Characteristics of pediatric chemotherapy medication errors in a national error reporting database

Is there a difference between an error and a mistake? In which category do we place performance deficits and equipment failures? The article by Rinke et al. regarding the characteristics of pediatric chemotherapy medication errors contains sobering, but bias-laden, conclusions that we believe misrepresent the practice of pediatric oncology.1

The primary conclusion that “pediatric chemotherapy errors often reach the patient” failed to emphasize that only 310 errors were reported in 6 years. Considering the annual number of pediatric cancer patients undergoing treatment, in addition to all nononcologic uses of chemotherapy, 50 chemotherapy errors per year most likely represents a small portion of the intended study population. Therefore, it is not valid, based on this study, to reach any conclusion regarding the overall frequency with which pediatric chemotherapy errors reach the patient.

Our pediatric oncology practice monitors chemotherapy errors, “near-miss” events (errors in chemotherapy ordering and production that do not reach the patient), and total doses, allowing us to accurately tabulate error frequencies. Our practice incorporates a periodic analysis of “near-miss” trends with the subsequent implementation of system remedies designed to prevent future errors. By using this method over the last 5 years, which includes more than 30,000 total doses of chemotherapy, we have achieved an average annual chemotherapy “near-miss” dosing error rate of 0.44% and “near-miss” production error rate of 0.83%, with only 0.027% of errors reaching the patient. Although the only acceptable error rate is 0, we believe our rates are significantly lower than all those previously reported.2 We attribute our relative success to incorporating a routine that assumes human error occurs and uses humans, not computers, to prevent them.

The process of giving children chemotherapy from the time of protocol verification through ordering, mixing, delivering, and administering is incredibly complex. Each step of this process is susceptible to performance deficits and equipment failures. Are these really the types of mistakes inferred when investigators report chemotherapy errors and how are they preventable? Every institution experiences equipment problems, delays in chemotherapy delivery, and mistakes in administration. Rapid, appropriate responses to these unavoidable variances are critical and most likely made more difficult to implement when self-reported to a national data tracking base available only by subscription. Our experience suggests that careful institutional monitoring, free from reporting and sponsorship bias, is an effective way to minimize chemotherapy medication errors. As Battista stated, an error does not become a mistake until you refuse to correct it.

David S. Dickens MD*, Diane Sinsabaugh RPh*, James B. Fahner MD*, * Pediatric Hematology/Oncology and Bone Marrow Transplantation, Helen DeVos Children's Hospital, Grand Rapids, Michigan.