• cost effectiveness
• screening
• women
• Chlamydia trachomatis

Screening women for asymptomatic Chlamydia trachomatis (CT) infections is indicated to prevent the spread of CT and the development of complications such as pelvic inflammatory disease (PID), chronic pelvic pain, ectopic pregnancy, tubal infertility, and neonatal pneumonia (major outcomes averted; MOA). Cost effectiveness presents an important aspect in the decision making regarding actual implementation. Recently, in this journal Van Valkengoed et al published a paper on the cost effectiveness of systematic screening among women in Amsterdam (Netherlands), using pharmacoeconomic modelling.¹ Using the same model, results on the cost effectiveness of an opportunistic screening in the same city have also been published.² Specific model assumptions differed in both publications. The aim of this letter is to compare cost effectiveness of systematic and opportunistic screening using similar model assumptions and correcting for potential biases.

Opportunistic screening was done during May 1996 to May 1997 in a pilot study.³ Women visiting the participating GPs were eligible for screening if they considered themselves heterosexually active, were aged 15–40 years, and did not visit their GP for sexually transmitted disease complaints (participation among women: 96% compared with 50% in the systematic screening). In this letter we report on the age group 15–30. Obviously, the effectiveness of this type of screening depends on the frequency of visiting the GP; 87% of Dutch women aged 15–30 visit the GP at least once per year.² As in the systematic universal screening, testing was done with ligase chain reaction (LCR) on urine. Participating GPs in the opportunistic screening had an over-representation compared to the general Amsterdam situation of participants from Caribbean and Surinam ethnicity with relatively high CT prevalence.³ To enhance valid comparison with the systematic screening, asymptomatic CT prevalence rates in the opportunistic screening were recalculated standardising for the distribution of the Amsterdam population over the ethnic groups of Caribbean, Surinam, and other (source: Statistics Amsterdam).

Parameters in the pharmacoeconomic model were kept similar to the previous paper in this journal, except for the probability of PID after asymptomatic infection.¹ For this probability we applied 20% compared to 10% in the paper by Van Valkengoed et al.¹ We even consider 20% as a very conservative estimate for the risk of PID in our model.⁴ Cost effectiveness was estimated as net costs per MOA in baseline analysis using assumptions above and sensitivity analysis (PID risk at 10%, high performance testing and pooling).^1,5

In the baseline analysis cost effectiveness is US$5300 per MOA for systematic screening of women aged 15–25 and $1400 for opportunistic screening of that same age group. Including sensitivity analysis, cost effectiveness of systematic screening ranges from $2000–$11 100 per MOA (see table 1). For opportunistic screening this range is $500–$4100 per MOA. For the age group of 15–30, cost effectiveness is estimated to be generally slightly less favourable.

We conclude that opportunistic instead of systematic screening reduces net costs per MOA up to 75% (age groups 15–25) and by approximately 50% (age groups 15–30) over a range of plausible assumptions. Opportunistic CT screening in Amsterdam is therefore more attractive than systematic screening from a pharmacoeconomic point of view. Obviously, pharmacoeconomics only present one aspect in decision making concerning CT screening, others being, for example, implementation issues and budgetary constraints.