Objective Approximately 50% of patients with major depressive disorder (MDD) prematurely discontinue their antidepressant medication within 6 months, increasing risk of relapse. Pharmacogenomic (PGx) testing may improve medication adherence by informing treatment based on gene-drug interactions (GDI). Here, we evaluated whether PGx informed treatmentimproved medication adherence in MDD patients. Methods This was an observational, retrospective claims study of adult MDD patients who received a weighted multi-gene PGx test between 1 Jan 2015 and 30 September 2021 and switched medication. PGx results were linked with de-identified administrative claims data from the Optum Labs Data Warehouse. The PGx test report organized psychiatric medications into three categories: no known GDI (congruent), moderate GDI (congruent), and significant GDI (incongruent). Patients were assigned to the following groups based on the medication with the worst congruency 90 days pre- and post-PGx testing: incongruent-to-congruent, no-change-in-congruency, and congruent-to-incongruent. Medication adherence (proportion of days covered [PDC]) and discontinuation (a ≥ 45-day gap in medication fills) were assessed using pharmacy fill data during the 180 days following the date of medication switch (index date). Results Among 6224 patients with PGx testing, those in the incongruent-to-congruent group had the highest adherence (mean PDC 0.65, SD 0.33), compared to the congruent-to-incongruent (mean PDC 0.58, SD 0.34) (p < 0.05) and no-change-in-congruency groups (mean PDC 0.61, SD 0.34) (p < 0.05). The incongruent-to-congruent group also had the lowest discontinuation rate (46%) compared to the congruent-to-incongruent (55%) (p < 0.05) and no-change-in-congruency groups (50%) (p < 0.05). Conclusions Using PGx informed medication selection can improve medication adherence and reduce discontinuation among MDD patients.

Purpose: To compare pregnancy outcomes and management between patients screening positive for five microdeletions (microdeletion screen-positive, MDS+) and patients screening negative (microdeletion screen-negative, MDS-). Patients and Methods: Patients who received a prenatal cell-free DNA (pcfDNA) test that screens for microdeletions 22q11.2, 15q11.2, 1p36, 4p, and 5p and results were linked to de-identified insurance claims. Diagnosis and procedure codes were used to assess outcomes. Logistic and Poisson regression with adjustment for prior high-risk pregnancy and payer type were used to compare pregnancy outcomes and management in those with MDS+ and MDS-results. Results: A total of 119 MDS+ patients and 287,169 MDS-patients were eligible for analysis. During pregnancy, MDS+ patients were more likely than MDS-patients to have polyhydramnios (18.5% in MDS+ vs. 3.4% in MDS-; OR=6.3 [95% CI: 3.8-10.1]; p<0.001) and fetal growth restriction (22.2% in MDS+ vs. 10.4% in MDS-; OR=2.4 [95% CI: 1.5-3.7]; p<0.001). MDS+ patients were more likely to experience pregnancy loss (OR=3.2; 95% CI: 1.3-6.3; p=0.004) or terminate the pregnancy (OR=22.2 [95% CI: 8.6-46.8]; p<0.001). Among patients with a live birth, MDS+ were more likely to have a preterm delivery (25.0% in MDS+ vs. 15.4% in MDS-; OR=1.8 [95% CI: 1.0-3.0]; p=0.04). MDS+ patients had increased pregnancy management, including more invasive diagnostic testing (13.4% in MDS+ vs. 0.5% in MDS-; OR=31.0 [95% CI: 17.5-51.1]; p<0.001) and more frequent echocardiograms (RR=1.5 [95% CI: 1.2, 1.9]; p<0.001) compared to MDS-pregnancies. Conclusion: MDS+ pregnancies had elevated rates of ultrasound abnormalities, pregnancy loss, preterm birth, and pregnancy management as compared to an MDS-control group. These findings support the clinical utility of microdeletion screening in prenatal care.

