We thank Drs Baird and Webb for their considered response1 to our article2 and for opening up a constructive debate on provision of different methods of emergency contraception (EC).

We are aware of the evidence regarding the relative efficacies of the three EC methods and that Faculty of Sexual & Reproductive Healthcare (FSRH) guidance3 is that all women, subject to eligibility, should be offered all options from which to make their own informed choice. However, services do work within constraints that inevitably impact on what can be provided and that must be considered in the development of protocols. As we have indicated, our study was a pilot testing a pragmatic approach to method selection and was subject to restrictions requested by our local Medicines Management team. It also predated current FSRH guidance.

When ulipristal acetate (UPA) was granted a UK licence in May 2009 our Primary Care Trust (PCT) placed it on their ‘Red List’ of drugs (i.e. not funded). Four years on it is still there, albeit with exceptional use in women presenting between 72 and 120 hours after unprotected sexual intercourse (UPSI) now permitted. Although, as a specialist service we are not constrained by the Red List, we are strongly encouraged to comply where possible to ensure consistency and parity of provision between our service and primary care. Our audit in 2010 of 136 women attending for EC demonstrated that 90% present within 72 hours of UPSI. Complying with the Red List would mean that only 10% of our attenders would be eligible for UPA. The pilot protocol was developed against this restrictive background, and whilst we fully appreciate the fundamental concern of Drs Baird and Webb about using risk assessment in decision-making, if use is restricted there have to be some eligibility criteria.

Encouragingly, the results from Dr Baird's service demonstrate that when women are provided with free choice they do not all choose UPA. However, the use of this method has increased at the expense of levonorgestrel (LNG) with additional cost implications.4 We would suggest that for some services this may be a cause for concern.

Faculty guidance recommends that women should be offered all three EC options from which they can make an informed decision. Whilst there are many factors that may influence this decision, we believe that for most women the principal one against which other factors are considered is the likelihood of pregnancy. No one would argue against an intrauterine device (IUD) being the most effective option. However, it is an invasive procedure that many women are reluctant to pursue. In making their decision and weighing up the ‘pros and cons’, surely the concern most likely to overcome women's reluctance is how likely they are to become pregnant? We are therefore obliged to discuss the IUD to enable a woman to make an informed choice. Whilst we accept that any method of risk assessment will not give an absolute answer, a relative risk can be provided. Estimates of the probability of conception relative to intercourse on a given cycle day do suggest a definite peak (i.e. a high-risk time), which for women with regular cycles is Day 13.5 Whilst not perfect, our method of risk assessment did at least permit more widespread use of UPA and hopefully targeted those women at the highest risk of pregnancy and most likely to benefit. The efficacy of hormonal methods of contraception is based on ratios between observed and expected pregnancies and therefore also relies on estimates of the risk of conception.

We are in complete agreement with the comments regarding use of UPA in the under-18s and reiterate that the exclusion of this group from the pilot was a condition requested by Medicines Management and does not indicate an inconsistency in our general approach to this age group. Since November 2012, UPA has been provided under a patient group direction (PGD), which has no age restrictions. We have also modified the protocol to include body mass index (BMI) in our discussion on method selection as there is some evidence that both hormonal methods, but particularly LNG, are less effective with a BMI>25.6 ,7

In the 8 months following the introduction of the PGD there have been 400 attendances for EC. Methods chosen were LNG 63%, UPA 24% and IUD 13%. Our use of UPA is therefore slightly higher than the 18.7% reported by Dr Baird4 when women are given free choice. Whilst there was a fall in IUD usage in eligible women from 15% to 8.7% during the pilot study when UPA was introduced, further staff training prior to introducing the PGD has returned our IUD usage rates to baseline. In the subgroup of patients aged under 18 years (80 patients) usage has been LNG 79%, UPA 17% and IUD 4%.

The National Health Service is going through a time of great financial hardship and services are being rationalised in all areas. Our algorithm was developed against a background in which unrestricted use was not permitted in order to promote the copper IUD as the most effective form of postcoital contraception for all eligible women, with UPA second line for those at most risk.