Journal of Neonatal Nursing xxx (xxxx) xxx Contents lists available at ScienceDirect Journal of Neonatal Nursing journal homepage: www.elsevier.com/locate/jnn Extracting the frequent sequential patterns among the factors associated with neonatal birthweight Donald Douglas Atsa’am a,*, Temidayo Oluwatosin Omotehinwa b, Samuel Nii Odoi Devine c, Emmanuel Awuni Kolog d,f, Oluwaseun Alexander Dada e,g a Department of Computer Science and Informatics, Faculty of Natural and Agricultural Sciences, University of the Free State, QwaQwa Campus, South Africa b Department of Mathematics and Computer Science, Federal University of Health Sciences, Otukpo, Nigeria c Department of Information and Communication Technology, Presbyterian University College, Abetifi, Ghana d Department of Operations and MIS, University of Ghana, Accra, Ghana e Department of Computer Science, University of Helsinki, Helsinki, Finland f Centre for Multidisciplinary Research and Innovation, Abuja, Nigeria g The School of Software, Lekki-Lagos, Nigeria A R T I C L E I N F O A B S T R A C T Keywords: The objective of this study was to employ the association rules mining technique to find the frequent co- Birthweight occurrences among maternal factors associated with birthweight. Secondary data consisting of 189 records Factors of birthweight with predictors that assess the risk factors that influence low neonatal birthweight were employed for association Association rules rules mining. The extracted rules show that the different patterns of co-occurrences among these factors: a Co-occurrence of factors Pregnant mother mother’s pre-pregnancy weight, presence/absence of uterine irritability, attendance to antenatal care visits in the first trimester, hypertension history, maternal age, and a history of premature labor often lead to either a low or normal neonatal birthweight. The extracted rules could serve as a reference tool in the education, treatment, and care of pregnant women to ensure positive outcomes related to birthweight. Flowing from these rules, pregnant mothers should be educated about the relationship between their attributes and the probable birthweight of their unborn babies. 1. Introduction and Makari, 2012; Hilaire et al., 2021). Chromosomal abnormality and cardiopathy in babies with low birthweight have been found responsible One of the most important measurements taken of a newborn is the for mortality (Hilaire et al., 2021). The prevalence of issues related to birthweight. Other measurements include the head and abdominal cir- birthweight varies from one region to another. One of the issues related cumferences. Birthweight is a strong indicator of the health status of a to birthweight is pre-eclampsia and eclampsia. According to a study by newborn and that of the mother. The goal of neonatal birthweight Macedo et al. (2020), the worldwide prevalence of measurement is to determine whether there will be a need for emer- pre-eclampsia/eclampsia is 6.7%. Low- and middle-income countries gency interventions (Gladstone et al., 2021). Some emergency in- have the highest prevalence of pre-eclampsia, at 11.5% and 10.6% terventions include continuous positive airway pressure or positive respectively. In Haiti, a high incidence of pre-eclampsia was found to be end-expiratory pressure, and the administration of oral immunoglob- responsible for low birthweight (Hilaire et al., 2021). Pre-eclampsia ulin to treat necrotizing enterocolitis (Foster et al., 2016). Generally, often leads to preterm labor and delivery. Babies born prematurely newborns with low birthweight require close monitoring, as they are have a high risk of cerebral palsy, breathing and feeding difficulties potentially susceptible to postnatal morbidity and mortality. Morbidity because of non-fully developed organs (ACOG, 2020). Pre-eclampsia has such as macrocephaly, microcephaly, cardiopathy, chromosomal ab- been associated with a reduction in fetal growth and women with normality, pneumonia, anemia, bronchopulmonary dysplasia, and uro- pre-eclampsia are about 4 times more likely to deliver small for gesta- genital infections have been associated with low birthweight (Groothuis tional age (SGA) newborns and/or preterm (Ødegård et al., 2000; Xiao * Corresponding author. E-mail addresses: donatsaam@alumni.emu.edu.tr (D.D. Atsa’am), temidayo.omotehinwa@fuhso.edu.ng (T.O. Omotehinwa), samuel.nodevine@presbyuniversity. edu.gh (S.N.O. Devine), eakolog@ug.edu.gh (E.A. Kolog), alexander.dada@helsinki.fi (O.A. Dada). https://doi.org/10.1016/j.jnn.2022.11.016 Received 6 September 2022; Received in revised form 23 November 2022; Accepted 23 November 2022 1355-1841/© 2022 Neonatal Nurses Association. Published by Elsevier Ltd. All rights reserved. Please cite this article as: Donald Douglas Atsa’am, Journal of Neonatal Nursing, https://doi.org/10.1016/j.jnn.2022.11.016 D.D. Atsa’am et al. J o u r n a l o f N e o n a t a l N u rsing xxx (xxxx) xxx et al., 2003). Hoodbhoy et al. (2021) affirmed that children exposed to 1.1. Related literature fetal pre-eclampsia suffer cardiovascular dysfunction. There are other socio-economic issues related to low birthweight such as prenatal Several studies have been conducted to identify and understand smoking, poor nutrition and ante-natal care. The global