This was an institutional-based prospective cross-sectional study of admitted neonates at South Gondar public hospitals in neonatal intensive care units (NICUs) from 02 October 2021 to 30 June 2022. In the study area, South Gondar Zone, there are eight public hospitals, namely, Debre Tabor Specialized and Comprehensive Hospital, Addis Zemen District Hospital, Ebnat District Hospital, Mekane-Eyesus District Hospital, Andabet District Hospital, Wogeda District Hospital, Nefas Mewucha District Hospital, and Tach Gayint District Hospital, through which neonatal care services were delivered. There are maternal and child health (MCH), inpatient, outpatient (OPD), emergency, and referral services in all the woredas in which the data were collected.

In addition to the zonal-level health care delivery system, the woreda level provides services in its catchment area. There is at least one health center and one district hospital in each woreda. At each district hospital, there is at least one tertiary neonatal nurse (a Bsc neonatal nurse professional) providing neonatal care to general practitioners (GPs) in the neonatal intensive care unit (NICU).

Debre Tabor Comprehensive and Specialized Hospital at the zonal level and the remaining seven public district hospitals were clustered in the zone and each of the woreda, respectively, within their catchment area. The data collection procedure was conducted at each cluster proportional to size (PPS) based on the annual neonatal case flow data. All neonates aged 48–72 h were selected from among all admitted neonates in this study.

The single population proportion was calculated with the following assumptions: 95% confidence interval (CI), 5% margin of error, and 50% incidence of PHG.n=(Zα/2)2P(1−P)d2Where

N = minimum sample size required for the study

Z = standard normal distribution (Z = 1.96) with a confidence interval of 95%

P = proportion of the PTHG (0.5)

d = is a tolerable margin of error (d = 0.05). The sample size calculation to determine the incidence of PHG by using the above formula was 384.

By adding a 10% nonresponse rate, the final sample size (N) was 422.

A cluster sampling technique proportional to size (PPS) was employed among admitted neonates in the intensive care unit (NICU) at public hospitals in South Gondar. Before a maternal interview, the data collectors ascertained the case in the neonatal intensive care unit (NICU) and extracted information from the checklist for a neonatal biophysical profile, such as cause of admission, gestational age, mode of delivery, presentation, rupture of membranes, duration of labor and other information from the labor delivery room referral report form.

The data were extracted (collected) by BSc neonatal nurses and health information technicians or delivery ward coordinators for verification of the data at each health facility. The data collectors and supervisors had one day of training from a principal investigator about the tool for data collection and abstraction.

The random glucose level of the neonates (RBS) was measured by using a glucometer device with blood glucose test strips, and the selected study participants were tested for hypoglycemia by piercing the outer part of their heel as per routine measurement procedures in each hospital. The measurements were taken twice a day at 6 AM in the morning and at 6 PM in the afternoon within 48–72 h of neonatal age. Two RBS measurements, morning and afternoon, were subsequently recorded on the data collection form.

Clinical information such as gestational age (GA), mode of delivery, duration of labor, and causes of admission, such as jaundice, prenatal asphyxia (PNA), and early-onset neonatal sepsis (EONS), was collected from individual neonates (Neonatal Chart). The maternal interview was conducted after taking her neonatal RBS record in the mother's waiting room at her convenient time and with her verbal consent.

The tool was formulated after referring to various pieces of literature. The interview questionnaire consisted of information on socio-demographic characteristics; pregnancy and obstetric factors; medical and other factors; and neonatal profiles from labor delivery referral forms and admission history (4, 12, 16, 17, 21–23).

Data accuracy, completeness, and timeliness were checked by the following activities. The training of the data collectors and supervisors for one day regarding the objective of the study, the data collection tool, and the methods of data collection, checking the completeness of the data collection tool, and maintaining confidentiality by coding and limiting access to the collected pieces of information was completed.

The data collection tool was pretested by Debre Tabor Comprehensive Specialized Hospital. A pretest was administered, questions that were not clear were corrected, and a discussion was held with the data collectors and supervisors on the problems they encountered in completing the checklist. The data were checked for completeness before data entry into epi version 4.2. Proper coding of the data and categorization of outcome variables into persistent hypoglycemia and not were maintained for the quality of the data to be analyzed. Similarly, temperature data were categorized as hypothermic or normothermic according to the operationalized definition used in this study. Double data entry was performed for validity, and the data were compared to the original data. Simple frequencies and cross-tabulations were used to check missing and outlier variables, which were cross-checked with hard copies of the collected data.

