Generalisable long COVID subtypes: Findings from the NIH N3C and RECOVER programmes

Background Stratification of patients with post-acute sequelae of SARS-CoV-2 infection (PASC, or long COVID) would allow precision clinical management strategies. However, long COVID is incompletely understood and characterised by a wide range of manifestations that are difficult to analyse computationally. Additionally, the generalisability of machine learning classification of COVID-19 clinical outcomes has rarely been tested.Methods We present a method for computationally modelling PASC phenotype data based on electronic healthcare records (EHRs) and for assessing pairwise phenotypic similarity between patients using semantic similarity. Our approach defines a nonlinear similarity function that maps from a feature space of phenotypic abnormalities to a matrix of pairwise patient similarity that can be clustered using unsupervised machine learning.Findings We found six clusters of PASC patients, each with distinct profiles of phenotypic abnormalities, including clusters with distinct pulmonary, neuropsychiatric, and cardiovascular abnormalities, and a cluster associated with broad, severe manifestations and increased mortality. There was significant association of cluster membership with a range of pre-existing conditions and measures of severity during acute COVID-19. We assigned new patients from other healthcare centres to clusters by maximum semantic similarity to the original patients, and showed that the clusters were generalisable across different hospital systems. The increased mortality rate originally identified in one cluster was consistently observed in patients assigned to that cluster in other hospital systems. Interpretation Semantic phenotypic clustering provides a foundation for assigning patients to stratified subgroups for natural history or therapy studies on PASC.

Applying the National Standards of Perioperative Care of Older People Undergoing Emergency Laparotomy (Nela)

Introduction There are well documented in-equalities for outcomes for surgical intervention associated with Age and Frailty including emergency laparotomy. NELA data has shown over half of such patients are over 65 years old about one fifth are over 80. These patients having significantly higher mortality, longer hospital stays and it has also shown frailty to be an independent marker of poor outcomes. Through application of key standards these outcomes have improved however input from "consultant geriatrician-led MDT" remains stubbornly low nationally. Aims To improve local Trust performance in meeting the NELA standard: "Peri-operative assessment by a member of the Geriatrician-Led MDT for frail (CFS 5+) patients 65 or older" to >80% (Green: >=80%, Amber: 50 - 79% Red: <50%) of estimated 100 patients per year. Methods 1. Proactive case finding with general surgical teams;2. Engagement with Emergency Surgical Committee and NELA leads;3. Improved our own electronic referral system;4. Assist in development of electronic booking system with emergency laparotomy cases Results We showed a significant improved in meeting the NELA standard from the red zone (Mean: 33% range 5% to 35%) into the amber with a of mean 60% (quartile range 52% to 78%) but still remains below our target with significant quarterly variation seen. All referrals and assessment remain post-intervention. Limitations in measures: Large variations in Frailty assessment and referral process (prospective Vs retrospective) Process rather than a Quality measure No balancing measures - Is there Reduced service elsewhere? Conclusions Following a number of change ideas and despite challenging COVID related staffing issues we showed that a combination of key stakeholder engagement, proactive case-finding and improved referral processes we have improved Geriatrician input in frail patients undergoing emergency laparotomy. We suspect due to the non-systematic assessment of frailty that we may be missing some patients and or seeing late in care pathway.

FACTORS ASSOCIATED with SEVERE COVID-19 OUTCOMES in PATIENTS with IDIOPATHIC INFLAMMATORY MYOPATHY: RESULTS from the COVID-19 GLOBAL RHEUMATOLOGY ALLIANCE PHYSICIAN-REPORTED REGISTRY

