Vaccines for preventing hepatitis B in health‐care workers



Hepatitis B virus (HBV) causes acute and chronic liver diseases. Hepatitis B vaccination is recommended for health‐care workers.


To assess the beneficial and harmful effects of hepatitis B vaccination in health‐care workers.

Search methods

We searched the trial registers of The Cochrane Hepato‐Biliary Group, The Cochrane Library, MEDLINE, and EMBASE to February 2003.

Selection criteria

Randomised trials comparing any dose, injection route, injection site, or schedule of hepatitis B plasma‐derived vaccines (PDV) or recombinant vaccines (RV) versus placebo, no intervention, or another hepatitis B vaccine in health‐care workers.

Data collection and analysis

Two reviewers extracted the data independently. The reviewers assessed the methodological quality of the trials regarding generation of the allocation sequence, allocation concealment, double blinding, and follow‐up. The results were presented as relative risk (RR) with 95% confidence intervals (CI).

Main results

We identified 21 randomised trials, all with one or more methodological weaknesses. Four trials demonstrated that PDV versus placebo significantly decreased hepatitis B events at maximum follow‐up (RR 0.51, 95% CI 0.35 to 0.73). RV did not differ significantly from PDV in eliciting a protective hepatitis B surface antibody (anti‐HBs) level in two trials. Both vaccines were well tolerated. Low‐dose vaccine (1 or 2 µg) by the intradermal route resulted in significantly more participants without protective anti‐HBs level compared with high‐dose (10 or 20 µg) by the intramuscular route (RR 1.41, 95% CI 1.13 to 1.76). The intradermal route caused significantly more local adverse events, while the intramuscular route caused significantly more systemic adverse events. The gluteal injection produced significantly more participants without protective anti‐HBs level than the deltoid injection. The prevalence of anti‐HBs seroconversion by rapid vaccination (0, 1, and 2 months) was significantly lower than that by standard vaccination (0, 1, and 6 months). Booster vaccinations with different RV doses (2.5, 5, 10, 20, or 40 µg) produced similar prevalence of anti‐HBs seroconversion in three trials assessing participants who did not respond to previous HBV vaccination.

Authors' conclusions

PDV significantly prevents hepatitis B events. RV seems to be able to elicit similar protective anti‐HBs levels. The intramuscular route with 20 µg RV was significantly more effective compared with the intradermal route with 2 µg RV as was the standard schedule compared with a rapid schedule and deltoid intramuscular injection compared with the gluteal intramuscular injection. It is unclear if booster vaccination of non‐responders offers higher anti‐HBs seroconversion and hepatitis B vaccine prevents the infection of hepatitis B mutants in health‐care workers.


Wendong Chen, Christian Gluud


Plain language summary

Hepatitis B vaccine prevents hepatitis B infection in health‐care workers

Plasma‐derived vaccine significantly prevents hepatitis B events in health‐care workers. Recombinant vaccine does not significantly differ from plasma‐derived vaccine in eliciting protective hepatitis B surface antibody levels. Both vaccines are well tolerated. The intramuscular route causes significantly more systemic adverse events, and the intradermal route causes significantly more local adverse events. The deltoid injection seems more effective than the gluteal injection in eliciting antibodies. The standard vaccination schedule (0, 1, and 6 months) elicits a better antibody response than a rapid vaccination schedule (0, 1, and 2 months). It is unknown whether hepatitis B vaccine protects health‐care workers from infection of mutated hepatitis B virus.


Wendong Chen, Christian Gluud

Reviewer's Conclusions

Authors' conclusions 

Implications for practice 

‐ PDV significantly decreased the hepatitis B events without severe adverse events. - ‐ There was no evidence to support or dispute the effects of RV to decrease hepatitis B events. - ‐ RV seems to elicit protective anti‐HBs level in health‐care workers comparable to PDV. - ‐ The intradermal route with 2 µg vaccine was not able to elicit the same anti‐HBs response as the intramuscular route with 20 µg vaccine. - ‐ The gluteal intramuscular injection was not able to elicit the same anti‐HBs response as the deltoid intramuscular injection. - ‐ The rapid vaccination schedule (0, 1, and 2 months) was not able to elicit the same anti‐HBs response as the standard vaccination schedule (0, 1, and 6 months). - ‐ Booster vaccination with high dose vaccine did not work well in non‐responders and the proportion with anti‐HBs seroconversion was not significantly associated with the vaccine dose. - ‐ The same type of vaccines produced by different companies or in different countries did not differ significantly regarding anti‐HBs seroconversion. - ‐ Vaccinated health‐care workers might still face the risk of HBV mutant infection.

Implications for research 

‐ No randomised trials have compared RV versus PDV for the prevention of hepatitis B events in long‐term follow‐up. - ‐ Long‐term persistence of anti‐HBs elicited by RV and PDV needs to be assessed. - ‐ No randomised trials have assessed the intradermal route with a relative higher dose of vaccine, eg, 5 µg. - ‐ A new type of vaccine (with the new adjuvant MF59) needs to be studied in naive as well as non‐responding health‐care workers. - ‐ The current HBV vaccines need to be evaluated in preventing HBV mutant infection among health‐care workers. - ‐ We found relatively poor standards of trial reporting. Initiatives to improve reporting, such as CONSORT (www.consort‐, should be followed in the reporting of future trials.

Get full text at The Cochrane Library