prevalence of neonatal birthweight. Some of such studies have focused on the char- smoking in pregnant women is 1.7% and 72.5% of these women are acteristics of pregnant mothers that influence birthweight (Balogun daily smokers (Lange et al., 2018). et al., 2020). Afaya et al. (2021) observed that the phenomenon of low To improve the quality of care received by infants with low birth- birthweight (LBW) of neonates over the past decade has not seen any weight, the World Health Organization published guidelines on optimal decline in sub-Saharan Africa. In the same vein, Goldenberg and Cul- feeding of infants with low birthweight in middle-income countries hane (2007) held that studies related to low birthweight are a good (WHO, 2011). There are other guidelines for the management of infants measure to address child mortality or morbidity. Extant literature has with low birthweight in other regions. For example, the government of established a relationship between a mother’s socio-demographic Swaziland through the ministry of health partnered with the United characteristics, lifestyle, and neonatal birthweight (Atsa’am et al., Nations International Children’s Emergency Fund (UNICEF) and WHO 2022; Balogun et al., 2020; Singh et al. 2009; Wang et al., 2020). Many to develop the national neonatal care clinical guidelines (UNICEF, researchers have focused on comparing one or more maternal factors 2018). The guidelines specify how infants with low birthweight should and their contribution to birthweight. A review of studies focused on be managed (UNICEF, 2018). identifying mother’s characteristics, lifestyle, attributes, and their effect According to the World Health Organization (WHO, 2022a), high on the weight of the new-born are presented in this section. birthweight as defined in the 2022 version of the International Classi- fication of Diseases (ICD-11) mortality and morbidity statistics, is a 1.2. Association between a mother’s age, initial weight, and neonatal birthweight greater than 4000g (4 kg). It is associated with a gestation birthweight period greater than 42 weeks. On the other hand, low birthweight ranges between 1500 and 2499g (1.5–2.499 kg) and is associated with a short Studies indicate that a mother’s age (Balogun et al., 2020) and gestation period (WHO, 2022b). Apart from short gestation, low birth- weight at their last menstrual period before pregnancy (Shin et al., weight could be associated independently with Intrauterine Growth 2013) influence the birthweight of their child (Singh et al., 2009). Retardation (IUGR) or a combination of both . Low birthweight resulting Metgud et al. (2012) reported that as maternal age increases, the from IUGR could lead to permanent growth and the cognitive deficiency chances of LBW also increase, at a prevalence rate of 22%. Senthilkumar (Kono, 2021; Zimmerman, 2018). and Paulraj (2015) used a dataset collected from a hospital in Massa- The effects of low birthweight on the neonate’s health, development, chusetts, USA to predict the risk factors associated with LBW. The study and the health of the mother are of great importance to public health. reported that a mother’s weight before pregnancy and their age are Further, the frequency of occurrence of low birthweight in each society useful in predicting LBW. Further, in ranking the factors useful in pre- is very important to public health. In 2015, about 20.5 million babies dicting the likelihood of a LBW, a mother’s last weight before pregnancy suffered from low birthweight globally (UNICEF-WHO, 2020). This ranked top, with 100.00%, while a mother’s age ranked second, with represents about 15% of the total births in 2015. Of the 20.5 million 98.00%. This implies that a mother’s weight and age are highly influ- babies with low birthweight, 1 million (7.2%) were born in more ential to birthweight. Additionally, the weight or body mass index (BMI) developed regions such as North America, Europe, Japan, Australia, and of a mother before pregnancy was observed as a factor that can influence New Zealand. About 12.8 million (17.3%) were born in Asia, 1.4 million the weight of their neonate (Goldenberg and Culhane, 2007). Specif- (9.9%) in Oceania, and 5.7 million in Africa (Blencowe et al., 2019; ically, mothers with a high BMI usually have neonates with a high UNICEF-WHO, 2020). The prevalence of low birthweight in both birthweight and vice-versa (Goldenberg and Culhane, 2007; Sutan et al., developed and developing countries are said to be on the increase. For 2014). Related to this, Mayo Foundation for Medical Education and instance, the prevalence of low birthweight in southern Europe is 7.6% Research ([MFMER], 2022) outlined pre-pregnancy BMI in the and 4.68% in eastern Europe (Erasun et al., 2021). A study by DeMarco following categories: below 18.5 [underweight], 18.5–24.9 [healthy et al. (2021) reported a 53% increase in the prevalence of low birth- weight], 25–29.9 [overweight], and 30 or more [obese]. weight among adolescent pregnancies in Canada. In a survey conducted in 35 sub-Saharan countries in Africa including South Africa, Namibia, 1.3. Association between a mother’s smoking status, history of Angola, Cameroun, Ethiopia, Kenya, Nigeria, Senegal, and so on, find- hypertension, and neonatal birthweight ings