Registry Data Analysis and Reporting
Monash Clinical Registries utilise the expertise of senior and experienced statisticians, data scientists and data analysts for all formal reporting and more complex research-related activities. Subscribing to the principles of Reproducibility, Accuracy, Consistency and Validity (RACV), the team also works closely with registry practitioners via a hub and spoke model, to elevate the level of statistical literacy among staff and improve on the quality of reporting.
The team leads methodological and applied research, including simulation studies to improve the application of outlier detection methods for rare/sparse disease outcomes, geo-spatial Bayesian hierarchical models for areal data, risk-adjusted cusum models, machine learning techniques to predict and classify health outcomes and reporting checklist for registry data. The team regularly conducts workshops on ‘Clinical Registry Data Analysis using Stata’ and is currently planning newer workshops on machine learning, Bayesian statistics, developing dashboards and automating reports using R.
Our Team
Our services
SPECIFIC ACTIVITIES
- Providing statistical support to produce annual and site reports which are statistically sound and in a timely manner
- Troubleshooting existing data analytics problems
- Developing guidelines and operating procedures for the design and analysis of registry data
- Undertaking quality control audits of data/analysis from registry data
- Developing descriptive and inferential dashboards
- Automating annual and benchmarked site reports
- Planning and conducting training courses and workshops
ADVANCED SERVICES
Current key projects
- Development and deployment of machine learning models tailored for lung cancer and breast device registries
- Establishment of a simulation framework aimed at assessing and contrasting statistical models dedicated to outlier detection
- Utilisation of Bayesian spatio-temporal models for the integrated analysis of various clinical registry quality indicators
- Creation of a comprehensive checklist to enhance the statistical reporting of clinical registry data
- Development of both descriptive and inferential dashboards for managing diabetes registries
- Automation of the entire reporting process, from data collection and extraction to the generation of validated reports, for various Clinical Quality Registries
- Verification of the efficacy of statistical methods across diverse software platforms and environments