The UK is a leading example of honest, transparent and resolute approach to healthcare deficiencies. Recently, there has been a succession of independent inquiries into unsafe maternity services in some NHS hospitals namely, Morecambe Bay, Shropshire and East Kent. The grief of families who experience loss of life or serious morbidity is immense. An excessive number of adverse events in a particular hospital and poor practices rightly demand an independent transparent investigation. However, one may get an impression that the maternity services in the UK are particularly substandard or wanting, thereby increasing the anxiety amongst the service users and distrust when outcome is poor. The censure and dissatisfaction are particularly perplexing, given that British midwives and obstetricians are rigorously trained to provide skilled professional care of a high standard. This article explores the reasons behind this paradox and considers solutions.
Uniquely, the maternity services deal with a natural joyous welcome event in life. But, being born is considered the most dangerous time in one's life. There is a difficult balance to be struck between not making the childbirth very unnatural (highly medicalised) and delivering near-perfect outcomes if feasible. Moreover, medical interventions themselves have some risks. A critical look at the public expectation, communication, politics and the media coverage seems apt.
Improvements in maternity care
Substantial advances have occurred in the science of healthcare of mothers and babies over time. The NHS is said to have been one of the safest places in the world to have a baby; with stillbirths, perinatal mortality and maternal mortality down (from an already low level) by 21%, 15% and 14% respectively over the last 10 years, despite increase in some risk-factors (Government Equalities Office, 2020). The General Medical Council, 2015) and the Nursing and Midwifery Council ([NMC], 2015) have given guidance on openness, honesty and the professional duty of candour. There is also a statutory duty to take certain steps when a ‘notifiable incident’ occurs. For many years, the Clinical Negligence Scheme for Trusts has been setting standards for the delivery of safer maternity care (Maternity incentive scheme, 2020). The Morecambe Bay inquiry (2015) examined maternity services between 2004–2013 by the University Hospitals of Morecambe Bay NHS Foundation Trust and came up with a raft of recommendations. The failures at the Shrewsbury and Telford Hospital Trust are again largely contemporaneous or predate those at Morecambe Bay. All NHS trusts have been improving organisational structures, increasing safety protocols, guidelines, team training, continuous professional development, staff appraisals, adverse event reporting/investigation, openness and improved communication with the families. In 2015, the Royal College of Obstetricians and Gynaecologists (RCOG) launched a national quality improvement programme to reduce the number of babies who die or are left disabled during term labour (Each Baby Counts, 2015). The Healthcare Safety Investigation Branch ([HSIB], 2017) of the NHS has a major focus on maternity adverse events. However, one may have to be patient about the degree and timescale of the results of the improvements. Invariably, there are likely to be instances of ‘falling short’ with a scope for correction. However, these may best not be portrayed as failure or unwillingness to implement the recommendations of the inquiries.
Adverse events in maternity care
The reasons for adverse events in the maternity care can be broadly divided into following four categories:
- Unavoidable adverse events: these constitute the majority and are generally due to catastrophic clinical conditions, inadequate time/opportunity to intervene and untreatable or nonresponding clinical condition despite optimal care. These are slowly decreasing in numerical terms with advances in the medical knowledge
- Negligent management (eg injurious procedure due to inappropriate skill and systemic issues): fortunately, this mostly remains a rare cause. Often it is a one-off error of judgement. Recurrent issues with particular health workers need retraining, supervision and rarely disciplinary measures. Some systemic issues and substantially inadequate provision of facilities by the hospital management may fall in this category but often this is not very clear-cut as there is always a limitation on resources
- Human factors or errors (suboptimal care): this seems a significant category and may stubbornly remain proportionately unchanged, even when the overall quality of care is improving. A recent study reported that 75% of the perinatal deaths in the midwife-led maternity care were preventable with appropriate care standards (Rowe et al, 2020). Such a rhetoric is common but needs a lot more qualification. It is to be expected that in retrospect (with the benefit of hindsight), with presumption of all the necessary resources, flawless science and banishing of the human factors (not feasible), 75% of all adverse events may always appear to be ‘preventable’. But it does not follow that they can always be prevented given the above constraints. The important point these studies omit to mention is the continuing overall improvement in outcomes which is a slow process and difficult pinpoint and is not newsworthy (publication bias). There is also the principle of ‘satisficing’ which contributed to a Nobel prize in economics (Simon, 1956; Wikipedia, 2021). This principle is particularly important in the healthcare provision to enable the best utilisation of the available resources (Coiera, 2011). Maximal goal is generally unachievable and there are disadvantages of excessive medicalisation which most women probably do not want. Even learning from mistakes may not always offer fail-proof or quick-fix solutions.