BACKGROUND:Pharmacogenomic (PGx) testing can help improve response and remission rates for patients with major depressive disorder (MDD) and at least one treatment failure. To investigate real-world outcomes, we examined 1) significant gene-drug interactions (GDIs) and 2) healthcare resource utilization (HRU) in a large US insurance claims dataset. METHODS:Weighted multigene PGx testing results in adult patients with MDD were linked with deidentified US claims data. The PGx test report organized medications as congruent (no known or moderate GDI) or incongruent (significant GDI). Medication claims data before and after PGx testing was used to categorize patients as no change in congruency, incongruent-to-congruent, or congruent-to-incongruent. HRU (hospitalizations and emergency department visits) was compared in the 180 days before and after PGx testing. RESULTS:A total of 20,933 patients met inclusion criteria; 16,965 of whom filled medication prescriptions before and after PGx testing. After PGx testing, the proportion of patients filling prescriptions with significant GDIs was reduced (26.1% pretesting vs 15.9% posttesting). All HRU was significantly reduced ( P < 0.001) after PGx testing except for nonpsychiatric hospitalizations ( P > 0.05). Psychiatric hospitalizations were significantly reduced after PGx testing in the incongruent-to-congruent and no change in congruency categories ( P < 0.001), but not in the congruent-to-incongruent category. Conversely, emergency department visits were significantly reduced after PGx testing in all congruency categories ( P < 0.005) and did not differ when compared across congruency categories. CONCLUSIONS:After PGx testing, patients with MDD had decreased prescribing of medications with significant GDI and reduced HRU. PGx testing may have influenced these outcomes, but the retrospective study design limits clarity on its impact.

Pharmacotherapy is one modality recommended to treat postpartum depression (PPD), but treatment patterns are not well characterized. In this study we characterized psychiatric medications used to treat PPD in real-world settings. Two cohorts of patients diagnosed with PPD within 180 days of delivery between October 2015 and January 2022 were retrospectively studied using two U.S. claims databases (Symphony Health [SH], Myriad Genetics-Komodo Health [MGKH]). Prescription fills of select psychiatric medications in the 365 days after PPD diagnosis were assessed using pharmacy claims. The two cohorts (SH, MGKH) included 124,742 and 22,141 patients with PPD, respectively. Most patients with PPD (SH: 64.9

OBJECTIVE:Outcomes in pregnancies with rare autosomal aneuploidies (RAAs) are poorly characterized, with most studies having small sample sizes. Here, we describe outcomes and management in a large cohort of pregnancies that screened positive for an RAA (RAA+). METHODS:Results of prenatal cell-free DNA screening were linked to de-identified insurance claims data. Diagnosis and procedure codes were used to estimate pregnancy outcomes and management. Relevant covariates in comparative analyses were adjusted using propensity-score matching. Outcomes were statistically compared using Mantel-Haenszel and McNemar's tests. RESULTS:Among 682 RAA+ pregnancies, the rate of live birth was significantly lower (56.7% vs. 78.7%; p < 0.001), and the rates of miscarriage and preterm birth were significantly higher (14.8% vs. 3.2%, p < 0.001; 18.5% vs. 8.9%, p < 0.001; respectively), compared to pregnancies with RAA- results. In pregnancies that screened positive for a rare autosomal trisomy (RAT+) and in which the RAT+ results were known, ultrasounds (mean: 3.7 vs. 2.5, p = 0.002), and pregnancy-specific visits (mean: 6.6 vs. 5.1; p = 0.007) were more frequent compared with pregnancies in which the RAT+ result was unknown. CONCLUSION:Pregnancies with RAA+ results had higher rates of adverse outcomes compared with those with RAA- results, and awareness of RAA+ results was associated with more intensive monitoring.

10527 Background: Guidelines recommend individuals with ≥20% lifetime risk of breast cancer (BC) undergo enhanced management, including annual screening mammography (SM) as early as age 30, annual breast MRI, and genetic counseling (GC). Lifetime BC risk can be estimated using a validated risk model, such as Tyrer-Cuzick (TC). A risk predictor that combines TC with a polygenic risk score (“combined risk score” or CRS) significantly improves risk prediction over TC alone. Little is known about how clinicians manage patients based on CRS. Here, we describe management following receipt of CRS results. Methods: De-identified administrative claims data from the Optum Labs Data Warehouse were linked with de-identified TC and CRS results originally provided to ordering clinicians. Patients were divided into four subgroups based on their lifetime risk predicted by CRS and by TC alone (high risk: ≥20%, average risk: <20%): A (CRS-high/TC-high), B (CRS-high/TC-avg), C (CRS-avg/TC-high), and D (CRS-avg/TC-avg). Patient management claims were compared 360 days before and after the CRS test date and differences evaluated with McNemar’s test. Differences in post-test management were analyzed using multivariable logistic regression with adjustment for clinical factors. Results: Before receiving risk results, 11.9%, 7.7%, 8.8%, and 5.0% of patients in A, B, C, and D received a SM before age 40, respectively; 3.4%, 2.6%, 2.3%, and 0.9% received an MRI, respectively; and 4.3%, 4.7%, 6.0%, and 3.6% received GC, respectively. After receiving risk results, SM in those under age 40 significantly increased 1.6-2.2-fold in A, B, and C; MRI significantly increased 4.7-5.6-fold in A, B, and C; and GC significantly increased 1.9-2.3-fold in A and B. SM, MRI, and GC did not increase in D. Prophylactic mastectomy occurred in only 0.17% of all patients after receiving results. After adjustment for confounders, those in A, B, and C were 3.8-5.2 times more likely to undergo SM; 11.6-23.1 times more likely to undergo MRI; and 2.0-2.9 times more likely to undergo GC, compared to those in Group D (all p<0.001). Conclusions: Patients with a ≥20% lifetime risk for BC were more likely to undergo enhanced management compared to those with a lifetime risk <20%, regardless of whether their risk was based on the CRS or on TC. These results suggest that clinicians recommended management aligned with guidelines for those with ≥20% lifetime risk, even when such risk was predicted by the CRS. [Table: see text]

