Deep phenotyping of 34,128 patients hospitalised with COVID-19 and a comparison with 81,596 influenza patients in America, Europe and Asia: an international network study

Edward Burn, Seng Chan You, Anthony G. Sena, Kristin Kostka, Hamed Abedtash, Maria Tereza Fernandes Abrahão, Amanda Alberga, Heba Alghoul, Osaid Alser, Thamir M. Alshammari, María Aragón, Carlos Areia, Juan M. Banda, Jaehyeong Cho, Aedín C. Culhane, Alexander Davydov, Frank DeFalco, Talita Duarte‐Salles, Scott L. DuVall, Thomas Falconer, Sergio Fernández‐Bertolín, Weihua Gao, Asieh Golozar, Jill Hardin, George Hripcsak, Vojtech Huser, Hokyun Jeon, Yonghua Jing, Chi Young Jung, Benjamin Skov Kaas‐Hansen, Denys Kaduk, Seamus Kent, Yee-Suk Kim, Spyros Kolovos, Jennifer C. E. Lane, Hyejin Lee, Kristine E. Lynch, Rupa Makadia, Michael E. Matheny, Paras Mehta, Daniel R. Morales, Karthik Natarajan, Fredrik Nyberg, Anna Ostropolets, Rae Woong Park, Jimyung Park, Jose Posada, Albert Prats‐Uribe, Gowtham Rao, Christian Reich, Yeunsook Rho, Peter R. Rijnbeek, Lisa M. Schilling, Martijn J. Schuemie, Nigam H. Shah, Azza Shoaibi, Seok Young Song, Henry M. Spotnitz, Marc A. Suchard, Joel N. Swerdel, David Vizcaya, Salvatore Volpe, Haini Wen, Andrew E. Williams, Belay Birlie Yimer, Lin Zhang, Oleg Zhuk, Daniel Prieto‐Alhambra, Patrick RyanmedRxiv 2020 · DOI

TL;DR

This study provides a comprehensive "deep phenotyping" of over 34,000 patients hospitalized with COVID-19 across the US, South Korea, and Spain, detailing their medical histories and medications. By comparing these patients to over 81,000 hospitalized with influenza, the research reveals that COVID-19 patients were generally younger, healthier, and more likely to be male, suggesting that existing influenza-focused vulnerability strategies may need to be adapted for COVID-19.

What Problem Does This Paper Address?

The COVID-19 pandemic presented a significant challenge to healthcare systems worldwide, and understanding the characteristics of hospitalized patients was crucial for effective response and resource allocation. While early studies provided some insights, there was substantial uncertainty regarding the prevalence of comorbidities and prior medication use among COVID-19 patients. Furthermore, it was unclear how these patients systematically differed from those hospitalized during previous influenza seasons. This knowledge gap made it difficult to tailor public health interventions and clinical management strategies effectively. This paper aims to bridge this gap by providing a detailed, multi-continental characterization of COVID-19 patients and a direct comparison to influenza patients.

Methods

The study employed a retrospective cohort design, leveraging routinely collected electronic health records (EHRs), hospital billing data, and insurance claims. A key methodological aspect was the use of the Observational Medical Outcomes Partnership (OMOP) Common Data Model (CDM). This standardized data model allowed for the harmonization of data from diverse sources across different countries, enabling consistent analysis without sharing raw patient-level data. The researchers extracted a wide range of aggregate characteristics, including demographics, pre-existing conditions (identified using SNOMED codes), and medication use (categorized using Anatomical Therapeutic Chemical (ATC) groupings). Medication use was analyzed over different time windows: a year prior to hospitalization, 30 days prior, and on the day of hospitalization. Standardized Mean Differences (SMDs) were calculated to compare characteristics between COVID-19 and influenza cohorts.

Datasets

The study utilized data from multiple sources across three continents:

  • United States:
    • Columbia University Irving Medical Center (CUIMC)
    • Premier Healthcare Database (PHD)
    • UCHealth System Health Data Compass (UC HDC)
    • Department of Veterans Affairs (VA OMOP)
  • South Korea:
    • Health Insurance Review & Assessment (HIRA)
  • Spain:
    • The Information System for Research in Primary Care (SIDIAP)
    • HM Hospitales (HM)

These datasets were used to identify:

  • 34,128 individuals hospitalized with COVID-19 (December 2019 - April 2020).
  • 81,596 individuals hospitalized with influenza (September 2014 - April 2019, with a subset from 2009-2010).

Key Insights and Ideas

  • COVID-19 patients are generally "healthier" than influenza patients: A surprising finding was that COVID-19 patients, on average, presented with fewer comorbidities and lower medication use compared to those hospitalized with influenza in previous seasons. This challenges the initial assumption that severe COVID-19 would disproportionately affect individuals with extensive pre-existing conditions, similar to severe influenza.
  • Demographic differences are significant: COVID-19 hospitalizations skewed male, particularly in the US and Spain, and were generally younger than influenza hospitalizations. This contrasts with some influenza seasons where older individuals are more severely affected.
  • Geographic variation is substantial: The study highlighted significant differences in patient characteristics across the contributing databases and countries, particularly in South Korea where patients were predominantly female and younger. This underscores the importance of context-specific analysis.
  • Medication use patterns offer clues: While overall medication use was lower in COVID-19 patients, specific drug classes, like those acting on the renin-angiotensin system, were still commonly used, raising questions about their impact and management during COVID-19 infection.