revealed a 9.7% prevalence of low birthweight (Tessema et al., 2021). This is higher than the prevalence rates in eastern (4.68%) and Several studies have separately established a strong relationship southern Europe (7.6%). As of 2016, the prevalence rate of low birth- between maternal smoking during pregnancy and low birthweight weight in Taiwan was 8.4% (Waits et al., 2021). Adverse birthweight has Bernstein et al. (2005); Cliver et al. (1995); Cnattingius (2004); Xi et al. a debilitating effect on the health and development of the child which (2020). The separate findings from these studies infer that the more the could result in death. A study conducted by Woelile et al. (2021) in foetus is exposed to primary or passive smoking by the mother, the Ethiopia revealed that two out of every seven low birthweight neonates higher the chances of a low birthweight. die during follow-up. Steer et al. (2004) found an association between low birthweight and Considering the importance of birthweight, the objective of this maternal hypertension during pregnancy. The researchers argued that study was to deploy the association rules mining technique on a sec- birthweight is optimal when the highest diastolic blood pressure of the ondary birthweight dataset to determine how maternal attributes asso- mother is between 70 and 90 mmHg during pregnancy. Thus, mothers ciated with birthweight co-occur. The study extracted association rules who experience high or low blood pressure during pregnancy are likely that show the most frequent co-occurrences among the factors associ- to deliver a baby with LBW. In a related study, Metgud et al. (2012) ated with birthweight. Association rules mining is an unsupervised observed that neonates delivered by mothers who were exposed to machine learning activity that is anchored on the theory that if item X passive smoking and experienced pregnancy-induced hypertension were occurs then item Y is very (or less) likely to occur (Atsa’am and Wario, likely to have a LBW. Furthermore, a study conducted to determine the 2022a). The extracted rules provide useful insight on the maternal fac- factors associated with LBW in Malaysia identified young maternal age tors that frequently occur together to lead to either low or adequate and hypertension as having a significant association with LBW (Sutan birthweight. This can be a reference tool for pregnant mothers and et al., 2014). The authors employed a multivariate conditional logistic health practitioners on how adverse birthweight outcomes can be regression approach to identify the associated outcomes. When adjusted avoided. with mother’s age, gestational age and history of previous LBW, 2 D.D. Atsa’am et al. J o u r n a l o f N e o n a t a l N u rsing xxx (xxxx) xxx hypertensive mothers were found to be 4.52 times more likely than Ang, 2007). The technique was originally developed for finding the most non-hypertensive mothers to deliver a baby with LBW (Sutan et al., frequent purchasing patterns of customers from historical transactions 2014). In a related study, Khan and Jamal (2003) reported that young data. Subsequently, studies in other fields such as healthcare (Atsa’am & maternal age and hypertension independently affect birthweight Wario, 2022a), security (Atsa’am et al., 2022), and sociology (Olaleye negatively. et al., 2022) have applied association rules to extract the frequently occurring patterns among domain variables. 1.4. Association between a mother’s number of premature labors, Mathematically, a rule is represented as A ⇒ B, where A and B are physician visits during first trimester, and neonatal birthweight items that frequently occur together in the dataset, and A ∩ B = ∅. The item(s) on left-hand side (LHS) is called the antecedent while the right- Preterm birth has been defined as that before 37 weeks of gestation hand side (RHS) is called the consequent (Atsa’am & Wario, 2022b). A while early preterm birth takes place before 32 weeks of gestation (Yang combination of the LHS and the RHS is referred to as an itemset. The et al., 2016). A study by Yang et al. (2016) found that pregnant women association rule depicted above implies that “if item A occurs in a who experienced either preterm or early preterm delivery were at a high transaction, then item B will also likely occur in the same transaction” risk of experiencing a recurrence in subsequent pregnancies. Other (Zhou and Yau, 2007). The strength of a rule is measured by the factors such as old age, underweight BMI, smoking, and short intervals following criteria (Atsa’am & Wario, 2022a; Atsa’am & Wario, 2022b; between pregnancies were also associated with recurring preterm de- Goh and Ang, 2007; Zhao and Bhowmick, 2003): liveries and low birthweight (Fuchs et al., 2018; Simonsen et al., 2013; Support: This measures how frequent an itemset appears in the Spong, 2007). Furthermore, a new guideline for a positive pregnancy dataset with respect to the total number of transactions in the dataset. It outcome has recommended a total of at least eight ante-natal care visits is given in Equation (1). (WHO, 2016). According to WHO (2016), the first among these visits number of transactions containing both A and B P(A ∩ B) should take place within the first 12 weeks of gestation (first