Spark-owned data and analytics unit Qrious is helping to optimize services at Auckland Hospital with a new integrated operations center.
The initiative provides a central space to coordinate and monitor rapid activity at the 710-bed Auckland Hospital. Clinical nurse managers, patient flow facilitators, charge nurses and senior duty managers are now consolidated in the new 24-hour centre.
Auckland District Health Board (ADHB) director of data and analytics Ali Khan said the aim was to bring together the right people and the right technology to make smarter and faster decisions.
Developed using Agile principles, the deployment gathered aggregated information from five different source systems. Product dashboards are refreshed every 90 seconds, 24/7 with up-to-date data.
Qrious and ADHB staff built four dashboard views covering bed, general hospital, ward, and staff statuses and trained the hospital’s health intelligence team to maintain the solution in-house.
Qrieux designed the solution using Microsoft Azure cloud services providing a solid, secure and highly available foundation for these critical services. From there, ADHB could use Azure Analysis Services and PowerBI to provide dashboard-based monitoring and management.
The new dashboards are projected onto walls and displayed on digital screens in the operations center and are accessible via smartphone or tablet so that hospital staff can access information remotely.
Khan leads the multifaceted team within ADHB’s Health Information Technology (HIT) directorate, which works with staff from all other directorates.
“The initial Qrious project implemented PowerBI, some Azure components such as analytics services and logic apps, all managed through DevOps processes,” he said. Reseller news.
“This enabled a real-time view of hospital operations for our integrated operations center. Since then, we’ve added significant new capabilities and content and are now building a full Azure technology stack that includes Delta Lake, Synapse/ADF, DataBricks, Embedded PowerBI, and full PowerApp capability.”
Examples of products and services provided include real-time operational information on bed capacity, theater staffing and utilization, cost of elective services, and IMT incident management team support covering the response to COVID-19 and the deployment of vaccination.
More automation is being added to allow new data to be introduced into the ecosystem at little or no cost to the end user or the efforts of the ADHB team, he said. This is achieved through the implementation of “DataOps” centric solutions running on modern cloud technologies.
DataOps, a mirror of DevOps, is a set of Agile practices, processes, and technologies to improve quality, speed, and collaboration to deliver continuous improvement in data analytics. An example is using code to query a data model repository to automatically generate and schedule PowerBI datasets in ADHB’s presentation layer.
“This new capability significantly reduces friction and delivery costs and allows us to drive data delivery through a model-based approach,” Khan said.
“Finally, a particular area of growth for us is the practical adoption of artificial intelligence and machine learning.”
The project was timely, with the new dashboards going live in late 2019, just before the COVID-19 threat emerged.
Even so, it’s sometimes difficult to understand the tangible connection a computer system can have to patient care, Khan said.
“But you cannot underestimate the effect that making smarter and faster operational decisions has on patient well-being. We are having a positive impact on the services we provide at Auckland Hospital .”
Operations staff, for example, can see the activities of different functions and the flow effect if a team is slowed down.
“By reducing silos and displaying a consolidated view, they are better equipped to solve problems and improve day-to-day hospital operations.”
Change of team
Dashboards are also now a starting point for team meetings and transfers for nursing and service staff, who stand around them in the operations center before shifts.
“With access to live information, staff can identify issues in real time and act quickly,” Khan said.
The system’s mobile capabilities not only provide actionable insights to staff, but also peace of mind to senior managers, who can remotely check on the health of the facility.
Khan said companies were struggling to realize the full potential of analytics and keep up with demand. The data team never fully meets the parameters of business demand, including cost, which is low to zero, and speed.
Companies must overcome these challenges.
“At ADHB, we have a comprehensive data and analytics strategy in response to our business needs and also for our digital transformation,” he said. “True self-service analytics is at the heart of this strategy enabling business analysts to use the same tools as our team and work with data at all levels in our data environments.”
The DataOps approach, he said, was different from the traditional approach where data was only visible to business teams once it had been massaged, incorporated and transformed into data platforms.
“DataOps is central to our new delivery model, with true self-service and automation being key enablers of business value. This approach focuses on building data products, prototyped by or in collaboration with the business in secure, modern self-service environments powered by automation and data discovery. »
A new psychology
It changes the psychology of the relationship between business teams and IT, he said, creating better buy-in and trust while driving data innovation that was not possible in business and project environments. traditional.
“A lot of the biggest challenges aren’t necessarily technical, but rather related to people, process and culture,” Khan said.
“The leadership team needs to be aligned and genuinely committed to data transformation and creating a culture of change – with clear direction, expectations and business processes established across the rest of the organization.”
It also involves a journey with no real end.
The ADHB conducts over 1000 sleep studies per year, each requiring several hours of manual review by an expert clinician to annotate the stage of sleep and the various respiratory events that occur.
“We want to use AI to help diagnoses of obstructive sleep apnea-hypopnea syndrome, by labeling various sleep stages and respiratory components that occur in sleep studies,” Khan said.
“The labor-intensive nature of labeling combined with the large number of studies that occur make it an attractive task for automation.”
Qrious data scientists are now engaged in building AI models to perform sleep staging and sleep apnea detection tasks.
In 2020, Microsoft also helped engineering firm Asbuilt deliver a digital twin of Auckland Hospital, providing a digital footprint of all physical assets in the hospital to help hospital managers understand resources available and how to optimize the way the spaces were used.
Join the newsletter!
Error: Please verify your email address.
Tags business intelligencehealthmicrosoft azurhospitalsdistrict health boardsparkQriousanalyticsADHB