postnatal vitamin k
sara wickham

A version of this article is also published in
Midwifery Today,

When a woman gives birth in the Western world today,
she has a series of choices to make. Some of these
choices will be more explicit than others; some will
be intrinsically linked to other decisions she has
made previously. Many of the choices a woman makes
will be framed by the person she chooses to care for
her during childbirth, who will outline the decision
for her when offering her the information relating to
the issue. The personal philosophy of the attendant
goes a long way to help this framing of the question;
we are only human, and our own beliefs form an
important part of the messages we convey to women
about the ability (or otherwise) of their and their
babies' bodies to keep them healthy.

Deciding whether or not to have vitamin K given to
her newly born baby is one of the choices a woman
must make. It is also one of those areas where the
evidence and current thinking is dominated by the
medical model. How can we bring all of the different
philosophies and evidence together in order to make
sense of this for ourselves - and for women - and to
begin building a theory which is useful for women and
for midwifery?

In the UK, the standard information given to women
in hospitals goes something like this:

"In this hospital we offer all babies vitamin K. The reason for this is that all babies are born with low levels of vitamin K. Babies need vitamin K to prevent haemorrhagic disease, which can cause serious complications. There are also low levels of vitamin K in breastmilk, so if you choose oral vitamin K and are breastfeeding we will give your baby three doses of vitamin K rather than the one which we give to formula-fed babies. (Formula contains the high amounts of vitamin K which babies need.)"

This may be followed by a further explanation of
the possible routes via which vitamin K can be given.
The attendant may also expand on some of the research
into the different routes, notably to outline the
question surrounding the risk of cancer. In my
experience, the information the woman is given about
this aspect of vitamin K administration varies widely
between attendants and institutions, while the first
part outlined above has become fairly standard.


A Question of Philosophy

An interesting starting point in this area is an
analysis of the statements which are made - with the
best intentions - by the medical model and repeated
by many of those who attend childbearing women.

1. All babies are born with low levels of vitamin K.

Some fairly obvious questions are raised by this statement:

· What is a 'low' level of vitamin K?
· Semantically, can all babies have low levels?
· Low in relation to what?
· How do we define low levels and normal levels?
· Surely someone needs to have a normal level
against which this is measured?
· Who has normal levels of vitamin K?

Actually, although the 'all babies have a low level'
argument is heard in practice and not something which
I have ever seen analysed 'in the literature', babies
are deemed to have low levels of vitamin K relative
to adult levels. Babies also have large heads
relative to adult head size, but this is not perceived
as pathological. It is deemed a good thing, because
the human brain needs to be large at birth. Yet the
fact that relative vitamin K levels differ between
newborn and adult is perceived as pathological. Why?

Philosophically, the question is raised that, if all
babies have what is perceived as a 'low' level of
vitamin K, then surely this must become the 'normal
level' of vitamin K for babies to have. Even if
proponents of vitamin K think that this is 'too low'
a level for some reason, they need to state this,
rather than telling women their baby is deficient
in an essential substance. Doesn't this just feed
back into the idea that women are relatively
inefficient at making babies and need to be
supplemented by the skills and technology of
hospitals and doctors?

2. Babies need vitamin K to prevent haemorrhagic
disease, which can cause serious complications.

This may also not be as clear-cut as some people
suggest; this seems to suggest that all babies are
at equal risk from haemorrhagic disease, when this
is not the case. It would seem to me that it would
be clearer and more helpful to give women reliable
estimates of the relative risks in this situation;
this is discussed below, in relation to some of the
reseach which has been carried out in this area.

3. There are also low levels of vitamin k in
breastmilk. (Formula contains the high amounts
of vitamin K which babies need.)
Again, 'low' levels in relation to what? To formula?
Which came first? Surely we can't be using
artificially treated and processed cow's milk
as the baseline against which to measure the
constituents of human breast milk? Yet this is
the undertone of the statements which are being
offered to some women. What does this suggest to
women about their body's ability to feed and
nurture their baby? What impact does it have
on breastfeeding rates? In any case, the research
which first suggested that breastmilk was relatively
low in vitamin K was carried out at a time when
women were told to restrict the number of feeds,
apply limits to the time the baby spent on each
breast and, in some areas, to express colostrum
without giving this to the baby. The net result
of this was a reduction in the amounts of fat-rich
colostrum and hindmilk which babies were recieving.
Vitamin K is fat soluble and so is found mostly in
colostrum and hindmilk. Would a study conducted on
women who were breastfeeding now show different results?