trimester). Support= = (1) total number of transactions N At each visit, the pregnant women should be counselled on diet and nutrition, smoking and substance use, prevention of malaria and HIV, Confidence: This measures the likelihood that item B will occur and so on (WHO, 2016). These have the potential to ensure that a whenever item A occurs. It is given in Equation (2). pregnant woman and her baby stay healthy, ultimately resulting in a number of transactions containing both A and B P(A ∩ B) normal birthweight. Confidence= =total number of transactions with A P(A) The reviewed literature shows that the association between various (2) maternal attributes and neonatal birthweight have been examined in previous studies. However, these studies were largely concerned with It is instructive to note that support and confidence are originally how maternal attributes independently affect birthweight. Furthermore, reported as ratios (Atsa’am & Wario, 2022b); however, these are the most frequent sequential patterns of these factors have not been interpreted better in percentages by multiplying the values by 100. reported. Clearly, a study that examines the frequently occurring pat- Lift: is the ratio of the observed frequency of co-occurrence of the terns among maternal attributes and their effect on birthweight was items and the expected frequency. It is given in Equation (3). lacking. Therefore, the present study sought to address this gap by confidence P(A ∩ B) extracting the association rules that show the maternal attributes that Lift= = (3) expected confidence P(A).P(B) frequently co-occur and the effect on birthweight. Lift can take a score between zero and infinity. When comparing two 2. Materials and methods or more rules, the stronger rule is that which maximizes the values of support, confidence, and lift (Atsa’am & Wario, 2022a; Olaleye et al., 2.1. Data 2022). The research data originated from a medical facility in Springfield, 2.3. Data and variables preprocessing Illinois, USA. It consists of 189 records with eight predictors that assess the risk factors that influence low neonatal birthweight. There is a class The research data was preprocessed and transformed into a trans- variable that indicates the birthweight of the neonate in grams. The data action format prior to the deployment of association rules mining is secondary and freely available in one of the packages in R (Venables technique. For the categorical and binary variables (mother’s race, and Ripley, 2002). The data variables are given in Table 1. smoking status, history of hypertension, presence of uterine irritability), each category was represented as a separate item (variable) in the new format. For the numeric variables (mother’s age, weight, number of 2.2. Modeling tool: association rules previous premature labors, number of physician visits during the first trimester, birthweight), literature evidence was used to create categories Association rules mining is a technique that finds interesting re- that represent new items. For example, mother’s age was categorized lations (e.g., frequent patterns, causal structures) in a dataset (Goh and into three as ≤ 19, 20–34, and ≥35. The categorization is supported by WHO (2020) where pregnancy outcomes have been associated with Table 1 these age ranges. Further, the number of physician visits during the first Research data variables. trimester was categorized into two as “no visit”, and “one or more Variable Description Data Type visits”. This categorization is supported by a new WHO guideline that age Mother’s age (in years) Numeric encourages one ante-natal care visit during the first trimester (WHO, lwt Mother’s weight at the last menstrual period Numeric 2016). In the new format, data points were represented with the new race Mother’s race; with 1 = white, 2 = black, 3 = other Categorical category name corresponding to the original value. Blank spaces were smoke Mother’s smoking status during pregnancy; with 0 = does Binary entered where the category did not occur in the original data. The not smoke, 1 = smoke ptl Number of previous premature labors Numeric preprocessed data variables in transaction format are shown in Table 2. ht Mother’s history of hypertension; with 0 = no, 1 = yes Binary ui Presence of uterine irritability; with 0 = no, 1 = yes Binary 2.4. Rules mining ftv Number of physician visits during the first trimester Numeric bwt Birthweight in grams Numeric R programming language version 4.1.3 was utilized for mining the 3 D.D. Atsa’am et al. J o u r n a l o f N e o n a t a l N u rsing xxx (xxxx) xxx Table 2 3. Results Preprocessed data items. Item Description The preprocessed data consisted of 21 new set of variables, each representing an item in the birthweight data, as shown in Table 2. teenage-mother Pregnant mother aged below 20 years mother-aged-20-to-34-years Pregnant mother aged 20–34 years Table 3 presents the rules that show the frequent co-occurrences mother-aged-35-years-or-above Pregnant mother aged 35 or above among the factors associated with birthweight. It is instructive to note mother-below-40kg Weighs below 40 kg before pregnancy that the factors could be positive or negative. The rules show the mother-40kg-or-above Weighs 40 kg or above before pregnancy frequent patterns