A simulation-based bias analysis to assess the impact of unmeasured confounding when designing non-randomized database studies.

BACKGROUND: Unmeasured confounding is often raised as a source of potential bias during the design of non-randomized studies but quantifying such concerns is challenging. METHODS: We developed a simulation-based approach to assess the potential impact of unmeasured confounding during the study design stage. The approach involved generation of hypothetical individual-level cohorts using realistic parameters including a binary treatment (prevalence 25%), a time-to-event outcome (incidence 5%), 13 measured covariates, a binary unmeasured confounder (u1, 10%), and a binary measured 'proxy' variable (p1) correlated with u1. Strength of unmeasured confounding and correlations between u1 and p1 were varied in simulation scenarios. Treatment effects were estimated with, a) no adjustment, b) adjustment for measured confounders (Level 1), c) adjustment for measured confounders and their proxy (Level 2). We computed absolute standardized mean differences in u1 and p1 and relative bias with each level of adjustment. RESULTS: Across all scenarios, Level 2 adjustment led to improvement in balance of u1, but this improvement was highly dependent on the correlation between u1 and p1. Level 2 adjustments also had lower relative bias than Level 1 adjustments (in strong u1 scenarios: relative bias of 9.2%, 12.2%, 13.5% at correlations 0.7, 0.5, and 0.3, respectively versus 16.4%, 15.8%, 15.0% for Level 1, respectively). CONCLUSION: An approach using simulated individual-level data was useful to explicitly convey the potential for bias due to unmeasured confounding while designing non-randomized studies and can be helpful in informing design choices.

A collaborative study of the impact of N-nitrosamines presence and ARB recall on ARB utilization - results from IQVIA™ Disease Analyzer Germany.

PURPOSE: Regulators are increasingly concerned with  the impact of recalls on drug adherence. In 2018, N-nitrosamines impurities were detected in valsartan containing medical products. Concerned products were immediately recalled in July 2018 by regulatory agencies internationally. In Germany, recalls were issued for valsartan, losartan and irbesartan from July 2018 to March 2019. This study examined angiotensin II receptor blocker (ARB) utilization trends and switching patterns in Germany before and after July 2018. METHODS: Patients prescribed ARBs from January 2014 to June 2020 in general practices in Germany were included in a collaborative framework common protocol drug utilization study led by the US Food and Drug Administration. Trends in monthly and quarterly proportions of total ARB prescribing were analysed for individual ARBs using descriptive statistics and interrupted time series analysis. The rate of switching to an alternative ARB was analysed before and after the recalls. RESULTS: The proportion of valsartan prescriptions immediately decreased from 35.9 to 17.8% following the first recalls in July 2018, mirrored by an increased proportion for candesartan. Increased switching from valsartan to candesartan was observed. No increased switching was observed after losartan recalls, whereas for irbesartan, increased switching was observed 6-12 months after the last recall. Increased switching from ARBs to angiotensin-converting enzyme (ACE) inhibitors or ARB treatment discontinuations were not observed. CONCLUSION: This study showed that patients were able to continue ARB treatment despite the July 2018-March 2019 recalls, although many patients needed to switch to an alternative ARB. The duration of the impact of ARB recalls appeared to be limited.

Assessing the risk of intentional self-harm in montelukast users: an updated Sentinel System analysis using ICD-10 coding.

