The author, Monica Tolfari's experiences listening to women and their families in debrief and antenatal clinics between 2014 and 2018 have revealed that the impact of massive bleeds is enormous and lasts for years, causing fear not only for themselves but also for family and friends. This fully justifies historical efforts by obstetricians and midwives to prevent maternal morbidity and death from postpartum haemorrhage (Stallworthy, 1939; Brown, 1962; 1968; Chukudebelu et al, 1963; Embrey et al, 1963; Kemp, 1963; Patterson, 1964; O'Driscoll et al, 1969; Brinsden and Clark, 1979; Weeks et al, 2021).
Personal reflection
In their early careers, both authors observed minute blood losses after spontaneous labours. Later returning to work on a labour ward, Monica was surprised to find major postpartum haemorrhage being normalised by some midwives and obstetricians.
Keenly aware of professional accountability, the Duty of Candour and a responsibility to uphold the Nursing and Midwifery Council's (2018) Code, and with the awareness that women themselves had not changed over the years, the author concluded that changes had occurred within obstetric practice (Ford et al, 2007; Knight et al, 2009). To verify this, the author compared the licensed instructions in the box of synthetic oxytocin ampoules with the local syringe driver synthetic oxytocin regime. They were significantly different, but it was not clear why practice had been changed.
Discovery
After an extensive search online, the authors discovered that synthetic oxytocin, although licensed, is being used in an unlicensed manner in British midwifery, on explicit advice from the Royal College of Obstetricians and Gynaecologists (RCOG, 2001; 2008).
Without reference to the role of endogenous oxytocin in induced labour, the removal of all options to fine tune by 0.5 milliunits per minute dosage changes (Mylan Products Ltd, 2019) or informed patient consent for unlicensed dilutions, increments and the new maximum rate, the 2001 dosage advice was re-endorsed by an RCOG-published review commissioned by the National Institute for Health and Care Excellence (National Collaborating Centre for Women's and Children's Health, 2008).
Hypothesis
If licensed dilutions and dosage options for synthetic oxytocin (and one-to-one care and fine tuning) were employed, postpartum haemorrhage rates would fall below 7% overall, as recorded prior to 1980 (Brinsden and Clark, 1978; 1979), and hasten the reinstatement of express informed consent from women for any unlicensed synthetic oxytocin practices.
Freedom-of-information request
In 2019, the authors sent a freedom-of-information inquiry to all UK maternity units, requesting their infusion devices, synthetic oxytocin regimes and blood loss statistics.
The majority (>90%) of induction of labour and synthetic oxytocin regimes were returned. All but three followed the RCOG (2001) advice, expressed in a variety of formats. The guidelines contained no evidence of consent for unlicensed practice or fine-tuning. The makes and models of volumetric pumps and syringe drivers varied more than expected, highlighting a challenge to safer practice for midwives moving between hospitals. The overall rates of postpartum haemorrhage ranged from 15–50%, confirming the elevated postpartum haemorrhage rate and the widespread loss of licensed practice with synthetic oxytocin.
The literature
The authors reviewed key research before and after synthetic oxytocin was introduced (Archer, 1969), in order to understand the globally accepted licensed instructions as they still stand today (developed by manufacturer Novartis, pre 1977–2018). It should be noted that the British National Formulary (2021) mentions its oxytocin advice may differ from the manufacturers' guidelines.
This research included discussions between obstetricians about postpartum haemorrhage rates in medical journals before 1980, statistics from NHS Digital (2016), which were supplied by the maternity units, online synthetic oxytocin policies and regimes for compliance with the licence, pump specifications and a meeting with a manufacturer of motorised infusion devices, as well as an investigation of fetal and maternal acidosis. The key findings are shown in Box 1.
Box 1.Key findings from the literature review
- Published postpartum haemorrhage rates >1500ml range from 1–6% (maternity dashboards)
- Bovine post-pituitary extract contained oxytocin and vasopressin until these could be separated
- Pure bovine oxytocin (Archer, 1969) was used until synthetic oxytocin was developed for clinical use in intravenous infusions and oral preparations
- Postpartum haemorrhage rates were commonly 3–5% (Brinsden and Clark 1978; 1979)
- Postpartum haemorrhage fatalities occurred in the 1960s
- Desensitisation of oxytocin receptors occurs in the presence of excess oxytocin
- Quality control for error factors is intrinsic to infusion device manufacture
Berde (1965) commented that obstetricians calculate bovine oxytocin or synthetic oxytocin as ‘1 mg=450 000 milliunits’. He observed moderate lowering of blood pressure after experimental pharmacological doses of synthetic oxytocin. In the formal discussion, Berde acceded to Theobald that physiological dilutions better replicate hormonal action, including mild hypotension (Pinkerton, 1965).