of occurrences of these factors plus the birthweight white-race Mother is white race black-race Mother is black race outcome. Other-race Mother is other race Rule 1 indicates that the co-occurrence of these maternal factors, non-smoker Mother does not smoke during pregnancy namely, a pre-pregnancy weight below 40 kg, uterine irritability during smokes Mother smokes during pregnancy pregnancy, and not attending to ANC visit in the first trimester, often no-previous-premature-labor No previous premature labor(s) results to a low birthweight. one-previous-premature-labor One previous premature labor(s) two-or-more-previous-premature- Two or more previous premature labor(s) Rule 2 shows that the pattern of occurrence of the maternal factors labor where the pregnant mother has no history of hypertension, aged 20–34 no-history-of-hypertension No history of hypertension years, has a pre-pregnancy weight below 40 kg, and the presence of history-of-hypertension History of hypertension uterine irritability, often results to low neonatal birthweight. uterine-irritability-absent No uterine irritability uterine-irritability-present Uterine irritability present Rule 3 shows that a sequential pattern of these maternal factors, No-ANC-visit No physician visit in first trimester namely, maternal age of 35 years or above, a pre-pregnancy weight one-or-more-ANC-visits One or more physician visit(s) in first equal to or above 40 kg, and attendance to one or more ANC visit(s) trimester during the first trimester, often results to a normal birthweight. low-birthweight Birthweight below 2.5 kg Rule 4 established that a sequential co-occurrence of the following normal-birthweight Birthweight 2.5 kg or more factors often leads to a low birthweight: low pre-pregnancy weight (below 40 kg), a lack of ANC visits in the first trimester, and no previous association rules in this study. The apriori algorithm was invoked on the premature labor. preprocessed data, using the minimum thresholds of 0.01 and 0.5 for support and confidence, respectively. A support threshold of 0.01 means 4. Discussion that any rule extracted had the same pattern of risk factors occurring together (itemset) in at least 1% of the total records. Similarly, a con- The LHS of each rule show the patterns of co-occurrence of certain fidence threshold of 0.5 means that the rules extracted had the same maternal factors that often produce a specific birthweight outcome consequent in at least 50% of the records consisting of the same ante- (shown on the RHS). Each of the four rules has a confidence of 100%. cedents. The support and confidence thresholds effectively streamlined This indicates that in all the records of the research data where either the number of rules returned by the algorithm. More than 200 rules were low or normal birthweight was the consequent, the exact sequence of generated from the birthweight data. After eliminating the redundant, maternal factors could be observed. Furthermore, each rule has a lift weak, and repetitive rules, the four strongest rules presented in Table 3 score greater than 1.0. A lift score above 1.0 is an indication that both were retained. the LHS and RHS of the rule often occur together more than expected. The confidence and lift scores show that each of the four rules is strong and can be generalized. Table 3 It is instructive to note that three out of the four rules have their Association rules. consequent as low neonatal birthweight – Rules 1, 2, and 4. It could be Rule LHS RHS Support Confidence Lift observed that a lack of ANC visits appears in two out of these three rules – Rules 1 and 4. This underscores the importance of ANC visits in taming 1 {mother-below- => {low- 0.01 1.00 6.75 the possibility of a low birthweight. ANC visits to a health facility help to 40kg, uterine- birthweight} irritability- monitor the health conditions of pregnant mothers and provide them present, no- with the required support and care. However, inability of some mothers ANC-visit} to attend ANC visits has been attributed to their family economic status 2 {no-history-of- and level of education (Ghaemmaghami et al., 2013). hypertension, mother-aged- It could be observed that low maternal pre-pregnancy weight (below 20-to-34-years, 40 kg) appears in all the three rules where low birthweight is the mother-below- => {low- 0.01 1.00 3.20 consequent. This agrees with Shin et al. (2013) and Singh et al. (2009) 40kg, uterine- birthweight} who averred that a mother’s weight at their last menstrual period before irritability- pregnancy has an influence on the birthweight of their child. ANC visits present} 3 {mother-aged- are strongly recommended for this category of pregnant mothers where 35-years-or- they could receive counselling, support, and care required for pregnancy above, one-or- weight-gain. This is in line with Tela et al. (2019) who found that more-ANC- weight-gain during pregnancy is strongly associated with birthweight. visits, mother-40kg- {normal- 0.01 1.00 2.12 There are two negative attributes in Rule 3, namely, advanced maternal => or-above, one- birthweight} age (35 years and above) and a history of premature labor. However, the previous- rule shows that a co-occurrence of these negative factors with two premature- positive factors, namely, ANC visit(s) in the first trimester and a good labor} pre-pregnancy weight, often results to a normal birthweight. 