prof alec welsh - unsw and royal hospital for women - interpreting intrapartum fetal monitoring

Interpreting Intrapartum Fetal Monitoring

(Medical versus Legal ‘Evidence’)

Department of Maternal-Fetal Medicine, Division of Women’s & Children’s Health

Professor Alec Welsh GAICD MBBS MSc PhD MRCOG(MFM) FRANZCOG DDU CMFM

Professor in Maternal-Fetal Medicine School of Women’s & Children’s Health University of New South Wales Randwick, Sydney

Professor Maternal-Fetal Medicine Royal Hospital for Women Randwick, Sydney

We are professional ‘believers’…..

So where do I stand on CTGs then?

Professor Alec Welsh [email protected]

Previous presented arguments regarding the poor scientific basis for our current use and interpretation of EFM


Is it possible that we can abandon our implicit faith in CTGs in order to try to gain sufficient evidence to support their use and interpretation? Faith: strong belief in doctrines, based on spiritual conviction rather than proof or evidence Evidence: the available body of facts or information indicating whether a belief or proposition is true or valid


• The Issue and why it is so highly relevant • Summary of the usefulness of CTGs • Does ‘proof’ differ between the legal and medical

professions? • Peer practice test versus scientific evidence • Why are we where we are? • What on earth can we do about it?

Talk Outline



The issue and its relevance

Negligence bills faced by NHS risen 7-fold in last 5 years 23,000 outstanding claims NHS Bill for negligence in pregnancy has reached £1Billion One basic error accounts for a quarter of all payouts


• UK: Currently 25% of the NHS budget is taken up by litigation

• 6% of the NHS budget is taken up by electronic fetal monitoring: birth asphyxia litigation

• So does that mean we should: • Focus more effort upon EFM to make it

work better? • Abandon EFM as it isn’t working?

Cost of malpractice litigation


10 years: 300 claims of CTG ‘misinterpretation’ Total value £466Mill 37 closed Of the 131 open, some remain open for the rest of the claimant’s life

NHS Report 2012


In the text of the responses, all but one claim allege that abnormal CTGs had not been acted on

Misinterpretation? Themes?


27 deaths 117 neurological problems: cerebral palsy, quadriplegia, hemiplegia and developmental problems

Outcomes Stage of Labour


Typical case reports contained within this document…



Evidence for use of CTGs

Physiology: the “chronic or acute insult”

Uterine hypoxia secondary to uterine activity

Fetal umbilical vein hypoxia

Central fetal hypoxia

Fetal redistribution

Fetal peripheral anaerobic metabolism

Accumulation of fetal metabolites and acidaemia

Fetal hypoxic / acidaemic damage

Fetal Response

Fetal Damage


An element of physiological understanding has been there since the 1960s: Though previously intrapartum events were thought to be causative of Cerebral Palsy


• Blair, Stanley 1987: clinically observed perinatal signs of birth asphyxia RR CP 2.84 (1.85-4.37); ~8% of spastic CP caused by intrapartum asphyxia

• CDC: ‘People used to think that CP was mainly caused by lack of oxygen during the birth process. Now, scientists think that this causes only a small number of CP cases.

• Risk Factors: Low birthweight; prematurity; multiple births; ART; infections; jaundice; maternal medical conditions (e.g. thyroid disease, seizures); birth complications; genetic factors

CP Causation: current understanding


These observations have lead to the hypothesis that increased survival of premature, neurologically impaired infants may have masked an actual reduction in cerebral palsy among term infants as a result of the use of electronic monitoring and the avoidance of intrapartum asphyxia. A review of the medical literature, as well as a demographic analysis of term and preterm birth rates in the United States, refutes this hypothesis on four grounds. First, cerebral palsy prevalence has been separately analyzed in term infants and shows no change over 30 years.