In previous administration systems, infusions ran at 40 larger drops per minute and oxytocin concentration could be increased (Barley and Ripman, 1970). Normington (1972) researched three standardised dilutions.
Theobald et al (1948) noted that too concentrated a solution of bovine post-pituitary extract induced ‘irregularity of the fetal heart’, leading to their advocacy of well-diluted solutions of this ‘powerful hormone’ at 5 milli International Units (IU)/minute for all cases (Theobald, 1966a; b), including successful, labour induction to pre-empt eclamptic fits (Theobald, 1959). While respectfully acknowledging Calderyo-Barcia's maximum of 8 mIU/minute, Theobald (1965) asserts ‘…it has never been shown that the pharmacological drip is more effective than the physiological one, and it is certainly less safe’. No subsequent study of pharmacological doses confirms that there is a benefit in every measurable category.
Appreciating and analysing women's care needs
The term ‘patient’ in this article is used to draw attention to the status of a (healthy) pregnant woman now receiving synthetic oxytocin infusion, changing from normal to highly-dependent obstetric care because of the risks of synthetic oxytocin infusion (Pinkerton, 1964) and an epidural cannula and anaesthetic drugs (if chosen), during labour.
As unlicensed regimes flowed from experimental research that intentionally varied dilutions and increments from the advice licensed, midwives are empowered to inform families of the licensed alternatives and support women who, when choosing one of them now, have a wider choice of pain-relief during labour, and when an epidural cannot be offered following a caesarean section (Moodie and Moir, 1976; Walsh, 2009).
As some practitioners across the world work without electricity, the ‘recommended dosage range’ is based on the standard dilution for drip infusion: 5 IU in 500 ml (Mylan Products Ltd, 2019).
Oxytocinmeasures.com (2021) presents graphics, figures and tables related to synthetic oxytocin dosage calculations, including how to relate the full range of licensed dosages (for live pregnancy) by drip infusion to volumes-to-be-infused via infusion pump. The start rates are 0.5–4 mIU/minute and subsequent rate increases must range between 0.5 mIU and 2 mIU/minute only.
Calibration of synthetic oxytocin for intravenous infusion
Synthetic oxytocin by weight
IU define organic (bio-) chemicals; 1 mg synthetic oxytocin is equal to 600 IU biological activity.
Synthetic oxytocin by weight in diluent
Synthetic oxytocin 1 ml ampoules come as 5 IU or 10 IU concentrate, which contain 8.3–8.5 micrograms or 16.66–17 micrograms respectively, as shown in Box 2 (Kent Pharma Ltd, 2018; Mylan Products Ltd, 2019).
Box 2.How to calculate mIU in a drop of synthetic oxytocin1 ml = 20 drops. 1 International Unit (IU) = 1000 milli International Units (mIU)5 IU = 5000 mIU/500 ml = 10 mIU/ml10 mIU/20 drops = 0.5 mIU/dropIf 1 drop = 1 ml/20, then 1 drop = 0.05 ml8.3 micrograms/5000 mIU = 0.00166 micrograms/mIU0.00166/2 = 0.00083 micrograms per drop or 0.5 mIU/0.05 ml
Pre-prostaglandin era
Prostaglandins were being researched prior to 1977, when the current licensed instructions were published, and their use was not standard when reference to it appeared in the instructions in 1977. Hospitals were still using drip infusions, and so practice was altered by prostaglandins and motorised infusion pumps becoming available.
The instructions agreed by the UK licensing body balanced more synthetic oxytocin in less diluent for fewer drops per minute via giving sets with a controller, to address medical concerns about water intoxication. The instructions advise that if only irregular (or no) contractions occurred during the first 5 IU infused, the infusion should stop overnight. Water intoxication concerns (Feeney, 1982) were addressed by new medication for evacuating the uterus up to the second trimester. Buccal and intravenous synthetic oxytocin were administered experimentally for cervical ripening (Coltart and Nash, 1974).
Based on pre-synthetic oxytocin research, the design of treatment with synthetic oxytocin aims to promote ‘a contraction pattern similar to that of normal labour’ (Mylan Products Ltd, 2019), offering a straightforward system for measuring and calculating doses based on one accurately-calibrated mixture. The mixture was weak enough to individualise by fine-tuning in titration with the patient's own oxytocin and contractions, or be halved or doubled as clinically indicated (Mylan Products Ltd, 2019).
