Changes in error rates in the Australian key incident monitoring and management system program

Introduction: The Key incident monitoring and management system program (KIMMS) program collects data for 19 quality indicators (QIs) from Australian medical laboratories. This paper aims to review the data submitted to see whether the number of errors with a higher risk priority number (RPN) have been reduced in preference to those with a lower RPN, and to calculate the cost of these errors. Materials and methods: Data for QIs from 60 laboratories collected through the KIMMS program from 2015 until 2018 were retrospectively reviewed. The results for each QI were averaged for the four-year average and coefficient of variation. To review the changes in QI frequency, the yearly averages for 2015 and 2018 were compared. By dividing the total RPN by 4 and multiplying that number by the cost of recollection of 30 AUD, it was possible to assign the risk cost of these errors. Results: The analysis showed a drop in the overall frequency of incidents (6.5%), but a larger drop in risk (9.4%) over the period investigated. Recollections per year in Australia cost the healthcare industry 27 million AUD. If the RPN data is used, this cost increases to 66 million AUD per year. Conclusions: Errors with a higher RPN have fallen more than those with lower RPN. The data shows that the errors associated with phlebotomy are the ones that have most improved. Further improvements require a better understanding of the root cause of the errors and to achieve this, work is required in the collection of the data to establish best-practice guidelines.

Security of the Internet of Things Based on Game Algorithm

To explore the security mechanism of the Internet of Things (IoT) perception environment, we perform a security research on the IoT on the basis of game algorithm. The dynamic game method of node cooperation is used in the experiments. Firstly, multiple report nodes are merged into a game party, and the dynamic game for two parties is established with the detection node. In the environment where the malicious nodes are dominant, the detection nodes collaborate, and the state of the unknown nodes is conjectured by the reputation value of the reporting nodes. The high trust reference report is used for the modification and reduction the weight of malicious nodes in the overall report, for node merging, and finally for bias equilibrium. The results show that cooperative game can significantly improve the success rate of incident monitoring and reduce the number of forged reports.

The key incident monitoring and management system – history and role in quality improvement

Background: The determination of reliable, practical Quality Indicators (QIs) from presentation of the patient with a pathology request form through to the clinician receiving the report (the Total Testing Process or TTP) is a key step in identifying areas where improvement is necessary in laboratories. Methods: The Australasian QIs programme Key Incident Monitoring and Management System (KIMMS) began in 2008. It records incidents (process defects) and episodes (occasions at which incidents may occur) to calculate incident rates. KIMMS also uses the Failure Mode Effects Analysis (FMEA) to assign quantified risk to each incident type. The system defines risk as incident frequency multiplied by both a harm rating (on a 1–10 scale) and detection difficulty score (also a 1–10 scale). Results: Between 2008 and 2016, laboratories participating rose from 22 to 69. Episodes rose from 13.2 to 43.4 million; incidents rose from 114,082 to 756,432. We attribute the rise in incident rate from 0.86% to 1.75% to increased monitoring. Haemolysis shows the highest incidence (22.6% of total incidents) and the highest risk (26.68% of total risk). “Sample is suspected to be from the wrong patient” has the second lowest frequency, but receives the highest harm rating (10/10) and detection difficulty score (10/10), so it is calculated to be the 8th highest risk (2.92%). Similarly, retracted (incorrect) reports QI has the 10th highest frequency (3.9%) but the harm/difficulty calculation confers the second highest risk (11.17%). Conclusions: TTP incident rates are generally low (less than 2% of observed episodes), however, incident risks, their frequencies multiplied by both ratings of harm and discovery difficulty scores, concentrate improvement attention and resources on the monitored incident types most important to manage.