Seventy four percent (74%) of manually compounded sterile product errors are undetectable at final product verification and would reach the patient

Posted on March 16, 2016

By: Scarlett Eckert, Pharmacist Consultant

MD Anderson Cancer Center’s implementation of an IV workflow software system that uses gravimetric and barcode verification intercepted over 1,000 errors during manual compounding1. Over a yearlong study conducted from November 2012 to November 2013, 15,843 doses were prepared using this system and 1,126 errors (7.1%) were detected during the dose preparation process. During the same period, out of 51,037 manually prepared doses, 49 (0.096%) self-reported errors occurred. This is a significant difference in error capture rate and the article acknowledges that not all near misses that occurred during manual compounding were reported. Additionally, the 49 self-reported errors that were found at the final product check resulted in an increased cost of $56,000, plus additional production and verification time.   Of the 1,126 errors detected by the workflow system, 292 errors (26%) were detected at barcode scanning, 797 errors (71%) were detected by gravimetric weighing of prepared consumable prior to injecting into final bag and 37 errors (3%) detected at the vial reconstitution step. The system’s real time alerts detected and intercepted numerous errors during the vial preparation process, which allows correction before adding to the final administration container. 25 of the 37 detected vial reconstitution errors were vials requiring additional diluent, but 12 vials were reconstituted with too much diluent and required remaking at an additional cost of $12,395.

One of the most important findings of the study is that 834 errors (74%) would most likely never been caught by the pharmacist at the final product check. This is an alarming statistic, especially because of the difference in error rates between the study group and the self-reported errors. Even allowing for near misses that were intercepted and not reported, the real error rate in the manual compounding process is likely somewhere between 0.1% and 7.1%. It is logically difficult and statistically unlikely that we can accept an assumption that the entire 7% difference in error rates is accounted for by near misses.   For the sake of discussion, consider the following table, which shows the number of manual compounding errors that would escape detection for various possible error rates between 0.1% and 7%, calculated using the 74% statistic cited above.


51,037 Manually Compounded Doses
0.1 38
1 378
2 755
3 1133
4 1511
5 1888
6 2266
7 2644


For the errors that escape detection, the real impact on the patient must next be considered. How many of these errors result in harm? The harm spectrum is equally variable and ranges from no harm to death.   Using an average value for errors that result in an increased length of stay, reported by Schneider and Urbine as 1.6%2, manual compounding errors in this study leading to longer hospital stays could have occurred in as many as 42 patients. Continuing this extrapolation for increased length of stay (3.17 days) and cost ($1355/day) as reported by Schneider and Urbine, the increased cost of care on these patients is over $180,000.   Reimbursement payment penalties, costs for investigating these errors, and other costs associated with error management must be considered as well.

In addition to the 7.1% error reduction, implementing an IV workflow system allowed MD Anderson Cancer Center to improve their workflow, reducing technician IV compounding time by 34% and pharmacist verification time 37%. This system also allowed for improved vial management and decreased waste resulting in further reduced total costs.

This study of IV workflow software demonstrates that the manual IV compounding process, no matter how many checks and balances are in place, has the potential for compounding errors that cannot be detected by our traditional visual inspection methods. Implementing IV workflow software with barcode verification and gravimetric weight checks intercepts compounding errors, but still requires rework to correct the error, and also does not address the issue of possible contamination of the CSP with viable microorganisms.  Fully automated compounding systems, such as RIVA, use precise automated measuring methods and gravimetric checks to ensure the accuracy of each dose. In addition, because automated systems eliminate human interaction within the sterile compounding environment, a higher degree of sterility assurance can be achieved.

We cannot continue to use the same error-prone processes for compounding sterile preparations.   Better methods using barcoding, gravimetric verification, and automation are available and must be implemented as soon as possible to protect patients from harm.


1 Reece KM, Lozano MA, Roux R, Spivey SM. Implementation and evaluation of a gravimetric i.v. workflow software system in an oncology ambulatory care pharmacy. Am J Health Syst Pharm. 2016;73(3):165-73.

2 Urbine TF, Schneider PJ. Estimated cost savings from reducing errors in the preparation of sterile doses of medications. Hosp Pharm. 2014 Sep;49(8):731-9.

Announcements | Automation | Blog | ROI | Safety