Patient stability and fetal wellbeing are safeguarded by the expertise that was given to determining and wording the licensed instructions, as long as the infusion is managed accordingly.
In 1977, two changes were made to the licensed instructions for synthetic oxytocin by the manufacturer, Novartis. These were to delay synthetic oxytocin infusion for 6 hours after prostaglandins therapy, and to use ‘a motorised, variable-speed infusion pump’ if possible.
Dilution
In this article, the word dilution is used to signify synthetic oxytocin concentrate that is diluted for clinical use, whether or not the resulting infusible mixture is ‘physiological’ (licensed for pregnancy), or strong enough to be deemed a drug (pharmacological, unlicensed for pregnancy).
There are three licensed dilutions:
- 5 IU in 500 ml diluent (standard dilution)
- 10 IU in 500 ml (20 mIU/ml) after intrauterine death
- 20 IU in 500 ml, used to prevent or manage postpartum haemorrhage.
All unlicensed practices oblige clinical justification and valid informed patient consent.
Unlicensed synthetic oxytocin usage
The following practices with synthetic oxytocin are currently all unlicensed:
- Synthetic oxytocin concentrate that is not fully diluted before intravenous infusion
- Intravenous bolus of any dose for any reason at any speed
- Intramuscular injection >5 IU synthetic oxytocin (instead of 1 ml Syntometrine), including injection of 10 IU, 20 IU or 30 IU in theatre post-caesarean section (Svanstrom et al, 2008)
- 40 IU synthetic oxytocin in 500 ml diluent (instead of 20 IU in 500 ml)
When the licensed dilution has been used in labour (or none), 20 IU in 500 ml will control postpartum haemorrhage. To justify exceeding 20 mIU/minute during labour, an experienced obstetrician should examine the patient for clinical indications for this new unlicensed treatment. If it is justified, the dose must be separately prescribed to proceed.
De-sensitisation of oxytocin receptors
Theobald (1959) unwittingly elicited oxytocin receptor desensitisation before it was formally identified. So, in 2016, the authors privately asked Dr Gareth Leng, Professor of Experimental Physiology at Edinburgh University, whether too much synthetic oxytocin blocks the release of natural oxytocin. He replied ‘If you give too much [synthetic] oxytocin when the uterus is fully ready – when there are lots of [oxytocin] receptors – you will get a strong initial response but you won't sustain that response, because again the receptors will disappear [inside the cell]. You need to give enough to activate the receptors, but not so much that you desensitise them’ (private correspondence).
The authors suggest that the correct use of the recommended dosage options accommodates this sensitivity, which is obliquely raised by Weeks and Neilson (2015) in ‘Rethinking our approach to postpartum haemorrhage and uterotonics’.
Pump error factors
An error factor is the permitted percentage variation in accuracy during use, as specified by the hardware's manufacturer. Volumetric pumps are permitted to vary ±5%. At very slow rates, this permitted variance may increase to ±10%. For syringe drivers, the permitted variation is ±2%. Each pump at its best has a unique, inbuilt error factor, which, since synthetic oxytocin causes pain, adds another cause for caution during synthetic oxytocin administration, and fine-tuning with contractions.
The stronger the synthetic oxytocin concentration, the higher the mIU in the smallest volumes (after the decimal point) for volumes to be infused. When 10 IU in 1 ml synthetic oxytocin is added to 49 ml diluent (in syringe drivers), it provides 200 mIU/ml, which is 20 times stronger than licensed for live pregnancy (Oxytocinmeasures.com, 2021). The authors recommend that health professionals consider using 5 IU in 50 ml instead when using a syringe driver, which produces a dilution that is only 10 times the licensed dilution and may therefore be marginally safer, though this is still an unlicensed dilution. In general, the authors would not recommend the use of syringe drivers for the administration of synthetic oxytocin.
In volumetric pumps, slow volumes to be infused militate against unlicensed concentrations, where times would be lengthened to attempt fine-tuning. Dosage accuracy, safety and ease of titration with contractions, all increase with weaker dilutions.
Managing synthetic oxytocin by motorised infusion pump
Disable the bolus function
Motorised infusion devices vary in design depending on the make, so it is important to comply with each one's specific operational routines and limitations to administer synthetic oxytocin safely. On the pump, or in its software, the bolus option should be disabled permanently or routinely prior to synthetic oxytocin infusion, because with unlicensed dilutions even a failsafe-sized bolus may elicit an adverse maternal or fetal response (technical help or a password may be needed to do this).
