Magnitude of preterm hospital neonatal mortality and associated factors in northern Ethiopia: a cross-sectional study.

OBJECTIVE: This study aimed to assess the magnitude of preterm neonatal mortality in hospitals and associated factors in northern Ethiopia.
DESIGN: Institutional-based cross-sectional study.
SETTING: Comprehensive specialised hospitals in the Tigray region, northern Ethiopia. PARTICIPANTS: Preterm neonates admitted in Ayder and Aksum comprehensive specialised hospitals PRIMARY OUTCOME: Magnitude of preterm neonatal mortality. SECONDARY OUTCOME: Factors associated with preterm neonatal mortality RESULT: This study was conducted from 1 April 2019 to 15 May 2019 among 336 participants with a response rate of 96.8%. The magnitude of preterm neonatal mortality was 28.6% (95% CI: 24.0 to 33.7). In multivariable logistic regression, respiratory distress syndrome (adjusted odd ratio (AOR)=2.85; 95% CI: 1.35 to 6.00), apnoea of prematurity (AOR=5.45; 95% CI: 1.32 to 22.5), nulli parity (AOR=3.63; 95% CI: 1.59 to 8.24) and grand parity (AOR=3.21; 95% CI: 1.04 to 9.94) were significant factors associated with preterm neonatal mortality. However, receiving Kangaroo mother care (AOR=0.08; 95% CI: 0.03 to 0.20) and feeding initiated during hospitalisation (AOR=0.07; 95% CI: 0.03 to 0.15) were protective against preterm neonatal mortality.
CONCLUSIONS: The magnitude of preterm neonatal mortality in hospitals was still high. Interventions geared towards curbing preterm in-hospital neonatal mortality should strengthen early diagnosis and treatment of preterm newborns with respiratory distress syndrome and apnoea of prematurity; while concomitantly reinforcing the implementation of kangaroo care and early feeding initiation is important. © Author(s) (or their employer(s)) 2021. Re-use permitted under CC BY-NC. No commercial re-use. See rights and permissions. Published by BMJ.

As strength, the study addressed all comprehensive specialised hospitals found in the Tigray region, Ethiopia.

The study results were presented using veracious statistical methods.

Due to the cross-sectional nature of the study design, we are unable to establish cause–effect relationships.

Introduction

In the world, preterm-related complications are the leading causes of neonatal mortality1; especially in low-income and middle-income countries (LMICs).1–3 The global rate of neonatal mortality is 18 deaths per 1000 live births; 35% due to prematurity.2 According to the WHO report, 81.1% of preterm births occur in LMICs, this is due to low quality of care for pregnant mothers.4 5 Moreover, the neonatal mortality ranges from 19.9% to 69%.6 7 In sub-Saharan Africa and South Asia, prematurity is responsible for 0.75 million deaths of under 5 years of age; but the reduction rate of neonatal mortality is very low.1

Preterm newborns (born before 37 completed weeks of gestation)8 had high risk of mortality9 10 and morbidity11 as compared with full-term newborns. Studies showed that preterm newborns are more vulnerable to organ failure, neurodevelopmental and learning impairment, visual disorders and long-term cardiovascular and non-communicable diseases.12–14 Prematurity is responsible for 3.1% of the disability-adjusted life years occurring worldwide.15 Moreover, it is a major contributor to the loss of human potential16 and hospital admission.17

To achieve the global sustainable development goal (SDG) of decreasing the neonatal mortality rate and mortality of under 5 to at least 12 and 25 deaths per 1000 live births, respectively by 2030,18 better preventions and treatments of the leading causes of mortality are necessary. Furthermore, the WHO designed different strategies including essential newborn care, neonatal resuscitation, Kangaroo mother care (KMC), treatment of premature babies with complications and comprehensive neonatal intensive care.13 19 20 Even though, previous studies identified respiratory distress syndrome (RDS), asphyxia, maternal residency, gestational age, birth weight, jaundice, hypothermia, neonatal sepsis and maternal chronic disease21–32 as a factors of mortality, preterm neonatal death remains a global burden.33

