampsia, and twin pregnancies with selective intrauterine growth restriction, we show that the pathogenesis of preeclampsia is attributed to immature trophoblast and maldeveloped endothelial cells. Delayed epigenetic reprogramming during early extraembryonic tissue development leads to generation of excessive immature trophoblast cells. We find reduction of de novo DNA methylation in these trophoblast cells results in selective overexpression of maternally imprinted genes, including the endoretrovirus-derived gene PEG10 (paternally expressed gene 10). PEG10 forms virus-like particles, which are transferred from the trophoblast to the closely proximate endothelial cells. In normal pregnancy, only a low amount of PEG10 is transferred to maternal cells; however, in preeclampsia, excessive PEG10 disrupts maternal vascular development by inhibiting TGF-beta signaling.Conclusions: Our study reveals the intricate epigenetic mechanisms that regulate trans-generational genetic conflict and ultimately ensure proper maternal–fetal interface formation.
- 7 May 2024
- Published by Springer Nature in Genome Biology
- Vol. 25 (1) , 1-46
Open AccessDisruption of maternal vascular remodeling by a fetal endoretrovirus-derived gene in preeclampsia
Background: Preeclampsia, one of the most lethal pregnancy-related diseases, is associated with the disruption of uterine spiral artery remodeling during placentation. However, the early molecular events leading to preeclampsia remain unknown.Results: By analyzing placentas from preeclampsia, non-preeclampsia, and twin pregnancies with selective intrauterine growth restriction, we show that the pathogenesis of preeclampsia is attributed to immature trophoblast and maldeveloped endothelial cells. Delayed epigenetic reprogramming during early extraembryonic tissue development leads to generation of excessive immature trophoblast cells. We find reduction of de novo DNA methylation in these trophoblast cells results in selective overexpression of maternally imprinted genes, including the endoretrovirus-derived gene PEG10 (paternally expressed gene 10). PEG10 forms virus-like particles, which are transferred from the trophoblast to the closely proximate endothelial cells. In normal pregnancy, only a low amount of PEG10 is transferred to maternal cells; however, in preeclampsia, excessive PEG10 disrupts maternal vascular development by inhibiting TGF-beta signaling.Conclusions: Our study reveals the intricate epigenetic mechanisms that regulate trans-generational genetic conflict and ultimately ensure proper maternal–fetal interface formation.
- Open Access
- 1 March 2024
- Published by Taylor & Francis in Pragmatic and Observational Research
- Vol. ume 15, 65-78
Using Claims Data to Predict Pre-Operative BMI Among Bariatric Surgery Patients: Development of the BMI Before Bariatric Surgery Scoring System (B3S3)
Background: Lack of body mass index (BMI) measurements limits the utility of claims data for bariatric surgery research, but pre-operative BMI may be imputed due to existence of weight-related diagnosis codes and BMI-related reimbursement requirements. We used a machine learning pipeline to create a claims-based scoring system to predict pre-operative BMI, as documented in the electronic health record (EHR), among patients undergoing a new bariatric surgery. Methods: Using the Optum Labs Data Warehouse, containing linked de-identified claims and EHR data for commercial or Medicare Advantage enrollees, we identified adults undergoing a new bariatric surgery between January 2011 and June 2018 with a BMI measurement in linked EHR data ≤ 30 days before the index surgery (n=3226). We constructed predictors from claims data and applied a machine learning pipeline to create a scoring system for pre-operative BMI, the B3S3. We evaluated the B3S3 and a simple linear regression model (benchmark) in test patients whose index surgery occurred concurrent (2011– 2017) or prospective (2018) to the training data. Results: The machine learning pipeline yielded a final scoring system that included weight-related diagnosis codes, age, and number of days hospitalized and distinct drugs dispensed in the past 6 months. In concurrent test data, the B3S3 had excellent performance (R2 0.862, 95% confidence interval [CI] 0.815– 0.898) and calibration. The benchmark algorithm had good performance (R2 0.750, 95% CI 0.686– 0.799) and calibration but both aspects were inferior to the B3S3. Findings in prospective test data were similar. Conclusion: The B3S3 is an accessible tool that researchers can use with claims data to obtain granular and accurate predicted values of pre-operative BMI, which may enhance confounding control and investigation of effect modification by baseline obesity levels in bariatric surgery studies utilizing claims data. Plain Language Summary: Pre-operative BMI is an important potential confounder in comparative effectiveness studies of bariatric surgeries.Claims data lack clinical measurements, but insurance reimbursement requirements for bariatric surgery often result in pre-operative BMI being coded in claims data.We used a machine learning pipeline to create a model, the B3S3, to predict pre-operative BMI, as documented in the EHR, among bariatric surgery patients based on the presence of certain weight-related diagnosis codes and other patient characteristics derived from claims data.Researchers can easily use the B3S3 with claims data to obtain granular and accurate predicted values of pre-operative BMI among bariatric surgery patients.
- 1 March 2024
- Published by Taylor & Francis in Pragmatic and Observational Research
- Vol. ume 15, 65-78
Open AccessUsing Claims Data to Predict Pre-Operative BMI Among Bariatric Surgery Patients: Development of the BMI Before Bariatric Surgery Scoring System (B3S3)
Background: Lack of body mass index (BMI) measurements limits the utility of claims data for bariatric surgery research, but pre-operative BMI may be imputed due to existence of weight-related diagnosis codes and BMI-related reimbursement requirements. We used a machine learning pipeline to create a claims-based scoring system to predict pre-operative BMI, as documented in the electronic health record (EHR), among patients undergoing a new bariatric surgery. Methods: Using the Optum Labs Data Warehouse, containing linked de-identified claims and EHR data for commercial or Medicare Advantage enrollees, we identified adults undergoing a new bariatric surgery between January 2011 and June 2018 with a BMI measurement in linked EHR data ≤ 30 days before the index surgery (n=3226). We constructed predictors from claims data and applied a machine learning pipeline to create a scoring system for pre-operative BMI, the B3S3. We evaluated the B3S3 and a simple linear regression model (benchmark) in test patients whose index surgery occurred concurrent (2011– 2017) or prospective (2018) to the training data. Results: The machine learning pipeline yielded a final scoring system that included weight-related diagnosis codes, age, and number of days hospitalized and distinct drugs dispensed in the past 6 months. In concurrent test data, the B3S3 had excellent performance (R2 0.862, 95% confidence interval [CI] 0.815– 0.898) and calibration. The benchmark algorithm had good performance (R2 0.750, 95% CI 0.686– 0.799) and calibration but both aspects were inferior to the B3S3. Findings in prospective test data were similar. Conclusion: The B3S3 is an accessible tool that researchers can use with claims data to obtain granular and accurate predicted values of pre-operative BMI, which may enhance confounding control and investigation of effect modification by baseline obesity levels in bariatric surgery studies utilizing claims data. Plain Language Summary: Pre-operative BMI is an important potential confounder in comparative effectiveness studies of bariatric surgeries.Claims data lack clinical measurements, but insurance reimbursement requirements for bariatric surgery often result in pre-operative BMI being coded in claims data.We used a machine learning pipeline to create a model, the B3S3, to predict pre-operative BMI, as documented in the EHR, among bariatric surgery patients based on the presence of certain weight-related diagnosis codes and other patient characteristics derived from claims data.Researchers can easily use the B3S3 with claims data to obtain granular and accurate predicted values of pre-operative BMI among bariatric surgery patients.