Development and internal validation of the NOTICE assessment as a non-biological measure of prenatal nicotine, tobacco, and cannabis use and exposure.

OBJECTIVES: Accurate measurement of tobacco, nicotine and cannabis use and exposure during pregnancy is essential for research and clinical care, yet existing perinatal surveillance relies on simplified self-report that inadequately captures exposure patterns. Biological assays provide objective measurement but are invasive, costly, limited to short detection windows, and impractical for repeated or population-level use. This study aimed to develop and internally validate the Nicotine, Tobacco and Cannabis Use and Exposure (NOTICE) assessment as a low-burden, non-biological behavioural tool for use and exposure during pregnancy.

STUDY DESIGN AND SETTING: In a longitudinal cohort of Australian First Nations pregnant women, repeated NOTICE assessments were collected across pregnancy alongside exhaled carbon monoxide, saliva cotinine, and urinary free cotinine and total 11-Nor-9-carboxy-Δ9-tetrahydrocannabinol (THC-COOH), with and without creatinine correction. Weekly self-use and exposure scores were combined into a weighted, standardised composite score. Discriminative validity relative to biochemical reference standards was assessed using receiver operating characteristic analysis. Area under the curve (AUC) estimates with bias-corrected and accelerated bootstrap confidence intervals clustered by participant were calculated. Population-averaged associations were examined using generalised estimating equations.

RESULTS: The composite NOTICE score demonstrated very high discrimination for tobacco and nicotine exposure relative to creatinine-corrected cotinine ≥5 ng/mg cr (AUC = 0.99; 95% CI 0.98-1.00) and high discrimination for cannabis exposure relative to creatinine-corrected THC-COOH ≥80 ng/mg cr (AUC = 0.92; 95% CI 0.83-1.00). Discrimination was stable across alternative biomarker thresholds and consistently superior to exposure-only measures. Weekly scores demonstrated better discriminative performance than daily scores.

CONCLUSION: NOTICE is a low-burden behavioural assessment that demonstrates strong criterion-based alignment with established biochemical reference standards for measuring tobacco, nicotine and cannabis use and exposure during pregnancy. By integrating active use and environmental exposure across products and timeframes, the tool supports repeated measurement and surveillance of exposure patterns in settings where routine biological testing is impractical.