Evaluating a surveillance system: live-bird market surveillance for highly pathogenic avian influenza, a case study

INTRODUCTION: Highly pathogenic avian influenza H5N1 was first reported in poultry in Nigeria in February 2006. The only human case that occurred was linked to contact with poultry in a live bird market (LBM). LBM surveillance was instituted to assess the degree of threat of human exposure to H5N1. The key indicator was detection of H5N1 in LBMs. We evaluated the surveillance system to assess its operations and attributes. METHODS: We used the updated US Centers for Disease Control and Prevention (CDC) updated guidelines for evaluating public health surveillance systems. We reviewed and analyzed passive surveillance data for HPAI (January 2006-March 2009) from the Avian Influenza National Reference Laboratory, and live bird market surveillance data from the Food and Agriculture Organization of the United Nations, Nigeria. We interviewed key stakeholders and reviewed reports of live bird market surveillance to obtain additional information on the operations of the system. We assessed the key system attributes. RESULTS: A total of 299 cases occurred in 25 (72%) states and the Federal Capital Territory (FCT). The system detected HPAI H5N1 virus in 7 (9.5%) LBMs; 2 (29%) of which were from 2 (18.2%) states with no previous case. A total of 17,852 (91.5%) of samples arrived at the laboratory within 24 hours but laboratory analysis took over 7 days. The sensitivity and positive predictive value (PPV) were 15.4% and 66.7% respectively. CONCLUSION: The system is useful, flexible, complex and not timely, but appears to be meeting its objectives. The isolation of HPAI H5N1 virus in some of these markets is an indication that the markets are possible reservoirs of the virus in Nigeria. We recommend that the Federal Government of Nigeria should dedicate more funds for surveillance for HPAI as this will aid early warning and reduce the risk of a pandemic.