Death by diagnostic

According to a new study that extrapolates from data at four hospitals, about 300,000 errors in cancer diagnosis occur annually in the US and about 120,000 patients per year are harmed. The most common form of harm was delayed diagnosis or repeated testing –which is bad enough. But more substantial harm was reported as well, such as unnecessary hysterectomies and unnecessary cancer treatments.

The wild variability among the institutions is also troubling. Reports of harm were mainly based on each institution’s reporting. On the high side one hospital reported harm in 89% of cases where there was an error, another reported harm in 85%. On the low side one hospital reported harm in 20% of cases and another reported harm in zero percent! The authors recommend standardization and uniform reporting of errors. Good idea!

Receiving an incorrect cancer diagnosis is agonizing, expensive, and potentially dangerous. There is no consumer information available that would help a patient (or even a physician) select a lab that is less error prone. This study provides a hint at how the US can spend so much money on health care without having better health outcomes to show for it.

See Raab SS et al. Clinical impact and frequency of anatomic pathology errors in cancer diagnosis. Cancer. Advanced online publication October 10, 2005 and the commentary on MedPage Today.

October 11, 2005

Leave a Reply

Your email address will not be published. Required fields are marked *