Background: There is a paucity of data in the literature about the outcome of patients with idiopathic infammatory myopathy (IIM) who have been infected with SARS-CoV-2. Objectives: To investigate factors associated with severe COVID-19 outcomes in patients with IIM. Methods: Data on demographics, number of comorbidities, region, COVID-19 time period, physician-reported disease activity, anti-rheumatic medication exposure at the clinical onset of COVID-19, and COVID-19 outcomes of IIM patients were obtained from the voluntary COVID-19 Global Rheumatology Alliance physician-reported registry of adults with rheumatic disease (from 17 March 2020 to 27 August 2021). An ordinal COVID-19 severity scale was used as primary outcome of interest, with each outcome category being mutually exclusive from the other:a) no hospital-ization, b) hospitalization (and no death), or c) death. Odds ratios (OR) were estimated using multivariable ordinal logistic regression. In ordinal logistic regression, the effect size of a categorical predictor can be interpreted as the odds of being one level higher on the ordinal COVID-19 severity scale than the reference category. Results: Complete hospitalization and death outcome data was available in 348 IIM cases. Mean age was 53 years, and 223 (64.1%) were female. Overall, 167/348 (48.0%) people were not hospitalized, 136/348 (39.1%) were hospitalized (and did not die), and 45/348 (12.9%) died. Older age (OR=1.59 per decade of life, 95%CI 1.32-1.93), male sex (OR=1.63, 95%CI 1.004-2.64;versus female), high disease activity (OR=4.05, 95%CI 1.29-12.76;versus remission), presence of two or more comorbidities (OR=2.39, 95%CI 1.22-4.68;versus none), predni-solone-equivalent dose >7.5 mg/day (OR=2.37, 95%CI 1.27-4.44;versus no gluco-corticoid intake), and exposure to rituximab (OR=2.60, 95%CI 1.23-5.47;versus csDMARDs only) were associated with worse COVID-19 outcomes (Table 1). Conclusion: These are the frst global registry data on the impact of COVID-19 on IIM patients. Older age, male gender, higher comorbidity burden, higher disease activity, higher glucocorticoid intake and rituximab exposure were associated with worse outcomes. These fndings will inform risk stratifcation and management decisions for IIM patients.

FOLIC ACID and METHOTREXATE USE and THEIR ASSOCIATION with COVID-19 DIAGNOSIS and MORTALITY: AN ANALYSIS from the UK BIOBANK

Background: Folate metabolism is implicated in SARS-CoV-2 infectivity (Ref). Objectives: To determine if methotrexate (an antifolate) or folic acid prescription were associated with a lowered and increased risk, respectively, for COVID-19 diagnosis or mortality in a large population-based cohort (UK Biobank). Methods: Data from 380,380 UK Biobank participants with general practice prescription data were used. Criteria for COVID-19 diagnosis were 1) a positive SARS-CoV-2 test and/or 2) ICD-10 code for confrmed COVID-19 (U07.1) or probable COVID-19 (U07.2) in hospital records, or death records. This defnition identifed 26,003 individuals diagnosed with COVID-19 of whom 820 were known to have died from COVID-19. Logistic regression statistical models were adjusted for age group (4 categories), sex, ethnicity, Townsend deprivation index, BMI, smoking status, presence of rheumatoid arthritis, sickle cell disease, use of anti-convulsants, statins and iron supplements. Results: Compared with people prescribed neither folic acid nor methotrexate, people prescribed folic acid supplementation had increased risk of diagnosis of COVID-19 (OR 1.51 [1.42;1.61]). The prescription of methotrexate with or without folic acid was not associated with COVID-19 diagnosis (P≥0.18). Compared with people prescribed neither folic acid nor methotrexate, people prescribed folic acid supplementation had an increased risk of death after a diagnosis of COVID-19 (OR 2.64 [2.15;3.24]) in a fully adjusted model. The prescription of methotrex-ate in combination with folic acid was not associated with an increased risk for death after a diagnosis of COVID-19 (1.07 [0.57;1.98]). (Table 1) Conclusion: We report increased risk for COVID-19 diagnosis and COV-ID-19-related death for people prescribed folic acid supplementation. The prescription and use of supplemental folic acid may confer risk of infection with the SARS-CoV-2 virus as well as the risk of death resulting from COVID-19. Our results also suggest that methotrexate might attenuate an increased risk for COVID-19 diagnosis and death conferred by folic acid.

CHARACTERISTICS ASSOCIATED with POOR COVID-19 OUTCOMES in PEOPLE with PSORIASIS and SPONDYLOARTHRITIS: DATA from the COVID-19 PSOPROTECT and GLOBAL RHEUMATOLOGY ALLIANCE PHYSICIAN-REPORTED REGISTRIES

