Author Identifier

Annie Yen-Hua Chang: https://orcid.org/0000-0002-9342-064X  

Date of Award

2026

Keywords

Neonatal Abstinence Syndrome, NAS, neonates, opioid exposed infants, substance exposed infants, Finnegan Neonatal Abstinence Scoring Tool, FNAS, Modified Finnegan Score, mFNAS, Lipsitz tool, Neonatal Withdrawal Inventory, NWI, clinical assessment tools, psychometric properties, validity, reliability, inter-rater reliability, internal consistency

Document Type

Thesis

Publisher

Edith Cowan University

Degree Name

Master of Nursing (Research)

School

School of Nursing and Midwifery

First Supervisor

Maggie Zgambo

Second Supervisor

Linda Coventry

Third Supervisor

Lukman Thalib

Fourth Supervisor

Evalotte Morelius

Abstract

Background: Neonatal Abstinence Syndrome (NAS) is a withdrawal condition affecting newborns exposed to opioids and other substances in utero. Accurate assessment is essential for guiding clinical management; however, existing NAS assessment instruments vary widely in reliability, validity, and feasibility. This research has two aims: 1) to systematically evaluate the psychometric properties of existing NAS assessment instruments using the COnsensus based Standards for the selection of Health Measurement Instruments (COSMIN) risk of bias checklist; and 2) to validate the reliability, validity, and feasibility of the Simplified Finnegan Neonatal Abstinence Scoring Tool (sFNAST), an 8-item instrument with binary scoring, within an acute neonatal clinical setting in Western Australia.

Methods: A COSMIN-guided systematic review pathway was conducted to appraise the published NAS instruments against the COSMIN risk of bias checklist and COSMIN 10 measurement property criteria. A prospective observational validation study was undertaken in a tertiary neonatal intensive care unit (NICU) in Western Australia, to evaluate the properties of the 8 item instrument (sFNAST), which had been modified from the original 21-item Finnegan’s Neonatal Abstinence Scoring System (mFNAS). The study involved three methods: 1) a test retest reliability assessment by 52 registered nurses using standardised two video-based scenarios scored two weeks apart; 2) an inter-rater reliability assessment based on 100 set paired bedside scoring observations, performed by nurses on duty, comparing the 8-item sFNAST with the 21-item modified Finnegan (mFNAS); and 3) a feasibility and usability questionnaire completed by the same 52 nurses. Data were analysed using Cohen’s kappa (k), the Prevalence-Adjusted and Bias-Adjusted kappa (PABAK), Spearman’s (1904) correlation coefficient, and Cronbach’s (1951) alpha (α).

Results: The systematic review identified 12 NAS instruments. Most demonstrate low-quality evidence crossed key psychometric domains, particularly development, content validity, and construct validity, highlighting significant limitations in existing instruments. Two instruments: the new NAS (Kocherlakota et al., 2020) and the sFNAST (Devlin et al., 2020) achieved moderate-quality evidence (Level B recommendation). The observational validation study evaluated the psychometric properties of the sFNAST and demonstrated substantial to almost perfect test–retest reliability for most items (κ = 0.76–1.00) and strong inter-rater reliability (κ = 0.77–0.96). To assess test-retest reliability, video-based scenarios were used: one depicted moderate NAS (scenario 1) and another depicted severe NAS (scenario 2). Greater variability was observed in the severe NAS scenario (κ = 0.32–1.00), particularly for behaviourally expressed signs such as muscle tone and tremors, which can be interpreted differently by observers. In contrast, the moderate NAS scenario showed substantial reliability (κ = 0.76–1.00).

In addition, total scores for the sFNAST and mFNAS were strongly correlated (rs = 0.862, p < 0.001). Cut-off scores of 3 to 4 and 5 on the sFNAST corresponded best with mFNAS scores of 8 to 11 and 12, respectively, yielding the highest agreement. Internal consistency was moderate (Cronbach’s α = 0.55–0.64). More than 90% of nurses rated the sFNAST as easy to use, time-efficient, and clinically relevant.

Conclusion: This study demonstrated that the 8-item sFNAST has acceptable reliability, construct validity, feasibility, and usability in a tertiary NICU setting, supporting its potential for standardised NAS assessment. Further multi-centre research is needed to confirm generalisability, examine performance across diverse neonatal populations and settings, and establish validity for key clinical outcomes.

Access Note

Access to this thesis is embargoed until 25th October 2027 

Available for download on Monday, October 25, 2027

Share

 
COinS
 

Link to publisher version (DOI)

10.25958/hak6-tk43