Abstract:

Objective

To compare the difference of data quality and early warning effectiveness of infectious diseases between manual attendance and intelligent attendance， in order to provide empirical reference for the construction of intelligent infectious disease syndromic surveillance.

Methods

Two primary schools A and B in a city were selected to collect absenteeism data for the 2021-2022 school year by the two methods： Face recognition attendance （standard： absence time ≥1 hour）， the all-cause absenteeism rates of grades 1-2 （DARL）， grades 3-6 （DARH） and the whole school （DARW1） were collected； School doctor attendance （standard： absent for a whole day）， the all-cause （DARW2） and sickness （DARW3）absenteeism rate of the whole school were collected.

Results

DARW2 and DARW3 comprised 32.6% and 25.2% of DARW1， respectively， and DARW3 comprised 77.3% of DARW2.In School A， DARW1 was significantly correlated with DARW2 （r=0.256， P＜0.001）， DARW1 was significantly correlated with DARW3（r=0.243， P＜0.001）， and DARW2 was significantly correlated with DARW3 （r=0.954， P＜0.001）； In School B，DARW1 and DARW2 （r=0.800，P＜0.001）， DARW1 and DARW3 （r=0.790，P＜0.001）， and DARW2 and DARW3 （r=0.964，P＜0.001） were also significantly related.The early warning sensitivity of DARL， DARH，DARW1， DARW2 and DARW3 was 97.0%， 95.3%， 100%， 100%， 100%， the specificity was 88.5%， 91.7%，81.8%， 80.3%， 78.4%， and the Yoden index was 85.5%， 87.0%， 81.8%， 80.3% and 78.4%， respectively.

Conclusions

Compared with manual attendance， smart attendance-based surveillance system has higher data quality and thus better surveillance accuracy.Moreover， adjusting absenteeism time and calculating absenteeism rate by grade segment could further improve the effectiveness of smart attendance-based surveillance system in early warning of infectious disease outbreaks.

Key words: School - age children, Absenteeism, Smart attendance, Infectious diseases, Syndromic surveillance