Singh · BMJ quality & safety 2014 · secondary synthesis and population extrapolation · n=3 studies

The frequency of diagnostic errors in outpatient care: estimations from three large observational studies involving US adult populations.

Cited 679 times in the scientific literature.

Level 4 - case-series / case-control

Secondary synthesis and population extrapolation from three observational cohorts.

PubMed 24742777 · doi:10.1136/bmjqs-2013-002627 · record verified 2026-08-29

What was done

Data were synthesized from three prior clinic-based observational studies of US adults that used chart review to confirm diagnostic errors, defined as missed opportunities to make a timely or correct diagnosis. Two studies used electronic trigger algorithms to detect unexpected return visits after primary care or lack of follow-up for abnormal colorectal cancer findings; the third evaluated consecutive lung cancer cases. Authors extrapolated these error rates to the US adult population, using primary care data to model acute conditions and cancer cohorts as proxies for serious chronic conditions.

What was found

Combining data from the three studies yielded an estimated outpatient diagnostic error rate of 5.08%, corresponding to approximately 12 million US adults each year. Based on previous work, the authors estimated that approximately half of these errors could potentially be harmful.

Why it matters

This analysis establishes a foundational national benchmark indicating that at least 1 in 20 US adults experiences an outpatient diagnostic error annually, identifying outpatient safety as a critical focus for healthcare policy.

Limits

The national rate is derived from an extrapolation of only three specific clinical cohorts (acute primary care return visits, colorectal cancer follow-up, and lung cancer cases), which may not represent all outpatient encounters. The abstract does not provide the underlying sample sizes or confidence intervals. Relying on electronic triggers and chart reviews captures only documented care and misses errors occurring outside participating records.

Cited by