Jaundice and its associated factors among neonates admitted to selected referral hospitals in southwest oromia, Ethiopia: Multi-center cross-sectional study

Background Jaundice is a common clinical problem during the first month of birth throughout the world. Mainly, it is the leading cause of neonatal morbidity and mortality in developing countries. Objectives The aimed of this studied was to assess predictors of jaundice among neonates admitted to selected referral hospitals in southwest Oromia, Ethiopia, 2021. Methods An Institutional based cross-sectional study was implemented among 205 admitted neonates at selected referral hospitals in southwest Oromia, Ethiopia from October 05 to November 5, 2021. Jimma medical center (JMC), Wollega University referral hospital (WURH), and Ambo University Referral hospital (AURH) were selected by simple random sampling technique. A pretested structured interviewer-administered questionnaire and medical record review was used to collect data. Both binary and multivariable logistic regression analyses were performed to identify factors associated with neonatal jaundice. Logistic regression analyses were performed to identify factors associated with neonatal jaundice. Statistical Significance was declared at P-value less than 0.05 in the final model, and if the confidence interval does not include the null hypothesis value. Results The prevalence of neonatal jaundice was 20.5% (95%CI: 1.74–1.85). The mean age of neonates was 8.6 ± 7.8 days. Traditional medicine use during current pregnancy (AOR: 5.62, 95%CI: 1.07, 9.52), Rh incompatibility (AOR: 0.045, 95%CI: 0.01, 0.21), gestational age (AOR: 4.61, 95%CI: 1.05, 10.3), premature rupture of membrane (AOR: 3.76, 95%CI: 1.58, 8.93) and hypertension (mother) (AOR: 3.99, 95%CI: 1.13, 14.02) were factors significantly associated with neonatal jaundice. Conclusion Neonatal jaundice was relatively higher in the current study. Traditional medicine use, Rh incompatibility, premature ruptures of membrane, hypertension, and preterm gestational age were factors associated with neonatal jaundice.


Background
Neonatal jaundice is the most common reason for hospitalization in the first week of life all over the world [1]. The term neonatal jaundice refers to the yellowish color of a newborn's skin and other membranes, which indicates high levels of unconjugated bilirubin in their blood [2,3].

Source population
All admitted neonates to NICU at referral hospitals in southwest Oromia and their mothers were our source population.

Study population
All randomly selected neonates admitted to NICU referral hospitals in southwest Oromia and their respective mothers who fulfilled inclusion criteria during the study period were our study population.

Eligibility criteria
All neonates admitted to NICU with age <28 days and their respective mothers were included in the study. The study excluded neonates with incomplete medical charts, extremely preterm neonates less than 28 weeks of gestational age, and neonates whose mothers were not present or deceased.

Sample size and sampling technique
The actual sample size was determined by using the single population proportion formula, where the following assumptions were considered, the prevalence of newborn jaundice was 37% (19) a study conducted in Mekelle, northern Ethiopia, with 95% confidence interval, and 5% margin of error. Because the overall population was 398, we used a correction formula and then added a 10% no response rate, yielding a final sample size of 208.

Sampling technique
Five referral hospitals are present in Southwest Oromia. Jimma medical center (JMC), Wollega University referral hospital (WURH), and Ambo University Referral hospital (AURH) were selected by simple random sampling technique. Total admissions of neonates in these three hospitals were 1644, 1620, and 1512 at JMC, WURH, and AURH respectively from the HMIS report in 2013E.C. From this monthly average number of neonates admitted to the neonatal intensive care unit at JMC, WURH, and AURH were 137, 135, and 126 respectively by dividing total admitted neonates into twelve months. By adding these average total number of neonates admitted in these three hospitals was 398 and was taken as a frame of reference.
The calculated sample size (208) was proportionally allocated by using a proportional allocation formula Where, the sample size for each hospital, n = a total sample size to be selected, N = total number of neonates admitted in three hospitals (398), and a total number of admitted neonates in each hospital. Accordingly, 72 neonates from JMC, 70 neonates from WURH, and 66 neonates from AURH were allocated.
The required number of individuals was selected by systematic random sampling method after calculating the interval (K) by dividing the total number of admitted neonates per month at each hospital to desired sample size for each hospital. Therefore, for JMC K = = 2, WURH K = = 2, and AURH K = = 2.
The first participant was selected by lottery method from the K participant of admission registration file of neonates admitted in NICU during the data collection period. Finally, data were collected from every K [2] individual from the randomly selected participants based on the admission sequence of the registration files, and neonates admitted during the data collection period were selected based on their case flow. If randomly selected neonates were not fulfilled inclusion criteria; the next neonates were selected. A total of 205 neonates were selected through the above procedure as indicated in Fig. 1.

