A cross-sectional design was used. It targeted NICU at Kirehe District Hospital, Rwanda. The study period was from February to March 2024.

The Eastern Province of Rwanda is home to Kirehe District Hospital, where the study was carried out. Serving as a referral center for a predominantly rural population, this hospital provides maternity and neonatal care services, including NICU facilities that can manage critical neonatal conditions. Its NICU offers the perfect environment for researching the variables influencing newborn admissions.

NICUDefinition

NICU admission was explicitly defined as the admission of any neonate within the first 28 days of life to the NICU for specialized medical care, regardless of whether the infant was born at Kirehe District Hospital or referred from another health facility.

The study population included all newborns delivered at or referred to Kirehe District Hospital during the study period from February to March 2024. The mothers of these newborns were interviewed to gather information on maternal factors.

InclusionCriteria

All newborns admitted to the NICU or other hospital wards during the study period who were born at the hospital or referred there met the inclusion criteria.

ExclusionCriteria

Newborns with missing or incomplete medical records, those admitted for less than 24 hours unless they were released or died within that time, newborns with more than 28 days of life, and mothers who either refused to participate or were unable to provide the necessary information were all excluded.

The sample size consisted of 340 neonates from the targeted population, who were delivered at or referred to Kirehe District Hospital were included in the census sampling method. This included 59 referrals from other facilities and 281 hospital deliveries. To gather information on maternal factors, the mothers of these newborns were also included.

A structured data abstraction was employed to extract pertinent clinical and demographic data from the medical records of both mothers and newborns. Referral status was retained and accounted as an analytic variable to contextualize admission prevalence. To maintain confidentiality each participant was given a unique identifier. Maternal socio-demographic information (age, education and marital status), obstetric and gynecologic history (prenatal care visits, gestational age, delivery type, gravida and parity) and neonatal clinical features and conditions (birth weight, birth age, asphyxia, respiratory distress) were important variables. Interview-administered questionnaires concentrating on prenatal care and delivery conditions were used to gather maternal health data.

SPSS software (version 18) was used to clean and analyze the data. Using frequencies and percentages, descriptive statistics compiled important maternal and neonatal characteristics. By dividing the number of admitted neonates by the total number of births during the study period, the prevalence of NICU admission was determined. While binary logistic regression was employed in multivariate analysis to identify significant risk factors, bivariate analysis using Chi-square tests investigated associations between maternal/neonatal factors and NICU admission. Statistical significance was set at p < 0.05, with 95% confidence intervals.

StatisticalModelingStrategy

Two multivariable logistic regressions models were constructed:

Model 1: Maternal and delivery-related factors only;

Model 2: Model 1 plus neonatal clinical diagnoses.

Collinearity was assessed using variance inflation factors (VIF). Multiple gestations were included to account for clustering by pregnancy.

A total of 340 neonates were included, with 39.7% (n = 135) admitted to the NICU and 60.3% (n = 205) not admitted. Among neonates, 82.2% were born at Kirehe District Hospital, Rwanda while only 17.4% were referred from other health facilities. 55.9% were female and 55.9% were born at term. Low birth weight (<2500 g) was observed in 60.0%, respiratory distress syndrome (RDS) in 29.4%, neonatal sepsis in 26.5%, hypoglycemia in 30.0%, birth asphyxia in 8.8%, and congenital anomalies in 7.4%. Most neonates (94.1%) had normal Apgar scores at 1 minute. Further details of the neonatal characteristics are detailed in Table 1.

Mothers were predominantly aged 20 - 34 years (50%), with 40% aged 35 years or older. Rural residency accounted for 65%, and 67.1% were married. Half of the mothers had primary education, 15% had gestational diabetes, 40% experienced antenatal infections, and 10% had multiple gestations. Delivery modes were equally split between the cesarean section and vaginal delivery. Further details of the maternal characteristics are detailed in Table 2.

Table 1. Neonatal characteristics (N = 340).

Table 2. Maternal characteristics (N = 340).

The NICU admission prevalence was 39.7%, reflecting that nearly 40% of neonates required intensive care during the study period.

3.3.1. Maternal Factors

Significant associations with NICU admission were found for maternal age (χ² = 12.6, p = 0.002), place of residence (χ² = 5.4, p = 0.02), gestational diabetes (χ² = 25.4, p < 0.001), maternal infection (χ² = 18.2, p < 0.001), multiple gestations (χ² = 14.8, p < 0.001), and mode of delivery (χ² = 6.9, p = 0.009). Younger mothers (<20 years) had lower admission rates compared to older groups. Rural-born neonates had higher admission rates than urban counterparts. Gestational diabetes and maternal infections markedly increased NICU admissions. Variance inflation factors (VIFs) for all variables were below 1.5, indicating no substantial multicollinearity among predictors. Multiple pregnancies and cesarean deliveries were also linked to higher admission rates. No significant association was observed for maternal education, marital status, or antenatal care visits. Further details of the maternal factors are detailed in Table 3.

