effectiveness Prepared by: David Pryce & Reshma Kolambkar - PowerPoint PPT Presentation

Diabetes Inpatient Surveillance Dashboard: Evaluation of effectiveness Prepared by: David Pryce & Reshma Kolambkar Business Analytics Services WSLHD Health Data Analytics October 2019

Aim • Evaluation of effectiveness of dashboard • Evaluation of impact on clinical activities

Introduction  Western Sydney has a high  Generating and maintaining prevalence of Diabetes. an updated list of these individuals was challenging.  Testing Diabetes is not a routine in all hospitals in WSLHD.  Mid-2016 - HbA1c Screening tests at Blacktown and Mt. Druitt ED commenced.  Revealed alarming rate – 17% Diabetes (HbA1c > 6.5 ) – 30% Pre-diabetes (HbA1c 5.7 - 6.4)

Background  Diabetes Inpatient Surveillance Dashboard developed  Improve identification of diabetes patients across Hospital  Provide Pop Health view if HbA1c Testing (>100k)  Enable improved management of inpatients by Diabetes Team via MOC

Data Source

Dashboard as a tool

Evaluation Method

Quantitative evaluation  Currently – Length of Stay (LOS) – Reduce in the LOS by ~2days – CMI – No signification improvement Clinical Indicators – 6 Months Before Comparison Indicator 6 months After (Average length of stay in days ) (05/05/2018 – 04/11/2018) (05/11/2018 – 04/05/2019) Primary Diagnoses is diabetes - Including 0.51 ↓ Gestational Diabetes 3.93 3.42 Primary Diagnoses is diabetes - Excluding 0.48 ↓ Gestational Diabetes 3.93 3.45 Identified diabetes in any diagnoses - 0.99 ↓ Including Gestational Diabetes 5.36 4.37 Identified diabetes in any diagnoses - 1.95 ↓ excluding Gestational Diabetes 7.19 5.24  In future, planning to consider HAC for diabetes complication

Qualitative evaluation

Qualitative evaluation

Evaluation Results  Improved identification diabetes patients  Improved clinical management of hyperglycaemia and hypoglycaemia  Improved identification of undiagnosed diabetes in pregnancy.  Reduce inpatient length of stay  Dashboard helped identifying data quality issues  User feedback - easy review of all diabetes patients across the whole hospital

Discussion  Better outcomes via reduced LOS  Need to analyse the Model of Care – to determine Improved evaluation metrics  Usefulness of the dashboard as a tool by end users  Staff are using the dashboard for MOC Operational and research purposes  Some better evaluation metrics could be TAT, Complications, eGFR, measures of Liver/Kidney conditions

Conclusion • The evaluation carried out showed a decrease in the Average length of stay • Feedback from clinician that dashboard helped to quickly identify patient and provide better care • Triangulation of data from different sources can provide a comprehensive view of diabetes patient journey in hospital. • Dashboard enables complex data set to be easily accessed by clinicians on one platform.

Thank You to the Project Team  Project Sponsor – Dr Tien-Ming Hng, Head of Endocrinology Department, Blacktown Mt Druitt Hospitals  BAS Team – Ching Luo (Senior Analyst Developer), Reshma Kolambkar (Senior Business Data Analyst)  BAS Manger – David Pryce

Feedback/Questions?

Thank You !!!