Longer Waits Linked to Missed Pediatric Specialist Visits in a 10-Year Singapore Audit

Scienmag reported on October 13, 2026 on a BMC Pediatrics study by Xi Yar Ong and Bao Bei Kelly Low of the General Pediatrics Service at KK Women's and Children's Hospital in Singapore ^[1]. The retrospective clinical audit covered 156,161 consecutive referrals to the hospital's General Pediatrics Specialist Outpatient Clinic between 1 January 2015 and 31 December 2024 ^[1]. Longer intervals between referral and appointment were associated with higher odds that the child never attended, and overall attendance was 63.0 percent ^[1].

What the audit found

The referred children had a median age of 4.2 years (interquartile range 1.2 to 9.5) ^[1]. Public polyclinics made 43.9 percent of referrals and the children's emergency department made 32.8 percent ^[1]. The median referral-to-appointment time was 21 days, with an interquartile range of 7 to 38 days ^[1].

The authors used multivariable logistic regression, which estimates how each factor relates to an outcome while statistically accounting for the others, and reported adjusted odds ratios (aOR) for non-attendance ^[1]. Each additional day of waiting carried an aOR of 1.018 (95 percent confidence interval 1.017 to 1.019; p less than 0.001), which the report translates into roughly 70 percent higher odds of a missed appointment per 30 extra days ^[1].

Other factors in the same model, as reported by Scienmag ^[1]:

• Subsidized referral status: aOR 2.007 (95 percent CI 1.929 to 2.087) ^[1].
• Referral from the children's emergency department: aOR 2.070 (95 percent CI 2.018 to 2.214) ^[1].
• Self-initiated referral by the family: aOR 0.304 (95 percent CI 0.287 to 0.322) ^[1].

The team also ran a receiver operating characteristic (ROC) analysis, which tests how well a single variable separates people who attend from those who do not; the area under the curve (AUC) for waiting time was 0.669 (95 percent CI 0.666 to 0.672), which the report describes as modest discriminatory performance ^[1]. An exploratory cutoff of 27.5 days identified non-attendance with a sensitivity of 58.3 percent, meaning the share of missed appointments it flagged, and a specificity of 71.3 percent, meaning the share of attended appointments it correctly left unflagged ^[1].

Why it matters for people who book appointments

The report argues that systems treating waiting time as a modifiable risk factor could recover appointments, a claim the audit's associational design does not itself test ^[1]. It points to subsidized and emergency-department referrals as profiles where reminders, SMS confirmations, caregiver education at the point of referral, or fast-track pathways might be prioritized, though none of these measures was evaluated in the audit ^[1]. It also notes what a missed specialist visit can mean: a condition that progresses unnoticed, a referring doctor who assumes care is underway, and a slot another child could have used ^[1].

The write-up describes the waiting list as one where "the queue itself is quietly dissolving a substantial fraction of the demand it exists to serve" ^[1]. Scheduling tools such as SmartMeet include email and text reminders and let guests reschedule themselves ^[SmartMeet]; this audit offers no evidence on whether such features change attendance ^[1].

What to watch next

The authors describe the 27.5-day threshold as exploratory and say it would need prospective validation before any operational use ^[1]. The reasons families gave for missing appointments, and clinical outcomes for children who did not attend, were not available to the authors ^[1].

Limitations

This is a retrospective audit of administrative data from a single clinic; as the report acknowledges, such a design can identify associations but not definitively explain them ^[1]. The waiting-time AUC indicates modest discrimination, and the cutoff is not a validated tool ^[1]. This summary relies on Scienmag's account rather than the full paper, and the emergency-department confidence interval is reproduced as that report gives it ^[1]. The authors declared no competing interests and no external funding, and the work was run as a clinical audit under SingHealth Central Institutional Review Board policy ^[1].