- Limitations of science or rarely misguided science: medicine is a soft science and far from flawless. Monitoring of the fetal health during antenatal period and labour is markedly constrained by the lack of direct access to the fetus. The intrapartum fetal heart rate (FHR) monitoring has been compared to checking just the pulse of a patient in an intensive care unit. Nevertheless, the correction of flawed science where possible seems crucial as discussed next.
Adverse events from unscientific intrapartum fetal monitoring and the way forward
The most frequent and serious criticism in independent inquiries and HSIB investigations concerns ‘mistakes’ in cardiotocography (CTG) and intermittent auscultation (IA) of FHR. Coroners have sometimes censured HSIB or rebuked health workers as narrowly missing criminal culpability (Dyer, 2020). The concerning frank or forthright tone (generally used for grassroots birth attendants) of this article reluctantly stems from repeated failure of subtle language. There is a common impression that if only the guidelines and science were followed, all ‘preventable’ perinatal deaths would be avoided. But how strong is the science and how scientific can the guidelines be? Probably the only high quality study of CTG training and practice shown to improve perinatal outcomes dates to 1998–2003 (Draycott et al, 2006) and taught the contemporary truthful pattern recognition of FHR decelerations as majority early and minority variable.
Science has been aptly defined as an unrelenting pursuit of truth. Notwithstanding, an ‘untruthful’ CTG pattern recognition of centrally important FHR decelerations has been adopted by the National Institute of Clinical Care and Excellence Committee ([NICE], 2007) led by Professor David James. This was borrowed from external mistaken system, replacing the previous scientific British practice disregarding repeated protests by British obstetricians at that time (Sholapurkar, 2019). The birth attendants may pay the price for this flawed pattern recognition which detected only about 30% of acidaemic babies under practice conditions and only 46% even on retrospective expert review (Clark et al, 2017) in contrast to the previous British practice in the study by Draycott et al (2006). This is no surprise given that the current pattern recognition is based on a false premise that head-compression decelerations do not occur; and the main cause of fetal hypoxaemia in labour is cord-compression, hence majority of FHR decelerations are variable (Frasch et al, 2009). Notwithstanding, it is known since the 1960s that 75% of cases of fetal hypoxaemia in labour are due to drop in uteroplacental perfusion during uterine contractions (Sholapurkar, 2013; Turner et al, 2020).
Systematised misinformation (akin to Orwellian reality control, double-think and language-control manipulating thought) cannot be argued to be acceptable, let alone useful. Even the so-called ‘physiological CTG interpretation’ practiced in many leading NHS hospitals has been shown to be ‘anti-physiological’ and unsafe (Sholapurkar, 2019). Birth attendants' intuition may have prevented much potential harm. Furthermore, prestigious journals should amend/retract ‘pathological science’ by prominent obstetricians/physiologists with a chain of untruths viz chemoreflex is an indefatigable guardian of fetal wellbeing, fetal hypoxaemia per se does not matter because all decelerations are due to hypoxaemia, timing of decelerations is irrelevant and await fetal decompensation (Lear et al, 2018). In contrast, birth attendants are (rightly) seriously criticised if they had ignored persistent hypoxaemic decelerations during investigation of adverse outcomes.