3.3.2. Neonatal Factors

Neonatal factors were significantly associated with admission to the neonatal intensive care unit (NICU). Male newborns accounted for a higher proportion of NICU admissions compared to females, though this difference was not statistically significant (51.9% vs. 48.1%; χ² = 2.5, p = 0.114). Preterm infants (<37 weeks) were predominantly admitted to the NICU compared with term and post-term infants (74.1% vs. 22.2% and 3.7%, respectively; χ² = 28, p < 0.001). Similarly, low birth weight (<2500 g) was strongly associated with NICU admission (88.9% vs. 11.1%; χ² = 35, p < 0.001). Birth asphyxia, congenital anomalies, and neonatal sepsis were also significantly associated with NICU admission, with higher proportions of affected newborns requiring intensive care (birth asphyxia: 22.2% vs. 0.0%, χ² = 5, p = 0.025; congenital anomalies: 18.5% vs. 0.0%, χ² = 5.6, p = 0.018; neonatal sepsis: 66.7% vs. 0.0%, χ² = 15.04, p < 0.001). Variance inflation factors (VIFs) for all variables were below 5, indicating no significant multicollinearity among predictors. These findings suggest that prematurity, low birth weight, birth complications, congenital anomalies, and neonatal sepsis are critical determinants of NICU admission. Further details of the neonatal factors are detailed in Table 4.

Table 3 and Table 4 summarize bivariate analyses.

Table 3. Maternal factors associated with neonatal admission.

*Significant at p < 0.05; VIF values < 5 indicate no multicollinearity.

Table 4. Neonatal factors associated with neonatal admission.

*Significant at p < 0.05; VIF values < 5 indicate no multicollinearity.

3.4.1. Maternal Factors

Table 5 presents the multivariate logistic regression analysis examining the association between maternal and delivery factors and NICU admissions. In Model 1, several maternal and delivery factors were independently associated with NICU admission. Compared with mothers aged < 20 years, maternal age of 20 - 34 years (aOR = 2.35; 95% CI: 1.20 - 4.62, p = 0.01) and ≥35 years (aOR = 4.79; 95% CI: 2.10 - 10.91, p = 0.001) significantly increased the odds of admission. Gestational diabetes was a strong predictor (aOR = 6.91; 95% CI: 3.15 - 15.17, p = 0.001), as were maternal infection (aOR = 3.54; 95% CI: 1.58 - 7.95, p = 0.002) and multiple gestations (aOR = 5.62; 95% CI: 2.25 - 14.04, p < 0.001). Cesarean delivery was associated with higher NICU admission compared with vaginal delivery (aOR = 2.51; 95% CI: 1.14 - 5.51, 0.002), and rural residence remained an independent risk factor (aOR = 2.12; 95% CI: 1.10 - 4.06, p = 0.002). Further details of the multivariate logistic regression of the maternal factors influencing neonatal admissions in NICU are detailed in Table 5.

3.4.2. Neonatal Factors

Preterm birth (<37 weeks) significantly increased NICU admission odds (aOR = 4.50, 95% CI: 2.10 - 9.61, p < 0.001). Low birth weight (<2500 g) was also a strong predictor (aOR = 5.80, 95% CI: 3.02 - 11.14, p < 0.001). Birth asphyxia (aOR = 3.25, 95% CI: 1.11 - 9.41, p = 0.031), RDS (aOR = 12.00, 95% CI: 6.00 - 24.01,), congenital anomalies (aOR = 2.75, 95% CI: 1.20 - 6.32, p = 0.019), and neonatal sepsis (aOR = 8.00, 95% CI: 3.80 - 16.57, p < 0.001) significantly elevated NICU admission risk. Post-term birth showed no significant associationIn the final multivariable logistic regression model, both maternal and neonatal factors were independently associated with NICU admission. Compared with mothers aged < 20 years, maternal age of 20 - 34 years (aOR = 2.1, 95% CI: 1.05 - 4.21) and ≥35 years (aOR = 4.12, 95% CI: 1.88 - 9.02) significantly increased the odds of neonatal admission. Gestational diabetes (aOR = 5.8), maternal infection (aOR = 3.2), and multiple gestations (aOR = 4.8) were strong maternal predictors of NICU admission. Neonates delivered by cesarean section also had higher odds of admission compared with vaginal delivery (aOR = 2.2), while rural residence showed a borderline association (aOR = 1.95, p = 0.057).