4 {mother-below- 40kg, no-ANC- Though the experimental dataset consists of several maternal attri- visit, butes, it could be deduced that uterine irritability, low pre-pregnancy no-previous- => {low- 0.01 1.00 1.89 weight, and lack of ANC visits are the most frequent negative factors premature- birthweight} that often co-occur with other attributes to influence low birthweight. labor} Pregnant mothers and maternal care givers ought to note the frequently 4 D.D. Atsa’am et al. J o u r n a l o f N e o n a t a l N u rsing xxx (xxxx) xxx occurring negative attributes that often result to an unwanted Blencowe, H., Krasevec, J., de Onis, M., Black, R.E., An, X., Stevens, G.A., Borghi, E., birthweight. Hayashi, C., Estevez, D., Cegolon, L., Shiekh, S., Ponce Hardy, V., Lawn, J.E., Cousens, S., 2019. National, regional, and worldwide estimates of low birth weight in 2015, with trends from 2000: a systematic analysis. Lancet Global Health 7 (7), 4.1. Limitations of the study e849–e860. https://doi.org/10.1016/S2214-109X(18)30565-5. Cliver, S.P., Goldenberg, R.L., Cutter, G.R., Hoffman, H.J., Davis, R.O., Nelson, K.G., 1995. The effect of cigarette smoking on neonatal anthropometric measurements. The findings of the present study are limited to the eight risk factors Obstet. Gynecol. 85 (4), 625–630. contained in the secondary data at our disposal. Furthermore, the Cnattingius, S., 2004. The epidemiology of smoking during pregnancy: smoking prevalence, maternal characteristics, and pregnancy outcomes. Nicotine Tob. Res. 6 sample size consists of only 189 observations on some pregnant women (Suppl. l_2), S125–S140. who attended antenatal care at a medical facility in Springfield, Illinois, DeMarco, N., Twynstra, J., Ospina, M.B., Darrington, M., Whippey, C., Seabrook, J.A., USA. It is not clear whether or not the findings could be generalized for 2021. Prevalence of low birth weight, premature birth, and stillbirth among pregnant adolescents in Canada: a systematic review and meta-analysis. J. Pediatr. all pregnant mothers from other regions of the world. These limitations Adolesc. Gynecol. 34 (4), 530–537. https://doi.org/10.1016/J.JPAG.2021.03.003. set the pace for future research with data that encompass a larger set of Erasun, D., Alonso-Molero, J., Gómez-Acebo, I., Dierssen-Sotos, T., Llorca, J., maternal factors across participants of diverse origins. Schneider, J., 2021. Low birth weight trends in Organisation for Economic Co- operation and Development countries, 2000–2015: economic, health system and demographic conditionings. BMC Pregnancy Childbirth 21 (1). https://doi.org/ 5. Conclusion 10.1186/s12884-020-03484-9. Foster, J.P., Cole, M.J., Seth, R., 2016. Oral immunoglobulin for preventing necrotizing enterocolitis in preterm and low birth weight neonates. In: Cochrane Database of Association rules mining was deployed on some existing maternity Systematic Reviews (Vol. 2016, Issue 4). John Wiley and Sons Ltd. https://doi.org/ data to extract rules that show the maternal attributes that frequently co- 10.1002/14651858.CD001816.pub3. occur and influence birthweight. The extracted rules show that a Fuchs, F., Monet, B., Ducruet, T., Chaillet, N., Audibert, F., 2018. Effect of maternal age on the risk of preterm birth: a large cohort study. PLoS One 13 (1), e0191002. mother’s pre-pregnancy weight, presence/absence of uterine irritability, https://doi.org/10.1371/journal.pone.0191002. attendance to antenatal care visits in the first trimester, history of hy- Ghaemmaghami, S.J., Nikniaz, L., Mahdavi, R., Nikniaz, Z., Razmifard, F., Afsharnia, F., pertension, age, and history of premature labor co-occur in different 2013. Effects of infants’ birth order, maternal age, and socio-economic status on patterns that lead to either a low or normal neonatal birthweight. These birth weight. Saudi Med. J. 34 (9), 949–953. PMID: 24043008. Gladstone, M.E., Salim, N., Ogillo, K., Shamba, D., Gore-Langton, G.R., Day, L.T., rules can be used as a reference tool in the implementation of programs Blencowe, H., Lawn, J.E., Shabani, J., Shirima, K., Tarimo, M.N., Mbaruku, G., and interventions for pregnant mothers to ensure positive pregnancy Masanja, H., Ruysen, H., Peven, K., Gordeev, V.S., Boggs, D., Kong, S., Baschieri, A., outcomes related to birthweight. There is the need for medical practi- et al., 2021. Birth weight measurement processes and perceived value: qualitative research in one EN-BIRTH study hospital in Tanzania. BMC Pregnancy Childbirth 21 tioners to educate pregnant mothers on how their attributes affect (1), 1–11. https://doi.org/10.1186/S12884-020-03356-2/FIGURES/2. neonatal birthweight. Further, the findings of this study should remind Goh, D.H., Ang, R.P., 2007. An introduction to association rule mining: an application in pregnant mothers that even when there are some positive attributes to counseling and help-seeking behavior of adolescents. Behav. Res. Methods 39 (2), 259–266. their credit, a co-occurrence with one or more specific negative attri- Goldenberg, R.L., Culhane, J.F., 2007. Low birth weight in the United States. Am. J. Clin. butes could negatively influence birthweight. The findings are