This site is intended for health professionals only

At the heart of general practice since 1960

Contraception update

Long-acting reversible contraception

In the second article in Professor John Guillebaud's series on contraception, he outlines the benefits of long-acting methods

Injections and implants


Implanon is a single 40mm x 2mm subdermal rod, which releases etonogestrel (the biologically active metabolite of desogestrel) over three years. Its failure rate (about 5:10,000) if inserted properly, and assuming insertion is not too late in a conception cycle, is unmatched, apart from abstinence and vasectomy. It is best to use an anovulant method before insertion.

Mean insertion and removal times are well under five minutes, but special insertion training is vital. Frequent or prolonged bleeds affect about a fifth of users at one year. However, there are no current worries over low oestradiol levels (see depot medroxyprogesterone acetate [DMPA] below), so the high amenorrhoea rate is a bonus.

With both DMPA and Implanon, if irregular bleeding is unacceptable consider a trial of a suitable 20-30µg COC (if not contraindicated), for around three cycles. Mercilon contains the same progestogen as Implanon and usually controls the bleeding while the tablets are being taken, with shedding by 'pharmacological curettage' between packs of the spotting-prone endometrium. Thereafter, the woman may (or may not) obtain an acceptable bleeding pattern – although she should be warned that it is unlikely to be as good as during the short-term COC. If the combined pill is WHO 3 for that woman, it may be better to use a similar cyclical course of a natural oestrogen, preferably by transdermal patches, since this route causes least alteration in clotting factors.

What about body weight? In studies, the blood levels of etonogestrel were lower in obese women, so it has been suggested that in those with the very greatest body mass (say well over 100kg) it may be appropriate to discuss replacing their Implanon early, especially if they are no longer oligo-amenorrhoeic in the third year (which suggests continuing anovulation). However there is still (2006) no evidence of excess conceptions in such cases, presumably because of the 'margin' in this method, having such high efficacy overall.

What about women taking enzyme inducer drugs (EIDs)? Organon's summary of product characteristics says that these may lower the blood levels of etonogestrel but there have been no specific interaction studies. Therefore women on short-term treatment with any of these drugs are advised to use a barrier method in addition and (because reversal of enzyme induction always takes time) for 28 days thereafter.

During long-term EID treatment, Organon recommends transfer to a non-hormonal method and removal of the Implanon. This seems a bit wasteful and may be resisted by satisfied users, and those in monogamous relationships may hate to use barriers long term. So, despite the absence of specific trials, one might instead, in my opinion, consider compensating for the enzyme induction by a daily Cerazette tablet or even (expensively) by a second Implanon. These prescriptions would be unlicensed and must be strictly on a named-patient basis. EID users do so well with DMPA or Mirena (see below) that these might be better and cheaper choices.

Long-term data on Implanon is reassuring so far, for both oestradiol and bone density: in comparative two-year studies both remained similar to those in copper IUCD users. By analogy, there are no worries yet on this account with Cerazette either – or with the intrauterine system (IUS) whose amenorrhoeic action is primarily at the end-organ level, the endometrium.

DMPA and Depo-Provera

DMPA and Depo-Provera, 150mg are given 12 weekly. Note that the same injection frequency is now advised even if the user is on an EID.

My advice for late injections is as follows.

• If seven days late, give the next dose plus condoms for the next seven days. WHO and The Faculty of Family Planning and Reproductive Health say this is acceptable even up to 14 days late.

• If eight to 14 days late and a pregnancy test is negative, give the next dose with emergency contraception, if sexually

exposed, plus condoms for the next seven days.

• More than 14 days late: usually no emergency contraception and usually delay the next dose until after a mutually agreed total of 14 days since the last unprotected sex

using either:

– abstinence or

– ultra-careful use of condoms or

– a POP.

The last is a most useful option. It can be given for as many days as necessary until you are confident you can give the next dose of DMPA. There is no evidence to suggest a POP would be anything but safe for a pregnancy, if the patient were to conceive.

Then, if a pregnancy test is negative, give the delayed dose plus condom advice or POP for a further seven days. A pregnancy test is generally repeated later, for confirmation.

