The Pakistan National Emergency Department Surveillance Study (Pak-NEDS): Introducing a pilot surveillance

BACKGROUND Evidence-based decision making is essential for appropriate prioritization and service provision by healthcare systems. Despite higher demands, data needs for this practice are not met in many cases in low- and middle-income countries because of underdeveloped sources, among other reasons. Emergency departments (EDs) provide an important channel for such information because of their strategic position within healthcare systems. This paper describes the design and pilot test of a national ED based surveillance system suitable for the Pakistani context. METHODS The Pakistan National Emergency Department Surveillance Study (Pak-NEDS) was pilot tested in the emergency departments of seven major tertiary healthcare centres across the country. The Aga Khan University, Karachi, served as the coordinating centre. Key stakeholders and experts from all study institutes were involved in outlining data needs, development of the study questionnaire, and identification of appropriate surveillance mechanisms such as methods for data collection, monitoring, and quality assurance procedures. The surveillance system was operational between November 2010 and March 2011. Active surveillance was done 24 hours a day by data collectors hired and trained specifically for the study. All patients presenting to the study EDs were eligible participants. Over 270,000 cases were registered in the surveillance system over a period of four months. Coverage levels in the final month ranged from 91-100% and were highest in centres with the least volume of patients. Overall the coverage for the four months was 79% and crude operational costs were less than $0.20 per patient. CONCLUSIONS Pak-NEDS is the first multi-centre ED based surveillance system successfully piloted in a sample of major EDs having some of the highest patient volumes in Pakistan. Despite the challenges identified, our pilot shows that the system is flexible and scalable, and could potentially be adapted for many other low- and middle-income settings.