Introduction

During the year 2018, about 2.5 million neonatal deaths and over 2 million stillbirths were reported among the 6.2 million deaths of children under 15 years globally [1]; 98% of these deaths occurred in low- and middle-income countries (LMICs) [2] of which a vast majority of the deaths were preventable [3]. In the regions with the highest mortality, perinatal deaths are poorly recorded and are therefore most likely to be under-reported [3]. Understanding the numbers and collecting information about where, and the underlying contributing causes especially those which are avoidable is crucial in the effort to prevent perinatal mortality [4].

There is evidence that perinatal audits and feedback can lead to important improvements in practice [5–7]. The practice of collecting accurate data on how many maternal deaths occur, where the women die, why they die, and what could be done differently to prevent similar deaths have been established in many countries globally [8], including Tanzania. Consequently, applying similar data collection processes as well as a systematic analysis of events and contributing factors leading to individual perinatal death is considered achievable and will help identify health system breakdowns and inspire local solutions to prevent such deaths in the future [9].

Causes of perinatal deaths usually are obstetric in origin and are strongly associated with the causes of maternal mortality and morbidity [10, 11]. As a result, factors associated with perinatal deaths may be an indirect measurement of the availability and quality of care provided to a woman during childbirth and to the neonate during the neonatal period [12, 13]

Lack or inadequate care during the antenatal period, prematurity, low birth weight, and primigravity have strongly been associated with perinatal mortality [14]. Obstetric complications including obstructed/prolonged labor, preeclampsia/eclampsia, and antepartum hemorrhage are among the identified causes associated with perinatal mortality [15]. The perinatal audit conducted by Kidanto et al. (2009) at Muhimbili National hospital found a high (about 80%) prevalence of suboptimal care majority of them thought to be the likely cause of the adverse perinatal outcomes [16]. The maternal and fetal monitoring practice during labor was the main suboptimal factor found in the study, also delay in referral and operative interventions were prominent [16].

There have been several efforts to improve maternity and newborn care in this setting, however, the stillbirths and neonatal death rates are still unacceptably high in the setting and the country. The three delay model, stipulating delays which are decision to seek care, reaching the health facility and receiving appropriate care, have been studied extensively. Despite the available knowledge, periodic systematic review and clinical feedback are the key to the improvement of local quality of care and inform others on the available robust solutions. We decided to conduct the perinatal audit in the rural referral hospital aiming at identifying potential gaps in care provision and health-seeking behavior to foster best practice.

Method

Results

There were 4,898 births and 1,175 neonatal admissions. Out of these, there were 135 stillbirths (2.8%) of which fresh stillbirths were 60 (44.4%), and 201 neonatal deaths (4.1% of the total hospital birth and 17.1% of the total neonatal ward admission). Out of all stillbirths and neonatal deaths, 86 stillbirths case notes (63.7% of all stillbirths) and 140 neonatal deaths (death occurring within one week of life) case notes (69.7% of all neonatal deaths) were retrieved and audited. From the case audit, the mean maternal age was 27 years, and almost half (48.4%) of mothers had completed primary education. About three-quarters (73.5%) were married or living with a partner, and more than 80% were housewives (Table 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Socio-demographic characteristics of women who encountered perinatal deaths.

https://doi.org/10.1371/journal.pone.0264904.t001

More than half of the stillbirths (57.8%) and early neonatal deaths (53.5%) were male gender. Most mothers (77.9%) who experienced stillbirths were multipara with a history of stillbirths, augmentation of labor; lower level of hemoglobin and had severe maternal complications. More than three-quarters (80.5%) of all women whose pregnancy ended up with stillbirths or neonatal deaths had at least one ANC attendance. Almost half (48.7%) of all stillbirths or neonatal deaths were term pregnancies and 70% were delivered vaginally (Table 2).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Distribution of maternal characteristics by baby conditions.

https://doi.org/10.1371/journal.pone.0264904.t002

A total of 30 stillbirths (34.9%) were believed to die during labor, 5 of them had audible fetal heart rate at admission. About a quarter (25.6%), of all stillbirths, had intrapartum complications, of which severe pre-eclampsia accounted for more than one-third (38.5%) (Table 3).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Characteristics of stillbirths.

https://doi.org/10.1371/journal.pone.0264904.t003

From all the neonates who died within the first week of life at HLH whose notes were reviewed and included in this study, 77 (55%) were neonates born at HLH while 63 (45%) were born outside HLH i.e. referral from other health facilities, birth before arrival and home deliveries. Birth asphyxia, defined as a Low Apgar score (< 7) at 5 minutes, Prematurity, and Meconium aspiration were the top three recorded leading causes of early neonatal deaths (Table 4).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Characteristics of early neonatal death.

