Medicines in maternal and neonatal care: current practice and future developments.

This is a report of a joint meeting of the Forum on Maternity and the Newborn with the Section of Pharmaceutical Medicine & Research of the Royal Society of Medicine, held on Thursday 13th April 2006.

The report is to be published in part in the Midwives Journal of the Royal College of Midwives. It is reproduced here with their consent and our thanks.

Chairs (morning session): Dr Richard Tiner, Medical Director, Association of the British Pharmaceutical Industry (ABPI), London and Dr Maggie Redshaw, Social Scientist, National Perinatal Epidemiology Unit, Oxford

Setting the scene
Dr Richard Tiner (RT), Medical Director, the Association of British Pharmaceutical Industries (ABPI)

Significant changes are coming in the pharmaceutical industry. I was a principal in general practice for 17 years before moving into the industry, and even then I was aware that on the recommendations of paediatricians children were being prescribed off-label or unlicensed medicines; the current rate for this is 10 to 20% in the UK, in general practice, 45% in general paediatric wards in Western Europe, and over 90% in neonatal intensive care units.

In December 2000 the European Union Council passed a resolution inviting the Commission to make appropriate proposals in respect of clinical research and development, ensuring that medicines for children are fully adapted to their specific needs, and taking into account the ethical aspects of clinical trials on children and international standards for the protection of minors with regard to medical research. The document Better Medicines for Children published in February 2002 gave hope for the rapid development of appropriate regulations, but when the following year it was announced that the European Parliament were required to make relevant impact assessments this led to delay, until the Regulation was eventually published in September 2004. It is hoped that it will be ratified some time in 2006, but it will fail unless an absolute majority is achieved; and the same applies to every nation in the European Union. (Post meeting note – an absolute majority has been achieved and the legislation will be in place by the end of 2006).

The Preamble to the Regulation notes inadequate dosage information for children and increased adverse drug reactions, some fatal, and ineffective treatment due to under-dosage. Suitable drug formulations and routes of administration are addressed; part of an adult tablet crushed into an orange drink turns it into an unlicensed medicine. The Regulation is intended to increase research, development, and the authorisation of medicines for use in children; information on medicines already authorised will be improved, and an important objective is to avoid subjecting children to unnecessary trials - including the trial which prescribing an unlicensed medicine represents.

Every new medicine and every new indication for the use of a medicine will require a Paediatric Investigation Plan (PIP) as part of the application - a serious wake-up call for the industry. Consideration may be waived or deferred if a drug is deemed to be unsuitable for children or until there is adequate information on the adult preparation. Each approved medicine will have to be marketed in all 25 member states of the EU. Major incentives for developing a PIP are a six-month extension to the patent for the medicine concerned, applied to adult as well as paediatric versions, and 10 years of exclusive data protection rights for the product - 12 years for orphan medicines. Eventual licensing will be based on approved research, an advantageous factor which is sure to be included in a product's data sheet.

Guidelines have been published defining five ages of childhood from preterm to adolescent up to 18 years; use and dosage recommendations will where appropriate have to recognise these as well as other differences between the child patients. Licensing an antibiotic will require all five ages to be considered; for schizophrenia only the adolescent range need be addressed. Ideally doses will be calculated in mgs/square metre, which best recognises the rapid changes in weight of the smallest children; the immature physiology and small blood volume of preterm neonates must be taken into account, as must the immaturity of the neonatal blood/brain barrier. The rapid drug clearance of infants, changes at puberty, and non-compliance by adolescents must not be overlooked.

The planned electronic patient record in the UK should be an ideal tool for meeting the obligation for an application to indicate how long term follow up for adverse drug reactions and efficacy will be undertaken; risk management plans will also have to be drawn up and published for all medicines. All drug trials completed by the time of the Regulation's establishment will have to be reported to the relevant authority within one year, and future studies will be overseen by the relevant national regulatory authority but may be co-ordinated by the European Paediatric Research Network which will be set up under the auspices of the European Medicine Agency (EMEA).

