Predicting COVID-19 incidence in war-torn Afghanistan: A timely response is required!

Dear Editor,

We read with great interest the article entitled "A quick prediction tool for unfavorable outcome in COVID-19 inpatients: Development and internal validation" by Salto-Alejandre et al. Authors of this article forecast the outcomes of COVID-19. The upsurge in COVID-19 cases, along with health issues in the war-torn country, stoked fears of a disaster amid a still-raging conflict. As a result, this paper will focus on the future prediction of COVID-19 cases in the war-torn country of Afghanistan, providing them with additional insight for taking necessary preventive measures.

In Afghanistan, it's important to remember that both natural and human cause the emergence of new diseases. Military conflicts dominated by war and regional tribal and religious struggle have been significantly influenced by wrecking infrastructure and collapsing healthcare systems, among other human activities. People who dwell in the conflicted area will face many short- and long-term impacts. Shortage of facilities, which is generally the case inside congested refugee settlements with sanitation difficulties, greater exposure to disease vectors, and collapse of healthcare infrastructure, contribute to disastrous outcomes when large populations are evacuated. In war-affected country Afghanistan, the COVID-19 pandemic wreaks havoc on the previously fragile public health systems, dramatically undermining the government. In recent weeks, the worrying resurgence has seen an upsurge in COVID-19 cases, with the overall number of cases surpassing 0.15 million and the death toll over 7000 so far. ,

Estimating newly reported cases and expected future ones are essential to sustain the health system's allocation of demand and resources. In both short and long-term situations, well-founded statistical modeling techniques are required to forecast the extent and kind of containment actions taken. The Exponential Smoothing (ES) model is a widely use forecasting model that only assigns positive weights to past and current values. This method is appropriate for forecasting time series with no trend or seasonal pattern. The mean of the ES may be altering slowly over time. It is weighted average measures with weights decreasing exponentially as values become older. ,

The ES formula is given below: F = αA −1 + (1−α) F −1, t > 0 (Where α is the smoothing factor, and 0 < α < 1).

Figs. 1,2 illustrates an estimate of COVID-19 cases and deaths for the following 12 months using month-by-month data (Feb-2020 to Aug-2021). The results of the ES model (Holt's linear trend) indicate that the number of COVID-19 infected cases each month might reach 31,757 (CI 95%: 8017–55,497), while the number of deaths may reach 1572 (CI 95%: 493–2651). According to the anticipated figure, the number of infected cases and deaths will surge over the forthcoming 12 months. Prediction errors were used to assess the model's validity. We evaluated the model's accuracy using the mean absolute error (MAE) and R2 parameters. The ES (Holt's linear trend) month-by-month registered cases model with R-squared (0.62) and MAE (6993.21), as well as the model of the month-by-month death with R-squared (0.65) and MAE (334.79), have been validated and successfully predicted.

Fig. 1

Prediction of the newly infected cases will increase over the forthcoming 12 months. The accuracy and validity of this exponential smoothing (ES) (Holt's linear trend) month-by-month registered cases model with R-squared (0.62) and MAE (6993.21).

Fig. 2

Prediction of the mortality rate will increase over the forthcoming 12 months. The accuracy and validity of this Exponential Smoothing (ES) (Holt's linear trend) month-by-month registered deaths model with R-squared (0.65) and MAE (334.79).

Afghanistan is currently experiencing one of the world's serious medical emergencies, owing to the country's current political unrest state. According to the World Health Organization (WHO), conflicts, environmental catastrophes, and civil and political unrest have resulted in the need for emergency care for more than 3.7 million Afghans. Statistics will assist public health practitioners and officials in understanding the locations, timings, and demographics of newly diagnosed cases. Public health responses to the COVID-19 outbreak and future pandemics may be impacted by accurate and timely infectious disease forecasts, which can help decide the most effective preventative and mitigation methods and when they should be deployed. To prevent the spread of disease, these data can also be utilized to strengthen public health advocacy and the nature and breadth of government programs. Forecasting COVID-19 cases are crucial to ensure healthcare coverage for Afghan people with unknown infectious diseases; otherwise, this pandemic will be more devastating in this impoverished region.

Funding

There is no role of any funding agency related to this study.

Declaration of Competing Interest

The authors declare that there is no conflict of interest or financial disclosure about this publication.

CRediT authorship contribution statement

Usman Ayub Awan: Conceptualization, Visualization, Data curation, Formal analysis, Writing – original draft, Formal analysis, Writing – review & editing. Muhammad Wasif Malik: Conceptualization, Visualization, Data curation, Formal analysis, Writing – review & editing. Muhammad Imran Khan: Writing – original draft, Formal analysis, Writing – review & editing. Aamer Ali Khattak: Conceptualization, Visualization, Data curation, Formal analysis, Writing – review & editing. Haroon Ahmed: Supervision, Writing – review & editing. Usman Hassan: Writing – original draft, Formal analysis, Writing – review & editing. Humera Qureshi: Writing – original draft, Formal analysis, Writing – review & editing. Muhammad Sohail Afzal: Supervision, Writing – review & editing.