Only about 1/3 of all cases labelled as hernia from billing codes actually had them when compared with CT scans.

Potential For Substantial Error When Using Administrative Data to Identify Disease Conditions

We assessed the relationship between billing codes for hernia and CT scan diagnosis for the same condition. Only about 1/3 of all cases labelled as hernia from billing codes actually had them when compared with CT scans.

The work highlights the unreliability of using billing codes for identifying disease entities in administrative data. Those codes are not designed to establish the presence of disease, only that the condition is something the clinician is considering when evaluating a patient.

For example, if a patient is seen for urinary symptoms, a presumptive diagnosis of UTI may be billed for even if testing shows there is no UTI. A researcher investigating UTI will erroneously classify that patient as having a UTI based on administrative billing codes.

These findings suggest that a large body of work that relies on ICD 9 or 10 billing codes in administrative data may have misleading findings.

• Source: https://academic.oup.com/bjs/article/112/7/znaf139/8204820

• Via: https://x.com/ehlJAMA/status/1945889237653741960