Prolonged transitional hypoglycemia (PTHG) was defined as a random blood glucose level (RBS) < 47 mg/dl measured by a tool labeled as Gmate-Voice at 6 AM on the morning and at 6 PM in the afternoon for any gestational age within 48–72 h of life after delivery (1, 10, 24, 25).

Hyperglycemia: The latest record of a random blood sugar level (RBS) > 125 mg/dl was obtained at 6 AM in the morning and at 6 PM in the afternoon within 48–72 h after delivery (25).

Normglycemic status: The latest record of a random blood glucose level (RBS) of 47 mg/dl–125 mg/dl was obtained at 6 AM in the morning and at 6 PM in the afternoon within 48–72 h after delivery (24, 25).

Hypothermia was defined as the most recent measurement of an auxiliary body temperature within 48–72 h of age, which was less than 36.5°C (23).

Neonate: Newborns within the first 28 days of age after live birth with a viable gestational age (GA > 30 weeks) were included in this study.

The data were entered and analyzed using EpiData 4.2 and SPSS version 23. Descriptive analysis was performed by computing proportions and summary statistics. The data are presented as simple frequencies, summary measures, tables, and figures. Missing values were analyzed by using multiple imputation techniques.

Binary logistic regression was used to analyze the outcome variables. Bivariate and multivariate analyses were performed to determine the associations between the outcome variables and each independent variable. The assumptions for binary logistic regression were checked. The goodness of fit was tested by the Hosmer–Lemeshow test and Omnibus test. All variables with P < 0.25 in the bivariate analysis were included in the final model of multivariate analysis to control for all possible confounders, and the variables were selected by the method.

The direction and strength of the associations were measured by odds ratios (ORs) with 95% confidence intervals (CIs). The adjusted odds ratio (OR) and 95% confidence interval (CI) were estimated to identify factors associated with persistent neonatal hypoglycemia via multivariate analysis via logistic regression. In this study, a P value < 0.05 was considered to indicate statistical significance.

Out of 422 mothers whose neonates were admitted to neonatal intensive care units in the study areas fulfilled the criteria for enrollment in the study. Approximately 284 (67.30%) of the participants were aged 20–34 years, whereas the majority of the mothers were married [392 (92.90%)] (Table 1).

The median and range of random blood glucose levels at 48–72 h of age were 62 gm/dl, and the RBS ranged from 35 gm/dl to a high level, which was not recorded by a glucometer. For instance, high RBS levels can reach more than 250 mg/dl.

The incidence of persistent hypoglycemia (PTHG) was 94 (23.5%) (19.3%–28%), 280 (70%) neonates were in a normoglycemic state, and approximately 26 (6.50%) were hyperglycemic. Figure 1 shows the incidence of persistent neonatal hypoglycemia.

Approximately 218 (56.41%) mothers had a duration of labor between 1 and 12 h, and the majority of them delivered their current pregnancy through SVD, accounting for 96 (70.26%) of the mothers. The majority of the patients had a rupture of membrane (ROM) 12 h or less [365 (93.84%) of the patients] (Table 2).

The majority of the 303 (71.8%) women did not have health problems during pregnancy, while the remaining 119 (28.2%) had different health problems during their latest pregnancy. Figure 2 shows maternal health problems during pregnancy.

Among the 119 women who had a health problem, 86 (72.27%) had hypertensive disorders during pregnancy and 2 had other disorders (1.68%). Figure 3 shows the types of maternal pregnancy health problems.

The most common cause of admission was hypothermia (100, 25%), the least common cause was other causes (20, 5%), and the majority of neonates were born at term (n = 232, 54.80%). The number of very-low-birth-weight neonates was the lowest among the groups, accounting for 10 (2.56%) (Table 3).

Binary logistic regression was used to identify factors associated with PTHG. There was no marked variation according to the number of deliveries, intentions to current birth, pregnancy disorders, educational status, or maternal age group, even though these factors were associated with each other according to bivariate analysis.

Preterm birth, hypothermia, perinatal asphyxia (PNA), early onset of sepsis (EONS), and clinical pathological jaundice were risk factors for PTHG, while spontaneous vaginal delivery (SVD) was protective.

Neonates with hypothermia were 4 times more likely to develop PTHG than were those with normothermic hypothermia (AOR = 4.41; 95% CI = 2.72–10.92); preterm birth was 3.5 times more likely to be associated with PTHG than was term birth (AOR = 3.5; 95% CI = 1.69–11.97); perinatal asphyxia was 2.5 times more likely to lead to PTHG than was no asphyxia (AOR = 2.5; 95% CI = 1.34–9.67); and neonates with pathological jaundice were almost 2 times more likely to have PTHG than were those with no jaundice problems (AOR = 2.3; 95% CI = 1.21–10.34).