The Regulation will apply in full to all 25 EU member states; however clinical trials in children are overseen by the Clinical Trials Directive, and here there is considerable variation. It is hoped that the UK's Medicines for Children Research Network, with its substantial government funding, will encourage collaboration between academic departments and the pharmaceutical industry, which should involve the network at an early stage. Child-specific protocols and formulations will need to be developed, and the number and size of blood samples kept to a minimum. "Would I be happy for my child to enter this study?" must define the standard. My organisation, the ABPI, believes that more paediatric clinical research in Europe and particularly the UK is certain to lead to an increase in the numbers of licensed medicines for children; the end of this year, when the Regulation is established, will be a significant point of departure.


European Commission Consultation document 2002. Pharmaceuticals : regulatory framework and market authorizations. Better Medicines for Children: Proposed regulatory actions on paediatric medicinal products.

Conducting randomised trials of medicines for use in pregnancy and the newborn.
Dr Peter Brockelhurst (PB), Director of the National Perinatal Epidemiology Unit (NPEU), Oxford

Randomised controlled trials (RCT) are carried out with the pregnant and the newborn by the NPEU; quite a number have been concerned with medicines. Here I shall be addressing the design of perinatal trials.

Some of the pitfalls of design are exemplified by the significance given to outcome in multiple pregnancies; it is obviously illogical, but not unknown, to assign the outcome to be fetal death when only one of the fetuses is lost, and outcomes are crucial in perinatal studies. It is essential to control appropriately for multiple pregnancy, treating each fetus as independent and using the worst outcome for each.

As obstetricians we make a major effort to prevent the morbidity and mortality associated with preterm birth. The Oracle trial studied the pregnancies of 11,000 women at risk of preterm birth; the outcomes when erythromycin or co-amoxiclav (Augmentin) or both were prescribed to delay preterm pre-labour with ruptured membranes were compared with placebo. Co-amoxiclav was significantly more effective than erythromycin in delaying labour in these 5000 women; but it is clear from the size of this trial that it was not designed to study this outcome (for which a group of 500 women would have sufficed) but the neonatal outcome measured by death, cerebral abnormality or lung disease, and for this composite primary outcome erythromycin was significantly superior. Also the incidence of necrotizing enterocolitis was significantly higher for co-amoxiclav compared with placebo. If we had used the surrogate trial outcome 'gestation at delivery' we would all be prescribing co-amoxiclav, not erythromycin to delay labour, and harming some babies. Note however that the conditions of the babies at birth is the short-term substantive outcome, and the results for these children at age seven years may be very different. This is an excellent example of the way in which outcomes can mislead by moving in opposite directions, an effect which can only be avoided by using large enough trial groups.

Another trial using 724 women was set up to compare the effectiveness of the oxytocin antagonist atosiban with beta-agonists, which have had a fatal side effect for some mothers, in delaying labour. The outcome that atosiban was equally effective was naturally welcomed on behalf of women, but the trial's neonatal outcome was ambiguous. The outcomes of a trial of dexamethasone administered to mothers to prevent lung disease in their premature babies seemed to be satisfactory in that rates for this condition were significantly reduced, and the death rate not increased - until it was noted that there was a significant increase in cerebral palsy in the long term, a harm which overwhelmed any benefit. Only long-term follow-up can provide valid conclusions for trials affecting children.

The importance of follow up and its difficulties cannot be over-emphasised. Nonetheless a group in the north of the country achieved a 100% follow-up rate of over 800 babies born before 32 weeks. When their follow up was incomplete they found 6.9% of babies with severe disability in the 709 who had been found without difficulty, but complete follow up revealed an incidence of over 55% in the 47 found with great difficulty; the most disabled babies had been the most difficult to find. This is not hard to explain: the time of parents caring for a child with cerebral palsy, feeding him and taking him to the various hospital clinics, is fully occupied; they are unlikely to give priority to completing long questionnaires for research purposes.

