Table 4

Examination of association according to Sir Austin Bradford Hill's nine causality criteria38

Criterion for causalityFindings and associated characteristics of current study
Strength of associationSignificant downward trend in stillbirth rates which had remained static for the past 20 years. Each of the high-uptake regions dropped to their lowest ever rates (table 2)
TemporalityReduced stillbirth rates followed implementation of accreditation training and protocol. E.g., Yorkshire and the Humber drop in 2012 followed training initiative in 2011
ConsistencyReduction in stillbirths shown in the three regions with high uptake (figure 3), and not in the regions with low uptake (figure 4). See also significant correlation in figure 1
PlausibilityConfidential case reviews have shown that unrecognised fetal growth restriction was the most common cause of stillbirth.9 Population based study has shown that IUGR is the strongest risk factor for stillbirth, and its antenatal recognition reduces risk10
Dose responseThe region with the most intensive training programme (West Midlands) had the steepest drop and downward trend in stillbirth rates (figure 3)
Experimental evidenceAntenatal recognition of intrauterine growth restriction results in earlier delivery10 and reduces stillbirth risk10 30 31
CoherenceImproved recognition allows the implementation of appropriate investigations15 29 and timely delivery10
SpecificityThe regions which demonstrated reduction in stillbirth rates during the period of investigation had high uptake in training and protocols, but no other known interventions which applied selectively to them and no other regions
AnalogyGrowth restriction associated with placental failure has been shown to lead to fetal death in various animal models