Data collection tool and procedure
Data were collected face-to-face using interview-based structured questionnaires and medical record review. The data collection tools were adopted and modified from a review of different kinds of literature [19][20][21][22][23][24]. The data collection tool has different sections such as socio-demographic factors, maternal characteristics, and neonatal factors for neonatal jaundice. Maternal obstetric and fetal related data that might not be addressed by interviews, such as oxytocin during labor, birth weight (kg), hypothermia, sepsis, Rh incompatibility, blood type incompatibility, mother's blood group, and neonate's blood group were collected from patient medical records.
Neonatal jaundice was collected from the medical chart after being diagnosed by the physician. The time onset of jaundice was collected from the mothers and medical chart for confirmation to classify jaundice. Using the time onset; jaundice was classified as pathological and physiological jaundice. Jaundice onset within 24 h births was classified as pathological and onset after 24 h was physiological jaundice [25].

Operational definition and definition of terms
• Preterm: neonate born less than 37wks gestational age • Extremely preterm: neonates less than 28 weeks gestational age [26].
• Low birth weight: Neonate with birth weight less than 2.50 kg • Neonatal jaundice: Neonates diagnosed as jaundiced by the physician • Pathological jaundice: Jaundice visible within 24 h s and persistent for more than one week in term (bilirubin >12 mg/dl) and more than 2 weeks in preterm (bilirubin level >15 mg/dl) [27]. • Premature rupture of membrane (PROM): Rupture of the fetal membranes after the 28th week of GA and before the onset of labor. • Birth injuries: Physical injuries experienced during childbirth, and can affect either the mother or the baby. • Critically ill: Mother is unable to respond to questions because she is suffering from a life-threatening multisystem condition that can result in substantial morbidity or mortality. • Traditional medicine use: described as any culturally used medication by mothers during her pregnancy for any purpose and not prescribed by health professionals.

Data processing and analysis
Data from three hospitals were checked for completeness, cleaned, and entered using Epi Data-Version 0.3.1 and exported to SPSS version 25 for analysis. To compare the prevalence of neonatal jaundice in three hospitals data were separately treated. For descriptive statistics and logistic regression, the data were mixed up and analyzed. Descriptive statistics were used to summarize the data using frequencies, percentages, and graphs. Both binary and multivariable logistic regression analyses were performed to identify factors associated with neonatal jaundice. The goodness of fit test designed by Hosmer and Lameshow was used to assess model fitness. The Hosmer-Lameshow goodness of fit test was used to determine model fitness, and the model was found to be properly fit with a p > 0.05. The variables in bi-variable analysis with p < 0.25 were entered into a multivariable logistic regression model. AOR with a 95% CI was used to determine the association between dependent and independent variables. The statistical significance of the association between dependent and independent variables was declared at p-value <0.05, and also statistically significant if the confidence interval does not include the null hypothesis value.

Data quality management
To assure the quality of data, clarity, and understandability a pretest was done on 5% of the sample size at Shambu general hospital, and necessary modifications were done. The data collection tool prepared in English was translated to Afaan Oromo and Amharic and then retranslated back to English to check its consistency. Data were collected by fluent speakers' local languages under regular supervision. Regular daily supervision was done for checking the consistency and completeness of the questionnaires on the daily basis by the supervisors.

Socio demographic characteristics of respondents
A total of 205 newborns were enrolled in the study, with a 100% response rate for those who had complete information on their laboratory records and medical record review chart.
From maternal socio-demographic characteristics of respondents, most mothers 142 (69.3%) were between 20 and 34 years with a mean age of 26 + 5.5 y.rs. More than half 116 (56.6%) of mothers were from rural areas. About two-thirds 150 (73.2%) of respondents were unemployed (as shown in Table 1).

