Predicting exclusive breastfeeding in maternity wards using machine learning techniques

Predicting mothers' exclusive breastfeeding for the first 6 months: Interface creation study using machine learning technique

BACKGROUND

Machine learning techniques (MLT) build models to detect complex patterns and solve new problems using big data.

AIM

The present study aims to create a prediction interface for mothers breastfeeding exclusively for the first 6 months using MLT.

METHOD

All mothers who had babies aged 6-24 months between 15.09.2021 and 15.12.2021 and to whom the surveys could be delivered were included. 'Personal Information Form' created by the researchers was used as a data collection tool. Data from 514 mothers participating in the study were used for MLT. Data from 70% of mothers were used for educational purposes, and a prediction model was created. The data obtained from the remaining 30% of the mothers were used for testing.

RESULTS

The best MLT algorithm for predicting exclusive breastfeeding for the first 6 months was determined to be the Random Forest Classifier. The top five variables affecting the possibility of mothers breastfeeding exclusively for the first 6 months were as follows: "the mother not having any health problems during pregnancy," "there were no people who negatively affected the mother's morale about breastfeeding," "the amount of water the mother drinks in a day," "thinking that her milk supply is insufficient," "having no problems breastfeeding the baby".

CONCLUSIONS

Using created prediction model may allow early identification of mothers with a risk of not breastfeeding their babies exclusively for the first 6 months. In this way, mothers in the risk group can be closely monitored in the early period.

The impact of the Baby-Friendly Hospital Initiative on breastfeeding rates at maternity units in France

BACKGROUND

The Baby-Friendly Hospital Initiative (BFHI) is associated with improved breastfeeding outcomes in many high-income countries including the UK and the USA, but its effectiveness has never been evaluated in France. We investigated the impact of the BFHI on breastfeeding rates in French maternity units in 2010, 2016 and 2021 to assess if the BFHI aids to reduce inequalities in breastfeeding.

METHODS

We examined breastfeeding in maternity units (exclusive, mixed and any breastfeeding) in mothers of singleton full-term newborns using the 2010 (n = 13 075), 2016 (n = 10 919) and 2021 (n = 10 209) French National Perinatal Surveys. We used mixed-effect hierarchical multinomial regression models adjusting for neonatal, maternal, maternity unit and French administrative department characteristics, and tested certain interactions.

RESULTS

The adjusted rate of exclusive breastfeeding was higher by +5.8 (3.4-8.1) points among mothers delivering in BFHI-accredited maternity units compared with those delivering in non-accredited units. When compared with average-weight newborns, this difference was sharper for infants with low birthweight: +14.9 (10.0-19.9) points when their birthweight was 2500 g. Mixed breastfeeding was lower by -1.7 points (-3.2-0) in BFHI-accredited hospitals, with no notable difference according to the neonatal or maternal characteristics.

CONCLUSION

Mothers delivering in BFHI-accredited maternity units had higher exclusive breastfeeding rates and lower mixed breastfeeding rates than those delivering in non-accredited maternity units. The positive impact of the BFHI was stronger among low-birthweight neonates, who are less often breastfed, helping reduce the gap for this vulnerable group while favouring mothers with higher education levels.

Partners with a Highly Favorable Attitude Toward Breastfeeding Contribute to Promoting Initiation and Length of Breastfeeding for More than 5 Months in a Population-Based Canadian Study

Introduction: Promoting maternal breastfeeding for at least 6 months is important to improve children physical health during infancy and neurodevelopmental outcomes in childhood. For this guideline to be followed, it is paramount to identify what factors best support the initiation and length of breastfeeding. This study estimates the contribution of various child- and parent-level factors, as well as the sociodemographic context in predicting maternal breastfeeding initiation and duration. Methods: This study draws on data from the Quebec Longitudinal Study of Child Development. Mothers who never breastfed when the infant was 5 months old (n = 630, 28.3%) were compared to mothers who breastfed for less than 5 months (n = 844, 38.0%) and mothers breastfeeding for more than 5 months (n = 749, 33.7%), using multivariable multinomial regression models. Results: Mothers with a partner showing a positive attitude toward breastfeeding were up to 13 times more likely to breastfeed their infant for more than 5 months. The positive attitude of partners toward breastfeeding was the strongest predictor of breastfeeding duration, followed by the maternal educational attainment and timing she returns to work. Most prenatal and perinatal child-level factors and the sociodemographic context predicted breastfeeding duration, but to a lesser extent. Discussion: This finding underscores the role of the partner's attitude in promoting initiation and length of breastfeeding. As such, educational campaigns and health practitioners could target both the mother and their partner in promoting breastfeeding.

An Exaggeration? Reality?: Can ChatGPT Be Used in Neonatal Nursing?

