Evaluation of a symptom-based surveillance model for Ebola virus disease at Indira Gandhi International Airport, Delhi, India 2014- 2015

Background Ebola virus disease (EVD) is a zoonotic disease that presents as an acute illness in humans with a case fatality rate ranging from 25% to 90% [1]. From 2014‐15, the EVD outbreak in Guinea, Liberia, Nigeria, and Sierra Leone resulted in 28,616 illnesses and 11,310 deaths [2]. Because of the initial difficulty in containing the EVD outbreak, the World Health Organization declared a Public Health Emergency of International Concern on 8 August 2014 in accordance with the 2005 International Health Regulations [3, 4]. To prevent EVD importation to India by people travelling from affected countries, the Government of India implemented the WHOrecommended screening guidelines for EVD. All passengers from EVD‐affected countries completed a self‐declaration form upon arrival and were screened using thermal scanners at points of entry, including at Indira Gandhi International (IGI) Airport, New Delhi, India. We evaluated the EVD surveillance system IGI Airport from August 2014 to October 2015 to assess the attributes and recommend measures to strengthen the system. Methods The evaluation was conducted at IGI Airport, New Delhi, from January to February 2016. The U.S. Centers for Disease Control and Prevention evaluation framework for public health surveillance was used to evaluate the EVD surveillance system attributes of usefulness, simplicity, flexibility, data quality, acceptability, timeliness, and stability [5]. We interviewed key informants at the Ministry of Health and Family Welfare (MOHFW), National Centre for Disease Control (NCDC), and IGI Airport including directors, immigration officers, medical officers, laboratory and data collection staff. We reviewed surveillance protocols and evaluated the operation of the system via reporting patterns from the airport and NCDC laboratory and oversight mechanism at the MOHFW. We reviewed 312 randomly selected self‐declaration forms, flight records, and suspected case records, and all daily and weekly surveillance reports and laboratory reports from 1 August 2014 to 31 October 2015. Passengers were classified into one of four categories based on the paper self‐declaration form: high risk if the person had history of travel to an EVD‐affected country (i.e., Guinea, Liberia, or Sierra Leone) in the past 21 days and signs or symptoms of EVD; medium risk if the person had history of travel to an EVD‐affected country in the past 21 days and contact with a known EVD case and no signs or symptoms of EVD; low risk if the person had history of travel to an EVD‐affected country in the past 21 days only; and no risk if the person had no history of travel to an EVD‐affected country in the past 21 days. A suspected case of EVD was defined as person who had an elevated body temperature by thermal scan or reported subjective fever or signs or symptoms of EVD (i.e., severe headache, fatigue, muscle pain, vomiting, diarrhoea, abdominal pain, unexplained haemorrhage) and history of travel to an EVD‐affected country in the past 21 days. High risk, suspected cases were referred for examination, Ebola virus testing, and hospital admission. Clinical samples were sent to NCDC, New Delhi, and National Institute of Virology, Pune, for Ebola virus testing. Results There were no non‐stop flights arriving at IGI Airport from the four EVD‐affected West African countries. Between 1 August 2014 and 31 October 2015, 798,756 passengers were screened at IGI Airport. A total of 18,518 (2.3%) reported travel to an EVD‐affected country in the past 21 days. There were 83 high‐risk suspected cases of EVD, 5 medium‐risk passengers, and 18,430 (99.5%) low‐risk passengers. One (1.2%) of the 83 high‐risk suspects had laboratory‐confirmed Ebola virus. The month with the largest number of passengers with an EVD exposure risk was December 2014; the number dramatically declined after March 2015 (Figure 1). The EVD surveillance system was simple and flexible. The selfdeclaration form was concise, uniform, but only available in English. All informants found the form easy to use, including recording passenger temperatures. The MOHFW used the surveillance data collected and analysis to make changes to the system in real time. The number of airlines identified for passenger screening was revised twice during the surveillance period, and the reporting frequency was revised from daily reporting to weekly reporting when the number of passengers at risk of exposure decreased after March 2015. Screening was initially completed after immigration, but it was changed after 2 weeks prior to immigration to make the process more efficient. When the number of passengers increased, the number of screening counters were increased through the deployment of additional medical officers. Staffing was maintained by commissioning medical officers from New Delhi‐area municipal hospitals. Of the 312 self‐declaration forms we reviewed, 98% (n=306) were complete. All 249 daily and 28 weekly reports reviewed were found to be complete for the 17 daily and 10 weekly variables collected, and 100% of reports prepared by IGI Airport were sent to the MOHFW and received within the specified timeframe. All 83 high‐risk suspect case files and laboratory results were complete. The average time for reports from the designated laboratories to IGI Airport was less than 24 hours. Our evaluation had limitations. We were not able to interview cabin crews or airport ground staff due to security and unavailability. Contracted medical officers were not available for interview. Records to assess down time of the thermal camera and computers were not available for review. We could not calculate sensitivity due to lack of a denominator nor could we calculate predictive value positive for cases. Conclusions Overall, the surveillance system implemented to prevent importation of Ebola virus to India was useful for detecting high‐risk, suspected cases. Surveillance allowed for the identification, diagnosis, treatment, and isolation of an imported case of EVD and prevented secondary transmission. The EVD surveillance system was simple, flexible, timely, and captured quality data. The surveillance system was well supported by the IGI Airport Immigration Department and NCDC laboratory. The use of the self‐declaration form in English posed problems for some passengers. During future airport surveillance system implementation, we recommend making forms available in local languages to encourage reporting by passengers.