Maternal obstetric characteristics
From obstetric characteristics of respondents, the majority of 191 (92.7%) of mothers were followed antenatal care. More than half 112 (54.6%) of mothers were the Primiparous. About 109 (53.2%) were prim gravida. Below ten percent 12 (5.9%) of mothers were taking traditional medicine. Nearly 3 (1.5%) of mothers had a history of blood transfusion during pregnancy. About 10 (4.9%) of mothers were delivered at home. Less than one-third 33 (16.1%) of mothers were used oxytocin for induction during labor. Mothers who had a history of premature rupture of the membrane were accounted for 39 (19%). The majority of mothers' blood groups were A+ and Rh-positive which was 76 (37.1%) and 186 (90.7%) respectively. Less than ten percent 13 (6.3%) of mothers were having a history of gestational diabetic Mellitus whereas 14 (6.8%) mothers were a history of gestational hypertension (as shown in Table 2).

Independent predictors of neonatal jaundice
To identify the association between neonatal jaundice and predictive variables, bi-variable logistic regression analysis was first done for all independent variables. A total of seventeen variables were found to be associated in the crude analysis were a candidate for multivariable analysis with a p-value <0. 25. Five variables, namely: traditional medicine use during pregnancy, Rh incompatibility, gestational age, premature rupture of membrane, and hypertension showed significant association with neonatal jaundice among neonates.
Neonates born to mothers who took unknown traditional medicine during pregnancy were 5.62 times more likely to have neonatal jaundice than those who were not taken traditional medicine (AOR = 5.62 95% CI: 1.07, 9.52). Neonates who were born from mothers who had a history of PROM were 3.76 times more likely to have neonatal jaundice than neonates born from mothers without PROM history (AOR = 3.76 95% CI: 1.58,8.93). Neonates who were born from hypertensive mothers were 3.99 times more likely to have neonatal jaundice as compared with neonates born to non-hypertensive mothers (AOR = 3.99 95% CI: 1.13,14.02). The risk of  Table 4).
The observed finding was lower than studies done in, Bloemfontein (55.2%) [29], Rwanda (44.3%) [30], Nigeria (32.6%) [31], Nepal (39.95%) [32], and Mekelle Northern Ethiopia (37.3%) [33]. Possible reason for lower prevalence observed compared to in Malaysia, Nepal and Nigeria was due to difference in methodology, sample size and study population. The variability of current study from Rwanda study conducted in Rwanda was retrospective over two years of admitted neonates. The variation between current study and Mekelle Ethiopia might be the current study was done only at referral hospitals where more obstetric care is expected and study period for Mekelle was three month whereas the current was only one month study period.
The current study was higher than study done in Congo (7.2%) [34], and Nairobi Kenya (16%) [35]. Possible reason might be a methodological variation, and if bilirubin level 5 mg/dl and exclude the clinical jaundiced with low bilirubin level as compared to our study setting. The reason for Windhoek Namibia might be methodological variation that study in Namibia excludes neonates with hemolysis, unknown date of birth, delayed in transient and being icteric (yellow sclera and skin). The difference for Brazzaville Congo might be Study setting in one hospital and they includes only if bilirubin level 5 mg/dl and excluding the clinical jaundiced with low bilirubin level comparatively the current study includes both clinical jaundiced with low bilirubin and high level of bilirubin as jaundiced.
Traditional medicine use during pregnancy, Rh incompatibility, and preterm gestational age, premature rupture of membrane and hypertension during pregnancy were factors associated with neonatal jaundice. Rh incompatibility was found to be significantly associated with neonatal jaundice. This finding was in line with study done in Turkey on factors affecting neonatal jaundice 80 (11.3%) of neonatal jaundice was Rh incompatibility [36]. In addition, this study was in line with previous studies conducted in Mekelle Northern Ethiopia [37]. The possible explanation may be Rh incompatibility between maternal and fetal cause hemolytic. This develop DOL-duration of labor, PROM-premature ruptures of membrane, SVD-spontaneous vaginal delivery, ANC-Antenatal care.
when mother Rh negative and fetus Rh positive that results in formation of antibodies in maternal blood which attack fetal red blood cells. The fetal red blood cells broken down can develop hyperbilirubinemia [38]. According to our study, unknown traditional medicine use during pregnancy was significantly associated with neonatal jaundice. This finding was in line with the studies conducted in Malaysia [39], and Nigeria [40]. The possible explanation could be traditional medicine used during pregnancy could cross the placenta and affect fetal liver that might result in decreased bilirubin clearance from the body and cause hyperbilirubinemia [41,42]. Medicinal plants and herbal remedies contain substances that can be toxic to the human body and the fetus. Potential effects of indiscriminate use of medicinal plants are embryotoxicity, teratogenic, and abortifacient effects [43].
Preterm gestational age was significantly associated with neonatal jaundice. This finding was in line with studies conducted in Iran [44,45], Rwanda [30], Nigeria [46], Benin Nigeria [47] and Denmark [48]. The possible cause might be due to short life span of erythrocyte in premature neonates result in increased bilirubin production and reduced bilirubin elimination capacity [49].
According to our study, premature rupture of membrane was significantly associated with neonatal jaundice. Our finding was compatible with studies done in Turkey [36], Iran [50], and Indian [51]. The possible explanation might be premature rupture of membrane results in preterm delivery which result in immaturity of liver that cannot get rid of much bilirubin from the body [52].
According to the current study, gestational hypertension was significantly associated with neonatal jaundice. This finding was in line with studies done in Turkey [36], India [53], Iran [54] and China [55]. The possible reason is hypertension during pregnancy leads to cause abnormal coagulation profile with bleeding on babies which increase bilirubin level in the blood [56]. In addition, gestational hypertension might be result in prematurity which fails to conjugate normally produced bilirubin from red blood cell which results in jaundice [57].

