The feasibility of outpatient data used for respiratory syndromic surveillance in general hospital / 中华流行病学杂志

ABSTRACT

Objective To find out the data sources of respiratory syndromes and their components from the outpatients of general hospitals and to describe the time distribution and mutual relations of different respiratory syndromes. Feasibility of respiratory syndromes used for early warning surveillance on respiratory infectious disease was also under research. Methods Retrospective investigation on Hospital Information System (HIS) was implemented in a general hospital in Guangzhou, 2005, and data of outpatients was collected and classified into different syndromes. The respiratory syndromes with its time distribution similar to influenza like illness (ILI),were selected, and cross-correlation analyses were conducted to inveshgate the feasibility of respiratory syndromes for early warning surveillance on respiratory infection diseases (influenza as an example). Results Primary sub-classification of respiratory syndromes in outpatient department would include upper respiratory infection(URI)(51.20%), trachitis/bronchitis (18.80%), asthma ( 17.52% ), etc. Pulmonary infection accounted for only 2.26%. Time distributions of URI, trachitis/bronchitis, pulmonary infection, cough and asthma in outpatient department, X-ray tests and pneumonia/acute respiratory distress syndromes (ARDSs) in outpatient X-ray room were similar, with two peaks observed. Cross-correlation functions were calculated with the data sets of 1st-28th week.The most significant correlation was detected between the time series of outpatient pulmonary infections and ILIs moved 4 weeks backward (r＝0.739, P＜0.01 ), and that was detected between URIs and ILIs moved 5 weeks backward (r=0.714, P＜0.01 ). Correlation between X-ray tests,pneumonia/ARDSs in outpatient X-ray room and ILIs was the strongest when ILIs time series moved 1 week backward (r=0.858, P＜0.001; r=0.821, P＜0.001 ). Conclusion Outpatient data from HIS system in general hospital could be applied to syndromic surveillance on respiratory diseases. For early warning epidemics or outbreaks of influenza or other respiratory infectious diseases, data of outpatient pulmonary infection appeared to be the most feasible for its specificity and timeliness, followed by URI and cough. X-ray tests and pneumonia/ARDSs in outpatient X-ray findings were important supplementary to verify the respiratory disease epidemics or outbreaks for its good specificity, but with no advantage for early warning.