Leveraging Data Analytics and Tech in the Losartan Drug Recall Page ContentThe OpportunityIHiS was involved in the recent recall of three brands of high blood pressure medicine containing Losartan after they were found to contain higher than acceptable levels of nitrosamine impurity. Long-term exposure to nitrosamines at high quantities may potentially increase the risk of cancer. In response to the recall, three teams were called to action to support the healthcare clusters in the following priorities:The Solutions1. National-level Statistics Reporting for Active TrackingLeveraging IHiS’ data science expertise, the Health Insights team worked to identify affected patients and facilitate the conversion of their medication and the tracking of medication dispensing. They deep-dived into data from the National Electronic Health Record (NEHR) to provide statistics from across the public healthcare network to the Ministry of Health (MOH).Efforts involvedPerforming multiple iterations of data profiling and developing computational algorithms with MOHIdentifying data elements and defining how the data can be used by the Chief Pharmacist Office (CPO) to support the planning and tracking of the recallDefining the criteria to determine Losartan patients affected by the recall and the tracking of those who have converted to alternative medicationsDesigning and generating the recall report to facilitate analysis and planning work Close collaboration between IHiS’ Clinical Informatics Group (CIG) and the healthcare clusters was necessary to understand the recall operations that were performed on the ground and the context of introducing new institution drug codes for the purpose of drug returns and exchanges. In doing so, they successfully identified a list of drug codes and descriptions andworked with the Health Insights team to translate them into extraction rules, which were then used in NEHR to generate reports for supporting the drug recall.Their efforts were not without challenges. During the data extraction process, the team spent hours manually reviewingmore than 2,000 anonymised medication records of patients to ascertain their total daily dosage of Losartan. Thisis so that they could prioritise the conversion of patients who were on 100mg or more of the medication. The manual review was also necessary as computational rules may not have been applicable to all scenarios. For example, identifying patients for drug conversion based on their last prescription records would be inaccurate as it does not present a holistic view of the types of medication that patients are currently consuming.In the course of the recall exercise, the team regularly reviewed and updated the operation processes at all healthcare clusters to support better cohesion and integration across all teams. Enabling the sharing of near real-time information of patients and medication usage across the public healthcare network, our mission-critical efforts provided important statistics for MOH to plan, coordinate and track the recall with the healthcare clusters. Timely release of the statistics was also crucial. On 31 March, IHiS released the first report on the number of affected patients and the number of patients who have had their medication replaced three days after the recall announcement. The following day, IHiS released a report on the utilisation volumes of Losartan and alternative drugs for MOH to monitor patients who had started on alternative medication and the stock volume in the healthcare clusters.2. Ensuring Sufficient Medication SupplyOur Product Management & Delivery Group 2-Pharmacy team, together with vendors, ran specialised scripts within the pharmacy IT systems of the healthcare clusters to automate the processes of the rationing of medication and patient identification. The automation helped reduce the turnaround time taken by the healthcare clusters to contact affected patients and allowed front dispensing counters to inform patients on the spot about when they can collect their replacement medication subsequently.To cater to varying operational needs of the healthcare clusters, the IHiS team spent two to four weeks to help create, modify and test system scripts to generate dispensing reports. Patient safety is key. Where technology is concerned, the work continues with our healthcare clusters to manage the Losartan recall. Next UpIHiS is currently working to integrate a batch management functionality as part of a new national pharmacy management system. The upcoming platform will harmonise the dispensing of outpatient, discharge and A&E medication across healthcare clusters using consistent medical terminologies to achieve integration of pharmacy operations for best practices.In future recalls, this new feature will allow drugs to be recalled by batch numbers on the medication packaging instead of relying on date ranges (date the prescription was filed) for greater accuracy. Regular reviews will also be carried out on the data limitations within the NEHR and potential ways to improve our data analysis methodology.