Artificial intelligence (AI) represents a system endowed with the ability to derive meaningful inferences from a diverse array of datasets. Rooted in the advancements of machine learning models, AI has spawned various transformative technologies such as deep learning, natural language processing, computer vision, and robotics. This technological evolution is poised to witness a broadened spectrum of applications across diverse domains, with a particular focus on revolutionizing healthcare services. Noteworthy among these innovations is OpenAI's creation, ChatGPT, which stands out for its profound capabilities in intricate analysis, primarily facilitated through extensive language modeling. In the realm of healthcare, AI applications, including ChatGPT, have showcased promising outcomes, especially in the domain of neonatal nursing. Areas such as pain assessment, feeding processes, and patient status determination have witnessed substantial enhancements through the integration of AI technologies. However, it is crucial to approach the deployment of such applications with a judicious mindset. The accuracy of the underlying data must undergo rigorous validation, and any results lacking a solid foundation in scientific insights should be approached with skepticism. The paramount consideration remains patient safety, necessitating that AI applications, like ChatGPT, undergo thorough scrutiny through controlled and evidence-based studies. Only through such meticulous evaluation can the transformative potential of AI be harnessed responsibly, ensuring its alignment with the highest standards of healthcare practice.

Introduction of Breast Milk Substitutes During the First 3 Days of Life: Results of the Mexican National Survey of Demographic Dynamics, 2018

Background: The introduction of foods or fluids other than breast milk in the first few days after birth interferes with the establishment of breastfeeding. This study aimed to investigate the association of formula introduction during the first 3 days of life with maternal sociodemographic characteristics, hospital practices, and breastfeeding duration. Materials and Methods: Information from the National Survey of Demographic Dynamics, 2018, which includes 17,686 mother-baby pairs was analyzed. Mother-baby pairs were classified into categories according to breastfeeding duration: <5 months and ≥5 months. Statistical methods and a machine learning algorithm (Bayesian network, BN) were used to analyze the data. Results: In general, 3,720 (21%) mothers reported introducing formula during the first 3 days of life. A lower education level, lower sociodemographic stratum, living in a rural area, and considering oneself indigenous were factors associated with not introducing formula during the first 3 days of life. A total of 5,168 (29.2%) mother-baby pairs practiced breastfeeding for <5 months, and 12,518 (70.8%) for ≥5 months. Almost twice as many mothers who practiced breastfeeding for <5 months introduced formula during the first 3 days of life (31.7%) compared with those who practiced breastfeeding for ≥5 months (16.6%). The BN model can sufficiently predict cases with a breastfeeding duration ≥5 months (precision-recall curve area = 0.792). Discussion: Introducing formula during the first 3 days of life was associated with a shorter breastfeeding duration. BN analysis showed a probabilistic dependency between the type of delivery and variables associated with the establishment of breastfeeding.

[Breastfeeding in Spain and the factors related to its establishment and maintenance: LAyDI Study (PAPenRed)]

OBJECTIVE

To find out actual statistics on breastfeeding in Spain, as well as sociocultural and perinatal factors that affect its initiation and maintenance.

DESIGN

Prospective, multicentre, longitudinal, nationwide study (XXX study). SITE: Primary care paediatricians' office.

PARTICIPANTS

Cohort of newborns born between April 2017 and March 2018 in Spain who were followed up to two years of age in 8 visits.

MAIN MEASURES

Rates of different types of breastfeeding were analysed at each visit and variables related to gestation, delivery, neonatal period, social, economic and biological variables were also analysed.

RESULTS

Initial sample of 1946 (50.1% male). 90.7% decided to initiate breastfeeding at birth. Exclusive breastfeeding (EBF) was 66.4% at 15days and 35.2% at 6months. Any type of breastfeeding (total breastfeeding [TBF]) at 6months was 61.7%. Median survival from TBF was 6.0months (95%CI: 6.0-6.1). Variables related to EBF at 15days: previous children, mother's level of education, absence of illness during pregnancy, no separation of mother and child at birth, no dummy use, no nipple problems, and time of decision to breastfeed. Variables related to longer duration of TBF: relationship of parents older than 5years, no dummy use, co-sleeping at one month of life, deciding to breastfeed before pregnancy, receiving information on breastfeeding during pregnancy and using support from associations.

CONCLUSIONS

Early abandonment of breastfeeding is a major problem in Western societies. There are factors that can be worked on to improve outcomes.

A comparison of single- and double-threshold ROC plots for mixture distributions

The receiver operating characteristics (ROC) analysis is commonly used in clinical settings to check the performance of a single threshold for distinguishing population-wise bimodal-distributed test results. However, for population-wise three-modal distributed test results, a single threshold ROC (stROC) analysis showed poor discriminative performance. The purpose of this study is to use a double-threshold ROC analysis for the three-modal distributed test results to provide better discriminative performance than the stROC analysis. A double-threshold receiver operating characteristic plot (dtROC) is constructed by replacing the single threshold with a double threshold. The sensitivity and specificity coordinates are chosen to maximize sensitivity for a given specificity value. Besides a simulation study assuming a mixture of lognormal, Poisson, and Weibull distributions, a clinical application is examined by a secondary data analysis of palpation test results of the C7 spinous process using the modified thorax-rib static technique. For the assumed mixture models, the discrimination performance of dtROC analysis outperforms the stROC analysis (area under ROC (AUROC) increased from 0.436 to 0.983 for lognormal distributed test results, 0.676 to 0.752 for the Poisson distribution, and 0.674 to 0.804 for Weibull distribution).