Table 68, GRADE profile for predictive accuracy of computerised cardiotocograph interpretation to identify adverse outcomes - Addendum to intrapartum care: care for healthy women and babies

Warning: The NCBI web site requires JavaScript to function. more...

Table 68GRADE profile for predictive accuracy of computerised cardiotocograph interpretation to identify adverse outcomes

Quality assessment						Definition of outcome	Total number of CTGs	Measure of diagnostic accuracy (95% CI)				Quality
Number of studies	Design	Risk of bias	Inconsistency	Indirectness	Imprecision	Definition of outcome	Total number of CTGs	Sensitivity	Specificity	Positive likelihood ratio	Negative likelihood ratio	Quality
CTG interpretation identified as abnormal^a by a computer software program
1 (Chung 1995)	Retrospective cohort	Serious¹	No serious inconsistency	Serious²	Very serious³	pH < 7.15	73	87.50 (46.7 to 99.3)^b	75.40 (62.9 to 84.9)^b	3.55 (2.16 to 5.86)^b	0.17 (0.03 to 1.05)^b	Very low
CTG interpretation of an outcome as abnormal^c by a computer software program
1 (Nielsen 1988)	Retrospective cohort	Serious¹	No serious inconsistency	Serious²	Very serious⁴	1-minute Apgar score below 7 or acidosis (umbilical arterial pH < 7.15 or base excess below −10 meq/l), or primary resuscitation needed	50	68.8 (41.5 to 87.9)^b	94.1 (78.9 to 99.0)^b	11.7 (2.9 to 46.7)^b	0.33 (0.16 to 0.69)^b	Very low

CAS Cardiotocographic Assessment System; CI confidence interval; CTG cardiotocograph; FHR fetal heart rate

a

An abnormal trace was defined by one or more of the following criteria

tachycardia (fetal heart rate > 160 bpm) for more than 30 minutes during labour
bradycardia (fetal heart rate < 110 bpm) for more than 30 minutes during labour
low variation (standard deviation of the fetal heart rate of ≤ 3 bpm) for more than 60 minutes during labour
more than five late decelerations (minima of the FHR occurring 20–60 seconds after the maxima of the contraction) during labour
more than 10 variable decelerations (minima of the FHR occurring more than 20 seconds prior to, or 60 seconds after, the maxima of the contraction) during labour

b

Calculated by the 2017 NGA technical team

c

A computer system (CA) calculates the probability of the CTG belonging to a compromised infant by calculating a discriminant function, and a CTG is considered pathological if the probability is above 0.5. The computer system’s calculation of the probability of a compromised infant is for each CTG based on the experience from the other 49 CTGs, thus excluding the possibility of “self-recognition”

1

Selection of cases for assessment not well described and it is unclear whether a consecutive or random sampling approach was taken

2

The reference standard used was different to that specified in the guideline review protocol (arterial cord pH <7.05)

3

CI for the negative likelihood ratio crosses two boundaries (from very useful (< 0.1) to not very useful (> 0.5)

4

CI for the positive likelihood ratio crosses two boundaries (from very useful (> 10) to not very useful (< 5))

From: Appendix I, GRADE tables

Addendum to intrapartum care: care for healthy women and babies.

Clinical Guideline, No. 190.1.

National Guideline Alliance (UK).

London: National Institute for Health and Care Excellence (NICE); 2017 Feb.

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.