How long should a woman use DMPA, given the ongoing concern about low oestradiol levels? Much uncertainty persists, not all eliminated by the CSM circular (18 November 2004) which mainly recommends 'careful re-evaluation of risks and benefits in all those who wish to continue use for more than two years'.

In summary:

• if strong risk factors for osteoporosis

already exist, for all such women DMPA is WHO 3 or even 4

• in women under age 19, WHO classifies DMPA as WHO 2 because of concern that it may prevent achievement of peak bone mass

• UK advice (2004) is to use DMPA first line – but only after other methods have been discussed and are unsuitable or unacceptable

• DMPA is also WHO 2 above age 45 as ovarian failure may be beginning.

DMPA is very useful for fairly short-term use, after which switching to another long-term method is usual. If the woman wishes to use it for longer, even much longer, it is her right to decide to do so, after counselling about the uncertainty. This should be with formal two-yearly reassessment of alternatives but without imaging or blood tests unless clinically indicated. Remember DMPA is clearly safer, overall, than the COC.

Intrauterine contraceptive devices (IUCDs)

Any banded copper IUCD is in the gold standard category, and is first line unless an alternative is indicated. They are T-Safe Cu 380A or new variants with copper bands sunk into the plastic frame, branded as TT 380 Slimline or T-Safe Cu 380A QL Quick Load. These have a simpler loading system than the plastic 'hat' of the older T-Safe Cu 380A. It really is worth the effort to make this banded type of IUCD first choice, given that the licence is now for 10 years and the data supports its effectiveness until 12 years, even when fitted below the age of 40.

The main advantage is not its probable greater efficacy (in its latest form, with the copper bands near each tip of the T), but rather the fact that research in the past 50 years has clearly shown that most IUCD complications are related to insertion or reinsertion, and reduce in frequency with duration of use. Why use a five-year device when a 10-year will fit? This one usually passes through the cervical canal surprisingly easily, in all parities.

Nulliparity is not WHO 4 and in a mutually monogamous relationship, especially above the age of 35, it should be seen as only WHO 2 or maybe 3 for the IUCD method. The T-Safe Cu 380A and both its clones the Slimline and Quick Load, remain the first-choice devices even for nulliparae, although it may be necessary to use a smaller IUCD for a comfortable and satisfactory fitting.

How long can IUCDs be used for? UK practice since 1990 is that any copper IUCD fitted above age 40 can be used for the rest of the patient's reproductive life. Pre-medication with a NSAID should be routine for all insertions and local anaesthetic, a tiny injection at 12 o'clock on the cervix, always offered. I have never known a woman who gets that very unpredictable pain from the tenaculum not to appreciate this.

When should you use other IUCDs, such as Nova T380? In a randomised controlled trial the Nova T380, which has copper wire but no bands, was significantly less effective than the T-Safe Cu 380A (cumulative failure rate at three years 3.6 versus 1.7 for the T-Safe). However, I feel it might be appropriate for a nulliparous woman using it for emergency contraception and planning to have the device removed once established on a new method, such as DMPA. Another possible emergency contraception option for nulliparae is the Flexi-T 300, which is exceptionally small and has an easy push-in fitting technique with no separate plunger and which would be cheaper. However, it has been shown to have a highish expulsion rate and has no bands.

For long-term use, in my view the UT 380 Short (Nova T style but on a shorter stem) should usually be reserved for when the T-Safe Cu 380A or equivalent cannot be fitted for some reason (such as cavity sounding to less than 6cm). Also available now is the Flexi-T 380, on a slightly enlarged frame and with bands on its side arms, but otherwise identical in shape. We need more data on this. It might be very effective and usable for longer than its five-year licensed life, and it might be as easily inserted as the Flexi-T 300 and have an acceptable expulsion rate. If so, being copper banded, it might come to rival the T-Safe Cu 380A and its clones.

When should we use the banded GyneFix? This is not widely available, and specific insertion training is needed. It is useful in some cases.

• Special indications include a distorted cavity on ultrasound scan (if IUCD is usable at all), or a small uterine cavity sounding less than 6cm. (A rival option for the latter, however, is that new UT 380 Short or one of the Flexi-Ts).

• All users should be warned about unrecognised expulsion: being able to feel the threads is vital with GyneFix.

