Midwives’ visual interpretation of intrapartum cardiotocographs: intra- and inter-observer agreement
Article first published online: 5 SEP 2005
Journal of Advanced Nursing
Volume 52, Issue 2, pages 133–141, October 2005
How to Cite
Devane, D. and Lalor, J. (2005), Midwives’ visual interpretation of intrapartum cardiotocographs: intra- and inter-observer agreement. Journal of Advanced Nursing, 52: 133–141. doi: 10.1111/j.1365-2648.2005.03575.x
- Issue published online: 5 SEP 2005
- Article first published online: 5 SEP 2005
- Accepted for publication 15 December 2004
- fetal monitoring;
Aim. This paper reports an examination of intra- and inter-observer agreement in midwives’ visual interpretation of intrapartum cardiotocographs (CTGs).
Background. The issue of intra- and inter-observer agreement in the interpretation of CTG interpretation has serious implications for the validity of electronic fetal heart rate monitoring and subsequent decisions on intrapartum management. However, no studies were found that assessed intra- and inter-observer agreement in midwives’ interpretations of CTG tracings.
Methods. Twenty-eight midwives independently interpreted three intrapartum CTG tracings on two separate occasions using a self-administered Cardiotocograph Interpretation Skills Test. Inter-rater agreement in interpretation was assessed by cross-tabulating the two sets of raw data obtained at time 1 and time 2 and computing Cohen's Kappa (κ). Intra-rater agreement was assessed by computing κ for each rater with the two sets of raw data (time 1 and time 2) obtained from each individual. The data were collected in 2000.
Results. Overall intra-rater agreement ranged from ‘fair to good’ (κ = 0·48) to ‘excellent’ (κ = 0·92). Raters’ classifications altered in 18% (n = 5) of cases for the normal tracing, in 29% (n = 8) for the suspicious tracing and in 11% (n = 3) for the pathological tracing. Inter-rater agreement was fair to good, with κ statistics ranging from 0·65 to 0·74, respectively. Agreement was highest in the classification of decelerations (κ = 0·79) and lowest in the assessment of baseline variability (κ = 0·50). Overall inter-rater agreement was highest in the suspicious tracing (κ = 0·77, excellent) and lowest in the normal tracing (κ = 0·54, fair to good).
Conclusion. Inter- and intra-observer variability are intrinsic characteristics of the interpretation of intrapartum CTGs. Levels of agreement revealed degrees of variation that expose room for improvement. Efforts are needed to reduce inter- and intra-observer variation in interpretation of intrapartum CTG tracings. In addition, research should focus on the development and evaluation of non-invasive, low observer variability methods of intrapartum assessment of fetal well-being. The subjectivity of CTG interpretation and inconsistencies in interpretation should also be considered in intrapartum management, clinical audit and in medico-legal settings.