RCTs are rarely large enough to achieve the complete detection of disability. Case control studies are usually used to confirm or refute the association between drug exposure and later effects, and these are subject to bias. We cannot be too careful when prescribing drugs which will reach fetuses and newborns; pictures of the effects of thalidomide emphasise this, but the story of diethylstilboestrol is also a horrifying one; 5 million prescriptions were written for it to prevent recurrent miscarriage with no evidence of its efficacy. The daughters of women given the drug were subject to a high risk of vaginal cancer, and over half of the sons are infertile. Currently RCTs are being undertaken to study the effectiveness of progesterone in pregnancy to prevent premature birth - fetuses exposed to a powerful sex hormone during much of their intrauterine lives. As gestation at birth is the object of these inquiries only groups of about 400 women are being recruited, insufficient to demonstrate rare but potentially serious effects on the babies. Once again the lessons of history are not being learned. The essential vigilance must be active; subtle effects such as minor brain damage discovered years later may be associated with drug use in pregnancy. These are the reasons why perinatal research is different from any other.

The market for the drugs concerned is small and so necessarily is the potential profit for the pharmaceutical industry; for example only about 81,000 women will be given a tocolytic drug in the UK annually, and then only for 48 hours. Relative to the number of babies admitted to hospital, the number admitted to neonatal intensive care is small. The profit is in medication for common and/or chronic disease in adults. Naturally, given the disastrous history, all the parties concerned in this research are nervous; Oxford University has a blanket insurance for all its clinical trials except those concerned with pregnancy, children and contraception; we have to arrange separate insurance at substantial premiums for our RCTs, however small.

Increasing bureaucracy is a major threat to clinical trials, and whereas ethical approval has been streamlined recently, and similar changes are under way for research and development committee approval - currently a nightmare - these are taking their time. It is no surprise that these approvals are harder to get for trials in pregnancy and the newborn. In addition most of the drugs we use are not licensed, and very few have been rigorously tested. We do not usually get explicit consent from parents, and all this is a disincentive to undertaking RCTs, putting pressure on clinicians to use biased and unreliable observational data which are liable to give misleading results.

The long overdue Medicines for Children Research Network offers hope for the future prospects of well-designed studies of medicines for children and research into prevention, diagnosis and treatment; its objectives support our intentions:
· to improve the care of children and their families,
· to improve the speed, co-ordination and quality of research
· integration in terms of information needs · the formulation of research
· wider participation. For us this last would imply that around 10% of babies receiving unlicensed drugs are enrolled in research, a not unreasonable target when you consider that virtually all children with cancer are so involved.


Kenyon S.L, Taylor D.J., Tarnow-Mordi W.  (2001)  Broad spectrum antibiotics for spontaneous preterm labour: the MRC ORACLE II randomised trial.  Lancet 357; (9261) : 989-994

Kenyon S.L., Taylor D.J., Tarnow-Mordi W. (2001)  Broad spectrum antibiotics for preterm prelabour rupture of fetal membranes: the MRC ORACLE I randomised trial.  Lancet 357; (9261): 979-988.

Consumer collaboration in clinical trials in maternity care
Mrs Gill Gyte (GG), National Childbirth Trust (NCT) antenatal teacher and Consumer Panel Co-ordinator for the Cochrane Pregnancy and Childbirth Group

Brief history of consumer collaboration in clinical trials

At a NCT conference in 1992, Sir Iain Chalmers, then clinical director of the NPEU, reminded NCT members “Because mothers and babies have the most to gain from maternity care that has been evaluated for effectiveness and safety, they have the strongest vested interest in promoting evaluative research”. The NCT response was to increase its involvement in research, and the NPEU collaboration with consumers has continued to be significant. The Association for the Improvement of the Maternity Services (AIMS) had led the way for lay support in clinical trials since 1985, with the trial of chorionic villus sampling (CVS) versus amniocentesis.

INVOLVE (formerly Consumers in NHS Research), assists collaboration between consumers and professionals in all areas of research, and encourages consumers to become effectively involved.

As consumer referees with Cochrane we take part in the editorial process of publishing their systematic reviews in maternity care; the consumer perspective on protocols, particularly when commenting on outcomes, is proving very useful.

Some examples of consumer collaboration in clinical trials in maternity care
Among trials which have benefited from consumer collaboration are:-
1) The MAIN trial (treatments for inverted nipples), a trial for which uniquely the NCT recruited women, and the outcome of which enabled breastfeeding counsellors to dispense with a variety of ineffective measures.
2) ORACLE (antibiotics in threatened pre-term labour; link above), here the NCT was instrumental in helping to write a results leaflet specifically for parents: now 14 years on I am collaborating in a follow up study comparing children's health and school performance at age seven (Kenyon et al. 2001).
3) Hands on or Poised: The HOOP study (care of the perineum) (McCandlish et al. 1998).
4) The CAESAR trial (caesarean section techniques).

