SIIM: Why medicine needs business analyticsnow more than ever
WASHINGTON, D.C.—As pressure grows for hospitals to provide better care at lower costs to more patients, providers can no longer linger in the the dark ages. A business analytics movement in healthcare could deliver substantial savings for providers and improved outcomes for patients, according to a June 3 session on business analytics at the annual meeting of the Society for Imaging Informatics in Medicine (SIIM).
Business analytics, or business intelligence, describes the use of data, statistics and modeling to improve understanding, empower informed decision making and optimize organizational processes, explained Christopher D. Meenan, CIIP, from the University of Maryland Medical System in Baltimore.
“Business intelligence is nothing new. In fact, it has been used for years in almost every industry but healthcare,” added Katherine P. Andriole, PhD, from the department of radiology at Brigham and Women’s Hospital and Harvard University in Boston. The purpose of business intelligence, Andriole argued, is to enable better decision making via straightforward metrics and visual tools.
Business analytics begins with the aggregation of data. At this point, however, is where many practices veer off the intelligent road, collecting databases worth of data without prioritizing or organizing their metrics. “We suffer from the disease of information overload,” insisted Paul G. Nagy, PhD, from Johns Hopkins University in Baltimore.
Although industries apply an array of methods for capturing and interpreting data about an organization’s processes, key performance indicators (KPIs) have proved the most effective for the healthcare industry, the presenters stated. Understanding what a given institution is doing well and where it can improve requires identifying a few of the most important metrics to focus on and measure. Common examples provided by Meenan, Andriole and Nagy included radiology report turnaround time, productivity, patient safety, medical errors and customer satisfaction.
“Radiology has abundant data sources that make it a good candidate for business analytics,” Meenan said. The essence of business intelligence, though, is to manage these data using graphs and visualizations that provide simple depictions of performance, which can then be honed in on to grasp in detail what is going right and wrong in the department.
In the measurement of report turnaround time, for example, a simple graph can allow a practice to establish a baseline, against which it can compare future trends: Has the group’s signing of reports become quicker or slower? Honing in, then, shows what went right and what went wrong: Were a few individuals responsible for slowdowns? Is reporting slower at a certain time of day or for specific types of studies?
The typical way to render these readable depictions is via electronic dashboards, which can range from simple graphs to color-coded spectra indicating whether a group is above, below or at its baseline or target KPI measure. According to Nagy, “The purpose of a dashboard is to make information and data actionable to change individuals’ behavior. Business intelligence gives us the opportunity to create a macroscope, to help us step back and see what is going on—to make sure we are on the right path.”
Andriole emphasized the importance of including all stakeholders in the KPI process, from financial and operational executives to radiologists, IT staff and technologists. Moreover, from the get go, the data should be normalized and scrutinized so that it is “presented in a consistent, validated and unified format. I cannot emphasize that enough,” Andriole proclaimed.
Nagy pointed to several factors that determine a robust KPI or metric: the indicator should be important and patient-focused; repeatable and therefore measurable and manageable; and the KPI should be a metric that delivers actionable information. Moreover, Nagy adds, “A KPI should be chosen where you have identified a gap in performance. If your dashboards are all green, that’s not business analytics, that’s called marketing.”
While beginning with a few, definitively tractable metrics is essential, the ultimate aim is to develop the organization’s business intelligence or "information metabolism," Nagy indicated. The maturation of business analytics should follow an order, from standard reporting to the creation of ad hoc reports. The process should then proceed to effective drilling down for problem solving, followed by setting alerts that efficiently point managers to trouble spots.
As a practice becomes more advanced, statistical analysis, forecasting, pre-modeling and eventually optimizing the institution’s decision making and distribution of resources become long-term catalysts of information metabolism. Accompanying this progress, the presenters noted, should be the inclusion of additional KPIs that provide more and more comprehensive details about the department.
“All this measurement is sort of Orwellian,” Andriole admitted, “but it’s an effective way for a department to get a good view of what’s going on.” Nagy emphasized that transparency should overarch any practice’s business analytics program, as it creates buy-in among staff and contributes to autonomous performance improvement.
