For reference: Technical notes on obstructed labour and assisted delivery in rural Tanzania

Obstructed Labour and Assisted Delivery in rural Tanzania

Notes for professionals – and the technically-informed curious.

The puzzle of obstructed labour in rural Tanzania

The strange facts needing to be reconciled are:

1.      1. Most of the emergency CS’s in the UK are due either to a CTG diagnosis of fetal distress, or to a long, slow, seemingly dysfunctional labour; or both. Long slow labours are typically the product of induction of labour, or are primips in labour where the diagnosis of active labour was made too early. Fetal distress mainly comes from CTGs.

Where we have no induction, nor CTG, nor anxious primips coming in too early, the CS rate in Berega should therefore be much lower than the UK. However, it seems, if anything, to be higher – more than 30% of deliveries. This is presumably because we are skimming off the caesareans from a much larger population than just those who give birth at Berega.

2.       2. Obstructed labour is far and away the most common indication for CS. A few CS are done for eclampsia, twins, breech, post-dates with very high head, etc. However, far and away the most frequent indication for CS (90% in my two months here) was obstructed labour.

This is wildly different from the UK, where although the CS rate is more than 25% of pregnancies, only very few of these – 1 or 2% of deliveries if that – have full-blown obstructed labour.

3.       3. You may conjecture that this might be due in part to over-diagnosis of obstruction. However, when women get obstructed labour in Berega, the diagnosis is typically barn-door. The head is very high abdominally: often 3-4/5 palpable despite eg 14 hours or more of active labour – even in multips with several previous normal deliveries of seemingly similar-sized babies. In the UK, obstruction can be difficult to diagnose, as the head is often well down, and the diagnosis is less obvious.

4.       4. In obstructed labour in Tanzania, caput seems to be an important sign. They also have the classical signs, (see below). However, the heads seem to get more severe caput than they do in the UK – perhaps just a feature of longer labours not dealt with earlier? Or something to do with the traditional medicines taken?

5.       5. When reviewing antenatal women, the head just about never engages. It is either free or 5/5 palpable right up to labour.

6.       6. Obstruction is almost always in the first stage. In two months here, I have seen around 70 CS, but only three second stage obstructed labours. Two were an easy Kiwi rotations, in the presence of a tired mother, and an OT or oblique OP position, where there was minimal caput and moulding, and decent descent of the head with contraction. The other was the opposite – lots of caput, (difficult to determine position), and stuck right at the outlet. In this one, a Kiwi pull failed, and CS was needed. The baby was born in poor shape with a very sausage-shaped head.

7.       7. I did one other successful Kiwi – for eclampsia, a delightfully easy delivery, making the Kiwi rate around 2% in my time at Berega. (3 of them successful, one not.) I did no forceps, nor (surprisingly), did I see an indication.

Again, this is in sharp contrast with the UK, where the combined vacuum + forceps rate is often 15% or more of deliveries. If we are indeed dealing with the problems arising from a much bigger population, then we should have a higher instrumental delivery rate than the UK, not a lower one.

It is true that many instrumental deliveries are done for CTG problems in the UK, but many also are done for failure to deliver, often associated with malposition. Even when the baby is OA, the need for forceps to deliver because of a tight fit is common. Here I did not see it once.

8.       8. In Berega, fetal distress is rarely an indication for CS or assisted delivery – probably related to the lack of CTGs. FH auscultation is done badly. Nevertheless, for all that, when the baby was actually born vaginally, it was rare during my two-month stay, (possibly once or twice a month), for it to need resuscitation.

However, I did run an intensive update on resus, and I have no idea what the death rate from vaginal delivery was before I came.

9.       9. According to the Berega records, the death rate of babies in obstructed labour was horrifically high before I arrived. The combined perinatal mortality rate over the 7 months 1^st Dec 2012 to 30^th June 2013 was 9% of deliveries. In the UK, it is below 1%.

Some of this was poor resuscitation of babies born in poor condition, often after a GA caesarean*. (Despite annual resuscitation updates from a highly skilled paediatrician, the skills obviously waned without constant reinforcement – a salutary lesson.)