BACKGROUND: Montelukast prescribing information includes a Boxed Warning issued in March 2020 regarding neuropsychiatric adverse events. A previous Sentinel System study of asthma patients from 2000 to 2015 did not demonstrate an increased risk of intentional self-harm measured using the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes, with montelukast compared to inhaled corticosteroids (ICS). METHODS: Using a new user cohort study design, we examined intentional self-harm events in patients aged 10 years and older who were incident users of either montelukast or ICS as monotherapy, with a diagnosis of asthma, between October 1, 2015, to June 30, 2022, in the Sentinel System. We measured intentional self-harm using ICD-10-CM codes, which may have better accuracy for capturing suicide attempts than ICD-9-CM codes. We used inverse probability of treatment weighting to balance baseline covariates. We performed subgroup analyses by age group, sex, psychiatric history, and pre/post Boxed Warning era and conducted sensitivity analyses varying type of care setting of the outcome and exposure episode gaps. RESULTS: Among 752,230 and 724,855 patients in the montelukast and ICS exposure groups respectively, we found no association between montelukast use and self-harm compared to ICS use [Hazard Ratio (95% Confidence Interval): 0.96 (0.85, 1.08)]. This finding was consistent across all subgroups, and sensitivity analyses. CONCLUSION: Our results cannot exclude other neuropsychiatric idiosyncratic reactions to montelukast. Compared to the previous Sentinel study, this study identified about double the rate of self-harm events, suggesting a greater sensitivity of ICD-10 codes for measuring self-harm than ICD-9.

Social media for early characterization of pandemic symptoms: A qualitative analysis of patient-reported COVID-19 experiences.

BACKGROUND: Patients use social media forums to discuss their medical history and healthcare experiences, providing early insight into real-world patient experiences. We analyzed COVID-19 patient experiences from Reddit social media posts. METHODS: We extracted Reddit Application Programming Interface data for the subreddit/COVID-19 positive from March to August 2020 and selected users tagged as "Tested Positive" or "Tested Positive- Me" flair and who posted at least thirty times in any calendar month, excluding users who explicitly stated location outside of the U.S. For tested-positive patients (users), we created and reviewed individual case profiles summarizing their COVID-19 symptoms, testing, and medications or treatments. Data were imported to Nvivo qualitative analysis software and qualitative coding was conducted. FINDING: There were 31 759 posts and comments from 720 users in March to May 2020 (Q1) and 40 446 posts and comments from 1649 users from June to August 2020 (Q2). Final count of "Tested Positive" was 1296 users (280 in Q1 and 1016 in Q2). Across both quarters, frequently reported symptoms included sore throat, headaches, fevers, or chills. Loss of sense of smell or taste were reported by users in early March, prior to the inclusion of this symptom to the CDC list in April and GI-related symptoms and fatigue were reported in the March to May data, before they were added as a COVID-19 associated symptom in July 2020. Users also reported in-depth descriptions of their symptoms, motivations for testing, and long-term impacts such as post-viral fatigue. INTERPRETATION: Social media data can potentially serve as an early surveillance data source in a pandemic and offer preliminary insights into patient disease experiences.

Biosimilars approvals by thirteen regulatory authorities: A cross-national comparison.

Biosimilars are biological medicines highly similar to a previously licensed reference product and their licensing is expected to improve access to biological therapies. This study aims to present an overview of biosimilars approval by thirteen regulatory authorities (RA). The study is a cross-national comparison of regulatory decisions involving biosimilars in Argentina, Australia, Brazil, Chile, Canada, Colombia, Europe, Hungary, Guatemala, Italy, Mexico, Peru and United States. We examined publicly available documents containing information regarding the approval of biosimilars and investigated the publication of public assessment reports for registration applications, guidelines for biosimilars licensing, and products approved. Data extraction was conducted by a network of researchers and regulatory experts. All the RA had issued guidance documents establishing the requirements for the licensing of biosimilars. However, only three RA had published public assessment reports for registration applications. In total, the investigated jurisdictions had from 19 to 78 biosimilars approved, most of them licensed from 2018 to 2020. In spite of the advance in the number of products in recent years, some challenges still persist. Limited access to information regarding the assessment of biosimilars by RA can affect confidence, which may ultimately impact adoption of these products in practice.

A cohort study to assess risk of cutaneous small vessel vasculitis among users of different oral anticoagulants.

