[Quality of information on adverse events provided by the surgical patient]

To determine whether patients are a good information source on the occurrence of adverse events (AE).

Analytical retrospective study using in-depth interviews in a double-blind protocol, and in parallel, to ensure whether the patient had actually suffered an AE. The Harvard method was also applied to review the medical records using a screening guide. Agreement between the physician and patient point of view of the surgery outcome was also estimated.

A total of 28 randomly selected surgical patients discharged from a general hospital were interviewed. Ten patients (28% of the total suffering an AE yearly) who had experienced an AE, confirmed after a medical record review, and 18 patients who did not suffer an AE.

Intraclass correlation coefficient for the agreement between the medical criterion and the patient point of view was 0.35 (95% CI; 0.2-0.6), and the number of correct classifications was 20/28 (71%, 95% CI; 51-86). Reporting an error reduces the likelihood of the hospital being considered as safe (Fisher's exact p=0.012). Errors were attributed to workload and to the intrinsic randomness of human activity.

Patients can contribute in identifying an AE affecting them in a reasonable manner, providing us with additional information for enhancing patient safety and the quality of medical records.