Table 1.
Example Candidate Undesirable Diagnostic Events (UDEs)
| Condition | Context (setting and timing) | Readily clinically diagnosable | Objective, valid, and widely used reference standard | Prone to error | Timely diagnosis likely to have a positive effect on patients’ health | Sufficiently common | Diagnostic process definable | Balance measure available |
|---|---|---|---|---|---|---|---|---|
| Rheumatoid arthritis | Outpatient >12 months |
No (presentation varied, often over time) |
Yes | Yes | Yes | Yes | No (given varied presentation, many different routes to diagnosis) |
No |
| Sepsis | Inpatient >12 h |
Yes | No (reference standard is usually clinical diagnosis of sepsis, subject to marked variation and false positive/negatives) |
Yes | Yes | Yes | Maybe (many patients at risk for sepsis from many conditions, making diagnostic process heterogeneous) |
Yes (blood culture rate, antibiotic days) |
| Bacterial meningitis |
Emergency department/inpatient
>1 visit |
Yes | Yes | Yes | Yes | Yes | Yes |
Yes
(lumbar puncture rate, empiric antibiotic use) |
| Tuberculosis | Diagnosed on autopsy | Yes | Yes | Yes | Yes | Yes | Yes |
Yes
(patient isolation days) |
| Spinal epidural abscess |
Emergency department/inpatient
>2 visits |
Yes | Yes | Yes | Yes | Yes | Yes |
Yes
(MRI rate) |
| Colorectal cancer | Outpatient >12 months after + screening test |
Yes | Yes | Yes | Yes | Yes | Yes | No |
A number of clinical conditions and contexts (timing and setting) are presented and evaluated using the defined criteria. Examples that meet all seven criteria are in bold font