By T. Forcht Dagi, MD, DMed Sc, MPH
Note: This is the third installment of a three-part series. Part I provides an overview of data dashboards in the clinical setting; Part II reviews the history and evolution of dashboards; and Part III advises best practices for undertaking a dashboard project.
A dashboard project is not something to be undertaken lightly. They require a tremendous commitment of time and resources. Planning is key. A pivotal question is how much information is needed, and how much underlying data are available for integration. While basic dashboards displaying static information are simple projects, they are often not worth doing. Interactive dashboards with moderate detail are moderately complex and offer limited benefits. The highly interactive dashboards which lead to rich but also complex data and data analytics, however, are extremely complex.
Planning starts with agreement around the objectives of the dashboard and its scope. Key Performance Indicators (KPIs) must be identified. They must matter to the primary users and be supported by the quantity and quality of data available. The involvement of the end user at the planning stage is hugely important. The difficulty of identifying useful KPIs and working with the supporting data should not be underestimated.
The development of new dashboards is best pursued in phases. During the discovery phase, requirements and specifications are gathered and established. The pertinent data elements are identified. These elements need to map to the KPI.
The second, or design phase designs the dashboard around the content and appearance desired. During the third, or integration phase, the final user interface is established; queries are developed to extract information from the correct databases; the data warehouse is constructed; the dashboard is laid out and published to a portal where it can be accessed by users; security measures are implemented and tested; and the dashboard is validated and tested for fitness for use.
In the fourth, or deployment phase, the dashboard is deployed in its intended environment and tested to scale. Security features, overall reliability and other measures of performance are also assessed. This phase tests the dashboard in real life or in what industry calls “the production environment.”
Finally, at the point a dashboard is ready for deployment, it must be re-introduced to the end users with adequate training, support and reinforcement.
The dashboard must be actively maintained after it has been deployed and adopted. Needs, interests and requirements will change. The dashboard must remain sufficiently flexible to allow for enhancements. Whether the vendor maintains the dashboard or the client is both an operational and a strategic decision. Ideally, the client will become self-sufficient in most matters. In the health care environment, however, particularly in the context of a single neurosurgical department or division, it might be more efficient to outsource this function.
Whether outsourced or not, full and archived documentation must be pursued, verified and validated during development and at each subsequent step, including after deployment. Indeed, a documentation audit must be undertaken before a dashboard project is considered complete. The documentation should include a complete record of the elements of the dashboard and the logic behind its interlocking components, and be capable of supporting maintenance and future enhancement. Most audits of this nature require outside consultants. Very few institutions have the internal capacity to perform adequate documentation audits.
The steps involved in developing and deploying a dashboard can be summarized as follows:
- Initial analysis and discovery of goals and needs
- Process development around desired outcomes
- KPI selection
- Comprehensive design and construction
- Testing and validation
- Documentation audit
- Deployment and integration
- Ongoing monitoring and maintenance
The project can be deemed successful only when the KPIs prove to be material, after the dashboard can be shown to display the information pertinent to the goals of the project in actionable terms, and after the dashboard is both adopted and integrated into workflow. It is impossible to overstate the importance of selecting the right metrics and building the right data architecture. The visual presentation comes third and the whole must be supported through effective implementation, launch and support.
Dashboards do not stand alone. In the clinical context their purpose is to catalyze and support focused changes in clinical practice and clinical workflow to improve patient care. The process of launching and deploying or implementing a dashboard must be accompanied by a commitment to engage the individuals who will enter the data, use the data, make the changes and monitor the effects. Continuing improvements in automation will facilitate data entry, but the iterative process of analysis and optimization is entirely human.
Effective dashboards are characterized by relevance; navigability for and applicability to the purposes designed; excellent performance, including speed, scalability and reliability, from the perspective of the user; and coherence. They extract data from repositories of electronically stored data and they facilitate analysis. They are adoptable and usable.
Dashboards are not the only way to drive change in clinical processes, nor even to collect and array data to improve patient outcomes. Nevertheless, so long as the goals are clear, the efforts required to develop and deploy them appreciated, the necessary resources available, and the applicable infrastructure standards met, well designed dashboards serve as excellent tools to manage change.
Dr. Teo Dagi, Dean of Health and Informatics Management, Bryan University
As Vice Chancellor and Dean of Health Informatics Management, Dr. T. Forcht Dagi oversees programs of study in Health Information Technology, Applied Health Informatics and Advanced Personal Training and Exercise Science at Bryan University. Dr. Dagi is a neurosurgeon who holds an AB from Columbia University, and MD and MPH from Johns Hopkins University. In addition, he earned an MBA with distinction from Wharton School of Business, and was awarded an honorary doctor of medical science degree by Queen’s University Belfast for his contributions to medicine and public service. In addition to his role at Bryan University, he serves as a Distinguished Scholar and Professor at Queen’s University Belfast, Northern Ireland, and lectures at Harvard University and Massachusetts Institute of Technology.
Bryan University offers future-proof, industry-strong online degree programs in health and legal services. The university is committed to the personal and professional development of each student, providing an expansive educational experience in which preeminent industry experts teach, mentor, and lead the way to future success. For more information about the university and its programs of study, visit www.BryanUniversity.edu.
Media Contact: Drew Whitney, drew.whitney@bryanuniversity.edu, 602.384.2587
The post Undertaking a Dashboard Project appeared first on Bryan University.
