Human adenovirus infection is a common cause of upper and lower respiratory illness and easily breaks out in schools and the army. In an adenovirus epidemic, a large number of samples would be collected for laboratory diagnosis, and it is urgent to optimize the current sampling strategy. We researched the application of laboratory detection in the adenovirus epidemic and optimized the range of laboratory pathogen detection in the adenovirus epidemic by summarizing previous theoretical achievements, research reports, and experts' opinions and by using mathematical model tools. Under certain assumptions, a susceptible-infectious-quarantined-recovered (SIQR) model was established to describe the adenovirus epidemic and optimize the range of laboratory pathogen detection. Some standards and implementation rules suggest that when the number of cases is less than 10 or 20, all patients should be sampled for laboratory examination, and when the number of cases is more than 10 or 20, at least 10 or 20 samples should be collected. In practice, the sampling range can be appropriately expanded. A total of 21 studies were analyzed, and the sampling rate of adenovirus infection was 31% (95%CI: 24%~38%). The mathematical model suggested that the screening of asymptomatic people in the latent stage can slow the spreading of the epidemic, but the detection range will be too large. These findings may be helpful for policymaking during an adenovirus epidemic and to avoid proceeding with laboratory testing blindly. Furthermore, it may also provide some guidance for optimizing the sampling strategy of other diseases, especially for respiratory tract infections.
Keywords: Human adenovirus; mathematical model; meta-analysis; sample size; systematic review.