Even if we accepted that the originators of these
statements felt sure that they were correct in their
assumptions, putting the first and last of these
statements together raises one of the most important
questions in this area:

· Babies have relatively low levels of vitamin K.
· There are relatively low levels of vitamin K
in breastmilk.

Well, either nature (or God/dess) has dealt us a
double whammy, or the vast majority of babies don't
need too much vitamin K. Personally, considering what
I know about the process of birth, I would suggest
the second is the more likely. In fact, the
co-existence of these statements simply seems to
reinforce the idea that babies may not need this
substance for the first part of their lives.
Perhaps the relative (to adults) lack of it serves
them well, possibly preventing the development of
clotting problems in the first few weeks of life?
Of course, it may also be that medical intervention
has reduced the levels of vitamin K in both cases
and that these would be higher both in babies and
breastmilk which was not interfered with.


Research: relative risks.

As above, it may be more helpful to give women
estimates of the relative risks and benefits in
this area than to give information based only on
the personal philosophy of the attendant. What is
the likelihood of a baby developing haemorrhagic
disease of the newborn (HDN) if a woman declines
vitamin K? A figure for this was calculated by
Von Kries and Hanawa (1993), who suggested that the
risk of late onset HDN without vitamin K is between
1 in 10,000 and 1 in 25,000. Late-onset HDN may be
a serious condition for those babies who develop it,
but if between 10,000 and 25,000 babies have to be
given vitamin K in order to prevent one case, we have
to ask whether it is worth it? Or, rather, each
woman has to ask herself whether it is worth it.
This is especially relevant when we consider that
we already have information on which babies may
benefit from vitamin K; those who have traumatic
births are far more likely to develop HDN than those
who are born physiologically or, at least, without
undue trauma.

The other side of this particular decision is related
to the risks of giving otherwise healthy babies a
substance which they do not need. The risk of
healthy babies developing cancer as a result of being
given vitamin K may be higher than the risk of
developing HDN without (Parker et al 1998, Passmore
et al 1998). Unfortunately, the studies which have
looked at this have tended to use retrospective
('backwards-looking') research designs, which are
never as reliable as forward-looking prospective
trials. The only way to reliably assess the level
of risk and benefit of having vitamin K through
scientific research would be to conduct a prospective,
randomised controlled clinical trial. This is echoed
by Slattery (1994) who argues that a trial is the
only way we can establish the real risks and benefits
of vitamin K for those babies at low risk of HDN.

However, there remain problems with the clinical trial
suggestion. In order to run such a trial, researchers
need to recruit hundreds or thousands of mothers and
babies. These enormous numbers are needed so that
the frequency of very rare or occasional outcomes can
be accurately estimated. The most efficient way of
recruiting such large numbers of women is to access
them through hospitals. Which, of course, means that
very few of the babies in the study will be
experiencing truly physiological birth - a problem
which exists in many areas of midwifery practice where
the research is only carried out on medically-managed
birth. How can we then relate the results of such a
trial to women who choose to birth
'the way nature intended'?


Other Practice Issues

It is also quite possible that issues other than the
type of birth impact upon this situation. I might be
going out on a limb here, but from my experience as
a midwife and researcher, I would suggest it is
extremely unlikely that the relationship between
vitamin K levels and HDN is a simple one. I can
think of several birth-related factors which might
affect this issue. For instance, we should ask a
woman what happened during the third stage of her
labour? Was the cord cut quickly, or was the baby
allowed as much time as she needed to regulate the
amount of blood she would keep? What difference
does this make to the amount of clotting factors
and other relevant components in the baby's blood?
What impact does the woman's diet during pregnancy
have on the situation? And what are the possible
reasons that nature intended babies to have low
levels of vitamin K? Could this be an important
part of our design? And for what purpose?

