Public health impact and cost-effectiveness of gonorrhoea vaccination: an integrated transmission-dynamic health-economic modelling analysis

Abstract

Background: Gonorrhoea is a rapidly growing public health threat, with rising incidence and increasing drug resistance. Evidence that the MeNZB and four-component serogroup B meningococcal (4CMenB) vaccines, designed against Neisseria meningitidis, can also offer protection against gonorrhoea has created interest in using 4CMenB for this purpose and for developing gonorrhoea-specific vaccines. However, cost-effectiveness, and how the efficacy and duration of protection affect a gonorrhoea vaccine's value, have not been assessed.

Methods: We developed an integrated transmission-dynamic health-economic model, calibrated using Bayesian methods to surveillance data (from the Genitourinary Medicine Clinic Activity Dataset and the Gonococcal Resistance to Antimicrobials Surveillance Programme) on men who have sex with men (MSM) in England. We considered vaccination of MSM from the perspective of sexual health clinics, with and without vaccination offered to all adolescents in schools (vaccination before entry [VbE]), comparing three realistic approaches to targeting: vaccination on attendance (VoA) for testing; vaccination on diagnosis (VoD) with gonorrhoea; or vaccination according to risk (VaR), offered to patients diagnosed with gonorrhoea plus individuals who test negative but report having more than five sexual partners per year. For the primary analysis, vaccine impact was assessed relative to no vaccination in a conservative baseline scenario wherein time-varying behavioural parameters (sexual risk behaviour and screening rates) stabilise. To calculate the value of vaccination per dose administered, the value of vaccination was calculated by summing the averted costs of testing and treatment, and the monetary value of quality-adjusted life-year (QALY) gains with a QALY valued at £20 000. Costs were in 2018-19 GB£, and both costs and QALYs were discounted at 3·5% per year. We analysed the effects of varying vaccine uptake (0·5, 1, or 2 times HPV vaccine uptake by MSM in sexual health clinics in England), vaccine efficacy (1-100%) and duration of protection (1-20 years), and the time-horizon considered (10 years and 20 years). In addition, we calculated incremental cost-effectiveness ratios for the use of 4CMenB using assumed vaccine prices.

Findings: VbE has little impact on gonorrhoea diagnoses, with only 1·7% of MSM vaccinated per year. VoA has the largest impact but requires more vaccine doses than any other strategy, whereas VoD has a moderate impact but requires many fewer doses than VoA. VaR has almost the same impact as VoA but with fewer doses administered than VoA. VaR is the most cost-effective strategy for vaccines of moderate efficacy or duration of protection (or both), although VoD is more cost-effective for very protective and long-lasting vaccines. Even under conservative assumptions (efficacy equivalent to that of MeNZB and protection lasting for 18 months after two-dose primary vaccination and 36 months after single-dose booster vaccination), 4CMenB administered under VaR would likely be cost-saving at its current National Health Service price, averting an estimated mean 110 200 cases (95% credible interval 36 500-223 600), gaining a mean 100·3 QALYs (31·0-215·8), and saving a mean £7·9 million (0·0-20·5) over 10 years. A hypothetical gonorrhoea vaccine's value is increased more by improving its efficacy than its duration of protection-eg, 30% protection lasting 2 years has a median value of £48 (22-85) per dose over 10 years; doubling efficacy increases the value to £102 (53-144) whereas doubling the duration of protection increases it to £72 (34-120).

Interpretation: We recommend that vaccination of MSM against gonorrhoea according to risk in sexual health clinics in England with the 4CMenB vaccine be considered. Development of gonorrhoea-specific vaccines should prioritise maximising efficacy over duration of protection.

Funding: Medical Research Council (UK), National Institute for Health Research (UK).

Figure 1

Model-structure diagram of the epidemiology of gonorrhoea and vaccination The population is divided into compartments representing different states of infection, with changes of state occurring through various processes. Individuals entering the sexually active population are uninfected, with those who receive adolescent vaccination (under the vaccination-before-entry strategy) entering the vaccine-protected stratum (1) and the remainder entering the unvaccinated stratum (2). When vaccine protection wanes, individuals are no longer protected and move from the relevant compartment in the vaccine-protected stratum to the corresponding compartment in the vaccine protection waned stratum (3). Individuals who become infected (4) pass through an incubating state, before either developing symptoms (5) and entering the symptomatic infection state, or remaining asymptomatic (6) and entering the asymptomatic infection state. Symptomatic individuals seek treatment (7) and enter the treatment state. Asymptomatic infections can be identified through screening followed by treatment (8), with individuals entering the treatment state, or there can be natural recovery (9), returning individuals to the uninfected state. All treated infections are cured (10, 11), with individuals who are vaccinated (under the vaccination-on-diagnosis or vaccination-on-attendance strategies) entering the vaccine-protected stratum (10) and those who are not vaccinated remaining in their current stratum (11)—infection does not confer natural immunity and recovered individuals are as susceptible as those never infected. Uninfected individuals who are vaccinated when screened (under the vaccination-on-attendance strategy) enter the uninfected state of the vaccine-protected stratum (12). Individuals leave the sexually active population through ageing from any state (13). Note that there are separate sets of compartments for groups of individuals with low sexual activity (darker grey layer) and high sexual activity (lighter grey layer), which have identical arrangements of compartments and vaccination-status strata; for clarity only flows in and out of the upper layer (the high sexual activity group) are shown.

