ObjectiveTo achieve autoscreen and intellectual judgement on suspected healthcareassociated infection(HAI) cases in the realtime HAI surveillance system. MethodsThe screening strategy and prewarning condition of suspected HAI cases were formulated according to Diagnostic Criteria of Healthcareassociated Infections and experience in manually looking up HAI in medical records. The system can extract infectionrelated information from hospital information system (HIS), lab information system (LIS), radiology information system (RIS), and electronic medical record (EMR) system. All the information can be sorted, integrated and analyzed according to the correlativity between extracted information and development of infection. Then the prewarning information of suspected infection cases are output according to different departments. The sensitivity and correctness of the system is validated through comparing with the result of retrospective study by manually looking up medical records.Results30-40 suspected HAI cases in the hospital with 3 600 beds are found by the system everyday, one infection management professional can deal with all suspected cases between 1 and 1.5 hours; then all the infection data can be obtained automatically.Comparing with the result of retrospective study by manually looking up medical records, HAI cases confirmed by the system increases by 20%.ConclusionRealtime HAI surveillance system has the features of timeliness, high efficiency, and correctness. Autoscreen and intellectual judgement can save time for infection management professionals to screen and diagnose infection, and improve efficiency and effectiveness in preventing and controlling infection.