Am J Obstet Gynecol 2003;188:628-33

Temporal and demographic trends in cerebral palsy - Fact and fiction: Steven L. Clark, MD, Gary D.V. Hankins, MD


Cerebral Palsy in Australia: current

data


Overall prevalence of CP by state


Influence of gestational age


Influence of plurality


Systematic review: 51 articles (481,753 infants) Low arterial cord pH associated (OR; 95% CI): • Neonatal mortality 16.9

(9.7-29.5) • HIE 13.8 (6.6-28.9) • IVH or PVL 2.9 (2.1-4.1) • CP 2.3 (1.3-4.2) Associations strong, consistent and temporal

This is not to say that perinatal acidaemia does not have consequences…


2012: 10 year observational study Oxford, UK. 51,519 singleton, term, non-anomalous live neonates with validated umbilical cord arterial pH values

Though those consequences may not be so clear…


• Median arterial pH 7.22 (IQR 7.17-7.27) • Absolute risk of adverse neurological

outcome: • 0.16% for pH 7.29 – 7.30 • 0.36% for pH <7.1 • 2.95% for pH <7.0

• Increased risk below 7.1 but absolute risks very low and most affected babies have a higher pH.

• Intrapartum fetal surveillance….follows NICE Guidelines…will lead to both a failure to prevent most adverse neurological outcomes and a higher obstetric intervention rate.


Acidaemia matters, but we just can’t ‘pick it’ with CTGs

Maybe it’s just all about ‘interpretation’ of CTGs?

Numerous classification schemes exist for CTGs (and change periodically)


ACOG 2009


Positive predictive value of non-reassuring CTG to predict CP >2500g is 0.14%: this means that out of 1,000 fetuses with a nonreassuring CTG only 1 or 2 will develop CP. False positive rate >99% It is estimated that only 4% of encephalopathy can be attributed solely to intrapartum events.

ACOG Guideline comments

In the presence of placenta praevia: Number of previous CS: Risk of placenta accreta (%):

1 3 2 11 3 40 4 61 5 67


So if we conduct caesarean sections on these 1000 women how many placenta accretas will we deal with in the future?

SOGC 2007


839 Pages!

UK National Institute for Health and Care Excellence (NICE)


NICE


RANZCOG 2014


RANZCOG 2014

Medicine is governed by evidence - unless you are talking about EFM


• Hon et al (1968) described early and later decelerations to have a ‘uniform’ pattern of a specific time relationship to contraction cycle.

• However it appears that ‘uniform’ for RCOG / NICE guidelines has been interpreted to mean identical i.e. ‘uniform in duration and depth and shape’.

• NICE 2007 categorically states that early decelerations are ‘truly uniform’.

Interpretation of CTGs has also changed over time: and not necessarily for any reason


• RCOG • Early and late decelerations must

be uniform in both length and depth, but without mentioning gradual shape

• Gibb and Arulkumaran • Described ‘early decelerations’

as ‘bell’ shaped (gradual) with no mention of ‘uniform’

• K2MS • “Early decelerations are uniform

in shape, have rounded nadir (U shaped) and have symmetry with contractions”

Differing descriptions of decelerations


Association of Atypical Decelerations with Acidemia. Cahill et al. Obstetrics & Gynecology 2012 5 year retrospective cohort ≥37wks 30 mins EFM before delivery 2 formally trained research nurses; ACOG Guidelines; blinded Atypical features: shoulders, slow return, variability within deceleration Acidemia: umbilical arterial cord pH ≤7.1

What about ‘Atypical Decelerations’


5388 Women Incidence of atypical variable features: Shoulders (n=2914; 54.1%) Slow Return (n=2618; 48.6%) Minimal Deceleration Variability (n=430; 8.0%) No Deceleration Variablity (n=4; 0.07%) Association with acidaemia: Shoulders AOR 1.06 (0.63-1.81) Slow Return AOR 0.91 (0.54-1.53) Minimal DV AOR 0.82 (0.43-1.55) No DV AOR 0.65 (0.27-1.55) Level of Evidence IIA

Results


Variable decelerations and 8 subtypes evaluated to discriminate tracings between babies with normal umbilical arterial gases and those with metabolic acidaemia Tracings from last 4 hours for n=3320 babies with base deficits <8mmol/L n=316 with >12mmol/L. Computerised pattern recognition ROC curves and Area Under the Curves (AUCs) for each deceleration type.