In other cases, a few babies died, even when I was there, from being un-resuscitatable after CS for obstructed labour. This presumably relates either to late presentation, or, more commonly, to failure to pick up profound fetal distress early enough.

10.   10. The huge majority of women with obstructed labour admit having taken local medicines.

11.   11. This all represents a hugely different pattern of obstruction in Tanzania from my previous observation of Zulu labour – also a Bantu race. (I worked in a rural hospital in KwaZulu-Natal for four years.) There, outlet obstruction is far more common, and first-stage obstruction is rare. Vacuum delivery was a common need – sometimes with symphysiotomy, (which was preferable to what would otherwise have been a dreadfully difficult CS, with the head crowning but undeliverable vaginally).

Zulu women in general have much more of a sticking-out bottom – rural Tanzanian women are often flat by comparison. (They are also thinner and less female-shaped than their urban sisters.) Zulus are also considerably taller, and often stocky. Rural Tanzanian women are typically below 160cm, and are almost always slender – sometimes markedly so.

Reflection

The differences in body-type between KwaZulu-Natal and rural Tanzania may be due racial differences – eg Tanzania includes Nilotic tribes who are traditionally slimmer. However, Zulus are Bantu, and their antecedents originated in East Africa just a few centuries ago. It may therefore be that malnutrition in puberty in Tanzanian girls causes failure of development of the full gynaecoid pelvis, perhaps related to insufficient oestrogen.

A second conjecture relating to the markedly different pattern of obstruction on Tanzania, is the common consumption of labour-enhancing local drugs. 

Clearly, when a primip not in labour is given such a drug, it may readily cause distress before it causes delivery. Similarly, in a multip given such a drug inappropriately in early labour, she may end with rupture.

However, I wonder whether these drugs also interfere with the natural mechanism of labour? We have already established that as many as 90% of women referred in from the community have already taken these substances. Some of these women should not have developed obstruction – eg previous quick normal deliveries of a similar-sized baby. Do these local medicines somehow cause a type of contraction which has all the harm of squeezing the baby and exhausting or rupturing the uterus, but without the enhancement of natural labour? Is the cervix or pelvis or pelvic floor somehow not ready?

Finally, we know that for every delivery we do in the hospital, at least ten occur in the community. Perhaps our obstructed labour workload comes from the arrival at hospital of those women from the villages who get stuck. Most other causes of problems in childbirth – eg fetal distress, breech – would not cause a woman to travel a long way on bad roads on a motor-bike. Obstructed labour, however, means that she will not otherwise deliver in the village, and so the only other option is to die.

An interesting mathematical conclusion thus arises: We could examine our CS indication rate over a long period, and then compare the rates of ‘eclampsia’ and ‘obstructed labour’. We could then compare this ratio to that in a major city. 

Assuming that most village-based eclampsias do not arrive at hospital, (a long way to bring a fitting lady on a motor-bike), but that most obstructed labours do, the comparison of the two rates may allow us to infer how many extra obstructed labour cases we are doing, by comparison with eclampsia.

It would thereby give us an rough idea of how many women are delivering in villages in our territory, from which we could calculate how many common problems must be occurring there – eg breech, twins, eclampsia, fetal distress, etc.

Suggestions for the management of labour in Tanzania

1.       1. The diagnosis of labour is critical. The partograph is meaningless if the woman begins it many hours too early, or many hours too late. Therefore staff should not only be taught but also tested and checked within in-service training, (IST), on the diagnosis of labour.

2.       2. Using cm of dilatation to diagnose labour can produce considerable inter-observer variation. Nevertheless, the midwife needs to be clear that the os is not just a multip’s os, but is in fact at least 3-4cm dilated. More important however, than trying to tell the difference between 2cm dilated and 3cm, the midwife needs to be absolutely clear that the cervix is fully effaced; and that contractions are coming regularly, at least every 5 minutes. They need physically to feel the entirety of the contraction with their hand.

At full-term, a working definition of active labour is the combination of:

·  

- cervix open at least two fingers; plus

- cervix fully effaced and thin; plus

- regular spontaneous contractions each lasting 40 secs or more, coming every 5 minutes or more 

3.  Once labour is diagnosed, the woman should be entered on a partograph. Use of the partograph also needs IST. Ideally, a woman should not be left alone once active labour is diagnosed. However, that is impractical, so a bare minimum of observations need to be insisted upon, to ensure that the woman at least is seen and checked every half an hour in the first stage. (And every 5 minutes in the second). 

