Value of the alert and action lines on the partogram.

A partogram based on a World Health Organisation model has been used for many years in the peripheral maternity clinics of Pikine, Senegal, to monitor labour. We have assessed the value of the partogram and efficacy of the alert and action lines. 1022 pregnant women were monitored by partogram during 4 months. The alert line was crossed in 100 (9.8%) of these cases and the frequency of neonatal resuscitation was higher for this group (relative risk 4.0, 95% confidence interval 2.3-7.1; p less than 0.0001), as was the number of "fresh" stillbirths (5.3, 1.8-15.6; p less than 0.01) (recent death may have occurred during labour). Among the women who crossed the alert but not the action line, neonatal resuscitation was also four times more likely than for the normal labour group (4.0, 2.1-7.6; p less than 0.001), and the fresh stillbirth rate was higher but not significantly so. For women who crossed both lines, the fresh stillbirth rate was ten times higher than for women in the normal labour group (9.9, 2.8-34.7; p less than 0.001). Crossing the alert line had a sensitivity of 27%, a specificity of 93%, and a positive predictive value of 17% for neonatal resuscitation. If the action line was chosen as the decision level, the positive predictive value remained the same but the sensitivity was only 8%. Health workers intervened (eg, artificial rupture of the membranes, administration of oxytocics) in half the dystocic cases. Of the women who did not receive any such treatment 44% crossed the action line compared with only 26% of those who did receive treatment (p = 0.06). The results show the usefulness and efficacy of the partogram and underscore the value of medical intervention as soon as the alert line is crossed.