Variable decelerations: do size and shape matter: Hamilton 2012


3 subtypes were discriminatory though none were good: • Prolonged duration AUC 0.61

P<0.0001 • Loss of internal variability AUC

0.5694 P<0.0001 • “Sixties” criteria AUC 0.5997

P<0.0001: 2 or more of: depth ≥60bpm; lowest value ≤60; duration ≥60s All other subtypes no better than chance

Hamilton 2012


• Only 16% of normal vaginal births will maintain normal fetal heart rate baselines with normal baseline variability and continue without decelerations throughout labour, thus remaining in the lowest risk level called Category 1 in the NICHD schema.

• Therefore as well as most normal vaginal births, the great majority of tracings from babies with metabolic acidaemia will be found in the middle levels.

Hamilton: observations in the atypical variable deceleration group


• Testing reproducibility of British experts’ illustrations of fetal heart rate decelerations by trained British Obstetricians and midwives • eFM, K2MS, Gibb, Arulkumaran

• Staff: • 38 Obstetric Consultants; 49 Registrars; 45 midwives

• 98-100% midwives; 80-100% registrars; 74-100% consultants categorised FHR decelerations differently from the 5 illustrations / interpretations

Sholapurkar 2013: Open Journal of Obstetrics and Gynaecology


• Consultants categorised FHR decelerations differently from the 5 experts illustrations / interpretations (p<0.00001)

• Early decelerations classified as atypical variable by 56% Consultants, 78% Registrars, 99% midwives: 85% classified as pathological

• Midwives and Registrars appeared to rigorously follow teaching of NICE guidelines.

• Consultants showed more flexibility and discretion based on previous experience and personal views

Sholapurkar 2013: Open Journal of Obstetrics and Gynaecology


The Cochrane Review: Held in high esteem by

many as the ‘highest form of

evidence’ available.


13 trials with >37,000 women; only 2 judged to be of high quality Compared with intermittent auscultation: • No significant improvement in overall perinatal death rate:

Risk Ratio 0.86 (0.59-1.23; n=33,513, 11 trials) • Halving of perinatal seizures: RR 0.50 (0.31-0.80, n=32,386, 2

trials) • Cerebral palsy rate: RR 1.75 (0.84-3.63, n=13,252, 2 trials) • Caesarean section rate: RR 1.63 (1.29-2.07, n=18,861, 11

trials) • Instrumental vaginal birth: RR 1.15 (1.01-1.33, n=18,615, 10

trials)

Cochrane: Alfirevic Z et al 2013


Continous cardiotocography during labour is associated with a reduction in neonatal seizures, but no significant differences in cerebral palsy, infant mortality rate or other standard measures of neonatal well-being. However, continuous cardiotocography was associated with an increase in caesarean sections and instrumental vaginal births.

Cochrane: Authors’ conclusions


So: not functioning as intended on a population level, and lacking in ability to detect or exclude perinatal harm on an individual level. (AW comment)


Proof…

Legal proof

• Edwards J, Auchard J. Four trials. 2004 • US Presidential Candidate. 4 major successful trials in his

career before retirement from law aged 45 with an estimated personal wealth of up to $US70M

• “First, I had to become an overnight expert in fetal monitor readings…” • Closing statements: “…(the fetus) did everything she knew….At five-

thirty, she said ‘I need out’….at six the cries got weaker..and the cries heard were the cries of Jennifer C. dying…but now she speaks to you….(the jury) through me.”

• Secured a $6.5MUS verdict against the hospital

Closin A United States Plaintiff Attorney (and expert in EFM)


AW 2016: First, I had to become an overnight expert in Civil Law….

Balance of probabilities: burden of proof in civil trials. In a civil trial, one party's case need only be more probable than the other. Much law is based upon individual cases, the presentation of perspectives in these cases and ? Oratory skills…


Case-based reasoning / Case-law

“Case-based reasoning is ethical decision making based on precedents set in specific cases, analogous to the role of case law in jurisprudence. An accumulated body of influential cases and their interpretation provide moral guidance. Case-based reasoning asserts the priority of practice over theory, rejects the primacy of principles, and recognises the emergence of principles from a process of generalisation from analysis of cases.” AW: lends itself to opinion and oration (akin to “n=1” medicine

Chapter 3: Ethics in Obstetrics & Gynaecology


Causation

The position in New South Wales at least, following Seltsam v McGuiness [2000] NSWCA 29 is that the Court must still be satisfied on a balance of probabilities that the risk of injury created by the lack of care is indeed the cause and the injury has happened because of the negligent medical care


Australian Legal Website (accessed 18/02/16)

As a lawyer we need to establish on the balance of probabilities, that is more than 50%, that the medical gynaecologist or obstetric specialist acted in a manner that was careless, out of line with medical peer group or neglectful causing damages to either or both the mother of the newly born child or the newborn.


