In tropical Hong Kong (HK), influenza virus has both winter and summer peaks while Respiratory Syncytial Virus (RSV) has a peak starting in spring and extending into the summer. They lead to increase in hospital admissions and visits to family clinics (FCs). Viral culture results (VCT) correlate well with the intensity of clinical services but they are not available real time. Rapid viral tests results (RVT) are available on the same day and can be used for real time predictions. This study reports for a period of 145 weeks (2.8 yrs beginning January 2007) showing close correlation of the RVT with culture results and also with the intensity of clinical services. A summary of the RVT are reported weekly to all public hospitals and two acute hospitals successfully use them to manage the winter influenza surge of 2009.
To evaluate the use of the RVT results in predicting hospital admissions and FCs visits and using these predictions to manage the winter influenza surge in hospital admissions.
All specimens for viral culture in HK are submitted to the government Public Health Laboratory Centre (PHLC) and the numbers for influenza and RSV are obtained from PHLC. There are 41 public hospitals accounting for 91% of hospital beds in HK. Rapid viral test by immunofluorence is done in five hospital laboratories for admissions with pneumonia. Since May 09, PCR is also available for the diagnosis of nH1N1. A summary of these RVT are reported weekly real time to all public hospitals. Admissions data is obtained from the public hospital computer system and the visits to FCs are from 104 sentinel clinics that reports workload to the government.
The two hospitals when alerted by the summary of the RVT will initiate a program including rapid discharges to long-term care facilities, doctor’s visits to old age homes to reduce hospital admissions and emergency room (ER) doctors having rights of admission to any ward with available beds.
The RVT results (% positive) correlates closely with the number of weekly positive cultures of the PHLC for the 145 weeks (r = 0.85; p < 0.001). There is significant correlation of RVT for influenza and admissions for children (Influenza A: r = 0.94, p < 0.001) and the elderly (Influenza A: r = 0.77, p < 0.001). They are also significantly correlated with FCs visits (r = 0.72, p < 0.001). The RVT for RSV correlates with admissions for acute bronchiolitis (r = 0.63, p < 0.001).
During the winter surge, % of ER admissions in the 15 acute hospitals is similar at 12%. The two acute hospitals with surge management have an occupancy rate of 81% and 84% while the others have a mean of 105% (p<0.001, chi-square test).
The RVT is a good predictor of clinical services intensity and real time feedback can be effective in helping to manage the surge in clinical services. This should be the first study showing the effective use of rapid viral tests results for managing clinical services.