Everyone Wants Analytics--Whatever That Is...

by Karen Wolfe

It seems everyone in Workers’ Compensation wants analytics. At the same time, a lot of confusion persists about what analytics is and what it can contribute. Expectations are sometimes unclear and often unrealistic. Part of the confusion is that analytics can exist in many forms.

Analytics

Analytics is a term that encompasses a broad range of data mining and analysis activities. The most common form of analytics is straightforward data analysis and reporting. Other predominate forms are predictive modeling and predictive analytics.

 

Most people are already doing at least some form of analytics and portraying their results for their unique audiences. Analytics represented by graphic presentations are popular and often informative, but they do not change behavior and outcomes by themselves.

Predictive modeling
Predictive modeling uses advanced mathematical tools such as various configurations of regression analysis or even more esoteric mathematical instruments. They look for statistically valid probabilities about what the future holds within a given framework. In Workers’ Compensation, predictive modeling is used to forecast which claims will be the most problematic and costly from the outset of the claim. Predictive modeling is the most sophisticated and usually the most costly predictive methodology.

Predictive analytics
Predictive analytics lies somewhere between data analysis and predictive modeling. It can be distinguished from predictive modeling in that it uses historic data to learn from experience what to expect in the future. It is based on the assumption that future behavior of an individual or situation will be similar to what has occurred in the past.

Credit score
One of the most well-known applications of predictive analytics is credit scoring used throughout the financial services industry. Analysis of a customer’s credit history, payment history, loan application, and other conditions is used to rank-order individuals by their likelihood of making future credit payments on time. Those with the highest scores are ranked highest and are the best risks. That is why a high credit risk score is important to purchasers and borrowers.

Similarly, Workers’ Compensation claim data can be collected, integrated, and analyzed from bill review, claims system, utilization review, pharmacy (PBM), and claim outcome information to score and rank-order treating physicians performance. Those with the highest rank are the most likely to move the injured worker to recovery more quickly and at the lowest cost.

Making analytics predictive
Both predictive modeling and predictive analytics deal in probabilities regarding future behavior. Predictive modeling uses statistical methods and predictive analytics looks at what was, is, and therefore, probably will be. For predictive analytics, it is important to identify relevant variables that can be found in the data and take action when those conditions or events occur in claims.

Industry research

One way to find critical variables is to review industry research. For instance, research has shown when there is a gap between the date of injury and reporting or the first medical treatment, something is not right. That gap is an outlier in the data that predicts claim complexity.

Data search
Another way to identify key variables is to search the data to find the most costly cases and then look for consistent variables among them. Each book of business may have unique characteristics that can be identified in that manner.

Actionability
Importantly, predictive analytics can be used concurrently throughout the course of the claim. The data is monitored electronically to continually search for outlier variables. When predictive outliers occur in the data, alerts can be sent to the appropriate person so that interventions are timely and more effective.

For example, to evaluate medical provider future performance, select data elements that describe past behavior. Look at past return to work patterns and indemnity costs associated with providers. If a provider has not typically returned injured workers to work in the past, chances are pretty good that behavior will continue.

Where to begin

For organizations looking to implement analytics, those who have already made the plunge suggest starting by taking stock of your organization’s current state. “The first thing you need to know is what is happening in your population,” says Rishi Sikka, M.D., senior vice president of clinical transformation for Advocate Health Care in Illinois. “Everyone wants to do all the sexy models and advanced analytics, but just understanding that current state, what is happening, is the first and the most important challenge.”

Best results
It is important to note, that the accuracy and usability of results will depend greatly on the quality of the data analyzed. To get the best and most satisfying results from predictive analytics, cleanse the data by removing duplicate entries, data omissions, and inaccuracies.

Analytics-informed medical management
For powerful medical management informed by analytics, identify the variables that are most problematic for the organization and continually scan the data to find claims that contain them. Then send an alert. Structuring the outliers, monitoring the data to uncover claims containing them, alerting the right person, and taking the right action is a powerful medical management strategy.

Karen Wolfe is the founder and President of MedMetrics®, LLC, a Workers’ Compensation, analytics-Informed medical management and technical services company. MedMetrics analyzes and scores medical provider performance and offers other online apps that link analytics to operations, thereby making them actionable. karenwolfe@medmetrics.org

 

How to Operationalize the Insights of Analytics

by Karen Wolfe

Organizations are anxious to execute analytics, but their leaders are baffled about how to apply the knowledge gained. Portraying colorful graphics depicting the results of analytics in executive reports has zero effect on costs or outcomes. In order for analytics to have an impact, they must be fused into the operational process.

Terminology

Simply stated, analytics is the term used to describe data analysis of any kind. Analytics should not be confused with predictive modeling which is also analyzing data, but the goal of predictive modeling is to predict what is likely to happen when a specific set of circumstances occurs. Stated differently, predictive modeling predicts what claims are at risk for specific costs and conditions.  In predictive modeling, highly advanced statistical tools are used to identify the set of conditions that are then considered predictive. Predictive modeling is one form of analytics.

Analytics is the broader term applied to data analysis. Aside from predictive modeling, it is designed to provide the organization with knowledge and insight into their business processes. Garden variety analytics are used to identify trends and cost drivers. However, neither predictive modeling nor any other analytics can change organizational behavior or outcomes.

Operationalize analytics

Analytics (data analysis) can be powerful as a means of understanding business processes, organizational strengths, and especially cost drivers, but that is not enough. Analytics offers understanding, but that is only the first step. To impact the organization, its workers, and its clients, the insights gained from analytics must be transformed into timely operational initiatives and enforced through work-in-process electronic tools.
 