PURPOSE: Cutaneous small vessel vasculitis (CSVV) was identified as a safety signal among patients treated with direct oral anticoagulants (DOAC). This study aimed to determine if CSVV risk differed among patients with atrial fibrillation (Afib) who newly initiated warfarin or a DOAC. METHODS: We identified enrollees aged ≥21 years diagnosed with Afib who newly initiated rivaroxaban, dabigatran, apixaban, and warfarin in the Sentinel Distributed Database from October 19, 2010 to February 29, 2020. We selected and followed patients who did not have evidence of the following in the 183 days prior to initiating treatment: CSVV diagnosis, dispensing of other study drugs, select autoimmune diseases or autoimmune medications, cancer diagnoses or chemotherapeutic treatment, kidney dialysis or transplant, alternative anticoagulation indications, or an institutional (nursing home, hospice, hospital) stay on the treatment initiation date (index date) until CSVV outcome or pre-specified censoring. We conducted 1:1 propensity score matching in six comparisons. RESULTS: CSVV incidence rates for DOACs and warfarin ranged from 3.3 to 5.6 per 10 000-person years in our matched Afib population. The adjusted CSVV hazard ratio (HR) and 95% confidence interval (CI) was 0.94 (0.64, 1.39) for rivaroxaban versus warfarin; 1.17 (0.67, 2.06) for dabigatran vs. warfarin; 0.85 (0.62, 1.16) for apixaban vs. warfarin; 0.86 (0.49, 1.50) for rivaroxaban vs. dabigatran; 0.99 (0.68, 1.45) for rivaroxaban versus apixaban; and 1.70 (0.90, 3.21) for dabigatran versus apixaban. CONCLUSION: We did not find significant evidence of differential CSVV risk in pair-wise comparisons of DOACs and warfarin.

A General Propensity Score for Signal Identification Using Tree-Based Scan Statistics.

The tree-based scan statistic (TreeScan; Martin Kulldorff, Harvard Medical School, Boston, Massachusetts) is a data-mining method that adjusts for multiple testing of correlated hypotheses when screening thousands of potential adverse events for signal identification. Simulation has demonstrated the promise of TreeScan with a propensity score (PS)-matched cohort design. However, it is unclear which variables to include in a PS for applied signal identification studies to simultaneously adjust for confounding across potential outcomes. We selected 4 pairs of medications with well-understood safety profiles. For each pair, we evaluated 5 candidate PSs with different combinations of 1) predefined general covariates (comorbidity, frailty, utilization), 2) empirically selected (data-driven) covariates, and 3) covariates tailored to the drug pair. For each pair, statistical alerting patterns were similar with alternative PSs (≤11 alerts in 7,996 outcomes scanned). Inclusion of covariates tailored to exposure did not appreciably affect screening results. Inclusion of empirically selected covariates can provide better proxy coverage for confounders but can also decrease statistical power. Unlike tailored covariates, empirical and predefined general covariates can be applied "out of the box" for signal identification. The choice of PS depends on the level of concern about residual confounding versus loss of power. Potential signals should be followed by pharmacoepidemiologic assessment where confounding control is tailored to the specific outcome(s) under investigation.

Risk of Covid-19-Related Hospitalization and More Severe Outcomes in Medicare Beneficiaries Treated with Renin-Angiotensin-Aldosterone System Inhibitors for Hypertension.