Background: Some factors associated with severe COVID-19 outcomes have been identifed in patients with psoriasis (PsO) and infammatory/autoimmune rheumatic diseases, namely older age, male sex, comorbidity burden, higher disease activity, and certain medications such as rituximab. However, information about specifcities of patients with PsO, psoriatic arthritis (PsA) and axial spondyloarthritis (axSpA), including disease modifying anti-rheumatic drugs (DMARDs) specifcally licensed for these conditions, such as IL-17 inhibitors (IL-17i), IL-23/IL-12 + 23 inhibitors (IL-23/IL-12 + 23i), and apremilast, is lacking. Objectives: To determine characteristics associated with severe COVID-19 outcomes in people with PsO, PsA and axSpA. Methods: This study was a pooled analysis of data from two physician-reported registries: the Psoriasis Patient Registry for Outcomes, Therapy and Epidemiology of COVID-19 Infection (PsoProtect), comprising patients with PsO/PsA, and the COVID-19 Global Rheumatology Alliance (GRA) registry, comprising patients with PsA/axSpA. Data from the beginning of the pandemic up to 25 October, 2021 were included. An ordinal severity outcome was defned as: 1) not hospitalised, 2) hospitalised without death, and 3) death. A multivariable ordinal logistic regression model was constructed to assess the relationship between COVID-19 severity and demographic characteristics (age, sex, time period of infection), comorbidities (hypertension, other cardiovascular disease [CVD], chronic obstructive lung disease [COPD], asthma, other chronic lung disease, chronic kidney disease, cancer, smoking, obesity, diabetes mellitus [DM]), rheumatic/skin disease (PsO, PsA, axSpA), physician-reported disease activity, and medication exposure (methotrexate, lefunomide, sulfasalazine, TNFi, IL17i, IL-23/IL-12 + 23i, Janus kinase inhibitors (JAKi), apremilast, glucocorticoids [GC] and NSAIDs). Age-adjustment was performed employing four-knot restricted cubic splines. Country-adjustment was performed using random effects. Results: A total of 5008 individuals with PsO (n=921), PsA (n=2263) and axSpA (n=1824) were included. Mean age was 50 years (SD 13.5) and 51.8% were male. Hospitalisation (without death) was observed in 14.6% of cases and 1.8% died. In the multivariable model, the following variables were associated with severe COVID-19 outcomes: older age (Figure 1), male sex (OR 1.53, 95%CI 1.29-1.82), CVD (hypertension alone: 1.26, 1.02-1.56;other CVD alone: 1.89, 1.22-2.94;vs no hypertension and no other CVD), COPD or asthma (1.75, 1.32-2.32), other lung disease (2.56, 1.66-3.97), chronic kidney disease (2.32, 1.50-3.59), obesity and DM (obesity alone: 1.36, 1.07-1.71;DM alone: 1.85, 1.39-2.47;obesity and DM: 1.89, 1.34-2.67;vs no obesity and no DM), higher disease activity and GC intake (remission/low disease activity and GC intake: 1.96, 1.36-2.82;moderate/severe disease activity and no GC intake: 1.35, 1.05-1.72;moderate/severe disease activity and GC intake 2.30, 1.41-3.74;vs remission/low disease activity and no GC intake). Conversely, the following variables were associated with less severe COVID-19 outcomes: time period after 15 June 2020 (16 June 2020-31 December 2020: 0.42, 0.34-0.51;1 January 2021 onwards: 0.52, 0.41-0.67;vs time period until 15 June 2020), a diagnosis of PsO (without arthritis) (0.49, 0.37-0.65;vs PsA), and exposure to TNFi (0.58, 0.45-0.75;vs no DMARDs), IL17i (0.63, 0.45-0.88;vs no DMARDs), IL-23/IL-12 + 23i (0.68, 0.46-0.997;vs no DMARDs) and NSAIDs (0.77, 0.60-0.98;vs no NSAIDs). Conclusion: More severe COVID-19 outcomes in PsO, PsA and axSpA are largely driven by demographic factors (age, sex), comorbidities, and active disease. None of the DMARDs typically used in PsO, PsA and axSpA, were associated with severe COVID-19 outcomes, including IL-17i, IL-23/IL-12 + 23i, JAKi and apremilast.

EQUITY CONSIDERATIONS in COVID-19 VACCINATION STUDIES of INDIVIDUALS with AUTOIMMUNE INFLAMMATORY RHEUMATIC DISEASES

Background: Individuals with autoimmune infammatory rheumatic diseases (AIRDs) have an increased baseline risk of severe COVID-19 infection. Intersection of inequity factors may result in more severe adverse effects through influencing opportunities for health. We sought to examine the extent to which populations experiencing inequities were considered in studies of COVID-19 vaccination in individuals with AIRDs. Objectives: The objective of this study is to assess how health equity is considered in studies of COVID-19 vaccination studies in individuals with AIRDs. Methods: All studies (N=19) from an ongoing Cochrane living systematic review on the effects of COVID-19 vaccination in people with AIRDs were included. We identifed inequity factors using the PROGRESS-Plus framework which stands for Place of residence, Race/ethnicity, Occupation, Gender/sex, Religion, Education, Socioeconomic status, and Social capital. Age, multimorbidity, and health literacy were also assessed as 'Plus' factors. We applied the framework to assess equity considerations in relation to differences in COVID-19 baseline risk, description of participant characteristics, controlling for confounding factors, subgroup analysis and applicability of study fndings. RESULTS: All nineteen studies are cohort studies that followed individuals with AIRDs after COVID-19 vaccination. Two articles (11%) described differences in baseline risk for COVID-19 across age. All nineteen studies described participant age and sex, with race/ethnicity and multimorbidity described in four (21%) and occupation in one (5%). Seven studies (37%) controlled for age and/or sex as confounding factors. Eleven studies (58%) conducted subgroup analysis across at least one PROGRESS-Plus factor, most commonly age. Eight studies (42%) discussed at least one PROGRESS-Plus factor in interpreting the applicability of results, most commonly age (32%), then race/ethnicity and multimorbidity (11%). Conclusion: It is unknown whether COVID-19 vaccine studies on individuals with AIRDs are applicable to populations experiencing inequities, as key inequity factors beyond age and sex have little to no reporting or analysis. Future COVID-19 vaccine studies should report social characteristics of participants consistently, facilitating informed decisions about the applicability of study results to the population of interest.