Humans inter pret complex patter ns by conceptualisation and meaningful chunking. When a false pattern recognition is enforced, it conflicts with natural learning and outcome-feedback. There are direct and indirect undesirable consequences. A heavy bag of wrong tools is burdensome and can dent the job. The grassroots clinicians have no choice but to constantly improvise increasing the element of intuition or art. Midwives have described a ‘reverse CTG interpretation’ where often they feel it necessary to decide the clinical action first and then customise the CTG interpretation to suit it (Sholapurkar, 2019). This involves the need to frequently ‘downstage’ the CTG, with consequent false-alarm fatigue and errors thereof. It is true that some mistakes in the CTG interpretation may in retrospect appear very ‘gross’ (which of course need to be addressed); but when the clinicians carry the excessive unwarranted cognitive burden of bad science, their intuition as well as attention can fail.
A recurrent theme from investigations has been the CTG misinterpretations by midwives and obstetricians of all grades (tip of the iceberg of all mistakes). But this is simply an observation and not a lesson. A lesson could be obtained if a critical analysis is attempted as to ‘why’. Rigorous and repeated training of fundamentally false science is problematic because with the enforcement of the current ‘pre-Copernican’ pattern recognition, most expert groups have repeatedly given incorrect teaching examples of FHR decelerations, even the serious pathological ones (Sholapurkar, 2013), a very common problem for more than a decade. Majority of experts have been completely silent on this issue, sidestepping acceptance of mistakes impacting on patients and birth attendants. Is there selective immunity and anonymity to admitting errors?
Similarly, there have been critical deficiencies in the guidelines of IA forcing errors by midwives (Sholapurkar, 2020). In addition, arduous unproven frequency (at the expense of quality) of IA in the first and second stage is demanded, a recipe for non-compliance. Overzealous demands/standards reflect a temptation to create a mirage of ‘certainty’ when it may not exist in clinical science. Fortunately, improvements are occurring with more to come (Sholapurkar, 2020). All these factors need consideration when birth attendants make mistakes in CTG or IA interpretation. The ESMiE enquiry mainly faults grassroots birth attendants (Rowe et al, 2020) without critiquing science or guidelines. Recently, the Care Quality Commission ([CQC], 2020) rightly and promptly investigated a university hospital following a cluster of intrapartum asphyxia cases requiring therapeutic cooling. It mandated a raft of multisystem improvements but notably none questioning possible flaws in the practice of intrapartum fetal monitoring, the main problem being investigated. The CQC and HSIB do seem to have the expertise to examine any flawed science and raise concerns to the guideline makers.
The RCOG (2015) Each Baby Counts initiative had aimed to halve the perinatal adverse events by 2020 but, instead, these have gone up between 2015–2019. All such enquiries, although important, comprise retrospective unblinded, uncontrolled methodology and hence apparent conclusions need to be weighed carefully (eg exclusively heavy emphasis on risk categorisation). The national Maternity Transformation Programme (2016) has ambition to halve the perinatal mortality, maternal mortality and brain injury by 2025. For more than a decade, intensive resources have gone into CTG training, mandatory yearly team retraining, rigorous detailed clinical protocols/documentations, and organisation/culture changes with disappointing outcome improvements. In this regard, adopting truthful science will pay most dividend without much cost.
Can silence be complicity or collusion? The subject of ‘intrapartum fetal monitoring’ is very complex and controversial and hence most developed countries like the US, Canada and Australia-New Zealand have separate guidelines for it (unlike the UK). This has the advantage that most members of the guideline committee have special interest and expertise in the field of intrapartum fetal monitoring. With a serious intent to improve patient care, the NICE should also issue a separate dedicated guideline of ‘intrapartum fetal monitoring’ rather than subsuming it in an all-encompassing ‘care in labour’ guideline, thus leaving it to a largely generalist panel as in 2007 and 2014. Truthful pattern recognition doesn't promise a panacea but it would be science. Progress can then occur rather than stagnation or regression. It should substantially reduce errors in CTG interpretation directly and indirectly.
