Delay in seeking appropriate care: Phase-one delay
As shown in Table
4, phase-one delay (delay in decision-making to seek appropriate care) was more frequent than other phases of delay and was identified in 39% (98/250) of deaths. The most prominent reason for this delay was failure to recognize or report danger signs that occurred during pregnancy or labour. This delay was identified in 81 cases. The most frequently reported danger sign consisted of poor or absent fetal movements and signs of preterm labour such as preterm contractions or preterm rupture of membranes.
Some women were not aware of the severity of the danger signs and preferred to stay at home waiting for spontaneous resolution of these problems. Others had asked for advice from a partner, relative, neighbour or friend or had sought help from private pharmacies or traditional healers. The decision to seek care at health facilities had often been taken late, when danger signs had worsened or failed to subside.
Five women, who sought informal care providers, received traditional medicines because the main cause of danger signs was attributed to ifumbi. This was vaguely described as a clinical condition with abnormal symptoms varying from simple discomfort to major complications. The users of traditional medicines believed that ifumbi could be effectively cured by traditional medicine.
Four women had also used traditional medicines during labour, both before and after admission. The women considered these as effective stimuli of uterine contractions. Some also believed in the potential protection of these medicines against any witchcraft that could negatively influence maternal and fetal outcomes. Case study A (Additional file
1) illustrates a situation where the lack of knowledge on the severity of danger signs, combined with the utilization of traditional medicines, contributed to a phase-one delay.
Poor uptake of care by registered healthcare providers was noted in 16 cases. In two cases, the pregnancies were reportedly unintended and no antenatal care was sought. In two other cases, women were single mothers and did not have time to seek antenatal care due to a heavy workload to earn their living. Six women, including two with poor obstetrical histories (previous perinatal death), did not attend follow-up visits during pregnancy. Three of them explained that they could not return because they had not been clearly briefed by the healthcare provider about the need to attend. Others failed to give clear reasons for their non-attendance at follow-up visits. Five women initially wanted to deliver at home and came late to the health facility, when complications had already occurred. One woman refused the induction of labour suggested by her doctor despite having preterm rupture of the membranes and oligohydramnios.
Delay in receiving appropriate care at a health facility: Phase-three delay
Various circumstances that led to the phase-three delays – in relation to suboptimal care received once at a health facility – were found in 37% (93/250) of deaths, as shown in Table
4.
Before admission, most cases of death were related to delays in women’s referrals from health centres or private clinics to district hospitals or from the district to the tertiary referral hospital. Such delays were identified in 12 cases, where death was subsequent to intrapartum-related hypoxia, preterm rupture of membranes, or preterm labour. One of these deaths was related to uterine rupture and occurred at a health centre after prolonged labour. In another case, a woman who had signs of preterm labour reportedly had a long waiting time (4 h) at a health centre before being examined and referred to the district hospital.
Late diagnosis and failure to diagnose or perform proper management were other common inadequacies in the practices of healthcare providers. In five cases of poor obstetric history (previous perinatal death), these women did not receive any special attention during pregnancy. In one case, gestational diabetes was diagnosed late in the pregnancy by providers at a district hospital in a woman who had initially been attending antenatal care clinics at a health centre. In another case, a woman referred from a health centre for poor progress of labour also had an umbilical cord prolapse, which was detected upon arrival at the hospital.
Similar delays in diagnosis or management had occurred for some women presenting with hypertension or preeclampsia. Another woman who also had gestational diabetes had received insufficient treatment at a private clinic. Substandard provision of care was observed in another woman at 37 weeks of gestation who was treated by a nurse at a health centre with an anti-inflammatory drug for one week due to multiple joint pain. Another case was related to umbilical cord prolapse that occurred at a health centre after an amniotomy was performed.
After admission, insufficient labour and fetal monitoring, and delay in taking appropriate action during the management of pregnancy or labour complications were the most frequent factors related to the phase-three delay. These inadequacies were found in 40 cases where the deaths were linked to various conditions, mainly prolonged or obstructed labour, preterm labour, maternal hypertension, preeclampsia and placenta abruption. These conditions also included two cases of uterine rupture that occurred in hospital. There were difficulties in handling one case of breech presentation and another case of vacuum-assisted delivery. In both cases, the doctor was called to assist after unsuccessful efforts by a midwife or nurse to perform the delivery. A caesarean section was delayed for 30 min because two other caesarean sections were ongoing and no other operating theatre was available in the hospital.
Non-adherence to the national guidelines was another barrier that delayed access to appropriate care in hospital after admission. An accidental double-dose of intravaginal misoprostol (100 mcg) was repeatedly administered to two women for labour induction. This could have led to the hyper-stimulation of uterus and fetal heart, causing or aggravating fetal hypoxia, which was the cause attributed to the death.
An attempt at vaginal delivery of a twin pregnancy was made in another woman, where ultimately a caesarean section was urgently performed due to the failure of delivery by vaginal route, which subsequently caused fetal asphyxia. An early caesarean section would have prevented this asphyxia, because the first twin had presented in a breech position. The national guidelines suggest a caesarean where the first twin has not presented in a cephalic position.
Late diagnosis of some pregnancy-related complications was another contributing factor to the phase-three delay, which was identified in two cases of placenta abruption and one case of preeclampsia. In these cases, the delay in making the diagnoses also affected the initiation of adequate treatment, as depicted in case study C (Additional file
3).
After delivery, insufficient newborn resuscitation and management of neonates with severe intrapartum-related hypoxia constituted major causes of suboptimal care. In some instances, resuscitation was not promptly initiated within the first minute after birth.
Some equipment, including suction tubes and facemasks for ventilation, were used inappropriately or were not in good condition when resuscitation began. In two cases, ventilation with bag and mask was applied in two neonates before proper skin drying and stimulation. Upon arrival to the neonatal intensive care unit, they were still wrapped in the wet cloths used when drying the skin at birth. In one of these cases, the endotracheal intubation was delayed due to the limited skills of the healthcare provider who was performing the advanced resuscitation.
The management of some cases of intrapartum-related hypoxia was temporarily hampered by a lack of second-line treatment such as phenytoin. Some of these cases also required the use of a ventilator machine, equipment that was not available in one of the two study hospitals. Two cases of death occurred while the transfer of the infant to a higher level of care was being discussed.
Another important suboptimal care factor consisted of the irregular monitoring of vital signs. This was observed in three preterm babies who developed hypothermia and bradycardia. In the three remaining cases, suboptimal care factors were related to inadequate management of some neonatal emergency conditions, including jaundice, feeding of a sick preterm baby, and delay in blood transfusion. The jaundice had eventually evolved into kernicterus. The inappropriate enteral feeding of one preterm baby presenting with food intolerance and frequent vomiting led to pulmonary aspiration.