Sleep quality in children with primary ciliary dyskinesia (PCD)

Introduction: PCD is a rare, progressive disease resulting in upper and lower respiratory tract manifestations that increase the risk of sleep disordered breathing. This study is the first to characterize sleep quality in Australian children with PCD and examine it is relationship to mood and health-related quality of life (HrQOL). Methods: Clinically stable children with PCD (1-18 years of age) were recruited. Subjective sleep quality was assessed with the SDSC, PDSS and OSA-18 questionnaires. HrQOL and depressive symptoms were assessed via age-appropriate QOL-PCD and CDI questionnaires. Demographic data including passive smoke exposure was recorded. Pulmonary function testing and ENT assessments were performed. Children underwent overnight polysomnography including transcutaneous CO2 and video monitoring (50% studies complete due to Covid restrictions). Results: Twenty-two participants (10 female) aged 8.1 ± 5.1 (mean ± SD) years were recruited. Mean(±SD) FEV1 was 78.2 ± 21.2%. Ninety-two percent of children assessed were diagnosed with chronic rhinosinusitis and 35% exposed to regular passive cigarette smoke. Polysomnography identified sleep fragmentation in 90% of studies and one case of mild obstructive sleep apnoea. Subjective sleep questionnaires revealed 76% of parents and 50% of children reported clinically significant scores indicating sleep disturbance and excessive daytime somnolence respectively. Twenty-seven percent of parents reported a moderate-severe impact of sleep disturbance on QOL. Forty-six percent of children had elevated depression scores. HrQOL and mood scores were correlated with poor subjective sleep quality. Children exposed to passive cigarette smoke had poorer subjective sleep quality and lower HrQOL and mood. Conclusion: Even in periods of clinical stability, children with PCD exhibit poor sleep quality and excessive daytime sleepiness and this is associated with lower mood and HrQOL. Based on these findings, we recommend routine screening for mental health and symptoms of sleep disturbance in children and adolescents with PCD. The continued education of children and families regarding the effects of cigarette smoke is advised.

Assessing Past vs Present COVID-19 Infection: A Survey of Criteria for Discontinuing Precautions in Asymptomatic Patients

Background. COVID-19 patients can remain positive by PCR-testing for several months. Pre-admission or pre-procedure testing can identify recovered asymptomatic patients who may no longer be contagious but would require precautions according to current CDC recommendations (10 days). This can result in unintended consequences, including procedure delays or transfer to appropriate care (e.g., psychiatric or post-trauma patients requiring admission to COVID-19 units instead of psychiatric or rehabilitation facilities, respectively). Methods. We conducted a structured survey of healthcare epidemiologists and infection prevention experts from the SHEA Research Network between March-April, 2021. The 14-question survey, presented a series of COVID-19 PCR+ asymptomatic patient case scenarios and asked respondents if (1) they would consider the case recovered and not infectious, (2) if they have cleared precautions in such cases, and if so, (3) how many transmission events occurred after discontinuing precautions. The survey used one or a combination of 5 criteria: history of COVID-19 symptoms, history of exposure to a household member with COVID-19, COVID-19 PCR cycle threshold (CT), and IgG serology. Percentages were calculated among respondents for each question. Results. Among 60 respondents, 56 (93%) were physicians, 51 (86%) were hospital epidemiologists, and 46 (77%) had >10y infection prevention experience. They represented facilities that cumulatively cared for >29,000 COVID-19 cases;46 (77%) were academic, and 42 (69%) were large ( >400 beds). One-third to one-half would consider an incidentally found PCR+ case as recovered based on solo criteria, particularly those with two consecutive high CTs or COVID IgG positivity recovered (53-55%) (Table 1). When combining two criteria, half to four-fifths of respondents deemed PCR+ cases to be recovered (Table 2). Half of those had used those criteria to clear precautions (45-64%) and few to none experienced a subsequent transmission event resulting from clearance. Conclusion. The majority of healthcare epidemiologists consider a combination of clinical and diagnostic criteria as recovered and many have used these to clear precautions without high numbers of transmission.