In Ethiopia, approximately 42% of mortality among under 5 is attributed to neonatal deaths.34 Although the federal health ministry designed and implemented different strategies such as antenatal care (ANC) service and family planning service to prevent preterm birth and also in 2015 developed another strategy to reduce neonatal mortality,35 preterm mortality is still high in our country.36 According to the Ethiopian Mini Demographic Health Survey (EDHS) 2019 report, the neonatal mortality rate is 30%,37 which is increased from EDHS 2016 report (29%).38

To facilitate the reduction rate of this mortality, extensive and ongoing studies are needed to identify factors. However, in Ethiopia little is known about the magnitude of preterm in hospital neonatal mortality and its associated factors. Therefore, to support the achievement of the SDG plan, we aimed to assess the magnitude of preterm neonatal mortality in the hospitals and its associated factors in northern Ethiopia.

Methods and materials

Study setting and design

An institutional-based cross-sectional study was conducted from 1 April 2019 to 15 May 2019 in the comprehensive specialised hospitals of Ayder and Aksum, Tigray, Ethiopia. The Tigray region is one of the nine federal administrative regions in Ethiopia and has two comprehensive specialised hospitals. The Tigray region has an estimated total population of 5 377 144: 2 651 167 males and 2 725 977 females. Among the total population, 159 164 are under 1 year of age. There are 2 comprehensive specialised hospitals, 15 general hospitals, 23 primary hospitals, 245 health centres and 750 health posts.39

Ayder comprehensive specialised hospital provides its service to more than 9 million people in its catchment areas of the Tigray, Afar and northern parts of the Amhara regional state. It has a total capacity of about 500 inpatient beds in all departments and other specialty units, including 45 neonatal beds in the neonatal intensive care unit (NICU) and more than 170 000 patient flows per year. The hospital has more than 2165 staffs. Aksum comprehensive specialised hospital provides its service to a population of over 3.6 million from the central, north-west and western zones of the Tigray regional state. It has a total capacity of 173 beds, including 13 neonatal beds.

Source and study population

All preterm neonates admitted in Ayder and Aksum comprehensive specialised hospitals were source population. The study population was preterm neonates admitted in these hospitals from 1 February 2017 to 30 January 2019.

Eligibility criteria

Neonates born before their 37 weeks of gestation were included in this study.

Preterm newborns with an incomplete record for the outcome variable were excluded.

Sample size determination and sampling technique

The sample size was estimated using a single population proportion formula with a CI of 95%, α=0.05, and p=28.8% taken from a study conducted in Gonder, Ethiopia.40 By considering a 10% non-response rate, the total sample was 347. A total of 1242 preterm neonates were admitted to these hospitals from 1 February 2017 to 30 January 2019. Participants were proportionally selected using a simple random sampling technique from both hospitals.

Operational definitions

RDS: which is characterised by grunting while breathing, rapid or shallow breathing, and flaring of the nostrils.41

Apnoea of prematurity: respiratory pauses >20 s or pauses <20 s that are related to bradycardia (<80 beats/min), central cyanosis and/or oxygen saturation <85% in neonates born at <37 weeks gestation and without causal disorders that induce apnea.41

Perinatal asphyxia: an Apgar score that remained less than 7 (at 5 min after birth) and evidence of acute hypoxic compromise with acidaemia.42

Data collection procedure and tool

The lists of participants were obtained from NICU health management system registration book. To collect the data from the charts, we used a pretested adapted English version data extraction checklist. Lastly, the data were scouted by two data collectors from randomly selected charts.

Data quality control

We conducted a pretest on 5% (15) of the sample in Mekelle General Hospital. A single imputation was used to manage missing values. After 2 days of training, two data collectors (BSc nurses) and one supervisor (MSc in paediatrics and child health nursing) participated in the data collection. The twofold data entry was done and the consistency of the entered data was verified by comparing the two distinctly entered data. The content validity and inter-rater reliability tests were conducted to check the validity and reliability of the tool, respectively.