https://doi.org/10.1371/journal.pone.0264904.t004

The factors found to have likely contributed to perinatal deaths during the audit included inadequate or late ANC booking, 60 women (26.5%), home deliveries, 20 women (8.8%), and late arrival at the hospital, 17 women (7.5%), Table 5. There were high levels of inadequate documentation (62%), followed by inadequate labor monitoring (12%) Table 5, which are providers’ factors likely to have contributed to stillbirths and early neonatal deaths.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 5. Identified contributing factor.

https://doi.org/10.1371/journal.pone.0264904.t005

Discussion

This perinatal audit was aimed at highlighting the current situation, gaps, and opportunities to inform future interventions. This review report on the 86 stillbirths and 140 early neonatal deaths audited during the study period. The majority of women who experienced stillbirths were found to have a history of stillbirths, lower hemoglobin levels and suffered severe maternal complications like severe preeclampsia. Apgar scores less than 7 at 5 minutes, prematurity, and meconium aspiration were the top three causes of neonatal deaths. Inadequate or late antenatal care, delivery outside the health facility that included home and birth before arrival to HLH, inadequate labor monitoring were modifiable factors found to have contributed to both stillbirths and neonatal deaths.

Studies from Tanzania report stillbirths and neonatal deaths rate at 4.2% and 3.4% at referral hospitals [15, 21]. Our study reports low rates of stillbirths but high neonatal deaths (17.1%) when compared to hospitals of the same level in the country. The high neonatal death rates may be accounted for by the high number of referrals from the surrounding facilities in the region and neonates delivered at home or before arriving at the hospital. Similarly, other studies found that women delivering at home have an increased chance of losing their newborns as compared to those delivered at the health facility [22, 23]

Complications such as anemia in pregnancy, and cord around the neck are among the identified causes of most stillbirths, similarly to other studies from Tanzania and Italy [14, 24]. This study found prematurity, birth asphyxia, and meconium aspiration as the three leading causes of neonatal deaths similar to studies in other settings [25, 26]. Correspondingly with other studies [27, 28], in our study, we found that old age, multiparity, history of stillbirths, and induction of labor were high among women who experienced stillbirths. The findings may indicate the inadequacy of maternity care and the need for specialized care for women with a history of previous pregnancy loss.

This audit identified family/woman-related contributing factors namely inadequate antenatal care which was not attending ANC at all, late ANC booking, or very few visits less than recommended 4 ANC visits. Late arrival at the health facility when in labor and home delivery was also observed. Several studies have reported the highest odds for perinatal mortality for women with inadequate ANC visits [29–31], these findings from other studies concur with our study findings. In many Sub-Saharan African countries, the leading risk factors for perinatal deaths include pregnancy-related maternal complications, preterm births, birth complications, and neonatal infections [32, 33]. Most of these risk factors can be minimized or prevented through ANC interventions. Home delivery was found to have a high risk for perinatal mortality in our study similar to studies from other settings [34, 35]. The majority of home births are not attended by qualified birth attendants thus labor is not adequately managed leading to adverse maternal and perinatal outcomes.

Unfortunately, facility birth does not necessarily guarantee perinatal survival if the services are not optimal [36]. This audit highlights several facility/ providers’ factors thought to have contributed to perinatal deaths. Inadequate labor monitoring with inappropriate labor augmentation was found to have contributed to perinatal deaths. In some case notes, it was observed that there was a wide time range between decision and action which might have led to the poor outcomes. The audit encountered a high level of missing information and poor documentation. Likewise, partograph were not used in most cases and where used, mostly were incomplete. Our study findings concur with other studies reporting similar observations [15, 17, 37, 38]. Inadequate use of partogram and poor documentation has been reported to be associated with poor perinatal outcome [39].

The findings from this audit were communicated with the care providers at this hospital, several action plans were developed for implementation. We believe this will be followed up and measured in another perinatal audit. It is known that cheap and locally affordable interventions can save many lives [40]. Furthermore, commitment coupled with periodic audits and feedback have a significant contribution to the survival of newborns.

The main limitation of this audit is a high rate of missing information and failure to verify the information because the data were collected retrospectively. We did not include birth weight to ascertain maturity and only used gestational age which sometimes may be uncertain. Additionally, the audit lacks the opinions of both women who experienced perinatal loss and healthcare providers who provided care. However, the findings can be regarded as an eye-opener and can be used to facilitate a further review, set goals towards the improvement of maternity care provision in similar settings.

Conclusions

This audit shows that there are high rates of preventable intrapartum stillbirths and early neonatal deaths. Both women and providers’ factors were found to have contributed to the stillbirths and neonatal deaths. There is a need to encourage women to adequately attend antenatal care, utilize health facilities during birth, and improve maternity and neonatal care at the health facilities.

Acknowledgments

The authors would like to acknowledge women who gave birth or brought their sick neonates at Haydom Lutheran Hospital, research assistants, and staff of Haydom Lutheran Hospital who made this audit possible.