BACKGROUND: There are theoretical concerns that angiotensin-converting enzyme inhibitors (ACEIs) and angiotensin receptor blockers (ARBs) could increase the risk of severe Covid-19. OBJECTIVE: To determine if ACEIs and ARBs are associated with an increased risk of Covid-19 hospitalization overall, or hospitalization involving intensive care unit (ICU) admission, invasive mechanical ventilation, or death. DESIGN: Observational case-control study. PARTICIPANTS: Medicare beneficiaries aged ≥ 66 years with hypertension, treated with ACEIs, ARBs, calcium channel blockers (CCBs), or thiazide diuretics. MAIN MEASURES: Adjusted odds ratios (OR) and 95% confidence intervals (CI) for the outcomes of Covid-19 hospitalization, or hospitalization involving ICU admission, invasive mechanical ventilation, or death. RESULTS: A total of 35,300 cases of hospitalized Covid-19 were matched to 228,228 controls on calendar date and neighborhood of residence. The median age of cases was 79 years, 57.4% were female, and the median duration of hospitalization was 8 days (interquartile range 5-12). ACEIs and ARBs were associated with a slight reduction in Covid-19 hospitalization risk compared with treatment with other first-line antihypertensives (OR for ACEIs 0.95, 95% CI 0.92-0.98; OR for ARBs 0.94, 95% CI 0.90-0.97). Similar results were obtained for hospitalizations involving ICU admission, invasive mechanical ventilation, or death. There were no meaningful differences in risk for ACEIs compared with ARBs. In an analysis restricted to monotherapy with a first-line agent, CCBs were associated with a small increased risk of Covid-19 hospitalization compared with ACEIs (OR 1.09, 95% CI 1.04-1.14), ARBs (OR 1.10, 95% CI 1.05-1.15), or thiazide diuretics (OR 1.11, 95% CI 1.03-1.19). CONCLUSIONS: ACEIs and ARBs were not associated with an increased risk of Covid-19 hospitalization or with hospitalization involving ICU admission, invasive mechanical ventilation, or death. The finding of a small increased risk of Covid-19 hospitalization with CCBs was unexpected and could be due to residual confounding.

Risk of Stroke and Bleeding in Atrial Fibrillation Treated with Apixaban Compared with Warfarin.

BACKGROUND: A previous FDA study reported a favorable benefit risk for apixaban compared with warfarin for stroke prevention in older non-valvular atrial fibrillation (NVAF) patients (≥ 65 years). However, it remains unclear whether this favorable benefit risk persists in other populations including younger users. We examined if a similar benefit risk was observed in the Sentinel System and if it varied by age group. OBJECTIVE: To examine the risk of ischemic stroke, gastrointestinal (GI) bleeding, and intracranial hemorrhage (ICH) in apixaban users compared with warfarin users in Sentinel Distributed Database (SDD). DESIGN AND PARTICIPANTS: A retrospective new user cohort study was conducted among patients, 21 years and older initiating apixaban and warfarin for NVAF, between December 28, 2012, and June 30, 2018, in the SDD. MAIN MEASURES: Cox proportional hazard regression was used to estimate the hazard ratios (HR) and 95% confidence intervals (95% CI) for each outcome (ischemic stroke, GI bleeding, and ICH) in propensity score matched apixaban users compared with the warfarin users. Subgroup analyses by age (21-64, 65-74, and 75+ years) were conducted. KEY RESULTS: After matching, 55.3% and 58.4% (n = 55,038) of the apixaban and warfarin users were included in the main analysis. GI bleeding was the most common outcome. The HR (95% CI) for GI bleeding, ICH, and ischemic stroke in apixaban users compared with warfarin users were 0.57 (0.50-0.66), 0.53 (0.40-0.70), and 0.56 (0.45-0.71) respectively. The reduced risk of these outcomes in apixaban compared with warfarin users persisted across age groups. CONCLUSION: In NVAF patients of all ages initiating either apixaban or warfarin for stroke prevention in the Sentinel System, apixaban was associated with a decreased risk of GI bleeding, ICH, and ischemic stroke compared with warfarin. Among patients less than 65 years of age, apixaban use was associated with a decreased risk of GI bleeding and ischemic stroke.

Understanding utilization patterns of biologics and biosimilars in the United States to support postmarketing studies of safety and effectiveness.

