Study area and period

The study was conducted in four public hospitals, administered by the Addis Ababa City Administration health bureau. The population of the city was 7.8236 Million (2019 –Estimation) [23]. The Addis Ababa City Administration has twelve referral hospitals, three primary hospitals, 106 public health centers and 722 private medium and higher clinics, and 48 private hospitals. These health facilities have their own different departments like Gyn/Obs, surgery, pediatrics, ophthalmology, and internal medicine. The selected hospitals have a pioneer maternal and neonatal health services [24]. The study was conducted from March to April 2024.

Study design

An institution based cross sectional study design was used.

PopulationSource population

The source population was represented by all neonates, with their mothers, delivered and admitted to NICU in Addis Ababa city public hospitals.

Study population

All neonates, with their mothers, delivered and admitted to NICU in the selected public hospitals under Addis Ababa City Administration Health Bureau.

Eligibility criteriaInclusion criteria

All newborns, with their mothers, delivered and admitted to NICU in the selected public hospitals.

Newborns who were born after 28 weeks of gestation.

Exclusion criteria

Incomplete documentation (no maternal or fetal measurement parameters).

Mothers who took general analgesia and seriously ill.

Neonates with congenital heart defects.

Sample size determination

The sample size was calculated for both first and second objectives and the sample size calculated for the first objective was larger than the sample size calculated for the second objective. By using a single-population proportion formula with assumptions; the proportion (P) = 28.35% [13], Z-the standard normal distribution value at 95% confidence level of Zα/2 = 1.96, 5% of absolute precision, and 10% non-response rate, sample size was: -

$$\:n=\frac\:)}^*P\right(1-P)}^}$$

$$ \right)}^2}*(0.2835*(1 - 0.2835))} \over \right)}} = 312.134 \sim 312$$

By taking 10% non-response rate, 10% * 312 = 31, the final sample size was 343.

Sampling procedure

The study was carried out in four selected public hospitals namely Gandhi memorial hospital, Zewditu memorial hospital, Abebech Gobena memorial hospital, and Tirunesh Beijing hospital. The final sample size was distributed for the above-mentioned public hospitals proportionately based on the number of last month neonates delivered and admitted to NICU in each selected hospital (by considering the last one month’s prior report). A systematic random sampling technique was employed to approach the study subjects. The sampling fraction (Kth) was calculated, and then every 2 intervals the study subjects were approached.

Study variablesDependent variableIndependent variables

Socio-demographic characteristics:

Age, residence, marital status, educational status, income level.

Antepartum related characteristics:

Parity, Antenatal care (ANC), number of ANC visits, previous delivery history, medical complications, bad obstetric history, years of birth spacing.

Intrapartum related characteristics:

Time of membrane rupture, type and duration of labor, mode of delivery, place of delivery, fetal presentation, amniotic fluid, cord accident, time of delivery.

Neonatal related characteristics:

Maternal health behavioral related factors:

Operational definition

Birth asphyxia: Neonates born in the studied hospitals and diagnosed as asphyxia by an attendant professional with an Apgar score of < 7 within the first 5 min. The definition of neonatal asphyxia of AAP is mainly used to diagnose sever asphyxia. We used the above operational definition (Apgar score of < 7 within the first 5 min) because it enabled us to include all mild, moderate and severely asphyxiated newborns.

Prolonged labor: labor exceeding 12 h in primigravida or 8 h in multipara mothers after the latent phase of the first stage of labor.

Premature rupture of membranes (PROM)): rupture of the membrane of the amniotic sac and chorion that occur > 1 h before the onset of labor.

Data collection tools and procedures

A structured questionnaire was developed first in English language, after reviewing pertinent literature that have been published previously [12, 13, 22, 25]. A pretested, structured interviewer- administered questionnaire was used to collect data on maternal socio-demographic profile. Data related to antepartum (such as parity, antepartum hemorrhage, and antenatal visits), intrapartum (such as fetal presentation, mode of delivery, meconium-stained amniotic fluid, and premature rupture of membrane), and neonatal factors (such as gestational age, birth weight, and sex) and also maternal health behaviors, were extracted using a pretested structured checklist from the medical records of the neonates and their mothers. First five-minutes Apgar score was collected by the data collectors for every newborn. Data collection process was conducted by trained bachelor of science degree holder midwives, nurses and supervised by master of science degree holders.

Data quality management

Three days training was given to the data collectors and supervisors on the purpose of the study, the methods and tools of data collection, and how to reduce the possibility of bias. The supervisors and the investigators were followed and coordinated the field work throughout the data collection period; every completed data collection form was checked for consistency and completeness by investigators and the supervisors. Pretest was done on 5% of the same source population at the non-selected health institution, which was not included in the final sample size and then based on the findings of the pretest the questionnaire was modified as necessary. Data were double-entered into Epi-info version 7.2 to ensure their quality.

Data processing and analysis

The collected data were coded and entered into Epi info version 7.2, and exported to SPSS version 24 for analysis. Bivariable and multivariable logistic regression analyses were done. Tables, figures, pie chart and text were used to present the results of the analyzed data. Independent variables with P-value < 0.25 in the bivariable logistic regression analysis were considered for multivariable logistic regression analysis, and P-value < 0.05 was used as cut off point to declare the statistical significance. Multicollinearity was checked. Finally, model fitness was checked. The Hosmer Lemeshow test of goodness of fit, which considers good fit at P-value ≥ 0.05, was used to determine the final model’s goodness of fit.