Coding long COVID: characterizing a new disease through an ICD-10 lens.

BACKGROUND: Naming a newly discovered disease is a difficult process; in the context of the COVID-19 pandemic and the existence of post-acute sequelae of SARS-CoV-2 infection (PASC), which includes long COVID, it has proven especially challenging. Disease definitions and assignment of a diagnosis code are often asynchronous and iterative. The clinical definition and our understanding of the underlying mechanisms of long COVID are still in flux, and the deployment of an ICD-10-CM code for long COVID in the USA took nearly 2 years after patients had begun to describe their condition. Here, we leverage the largest publicly available HIPAA-limited dataset about patients with COVID-19 in the US to examine the heterogeneity of adoption and use of U09.9, the ICD-10-CM code for "Post COVID-19 condition, unspecified." METHODS: We undertook a number of analyses to characterize the N3C population with a U09.9 diagnosis code (n = 33,782), including assessing person-level demographics and a number of area-level social determinants of health; diagnoses commonly co-occurring with U09.9, clustered using the Louvain algorithm; and quantifying medications and procedures recorded within 60 days of U09.9 diagnosis. We stratified all analyses by age group in order to discern differing patterns of care across the lifespan. RESULTS: We established the diagnoses most commonly co-occurring with U09.9 and algorithmically clustered them into four major categories: cardiopulmonary, neurological, gastrointestinal, and comorbid conditions. Importantly, we discovered that the population of patients diagnosed with U09.9 is demographically skewed toward female, White, non-Hispanic individuals, as well as individuals living in areas with low poverty and low unemployment. Our results also include a characterization of common procedures and medications associated with U09.9-coded patients. CONCLUSIONS: This work offers insight into potential subtypes and current practice patterns around long COVID and speaks to the existence of disparities in the diagnosis of patients with long COVID. This latter finding in particular requires further research and urgent remediation.

Geographically skewed recruitment and COVID-19 seroprevalence estimates: a cross-sectional serosurveillance study and mathematical modelling analysis.

OBJECTIVES: Convenience sampling is an imperfect but important tool for seroprevalence studies. For COVID-19, local geographic variation in cases or vaccination can confound studies that rely on the geographically skewed recruitment inherent to convenience sampling. The objectives of this study were: (1) quantifying how geographically skewed recruitment influences SARS-CoV-2 seroprevalence estimates obtained via convenience sampling and (2) developing new methods that employ Global Positioning System (GPS)-derived foot traffic data to measure and minimise bias and uncertainty due to geographically skewed recruitment. DESIGN: We used data from a local convenience-sampled seroprevalence study to map the geographic distribution of study participants' reported home locations and compared this to the geographic distribution of reported COVID-19 cases across the study catchment area. Using a numerical simulation, we quantified bias and uncertainty in SARS-CoV-2 seroprevalence estimates obtained using different geographically skewed recruitment scenarios. We employed GPS-derived foot traffic data to estimate the geographic distribution of participants for different recruitment locations and used this data to identify recruitment locations that minimise bias and uncertainty in resulting seroprevalence estimates. RESULTS: The geographic distribution of participants in convenience-sampled seroprevalence surveys can be strongly skewed towards individuals living near the study recruitment location. Uncertainty in seroprevalence estimates increased when neighbourhoods with higher disease burden or larger populations were undersampled. Failure to account for undersampling or oversampling across neighbourhoods also resulted in biased seroprevalence estimates. GPS-derived foot traffic data correlated with the geographic distribution of serosurveillance study participants. CONCLUSIONS: Local geographic variation in seropositivity is an important concern in SARS-CoV-2 serosurveillance studies that rely on geographically skewed recruitment strategies. Using GPS-derived foot traffic data to select recruitment sites and recording participants' home locations can improve study design and interpretation.

Identifying who has long COVID in the USA: a machine learning approach using N3C data.

BACKGROUND: Post-acute sequelae of SARS-CoV-2 infection, known as long COVID, have severely affected recovery from the COVID-19 pandemic for patients and society alike. Long COVID is characterised by evolving, heterogeneous symptoms, making it challenging to derive an unambiguous definition. Studies of electronic health records are a crucial element of the US National Institutes of Health's RECOVER Initiative, which is addressing the urgent need to understand long COVID, identify treatments, and accurately identify who has it-the latter is the aim of this study. METHODS: Using the National COVID Cohort Collaborative's (N3C) electronic health record repository, we developed XGBoost machine learning models to identify potential patients with long COVID. We defined our base population (n=1 793 604) as any non-deceased adult patient (age ≥18 years) with either an International Classification of Diseases-10-Clinical Modification COVID-19 diagnosis code (U07.1) from an inpatient or emergency visit, or a positive SARS-CoV-2 PCR or antigen test, and for whom at least 90 days have passed since COVID-19 index date. We examined demographics, health-care utilisation, diagnoses, and medications for 97 995 adults with COVID-19. We used data on these features and 597 patients from a long COVID clinic to train three machine learning models to identify potential long COVID among all patients with COVID-19, patients hospitalised with COVID-19, and patients who had COVID-19 but were not hospitalised. Feature importance was determined via Shapley values. We further validated the models on data from a fourth site. FINDINGS: Our models identified, with high accuracy, patients who potentially have long COVID, achieving areas under the receiver operator characteristic curve of 0·92 (all patients), 0·90 (hospitalised), and 0·85 (non-hospitalised). Important features, as defined by Shapley values, include rate of health-care utilisation, patient age, dyspnoea, and other diagnosis and medication information available within the electronic health record. INTERPRETATION: Patients identified by our models as potentially having long COVID can be interpreted as patients warranting care at a specialty clinic for long COVID, which is an essential proxy for long COVID diagnosis as its definition continues to evolve. We also achieve the urgent goal of identifying potential long COVID in patients for clinical trials. As more data sources are identified, our models can be retrained and tuned based on the needs of individual studies. FUNDING: US National Institutes of Health and National Center for Advancing Translational Sciences through the RECOVER Initiative.

Regulatory Sandboxes: A Cure for mHealth Pilotitis?

Mobile health (mHealth) and related digital health interventions in the past decade have not always scaled globally as anticipated earlier despite large investments by governments and philanthropic foundations. The implementation of digital health tools has suffered from 2 limitations: (1) the interventions commonly ignore the "law of amplification" that states that technology is most likely to succeed when it seeks to augment and not alter human behavior; and (2) end-user needs and clinical gaps are often poorly understood while designing solutions, contributing to a substantial decrease in usage, referred to as the "law of attrition" in eHealth. The COVID-19 pandemic has addressed the first of the 2 problems-technology solutions, such as telemedicine, that were struggling to find traction are now closely aligned with health-seeking behavior. The second problem (poorly designed solutions) persists, as demonstrated by a plethora of poorly designed epidemic prediction tools and digital contact-tracing apps, which were deployed at scale, around the world, with little validation. The pandemic has accelerated the Indian state's desire to build the nation's digital health ecosystem. We call for the inclusion of regulatory sandboxes, as successfully done in the fintech sector, to provide a real-world testing environment for mHealth solutions before deploying them at scale.