Experiential evidence might also help move this
debate forward. I worked in a community midwifery
practice at a time when the decision was made to
increase the (oral) vitamin K given to breastfed
babies from one to three doses. The first dose
was given at birth, and the second on the seventh day
postpartum. While we would generally stop seeing
women on the tenth day postpartum, the other midwives
and I noticed that, almost as soon as this new policy
became practice, we suddenly had moderate numbers
of women who were not 'discharged' from midwifery
care until the 12th or 13th day. Analysis of the
records showed that the majority of these women had
babies who were becoming jaundiced on the 8th or 9th
day - following their second dose of vitamin K.
Is this coincidental? I have talked to other
midwives on Internet discussion lists who have had
the same experience; at least one of whom has
suggested that perhaps babies cannot handle the
increased prothrombin which comes about as a result
of receiving vitamin K. Perhaps this would explain
why babies are born with their relatively low levels?

Von Kries (1998) summarises some of the recent history
of vitamin K, which in some areas was not given until
the early 1980's as late-onset HDN had not been a
problem until then. This in itself should raise
concerns. Surely if all babies were pathologically
deficient in vitamin K, someone would have noticed
in these areas sometime before the 1980's? How does
the increase (in some areas) of late-onset HDN relate
to the changes in the practices women experience
during childbirth? Did the 'need' for routine
vitamin K increase alongside increasing
medicalization of birth?

Von Kries also points out that some of the babies
who were diagnosed as having HDN caused by vitamin K
deficiency actually have HDN caused by underlying
cholestatic disease. Even if giving vitamin K to
these babies could prevent the development of HDN,
we should not be saying that they have HDN because
they were not given vitamin K. This just confuses
and confounds the issue and is a bit like saying that
someone who was hit on the head by a block of wood
has a headache because they weren't given an aspirin.
The idea of giving vitamin K to all babies may then
be akin to the suggestion that we should all take an
aspirin before going outside just in case we are hit
on the head by a block of wood.

Ultimately, women may want to hear this kind of
midwifery model perspective and the medical model
viewpoint in order to choose. This may be one way
in which they can be sure their choice is informed.
There is still work to be done in developing a
midwifery model perspective on vitamin K, much of
which involves further analysis and research into
the area. Yet for those who believe in the midwifery
model some relatively simple truths remain.
Babies are born with pretty much everything they need.
The length of their umbilical cords almost always
enables them to reach their mother's breast to suckle
while their placenta is still attached inside her
uterus. That's not a coincidence. Neither is the
way the hormones of labour help the mother and baby
to fall in love with each other. For the majority
of babies, birth works very well.

For the minority of babies who are at increased risk
of HDN, vitamin K may well be a good idea. But a
clinical trial on vitamin K carried out on low-risk
babies born to women experiencing medical management
may not be helpful to those women who choose
out-of-hospital birth. Now is the point at which we
have to ask ourselves whether we really believe that
babies are born with less vitamin K than they need,
or whether there might be another explanation.


References

Parker L, Cole M, Craft AW and others (1998)
Neonatal vitamin k administration and childhood
cancer in the north of England.
British Medical Journal, 1998; 316:189-93.

Passmore SJ, Draper G Brownbill P and others (1998)
Case-control studies of relation between childhood
cancer and neonatal vitamin K administration;
retrospective case-control study.
British Medical Journal, 1998; 316:178-84.

Von Kries (1998)
Neonatal vitamin K prophylaxis; the Gordian knot
still awaits untying.
British Medical Journal, 1998; 316:161-162.

Von Kries R and Hanawa Y (1993)
Neonatal vitamin K prophylaxis. Report of scientific
and standardization subcomittee on perinatal
haemostasis.
Thrombosis and Haemostasis, 1993: 69: 293-95.

Slattery (1994)
Why we need a clinical trial for vitamin K
British Medical Journal, 1994; 308:908-910.