In a recent study posted to the medRxiv* preprint server, researchers longitudinally assessed the long coronavirus disease (COVID) evidence gap by comparing self-documented data and clinical codes of long COVID with the electronic health records (EHRs) of the participants.
Study: The long COVID evidence gap: comparing self-reporting and clinical coding of long COVID using longitudinal study data linked to healthcare records. Image Credit: Pressmaster/Shutterstock
*Important notice: medRxiv publishes preliminary scientific reports that are not peer-reviewed and, therefore, should not be regarded as conclusive, guide clinical practice/health-related behavior, or treated as established information.
Background
The terminology ‘long COVID’ was established during the spring season of 2020 for individuals with symptoms persisting beyond the acute phase of COVID-2019 (COVID-19); however, clinical long COVID codes were created in December of 2020 for persistent COVID-19 symptoms and patient referrals in the EHRs. Analyzing EHRs at a population level has improved understanding of long COVID epidemiology; however, there have been concerns over EHR data completeness regarding long COVID.
Longitudinal population studies (LPS) have been conducted in the United Kingdom (UK) to obtain self-documented COVID-19 and long COVID data from the initial period of 2020 and uploaded the data in the United Kingdom longitudinal linkage collaboration (LLC) database, wherein the data is linked to the electronic health records of the individuals. Comparative evaluations of longitudinal population studies reported long COVID data with long COVID data in the electronic health records of the individuals could further improve understanding of long COVID epidemiology.
About the study
In the present study, researchers investigated probable discrepancies in perceived and documented long COVID data.
Data from 10 UK longitudinal population-based studies (LPS), uploaded to the UK LLC database, of 6,412 individuals whose COVID-19-associated survey information was linked with their electronic health records were analyzed. Self-documented long COVID was described as documenting the presence of COVID-19 symptoms for ≥4.0 weeks, based on the national institute for health and care excellence (NICE) criteria. Seven longitudinal population studies obtained data on debilitating COVID-19 symptoms only, whereas the other three studies obtained data on any persistent symptom related to COVID-19.
The team identified long COVID-associated health interactions using the international classification of diseases, 10th revision (ICD-10) through August of 2022, listed in the national general practice extraction service data for pandemic planning and research (GDPPR) dataset for COVID-19-associated prime care records of English individuals. In addition, data on COVID-19-associated secondary care records were obtained from the national hospital episodes statistics (HES) database.
Results
Of 6,412 individuals with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection symptom data linked to their EHRs, 14% (n=898) of individuals self-documented long COVID in the longitudinal population study surveys, among which only 5.0% (n=42) of individuals had long COVID evidence in their electronic health records.
Among individuals documenting debilitating long COVID, the percentage was marginally higher (6.0%). The codes were provided within mean durations of four months and five months of COVID-19 symptom documentation, respectively. The probability of being assigned long COVID codes was greater for middle-aged individuals and lesser for older and younger individuals.
Whites showed a greater likelihood of receiving long COVID-related codes than other individuals. No sex-based differences were observed in coding probability. The team found weak evidence for individuals of greater socioeconomic status having a greater probability of long COVID evidence in their EHRs. The absolute percent differences in long COVD coding in the EHRs between individuals with sociodemographic diversity and self-documented long COVID data (n ≤898) for females (vs. males) and Whites (vs. other ethnicities) were 0.5% and 5.8%, respectively.
Stratified by age, the percentage differences for tertile 2.0 individuals with a mean age of 46 years versus tertile 1.0 individuals with a mean age of 25 years and versus tertile 3.0 individuals with a mean age of 63 years were 3.8% and 3.4%, respectively. By socioeconomic position, using the multiple deprivation index, the corresponding percent difference for tertile 2.0 individuals (versus tertile 1.0, most socioeconomically deprived individuals) was 0.4%, and for tertile 3.0 least deprived individuals (versus tertile 1.0, most socioeconomically deprived individuals) was 1.7%.
Conclusion
Overall, the study findings showed a notable discrepancy between long COVID as discerned and documented by LPS participants, and long COVID evidence in the EHRs, patterned by ethnic background and probably by socioeconomic status. However, self-documented symptoms might not be reflected in coded EHRs due to varied care-seeking behavior among individuals and coding practices. The findings indicate a considerable unmet need in maintaining patient records of difficulties in access to health services and suboptimal identification and response to illnesses among the individuals when they seek care.
*Important notice: medRxiv publishes preliminary scientific reports that are not peer-reviewed and, therefore, should not be regarded as conclusive, guide clinical practice/health-related behavior, or treated as established information.
Inequities in COVID-19 vaccination coverage for older adults highlighted in global study