based on Nutr. 85 (2), 584S–590S. some secondary data from a medical facility in the USA where the Groothuis, J.R., Makari, D., 2012. Definition and outpatient management of the very low-birth-weight infant with bronchopulmonary dysplasia. Adv. Ther. 29 (4), participants were limited to certain demographics. A similar study is 297–311. https://doi.org/10.1007/s12325-012-0015-y. required on some data with a more representative demographics across Hilaire, M., Andrianou, X.D., Lenglet, A., Ariti, C., Charles, K., Buitenhuis, S., Van participants. Brusselen, D., Roggeveen, H., Ledger, E., Denat, R.S., Bryson, L., 2021. Growth and neurodevelopment in low birth weight versus normal birth weight infants from birth to 24 months, born in an obstetric emergency hospital in Haiti, a prospective cohort Funding study. BMC Pediatr. 21 (1), 1–16. https://doi.org/10.1186/S12887-021-02605-3/ FIGURES/8. Hoodbhoy, Z., Mohammed, N., Rozi, S., Aslam, N., Mohsin, S., Ashiqali, S., Ali, H., No funding to declare. Sattar, S., Chowdhury, D., Hasan, B.S., 2021. Cardiovascular dysfunction in children exposed to preeclampsia during fetal life. J. Am. Soc. Echocardiogr. 34 (6), 653–661. https://doi.org/10.1016/J.ECHO.2021.01.008. Declaration of competing interest Khan, N., Jamal, M., 2003. Maternal risk factors associated with low birth weight. J. Coll. Phys. Surg. Pakistan: JCPSP 13 (1), 25–28. None. Kono, Y., 2021. Neurodevelopmental outcomes of very low birth weight infants in the neonatal research network of Japan: importance of neonatal intensive care unit graduate follow-up. Clin. Exp. Pediatr. 64 (7), 313–321. https://doi.org/10.3345/ References cep.2020.01312. Lange, S., Probst, C., Rehm, J., Popova, S., 2018. National, regional, and global ACOG, 2020. Gestational hypertension and preeclampsia: ACOG practice bulletin, prevalence of smoking during pregnancy in the general population: a systematic number 222. Obstet. Gynecol. 135 (6), e237–e260. https://doi.org/10.1097/ review and meta-analysis. Lancet Global Health 6 (7), e769–e776. https://doi.org/ AOG.0000000000003891. 10.1016/S2214-109X(18)30223-7. Afaya, A., Afaya, R.A., Azongo, T.B., Yakong, V.N., Konlan, K.D., Agbinku, E., Agyabeng- Macedo, T.C.C., Montagna, E., Trevisan, C.M., Zaia, V., de Oliveira, R., Barbosa, C.P., Fandoh, E., Akokre, R., Karim, J.F., Salia, S.M., Kaba, R.A., Ayanore, M.A., 2021. Laganà, A.S., Bianco, B., 2020. Prevalence of preeclampsia and eclampsia in Maternal risk factors and neonatal outcomes associated with low birth weight in a adolescent pregnancy: a systematic review and meta-analysis of 291,247 adolescents secondary referral hospital in Ghana. Heliyon 7 (5), e06962. worldwide since 1969. Eur. J. Obstet. Gynecol. Reprod. Biol. 248, 177–186. https:// Atsa’am, D.D., Gbaden, T., Wario, R., 2022. Association rules on attributes of illicit doi.org/10.1016/j.ejogrb.2020.03.043. drugs, suspect s demographics and offence categories. J. Drug Issues. https://doi. Metgud, C.S., Naik, V.A., Mallapur, M.D., 2012. Factors affecting birth weight of a ’ org/10.1177/00220426221140010. newborn–a community based study in rural Karnataka, India. PLoS One 7 (7), Atsa’am, D.D., Wario, R., 2022a. Association rules on the COVID-19 variants of concern e40040. to guide choices of tourism destinations. Curr. Issues Tourism 25 (10), 1536–1540. MFMER, 2022. Pregnancy Weight Gain: What’s Healthy? https://www.mayoclinic.org/h https://doi.org/10.1080/13683500.2021.1951182. ealthy-lifestyle/pregnancy-week-by-week/in-depth/pregnancy-weight-gain/art Atsa’am, D.D., Wario, R., 2022b. Association rules extraction from the coronavirus -20044360. disease 2019: attributes on morbidity and mortality. Int. J. Healthc. Inf. Syst. Inf. 17 Ødegård, R.A., Vatten, L.J., Nilsen, S.T., Salvesen, K.Å., Austgulen, R., 2000. (1), 1–10. https://doi.org/10.4018/IJHISI.302652. Preeclampsia and fetal growth. Obstet. Gynecol. 96 (6), 950–955. https://doi.org/ Atsa’am, D.D., Wario, R., Balogun, O.S., 2022. Antenatal care visits: a moderator of the 10.1016/S0029-7844(00)01040-1. association between a mother’s age and the neonate’s birth weight. Int. J. Childbirth Olaleye, S.A., Ukpabi, D.C., Olawumi, O., Atsa’am, D.D., Agjei, R.O., Oyelere, S.S., 12 (1), 15–22. https://doi.org/10.1891/IJC-2021-0019. Sanusi, I.T., Agbo, F.J., Balogun, O.S., Gbadegeshin, S.A., Adegbite, A., Kolog, E.A., Balogun, O.S., Atsa’am, D.D., Akingbade, T.J., Kolog, E.A., Agjei, R.O., Devine, S.N.O., 2022. Association rule mining for job seekers’ profiles based on personality traits 2020. Modeling the effect of mother s characteristics on the weight of a newborn. and Facebook usage. Int. J. Bus. Inf. Syst. 40 (3), 299–326. ’ Int. J. Childbirth 10 (4), 206–216. Senthilkumar, D., Paulraj, S., 2015. Prediction of low birth weight infants and its risk Bernstein, I.M., Mongeon, J.A., Badger, G.J., Solomon, L., Heil, S.H., Higgins, S.T., 2005. factors using data mining techniques. March. In: Proceedings of the 2015 Maternal smoking and its association with birth weight. Obstet. Gynecol. 106 (5 Part International Conference on Industrial Engineering and Operations Management, 1), 986–991. pp. 186–194. 