Progressive BI analytics routed in complex data modeling at healthcare leader

Background

The leading international healthcare group recently implemented a global patient safety management system. This system captures all patients’ incidents that take place within the hospitals to ensure compliance and assist audits.

Challenge

The system's reporting capability was insufficient as comprehensive information about the most commonly occurring events, their frequency per hospital and their reporting amongst the different management divisions was missing. The customer required reporting that offers immediate, full visibility into the number of events at any point in time to improve patient safety. Detailed analyses of prevalence and place of different events, and prevention of these events were a desired outcome of this project.

Approach

The customer engaged Keyrus to design and implement rigorous data analyses and reporting that enable access to real-time patient safety information. The group was using world-leading BI tool, Tableau, of which Keyrus is the only Premier Partner on the African continent. The solution involved complex data modelling as the previous patient safety management system was designed as an operational and not a reporting system. Keyrus team worked closely with the customer’s staff to define and develop meaningful KPI’s to drive the right decisions and actions.

Key results

Benefits

The customer now has accurate, real-time information at their fingertips. The customer is able to see when events tend to occur, at which facilities, contributory factors, circumstances, patient outcomes and risk factors. Perhaps most importantly, they can pick up trends which enable them to put processes and interventions in place to further improve patient safety. The group also now boasts a clinical decision support dashboard for items that require management or operational action. This shows new actions, those in progress and overdue actions, which are flagged. Statuses are color-coded for ease of use. Each incident is tracked until it is closed, together with what caused it and how it could be avoided in future. Once sufficient data records have been built there may be future scope to run machine learning (ML) algorithms to predict incidents based on common trends.

Share this key play