Chapter 14:         Management of the Fetus with CPC –

 Comparison of the Ethical and Economic Implications of Different Protocols 

 

The most frequently asked question when a CPC is found is “should an amniocentesis be performed?”  Many papers of the 1990’s suggest a protocol of amniocentesis for isolated CPC (Perpignano150, Porto154, Kennedy110, Burrows34, Nava125, Kupfermine116 and Walkinshaw193).  Derivation of risk modification by the Bayes Theorem has been used in the previous sections of this thesis and this was sufficient to prove that in an otherwise low-risk patient, the risk of amniocentesis is higher than the age-related risk of T18 in cases of isolated CPC, regardless of the characteristics of the cyst.  However in this section, the information collated within this study is used to answer this question in another fashion, one relating to both the ethics of fetal loss and to economic considerations.  This approach might have a stronger impact as it more graphically represents the effects of a clinical decision.  

 

14.1         Three Suggested Protocols

There have been essentially three protocols advocated when a CPC has been discovered, which represent real clinical options.   The three protocols are:

 

Protocol  1.        Do nothing.

Protocol  2.        Amniocentesis whenever CPC is found.

Protocol  3.        Amniocentesis only when CPC is seen with other abnormalities.

14.2         Outcomes – Ethical and Financial

The success of prenatal screening protocols is usually judged in terms of the detection of aneuploid fetuses against a false positive rate.  From this comes the balance of risks discussed earlier.  For this discussion however, this constitutes only part of the calculation.  We can ask other questions.

 

The most important is “How will these protocols affect the rate of live births of T18?”  Given that the natural attrition rate of T18 is so high, how much does termination of prenatally detected T18 affect the actual birth rate?  A reduction in live births would represent a positive outcome in that concern for treatment of perinatal and neonatal distress and congenital malformations would be obviated. 

 

But how different are the results of the protocols on the live-birth numbers, and what is the number of otherwise chromosomally normal, which is to say healthy, wanted fetuses that will be lost due to the hazards of amniocentesis?  Will these numbers balance the risks in favour of or against amniocentesis in either of the Protocols which employ it?    

 

What we are asking really can be framed in terms of an ethical dilemma. We will take an essentially Utilitarian definition of ethics for this analysis.  Briefly then the question may be posed this way:

 

“What benefit or good is achieved, balanced against what harm is done?”

 

 

A ratio will be used to analyse the “negative” fetal loss against the “positive” outcomes described above.  This will be expressed in the following form:

 

Fetal Outcome (FO) : Fetal Loss (FL). 

 

 

For the purposes of this discussion, a FO:FL ratio of  around 1 would be considered ethically dubious – one normal fetus is lost for every abnormal live-birth prevented - while a ratio well below 1 would considered unethical.  FO:FL ratios well above 1 would be considered the most ethical as they indicate the minimization harm relative to the maximization of good.  

 

Also, we should ask what the financial cost is to the country’s health budget.  Real money is spent on these procedures, money that has generated from the public’s income tax.  Is it being spent wisely?  Calculations are made on the money spent in each protocol to detect each chromosomal abnormality and to prevent each aneuploid birth.  How do the protocols compare on a “value for money” basis?  A ratio of the protocols’ costs will be generated and discussed. 

 

14.3         Information Established in This Research

The argument which follows is based on the information garnished from the literature and discussed earlier in this thesis.  This information is summarised below:

14.4         200 Cases of T18

It is possible to imagine a hypothetical group of 200 16-week (the optimal age for amniocentesis) fetuses with T18. Given that the incidence of T18 is 1/2311, this group would come from a population of 462,200 women. In this group there 462,000 women who do not have a fetus with T18, but of course they might be affected by other malformations or other chromosomal problems. In the following argument, these 200 hypothetical fetuses are subjected to the three protocols described earlier. The calculations are tabulated in Tables 24 - 26 and graphed in Figs ...

14.5         Protocol 1 – Do Nothing

Those of the opinion that CPC are a normal anatomical variant in most cases might advocate this protocol. They might hold that T18 has a very high attrition rate throughout pregnancy, and as it is practically a lethal disease, nothing further need be done until a clinical issue arises. This does not equate to the protocol of those who do not report the presence of CPC, as it ignores the complicating presence of other abnormalities which might force the issue of a need for amniocentesis. This protocol is in fact the protocol that which was necessarily in effect in the era before ultrasound became available at all.