Some benefits of consumer collaboration in clinical trials in maternity care
Consumers can help to write the information inviting people to participate. In maternity care appropriate timing is important - for example, not to make the first mention a trial during labour. The NCT collaborated with AIMS in producing a Charter for Ethical Research in Maternity Care, proposing that participation should be discussed during pregnancy, and emphasising the need for sensitivity when the subject is raised again in labour.

Women need to know the outcome measures of the trials in which they are taking part; in one trial of induction of labour, the effectiveness of the drug to stimulate labour used was not an outcome measure and consumer input at the planning stage would have addressed this. Essential long-term outcomes for any study of caesarean section should include the potential adverse effects in future pregnancies. However, the inclusion of too many outcomes in the analysis of a trial would render it impractical, a case for compromise in collaboration.

NCT Research Networkers
The NCT Research Network aims to represent the interests of women invited to take part in clinical trials, and other types of research, but the pressures on the network and NCT can be considerable. It is important that NCT support for a trial should not lead women to regard it as a stamp of approval; women need to reach their own decisions. Prior to the ORACLE trial, another consumer organisation criticised NCT involvement, saying it was 'the attractive icing on a toxic cake'. This stimulated discussion on the judgments needed when a consumer organization supports a research trial, and whether it should have its name on the information leaflet or not.

NCT guidelines for collaboration
Before engaging in collaboration on a research project, the NCT has questions to ask:
· What are its purpose and importance of the research question?
· What is the probability that the purpose will be achieved?
· Has the subject already been researched?
· Is perceived opinion ready for it?
· Who is undertaking the research, and how is it funded?
· Who has approved it, and what are the safeguards for participants?
· What will be required of the participants?

Sometimes consumers' suggestions are not incorporated and this can give rise to dilemmas about continuing involvement. One trial did not get NCT support when it proposed to compare only the effectiveness of two narcotics during labour. The NCT suggested a third arm to include doula support or a water pool, as the high rate of narcotic use (67%) implied that women were not getting good support during labour. This was considered but declined, and NCT withdrew from the trial because we felt it was not asking the right question. Consumers also withdrew from a trial on retained placenta, in which the researchers planned to intervene 30 minutes after a birth, while we suggested an interval of one hour to enable the placenta to deliver. We would also have major doubts about collaborating in any research unless there is substantial uncertainty around the benefits and harm (equipoise); the NCT considers that there is sufficient data on the adverse effects of caesarean section (CS) to say that equipoise does not exist for a study of CS on demand. In every case the NCT requires women to get good information when being asked to participate.

Our purpose in participation is to ensure that clinical trials address issues important to women, and that women taking part are fully informed. Consumers will need training and support, and their involvement needs to be funded. All need to be aware of the responsibilities and burdens which involvement brings.


Kenyon SL, Taylor DJ, Tarnow-Mordi W. Broad-spectrum antibiotics for spontaneous preterm labour: the ORACLE II randomised trial. Lancet 2001;357: 989–994

McCandlish R, Bowler U, van Asten H, Berridge G, Winter C, Sames L, Garcia J, Renfrew M, Elbourne D. A randomised controlled trial of care of the perineum during second stage of normal labour. Br J Obstet Gynaecol. 1998 Dec;105(12):1262-72.


Wendy Savage: Has the long-term outcome of giving steroid by injection to mothers to prevent lung disease in their premature babies been researched? Has this treatment become protocol driven? PB: Observational data has determined a single dose regime, and as it is often impossible to know when a baby will be born, thresholds are changing, and many babies will be will not have been exposed to steroids. "Mont" Liggins first reported New Zealand research on this topic in 1972; the RCOG published a revised guideline in the 2004. The only adverse finding in a 30-years follow up has been some disturbance of glucose homeostasis, which may predispose to cardiovascular disease. Currently practice reflects the view that the benefit of this treatment outweighs the potential harms.