Key Results

  • Patient Cohort: 34,128 COVID-19 patients and 81,596 influenza patients were analyzed.
  • Demographics: COVID-19 patients were more typically male (e.g., 94% in VA OMOP) and younger compared to influenza patients. South Korean patients (HIRA) were predominantly female (60%).
  • Comorbidities: Prevalence of conditions like asthma, diabetes, and hypertension varied significantly across data sources but were generally comparable or lower in COVID-19 patients compared to influenza patients. For instance, heart disease prevalence in the US ranged from 28% to 48% for COVID-19 patients, while it was higher in some influenza cohorts.
  • Medication Use: Between 14% and 33% of COVID-19 patients were taking drugs acting on the renin-angiotensin system in the 30 days prior to hospitalization. Overall medication use was lower in COVID-19 patients compared to influenza patients.

Limitations and Future Work

  • Data Quality: The study acknowledges that the reliance on routinely collected data means potential data quality issues, such as underestimation of conditions due to missing codes or misclassification of medication use (e.g., prescribed vs. actually taken).
  • Snapshot in Time: The characterization is based on the point of hospital admission, providing a snapshot rather than a complete picture of disease progression or community-level infection.
  • Comparison Nuances: Comparing COVID-19 patients to influenza patients across different time periods has inherent limitations, as clinical practices, diagnostic criteria, and even the nature of influenza strains can change over time.
  • Future Work: The authors suggest that while existing influenza vulnerability strategies are a good starting point, they need to be broadened to account for the specific characteristics of COVID-19 patients. Further research is needed on the benefits and harms of continued use of medications commonly taken by COVID-19 patients, especially those with potential implications for COVID-19 susceptibility or severity. The study also highlights the need for further research into the consequences of temporarily discontinuing certain medications for cardiovascular risks and mortality.

Summary updated

Abstract

Background In this study we phenotyped individuals hospitalised with coronavirus disease 2019 (COVID-19) in depth, summarising entire medical histories, including medications, as captured in routinely collected data drawn from databases across three continents. We then compared individuals hospitalised with COVID-19 to those previously hospitalised with influenza. Methods We report demographics, previously recorded conditions and medication use of patients hospitalised with COVID-19 in the US (Columbia University Irving Medical Center [CUIMC], Premier Healthcare Database [PHD], UCHealth System Health Data Compass Database [UC HDC], and the Department of Veterans Affairs [VA OMOP]), in South Korea (Health Insurance Review & Assessment [HIRA]), and Spain (The Information System for Research in Primary Care [SIDIAP] and HM Hospitales [HM]). These patients were then compared with patients hospitalised with influenza in 2014-19. Results 34,128 (US: 8,362, South Korea: 7,341, Spain: 18,425) individuals hospitalised with COVID-19 were included. Between 4,811 (HM) and 11,643 (CUIMC) unique aggregate characteristics were extracted per patient, with all summarised in an accompanying interactive website (http://evidence.ohdsi.org/Covid19CharacterizationHospitalization/). Patients were majority male in the US (CUIMC: 52%, PHD: 52%, UC HDC: 54%, VA OMOP: 94%,) and Spain (SIDIAP: 54%, HM: 60%), but were predominantly female in South Korea (HIRA: 60%). Age profiles varied across data sources. Prevalence of asthma ranged from 4% to 15%, diabetes from 13% to 43%, and hypertensive disorder from 24% to 70% across data sources. Between 14% and 33% were taking drugs acting on the renin-angiotensin system in the 30 days prior to hospitalisation. Compared to 81,596 individuals hospitalised with influenza in 2014-19, patients admitted with COVID-19 were more typically male, younger, and healthier, with fewer comorbidities and lower medication use. Conclusions We provide a detailed characterisation of patients hospitalised with COVID-19. Protecting groups known to be vulnerable to influenza is a useful starting point to minimize the number of hospital admissions needed for COVID-19. However, such strategies will also likely need to be broadened so as to reflect the particular characteristics of individuals hospitalised with COVID-19.
BibTeX
@misc{burn2020deep,
  author = {Edward Burn and Seng Chan You and Anthony G. Sena and Kristin Kostka and Hamed Abedtash and Maria Tereza Fernandes Abrahão and Amanda Alberga and Heba Alghoul and Osaid Alser and Thamir M. Alshammari and María Aragón and Carlos Areia and Juan M. Banda and Jaehyeong Cho and Aedín C. Culhane and Alexander Davydov and Frank DeFalco and Talita Duarte‐Salles and Scott L. DuVall and Thomas Falconer and Sergio Fernández‐Bertolín and Weihua Gao and Asieh Golozar and Jill Hardin and George Hripcsak and Vojtech Huser and Hokyun Jeon and Yonghua Jing and Chi Young Jung and Benjamin Skov Kaas‐Hansen and Denys Kaduk and Seamus Kent and Yee-Suk Kim and Spyros Kolovos and Jennifer C. E. Lane and Hyejin Lee and Kristine E. Lynch and Rupa Makadia and Michael E. Matheny and Paras Mehta and Daniel R. Morales and Karthik Natarajan and Fredrik Nyberg and Anna Ostropolets and Rae Woong Park and Jimyung Park and Jose Posada and Albert Prats‐Uribe and Gowtham Rao and Christian Reich and Yeunsook Rho and Peter R. Rijnbeek and Lisa M. Schilling and Martijn J. Schuemie and Nigam H. Shah and Azza Shoaibi and Seok Young Song and Henry M. Spotnitz and Marc A. Suchard and Joel N. Swerdel and David Vizcaya and Salvatore Volpe and Haini Wen and Andrew E. Williams and Belay Birlie Yimer and Lin Zhang and Oleg Zhuk and Daniel Prieto‐Alhambra and Patrick Ryan},
  title = {{Deep phenotyping of 34,128 patients hospitalised with COVID-19 and a comparison with 81,596 influenza patients in America, Europe and Asia: an international network study}},
  year = {2020},
  doi = {10.1101/2020.04.22.20074336},
  howpublished = {medRxiv},
}
Figure