Background The 2001 National Strategy for Sexual Health and HIV recommended 90% uptake of Hepatitis B (HBV) vaccine in non-immune MSM at first GUM clinic attendance. The HepB3 Survey reported 95% uptake in 2008 but recent surveillance using GUM Clinic Activity Dataset-v2 (GUMCADv2) coding shows <20% uptake. A detailed regional audit was designed to investigate this apparent drop in coverage.

Aim To determine HBV vaccination coverage in ‘first-attendee’ MSMs.

Methods All MSM ‘first-attendees’ at our service between January-March 2014 were identified. Patient records were reviewed for HBV screening, vaccine-offer, vaccine-uptake, HIV testing and coding accuracy up to 18 months from first-attendance. MSM were deemed ‘immune’ if surface antibody (sAb) >10 mIU/ml, core antibody positive, or self-reported vaccination-status was ‘Fully-vaccinated’ and no serology was done; and eligible for vaccination if sAb ≤10 mIU/ml, or if they reported ‘Partially-vaccinated’, ‘Never-vaccinated’, or ‘Don’t know’ and no serology was done.

Results We identified 115 MSM ‘first-attendees’ (13 HIV+). 41% only attended once. Regarding vaccination-status: 41/95 (43%) reported ‘Fully-vaccinated’, 29/95 (30%) ‘Partially-vaccinated’, 12/95 (13%)’Never-vaccinated’, 11/95 (12%) ‘Don’t-know’, 1/95 (1%) ‘Chronic-HBV’ and 1/95 (1%) ‘Cleared-HBV’. 48/103 (47%) were deemed immune and 46/103 (45%) eligible. 36/46 (76%) of eligibles were offered vaccination; 2/36 (6%) declined, reporting ‘not at risk’. 3/32 (9%) who accepted vaccination pending sAb levels did not return for it. 31/46 (67%) of eligibles received ≥1 dose of vaccine, 28/46 (61%) within 42 days of first-attendance. Reasons for non-offer were not recorded. 75% of first-doses were coded. Only 15% of ‘immune’ patients were coded as such (P2I). HIV-test uptake was 99% and coding accuracy was 97%.

Discussion We found below-target levels of HBV vaccination-coverage and incomplete coding of immunity/vaccination. Failure to code P2I for ‘immunes’ will increase the apparent ‘eligibles’ denominator in GUMCADv2 algorithms, generating incorrectly low vaccination-coverage figures. Reduced offer-rate may contribute to low vaccination-coverage and should be reviewed locally. Further regional audits may be required. Significant improvements in coding are essential for accurate surveillance of HBV vaccination-coverage using GUMCADv2.