BACKGROUND: Differences in sexual behaviour are partly responsible for significant inequalities in human papillomavirus (HPV)-related diseases between sub-populations. Our aim was to illustrate how differential HPV vaccine uptake by sexual behaviour can impact population-level vaccination effectiveness and inequalities in HPV prevalence. METHODS: We used a multi-type individual-based transmission-dynamic model of HPV infection. The modelled population is stratified into 4 sexual activity levels (low=L0, high=L3). The model was calibrated to Canadian epidemiological data, including the proportion of sexually active individuals by gender and type-specific HPV prevalence by level of sexual activity. Vaccine uptake by sexual activity level was varied in one-way sensitivity analyses. Uncertainty in model predictions is presented using the median [10-90th percentiles] of simulations. RESULTS: In our pre-vaccination calibrated modelled population, the most sexually active risk groups (L2-L3) account for 43% of prevalent HPV-16/18 infections while only representing 20% of the population. Under base case assumptions (age at vaccination=12 years, vaccine efficacy=100%, average duration of protection=20 years, overall coverage=50%), vaccinating girls is predicted to reduce HPV-16/18 prevalence by 56% [46-69] at equilibrium (70 years post-vaccination) when uptake is uniform across sexual activity levels, and by 35% [28-40] when uptake is maldistributed (L0: 100%, L1: 55%, L2: 0%, L3: 0%). At uniform 50% vaccination coverage across risk groups, HPV-16/18 prevalence is predicted to be reduced by 62% [52-73], 74% [57-82], 40% [24-60], and 28% [7-54] at equilibrium among women in levels of sexual activity L0, L1, L2 and L3, respectively. CONCLUSIONS: A low vaccine uptake in girls at highest risk of future HPV acquisition may substantially limit population effectiveness of vaccination. Vaccination effectiveness is lower in more sexually active groups due to smaller herd effects. Uniform vaccination coverage across sub-populations may not be able to decrease existing inequalities in HPV infection and disease unless coverage is high enough to produce herd effects in higher risk groups.
BACKGROUND: Differences in sexual behaviour are partly responsible for significant inequalities in human papillomavirus (HPV)-related diseases between sub-populations. Our aim was to illustrate how differential HPV vaccine uptake by sexual behaviour can impact population-level vaccination effectiveness and inequalities in HPV prevalence. METHODS: We used a multi-type individual-based transmission-dynamic model of HPV infection. The modelled population is stratified into 4 sexual activity levels (low=L0, high=L3). The model was calibrated to Canadian epidemiological data, including the proportion of sexually active individuals by gender and type-specific HPV prevalence by level of sexual activity. Vaccine uptake by sexual activity level was varied in one-way sensitivity analyses. Uncertainty in model predictions is presented using the median [10-90th percentiles] of simulations. RESULTS: In our pre-vaccination calibrated modelled population, the most sexually active risk groups (L2-L3) account for 43% of prevalent HPV-16/18 infections while only representing 20% of the population. Under base case assumptions (age at vaccination=12 years, vaccine efficacy=100%, average duration of protection=20 years, overall coverage=50%), vaccinating girls is predicted to reduce HPV-16/18 prevalence by 56% [46-69] at equilibrium (70 years post-vaccination) when uptake is uniform across sexual activity levels, and by 35% [28-40] when uptake is maldistributed (L0: 100%, L1: 55%, L2: 0%, L3: 0%). At uniform 50% vaccination coverage across risk groups, HPV-16/18 prevalence is predicted to be reduced by 62% [52-73], 74% [57-82], 40% [24-60], and 28% [7-54] at equilibrium among women in levels of sexual activity L0, L1, L2 and L3, respectively. CONCLUSIONS: A low vaccine uptake in girls at highest risk of future HPV acquisition may substantially limit population effectiveness of vaccination. Vaccination effectiveness is lower in more sexually active groups due to smaller herd effects. Uniform vaccination coverage across sub-populations may not be able to decrease existing inequalities in HPV infection and disease unless coverage is high enough to produce herd effects in higher risk groups.
Authors: Helen C Johnson; Erin I Lafferty; Rosalind M Eggo; Karly Louie; Kate Soldan; Jo Waller; W John Edmunds Journal: Lancet Public Health Date: 2017-12-19