Authors of the paper by Kounali et al. have repeatedly made the same challenge to the validity of our model-based method for estimating chlamydia prevalence from surveillance data, to which we have repeatedly responded, with evidence [Reference White and Lewis1, Reference White and Lewis2] – none of which they have acknowledged.
Kounali et al., a group which includes representation of England's National Chlamydia Screening Programme (NCSP), assert ‘It is not possible… to make inferences about CT prevalence, or changes… over time, without information on [reason for testing]’, yet despite the lack of such data NCSP states that ‘modelling suggests that the level of testing that has been achieved in England… will probably have resulted in reductions in prevalence’ [3]. Regarding Kounali et al.'s citations in support of their assertion, we have already responded to Soldan et al.'s letter [Reference White and Lewis1, Reference White and Lewis2] and the commentary of Low and Smid [Reference White and Lewis1], and we cited the paper by Miller in our original paper [Reference Lewis and White4].
Estimating chlamydia incidence and prevalence from surveillance data is a subject of active debate, and methods have been described in several papers by ourselves and others [Reference Lewis and White4–Reference Lewis and White6]. We would direct readers to our original paper (and accompanying computer code) describing our model and its testing, validation, and sensitivity analysis considering unrecorded information, so they may evaluate our method for themselves [Reference Lewis and White4].
The point on which we all agree is the importance of understanding testing behaviour and individuals' reasons for chlamydia testing. We hope that more-detailed surveillance data and population-based studies such as the fourth National Study of Sexual Attitudes and Lifestyles will provide the information required to better-understand patterns of chlamydia infection and optimise the effectiveness and cost-effectiveness of chlamydia control.
Financial support
PJW acknowledges joint Centre funding from the UK Medical Research Council and Department for International Development (grant number MR/R015600/1). PJW and JL acknowledge funding from the UK National Institute for Health Research (NIHR) Health Protection Research Unit (HPRU) in Modelling Methodology at Imperial College London, in partnership with Public Health England (grant number HPRU-2012-10080). The views expressed are those of the authors and not necessarily those of the Department of Health, Department for International Development, MRC, NHS, NIHR, or Public Health England.
Conflict of interest
None.