PURPOSE: To describe utilization of filgrastim and infliximab, the first two products with biosimilars approved in the United States. METHODS: We identified use of filgrastim (reference, tbo-filgrastim, and filgrastim-sndz) and infliximab (reference, infliximab-dyyb, and infliximab-abda) in the Sentinel Distributed Database using Healthcare Common Procedure Coding System (HCPCS) codes and National Drug Codes (NDCs) from January 2015 to August 2018. We calculated the proportion of use by code type and assessed uptake over time. We compared baseline patient characteristics and treatment indications. Among patients with >1 exposure episode, we characterized gaps between episodes. RESULTS: Use was identified primarily via HCPCS codes (filgrastim: 86.4%-97.7%; infliximab: 87.8%-100%) although some was identified via NDCs (filgrastim: 2.2%-13.5%; infliximab: <0.1%-6.5%). Filgrastim reference product use declined from 89.4% in January 2015 to 30.3% in June 2018, with corresponding increases in filgrastim-sndz (0% to 49.3%) and tbo-filgrastim (10.6% to 20.4%). Infliximab biosimilar uptake was low (9.7% in June 2018). We identified 94 846 filgrastim reference product, 27 143 tbo-filgrastim, and 38 264 filgrastim-sndz users. For infliximab, we identified 125 412 reference product, 1034 infliximab-dyyb, 49 infliximab-abda, and 4855 undetermined biosimilar users. Patients receiving filgrastim products were largely similar, but differences in age, sex, and indication were observed across infliximab product users. The median exposure episode gap ranged from 1 to 3 days for filgrastim and 48 to 50 days for infliximab. CONCLUSION: Use of biosimilar filgrastim has increased in the United States, but infliximab biosimilar use remains low. Data on identification of biosimilars in claims data and observed gaps between exposure episodes can be used to support drug safety studies of biosimilars.

Use of tumor necrosis factor-alpha inhibitors during pregnancy among women who delivered live born infants.

PURPOSE: To describe the use of tumor necrosis factor-alpha inhibitors (TNFis) among pregnancies ending in a live birth and with a diagnosis of ankylosing spondylitis (AS), Crohn's disease (CD), juvenile idiopathic arthritis (JIA), psoriasis (PsO), psoriatic arthritis (PsA), rheumatoid arthritis (RA), or ulcerative colitis (UC). METHODS: We identified pregnancies among women aged 15 to 54 years between 01/01/2004 and 09/30/2015 from 16 health plans participating in Sentinel. We inferred indication using ICD-9-CM codes in the 183-day period before conception. We assessed proportion of infliximab, etanercept, adalimumab, certolizumab pegol, and golimumab by calendar year, indication, and maternal age, and compared them to proportions in an age-matched, indication-matched, and date-matched non-pregnant cohort. RESULTS: Among 19 681 pregnancies with at least one chronic inflammatory condition, 2990 (15.2%) received a TNFi. In both pregnancies and matched non-pregnant cohort, TNFi use was highest (34.4%; 55.8%) for PsA patients and lowest (6.2%; 13.4%) for PsO patients. Etanercept was most frequently used among AS/JIA/PsA/PsO/RA patients, while infliximab was the preferred TNFi for CD/UC patients. Except for infliximab and certolizumab, TNFi use during pregnancy decreased after the first trimester. Pregnancies among older pregnant women (45-54 years) were more likely to be treated compared with the matched non-pregnant cohort. CONCLUSION: There was a preference for etanercept among pregnancies with AS/JIA/PsA/PsO/RA, despite the availability of other TNFis. Decline in TNFi use after the first trimester may be related to the desire to reduce TNFis transplacental transfer and to minimize infection risk to the fetus or baby associated with live vaccine immunizations after birth.

Development of an algorithm to detect methotrexate wrong frequency error using computerized health care data.