According to Rachel Alt-Simmons, SAS, “As competitive pressures increase the need for organizations to master analytics, internal analytic teams have increased their statistical sophistication, but are struggling to operationalize their insight.”[1] The problem is they are missing the very significant step of translating the findings to the operational process.

 

Actionable analytics

Analytics, regardless of the variety, must be linked to operations to make them actionable. The dots must be connected between analysis, decision, and action. The way to do that is to translate knowledge to action using designed technology.

Technology-powered

Workers should not be expected to interpret sophisticated mathematical analyses, but they can act on the derived re-portrayed information. An example is comprehensive data analysis of medical provider performance re-presented as a score or rank compared to their peers. Rather than struggling with multiple analytic indicators of performance, workers should make informed decisions based on interpreted, understandable information found immediately at hand.

Analytic delivery framework

Rules-based technology combined with continuously monitored historic and current data, can send workers early notification of adverse conditions in a claim. Workers can be alerted of poorly performing providers, questionable prescriptions, severe diagnoses, comorbidities, cost benchmarks, and a myriad of other conditions of known risk as they occur in claims. Moreover, the technology can enforce organizational standards by including action steps (procedures) with the alerts.

Organizations using sophisticated predictive modeling initiatives should also take the next step by applying their results to the analytic delivery framework. Regardless of the level of statistical sophistication, the information derived must be delivered in a practical way to claims adjusters, nurse case managers, and others who make decisions and take action regarding claims. Only then will analytics empower workers, impact costs, and improve outcomes.

Analytics inspired—technology powered

Results of analytics must be implemented consistently and structured so that the intended cost control initiatives are achieved. Analyzing the data and delivering the results of analysis to workers will inform decisions and actions, thereby creating maximum value for customers, constituencies, and the organization itself.
 

Learn about MedMetrics analytic delivery framework or contact karenwolfe@medmetrics.org

 

---

[1] Alt-Simmons, R. Balancing Creativity and Control: Bringing Process Discipline to Predictive Analytics. SAS. January 21, 2013.

Predictive Analytics in WC Made Easy and Affordable

By Karen Wolfe

It’s a safe bet that claims will not have a happy ending if the treating physician has a history of being associated with poor claim outcomes. In fact, physicians rated poorly in analytic studies based on past performance are 100% predictive of high costs and inferior outcomes in future claims where they are involved. The question is, how can those providers be identified so they can be avoided?

Applying analytics

Whether the cause of poor performance is misunderstanding Workers’ Compensation or deliberate fraud, the claim results will be dismal. Nevertheless, in order to analyze provider performance, one must know where to find the data, what to look for, and how to apply the knowledge gained from analysis to achieve improved results.

Data can offer a clear picture of actual provider performance. Evaluating physician and other provider performance is a matter of scrutinizing the data using industry research to learn what to look for. In fact, leveraging published industry research is the way to skip the laborious and expensive regression analyses and other predictive modeling methods.

Industry research reveals what to look for

Exposing substandard providers is a matter of integrating and analyzing the data to understand the course of the claim and the providers who were involved. Selecting the data items to monitor can be guided in the first instance by industry research. Organizations such as NCCI (National Council on Compensation Insurance), CWCI (California Workers’ Compensation Institute), WCRI (Workers’ Compensation Research Institute) continually publish their research based on data they collect from members. These organizations offer research regarding medical issues causing cost escalation in the industry, and usually make results available from their individual websites.

Search

Academia and other organizations produce and publish research, as well. The best way to access other research is to use Google or other search engines to find research studies regarding specific issues and interest areas. For instance, if the concern is low back pain, simply use Google to find research and scholarly articles on the topic as it relates to Workers’ Compensation.

Indicators of performance

When the indicators of performance are identified, they can be tagged in the data to analyze individual providers. Providers associated with a preponderance of negative indicators will fall into the lowest class category. On the other hand, those whose results are exemplary will rise to the top—best in class.

Where to find the data

Billing data tells the story of diagnoses, treatments and the billed amounts. However, billing data by itself is never broad enough in scope to evaluate providers because it tells only a part of the story. Claim adjudication level data tells another part of the story. It describes the actual paid amounts, return to work, the amount of indemnity paid, and whether legal was involved. But there is more.

Analyzing PBM (Pharmacy Benefit Management) data is imperative. Overuse of prescribed narcotic pain relievers is now a major concern in Workers’ Compensation medical management. Prescribing excessive opioids is unconscionable, but the guilty are often not identified and avoided as they could and should be.

Provider performance should be scored by claim outcome combined with costs and other factors. Unless the initial injury was catastrophic, return to work following a workplace injury is often a function of medical management that should be measured. Analyzing multiple data indicators from disparate data sources is powerful in describing physician performance. It is also objective and fair.

Integrating the data for analysis

Any one Workers’ Compensation data source by itself is inadequate for the purpose of evaluating provider influence. Only the broad scope of data concerning a claim can provide a clear picture of the claim and provider culpability in outcome. Therefore, collecting the data from its various sources (billing or bill review, claim adjudication systems, and pharmacy data), then integrating current and historical data are crucial steps in provider performance analytics. The next steps are identifying, evaluating, and monitoring the data elements that are indicators of performance both from the medical and Workers’ Compensation viewpoints using research as a guide.

Link analytics to operations

Analytics results of any variety that remain in graphic form, in a brochure, or pinned to a wall are useless in the effort of actually containing costs. The findings must be functionally applied to operations to make them actionable. Information regarding best (and worst) in class doctors identified through the methods discussed here must be made available to network managers and others in a usable form. Moreover, the information should be specific, current, dynamic, easily accessible, and contain objective supportive detail. The work of analytics is not complete until its results are operationalized and actionable.

Learn more about MedMetrics provider performance analytics or contact karenwolfe@medmetrics.org.