Improving disease surveillance capabilities through a public health information affinity domain

Abstract

In March 2009, deaths from influenza like illness began to mount in Mexico and the United States. In April, the National Respiratory Disease Institute [1] reported a 100 percent increase in patients checking in for atypical pneumonia. Samples were sent to laboratories in Canada and the United States. On April 17, a national influenza alert was announced; on April 23, the new influenza virus was officially recognized [2]. In Mexico City, where a three week surge of influenza accounted for 90,000 visits in 220 health units and 20 hospitals [3], Mayor Marcelo Ebrard ordered the temporary closure of schools and commercial establishments. In response to the pandemic, IBM collaborated with the Ciudad de México Gobierno del Distrito Federal Secretaria de Salud del Distrito Federal (Secretaria de Salud of GDF) [4] on software to standardize influenza reporting and improve situational awareness [5]. Secretaria de Salud of GDF installed IBM's Public Health Information Affinity Domain (PHIAD) that uses Health Information Exchange technology and standards to enable rapid data sharing of clinical surveillance data with public health officials [6]. De-identified 2009 H1N1 positive laboratory results from the Institute of Epidemiological Reference and Diagnosis [7] were provided by the Secretaria de Salud of GDF for import into PHIAD; each record was transformed into a standard Integrating the Healthcare Enterprise XD-LAB document. The standardized data was analyzed with Spatiotemporal Epidemiological Modeler (STEM), an open source software framework for infectious disease modeling and forecasting [8, 9]. Using STEM, we made quantitative measures of the policy effects of school and commercial closures in Mexico City on the transmission rate of the 2009 H1N1 virus. © 2010 ACM.