5 D.D. Atsa’am et al. J o u r n a l o f N e o n a t a l N u rsing xxx (xxxx) xxx Shin, Y.H., Choi, S.J., Kim, K.W., Yu, J., Ahn, K.M., Kim, H.Y., et al., 2013. Association Wang, S., Yang, L., Shang, L., Yang, W., Qi, C., Huang, L., Xie, G., Wang, R., Chung, M.C., between maternal characteristics and neonatal birth weight in a Korean population 2020. Changing trends of birth weight with maternal age: a cross-sectional study in living in the Seoul metropolitan area, Korea: a birth cohort study (COCOA). J. Kor. Xi’an city of Northwestern China. BMC Pregnancy Childbirth 20 (1), 1–8. Med. Sci. 28 (4), 580–585. WHO, 2011. Guidelines on Optimal Feeding of Low Birth-Weight Infants in Low- and Simonsen, S.E., Lyon, J.L., Stanford, J.B., Porucznik, C.A., Esplin, M.S., Varner, M.W., Middle-Income Countries. World Health Organization. https://www.who.int/publi 2013. Risk factors for recurrent preterm birth in multiparous Utah women: a cations/i/item/9789241548366. historical cohort study. BJOG An Int. J. Obstet. Gynaecol. 120 (7), 863–872. https:// WHO, 2016. New Guidelines on Antenatal Care for a Positive Pregnancy Experience. doi.org/10.1111/1471-0528.12182. November 7. World Health Organization. https://www.who.int/news/item/07-11- Singh, G., Chouhan, R., Sidhu, K., 2009. Maternal factors for low birth weight babies. 2016-new-guidelines-on-antenatal-care-for-a-positive-pregnancy-experience. Med. J. Armed Forces India 65 (1), 10–12. WHO, 2020. Adolescent Pregnancy. January 31. World Health Organization. htt Steer, P.J., Little, M.P., Kold-Jensen, T., Chapple, J., Elliott, P., 2004. Maternal blood ps://www.who.int/news-room/fact-sheets/detail/adolescent-pregnancy. pressure in pregnancy, birth weight, and perinatal mortality in first births: WHO, 2022a. ICD-11 for Mortality and Morbidity Statistics. World Health Organization. prospective study. BMJ 329 (7478), 1312. https://icd.who.int/browse11/l-m/en#/http://id.who.int/icd/entity/153712793. Spong, C.Y., 2007. Prediction and prevention of recurrent spontaneous preterm birth. WHO, 2022b. ICD-11 for Mortality and Morbidity Statistics. World Health Organization. Obstet. Gynecol. 110 (2), 405–415. https://doi.org/10.1097/01. https://icd.who.int/browse11/l-m/en#/http://id.who.int/icd/entity/2041060050. AOG.0000275287.08520.4a. Woelile, T.A., Kibret, G.T., Workie, H.M., Amare, A.T., Tigabu, A., Aynalem, Y.A., Sutan, R., Mohtar, M., Mahat, A.N., Tamil, A.M., 2014. Determinant of low birth weight Chanie, E.S., Birlie, T.A., 2021. Survival status and predictors of mortality among infants: a matched case control study. Open J. Prev. Med. 4 (3), 91–99. https://doi. low-birth-weight neonates admitted to the neonatal intensive care unit at felege org/10.4236/ojpm.2014.43013. hiwot comprehensive specialized hospital, Bahir dar, Ethiopia, 2020. Pediatr. Health Tela, F.G., Bezabih, A.M., Adhanu, A.K., 2019. Effect of pregnancy weight gain on infant Med. Therapeut. 12, 451. https://doi.org/10.2147/PHMT.S323526. birth weight among mothers attending antenatal care from private clinics in Mekelle Xi, C., Luo, M., Wang, T., Wang, Y., Wang, S., Guo, L., Lu, C., 2020. Association between City, Northern Ethiopia: a facility based follow-up study. PLoS One 14 (3), maternal lifestyle factors and low birth weight in preterm and term births: a case- e0212424. https://doi.org/10.1371/journal.pone.0212424. control study. Reprod. Health 17 (1), 1–9. Tessema, Z.T., Tamirat, K.S., Teshale, A.B., Tesema, G.A., 2021. Prevalence of low birth Xiao, R., Sorensen, T.K., Williams, M.A., Luthy, D.A., 2003. Influence of pre-eclampsia on weight and its associated factor at birth in Sub-Saharan Africa: a generalized linear fetal growth. J. Matern. Fetal Neonatal Med. 13 (3), 157–162. https://doi.org/ mixed model. PLoS One 16 (3), e0248417. https://doi.org/10.1371/JOURNAL. 10.1080/jmf.13.3.157.162. PONE.0248417. Yang, J., Baer, R.J., Berghella, V., Chambers, C., Chung, P., Coker, T., Currier, R.J., UNICEF-WHO, 2020. UNICEF-WHO Low birth weight estimates. In: Definitions. Druzin, M.L., Kuppermann, M., Muglia, L.J., Norton, M.E., Rand, L., Ryckman, K., https://www.unicef.org/media/96976/file/UNICEF-WHO-Low-Birth weight-esti Shaw, G.M., Stevenson, D., Jelliffe-Pawlowski, L.L., 2016. Recurrence of preterm mates-2000-2015.pdf. birth and early term birth. Obstet. Gynecol. 128 (2), 364–372. https://doi.org/ UNICEF, 2018. Neonatal care clinical guidelines. In: Ministry of Health Eswatini. https:// 10.1097/AOG.0000000000001506. www.unicef.org/eswatini/media/631/file/UNICEF-Sd-Neonatal-Guidelines-report Zhao, Q., Bhowmick, S.S., 2003. Association Rule Mining: A Survey. Nanyang -2018.pdf. Technological University, Singapore, p. 135. Venables, W.N., Ripley, B.D., 2002. Modern Applied Statistics with S, 4th ed. Springer, Zhou, L., Yau, S., 2007. Efficient association rule mining among both frequent and New York. infrequent items. Comput. Math. Appl. 54 (6), 737–749. Waits, A., Guo, C.Y., Chien, L.Y., 2021. Inadequate gestational weight gain contributes to Zimmerman, E., 2018. Do infants born very premature and who have very low birth increasing rates of low birth weight in Taiwan: 2011–2016 nationwide surveys. weight catch up with their full term peers in their language abilities by early school Taiwan. J. Obstet. Gynecol. 60 (5), 857–862. https://doi.org/10.1016/J. age? J. Speech Lang. Hear. Res. 61 (1), 53–65. https://doi.org/10.1044/2017_ TJOG.2021.07.013. JSLHR-L-16-0150. 6