Background:A breast cancer (BC) risk predictor that combines the Tyrer-Cuzick (TC) risk model with a polygenic risk score (“combined risk score” or CRS) has been shown to significantly improve risk prediction over TC alone. Guidelines recommend that individuals predicted to have ≥20% remaining lifetime risk of BC undergo enhanced management, including annual screening mammography (SM) as early as age 30, annual breast MRI, and genetic counseling (GC). However, little is known about whether and how clinicians manage patients based on risk predicted by CRS. Here, we describe the uptake of mammography, breast MRI, and GC following receipt of CRS results. Methods:De-identified administrative claims data from the Optum Labs Data Warehouse were linked with de-identified TC and CRS results originally provided to ordering clinicians between August 1, 2017 and November 30, 2021. Patients were included in the cohort if they had continuous medical and pharmacy enrollment for ≥360 days prior to and ≥360 days after the CRS result and were ≥18 years of age. Patients were excluded if they had a history of breast malignancy, BC, or any metastatic cancer during baseline; evidence of BC prevention measures (tamoxifen, raloxifene, aromatase inhibitors, pre-menopausal ovarian function suppression, or mastectomy) prior to receiving CRS results; or evidence of any cancer diagnosis in the ≥360 days after receiving CRS results. Patients were divided into four subgroups based on their lifetime risk predicted by CRS and by TC alone (high risk (+): ≥20%, average risk (-): <20%): CRS+ TC+, CRS+ TC-, CRS- TC+, and CRS- TC-. Patient management claims were assessed in the ≥360 days after the CRS test date. Results:The study cohort consisted of 8,662 patients: 2,443 (28.0%) CRS+ TC+, 696 (8.0%) CRS+ TC-, 856 (9.9%) CRS- TC+, and 4,687 (54.1%) CRS- TC-. Mean lifetime risk of breast cancer predicted by the CRS in each group was 29.9% (SD 7.9%), 24.5% (SD 4.3%), 16.4% (SD 2.6%), and 11.2% (SD 4.4%), respectively. Among the 3,309 patients with ≥20% lifetime risk predicted by TC (TC+), 856 (25.9%) had a lifetime risk <20% predicted by the CRS (CRS-). Conversely, among the 5,383 patients with lifetime risk <20% predicted by TC (TC-), 696 (12.9%) had a lifetime risk ≥20% predicted by the CRS (CRS+). Following receipt of CRS results, compared to the CRS- TC-group, the rate of SM in patients under age 40 years was 2.4, 2.2, and 1.9 times higher in the CRS+ TC+, CRS+ TC-, and CRS- TC+ groups, respectively; the rate of breast MRI was 13.8, 8.3, and 7.9 times higher in the CRS+ TC+, CRS+ TC-, and CRS- TC+ groups, respectively; and the rate of GC was 3.2, 2.9, and 2.4 times higher in the CRS+ TC+, CRS+ TC-, and CRS- TC+ groups, respectively. Conclusions:For a substantial proportion of patients, the CRS predicted a different risk level compared to TC, suggesting that these patients should be managed differently based on their CRS result. Patients with a ≥20% lifetime risk for BC were more likely to undergo enhanced management compared to those with a lifetime risk <20%, regardless of whether their risk was based on the CRS or on TC. These results suggest that clinicians recommended management aligned with guidelines for those with ≥20% lifetime risk, even when such risk was predicted by the CRS. Citation Format: Katie Johansen Taber, Sarah Ratzel, Elisha Hughes, Alexander Gutin, Jeff Jasper, D. Claire Miller, Devika Chawla, Brady DeHart, Laura Becker, Pamela Morin, Julia Certa, Allison Kurian. Association of polygenic-based breast cancer risk prediction with patient management [abstract]. In: Proceedings of the San Antonio Breast Cancer Symposium 2024; 2024 Dec 10-13; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2025;31(12 Suppl):Abstract nr RF1-06.