Titrate with contractions
Licensed dosage infusion rates are to be strictly limited by the evoked contractions. The standard dilution (5 IU in 500 ml of diluent) offers 20-minute volumetric pump settings: 10 mIU (1 ml), 20 mIU (2 ml), 30 mIU (3 ml) and so on up to 400 mIU (20 ml) (Oxytocinmeasures.com, 2021). This regime is suitable for a trial of synthetic oxytocin infusion to induce contractions, as is 2.5 IU synthetic oxytocin in 500 ml, which the World Health Organization (2017) recommend as a possible starting dilution. This half-strength dilution should be considered for sensitive, parous, young, or small women (Mansy, 2017; Oxytocinmeasures.com, 2021).
Challenges to homeostasis
When chronic fetal acidosis is already present (because of placental dysfunction in elevated blood pressure during pregnancy, other causes of fetal growth retardation and other clinical pictures), Bobrow and Soothill (1999) warn that ‘anything that causes hypotension…will reduce the maternal blood supply and so oxygen delivery to the uterus’. Mechanically lowering blood pressure when an epidural first takes effect has been considered an advantage by obstetricians to hypertensive women approaching labour, but when normotensive patients' blood pressure falls, plain intravenous fluid is rushed in to restore it. The estimated blood volume per kilogram body weight in women is 65 ml according to Manuelsweb.com (2017). This takes haemodilution beyond the physiological plasma increase of 40-50% by term (Hytten and Paintin, 1963), adding to the cardiac stress of labour.
Myometrial hypoxia during contractions temporarily reduces oxygen to the fetus and naturally causes physiological pain. Repetitious, unnaturally severe contraction pain from pharmacological synthetic oxytocin is likely to lower blood pressure (Berde, 1965). This is a more profound physiological trauma that necessitates epidural analgesia to prevent shock (Ebirim et al, 2012) and potentially overworks the myometrium (Brinsden and Clark, 1978). Maternal and fetal acidosis may ensue. Research into the long-term effect of treating the uterus this way is lacking.
Metabolic acidosis causes vasodilation (Salameh et al, 2014) and lower ‘intracellular calcium concentration … could explain the reduction of smooth muscle tone’ (Aalkjaer and Peng, 1997; Aguilar and Mitchell, 2010).
Oxytocin receptor desensitisation, and vasodilatory influences correlate with atony after synthetic oxytocin regimes, which overwork the myometrium, making postpartum haemorrhage more likely (Brinsden and Clark, 1978).
In a literature review, Budden et al (2014) observed that ‘previous studies suggest that high-dose regimens can lead to shorter induction to delivery time and fewer failed inductions, but at the expense of increased rates of hyperstimulation and fetal distress, requiring cessation of oxytocin, caesarean section, instrumental birth, and postpartum haemorrhage’. The anaesthetists Griffiths and Campbell (2015) state that ‘local anaesthetics can accumulate in the fetus due to ‘ion trapping’ if the fetus becomes acidotic…when the decreased pH in the fetus produces an increased proportion of ionised drug’.
When analysing their comprehensive study, Maintenance of Labour, Beazley et al (1975) found 7 mIU/minute was the optimum synthetic oxytocin rate, with no caesarean sections for acidotic fetal distress (Steer et al, 1975). Clearly, inhibition of maternal and fetal acidosis is clinically desirable, being favourable to sustaining individual variations in time to delivery.
Managing synthetic oxytocin infusion
Synthetic oxytocin must be administered in a hospital, with one-to-one care, because known risks and unpredictable changes may demand prompt action from professionals (Dupont et al, 2017). Infusion rates are per minute to facilitate quick correction (allaying fetal distress) if uterine relaxations after contractions last less than 60 seconds (MacKenzie et al, 1989; Pehlivanoglu et al, 2013).
Midwives should compare the length of palpated contractions with those recorded on the tocograph trace, and use only the clinical finding to determine whether or not to increase the synthetic oxytocin infusion rate (Bhogal, 2017). Rate increases are guided by the frequency of contractions (not by the clock), as these prompt greater endogenous oxytocin and prostaglandins output (Vrachnis et al, 2011).
To leave the patient (professionally) unattended, midwives should record turning off the synthetic oxytocin infusion. On return, they should assess, time and record the contractions, the fetal heart (using Pinard's stethoscope), maternal pulse and blood pressure. If all are satisfactory, the midwife should choose a slightly lower rate of synthetic oxytocin than when paused, to make way for endogenous oxytocin and enhance these contractions safely. They should also record the time and infusion rate after restarting, and palpate the next few contractions for strength, length and frequency.
Establishing and completing labour
Despite the manufacturer's warning to space increments at least 20 minutes apart, quarter-hourly changes persistently appear in trials. This practice was proven by Loscul et al (2015) to raise postpartum haemorrhage rates.