Professor Lesley Page (London) expressed concern about the representation of multi-ethnic communities in trials; often there are language problems, and the mobility of populations makes follow up difficult. RT: These groups are seriously under-represented in trials, which leaves important questions unanswered. Face to face follow-up of large groups is impractical, so that we rely more and more on questionnaires with all the problems of literacy and language which they involve. For these reasons we are excluding London from a large trial which we are planning. GG: I was horrified recently to learn that the overspend in the clinical side of the NHS will impact adversely on research. Not only is there too little money for research funding, but also a mystery surrounds the choices of which trials will be funded.


Liggins GC, Howie RN. A controlled trial of antepartum glucocorticoid treatment for prevention of the respiratory distress syndrome in premature infants. Pediatrics 1972; 50:515-525.

Royal College of Obstetricians and Gynaecologists Guideline No. 7, Revised February 2004. Antenatal corticosteroids to prevent respiratory distress syndrome. Read on line

Nursing issues in prescribing
Ms Lynne Paterson (LP), Neonatal Nurse Consultant, South Tees Acute Hospitals NHS Trust, Middlesborough

Research evidence is scanty in this area. Most nurse prescribing papers are qualitative in nature and although they give good information as to how nurse prescribers have experienced their new role, they provide little in the way of concrete data as to its effects within the neonatal or wider paediatric field.

Historically nurse prescribing first appeared as an item in the Cumberlege report of 1986, with reference to district nurses. Three years later the Crown report expressed the view that it would enhance patient compliance; after pilot studies it was implemented nationally using a limited formulary, and later limited powers to prescribe depending on a nurse's qualifications. The prescription of controlled drugs or unlicensed medications was not permitted. Subsequently minor injuries, palliative care, and health promotion e.g. immunisation were included, and Patient Group Directives (PGDs) enabled non-prescribers to prescribe items already approved by “Independent” prescribers. The formulary has been extended to include 240 prescription-only medicines.

At present the government is funding training for nurse prescribers, pharmacists and other health professionals, but is well short of its target for numbers of the trained. There was no provision for prescribing in acute care; the various tiers of prescribers with varying rights to prescribe have been confusing, and a one-course-fits-all training unsatisfactory, with universities offering face-to-face or distance learning packages. In 2004 almost half of the trained nurses were not prescribing, and community nurse practitioners were prescribing on average less than once weekly; prescribers have been frustrated by the limited formulary and excessive paperwork. Improvements in the current arrangements are said to be imminent, with the opportunity to prescribe from the whole of the British National Formulary, although the Nursing and Midwifery Council (NMC) advises nurses to stick to their specialist areas.

You might suppose that in a neonatal intensive care unit (NICU) such as the one in which I work there would be no difficulty in getting a doctor to write a prescription; however, it is remarkable how seldom SPRs and consultants are available, which can leave a child who requires opiate analgesia in pain, another, asphyxiated, waiting for intubation and ventilation, and another with acute sepsis waiting for essential antibiotics to be prescribed. But when it is a matter of resuscitation nurses are expected to administer emergency drugs - a huge paradox. The drugs legally available to us Advanced Neonatal Nurse Practitioners (ANNPs) are irrelevant to our work, and any other drugs or fluids have to be prescribed as a “Supplementary” prescriber from a clinical management plan. PGDs do not answer the need. Corners are being cut, contrary to the regulations of the NMC.

A modest number of ANNPs responded to a questionnaire I sent round, revealing the following well-known illegal practices: getting a drug chart written and signed before leaving to retrieve a baby from another hospital or from the community; getting drugs prescribed before departure; getting telephone authorisation from a NICU doctor; covering junior medical staff and telling them what to prescribe or correcting their prescriptions; inserting intravenous or umbilical artery cannulae, and running fluids through without actually prescribing them. Some of these situations require a nurse to ask a doctor to prescribe later.

Pharmacists, doctors and nurses all have anxieties around nurse prescribing; one year's post-registration training in the specialist area of neonatal intensive care nursing cannot be adequate for safe prescribing. The NMC recommend that this gap in knowledge is remedied on the job. Training is not yet tailored for the improvement of patient care; equipping nurses to undertake a multiplicity of irrelevant procedures is pointless. In the USA and Canada most ANNPs are educated to Master's Degree standard including pharmacology and pharmacokinetics, prescribing is restricted to senior staff and monitored by the Drug Enforcement Agency as they are permitted to prescribe controlled drugs, and their in-service training is supervised.