Data processing and analysis

Data were coded and entered by Epi-Data manager V.4.4.2.1 and exported to Stata statistical software V.14 for clearance and analysis. Descriptive statistics frequency and percentage were conducted. Bivariate and multivariable analysis was performed to see the statistical association between the outcome and independent variables. Factors with a p value <0.2 in the bivariate analysis were entered into multivariable logistic regression. Variables with a p value <0.05 in multivariable analysis were considered statistically significant factors. Lastly, we used an adjusted OR with 95% CI to determine the association.

Patient and public involvement

The patient and/or the public were not involved in the design, development, analysis and publication of this study.

Result

Of 347 preterm newborns, 336 were eligible and included in this study with a response rate of 96.8%.

Neonatal, maternal and obstetric-related characteristics

Two hundred and eighty-one (83.63%) mothers delivered in the hospital. Two hundred and sixty nine (80.06%) mothers were found in the age range of 20–35 years; a median age of 27 years (IQR: 22–30). One hundred and eighty five (55.06%) mothers were primipara and 330 (98.21%) mothers had follow-up of ANC. Only 34 (10.12%) mothers had obstetric complications during their index pregnancy (table 1).

Table 1

Sociodemographic and obstetric characteristics related to mothers of preterm newborns admitted to the NICU of the Ayder and Aksum comprehensive specialised hospitals, northern Ethiopia, 2019 (n=336)

Characteristics	Frequency	Per cent
Maternal age in years
Less than 20	27	8.04
Between 20 and 35	269	80.06
35 and above	40	11.90
Place of delivery
Home	16	4.76
Health centre	39	11.61
Hospital	281	83.63
Maternal residency
Urban	191	56.85
Rural	145	43.15
Parity
Nulli parity	185	55.06
Multiparty	111	33.04
Grand parity	40	11.90
ANC visit
One	5	1.5
Two	39	11.6
Three	161	47.9
Four and above	122	36.3
Type of pregnancy
Singleton	196	58.33
Multiple	140	41.67
Mode of delivery
Spontaneous vaginal delivery	268	79.76
Caesarean section	68	20.24
Obstetric-related complication during index pregnancy
Yes	34	10.12
No	302	89.88
Mother received corticosteroid before delivery
Yes	291	86.61
No	45	13.39
Maternal serostatus (HIV)
Positive	15	4.46
Negative	321	95.54

ANC, antenatal care; NICU, neonatal intensive care unit.

Among 336 preterm newborns, 191 (56.85%) were males and 260 (77.38%) were within gestational age between 28 and 32 weeks; but not including 32. Feeding was initiated for 218 (64.88%) of preterm newborns during their hospital stay and only 118 (35.12%) of the neonates received KMC. Three hundred and nine (91.96%) neonates had low birth weight (<2500 gram) and approximately 147 (43.75%) and 242 (72.02%) preterm neonates had RDS and sepsis, respectively (table 2).

Table 2

Characteristics of preterm neonates admitted in NICU of Ayder and Aksum comprehensive specialised hospitals, northern Ethiopia, 2019 (n=336)

Characteristics	Frequency	Per cent
Place of admission
Ayder	279	83.03
Aksum	57	16.97
Sex of the neonate
Male	191	56.85
Female	145	43.15
Gestational age at birth in weeks
Very preterm (28 to <32 weeks)	260	77.38
Moderate or late preterm (32 to <37 weeks)	76	22.62
Birth weight at birth (in grams)
Less than 2500	309	91.96
2500 and above	27	8.04
Feeding initiated during hospitalisation
Yes	218	64.88
No	118	35.12
Newborn received Kangaroo mother care
Yes	118	35.12
No	218	64.88
Weight for gestational age at birth
Small	31	9.23
Appropriate	305	90.77
Hypoglycaemic diagnosed at admission
Yes	24	7.14
No	312	92.86
Hypothermia diagnosed at admission
Yes	154	45.83
No	182	54.17
Newborn with clinical diagnosed sepsis
Yes	242	72.02
No	94	27.98
Newborn diagnosed with apnoea of prematurity
Yes	14	4.17
No	322	95.83
Perinatal asphyxia diagnosed at birth
Yes	14	4.17
No	322	95.83
Newborn diagnosed with jaundice
Yes	37	11.01
No	299	88.99
Newborn diagnosed with respiratory distress syndrome
Yes	147	43.75
No	189	56.25