BackgroundMachine learning offers quantitative pattern recognition analysis of wearable device data and has the potential to detect illness onset and monitor influenza-like illness (ILI) in patients who are infected. ObjectiveThis study aims to evaluate the ability of machine-learning algorithms to distinguish between participants who are influenza positive and influenza negative in a cohort of symptomatic patients with ILI using wearable sensor (activity) data and self-reported symptom data during the latent and early symptomatic periods of ILI. MethodsThis prospective observational cohort study used the extreme gradient boosting (XGBoost) classifier to determine whether a participant was influenza positive or negative based on 3 models using symptom-only data, activity-only data, and combined symptom and activity data. Data were collected from the Home Testing of Respiratory Illness (HTRI) study and FluStudy2020, both conducted between December 2019 and October 2020. The model was developed using the FluStudy2020 data and tested on the HTRI data. Analyses included participants in these studies with an at-home influenza diagnostic test result. Fitbit (Google LLC) devices were used to measure participants’ steps, heart rate, and sleep parameters. Participants detailed their ILI symptoms, health care–seeking behaviors, and quality of life. Model performance was assessed by area under the curve (AUC), balanced accuracy, recall (sensitivity), specificity, precision (positive predictive value), negative predictive value, and weighted harmonic mean of precision and recall (F2) score. ResultsAn influenza diagnostic test result was available for 953 and 925 participants in HTRI and FluStudy2020, respectively, of whom 848 (89%) and 840 (90.8%) had activity data. For the training and validation sets, the highest performing model was trained on the combined symptom and activity data (training AUC=0.77; validation AUC=0.74) versus symptom-only (training AUC=0.73; validation AUC=0.72) and activity-only (training AUC=0.68; validation AUC=0.65) data. For the FluStudy2020 test set, the performance of the model trained on combined symptom and activity data was closely aligned with that of the symptom-only model (combined symptom and activity test AUC=0.74; symptom-only test AUC=0.74). These results were validated using independent HTRI data (combined symptom and activity evaluation AUC=0.75; symptom-only evaluation AUC=0.74). The top features guiding influenza detection were cough; mean resting heart rate during main sleep; fever; total minutes in bed for the combined model; and fever, cough, and sore throat for the symptom-only model. ConclusionsMachine-learning algorithms had moderate accuracy in detecting influenza, suggesting that previous findings from research-grade sensors tested in highly controlled experimental settings may not easily translate to scalable commercial-grade sensors. In the future, more advanced wearable sensors may improve their performance in the early detection and discrimination of viral respiratory infections.

22q11.2 deletion syndrome occurs in approximately 1 in 2,000-4,000 births. Prenatal cell-free DNA screening (pcfDNA) can detect fetuses affected by deletions as small as 2.5 Mb. Fetal fraction amplification (FFA), which has been shown to yield an average fetal fraction (FF) in excess of 20%, may further enhance pcfDNA detection of these deletions. Positive predictive values (PPV) of pcfDNA screening for 22q11.2 microdeletion have been reported between 20%-50%. Here, we sought to describe the impact of FFA on the PPV of 22q11.2 microdeletion screening using a whole-genome sequencing (WGS)-based pcfDNA platform. We retrospectively analyzed data from patients who underwent WGS-based pcfDNA screening with FFA (Prequel™, Myriad Genetics, Inc.) between 8/20-10/22. For screen-positive cases, pregnancy outcome data were requested via a routine HIPAA-compliant process. All samples with diagnostic confirmation were used to calculate PPV, defined as: true positives/(true positives + false positives). Confidence interval (CI) was estimated using the Exact Binomial Test. Complete outcomes were obtained for 54 cases that screened positive for 22q11.2 microdeletion; 21 underwent molecular diagnostic testing. All 21 were confirmed as true positives, for a PPV of 100% (21/21; 95% CI 83.9%-100%). Among the 33 cases that declined molecular confirmation, 17 had ultrasound findings that are either strongly or moderately associated with 22q11.2 deletion syndrome (e.g., cardiac defects, polyhydramnios, skeletal defects, intrauterine growth restriction). The average FF of true positive samples was 23.0%. For screen-positive patients who declined diagnostic testing, the average FF was 21.6%. The increased FF levels conferred by FFA yield a PPV for 22q11.2 microdeletion higher than any previous studies have reported, and comparable to that for common autosomal trisomies. Because respective FF levels were comparable in patients with and without diagnostic confirmation, PPV is likely to be comparable in patients with and without diagnostic confirmation.