I believe that nurse prescribing in the neonatal field and probably in paediatrics should be limited to experienced practitioners with core qualifications, for which there must be an adequate training programme. I have not taken the specialist training currently available, knowing it to be inappropriate; furthermore there is no incentive to undertake it, as it does not attract extra remuneration. Trusts can appoint their own nurse prescribers, and their policies for this are inconsistent. All this has introduced a systemic error which is likely to increase medication mistakes in nurse prescribing in the UK, despite the government's plans and pledges in the National Service Framework for Children, and Building a safer NHS for patients - improving medication safety.

I have gathered comments from my colleagues around the country, among them:

· "Nurse prescribing would provide patients with better access to timely appropriate care";
· "There would be improvement in babies receiving life-saving medications";
· "ANNPs, since they are in post for substantial periods, have more experience in neonatal dosage than some junior doctors, and are more familiar with the local policies and protocols".

The anticipated improvements have yet to arrive; nurse prescribers in neonatal care must be trained appropriately and able to prescribe from an appropriate formulary. Nurse prescribing has been implemented in a complicated and piecemeal fashion. Neonatal care has not implemented prescribing to a large extent due to the current regulations around unlicensed and off-label drug use, and as a result some nurse-led care has been affected. We can learn from other countries as to how this could be developed more effectively in the UK.


Cumberlege Report. Neighbourhood nursing: a focus for care. London: HMSO, 1986

Crown Report. Report of the Advisory Group on Nurse Prescribing. London: Dept. of Health, 1989

National service framework for children, young people and maternity services: Medicines for children and young people. Department of Health, Department for Education and Skills, London 2004.

Building a safer NHS for patients: Improving medication safety. Department of Health, London 2004

Mothers, breastfeeding & medication: What is good practice?
Ms Louise M Long (LL). Midwifery Lecturer, Kings College London

The great majority of prescription and over-the-counter medications are compatible with breastfeeding since they appear in very small amounts in human milk, usually less than 1% of the maternal dose, which is not clinically significant for the baby (Riordan & Auerbach 1998). "Medications and mothers' milk" (Hale 2002) is a much favoured and fairly comprehensive source of information on this topic.

Of course good practice requires and reliable research supports breastfeeding being encouraged, supported and continued at least for the first six months of life, and its benefits usually outweigh any risks to the infant's well-being; undue caution on the part of health professionals may lead to medication being used as an excuse to advise discontinuing breastfeeding, and mothers will often continue despite rather than because of medical advice. Such advice ignores the sad fact that we are steeped in a bottle feeding culture, with only 24% of mothers achieving the 6-month standard.

Medication should only be used if it is absolutely necessary and should if possible be delayed; when prescription is unavoidable a drug with a short half-life should be chosen and sustained-release preparations avoided. Drugs should be selected for which there is reliable and reassuring breastfeeding data, and given in the lowest effective dose for the shortest possible course. Usually the lowest amount will get into the milk if taken immediately after a feed or before the infant has a long sleep. A mother who needs a drug that is clearly contraindicated during lactation has to learn about breastfeeding interruption and how to express her milk; she should be referred to a healthcare worker known to have the appropriate expertise. The NCT and the La Leche League are reliable sources of counselling.

Hale defines five categories of drugs which range from the safest (L1), which have been taken by large numbers of breastfeeding mothers without any observed increase in adverse effects in the infant, and of which controlled studies have failed to demonstrate a risk to the infant or show that the possibility of harm is remote, or that the product is not orally bioavailable in an infant. In the highest category of risk, L5, the contraindicated, there is documented risk to the infant based on human experience or the medication could be expected to cause significant harm. The risk of use clearly outweighs any possible benefit from breastfeeding.