NICU, neonatal intensive care unit.

Magnitude of preterm hospital neonatal mortality

In this study, the magnitude of preterm in-hospital neonatal mortality was 28.6 with 95% CI (24.0 to 33.7). Among the neonates who died, 88 (91.7%) of them were within the first 7 days of life.

Factors associated with preterm neonatal mortality

In the bivariate analysis gestational age, KMC, feeding status, birth weight, maternal residency, hypoglycaemic at admission, sepsis, RDS, parity and apnoea of prematurity were significantly associated with preterm neonatal mortality. However, in the multivariable analysis RDS, parity, KMC, feeding status and apnoea of prematurity were still statistically significant factors.

The odds of mortality for preterm newborns who received feeding during their hospital stay were reduced by 93% (adjusted odd ratio (AOR)=0.07; 95% CI: 0.03 to 0.15) compared with their counterparts. Similarly, providing KMC reduces the risk of death by 92% (AOR=0.08; 95% CI: 0.03 to 0.20) compared with their comparison group. Preterm newborns diagnosed with RDS and apnoea of prematurity had 2.8 (AOR=2.85; 95% CI: 1.35 to 6.00) and 5.4 times (AOR=5.45; 95% CI: 1.32 to 22.5) higher mortality risk compared with their opposite groups, respectively. Furthermore, neonates born to nulliparous mothers had 3.6 times higher risk of death compared with neonates born to multipara mothers (AOR=3.63; 95% CI: 1.59 to 8.24) and neonates born to grand para mothers had 3.2 times (AOR=3.21; 95% CI: 1.04 to 9.94) higher risk of mortality compared with those born to multipara mothers (table 3).

Table 3

Factors associated with preterm in-hospital neonatal mortality in Ayder and Aksum comprehensive specialised hospitals, northern Ethiopia, 2019 (n=336)

Characteristics	Mortality		COR (95% CI)	AOR (95% CI)
	Yes	No
Feeding initiated during hospital stay
Yes	22	196	0.07 (0.04 to 0.12)	0.07 (0.03 to 0.15)*
No	74	44	1
Maternal residency
Urban	46	145	1
Rural	50	95	1.66 (1.03 to 2.67)	1.68 (0.85 to 3.34)
Hypoglycaemic at admission
Yes	3	21	2.97 (0.86 to 10.20)	1.00 (0.12 to 8.07)
No	93	219	1
Newborn with clinical diagnosed sepsis
Yes	76	166	1.69 (0.96 to 2.97)	1.31 (0.56 to 3.08)
No	20	74	1
Newborn diagnosed with RDS
Yes	67	80	4.62 (2.77 to 7.71)	2.85 (1.35 to 6.00)*
No	29	160	1
Newborn with apnoea of prematurity
Yes	8	6	3.54 (1.19 to 10.51)	5.45 (1.32 to 22.5)*
No	88	234	1
Newborn received KMC
Yes	7	111	0.09 (0.04 to 0.20)	0.08 (0.03 to 0.20)*
No	89	129	1
Birth weight at birth in grams
Less than 2500	93	216	3.44 (1.01 to 11.72)	3.03 (0.62 to 14.83)
2500 and above	3	24	1
Gestational age
Very preterm (28 to <32 weeks)	41	35	4.36 (2.54 to 7.49)	1.70 (0.77 to 3.74)
Moderate or late preterm (32 to <37 weeks)	55	205	1
Parity
Nulliparity	48	137	1.48 (0.890 to 2.48)	3.63 (1.59 to 8.24)*
Multiparity	38	73	1
Grand parity	10	30	1.56 (0.69 to 3.53)	3.21 (1.04 to 9.94)*

*Significantly associated factors.

