This report collates data collected from different sources using varied methodologies. Evidently, there are limitations: (a) sampling bias, e.g., blood donors constitute a convenient sample of self-selected healthy young adults, whereas testing of left-over sera in laboratory included hospitalized or clinic patients predominantly, and (b) incompleteness of data. Careful interpretation can, however, minimize the effects of underlying observational errors. The following is a summary of the seroepidemioologic patterns which can reasonably be deduced:-

• The HBV infection rate has been falling steadily in the last 2 decades.
• From all data collected, the HBsAg prevalence of adults in the general population has been falling. HBsAg prevalence of antenatal mothers is now 10%. It was >11% 10 years ago. The rate in selected healthy young adults is even lower - 7.3% in clients of PMPS, 6.2% in police officers, 5.5% in blood donors and 4% in university students. These figures should be interpreted with care, though, as they reflect prevalence findings in biased samples. For instance, the relatively high carrier rate of HBsAg in antenatal women is believed to have originated from the undue proportion of Mainland Chinese in the antenatal population.
• On implementation of the universal hepatitis B vaccination program for all newborns, the anti-HBs positivity rate of children climbed to >90% in those below age 2 in 1993. In those aged between 5 and 9, the rate was 50%.
• The HBsAg rate of drug users has all along been higher than that of other population groups. However, in line with general local pattern, there has been a drop of HBsAg prevalence in IVDU too, from 16.8% in 1985 to 8.8% in 1996. The rate of infection was lower in those with less than 5-year history of using drugs. One other contributing factor may have been the anti-AIDS education campaign against needle-sharing.
• The notification system confirms that HAV infection is punctuated by outbreaks occurring every few years. Prominent outbreaks occurred in 1972, 1975, 1988 and 1992.
  

  Year	HAV cases	total hepatitis cases
  1972	n/a	729
  1975	n/a	1761
  1988	1187	2398
  1992	3626	4357

  table 26- outbreaks of hepatitis

• A downward shift is noted in the age-stratified anti-HAV prevalence graphs. Whereas more than 90% of people aged over 40 carry anti-HAV both now and 15 years ago, the age at which infection occurs has been steadily rising over the years. Among people below the age of 20, only 15% in 1993, as opposed to 57% in 1978, have evidence of previous exposure to the virus.
• Blood donors constitute a convenient group of healthy young adults in the community for studying the prevalence of HCV infection. The overall prevalence is <0.5%, a figure which is supported by epidemiologic studies.
• The prevalence of anti-HCV is high in IVDUs and hemophiliacs in Hong Kong.
• HDV infection is uncommon in the general public. Its prevalence in the IVDU population has also fallen from 63.1% in 1985 to 28.8% in 1992/93.
• The data on HEV infection are limited. Available data do suggest the existence of the virus in the local community.