Predicting past and future SARS-CoV-2-related sick leave using discrete time Markov modelling
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (ScreenIT)
Abstract
Prediction of SARS-CoV-2-induced sick leave among healthcare workers (HCWs) is essential for being able to plan the healthcare response to the epidemic.
Methods
During first wave of the SARS-Cov-2 epidemic (April 23 rd to June 24 th , 2020), the HCWs in the greater Stockholm region in Sweden were invited to a study of past or present SARS-CoV-2 infection. We develop a discrete time Markov model using a cohort of 9449 healthcare workers (HCWs) who had complete data on SARS-CoV-2 RNA and antibodies as well as sick leave data for the calendar year 2020. The one-week and standardized longer term transition probabilities of sick leave and the ratios of the standardized probabilities for the baseline covariate distribution were compared with the referent period (an independent period when there were no SARS-CoV-2 infections) in relation to PCR results, serology results and gender.
Results
The one-week probabilities of transitioning from healthy to partial sick leave or full sick leave during the outbreak as compared to after the outbreak were highest for healthy HCWs testing positive for large amounts of virus (ratio: 3.69, (95% confidence interval, CI: 2.44–5.59) and 6.67 (95% CI: 1.58–28.13), respectively). The proportion of all sick leaves attributed to COVID-19 during outbreak was at most 55% (95% CI: 50%-59%).
Conclusions
A robust Markov model enabled use of simple SARS-CoV-2 testing data for quantifying past and future COVID-related sick leave among HCWs, which can serve as a basis for planning of healthcare during outbreaks.
Article activity feed
-
-
SciScore for 10.1101/2022.04.22.22274163: (What is this?)
Please note, not all rigor criteria are appropriate for all manuscripts.
Table 1: Rigor
Ethics Consent: All enrolled individuals provided a written informed consent that gave permission to link to the hospital administrative databases to obtain information from sick leave records. Sex as a biological variable not detected. Randomization not detected. Blinding not detected. Power Analysis not detected. Table 2: Resources
Antibodies Sentences Resources The SARS-CoV-2 antibody test used an assay with a sensitivity of 99.4% and a specificity of 99.1% (10). SARS-CoV-2suggested: NoneResults from OddPub: We did not detect open data. We also did not detect open code. Researchers are encouraged to share open data when possible (see Nature blog).
Results from LimitationRecognizer: We detected the following …SciScore for 10.1101/2022.04.22.22274163: (What is this?)
Please note, not all rigor criteria are appropriate for all manuscripts.
Table 1: Rigor
Ethics Consent: All enrolled individuals provided a written informed consent that gave permission to link to the hospital administrative databases to obtain information from sick leave records. Sex as a biological variable not detected. Randomization not detected. Blinding not detected. Power Analysis not detected. Table 2: Resources
Antibodies Sentences Resources The SARS-CoV-2 antibody test used an assay with a sensitivity of 99.4% and a specificity of 99.1% (10). SARS-CoV-2suggested: NoneResults from OddPub: We did not detect open data. We also did not detect open code. Researchers are encouraged to share open data when possible (see Nature blog).
Results from LimitationRecognizer: We detected the following sentences addressing limitations in the study:Strength and limitations: We used routinely collected sick leave data over a long period based on the administrative information from a large and systematically enrolled hospital HCW cohort, minimizing recall bias and ensuring high data quality. Using information from the administrative database in the hospital allowed us to have virtually complete information on sick leave and therefore increased the validity of our findings. The serology platform used in our study contained several SARS-CoV-2 proteins and was validated for high sensitivity and specificity. We have also been able to show the one-week and longer-term transition probabilities from one state to another state over the study period as well as attributable fraction of sick leave due to COVID-19, which adds the understanding on impact of infections of SARS-CoV-2 on sick leave beyond the previous evidence. Moreover, we have provided a novel example of using discrete-time Markov models for analyzing such data that is not commonly used in the current literature (15). There were some limitations in our study. First, we have only been able to measure the infections of SARS-CoV-2 through PCR and serology at the baseline instead of repeated measurement to identify the diseases states. However, we have followed up sick leave data which was a less biased measures for future infections and occurrence of diseases. Second, we were not able to test the HCWs that were not at work which might include a proportion of individuals t...
Results from TrialIdentifier: We found the following clinical trial numbers in your paper:
Identifier Status Title NCT04411576 Completed Current and Past SARS-CoV-2 Infection and COVID-19 in Health… Results from Barzooka: We did not find any issues relating to the usage of bar graphs.
Results from JetFighter: We did not find any issues relating to colormaps.
Results from rtransparent:- Thank you for including a conflict of interest statement. Authors are encouraged to include this statement when submitting to a journal.
- Thank you for including a funding statement. Authors are encouraged to include this statement when submitting to a journal.
- No protocol registration statement was detected.
Results from scite Reference Check: We found no unreliable references.
-
