All your points are valid and relevant. However we were working under limitations. The problem of medical errors is so huge, so ignored and so difficult to assess, we strongly felt that any attempt to define it however infantile is better than no attempt at all. We agree that the best way to accurately quantify errors is to study each prescription and use it as a denominator as you have mentioned, but practically this is extremely difficult. Secondly, the clinicians and staff were not blinded. This is a limitation of the study, but only serves to highlight that the problem is probably bigger than it appears.

The reason why more errors were by senior residents than junior residents were because in our hospital, most orders are written by the senior residents than junior resident.

Regarding the point about how many errors were intercepted, the data is as follows: 105 (22.97%) errors were detected immediately after being written, 63 (13.78%) were detected within 30min of occurrence, 194 (42.45%) were detected between 30min to 24hrs and 95 (20.78%) were detected after 24hrs of occurrence. Out of the 457 errors made on paper (planned), 341 (74.61%) were actually executed. In 375 (82.05%) cases no clinically significant outcome was noted.

Gouri Passi,
Mansi Parihar,
E-mail: [email protected]