PURPOSE: We validated an algorithm to detect frequency errors in computerized healthcare data and estimated the incidence of these errors in an integrated healthcare system. METHODS: We applied Sentinel System analytic tools on the electronic health records of Kaiser Permanente, Northern California, January 1, 2010, through May 30, 2015,to identify rheumatoid arthritis (RA) patients with new use of methotrexate (365-day baseline period). We identified potential methotrexate frequency errors using ICD-9 code 995.20 (adverse drug event), Current Procedural Terminology (CPT) code 96409 for injection of leucovorin and prescription refill patterns. We performed chart review to confirm the frequency errors, assessed performance for detecting frequency errors, and estimated the incidence of chart-confirmed errors. RESULTS: The study included 24,529 methotrexate dispensings among 3,668 RA patients. Among these, 722 (3%) had one dispensing and 23,807 (97.1%) had ≥2 dispensings during 1-year follow-up period. We flagged 653 (2.7%) with a potential medication error (46 with one dispensing and 607 with ≥2 dispensings). We sampled 94 for chart review, and confirmed three methotrexate errors. All three confirmed frequency errors involved a first methotrexate dispensing followed by injected rescue therapy, leucovorin, (positive predictive value, 60%; 95% confidence interval [CI], 15-95%). No potential errors were found among patients with ≥2 dispensings. We estimated the frequency error incidence among one methotrexate dispensing to be 0.4% (95%CI, 0.1% to 1.2%). CONCLUSION: Rescue therapy is a specific indicator of methotrexate overdose among first methotrexate dispensings. This method is generalizable to other medications with serious adverse events treated with antidotes.

The impact of the boxed warning on the duration of use for depot medroxprogesterone acetate.

OBJECTIVE: The objective of this study was to examine the impact of the Food and Drug Administration's boxed warning on the utilization of depot medroxyprogesterone (DMPA). METHODS: From the IMS Lifelink data (2001-2009), we identified DMPA and oral combined hormonal contraceptive (CHC) users without a prescription claim 6 months before and after the first and last claim. Episodes were defined as all contiguous claims with no more than 90-day DMPA or 30-day CHC between claims. Days' supply (CHC) and 90-day duration (DMPA) was used to determine episodes. We used interrupted time series to evaluate changes in the mean episode length and proportion of episodes >2 years before and after the Food and Drug Administration's November 2004 boxed warning. Stratified analyses by birth cohort were conducted. RESULTS: From 2001 to 2009, 126 528 DMPA and 651 356 CHC episodes were used for segmented regression. For the DMPA cohort, there was an immediate decline in the mean duration (-34.7 days [confidence interval: -45.4 to -24.1]) and episodes >2 years (-1.9% [confidence interval: -2.9% to -1.1%]) after the boxed warning. We did not observe any change in mean duration or episodes >2 years for the CHC cohort. The largest declines in mean duration and proportion >2 years were seen with the oldest women. CONCLUSION: We observed a modest decline in the mean duration and episodes >2 years for DMPA use immediately after the boxed warning not observed among CHC users. In the stratified analysis, we saw declines in the duration of use >2 years in all age groups, except adolescents who continue to use DMPA for longer than 2 years. Published 2017. This article is a U.S. Government work and is in the public domain in the USA.

Are Discrepancies in Waiting Time for Chest Pain at Emergency Departments between African Americans and Whites Improving Over Time?

BACKGROUND: One of the Healthy People 2010 goals was to eliminate racial disparities in the U.S health system. To date, we have limited knowledge about the impact of Healthy People on racial disparities at emergency departments (EDs). OBJECTIVE: We sought to investigate whether there has been an improvement in ED waiting time to see a physician for African Americans (AAs) compared to whites with chest pain symptoms that suggest acute coronary syndrome (ACS). METHODS: A retrospective analysis of the National Hospital and Ambulatory Care Survey data from 2004 to 2011 was conducted in adults with visits related to ACS. We compared covariate-adjusted odds ratios for race for each study year and 2011. In addition, adjusted average differences in waiting times (i.e., time to see a physician) for AAs and whites for each study year were compared. RESULTS: A total of 15,438 visits related to ACS symptoms were made during the study period. The waiting time for AAs (median, 33 min) was statistically longer compared to whites (median, 21 min). In addition, the adjusted waiting time for AAs was 30% longer compared to whites (95% confidence interval, 24-36%). Pairwise comparison of adjusted odds ratios between the year 2011 and other years was not significantly different (all p values = 0.32), suggesting no change in the difference in waiting times during the study period. CONCLUSION: Among patients presenting to the ED with symptoms suggesting ACS, AA compared to whites waited longer to receive care. In addition, this difference in waiting time persisted during the study period, even after the implementation of the Healthy People 2010 initiative. Additional research is warranted to investigate the underlying reasons for unequal care offered to AAs at EDs and the implications on disease outcome.