Whether conducting a statistical analysis of dozens of KPIs or peering at a five-metric dashboard, improving the department’s performance hinges on the capture of timely, accurate and relevant performance information. Or, as Nagy concluded (quoting deceased statistician W. Edwards Deming): “In God we trust, all others must bring data.”
Business analytics, or business intelligence, describes the use of data, statistics and modeling to improve understanding, empower informed decision making and optimize organizational processes, explained Christopher D. Meenan, CIIP, from the University of Maryland Medical System in Baltimore.
“Business intelligence is nothing new. In fact, it has been used for years in almost every industry but healthcare,” added Katherine P. Andriole, PhD, from the department of radiology at Brigham and Women’s Hospital and Harvard University in Boston. The purpose of business intelligence, Andriole argued, is to enable better decision making via straightforward metrics and visual tools.
Business analytics begins with the aggregation of data. At this point, however, is where many practices veer off the intelligent road, collecting databases worth of data without prioritizing or organizing their metrics. “We suffer from the disease of information overload,” insisted Paul G. Nagy, PhD, from Johns Hopkins University in Baltimore.
Although industries apply an array of methods for capturing and interpreting data about an organization’s processes, key performance indicators (KPIs) have proved the most effective for the healthcare industry, the presenters stated. Understanding what a given institution is doing well and where it can improve requires identifying a few of the most important metrics to focus on and measure. Common examples provided by Meenan, Andriole and Nagy included radiology report turnaround time, productivity, patient safety, medical errors and customer satisfaction.
“Radiology has abundant data sources that make it a good candidate for business analytics,” Meenan said. The essence of business intelligence, though, is to manage these data using graphs and visualizations that provide simple depictions of performance, which can then be honed in on to grasp in detail what is going right and wrong in the department.
In the measurement of report turnaround time, for example, a simple graph can allow a practice to establish a baseline, against which it can compare future trends: Has the group’s signing of reports become quicker or slower? Honing in, then, shows what went right and what went wrong: Were a few individuals responsible for slowdowns? Is reporting slower at a certain time of day or for specific types of studies?
The typical way to render these readable depictions is via electronic dashboards, which can range from simple graphs to color-coded spectra indicating whether a group is above, below or at its baseline or target KPI measure. According to Nagy, “The purpose of a dashboard is to make information and data actionable to change individuals’ behavior. Business intelligence gives us the opportunity to create a macroscope, to help us step back and see what is going on—to make sure we are on the right path.”
Andriole emphasized the importance of including all stakeholders in the KPI process, from financial and operational executives to radiologists, IT staff and technologists. Moreover, from the get go, the data should be normalized and scrutinized so that it is “presented in a consistent, validated and unified format. I cannot emphasize that enough,” Andriole proclaimed.
Nagy pointed to several factors that determine a robust KPI or metric: the indicator should be important and patient-focused; repeatable and therefore measurable and manageable; and the KPI should be a metric that delivers actionable information. Moreover, Nagy adds, “A KPI should be chosen where you have identified a gap in performance. If your dashboards are all green, that’s not business analytics, that’s called marketing.”
While beginning with a few, definitively tractable metrics is essential, the ultimate aim is to develop the organization’s business intelligence or "information metabolism," Nagy indicated. The maturation of business analytics should follow an order, from standard reporting to the creation of ad hoc reports. The process should then proceed to effective drilling down for problem solving, followed by setting alerts that efficiently point managers to trouble spots.
As a practice becomes more advanced, statistical analysis, forecasting, pre-modeling and eventually optimizing the institution’s decision making and distribution of resources become long-term catalysts of information metabolism. Accompanying this progress, the presenters noted, should be the inclusion of additional KPIs that provide more and more comprehensive details about the department.
“All this measurement is sort of Orwellian,” Andriole admitted, “but it’s an effective way for a department to get a good view of what’s going on.” Nagy emphasized that transparency should overarch any practice’s business analytics program, as it creates buy-in among staff and contributes to autonomous performance improvement.
Whether conducting a statistical analysis of dozens of KPIs or peering at a five-metric dashboard, improving the department’s performance hinges on the capture of timely, accurate and relevant performance information. Or, as Nagy concluded (quoting deceased statistician W. Edwards Deming): “In God we trust, all others must bring data.”