 

Due to T18’s high spontaneous pregnancy loss rate due mainly to intra-uterine fetal death, there would be only 60 live births from the T18 group. As T18 usually presents with many complicating abnormalities and including failure to thrive, only 50% or 30 babies would be alive after 10 days. Due to further implications of the abnormalities and failure to thrive, we would expect the survival of only 3 to 6 one-year-old T18 affected infants. Of course there would be no iatrogenic fetal loss unless amniocentesis was performed for another reason such as non-CPC fetal abnormality – though in this hypothetical group for statistical reasons we are assuming none were performed.

 

14.6         Protocol 2 – Amniocentesis for all Fetuses with CPC (including Isolated CPC)

This protocol is advocated by those who believe that CPC are so strongly associated with aneuploidy, either T18 or T21, that something must be done irrespective of the presence of other ultrasound detected abnormalities. As many fetal abnormalities are very subtle in these aneuploidies (such as small VSDs), some believe that a normal screening ultrasound cannot rule out the presence of abnormalities which would further indicate intervention.

 

They might also argue that T18 fetuses who reach normal delivery age usually present in distress. They are almost always growth-retarded and present with life-threatening abnormalities such as severe heart defects, diaphragmatic hernia and other issues. They may have abnormal CTGs. An emergency caesarian operation however is not advocated for fetuses with chromosomal abnormalities due to the unnecessary risk to the mother, the associated expenses incurred and the generally poor outcome of the fetus. So it would be preferable to know beforehand if a fetus presenting in distress at labour is chromosomally abnormal or not. Therefore, it might be argued, it would be best to make sure.  

 

If a mid-trimester screening ultrasound were performed at say 16-18wks on this population there would be, at a 1% incidence, 4,622 fetuses shown to have a CPC. As 21% of fetuses with T18 have CPC, there would be 42 cases of T18 included in that group. So this is the maximum number of T18 fetuses that could be caught in any CPC-based protocol net.

 

If all these fetuses with CPC were given an amniocentesis, irrespective of the presence of any other abnormality, those 42 with T18 would be identified and the pregnancies would be terminated (presumably). This would in turn reduce the non-CPC and therefore undetected T18 group to 158. Of those, 110 would die before birth. Of the live-born 48 fetuses with T18, 24 would die within 10 days of delivery, and only 2 to 5 would survive to one year. Thus the live-births of T18 babies would be reduced by a total of 12.

 

However, of the 4,580 Non-T18 fetuses also subjected to amniocentesis, approximately 23 would be lost due, assuming a best-case scenario (in the hands of an expert) loss rate of only 1/200, or 0.5%, associated with the interventional procedure.  (See Table 24)

 

 

14.7         Protocol 3 – Amniocentesis for Those with CPC and Other U/S Detected Abnormality(ies)

Those who believe that well-performed ultrasound is sufficiently accurate to detect other soft signs of aneuploidy in the majority of T18 fetuses, and those who wish to avoid unnecessary amniocenteses would advocate this protocol. 

 

Structural malformations, including those other than CPC, would be typically present within the entire group of 462,200 at a rate of about 3.5%; which means that 161 of the 4,622 fetuses identified in Protocol 2 with CPC would have other abnormalities.  A good quality ultrasound should have a sensitivity of 75% for detecting other abnormalities and thus detect 121 fetuses with multiple abnormalities of which CPC is one.  As 88% of the 42 fetuses in the T18 group with CPC would have other abnormalities detectable by ultrasound, that means 37 T18 fetuses would be found in this protocol. 

 

If these pregnancies with CPC and other malformations were given amniocentesis, and terminations performed for T18, the T18 pool would be reduced to 163, and again given the high spontaneous loss rate of T18 only 49 would survive until birth.  25 of these would die in the first 10 days, and only 2 - 5 would still be alive after one year.  The number of live born T18 babies would be reduced by a total of 11.

 

Of the 84 non-T18 fetuses with CPC and other malformation, the loss due amniocentesis would be between zero and one (actually 0.6) fetus.  This fetus would by definition of this Protocol have multiple anomalies. (See Table 24.)

 

We will assess this further with the FO:FL ratio later in this document.  

 

14.8         Which Protocol is More Effective?

The outcomes of the three protocols for T18 detection, reduction in live births, survival after 10 days, survival after 1 year, and iatrogenic fetal loss, from the discussion above in 14.5, 14.6 and 14.7 are tabulated below.

 

Protocol 1

Protocol 2

(CPC)

Protocol 3

(CPC+ Abns)

Total Population =  462,200

(@ 1%)Amnios = 4,622

(@3.5%)Amnios = 121

T18 Detected at 16 wks

 

@ 21% of 200 >>

42

@ 88% of 42 >>

37

T18 Not Detected (negative test)

200

158

(4580)

163

(84)

T18 Live Born

60

48

 

49

 

T18 Live After 10 days

30

24

 

25

 

T18 Live After 1 year

3 - 6

2 - 5

 

2 - 5

 

Reduction in T18 Live Births

 

12

 

11

Iatrogenic Loss of Normal Fetuses @ 0.5%

 

23

 

< 1 (0.6)

 

Table 24: Hypothetical group of 200 T18 fetuses and outcome related to amniocentesis protocol.

