Identification and correction of human errors in real-time – a blood banking surveillance system

Background Human errors are a major challenge in blood banking. Innovative ICT-based strategies for error interception and prevention must be developed and investigated for their efficacy. Materials and Methods A browser-based Blood Bank Management Information System (BBMIS) intercepting real-time human errors was implemented at The Indus Hospital Blood Centre, Karachi, Pakistan. Data were analysed for 6 blood bank, and three transfusion-related categories of errors intercepted, from December 2013 to December 2016, during which the blood centre drew 45 425 donors. Results A total of 413 errors were intercepted and corrected. There was an attempt to prepare 16 of 2199 quarantined components. Reagent quality assurance was not performed in 3 of 1126 instances. There were 169 misinterpretations of 113 536 blood groups, 44 of 90 850 mismatches at donor blood group reconfirmation and wrong blood group labelling in 98 of 141 417 cases. At unit verification, 26 of 28 581 units were not physically discarded. Wrong blood in tube was identified in 7 of 24 962 cases. Incorrect sample tube was selected for testing in 25 of 32 857 cases. There were 11 attempts to issue units to the wrong patient of 14 644 issues. Conclusion Auto-generated alerts on interception of human errors in real-time eliminate bias of under-reporting due to human dependent surveillance systems, thus, providing a better representation of the range of human errors in the blood bank and transfusions. The BBMIS was able to intercept errors at inception, halting further processing and prompting the user to take corrective actions, thus, enabling correction of all errors in the studied categories.