Given that the four strategies were equivalent in terms of effectiveness, the less costly strategy (cost per individual treated) was also considered to be the optimal strategy.
The results were presented for different rates of gonorrhoea prevalence and CIP resistance.
The most striking results were as follows.
With prevalence of gonorrhoea between 0 and 1% and with prevalence of CIP-resistance between 0 and 20%, the optimal strategy (lowest cost per patient treated) was ST1.
With prevalence of gonorrhoea between 2 and 3% and with prevalence of CIP-resistance between 0 and 5%, the optimal strategy was ST1.
With prevalence of gonorrhoea between 2 and 3% and with prevalence of CIP-resistance greater than 5%, the optimal strategy was ST3.
With prevalence of gonorrhoea between 3 and 10% and with prevalence of CIP-resistance between 0 and 20%, the optimal strategy was ST3.
With prevalence of gonorrhoea between 10 and 13% and with prevalence of CIP-resistance between 0 and 3%, the optimal strategy was ST2.
With prevalence of gonorrhoea between 10 and 13% and with prevalence of CIP-resistance greater than 3%, the optimal strategy was ST3.
With prevalence of gonorrhoea between 13 and 15% and with prevalence of CIP-resistance between 0 and 3%, the optimal strategy was ST2.
With prevalence of gonorrhoea between 13 and 15% and with prevalence of CIP-resistance greater than 3%, the optimal strategy was ST4.
Given the equivalence in terms of effectiveness among the strategies, the optimal strategy was not only the cheapest but also the one that yielded the lowest cost per case successfully treated.
The sensitivity analysis showed that the model outputs were sensitive to some parameters. For example, if the ratio of the costs of CIP to CEF were changed from 1:5 (as in the base-case) to 1:2, and the costs of the tests became equal, ST2 and ST4 were optimal for greater combinations of gonorrhoea prevalence and CIP-resistance prevalence than in the base-case.
However, if the ratio of the costs of CT to non-CT was changed from 1:1 to 1:3, then ST1 and ST3 became optimal for all combinations of gonorrhoea prevalence and CIP-resistance prevalence.
The Monte Carlo simulation showed that ST1 had the lowest mean cost per patient treated. Only when gonorrhoea prevalence was 2% and CIP-resistance prevalence was 10% did ST3 have a lower mean cost per patient treated. The probabilistic sensitivity analysis also showed that there was considerable overlap in costs across the two antimicrobial options.