 


 

Fig:  Effect of Amniocentesis.  Numbers show T18 fetuses detected (termination is assumed) after a mid-trimester anatomical survey and amniocentesis, the reduction in the number of live-born T18 fetuses, and the iatrogenic loss of normal fetuses.  Number in brackets is the number of amniocenteses performed.  (No Ix = Protocol 1,  Multiple Ax = Protocol 3, Amnio All = Protocol 2.)  

 


 

Fig   Fate of Undetected T18 Fetuses or each Protocol. Numbers are of surviving infants, given published mortality figures for this condition. (No Ix = Protocol 1,  Multiple Ax = Protocol 3, Amnio All = Protocol 2.)  

 

These figures show an improvement in the overall picture of T18 management in both protocols, with difficult births and neonatal dilemmas reduced by 20% (48/60) and 18% (49/60) respectively. 

 

The difference of 1-2% (1 live-birth) between Protocol 2 and Protocol 3 appears negligible.

 

After 1 year, the difference in live T18 babies is essentially identical. This reveals that the more aggressive Protocol 2 will have little to no effect on the support community for such babies.

 

 

14.8.1                          The Effects of Serendipitous Detection of Aneuploidy

 

 

 

 

Protocol 2

(CPC)

Protocol 3
(CPC+ Abns)

Non T18 fetuses

 

 

4,580

 

84

Other Chromosomal Abnormalities

@ 0.43%

20

@ 10%

8

Spontaneous Aneuploid Loss

@  50%

10

@ 50%

4

All Chromosomal Abnormalities (rate)

 

(42+20)

62

or 1.3%

 

(37+8)

45

or 37%

Reduction in Aneuploid Live-births

 

(12+10)

22

 

(11+4)

15

Table 25: Effects of serendipitous detection of chromosomal abnormalities.

 

In Protocol 2 also hidden within those 4,580 fetuses which do not have T18, are approximately 20 (@0.43%) fetuses with other types of chromosomal abnormalities.  These would be detected serendipitously by the amniocentesis protocol as we are effectively ignoring other abnormalities in this protocol.  As at the moment we are unaware of any influence of CPC on the relative risk, either positive or negative, of other abnormalities for sake we will consider it to be none for the sake of minimizing statistical complexity. 

 

From large studies we know that fully 33% of the aneuploidies that would be detected by a large number of amniocentesis will be of a type other than T18, and most of these of course will be T21.  This would explain the association of CPC with T21 that has been claimed by some authors who have advocated this protocol.  If the spontaneous loss rates were considered, and which are for T21, 31%, for T13, 71%, for Triploidy, 99%, for Turner’s, 52%, and other sex chromosome disorders about 3%, then perhaps 50% of those extra 20 fetuses would die anyway. Therefore while 62 chromosomally abnormal fetuses, or 1.3% of the all fetuses identified by Protocol 2 and given amniocentesis, the total reduction in the number of live-births of chromosomally abnormal fetuses (T18 plus all others) would only be 22. (Table 25)

 

 

Similarly in Protocol 3, serendipitous detection of other chromosomal abnormalities in the remaining 84 fetuses would yield another 8 (@10%) aneuploidies, of which half would die before birth.  In contrast to Protocol 45 (37%) of the fetuses identified by Protocol 3 would be aneuploid.  (Table 25) 

 

Only when unexpected detection of other aneuploidies is introduced into the equation, as shown in Table 25 are there noticeable differences between the protocols in the reduction of aneuploid births.

 

However, it is important to note that the rate of detection of aneuploid fetuses overall is much higher in Protocol 3 at 37%.   Protocol 2 only has a positive diction rate in 1.3% of cases.  This makes the former a much more "efficient" protocol for detecting aneuploidy, but is mere efficiency enough to make it more ethical?   