As blood volume is related to maternal stature, an allowable blood loss estimate, if stated in case notes antenatally (36 weeks), might sharpen focus towards minimising postpartum blood loss (Manuels web.com, 2017; Butterworth et al, 2018).
To the authors' knowledge, midwives in the UK work without fine-tuning options. Formal local permissions to fine-tune pump rates downwards during volume to be infused time segments are essential for improvements to postpartum haemorrhage rates to materialise. Volume alterations for fine-tuning will vary according to the dilution in use. Examples can be found at oxytocinmeasure.com (2021).
To avoid oxytocin receptor desensitisation and tetanic contractions (causes of fetal distress) during the establishment of labour, fine-tuning synthetic oxytocin up or down by 0.5 mIU/minute, as indicated by uterine relaxations lasting longer than 60 seconds, maximises fetal oxygenation and biochemical homeostasis (Pehlivanoglu et al, 2013).
Robust natural contractions often last longer than 60 seconds. Once labour is established, the licensed instructions advise that the infusion rate may, tentatively, be reduced by 0.5 mIU/minute more than once, to support the natural rhythm of only three (or maximum four) contractions in 10 minutes.
Total analgesia under epidural protects the patient from pain-consciousness, but also blocks sensory information complicit with effective fetal expulsion. With consent, epidural top-ups may be halved to allow sensation to return before pushing, for an unassisted vaginal birth.
Fitzpatrick and Walmsley (1965) estimated that expulsive contractions begin under the influence of an elevated supply of endogenous oxytocin. From experience with physiological synthetic oxytocin dilution, it can be discontinued, if clinically approved, without compromising progress in the second stage (JHP Pharmaceuticals, 2009). Therefore, that is the point at which the synthetic oxytocin infusion can be reduced by 2–4 mIU steps to end it prior to the birth, giving maternal oxytocin as much time as possible to adjust for the third stage to achieve natural haemostasis after expulsion of the placenta.
Intramuscular Syntometrine (1 ml) will assist separation of the placenta and lessen blood loss (Brown, 1968). During the first hour after the third stage, midwives' time would be well spent by checking the uterus manually, very regularly, to ensure that it is well contracted in all cases, or take appropriate remedial action to control bleeding (Hofmeyr et al, 2013).
Interpretation of the severity of impact of a woman's excessive blood loss is extrapolated from her allowable blood loss estimate, combined with other relevant clinical factors, to plan postnatal support (Thompson et al, 2010; RCOG, 2016; Association of Ontario Midwives, 2017).
Conclusions
The body is intelligent. Obstetric teams must respect and work with it, not against it, as it is affected by its external and internal environments. The uterus has the ability to start and complete labour by itself. Therefore, the authors strongly advocate for the full gentle ramp of infusion rates in the recommended dosage range to be re-embedded in local synthetic oxytocin policies, to restore the options missing from the RCOG steps regime.
Overall postpartum haemorrhage rates were <7% when bovine oxytocin or synthetic oxytocin was blended as intended with endogenous oxytocin to facilitate homeostasis and maximise the efficacy of contractions. Current clinical outcomes have failed to justify two decades of unlicensed practice with synthetic oxytocin, boosting postpartum haemorrhage rates in some hospitals to close to 50%.
Hospital guidelines disallow fine-tuning and are expected to be followed as ‘best practice’. A midwife will be sanctioned for departing from these guidelines, whereas licensed practice gives physiologically beneficial flexibility to adjust synthetic oxytocin up or down, as well as to stabilise or stop it.
Once upon a time, low postpartum haemorrhage rates were admired and coveted, but the 2001 change in synthetic oxytocin practice has not been undertaken with a readiness to reverse it, if proving detrimental to the ladies. The RCOG's advice removed too many infusion rates for meaningfully individualised care to be given. It does not take into enough consideration the reasons for varying levels of sensitivity to synthetic oxytocin, nor spontaneous endogenous oxytocin release.
Valid informed consent should also be obtained for the licensed dilution (and increment limits) or half-strength synthetic oxytocin dilution, because when optimally titrated, both are clinically effective to enhance, induce, and support as natural a labour as possible, and they significantly mitigate known pump error-factor slippage, that varies with non-standard synthetic oxytocin dilutions and each unique infusion device.
The authors are confident that without harm, or further research, licensed practice with synthetic oxytocin could be re-introduced to improve women's experience of labour, significantly diminish postpartum haemorrhage rates and maternal morbidity, and bring notable improvements to fetal outcomes.