Typical areas of concern are amphetamines, the antiarrhythmic cardiac drug amiodarone, the anti-depressant doxepin (Hale L5), where sertraline is to be preferred, and in category L4 oral or intravenous chloramphenicol (the eyedrops are safe), and the migraine remedy ergotamine tartrate, propranolol being preferred. Hale assigns lithium to his category L4, with the proviso that serum levels are monitored in the infant from the 10th day of life and throughout lactation, a tough requirement on the clinicians who have to take blood samples. Many psychiatrists are unenthusiastic about the alternatives to lithium, and its use must be balanced with the undeniable benefits of breastfeeding and of relief of the mother's disorder for mother-infant bonding. Hale has herbal remedies and teas in category L4 and ginseng in L3; La Leche recommends the major brands of herbal tea, taken in the weakest possible concentration and less than one litre daily. Products of unknown composition are to be avoided (Hale 2002, Mohrbacher & Stock 2003). Heparin and warfarin, contrary to a common misconception, are safe during lactation.

I cannot recommend the drug suppression of lactation otherwise than in cases of stillbirth. The present drug of choice is cabergoline (in preference to bromocriptine), although lactation may recur in a difficult third week of use (Spitz et al. 1998); compare this with the discomforts of the first two weeks when only a binder is used. The re-establishment of breastfeeding can be discounted when cabergoline has been taken, so it must never be prescribed when a woman who has chosen to bottle-feed may change her mind.

The undersupply of breast milk is almost invariably due to difficulties in the baby latching on to the breast and finding a good position. Medication to enhance lactation e.g. with domperidone should only be undertaken after a full history and attachment assessment is undertaken by an experienced healthcare professional. The situation is most likely when a baby too premature to breastfeed is eventually offered breasts which have not been primed by the essential early feeds.


Hale, Thomas.W. (2002) Medications & mothers' milk (11th edition). USA, Amarillo:Pharmasoft Publishing.

The American Academy of Pediatrics Committee on Drugs (2001) The Transfer of Drugs and other chemicals into Human Milk. Pediatrics 108:3 776-789 Read on line.

Mohrbacher N. & Stock J. (2003) The Breastfeeding Answer Book (3rd edition), USA, Illinois: La Leche League International

Riordan J and Auerbach KG. (1998) Breastfeeding & Human Lactation (2nd edition), London: Jones & Bartlett Publishers.

Spitz, A, Lee NC, and Peterson HB. (1998) Treatment for lactation suppression: little progress in one hundred years. Am. J. Obstet Gynecol 179:(6 Pt 1) 1485-1490

Useful Breastfeeding websites:
UNICEF UK Baby Friendly Initiative
Breastfeeding Online
Enhancing lactation with medication


LP: We are not insured vicariously by the Hospital Trust unless we are prescribers and fully qualified; I am not a prescriber. Rules around the right to prescribe vary in different Trusts.

LL: The Hale L1-5 categories refer exclusively to risk to the baby, so in the case of lithium they are no guide to effects on the mother's condition.

LL: Domperidone is a peripheral dopamine-receptor antagonist used empirically in the treatment of reflux oesophagitis in infants, newborns, and even premature babies. There has been some concern of the possible effect of dopamine antagonists on the functional maturation of central dopamine mechanisms in newborns. It crosses less freely than metoclopramide into breastmilk, so the possible risks to the infant are reduced. A milk domperidone concentration of 1.2ng/ml after a dose of 10mg three times per day has been demonstrated (da Silva 2001). This is compared with 125.7ng/ml of metoclopramide from milk samples taken 2 hours after a 10mg dose.

LP: The diagnostic abilities of AANPs and SHOs cannot be directly compared, because of the differences in training, though I believe that they are comparable. However the training for neonatal nurses to prescribe is currently inadequate. The forthcoming permission to prescribe from the BNF is hedged round with provisos.

LL: Caffeine ingested by mothers in tea or coffee can cause restlessness in their breastfed babies.

LL: The lay view of complementary therapies including herbal teas tends to be 'Natural, therefore OK'. They are best avoided when there are unknown factors.

LP: University courses in pharmacology for AANPs vary greatly, and need to to be standardised. There is little interest in providing for our small group. As for prescribing errors, I am chiefly interested in the effectiveness of the checks and balances which have been established. RT: The National Patient Safety Agency is gathering a large database on medical, including prescribing errors, so that it should be possible to identify those which have occurred in neonatal practice.