Neonatal clinical factors demonstrated particularly strong associations. Preterm birth (aOR = 4.5) and low birth weight (aOR = 5.8) markedly increased the likelihood of NICU admission, whereas post-term birth was not statistically significant. Birth asphyxia was independently associated with higher odds of admission (aOR = 3.25). The strongest predictors were respiratory distress syndrome (aOR = 12.0) and neonatal sepsis (aOR = 8.0), followed by congenital anomalies (aOR = 2.75), underscoring the dominant role of acute neonatal complications in determining NICU admission. Further details of the multivariate logistic regression of neonatal factors influencing neonatal admissions in NICU are detailed in Table 6.

Table 5. Multivariate logistic regression of maternal and delivery factors influencing NICU Admissions (Model 1).

*Statistically significant (p < 0.05).

Table 6. Multivariate logistic regression of maternal and neonatal factors influencing NICU admissions (Model 2).

*Statistically significant (p < 0.05).

3.4.3. Combined Maternal and Neonatal Factors

Integrated analysis confirmed that maternal age, gestational diabetes, infections, multiple gestations, mode of delivery, preterm birth, low birth weight, birth asphyxia, RDS, and neonatal sepsis were significant predictors of NICU admission. Older maternal age and pregnancy complications notably increased neonatal risk requiring intensive care. The integrated analysis shows that NICU admissions at Kirehe District Hospital, Rwanda are influenced by two factors including maternal and neonatal. Advanced maternal age, gestational diabetes, maternal infections, multiple gestations, and cesarean deliveries substantially increased the probability of neonatal admission. Neonatal complications, specifically prematurity, LBW, birth asphyxia, RDS, congenital anomalies, and sepsis, were strong predictors, with RDS and sepsis demonstrating very high risk. These results underline the necessity for strengthened antenatal care, early identification of high-risk pregnancies, and enhanced neonatal emergency readiness. The findings are consistent with regional evidence from sub-Saharan Africa, highlighting the importance of integrated maternal and neonatal interventions to reduce preventable NICU admissions and advance newborn survival.

The prevalence of NICU admissions in this study was 39.7%, indicating that nearly four in ten newborns required specialized neonatal care [24]. This relatively high proportion reflects the considerable burden of neonatal morbidity and the need for strengthened NICU infrastructure [24]. Similar NICU admission rates have been reported in other low- and middle-income countries (LMICs), ranging between 30% and 45%, suggesting that this is a common issue in resource-limited settings [5][25][26].

The study revealed a strong association between preterm birth (<37 weeks) and NICU admission. Preterm neonates were more than eight times more likely to be admitted than term neonates. This aligns with previous studies highlighting prematurity as the leading cause of neonatal mortality and morbidity globally, due to organ immaturity and related complications [12][26]-[29].

Low birth weight (<2500 g) was also significantly associated with NICU admission, with nearly seven times the odds of being admitted compared to neonates with normal birth weight. This finding is consistent with other studies that show low birth weight neonates are at higher risk for hypothermia, infections, and respiratory problems [3][29]-[31].

Other neonatal conditions significantly associated with NICU admission included birth asphyxia, respiratory distress syndrome (RDS), congenital anomalies, and neonatal sepsis [32][33]. Neonates with RDS had twelve times higher odds of admission, underscoring the severity of respiratory complications in newborns [32]. These findings support global literature that consistently points to these conditions as major contributors to NICU admissions and poor neonatal outcomes [5][31]

Conversely, newborn gender and Apgar score at 1 minute were not significantly associated with NICU admission, indicating that clinical complications and gestational maturity were more decisive predictors of NICU admission than basic demographic characteristics, an observation consistent with findings from other African and global studies [28][34].

Maternal characteristics were also found to significantly influence neonatal outcomes. Advanced maternal age (≥35 years) was associated with a fourfold increase in NICU admission risk compared to adolescent mothers. These findings align with global studies that associate advanced maternal age with complications such as gestational hypertension, diabetes, and poor placental function [3][35][36].

The presence of gestational diabetes and maternal infections significantly elevated NICU admissions, as neonates born to affected mothers faced nearly seven times and over three times the risk, respectively [37]. This aligns with studies that link hyperglycemia and peripartum infections to complications including macrosomia, hypoglycemia, and respiratory issues in neonates [12][28][31].

Multiple gestations and cesarean section deliveries were also associated with increased NICU admissions [32][33]. These findings are consistent with previous studies suggesting that multiple births are prone to preterm delivery and low birth weight, while cesarean sections often reflect underlying maternal or fetal distress necessitating neonatal support [5][25].

Furthermore, place of residence emerged as a significant determinant. Newborns from rural areas were more likely to be admitted to NICU than those from urban areas, possibly reflecting disparities in healthcare access, delays in care, and limited resources in rural settings. These findings are echoed in previous literature advocating for equitable health service access across geographic settings [26][38].

Finally, these findings highlight a dual approach including avoiding maternal complications to decrease preventable NICU admissions and ensure good neonatal care for those with direct clinical needs. Such organized interventions are fundamental to advancing neonatal survival and health outcomes in rural and resource-limited settings like Kirehe District, Rwanda.