Skip to main content
SearchLoginLogin or Signup

Review 1: "Healthcare Utilisation of 282,080 Individuals with Long COVID over Two Years: A Multiple Matched Control Cohort Analysis"

Both authors found that this preprint seemed reliable and had appropriate methodology for the primary research question, though recommended more details to be provided in the Methods section.

Published onFeb 20, 2024
Review 1: "Healthcare Utilisation of 282,080 Individuals with Long COVID over Two Years: A Multiple Matched Control Cohort Analysis"

RR:C19 Evidence Scale rating by reviewer:

  • Reliable. The main study claims are generally justified by its methods and data. The results and conclusions are likely to be similar to the hypothetical ideal study. There are some minor caveats or limitations, but they would/do not change the major claims of the study. The study provides sufficient strength of evidence on its own that its main claims should be considered actionable, with some room for future revision.

***************************************

Review: 

This preprint provides evidence that patients with long covid have higher use of healthcare services and incur higher healthcare costs than patients who do not have the condition. 

Methods:

  • There was no description of the datasets used. For example, what does the CVD-COVID-UK/COVID-IMPACT consortium entail? What is the COVID General Practice Extraction Service Data for Pandemic Planning and Research? Is it based on existing datasets such as CPRD? Also for non-UK readers a brief explanation of HES would be relevant. 

  • Additionally, SNOMED and ICD-10 codes should also be provided for the COVID only, no LC group.

  • I did not quite follow this assertion “If multiple records were found on the same day, it was counted as one consultation, admission or attendance, as appropriate”. So for example, if the patient had two outpatient appointments in one day (say to cardiology and then respiratory medicine), only one of these appointments was counted? Why is that?

  • Health care utilisation, with the probable exception of days in hospital, are count data. Why were rates and Poisson not used and averages and t-tests used instead?

  • It appears that National Tariff payment system schedules were used to value resource use. These schedules are a set of prices used by NHS commissioners and providers, but they do not reflect cost but rather payments. NHS Reference Costs should have been used instead. 

  • I’m unsure as to why for outpatient consultations, the authors only concentrated on high-frequency specialties and investigations for LC patient referrals. Why did they not include all visits? Once an HRG codes are obtained for each appointment, these are easily linked to NHS costs. Following on, what happened if the patient had an outpatient appointment under “Vascular Surgery” (not a high-frequency specialty) and as part of that appointment they had an MRI? Was the cost of the MRI included but not the appointment?

  • Given the lack of details on the datasets used, it is difficult to gauge the availability of certain data. For example, the authors reported they had lack of sufficient data to gage the organ-specific reasons for critical care. In datasets such as CPRD, the HES-linked dataset does have enough information that after uploading data to the NHS groupers (https://digital.nhs.uk/services/national-casemix-office/downloads-groupers-and-tools/hrg4-2021-22-national-costs-grouper) HRGs with organ-specific reasons are provided. 

  • Related to this, the authors do not detail the process by which resource use was converted to costs. For example, did they upload all the data into the NHS grouper, which then provided HRG codes for each hospital encounter, allow linkage to NHS tariffs?

  • The authors are implicitly claiming that if a patient did not, for example, appear in the outpatient dataset, this was missing data, and the value for that patient in that category was 0. I think the authors are being a bit harsh on themselves, with the probable cause for this being that the patient had no visits. The relevant sentences should therefore be re-written. 

  • Given that costs are presented per year, how did the authors account for censoring in patients with less than 1 year follow-up?

Results:

  • As per my comments in the methods, it would have been useful the population figures. 

  • It would be useful if measures of precision were provided alongside mean costs, e.g. SD or 95% Cis.

Discussion:

  • In the “strengths and limitations” discussion, what does EHR stand for

Minor comments:

  • In the third line, there is a typo: sentence should read “regardless of hospitalisation status”.

Comments
1
?
rajin daws:

Tube Jumpers is a multiplayer platformer game where players compete to survive rotating tubes, aiming to outmaneuver opponents in vibrant, fast-paced challenges with colorful visuals.