Single versus pairwise interpretation of cardiotochography, a comparative study from six Swedish delivery units

• The proportion of correctly individually classified CTG's was 75% and the proportion of correctly pairwise classified CTG's was 80% (difference 5%, p = 0.12).
• The baseline was high, 75% (95% CI: 72%–79%) individually correct classifications. This finding indicates a high standard of agreement in CTG classification in our study population, which was large and unselected, including both large and small delivery units.
• Differences in proportion of correctly classified CTG recordings between the delivery units were observed, indicating a potential area of improvement in CTG interpretation.

Objective:The aim of the study was to evaluate whether interpreting CTG pairwise brings about a higher level of correctly classified CTG recordings in a non-selected population of midwives and physicians.Study design:A comparative study.Setting:Five delivery units in Stockholm and one delivery unit in Uppsala, with 1589, 3740, 3908, 4539, 6438, and 7331 deliveries in 2011, respectively.Subjects:536 midwives and physicians classified one randomly selected CTG recording individually followed by a pairwise classification. The pairs consisted of two midwives (119 pairs) or one midwife and one physician (149 pairs), a total of 268 pairs.Main outcome measure:The proportion of individually correctly classified CTG recordings versus the proportion of pairwise correctly classified CTG recordings.Results:The proportion of individually correctly classified CTG's was 75% and the proportion of pairwise correctly classified CTG's was 80% (difference 5%, p = 0.12).Conclusions:There was no statistically significant difference when CTG's were classified pairwise compared to individual classifications. The proportion of individually correctly classified CTG's was high (75%). There were differences in the proportion of correctly classified CTG recordings between the delivery units, indicating potential areas of improvement.