Objective22q11.2 deletion syndrome (DS) is a serious condition with a range of features. The small microdeletion causing 22q11.2DS makes it technically challenging to detect using standard prenatal cfDNA screening. Here, we assess 22q11.2 microdeletion clinical performance by a prenatal cfDNA screen that incorporates fetal fraction (FF) amplification.MethodsThe study cohort consisted of patients who received Prequel (Myriad Genetics, Inc.), a prenatal cfDNA screening that incorporates FF amplification, and met additional eligibility criteria. Pregnancy outcomes were obtained via a routine process for continuous quality improvement. Samples with diagnostic testing results were used to calculate positive predictive value (PPV).Results379,428 patients met study eligibility criteria, 76 of whom were screen-positive for a de novo 22q11.2 microdeletion. 22 (29.7%) had diagnostic testing results available, and all 22 cases were confirmed as true positives, for a PPV of 100% (95% CI 84.6%-100%). This performance was based on cases that ranged broadly across FF (5.9%-41.1%, mean 23.0%), body mass index (22.3-44.8, mean 29.9), and gestational age at testing (10.0w-34.6w, median 12.7w). Ultrasound findings in screen-positive pregnancies were consistent with those known to be associated with 22q11.2DS.Conclusion22q11.2 microdeletion screening that incorporates FF amplification demonstrated high PPV across both general and high-risk population cohorts. What's already known about this topic?22q11.2 deletion syndrome is a serious condition characterized by congenital heart defects and several other abnormalities. The small size of the 22q11.2 microdeletion makes it technically challenging to detect using prenatal cell-free DNA.What does this study add?This study demonstrated a high positive predictive value (100%, CI 84.6%-100%) for 22q11.2 microdeletion using a prenatal cfDNA test that incorporates fetal fraction amplification. Performance was based on cases that broadly ranged across fetal fraction, BMI, GA, and indication for testing.

The burden of influenza-like illness (ILI) is typically estimated via hospitalizations and deaths. However, ILI-associated morbidity that does not require hospitalization remains poorly characterized. The main objective of this study was to characterize ILI burden using commercial wearable sensor data and investigate the extent to which these data correlate with self-reported illness severity and duration. Furthermore, we aimed to determine whether ILI-associated changes in wearable sensor data differed between care-seeking and non–care-seeking populations as well as between those with confirmed influenza infection and those with ILI symptoms only. This study comprised participants enrolled in either the FluStudy2020 or the Home Testing of Respiratory Illness (HTRI) study; both studies were similar in design and conducted between December 2019 and October 2020 in the United States. The participants self-reported ILI-related symptoms and health care–seeking behaviors via daily, biweekly, and monthly surveys. Wearable sensor data were recorded for 120 and 150 days for FluStudy2020 and HTRI, respectively. The following features were assessed: total daily steps, active time (time spent with >50 steps per minute), sleep duration, sleep efficiency, and resting heart rate. ILI-related changes in wearable sensor data were compared between the participants who sought health care and those who did not and between the participants who tested positive for influenza and those with symptoms only. Correlative analyses were performed between wearable sensor data and patient-reported outcomes. After combining the FluStudy2020 and HTRI data sets, the final ILI population comprised 2435 participants. Compared with healthy days (baseline), the participants with ILI exhibited significantly reduced total daily steps, active time, and sleep efficiency as well as increased sleep duration and resting heart rate. Deviations from baseline typically began before symptom onset and were greater in the participants who sought health care than in those who did not and greater in the participants who tested positive for influenza than in those with symptoms only. During an ILI event, changes in wearable sensor data consistently varied with those in patient-reported outcomes. Our results underscore the potential of wearable sensors to discriminate not only between individuals with and without influenza infections but also between care-seeking and non–care-seeking populations, which may have future application in health care resource planning. Clinicaltrials.gov NCT04245800; https://clinicaltrials.gov/ct2/show/NCT04245800

