Item-specific patterns in family physicians’ age estimates for developmental milestones and development-promoting activities: a cross-sectional study

Background Repeated contact with young children and families can offer family physicians opportunities for developmental surveillance and guidance, but evidence on their developmental timing expectations remains limited. We aimed to describe physicians’ age estimates for developmental milestones and development-promoting activities at the item level. Methods This cross-sectional survey included 310 physicians working or training in family medicine in Adana, Türkiye, during January–June 2024, recruited through voluntary convenience sampling. We described reported age distributions for each of the 20 Caregiver Knowledge of Child Development Inventory items. We also summarised response categories (earlier than, within or later than the original reference age ranges), distances from reference boundaries in months, self-reported training needs and responses after suspected delay. Associations between response categories and physician characteristics were explored using item-specific multivariable bias-reduced logistic models with Holm correction. Results Numerical age summaries included 6,181 of 6,200 responses. Median ages (first–third quartiles) were 12 (12–12) months for walking independently with good coordination, 11 (8–12) for first meaningful single words, 4 (3–6) for responding vocally to speech and 9 (3–18) for beginning to look at books together. For the latter two items, 82.9% and 51.6% of responses, respectively, were later than the original reference ranges. Across items, median distances outside the ranges were 1–6 months for earlier responses and 1–12 months for later responses. Social-emotional development was the most frequently reported training need (87.7%), and 88.8% of 240 physicians reporting previous suspicion of developmental delay reported referral to a paediatrician. In exploratory adjusted analyses, only the association between sex and responses for looking at books together remained significant after Holm correction (joint Holm-adjusted p = 0.0067; n = 294): men had higher odds of later rather than within-range responses than women (odds ratio 2.92; 95% CI 1.43–5.95). Conclusions Physicians’ age estimates showed item-specific patterns, including differences between milestones within the same developmental domain. These findings support examining timing expectations for individual milestones and development-promoting activities. Their implications for education, developmental surveillance and guidance for families require further investigation.