Most importantly, at these times, the FH needs to be auscultated – after a contraction. This again needs both in-service and classroom-based updates.

The normal observations of pulse, BP, and contractions can be recorded at appropriate intervals, depending on clinical circumstances.

4. In addition to in-service training, midwifes need regular in-service checks to ensure that they are able to tell the difference between a normal FH, and an abnormal one.

5. Whenever a labour is abnormal, the doctor needs not only to be informed, but to come and check the woman.

6. If second stage has not occurred by the transfer line, the woman should be transferred.

7. If no second stage has occurred by the action line, but the FH is still good, then the whole situation needs assessment. However, pay particular attention to the station of the head. If it is still 2/5 palpable or more in rural Tanzania, then obstructed labour is extremely likely. If the labour has been long, with good contractions, but the head is still high in the abdomen, then a vaginal examination should be done, but will probably just confirm the diagnosis of obstruction, and a CS should be ordered. (Rather than augmentation and/or giving time for the obstruction to become impacted.)

8. Membranes should only be artificially ruptured when not to do so would probably deny the woman the chance of a normal delivery. The typical situation in which ARM would be appropriate would be a slow labour in a multip, where contractions could be better; where the FH is good; where the head is not high abdominally; where the presentation is definitely uncomplicated cephalic; where the position is probably OA: and where plenty of time has already been given for SRM, so the next step would otherwise be to call the doctor.

Even then, ARM should not be performed in someone about to be transferred, as labour will get stronger. Hypoxia of the baby secondary to obstruction does not occur if the membranes are intact. Intact membranes also protect against cord prolapse, when the presenting part is ill-fitting.

Suggestions for the diagnosis of obstructed labour in Tanzania

When the doctor is called, s/he has to decide whether the labour is obstructed. A suggested guide is:

1. In rural Tanzania, it seems that the head stays high in the abdomen in obstructed labour. The combination of a long labour, (where diagnosis of labour had been good); plus good contractions actually palpated; plus high head abdominally, is very likely to mean obstruction.

2. As mentioned above, caput seems commoner and more severe in obstructed labour in Tanzania.

3. Additionally, the accompanying factors are typically: on vaginal examination there is an empty pelvic curvature at the back of the V/E, even though at the front of the vagina the head is easily palpable; no descent with a good contraction; moulding should be present, (if sutures can be felt); poorly applied swollen cervix; and eventually vulval oedema and haematuria.

4. Once the labour is actually obstructed, the baby will increasingly get hypoxic, and previously clear liquor will become increasingly meconium-stained.

Suggestions for the use of Kiwi Omnicup and Forceps in Tanzania

The Kiwi Omnicup has an important place in Africa, where it may not only prevent caesarean section, (and therefore problems in future pregnancy), but may save life. However, it use should be considerably more limited than is normal in the UK. My suggested rules are:

1.       Certain rules should be as for the UK: Fully dilated; Head not palpable per abdomen; head comes down with contraction; placement of the Kiwi on the flexion point; no other contra-indication to Kiwi such as HIV; etc.

2.       Kiwi Omnicup should never be used for (relatively) obstructed labour in OA position, I believe. A technical reason for saying this is that in OA position, the tendency of the omnicup is to increase deflexion, whereas in OP position, it corrects deflexion.

However, more important than this is the balance between risk of good and risk of harm. To do an omnicup delivery for slow progress in OA position (ie when trying to overcome a tight fit), would be to suggest that even though the head is correctly placed, by traction on just a small area of the baby’s scalp, you would hope to succeed in the large majority of cases, and only rarely get unfavourable consequences for the baby. That does not usually make sense to me.

I think that too often, the rural Tanzanian pelvis would be fooling you into thinking that vaginal delivery was possible, whereas in fact, the mother had only just made it to full dilatation. You would therefore fail too often, and the baby in these circumstances might have suffered from the intervention – eg cephalhaematoma, or even cerebral haemorrhage. Additionally, a failed Kiwi in such cases would make the CS more difficult, and therefore the baby would be more hypoxic.