Clinical and Sociodemographic Factors Associated With Attention-Deficit/Hyperactivity Disorder in Patients With Cystic Fibrosis.

BACKGROUND: There is scarce evidence on the epidemiology of attention-deficit/hyperactivity disorder (ADHD) in patients with cystic fibrosis (CF). OBJECTIVE: We employed stepwise logistic regression to examine the association between ADHD diagnosis and selected patient characteristics. METHODS: This was a cross-sectional analysis of inpatient and outpatient billing data for Medicaid beneficiaries with CF ages 3-18 years to obtain ADHD diagnosis prevalence and incidence estimates from 1999-2006. RESULTS: Annual ADHD prevalence increased 1.55-fold from 5.26% (95% CI: 5.25-5.27) to 8.16% (8.15-8.17), and annual ADHD incidence rose slightly from 1.70% (1.70-1.71) to 2.01% (2.00-2.01). As in the general population, males were significantly more likely to have a diagnosis of ADHD compared with females (odds ratio: 1.97 [CI: 1.49-2.60]), as were children with recent diagnoses of anxiety, emotional disorder, depression, adjustment disorder, and learning, motor, and communication disorders. Patients with ADHD diagnoses were also more likely to be in foster care (odds ratio = 4.36 [CI: 2.26-8.40]). Except for recent DNase use (odds ratio = 0.64 [CI: 0.43-0.93]), CF severity indicators and medications including pancreatic enzymes, inhaled tobramycin, inhaled or oral corticosteroids, inhaled bronchodilators, and oral antibiotics had no association with ADHD diagnosis. CONCLUSION: ADHD prevalence in CF increased during the study period. Clinical and sociodemographic determinants of ADHD diagnosis were similar to the general population, whereas treatment and severity of CF appeared to have little influence. Our findings warrant future research evaluating diagnostic protocols and assessment of safety and efficacy of ADHD treatment in children with CF.

Identifying COVID-19 cases and extracting patient reported symptoms from Reddit using natural language processing.

We used social media data from "covid19positive" subreddit, from 03/2020 to 03/2022 to identify COVID-19 cases and extract their reported symptoms automatically using natural language processing (NLP). We trained a Bidirectional Encoder Representations from Transformers classification model with chunking to identify COVID-19 cases; also, we developed a novel QuadArm model, which incorporates Question-answering, dual-corpus expansion, Adaptive rotation clustering, and mapping, to extract symptoms. Our classification model achieved a 91.2% accuracy for the early period (03/2020-05/2020) and was applied to the Delta (07/2021-09/2021) and Omicron (12/2021-03/2022) periods for case identification. We identified 310, 8794, and 12,094 COVID-positive authors in the three periods, respectively. The top five common symptoms extracted in the early period were coughing (57%), fever (55%), loss of sense of smell (41%), headache (40%), and sore throat (40%). During the Delta period, these symptoms remained as the top five symptoms with percent authors reporting symptoms reduced to half or fewer than the early period. During the Omicron period, loss of sense of smell was reported less while sore throat was reported more. Our study demonstrated that NLP can be used to identify COVID-19 cases accurately and extracted symptoms efficiently.