Strength of the study
A study was conducted at multicenter cross-sectional study. The study identifies new factors associated with outcome. The study is important for researchers as bassline to identify definite etiologies and underlying causes of neonatal jaundice.

Limitation of the study
As the study design was cross-sectional; it is difficult to form the causal relationship between neonatal jaundice and the associated factors. Due to small sample was obtained from each hospital; this study didn't compare among three hospitals regarding risk factors of neonatal jaundice.

Conclusion
The prevalence of neonatal jaundice in southwest Oromia referral hospitals was relatively higher. Preterm gestational age at birth Rh incompatibility, premature rupture of membrane, use of unknown traditional medicine during pregnancy, and gestational hypertension were factors significantly associated with neonatal jaundice. For Researchers, longitudinal studies should be important to identify definite etiologies and underlying causes of neonatal jaundice.

Ethical statement
The study was conducted after ethical approval from Research Ethical Committee of Jimma University with IHR/554/21. After permission secured from the hospital data collection was started. Written, informed consent was obtained from all participants by the local language prior to interview. Data was kept confidential. The rights to withdraw from the study were respected for all.

Informed consent
Informed consent was obtained from all the participants included in the study.

Author contribution statement
Gutu Belay: Conceived and designed the experiments; Performed the experiments; Analyzed and interpreted the data; Contributed reagents, materials, analysis tools or data; Wrote the paper. Asfaw Gerbi; Tesema Etefa: Conceived and designed the experiments; Performed the experiments; Wrote the paper. Teka Gebremariam: Performed the experiments; Analyzed and interpreted the data. Tsion Tilahun: Conceived and designed the experiments; Contributed reagents, materials, analysis tools or data; Wrote the paper. Emebet Chimdi: Analyzed and interpreted the data; Contributed reagents, materials, analysis tools or data.

Data availability statement
Data included in article/supplementary material/referenced in article.

Declaration of competing interest
The authors declare no conflict of interest.

Acknowledgments
First of all, we would like to praise our God, without his help all this would have been impossible. Also, we don't want to pass without acknowledging data collectors and study participants. We want to express our appreciation to our colleagues for their continuous encouragement.