Validation of Parent-reported Gestational Age Categories for Children Less Than 6 Years of Age Journal Articles uri icon

  •  
  • Overview
  •  
  • Research
  •  
  • Identity
  •  
  • Additional Document Info
  •  
  • View All
  •  

abstract

  • Background: Preterm birth is an important outcome or exposure in epidemiologic research. When administrative data on measured gestational age is not available, parent-reported gestational age can be obtained from questionnaires, which is subject to potential bias. To our knowledge, few studies have assessed the validity of parent-reported gestational age categories, including commonly defined categories of preterm birth. Methods: We used linked data from primarily healthy children <6 years of age in TARGet Kids! in Toronto, Canada, and ICES administrative healthcare data from April 2011 to March 2020. We assessed the criterion validity of questionnaire-based parent-reported gestational age by calculating sensitivity and specificity for term (≥37 weeks), late preterm (34–36 weeks), and moderately preterm (32–33 weeks) gestational age categories, using administrative healthcare records of gestational age as the criterion standard. We conducted subgroup analyses for various parent and socioeconomic factors that may influence recall. Results: Of the 4684 participants, 97.3% correctly classified the gestational age category according to administrative healthcare data. Parent-reported gestational age sensitivity ranged from 83.7% to 98.5% and specificity ranged from 88.3% to 99.8%, depending on category. For each subgroup characteristic, sensitivity and specificity were all ≥70%. Lower educational attainment, lower family income, father reporting, ≥1 year since birth, ≥2 children, lower parent age, and reported gestational diabetes and/or hypertension were associated with slightly lower sensitivity and/or specificity. Conclusions: In this linked cohort, parent-reported gestational age categories had high accuracy. Criterion validity varied minimally among some parent and socioeconomic factors. Our findings can inform future quantitative bias analyses.

publication date

  • November 2023