Background Previous research has estimated that >50% of individuals experiencing influenza-like illness (ILI) do not seek health care. Understanding factors influencing care-seeking behavior for viral respiratory infections may help inform policies to improve access to care and protect public health. We used person-generated health data (PGHD) to identify factors associated with seeking care for ILI. Methods Two observational studies (FluStudy2020, ISP) were conducted during the United States 2019-2020 influenza season. Participants self-reported ILI symptoms using the online Evidation platform. A log-binomial regression model was used to identify factors associated with seeking care. Results Of 1667 participants in FluStudy2020 and 47 480 participants in ISP eligible for analysis, 518 (31.1%) and 11 426 (24.1%), respectively, sought health care. Participants were mostly female (92.2% FluStudy2020, 80.6% ISP) and aged 18-49 years (89.6% FluStudy2020, 89.8% ISP). In FluStudy2020, factors associated with seeking care included having health insurance (risk ratio [RR], 2.14; 95% CI, 1.30-3.54), more severe respiratory symptoms (RR, 1.53; 95% CI, 1.37-1.71), and comorbidities (RR, 1.37; 95% CI, 1.20-1.58). In ISP, the strongest predictor of seeking care was high symptom number (RR for 6/7 symptoms, 2.14; 95% CI, 1.93-2.38). Conclusions Using PGHD, we confirmed low rates of health care-seeking behavior for ILI and show that having health insurance, comorbidities, and a high symptom burden were associated with seeking health care. Reducing barriers in access to care for viral respiratory infections may lead to better disease management and contribute to protecting public health.

Abstract Background Lost productivity from workplace absenteeism is a significant component of the economic burden of influenza [1]. Antiviral treatment together with influenza vaccination may help reduce work time lost from influenza illness [2]. In clinical trials, baloxavir, an oral single dose treatment for influenza, demonstrated comparable symptom resolution to oseltamivir taken twice daily for 5 days [3]. We examined baloxavir real-world outcomes using patient generated health data (PGHD) to investigate the association between antiviral use and workplace absenteeism. Methods Using data from a participatory influenza-like illness (ILI) surveillance program (ISP), we identified participants who self-reported ILI using the online Evidation platform during the 2019-2020 influenza season in the United States. Participants who self-reported treatment with baloxavir or oseltamivir were included. We conducted an ordinal logistic regression to estimate the odds of missing work due to ILI, while adjusting for use of over-the-counter (OTC) medication, age, number of symptoms, US region, and comorbidities. Results Of 3658 participants eligible for inclusion in the analysis, 3285 (89.8%) were prescribed oseltamivir and 373 (10.2%) were prescribed baloxavir. A majority of participants (81.7%) reported missing at least one day of work. In the ordinal logistic regression, use of baloxavir was associated with lower odds of absenteeism, compared to oseltamivir (odds ratio [OR]: 0.748, 95% confidence interval [CI]: 0.616, 0.907). In addition, a higher number of symptoms (OR 2.922, 95% CI: 2.454, 3.481), older age (OR: 1.425, 95% CI: 1.162, 1.747), and use of OTC medication (OR: 1.242, 95% CI: 1.086, 1.420) were associated with higher odds of absenteeism. Due to a high rate of missing data on the timing of antiviral use (24.5% missing), its association with absenteeism could not be adequately assessed. Conclusion Treatment of patients with ILI with single dose baloxavir was associated with reduced workplace absenteeism compared to oseltamivir after adjusting for measured confounders. This study provides useful insight into factors associated with ILI-related workplace absenteeism and the potential real-world utility of baloxavir. Disclosures Hao Xu, MSc, Hoffmann-La Roche Limited: Employee Vincent Ukachukwu, n/a, Roche Products Ltd: Employee Devika Chawla, PhD, Genentech: Employee.

OBJECTIVES To determine whether baloxavir use is associated with lower health care resource utilization (HCRU) and costs for secondary influenza complications post treatment compared with oseltamivir. STUDY DESIGN Retrospective cohort study. METHODS Patients filling a prescription for baloxavir or oseltamivir within 48 hours following an influenza-related outpatient visit were identified in the 2018-2019 influenza season from the US Truven MarketScan Research Databases and propensity matched 1:2 (baloxavir:oseltamivir). Outcomes were assessed 15 and 30 days after antiviral treatment and included all-cause, all respiratory-related, and select respiratory-related (influenza, asthma, chronic obstructive pulmonary disease, or infection) HCRU and costs. RESULTS The study included 5080 baloxavir-treated and 10,160 matched oseltamivir-treated patients. All-cause emergency department (ED) visits and inpatient hospitalizations were lower in baloxavir-treated patients, with a statistically significant difference in the percentage hospitalized at 30 days (0.3% vs 0.5%; P = .04). ED visits for all or select respiratory-related conditions were significantly reduced with baloxavir (P < .01 for all comparisons). Mean per-patient cost savings at day 30 for all-cause, all respiratory-related, and select respiratory-related conditions were $79, $50, and $51, respectively, despite slightly higher prescription costs for baloxavir. In high-risk patients (baloxavir: n = 1958; oseltamivir: n = 3949), the incidence of ED visits was significantly lower for all respiratory-related and select respiratory-related conditions (P < .01); cost savings with baloxavir in the high-risk cohort were substantially greater than in the overall cohort. CONCLUSIONS Treatment of patients with influenza with single-dose baloxavir was generally associated with lower HCRU and costs post treatment compared with oseltamivir, particularly in high-risk patients.