Figure 2

Simulations of gonorrhoea transmission over time in MSM in England, under different vaccination strategies Simulations are based on use of a vaccine providing 40% protection for 4 years, with vaccine uptake at the level of HPV vaccine uptake by MSM in sexual health clinics in England (ie, 33·0% [95% CI 32·7–33·3]). (A) Annual gonorrhoea diagnoses (note that the lines for the VaR and VoA strategies almost overlap). (B) Annual vaccine doses administered. (C) Cumulative value of vaccination per dose administered in sexual health clinics; note that there is no value calculated for the VbE strategy because we conducted the analysis taking the perspective of sexual health clinics and VbE is not provided by sexual health clinics. For panels A–C, lines represent medians and shaded regions represent 95% credible intervals. Note that panels A and B show undiscounted numbers while panel C shows discounted £ values. (D) Probability that each strategy is the most cost-effective over 20 years for vaccines ranging in efficacy (1–100%) and duration of protection (1–20 years); in all cases, either VoD or VaR is the most cost-effective strategy, and the dashed contour line shows where the two strategies have equal probability of being the most cost-effective, while the solid contour lines show where either the VoD strategy (upper right) or the VaR strategy (lower left) has a 95% probability of being the most cost-effective. MSM=men who have sex with men. VbE=vaccination before entry. VoD=vaccination on diagnosis. VaR=vaccination according to risk. VoA=vaccination on attendance.

Figure 3

Effects of vaccine uptake, efficacy, and duration of protection on the impact and cost-effectiveness of three vaccination strategies against gonorrhoea in men who have sex with men in England over 10 years (A) Total cases averted by vaccination. (B) Total number of vaccine doses administered. (C) Value of vaccination per dose administered. Columns of panels correspond to the three targeting strategies assessed (VoD, VaR, and VoA). Three levels of efficacy (20%, 40%, or 80%) and three durations of protection (2, 4, or 8 years) are represented along the x-axis. Different coloured bars correspond to three different levels of uptake: the low-uptake scenario (16·5% [95% CI 16·3–16·7]), which is half the level of HPV vaccine uptake by MSM in sexual health clinics in England (ie, 33·0% [32·7–33·3], labelled intermediate on the graph), and one-quarter the level of the high-uptake scenario (66·0% [65·4–66·6]). Points show medians and bars show 95% credible intervals. Note that panels A and B show undiscounted numbers while panel C shows discounted £ values. Also note that in panel B, the dashed line in the VoA graph shows the upper limit of the y-axis scales of the VoD and VaR graphs, and in panel C, the dashed lines in the VoD and VaR graphs show the upper limit of the y-axis scale of the VoA graph. VoD=vaccination on diagnosis. VaR=vaccination according to risk. VoA=vaccination on attendance.

Figure 4

Value of vaccination against gonorrhoea under three targeting strategies among men who have sex with men in England over 10 years (A) Heatmaps of median value per dose administered of vaccines providing 1–100% protection against infection for 1–20 years, with contour lines showing values of £50, £100, £150, and £200. (B) Impact of duration of protection on the median value per dose administered of vaccines with a range of efficacies. (C) Impact of efficacy on the median value per dose administered of vaccines with a range of durations of protection. Columns of panels correspond to the three targeting strategies assessed (VoD, VaR, and VoA). The lines plotted in panels B and C are transects through the heatmaps in panel A, with value represented on the vertical axis rather than by colour. Vaccine uptake is 33·0% (95% CI 32·7–33·3). VoD=vaccination on diagnosis. VaR=vaccination according to risk. VoA=vaccination on attendance.

Figure 5

Impact of 4CMenB vaccination over 10 years, how the cost per dose affects the probability of vaccination being cost-effective, and cost-effectiveness efficiency frontiers comparing targeting strategies (A) Cases of gonorrhoea in men who have sex with men in England following introduction of vaccination in 2022 under different targeting strategies (VoD, VaR, and VoA) using the 4CMenB vaccine, assuming the vaccine is 1, 1·5, 2, or 2·5 times as protective as the MeNZB vaccine, and has durations of protection after primary vaccination and re-vaccination as estimated by the Joint Committee on Vaccination and Immunisation for protection of infants against serogroup B meningococcal disease (ie, 18 months after two-dose primary vaccination and 36 months after re-vaccination). Lines represent medians and shaded regions represent 95% credible intervals. (B) Probability that vaccination is cost-effective (ie, its value exceeds its cost) with the same vaccine efficacies and durations of protection as in panel A. Vertical dashed lines show two alternative costs per dose of 4CMenB administered: £18, corresponding to the estimated National Health Service price of £8 per dose plus the £10 administration cost; or £85, corresponding to the current UK list price of £75 plus the £10 administration cost. (C) Cost-effectiveness efficiency frontiers comparing the three targeting strategies if 4CMenB is as protective as MeNZB, the durations of protection are the same as in panels A and B, and the cost per dose administered is £18 or £85. The black line shows the frontier, the numbered circles show the mean incremental costs and QALYs for each strategy, and the individual points show the uncertainty represented by the 1000 sets of sampled epidemiological and health-economic parameters. In all panels vaccine uptake is 33·0% (95% CI 32·7–33·3). Note that panel A shows undiscounted numbers while panel C shows discounted values. QALY=quality-adjusted life-year. VoD=vaccination on diagnosis. VaR=vaccination according to risk. VoA=vaccination on attendance.