So, are obstetricians to be ‘judged’ relative to medical peer group and not by true evidence? Is there any chance that a peer group could mislead?

Typical Medico-Legal Wording:

…Please advise whether the intrapartum fetal surveillance provided in relation to Patient xxx was reasonable and in accordance with practice in Australia accepted by a significant number of respected practitioners in the field as competent professional practice in the circumstances?


Maybe the two professions don’t think in the same way

Leigh: “ The lawyer will always look for reinforcement of what he knows already. We are the profession most inclined to confuse that which is with that which should be; because our authority is entirely man-made, as medicine was in the era governed by the authority of Galen for 1500 years. ”


Leigh:


“The enormous fiscal responsibility of a life that is destroyed at birth” “I am regularly told by experts that they cannot defend a failure to respond to a change in a trace, but they also know that many similar traces are followed by the delivery of a healthy baby and that the tolerance of some degree of uncertainty is, and is likely to remain, a constant feature of modern maternity practice.”


Proof…

Medical proof

Evolution of medical thinking: the past orators

• Traditional expert: sum total of experience • Danger: ‘n=1’ medicine or “I once had a case” • Nonevidentiary medicine:

• Individual clinical experience • Physiologic principles • Expert opinion • Understanding of professional values • Understanding of patient values

• “The Art of Medicine” - Oratory


Evolution of medical thinking: the present

Need to have an overall understanding of population medicine – hence ‘evidence based’ based medicine Translation of research findings into clinical practice Relegation of experience and ?wisdom Total culture of research and evidence-based practice, with associated terminology and statistics:



Terminology of proof / value of interventions

Terminology of proof / value of interventions

Sensitivity (true positive rate): proportion of positives that are correctly identified as such: completely sensitive means you identify all of them Specificity (true negative rate): proportion of negatives correctly identified as such: completely specific means you falsely label none

* Positive predictive value = number of true positives divided by number with test outcome positive ACOG: Positive predictive value of non-reassuring CTG to predict CP >2500g is 0.14%: Out of 1,000 non-reassuring CTGs only 1 or 2 will develop CP Negative predictive value – number of negative divided by number with test outcome negative


Levels of ‘Evidence’: e.g. NHMRC

Nb Expert Opinion is the lowest level of acceptable evidence, in the absence of research evidence… = RANZCOG level of evidence for its CTG recommendations


Receiver operating

characteristic (ROC) curve

This can be used to show how ‘good’ a test might be in the clinical situation A trade-off exists between setting ‘the bar’ such that you detect all true cases but include a large number of negative cases.


Sensitivity and specificity of intrapartum computerised FIGO criteria for CTG and fetal scalp pH during labour


Sensitivity and specificity of intrapartum computerised CTG: The ROC space

Detect all of them

Falsely identify none


Evolution of medical thinking: the future

• We are probably headed towards a more rational blend of: • “experience or thinking” medicine • “evidence-based” medicine • “individualised” medicine

• Numerous criticisms of the blanket application of EBM abound.


Evidence-based medicine versus Experience-based medicine

In the last three decades, evidence-based medicine (EBM) has become the gold standard for clinical practice. In fact, physicians who forgo evidence-based recommendations in favor of treatments supported by personal experience or undocumented recommendations make themselves more vulnerable to liability and subsequent indictment and may even appear arbitrary or unscientific. Nevertheless, EBM’s rise to prominence in clinical practice has stirred up some physician opposition, particularly from older health care professionals, who perhaps better recognize the growing divide in perceived value between the art of medicine and the science (a subtlety younger generations of physicians born into a system focused on EBM may not be able to appreciate as acutely).