ObjectiveTo identify potential risk factors for adverse long-term outcomes (LTOs) associated with COVID-19, using a large electronic health record (EHR) database.DesignRetrospective cohort study. Patients with COVID-19 were assigned into subcohorts according to most intensive treatment setting experienced. Newly diagnosed conditions were classified as respiratory, cardiovascular or mental health LTOs at >30–≤90 or >90–≤180 days after COVID-19 diagnosis or hospital discharge. Multivariate regression analysis was performed to identify any association of treatment setting (as a proxy for disease severity) with LTO incidence.SettingOptum deidentified COVID-19 EHR dataset drawn from hospitals and clinics across the USA.ParticipantsIndividuals diagnosed with COVID-19 (N=57 748) from 20 February to 4 July 2020.Main outcomesIncidence of new clinical conditions after COVID-19 diagnosis or hospital discharge and the association of treatment setting (as a proxy for disease severity) with their risk of occurrence.ResultsPatients were assigned into one of six subcohorts: outpatient (n=22 788), emergency room (ER) with same-day COVID-19 diagnosis (n=11 633), ER with COVID-19 diagnosis≤21 days before ER visit (n=2877), hospitalisation without intensive care unit (ICU; n=16 653), ICU without ventilation (n=1837) and ICU with ventilation (n=1960). Respiratory LTOs were more common than cardiovascular or mental health LTOs across subcohorts and LTO incidence was higher in hospitalised versus non-hospitalised subcohorts. Patients with the most severe disease were at increased risk of respiratory (risk ratio (RR) 1.86, 95% CI 1.56 to 2.21), cardiovascular (RR 2.65, 95% CI 1.49 to 4.43) and mental health outcomes (RR 1.52, 95% CI 1.20 to 1.91) up to 6 months after hospital discharge compared with outpatients.ConclusionsPatients with severe COVID-19 had increased risk of new clinical conditions up to 6 months after hospital discharge. The extent that treatment setting (eg, ICU) contributed to these conditions is unknown, but strategies to prevent COVID-19 progression may nonetheless minimise their occurrence.

Abstract Background Over 32 million cases of COVID-19 have been reported in the US. Outcomes range from mild upper respiratory infection to hospitalization, acute respiratory failure, and death. We assessed risk factors associated with severe disease, defined as hospitalization within 21 days of diagnosis or death, using US electronic health records (EHR). Methods Patients in the Optum de-identified COVID-19 EHR database who were diagnosed with COVID-19 in 2020 were included in the analysis. Regularized multivariable logistic regression was used to identify risk factors for severe disease. Covariates included demographics, comorbidities, history of influenza vaccination, and calendar time. Results Of the 193,454 eligible patients, 36,043 (18.6%) were hospitalized within 21 days of COVID-19 diagnosis, and 6,397 (3.3%) died. Calendar time followed an inverse J-shaped relationship where severe disease rates rapidly declined in the first 25 weeks of the pandemic. BMI followed an asymmetric V-shaped relationship with highest rates of disease severity observed at the extremes. In the multivariable model, older age had the strongest association with disease severity (odds ratios and 95% confidence intervals of significant associations in Figure). Other risk factors were male sex, uninsured status, underweight and obese BMI, higher Charlson Comorbidity Index, and individual comorbidities including hypertension. Asthma and overweight BMI were not associated with disease severity. Blacks, Hispanics, and Asians experienced higher odds of disease severity compared to Whites. Figure. Significant associations (odds ratio and 95% confidence intervals) with COVID-19 severity (hospitalization or death), adjusted for geographical division. Reference and abbreviation categories: Charlson comorbidity index (CCI) = 0; Age = 18-30; Sex = Female; Race/Ethnicity = White; Insurance = Commercial; Body mass index (BMI) = 18.5-25; Calendar time = 0-25 weeks; Chronic obstructive pulmonary disease (COPD). Conclusion Odds of hospitalization or death have decreased since the start of the pandemic, with the steepest decline observed up to mid-August, possibly reflecting changes in both testing and treatment. Older age is the most important predictor of severe COVID-19. Obese and underweight, but not overweight, BMI were associated with increased odds of disease severity when compared to normal weight. Hypertension, despite not being included in many guidelines for vaccine prioritization, is a significant risk factor. Pronounced health disparities remain across race and ethnicity after accounting for comorbidities, with minorities experiencing higher disease severity. Disclosures Shemra Rizzo, PhD, F. Hoffmann-La Roche Ltd. (Shareholder)Genentech, Inc. (Employee) Ryan Gan, PhD, F. Hoffmann-La Roche Ltd (Shareholder)Genentech, Inc. (Employee) Devika Chawla, PhD MSPH, F. Hoffmann-La Roche Ltd. (Shareholder)Genentech, Inc. (Employee) Kelly Zalocusky, PhD, F. Hoffmann-La Roche Ltd. (Shareholder)Genentech, Inc. (Employee) Xin Chen, PhD, F. Hoffmann-La Roche Ltd. (Shareholder)Genentech, Inc. (Employee) Yifeng Chia, PhD, F. Hoffmann-La Roche Ltd (Shareholder)Genentech, Inc. (Employee)

