Background
The first few days after childbirth are a critical period for establishing successful breastfeeding. During this time, healthcare providers have a limited opportunity to assess how well mothers and newborns are adapting to breastfeeding before they leave the hospital.
Although many mothers initiate breastfeeding successfully, some experience challenges that may lead to early discontinuation of exclusive breastfeeding after discharge. Identifying these mother-infant pairs before they leave the hospital would allow healthcare providers to offer additional education, lactation support, and follow-up care.
The LATCH breastfeeding assessment tool was developed as a standardized method for evaluating breastfeeding effectiveness. However, before it can be used to guide clinical decisions, researchers must determine whether early LATCH scores can accurately predict future breastfeeding outcomes. This study sought to evaluate the predictive value of pre-discharge LATCH scores for exclusive breastfeeding at six weeks postpartum among term mother-infant dyads.
The Research Challenge
The research team sought to determine whether pre-discharge LATCH scores could predict exclusive breastfeeding at six weeks postpartum and identify the optimal cutoff score for clinical use.
The study explored questions such as:
- Can a bedside breastfeeding assessment accurately predict breastfeeding success after hospital discharge?
- What LATCH score best distinguishes mothers who are likely to exclusively breastfeed from those who may require additional support?
- How accurate is the LATCH scoring system in predicting future breastfeeding outcomes?
- Which statistical measures best evaluate the performance of a clinical screening tool?
Beyond evaluating a single assessment tool, the project addressed a broader challenge frequently encountered in clinical research: how can researchers determine whether a screening instrument is accurate enough to support evidence-based patient care?
Data2Stats’ Role
Data2Stats provided comprehensive statistical support throughout the project, helping transform clinical assessment data into meaningful evidence that can improve breastfeeding support and maternal care.
The team supported the study through:
- Research consultation and statistical analysis planning;
- Sample size verification;
- Data validation, quality assessment, and coding guidance;
- Statistical analysis and interpretation;
- Receiver Operating Characteristic (ROC) curve analysis;
- Determination of the optimal cutoff using the Youden Index;
- Estimation of sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV);
- Preparation of publication-ready statistical tables; and
- Technical review and statistical reporting support.
By ensuring that the study applied appropriate diagnostic performance analyses, the project was designed to generate reliable evidence on the clinical utility of the LATCH assessment tool.
Statistical Methods Used
To answer the research questions, several statistical methods were employed.
Descriptive statistics were used to summarize participant characteristics and breastfeeding outcomes among mother-infant dyads.
Independent Samples t-tests and Mann-Whitney U tests were conducted to compare breastfeeding outcomes across participant groups, depending on data distribution.
To evaluate the predictive performance of the LATCH scoring system, Receiver Operating Characteristic (ROC) curve analysis was performed. The Area Under the Curve (AUC) was used to measure the tool’s ability to distinguish between successful and unsuccessful exclusive breastfeeding outcomes.
The Youden Index was then applied to identify the optimal LATCH score cutoff that balanced sensitivity and specificity, while additional diagnostic measures – including positive predictive value (PPV) and negative predictive value (NPV) – were calculated to assess the tool’s overall clinical performance.
What the Data Revealed
The study protocol has been developed to evaluate whether pre-discharge LATCH scores can predict exclusive breastfeeding at six weeks postpartum among term mother-infant pairs.
Upon completion, the planned analyses are expected to determine the diagnostic performance of the LATCH assessment tool by estimating its sensitivity, specificity, predictive values, and overall discriminatory ability.
The study also aims to identify an optimal LATCH score cutoff that can help healthcare providers recognize mothers and infants who may require additional breastfeeding support before hospital discharge.
If validated, the findings could provide valuable local evidence supporting the use of structured breastfeeding assessments to strengthen discharge planning, improve lactation counseling, and promote successful exclusive breastfeeding practices.
A Valuable Lesson in Clinical Research
Not every clinical assessment tool should be adopted simply because it is easy to use.
Before a screening tool becomes part of routine patient care, researchers must determine whether it accurately predicts the outcomes it is intended to measure.
This study highlights an important lesson in clinical research:
A useful screening tool is one that demonstrates both clinical relevance and statistical accuracy.
Measures such as sensitivity, specificity, positive predictive value, negative predictive value, and the Area Under the ROC Curve provide researchers with evidence about how well a tool performs under real clinical conditions.
By combining these statistical methods, researchers can determine whether an assessment tool is reliable enough to guide patient care, identify individuals who may benefit from additional support, and strengthen evidence-based clinical practice.
Outcome and Impact
The study protocol has been developed to evaluate the predictive value of pre-discharge LATCH scores for exclusive breastfeeding among term mother-infant dyads.
More importantly, the project demonstrates the value of rigorous statistical support in helping researchers:
- Evaluate the diagnostic performance of clinical assessment tools;
- Identify evidence-based cutoff scores for patient screening;
- Translate statistical findings into meaningful clinical recommendations; and
- Support earlier identification of mothers and infants who may benefit from additional breastfeeding assistance.
Upon completion, the findings are expected to contribute local evidence that strengthens breastfeeding support programs, improves discharge planning, and promotes evidence-based maternal and newborn care.
Key Takeaways
- Early postpartum assessments may help identify mother-infant pairs who need additional breastfeeding support.
- ROC curve analysis provides an objective way to evaluate the performance of clinical screening tools.
- Sensitivity, specificity, PPV, NPV, and AUC are essential measures for assessing diagnostic accuracy.
- Determining an evidence-based cutoff score improves the clinical usefulness of assessment tools.
- Rigorous statistical evaluation helps ensure that screening instruments support better maternal and newborn healthcare decisions.
At Data2Stats, we believe that statistical analysis is about more than evaluating numbers. It is about validating clinical tools that help healthcare professionals make informed decisions, deliver timely interventions, and improve health outcomes for mothers and newborns through evidence-based practice.