In contrast, spontaneous vaginal delivery (SVD) was protective and was 72% less likely to develop vaginal delivery than was cesarean section delivery (AOR = 0.72; 95% CI = 0.35–0.88) (Table 4).

The incidence of PTHG was 94 (23.5%; 19.3%–28%) in this study. According to the literature, under diagnosing or misdiagnosing neonatal hypoglycemia can lead to neonatal morbidity and mortality, which can be attributed to the aggravating factors of hypoglycemia. There have been underreporting or nearly absent studies about neonatal hypoglycemia globally and in Africa about persistent hypoglycemia.

The incidence estimates of hypoglycemia in the newborn depend both on the definition of the condition and the methods by which blood glucose concentrations are measured as well as the cutoff point of time for taking the measurement.

Studies Globally, conducted in Royal Alexandra Hospital, Canada, Denver Colorado, Waikato District Health Board, Hamilton, New Zealand, according to the American Association of Pediatrics in the entire population estimate, Boston, USA, showed that the incidence of neonatal hypoglycemia within 48 h of age ranged from 30%–63%, which was higher than that in this study (14, 16–19).

Unpublished findings indicated that 25% of neonatal hypoglycemic agents were diagnosed at St. Paul Millennium Hospital Medical College in Ethiopia (2).

The variation might be because the studies used transient hypoglycemia within 48 h after delivery, which in turn consisted of physiological hypoglycemia that subsequently returned to the normal glycemic state after the physiological process (48–72 h). This discrepancy might arise from the sample size variation and some clinical setup variation found among studies.

According to other reports, for early neonatal hypoglycemia, which was screened in Turkey, Denver Colorado, and Israel, the incidence of hypoglycemia was 12%–17%, and 12.1%, which was lower than that in this study (20, 26, 27).

This variation might arise from the definition of the condition and the methods by which blood glucose concentrations are measured as well as the cutoff point of time for taking the measurement.

Preterm birth, hypothermia, perinatal asphyxia (PNA), early onset of sepsis (EONS), and pathological jaundice were factors associated with PTHG, while spontaneous vaginal delivery (SVD) was protective.

Compared with normothermic neonates, neonates with hypothermia were 4 times more likely to develop PTHG; preterm birth was 3.5 times more likely to be associated with; and perinatal asphyxia was 2.5 times more likely to develop than it was in noasphyxia. Neonates with pathological jaundice were almost 2 times more likely to be PTHG agents than were those with no jaundice.

There are various risk factors for neonatal hypoglycemia, as reported in the literature, depending on the population and area of study, making it difficult to use these data consistently worldwide.

In line with these findings, in Israel and West Virginia University Hospital, hypothermia, low birth weight (<2,500 g), prematurity, small for gestational age (SGA), and perinatal asphyxia were found to be independent predictors of neonatal hypoglycemia (2, 13, 27).

This might be because hypothermia causes an increased demand for glucose metabolism for thermoregulation in addition to the demand for an extra uterine adaptation process. When the neonate is cold, he or she uses more glycogen to stay warm. Then, he utilized his glucose stores to stay warm, his blood sugar concentration decreased, and he became hypoglycemic and hypothermic.

Similarly, premature individuals are predisposed to developing hypoglycemia and associated complications due to their limited glycogen and fat stores, inability to generate new glucose via gluconeogenesis pathways, increased metabolic demand due to relatively greater brain size, and inability to mount a counter regulatory response to hypoglycemia.

In Ethiopia, the risk factors associated with neonatal hypoglycemia were birth weight, duration of labor, maternal age, duration of feeding initiation, hypothermia, and respiratory distress syndrome, and some of the factors leading to hypoglycemia were identified as consistent with the findings of this study (2).

These factors associated with hypoglycemia increase the risk of developing hypoglycemia because hypothermia, low birth weight, and respiratory distress syndrome affect the normal life of neonates from intrauterine to extra uterine life.

Conversely, this study revealed that spontaneous vaginal delivery (SVD) was protective and was 72% less likely to develop than cesarean section delivery was.

Preventing and treating PTHG is required because it leads to neurological damage, intellectual disability, epilepsy, personality disorders, impaired cardiac performance, and muscle weakness (23–29).

Preventing and treating PTHG prevents neurological damage, intellectual disability, epilepsy, personality disorders, impaired cardiac performance, and muscle weakness attributed to PTHG during the neonatal period. It highlights the consequences of adequate postnatal care in preventing persistent neonatal hypoglycemia. However, because of the cross-sectional nature of the study, we could not determine the causal relationships.