• epidemiology
• family planning service provision
• general practice
• hormonal contraception
• oral contraceptives

I read the article by Briggs and colleagues1 with interest as it is, to my knowledge, the only study to assess the impact of the implementation of the UK Medical Eligibility Criteria on general practice prescribing of combined hormonal contraceptives (CHCs) in the UK. However, I would like to point out a limitation of this study.

As the authors have stated, oral contraception can be accessed on prescription from a general practitioner (GP) and these data are captured in the therapy table in the General Practice Research Database (GPRD) – the data source used by the investigators. However, CHCs can also be obtained from National Health Service contraception clinics and these data are not recorded in a patient's prescription history in GPRD. Women registered with GPRD who have obtained their CHC from contraception clinics will therefore not be identified as CHC users as part of this study, leading to an underestimation in the number of CHC users aged 15–49 years. An estimated 413 000 women in England and 20 000 women in Wales received oral contraception from a contraceptive clinic in 2011/2012 and 2010/2011, respectively.2 ,3 Although these figures are not specifically for CHCs, they illustrate that community contraceptive clinics play a substantial role in the provision of contraception to women of childbearing age.

In addition, those patients prescribed CHCs by their GP, and therefore included in the study, are not likely to be sufficiently representative of those patients who are prescribed CHCs from clinics. The National Statistics Opinion Survey conducted during 2008/2009 showed that during the 5 years prior to interview, a higher percentage of patients aged 16–19 years accessed contraception through a community contraception clinic rather than their GP or community practice nurse.4 Conversely, a larger percentage of older women obtained family planning services from their GP.4 As the patients accessing community contraceptive services are in general younger, they may also be less likely to have certain Category 3 or 4 risk factors (e.g. current smokers and previous smokers aged 35 years or over). Furthermore, patients with risk factors such as hypertension, migraine, stroke, ischaemic heart disease, dyslipidaemia and migraine may be more likely to obtain contraception from their GP while being monitored for these conditions.

The impact of this limitation is likely to have influenced the estimation of the number of women using a CHC with a Category 3 or 4 risk factor in the whole UK population.