Healthcare has been identified as a sector that could benefit greatly from the use of advanced analytics, especially in the area of clinical outcomes improvement. Use cases such as readmissions reduction and population health analytics are high on the priority list of senior executives looking to navigate the transition to value-based care.
Yet, recent studies indicate that something is holding back the sector from realizing the full potential of analytics. A study by Health leaders Media on population health management indicated that while a full 69 percent of study participants were “fully committed” or are running pilots, many of them were in early stages of formulating strategies and making appropriate investments.
A similar study by HIMSS Analytics and BI vendor Qlik found that most healthcare providers were struggling with end-use adoption. The study further suggested that healthcare providers are barely scratching the surface when it comes to unlocking the inherent value of their clinical data. According to John Hoyt, EVP of HIMSS Analytics, “BI benefits are in quality improvements and driving out waste…and often, improving quality reduces unnecessary cost.” “While analytics is front and center in healthcare, maturity can vary by solution areas – driven in large part by local and regional pressures,” says David Bolton, Global Director of Healthcare Industry Solutions for Qlik.
So is it really a gloomy picture today for healthcare analytics?
Pockets of excellence – Adventist Health System
Florida-based Adventist Health System (AHS) serves more than 4.7 million patients annually through 44 hospital campuses across 10 states. At AHS, much effort has gone into laying the foundation for measurement and performance benchmarking at the facility and provider level in addition to tracking population health metrics across broader affiliations. The effort has been evolving for the last five years, according to Nick Scartz, Chief Analytics Officer for AHS. By using a combination of tools and platforms, including Qlik for BI and data discovery and Explorys for population health metrics, Scartz’s team has set up the framework and processes for a number of clinical and financial scorecards and dashboards. An example is the monthly “clinical close” for the entire system which drives accountability through a range of operational performance metrics. Scartz has now started hiring data scientists with PhD and master’s degrees to develop predictive models and heuristic data mining techniques to direct more analytics toward identifying the most effective interventions of individual patients across the continuum.
The journey of Adventist Health System is not atypical. According to the joint survey by HIMSS and Qlik initial priorities tend to be driven by the requirements of the Accountable Care Act (ACA), and other immediate priorities such as ICD-10 coding accuracies with near-term financial implications. Integration of BI and analytics tools with other systems tends to be another high-focus area, not surprisingly, since a robust data integration framework is critical to delivering real-time analytics and performance metrics.
An interesting aspect of Adventist Health System’s success with their BI and analytics program has been the support of senior management at the health system. While the joint survey by HIMSS and Qlik points to end-user adoption as the key challenge reported by participants, AHS seems to be ahead of the pack in this regard. Says Scartz, senior executives at AHS are fully bought into driving accountability through the tie-in of performance to operational metrics and the clinical close. Over the last three years, AHS has been able to reduce cost and length-of-stay for four of their highest volume and variable diagnoses by 10% and 4% respectively, controlling for case mix. AHS has also saved more than $15 million over two years by driving the adoption of disease specific evidence-based physician order sets.
Yet another interesting aspect is the way AHS has been approaching advanced analytics, specifically predictive models. By partnering with Cerner, their EMR vendor, Scartz and his team are developing customized models for AHS that integrate seamlessly into the EMR system and deliver actionable insights at the point of care.
Other examples: UPMC and University of Tennessee (UT) at Knoxville
Healthcare providers such as UPMC and the University of Tennessee Medical center are deploying analytics to bring about meaningful reductions in readmissions rates by stratifying patients based on risk profiles and targeting interventions accordingly. UPMC, in particular, has been an early adopter of risk-scoring of patients in a population for stratification and targeted interventions. More recently, they have been able to measure the impact of the interventions based on these models and have demonstrated a 2 percent reduction in readmissions rates to 13.5 percent. Along the way, they also learned about what cohorts respond best to interventions and can impact readmission rates. University of Tennessee has developed a process to put patient risk scores in the hands of clinicians on a daily basis to help determine appropriate intervention plans. UT has seen a 10 percent reduction in readmissions rates.
It’s early days yet
The surveys seem to suggest that the hype of analytics is overhanging reality by a significant margin. Analytics teams have to work with limited investment outlays at a time of margin pressures across the healthcare sector. Several other challenges exist, especially around data integration and data privacy protections for patients. However, it would be premature to draw conclusions about the state of healthcare analytics simply based on these surveys. Isolated pockets of excellence are coming to light progressively, and the collective experience of these health systems will drive wider adoption in the coming months.
This article is published as part of the IDG Contributor Network. Want to Join?