American Medical Association Journal of Ethics Jan 2011, Vol 13: No 1: 26-30


• Kesselheim et al: Characteristics of physicians who frequently act as expert witnesses in neurologic birth injury litigation: Obstet Gynecol 2006: • 827 cases 1990-2005 • 71 witnesses participated in 738 cases (89%) =>

$2.9BillionUS • Frequent plaintiff witnesses:

• Higher median case rate than defendant • Older (57.2 vs 50.8yrs) • Less likely to be board certified (38% vs 95%) • Fewer academic publications (5.0 vs 53.5)

The (Professional) Expert Witness


In the absence of any clear medical evidence for the predictive value of CTGs for perinatal consequence…

What would our peers do? What are the “Colleges” telling us to do?


The “Worst” CTGs


College Description of CTG Nature of CTG Recommendations

ACOG

Category III (abnormal)

Absent variability AND any of: recurrent late decelerations, recurrent variable decelerations, bradycardia, sinusoidal pattern.

Prompt evaluation: providing oxygen, repositioning, stopping syntocinon, treating hypotension + others. If not normalised, deliver

RANZCOG FSEP

Features likely to be associated with significant fetal compromise and require immediate management which may include urgent delivery

Prolonged bradycardia; Absent baseline variability; sinusoidal pattern; complicated variables with reduced or absent baseline variability

Immediate management which may include urgent delivery (Level 1 evidence inappropriately ascribed)

NICE\RCOG

Abnormal

Tachy >180 bpm; Non-reassuring after 30 minutes; Late decelerations > 30mins; bradycardia > 3mins

“Inbetween” CTGs (most that we deal with)


College Description of CTG Nature of CTG Recommendations

ACOG

Category II (Indeterminate)

All FHR tracings that aren't entirely normal or Category III (70-80% of all CTGs)

"Category II tracings may represent an appreciable fraction of those encountered in clinical care." Evaluation and surveillance and possibly other tests to evaluate well-being

RANZCOG FSEP

Features that may be associated with significant fetal compromise and require further action

Baseline fetal tachy; reduced or reducing variability; rising baseline; complicated variasbles; late or prolonged decels

Identify and treat reversible causes; consider further fetal evaluation; escalate to a more senior practitioner

NICE\RCOG

Non reassuring

Reduced variability; tachycardia; late decelerations

UK NICE / RCOG


Canadian College Guidelines 2007


Our College is guiding us:

Nb Alfirevic Cochrane cited as the evidence for this statement: however the Cochrane review describes population benefits of CTG over intermittent ausculatation: it does not give recommendations on management when the FHR is abnormal.



“with our training program, we hope to have evidence to prove vastly improved consistency of nomenclature….”

• We’re in a mess: • There is little clear evidence to support the

routine use of CTG (population level) • The association between CTG tracings and

outcome is poor (individual level) • The dissociation between ‘expert opinion’

and evidence is vast and inexcusable • Both the legal and medical professions are to

blame

Summary 1


Electronic Fetal Heart Rate Monitoring: Nice Idea, good background physiology:

BUT sadly just not a good test…


Where is the common ground for

these two?

Legal Reasoning: Akin to the “Art” of

practice

Medical Reasoning:

Akin to the “Science” of practice

Summary 2

• In the absence of clear evidence of the validity of EFM why can we not just say that CTGs show the fetal heart rate and that there is insufficient evidence to indicate fetal state and neonatal outcome?

• Can we just accept that the most grossly pathological CTGs are relevant, but the vast majority cannot be used to show causation? And if not, why not?

• Why do we have to seek to perpetuate the myth that CTGs have validity?

• Whose gain is this for? Our own sense of wellbeing; Rich medicolegal experts; Rich lawyers; Professional Organisations’ standing

Where to? Do we have a way out?


Questions & Discussion

What’s wrong with CTGs? • Insufficient understanding of the (patho-)physiologic background • A too indirect signal of the fetal condition • A number of technical pitfalls • Differences in recording techniques • Primarily qualitative information (pattern recognition) • Lack of uniform classification systems • Confusion due to the many influences on the fetal heart rhythm • Substantial intra- and inter-observer variation regarding interpretation • Low validity, high incidence of false-positive findings • Primarily screening method, too often applied as a diagnostic • Leads to an increase in artificial deliveries • Lack of agreement on how, when, and whom to monitor • Contributes to medico-legal vulnerability