The early warning and response systems in Syria: A functionality and alert threshold assessment

Abstract

Objectives: this study aims to provide an updated evaluation of the functional characteristics of the two Early Warning Systems (EWS) in Syria, EWARS (Early Warning, Alert, and Response System) and EWARN (Early Warning, Alert, and Response Network), and to test different alert threshold methods using World Health Organization guidelines against the data of selected diseases.

Methods: A retrospective analysis of EWARN and EWARS surveillance data assessed functional characteristics. The World Health Organization alert thresholds for measles, acute bloody diarrhea, acute jaundice syndrome, and severe acute respiratory infections were tested using three methods. Sensitivity, specificity, and Youden index determined threshold suitability for each syndrome.

Results: The annual average number of reported cases was 1,140,717 for EWARS and 10,189,415 for EWARN. This study found that the optimal alert thresholds varied among different diseases. The percentile method showed promising results with good sensitivity and specificity. For measles, the 85^th percentile threshold had the best results (Youden index = 0.443), whereas for acute bloody diarrhea, it was 75^th percentile (Y = 0.532) and for severe acute respiratory infections, it was 90^th percentile (Y = 0.653).

Conclusions: This study supports the use of adaptable disease-specific alert thresholds such as the percentile approach. Further research is required to develop statistical methods that can be applied to various early warning systems in conflict contexts.

Keywords: Alert threshold; EWARN; EWARS; Early warning system; Syndromic surveillance, conflict; Syria.

Figure 1

Population size and geographical coverage of the two Early Warning Systems in Syria until November 2024; Aleppo, and Damascus are the most populated cities. The areas shaded green are government-controlled areas with EWARS coverage and shaded blue (northeast Syria) and red (northwest Syria) with EWARN coverage. The map was created using Data Wrapper in June 2024.

Figure 2

Comparison of EWARS and EWARN between 2012 and 2023 regarding the number of sentinel sites, number of annual cases, completeness, and timeliness.

Figure 3

Visual representation of the tested threshold against the four diseases in Azaz district 2023, where the red line represents the threshold produced using different method, the blue line represents the total number of cases, and the gray area represents the lag window between the first alert and the peak of the cases. Also, an ROC curve was produced for each of the thresholds. ABD, acute watery diarrhea; AJS, acute jaundice syndrome; ROC, receiver operating characteristic; SARI, severe acute respiratory infection.