Note that the Multiloads are without any established advantages. Schering no longer markets the ineffective Nova-T 200.

Mirena IUS

This is a major advance in contraception technology, with the added value in relieving pain as well as menorrhagia. There are two more differences between Mirena and copper IUCDs.

• It is not a method of emergency contraception. Unlike the copper ion, the levonorgestrel hormone is not fast acting. As with Implanon, using an anovulant method up to insertion greatly helps the logistics of insertion timing.

• It should normally be replaced at the

licensed five years, or at four years if used as part of HRT, for which it is excellent.

When might an IUS be left in longer? If it is fitted above age 45 and longer use is requested for contraception (never as part of HRT), the 2005 NICE guideline permits use of the same IUS provided the woman does not have periods with the IUS in place until contraception is no longer needed. However, this should be done only on a named-patient basis after well-documented counselling. If the IUS is only used for menorrhagia and pain control, not contraception, the same IUS may of course stay in situ for as long as it helps the woman.

As with all intrauterine methods good training and attention to detail are crucial for insertion. The Faculty of Family Planning and Reproductive Health's recommendation for maintaining experience is a minimum of 12 insertions per year, of at least two types of IUCD/IUS.

What about Mirena and EIDs? Studies from the Margaret Pyke Centre show good effectiveness was maintained in 50 users of the IUS plus EIDs, and one definite pregnancy was reported. This is biologically plausible, since the levonorgestrel would be released in high concentration locally and should therefore still have its normal effects on the uterocervical fluid and also in impairing implantation. Therefore, pending more data, the use of Mirena by women on EIDs is a good choice, a WHO 2 situation. No added precautions need be taken, but with advice that there might be a minimal reduction in its amazing efficacy.

What is the risk of pelvic inflammatory disease? Neither IUCDs nor the IUS, intrinsically, increase this risk. It is essential to ensure the cervix is pathogen free, through verbal screening for mutual monogamy plus also (unless good verbal screening suggests otherwise) testing for chlamydia.

What about a history of a past ectopic pregnancy? Although anovulant methods would be even better, in my view this is at most WHO 3 for the IUS or banded IUCDs.

Which is better – female sterilisation, a new-type IUCD or the IUS? Peterson et al (1996) showed the failure rate of female sterilisation as performed in the US to be 14/1,000 at seven years – not different from the T-Safe Cu 380A and the IUS (about 14/1,000 and 11/1,000 failures respectively) by seven years. In the UK, the RCOG estimates the lifetime failure rate of the Filshie clip as better, 3-4/1,000, whereas it is 0.5/1000 for vasectomy with azoospermia.

Professor John Guillebaud is emeritus professor of family planning and reproductive health, University College London

Competing interests Professor Guillebaud has received research grants from pharmaceutical companies interested in contraception, and has received payments for lecturing and providing consultancy advice

This article is based on a presentation, Top tips on reproductive health, given by Professor Guillebaud at the Pulse Seminar on Clinical Challenges on Wednesday 26th November 2006 at the Royal College of Physicians, London

Next week: emergency contraception

key points

• Advice-givers should be competent to give information about the efficacy, risks and side-effects, advantages and disadvantages, and non-contraceptive benefits of each method

• Ignorance, especially about conditions not yet categorised on WHO's 1 to 4 scale by UKMEC should be admitted during a consultation between two experts on equal terms with the user, or couple ( )

• The COC is too often seen as synonymous with contraception. Providers everywhere should de-mythologise and promote the

long-acting reversible contraceptives – the injectables, implants and intrauterine methods – which are still efficacious when forgotten and comparable in efficacy to female sterilisation

• Iatrogenic pregnancies can result from errors and omissions on the part of service providers: especially, the omission of sufficient time for the new user. At least 20 minutes should be allocated to a practitioner or clinical team (nurse and doctor) for a new consultation UK NSF 2005 (see ServiceStandardsforWorkloadFINAL.pdf)

Rate this article 

Click to rate

  • 1 star out of 5
  • 2 stars out of 5
  • 3 stars out of 5
  • 4 stars out of 5
  • 5 stars out of 5

0 out of 5 stars

Have your say