The paradox of progress and preventable adverse events
With the welcome increasing safety culture and awareness, there are more protocols/guidelines, more monitoring, investigations, interventions as well as better record keeping and communication (enabled by digitisation). But this increasing complexity and digitisation come with their own brand of new errors and omissions which were not recognised as such previously. Thus, paradoxically as the quality and safety of healthcare improves, the same or more proportion of adverse events may still be deemed as preventable. Hence, a balanced communication about healthcare adverse events to the patients and public at large becomes vitally important. Unfortunately, in the media, a disparaging story sells much better (negativity bias) than a balanced nuanced one.
Public expectations, perception and criminal culpability
Public expectations rise over time with improving healthcare and the failures become more unacceptable and are often thought as entirely unavoidable. Politicians, managers, civil servants and medical leaders make high promises about the quality of care during a pitch for attention. They do not deliver the direct care. The understandable need to project a rigorous self-image impels the politicians, experts and investigators to pledge that ‘avoidable’ perinatal deaths will not be allowed to occur in future. It would be true that some baby deaths were ‘avoidable’ (Dyer, 2020) and the grief is undeniable. But the déjà vu follows. Honesty should include realism.
Recent episodes of public opprobrium and inquiries have been useful in shaking the system bringing about major changes and better provision of resources. However, it seems a pity if ‘naming and shaming’ on each serious event is deemed a necessary or desirable continuing way forward. Importantly, these inquiries have an opportunity (responsibility) to educate public about the nature of adverse events, the limitations of healthcare and importantly question any bad science. The openness and public dialectic about the adverse events should be accompanied by an honest message that medical errors cannot be completely eradicated but certainly minimised. The public will welcome such transparency and honesty. A ‘blame culture’ must be avoided if healthcare workers are expected to report errors openly. The unrealistic rhetoric of ‘perfect outcomes’ with clinician faulting should be resisted as it serves only short-term interests (the tragedy of belief commons). Health workers can feel vulnerable, helpless, resigned and, worse, sceptical or cynical. Cauldwel et al (2017) have opined about its damaging effects on staff morale, burnout and even recruitment, which is not in the long-term interest of the service or the patients. The talk of criminal culpability seems unhelpful if not a distraction. This requires health workers or (often hapless) managers to have shown a criminal disregard for patient safety or an intent to harm. Isolated gross human errors even when associated with very tragic outcomes need not imply criminal culpability.
The interim Ockenden Report (2020) into the Shropshire maternity services identifies seven domains of important essential actions for all maternity services. Many of these have already been implemented across the NHS, including the very recently created local maternity systems and maternity voice partnerships. It acknowledges that the Shropshire hospitals underwent external review and scrutiny by the CQC in 2015, 2018 and 2020, and by the RCOG in 2017; and have shown improvement over that period, although a few shortcomings were seen, not uniquely. There has been a criticism that the alphabet soup of NHS organisations – the inspectors, the regulators, the commissioners – have in past failed to protect the families (BBC, 2020). However, the regulators and inspectors should be patient, pragmatic and cautious not to impose a further plethora of additional standards. A wealth of standards leads to a poverty of their unfailing implementation (Coiera, 2011), apart from being highly resource intensive and at times counterproductive.
Conclusions
The grieving families following adverse events need honest explanation, continued support and a reassurance that lessons are being learnt. However, going forward, the current rhetoric that the ‘avoidable adverse outcomes or perinatal deaths will not be allowed to occur’ is neither realistic nor helpful. Pragmatism and perseverance seem important together with relevant number of clinical standards and investigating/regulating agencies. The interim Ockenden report (2020) comprehensively addresses cultural and organisational issues but correction of known flaws in the science of CTG interpretation and guidelines are urgently required to achieve the desired reduction in intrapartum hypoxic adverse events.
Key Points
- Avoidable adverse outcome in the maternity care results from human factors, negligence and limitations of science
- The safety culture, investigations and continuous improvements are paramount
- Reform of any unscientific intrapartum fetal monitoring seems crucial for patient safety
- Honest communication and support to the women and families, with grounded expectation from the healthcare services should be encouraged
- Improvements should occur in an organic way by continuous progress in science and clinical practice. The rising trend of critical repercussions from the public and media is worrying and needs thoughtful solutions