Abstract Background While COVID-19 carries substantial morbidity and mortality, the extent of long-term complications remains unclear. Reports suggest that acute lung damage associated with severe COVID-19 can result in chronic respiratory dysfunction. This study: (1) estimated the incidence of dyspnea and ILD after COVID-19 hospitalization, and (2) assessed risk factors for developing dyspnea and ILD in a real-world cohort of patients hospitalized with COVID-19 using US electronic health records (EHR). Methods Patients in the Optum de-identified COVID-19 EHR database who were hospitalized for COVID-19 (lab confirmed or diagnosis code) between February 20 and July 2020 and had at least 6 months of follow-up were eligible for analysis. Dyspnea and ILD were identified using diagnosis codes. The effects of baseline characteristics and hospitalization factors on the risk of incident dyspnea or ILD 3 to 6 months’ post discharge were evaluated. Results Among eligible patients (n=26,339), 1705 (6.5%) had dyspnea and 220 (0.8%) had ILD 3 to 6 months after discharge. Among patients without prior dyspnea or ILD (n=22,613), 110 (0.5%) had incident ILD (Table 1) and 1036 (4.6%) had incident dyspnea (Table 2) 3 to 6 months after discharge. In multivariate analyses, median (IQR) length of stay (LOS; 5.0 [3.0, 9.0] days in patients who did not develop ILD vs 14.5 [6.0, 26.0] days in patients who developed ILD; RR: 1.12, 95% CI: 1.08, 1.15; P=4.34 x 10-10) and age (RR: 1.02, 95% CI: 1.01, 1.03; P=4.63 x 10-3) were significantly associated with ILD. Median (IQR) LOS (5.0 [3.0, 9.0] days in patients who did not develop dyspnea vs 7 [4.0, 14.0] days in patients who developed dyspnea; RR: 1.04, 95% CI: 1.02, 1.06; P=8.52 x 10-4), number of high-risk comorbidities (RR: 1.18, 95% CI: 1.12, 1.24; P=3.85 x 10-9), and obesity (RR: 1.52, 95% CI: 1.25, 1.86; P=2.59 x 10-4) were significantly associated with dyspnea. Table 1. Selected Baseline Risk Factors for Incident ILD Table 2. Selected Baseline Risk Factors for Incident Dyspnea Conclusion In a real-world cohort, 4.6% and 0.5% of patients developed dyspnea and ILD, respectively, after COVID-19 hospitalization. Multivariate analyses suggested that LOS, age, obesity, and comorbidity burden may be risk factors for post-COVID-19 respiratory complications. Limitations included sensitivity of diagnosis codes, availability of labs, and care-seeking bias. Disclosures Kelly Zalocusky, PhD, F. Hoffmann-La Roche Ltd (Shareholder)Genentech, Inc. (Employee) Devika Chawla, PhD MSPH, F. Hoffmann-La Roche Ltd. (Shareholder)Genentech, Inc. (Employee) Margaret Neighbors, PhD, F. Hoffmann-La Roche Ltd (Shareholder)Genentech, Inc. (Employee) Shemra Rizzo, PhD, F. Hoffmann-La Roche Ltd. (Shareholder)Genentech, Inc. (Employee) Larry Tsai, MD, F. Hoffmann-La Roche Ltd (Shareholder)Genentech, Inc. (Employee)