Page 195 - 16Neonatal Jaundice_compressed

Neonatal jaundice

effective option. However, not all midwives do postnatal checks. It may be more useful to
consider the number of postnatal checks undertaken per day.
If we assume that each birth has, on average, three postnatal visits then this amounts to:

(690 000 × 3) ÷ 365 = 5670 postnatal visits per day
Community midwives would typically do 6–10 postnatal visits per day, which suggests that the
postnatal workload is managed by approximately 1000 midwives on any given day, which
might suggest that the service could actually be delivered with fewer than 9200 meters.
In interpreting this analysis, there are a number of caveats to be considered in addition to the
most important ones already highlighted concerning the lack of evidence. The analysis assumes
that 25% of infants will require a confirmatory TSB before consideration of phototherapy. If this
estimate were higher, then the total cost of the TCB strategy would be higher and the cost
threshold at which TSB would be the preferred option would consequently be lower. The
analysis also assumes that a move to more intensive testing does not lead to increased
phototherapy. This might seem a counter-intuitive assumption as the efficacy of more intensive
testing is ultimately predicated on not missing cases that could benefit from treatment.
However, intervention rates may also be influenced by recommendations on thresholds for
commencing treatment, where current practice varies. In some settings this might lead to lower,
but more targeted, intervention than currently occurs.
The analysis also assumes that the varous test strategies will not differ in terms of the amount of
testing undertaken and the number of follow-up home visits undertaken. Of course, it is possible
that the convenience of TCB could lead to additional ‘downstream’ costs not considered here.

An important assumption in this analysis is that phototherapy rates would not change if a more
intensive testing strategy were adopted. This is a strong assumption in the model since we do
not know how many more cases of hyperbilirubinaemia would be correctly identified by a
change in testing strategy.

C.9 Conclusion

Based on the published limitations of visual examination, the GDG strongly believes that a
more intensive testing strategy is required in order to improve outcomes in neonatal jaundice.
This will require more resources but, if this reduces the incidence of kernicterus by sufficient
numbers, it would be cost-effective to implement in the NHS. While the analysis presented here
is unable to demonstrate that this would be cost-effective, it does suggest that the actual number
of kernicterus cases needed for more intensive testing to be cost-effective is relatively small, for
example 0.49 cases per year if the TCB strategy could be delivered with 1000 meters up to 1.52
cases per year if 9200 meters are required. This is the cut-off above which the total cost of TSB
strategy is cheaper than TCB. It is important to remember that these values are based on strong
assumptions (for the lifetime QALYs lost through kernicterus and the lifetime cost of kernicterus)
that are supported by the GDG but are not based on externally verifiable evidence. The number
of cases of kernicterus that could be prevented is the critical unknown. However, reports from
the USA 240 have shown a reduction of four cases per 100 000 births after the mid 1990s.
Determining which intensive testing strategy is cost-effective depends crucially on the number
of meters that would have to be purchased in order to deliver TCB. The number of community
midwives involved in home visits on any one day is far smaller than the total number working
in the NHS at any one time. Therefore it seems plausible that the TCB strategy could be
delivered with a number of transcutaneous meters that is sufficiently low to meet the threshold
for cost-effectiveness. However, service delivery is not within the remit of this NICE guideline
and local commissioners may want to opt for the strategy they believe can be delivered most
cost-effectively in their area.

164