 

14.9         Which Protocol Can Be Ethically Justified?

 

Protocol 1

(Do Nothing)

Protocol 2

(CPC)

Protocol 3

(CPC+ Abns)

 

 

T18

All

T18

ALL

Fetal Outcome (Reduction in Livebirth)

0

12

22

11

15

Fetal Loss  (Iatrogenic Death)

0

23

23

0.6

0.6

FO:FL

1

0.52

0.95

18.3

25

 

Table 26: FO:FL Ratio

 

Protocol 1, or doing nothing, gives no benefit to the patient, nor does it actively harm as no fetal losses are incurred through amniocentesis.  However, in comparison with a successful protocol, this non-interventional stance may be ethically unsustainable, when there is the strong possibility of doing good and preventing harm using a minimally deleterious procedure for a large positive gain.  Such a gain may be found in the reduction in the number of unnecessary caesarians for fetal distress in aneuploid pregnancies and in the pain suffered, the risks undertaken, not to mention the financial costs.  Also it may be found in the reduction of the possibility of pain, suffering and extremely poor quality of life for the affected infant, and arguably there is the financial savings by reducing the cost of treatment and support for the affected infants.  It has a FO:FL ratio of 1, and is marked as yellow, indicating that it has a debatable ethical position.

 

In the days before ultrasound and amniocentesis this was the only option available, and therefore could not be unethical.  In present times, when sophisticated machines and procedures are available, it is probably unsustainable.

 

Protocol 2 allows for a reduction of 12 live-births of T18 fetuses by detecting those 42 positive cases which have CPC. This contrasts strongly with the loss by amniocentesis of 23 chromosomally normal fetuses.  FO:FL ratio in this instance is 0.52, which is certainly a very poor outcome by this definition and could not be considered the ethical choice.  The damage or harm done in this protocol is much higher then the benefits or the good done.

 

When one considers that Protocol 2 allows the detection of fetuses with other aneuploidies up to a total 62 and a reduction of live-births to 22, the situation is less clear-cut.  Considering ALL chromosomal abnormalities brings the FO:FL ratio to 0.95.  This almost the same result as Protocol 1, but still slightly less than 1.  Therefore while it also be considered to be an ethically arguable position, ho doe sit fare against Protocol 3.   

 

Protocol 3 does not detect as many aneuploidies in total as Protocol 2. However it is, as noted, a much more efficient protocol for the detection of aneuploidy and it also has an extremely low fetal iatrogenic loss rate.  For T18 Fetuses it prevents 11 live-births and causes 0.6 of a fetal losses due to amniocentesis.  That is to say, less than one, but not necessarily zero.  In real life this would mean a loss of EITHER one or zero.  

 

For the FO:FL ratio we will use 0.6.  This gives a ratio of 18.3.  This is much better than 1, being 18.3 times more “ethical” (if such a comment is sensible) than Protocol 1 and more than 35 times more ethical than Protocol 2 for the detection of T18.

 

When other aneuploidies are included the number of live-births is further reduced to 15.  This boosts the FO:FL ratio to 25.  This makes Protocol 3 even more ethical in this discussion’s definition.

 

Discussion: By using birth-rate analysis rather than the mere detection of aneuploidy, the real effect of these protocols can be seen.  A decision concerning the significance of this approach rests with those individuals performing amniocentesis.  The fetal loss rate with Protocol 2 is 38.3 times higher (23/0.6) than that for Protocol 3, with only minimal improvement in the various live-birth reduction outcomes. 

 

Performing 4,501 (38.2 times) extra amniocenteses and losing 23 fetuses is not offset by the detection of 5 extra cases of T18 in fetuses with ISOLATED CPC and the reduction of T18 live-births by one.  Should the figures of Nicolaides129, Ramsey157 and Tabor183 for amniocentesis loss rate be used, the fetal loss would be double what has been calculated here, and an even more substantial argument could be raised to criticise Protocol 2 on ethical grounds. 

 

14.10       Which Protocol is More Cost Effective?

What is the comparative cost to Medicare of these protocols?  Putting a dollar cost on The scheduled Health Insurance Commission Medicare 1996 fee for Amniocentesis is $46.60 and for Cytogenetics it is $327.60 – a total for the procedure (not including ultrasound control) of $374.20.  Using Protocol 2, where 4,622 amniocenteses would be indicated, a Medicare bill of $1,729,552 would be generated.  Using Protocol 3, which indicates 121 amniocenteses, the Medicare bill would be $45,278.  The cost of these protocols is analysed in Table 26.

 

 

Protocol 2

Protocol 3:

Ratio of P2/P3

Number of Amniocenteses

4,622

121

38.2:1

Total Cost @ $374.20 per Amnio

$ 1,729,552

$ 45,278

38.2:1

T18 Detected

42

37

1.13:1

Cost / T18 Detected

$41,179

$1,223

33.4:1

Cost / Reduction in each T18 Live-births

$144,129

$4,116

35:1

Extra Aneuploidies Detected

20

8

2.5:1

Cost / Extra Aneuploidy