Reclassification of Infectious Disease in the Mondo Disease Ontology

The Mondo Disease Ontology (Mondo) represents cross-species diseases, which integrates several source disease terminologies to represent cross-species diseases, and provides precise semantic mappings to the original sources. Mondo spans both rare and 'common' diseases, as well as monogenic, acquired, neoplasms, infectious diseases, and more. Mondo is a community resource and is continuously updated and iteratively curated. Recent efforts sought to improve the representation of viral infectious diseases in Mondo, to properly represent primary infections, diseases caused by reactivation of a latent virus, such as shingles and diseases caused by aftereffects of a primary infection such as long COVID-19. This included the addition of new classes and new relations (object properties), and the creation of new design patterns. © 2021 Copyright for this paper by its authors.

Becoming virtual: Lessons learned from moving wessex core surgical training deanery teaching on-line

Aim: Surgical training has been significantly impacted by COVID-19. Social distancing requirements mandated a change in face-to-face teaching with many deaneries adopting 'virtual' sessions. Surgical training does not immediately lend itself to e-learning owing to its hands-on nature. We describe our experiences in developing a virtual teaching program for Core Surgical Trainees within the Wessex Deanery. We provide tips, tricks, and pitfalls for educators to establish or improve similar programs. Method: From June 2020 monthly, in-person teaching was replaced with virtual sessions. Quantitative and qualitative feedback directed improvements in the program. In addition to knowledge-based lectures we integrated on-line learning tools (LapPass) and utilised surgical videos to ensure continued development of surgical skills. Mock MRCS and ST3 interviews were conducted remotely using 'break-out rooms. Where face-to-face teaching was essential (Boot Camp, Field Camp) safety was ensured with reduced numbers (split sessions), social distancing and appropriate PPE. Results: All trainees strongly agreed (67%) or agreed (33%) that virtual teaching works well. There were no significant differences in feedback scores compared with face-to-face teaching. Attendance increased by 42%. Interactivity was maintained with 'cameras on, mics off', polling apps and chat box function. Advantages include uploading webinars for future review, ability for educators to present from multiple locations, increased availability and breadth of speakers and reduced burden on clinical commitments. Conclusions: The COVID-19 pandemic has dictated an evolution of surgical teaching. Virtual teaching has many advantages over face-to-face and should continue to play a part in postgraduate medical education, even after social distancing restrictions are lifted.

Association of search query interest in fibromyalgia and its symptoms during the COVID-19 pandemic in Australia: An infodemiology study

Purpose: To correlate internet search activity for fibromyalgia (FM) and its symptoms in relation to case numbers during the COVID-19 pandemic. Methods: Search terms chosen for inclusion included 'fibromyalgia', 'pain', 'joint pain', 'muscle ache', 'extreme fatigue' and 'poor sleep', based on a review of the most common symptoms reported among patients with FM. Positive control terms of 'fever' and 'cough' were used and 'rash' was a negative control. Google Trends was used to determine the frequency of search terms as a relative search volume (RSV) and queries for selected search terms were set with filters for specific dates and location. COVID-19 case numbers in Australia were obtained from the Australian Department of Health. Results: The average 2020 FM search volume in Australia was slightly less than the previous three years (RSV 47 vs RSV 49-53). While an increase in COVID-19 cases in Australia in August and September correlated with increased search popularity, similar increases were demonstrated in 2017 and 2019. Similar trends were seen in New Zealand but not the United States or United Kingdom. The RSV of typical FM symptoms including 'pain', 'muscle ache' and 'poor sleep' remained stable throughout 2020. Increases in search volumes for 'extreme fatigue' were seen in February and September to November. Significant increases in the positive control terms occurred in March, coinciding with a rapid rise in COVID-19 cases, but negative and positive controls otherwise remained similar to previous years. Conclusion: Despite public health social distancing restrictions and high COVID-19 case numbers, infodemiology techniques did not detect any increase in fibromyalgia-related searches. This occurred despite seasonal variation in ' fibromyalgia ' searches in Southern Hemisphere countries and marked changes in positive control search terms. A large increase in disease activity appears not to have emerged, and patients with fibromyalgia and ' fibromyalgianess ' may have demonstrated surprising resilience during the COVID-19 pandemic.