AOR, adjusted odd ratio; COR, crude odd ratio; KMC, Kangaroo mother care; RDS, respiratory distress syndrome.

Discussion

In the present study, the magnitude of preterm hospital neonatal mortality was 28.6; 95% CI (24.0 to 33.7). This finding is similar to studies done in Iran (28.7%43 and 27.4%)28 Nigeria (27.7%),44 Ethiopia (25.2%)22 and 29.7%,21 and WHO and UNICEF multicounty report 29.3%.45 It was high compared with studies done in Iran (9.1%),30 Cameroon (15.7%),7 Ethiopia (18.2%46and 16.7%)47 and save the children multicounty report 15%.48 This could be due to variation in the study population, quality of care provided and difference in the availability of basic interventions, especially surfactant.

However, it was low compared with studies done in Ethiopia (34.9%24 and 36.1%),49 South Africa (64%),50 population-based study done in LMIC (37.5%)50 and Jordan (40%).51 This might be due to the improvement in care in recent years due to access to healthcare service, the accessibility of trained healthcare professionals and the behaviour of the community toward the search and utilisation of health in Ethiopia.52

Preterm newborns diagnosed with RDS had a higher risk of death. This finding was in line with studies done in Ethiopia (Gonder40 and Jimma24). Another study conducted in selected Ethiopian hospitals support this finding.53 This might be due to the inaccessibility of surfactant treatment in our country, the similarity in the study setting (specialised hospitals), and RDS also leads to acute complications such as pulmonary haemorrhage and intraventricular haemorrhage that increase the risk of mortality.54

In this study, KMC had a protective effect on preterm neonatal mortality. This finding was in line with the study done in Gonder, Ethiopia.40 This finding was also supported by a systematic review and meta-analysis study that showed that KMC reduces neonatal mortality.55 This might be due to the fact that KMC has the benefit of protecting the newborn from infection, effectively treating hypothermia, improving gastrointestinal function and cardiorespiratory stability and encouraging breast feeding.56

Preterm neonates born to nulliparous and grand-parity mothers had high mortality compared with neonates born to multiparty mothers. This finding was supported by studies done in northern Ethiopia,47 Uganda57 and Australia.58 This might be related to nulli and grandparous mothers who are at increased risk of unfavourable newborn outcomes and intrapartum complications.59 60

In the present study, preterm neonates who received feeding during their hospital stay had a low risk of mortality. The current finding was supported by studies conducted in Ethiopia.61 62 This correspondence might be due to colostrum which has an effect on reducing the risk of neonatal infections such as respiratory and gastrointestinal infections, and can also reduce the risk of hypoglycaemic and hypothermia.63

Apnoea of prematurity was also identified as a significant predictor of preterm death. It might be due to the potential it has to decrease systematic blood pressure that leads to hypoperfusion of the brain and hypoxic ischaemic injury, and occurred mainly with desaturation and bradycardic episodes.64

Each study has its own limitations. This study had some limitations. First, it was a cross-sectional study that does not show a cause and effect relationship. Second, due to the retrospective nature of the study, some variables were squandered, such as institutional-related factors. Lastly, although our study was a region-wide study, its generalizsability to other settings could be another limitation.

Conclusion

In the current study, the magnitude of neonatal mortality was still high. RDS, apnoea of prematurity, KMC, feeding status and mother parity were significantly associated with preterm in-hospital neonatal mortality. Therefore, to reduce the burden of this problem, it is better to encourage the implementation of KMC and early initiation of feeding. Also, it is better to strengthen early diagnosis and treatment of preterm neonates with RDS and apnoea of prematurity. Furthermore, we recommend prospective multicentre studies to identify institutional factors.