key: cord-0706551-34m5twz4
authors: Pathan, Sameer A.; Moinudheen, Jibin; Simon, Katie; Thomas, Stephen H.
title: COVID-19 cases presenting to the Emergency Department predict Qatar National COVID-19 trends and numbers
date: 2021-10-21
journal: Qatar Med J
DOI: 10.5339/qmj.2021.56
sha: ca328d58732a7e0489b41fe648164c0def5deb5e
doc_id: 706551
cord_uid: 34m5twz4

In this short communication, we summarized the analyses, models, and interpretations of the corporate department of emergency medicine's (CDEM) COVID-19 numbers and their relationship to predict the national COVID-19 trends and numbers in Qatar. Data included in this analysis were obtained between March 1, 2020 and July 31, 2021. It included the number of COVID-19 cases that presented to four major EDs under the Hamad Medical Corporation CDEM umbrella and published data from the Qatar Ministry of public health (MoPH). On plotting weighted scatterplot smoothing (lowess) trend lines, there were striking similarities between CDEM and national COVID-19 n curves for overall trends and peaks. In conclusion, CDEM COVID-19 spike may be useful to predict national COVID-19 spike in 2–3 weeks.

geography of Qatar's habitation. Qatar's confirmation of COVID-19 positivity is through a real-time reverse transcription-polymerase chain reaction (rRT-PCR) test. Qatar's number of new cases is published daily on the Ministry of public health (MoPH) website. 4 In May 2020, the HMC's CDEM Research Division noted that the national peak occurred approximately 2.5 weeks after peak COVID-positivity was noted in the HMC EDs. The pattern of approximately 2.5 weeks' interval between CDEM and national peaks also occurred in the Spring 2021 Qatari COVID peak. Plots of the daily CDEM COVID and national COVID numbers are combined in Figure 1 . Figure 1 depicts that for the May 2020 national COVID peak, there was an interval of 18 days between the HMC ED peak and the nationwide peak. Figure 1 also shows the 19-day period between the CDEM peak and national peak that was seen in the April/May 2021 time frame.

The striking similarity of the CDEM and national COVID n curves is emphasized when the national curve is "frame-shifted" 18 days to the left on the x-axis. In Figure 2 , a given "index" day's CDEM COVID n is plotted paired (i.e., in vertical alignment) with the nationwide COVID number for 18 days after the index date. In Figure 2 , the locally-weighted scatterplot smoothing (lowess) trend depicts a relationship between variables and foresees trends using the smooth line, lines to emphasize plots' similarities in overall trends and peaks. A simple regression model was generated to assess the precision of the predictive value of CDEM COVID diagnoses to forecast the national COVID numbers that would be seen 18 days later. The model demonstrated that simply knowing the CDEM COVID volume on a given day could predict national COVID levels in 2-3 weeks, with 90% accuracy (r 2 .88, p , .001). The coefficient for ED prediction of national COVID n was 3.9 (95% CI based on Huber-White sandwich method: 3.7 -4.1).

An overall plot of model-predicted vs. actual national COVID numbers in shown in Figure 3 . The figure shows that the ED case numbers tend to underestimate the national COVID numbers but also that the overall trends are similar. The lowess trend line for model-predicted vs. actually reported national COVID n is close to the 45-degree line of perfect prediction.

While Figure 3 demonstrates that the ED-based prediction tends toward occasional underestimation of national numbers, Figure 4 demonstrates the overall "fit" of the predicted vs. actual Qatar-wide COVID data (the green 45-degree line represents perfect fit and the model's lowess trend is the orange dashed line). We conclude that for COVID, the occurrence of ED peaks should be interpreted as predicting a likely national peak within 2-3 weeks. Furthermore, we found that the reasonably rapid resolution of the ED COVID numbers reliably predicts the national-level resolution within 2-3 weeks. 

Temporal trends and forecasting of COVID-19 hospitalisations and deaths in Scotland using a national real-time patient-level data platform: a statistical modelling study

A review of mathematical model-based scenario analysis and interventions for COVID-19

Mathematical modeling of the SARS-CoV-2 epidemic in Qatar and its impact on the national response to COVID-19

Ministry of Public health (MoPH), the State of Qatar, COVID-19 information

Scatterplot of ED-based model-predicted and actual national COVID n COVID-19 cases presenting to the Emergency Department predict Qatar National COVID-19 trends and numbers