3. The Kiwi omnicup comes into its own for correction of malposition – OP or OT, with or without asynclitism and/or deflexion. In prolonged second stage – or earlier if the baby is distressed – the Kiwi will very often turn the baby readily on the first pull.

Thereafter it will come quickly, as long as the pull is correct, the pushing is maintained, and the contractions are satisfactory. Such Kiwis will save unnecessary CS. Despite a small risk of cerebral bleed, they will often help prevent severe hypoxia in the baby, and the balance is therefore very much in their favour.

This is particularly the case, as there are some mothers where delivery should be easy, but malposition prevents it from being so – eg twins, grand multips. Nevertheless, such indications probably do not amount to much more than 1% of deliveries.

4. Remember that the first pull especially must pull the head into the levator ani, and into the woman’s perineal body. If the baby's head is pulled into the woman’s pubis bone, clearly it cannot advance. The perineal body, by contrast, is soft, and the levator ani naturally assist in the turning of the head. PULLING THE HEAD TOWARDS YOUR CHEST, YOU MEET HARD PUBIS. PULLING THE HEAD TOWARDS YOUR GROIN, YOU MEET SOFT MUSCLE. SO PULL DOWNWARDS!

The Kiwi should not be pulled above the horizontal until the head is half-born!! (It’s just physics – with less than half delivery of the head, if you pull above the horizontal, the fulcrum is passing through the pubic bone.)

5.The other important place for the Kiwi is when a quick delivery is required in the absence of prolonged labour – eg eclampsia; fetal distress; second twin; abruption. Such indications currently probably do not add to more than 2-3% of deliveries in rural Tanzania, but with better fetal monitoring, the diagnosis of fetal distress would increase. (And with CTGs, it would increase too much!).

In such situations, especially in multips, a Kiwi can be quick and simple, and will often save at least the brain, if not the life, of the baby. A CS for mother can also be avoided – eg distressed second twin with high head, eclampsia with poor maternal effort; etc.

In such a real crisis in a multip, you may get a good Kiwi delivery even at 8cm. In a grandmultip, she just needs to be in good labour, and you can push the cervix away!

6. I suggest that non-rotational forceps currently have little place, though this may change. They need the head to be OA, and they need good analgesia and a decent episiotomy. In the UK, one cautious and judicious pull with (eg) Neville-Barnes forceps can make the baby deliverable where this was uncertain. By contrast, it can confirm the diagnosis that the baby is stuck and CS is needed. In this case, the forceps can be used to disengage the head before they are taken off.

However, stuck OA head in second stage seems very rare in Berega – I have not seen one. Furthermore, where it does occur, I conjecture that it would be common for the head to be truly stuck and vaginal delivery impossible.

The place for Neville-Barnes forceps is thus very limited – for a woman who needs vaginal delivery to be expedited; where the head is OA; she is fully dilated with no head palpable abdominally and no other signs of obstruction, (in particular, be wary of caput); she has sufficiently good analgesia or does not need it; and the operator is skilled with the instrument.

Effectively, this amounts to certain clearly-defined situations: fetal distress not caused by obstruction; eclampsia; after-coming head of breech; abruption; poor maternal effort; the need to protect the mother – eg in heart disease; needing to avoid a Kiwi – eg HIV; etc.

In such circumstances, the availability of skilled forceps delivery is vital, so forceps still have an important place.

It is likely that as obstetrics improves in rural Tanzania, these indications will become more common.

*Post script

In Berega, GA CS is far too common. The typical reasons were either because of having run out of spinal lignocaine, or because of a cautious anaesthetist not wanting to give a spinal to a mum with a low Hb, in whom the ward staff had omitted to give an IV fluid bolus on the way to theatre.

The GA is done with a variety of induction agents, often added together ad hoc until she is asleep, then halothane plus mask. There are no facilities for intubation, and no antacids are given. Post op, patients are left alone, on their back, often deeply asleep for many hours, untended, with an airway in. 

When visitors ask what they can bring, one idea is battery-driven pulse oximeters and BP monitors, and then put them right in the hands of an anaesthetist you trust. Insist on their use when you do a CS. And regularly without embarrassment ask "What are her vitals?"