What Does a Fish Know?

“What does the fish know about the water in which it swims?” Albert Einstein.

An existential question
Leave it to Einstein to ask the existential question regarding what we know about the environment in which we live. Of course, he was referring to the universe, but his question might just as well be applied to the Workers’ Comp world. Those of us who have long-labored in Workers’ Comp assume we have full knowledge of it. But do we? We swim in such a huge vessel of information, can we actually know much of it?

The industry is swimming in data
The Workers’ Comp industry has successfully amassed vast amounts of data. Computers become available to manage claims in the late 1970’s and early 1980’s when digitalized data was first collected. However at that time, few companies could afford to computerize. Computers were bulky, costly and the software available to run them was scant. Not until the late 1980’s were simple DOS-based claims management systems available and more affordable.

Since that time, in just twenty-five years, the industry has collected immense amounts of claims-related data. Yet, we know little of it!

What don't' we know?
Mark Twain put it this way, “What gets us into trouble is not what we don’t know. It’s what we know for sure that just ain’t so”.

We assume we know the industry in which we work. Moreover, we feel sure of the operations and actions of our own organization, along with their effects. The truth is, we actually know little of the Workers’ Comp world in which we live, but we could know much more through analytics.

Analytics is a solution
Analytics can be a forbidding notion, but it simply means data analysis, analysis of the data to gain knowledge, find meaning and direction. Data, without analytics, is useless because in its natural form, data is fragmented bits of information. Relationships are unknown. Conclusions cannot be reached. Decisions are not supported. Predictions cannot be found. Raw data keeps us profoundly unaware of the world in which we live.

When analytics are applied, the data are summarized, calculated, calibrated, and re-presented so that relationships become apparent. Conclusions and decisions are quicker, easier, and more defensible. Predictions can be made based on past experience. As a result of analytics, processes can be made more standardized and cost efficient. Computer-aided management is made real.

More specific to Workers’ Comp medical analytics, data subjected to analytics can provide even more awareness and efficiency. The spotlight can be trained on providers, treatment pathways, and outcomes. Such analyses lead to derivative knowledge about the treatments and processes that lead to successful outcomes.

Reasons for resistance
The question then becomes, if analytics is so powerful, what has kept the Workers’ Comp industry from embracing it? Among the reasons the Workers’ Comp industry has been slow to embrace analytics, a couple can be noted here.

People have long discussed the data silos in Workers’ Comp. Data silos include claims systems, bill review systems, UR systems, provider network systems, and medical case management systems, along with digitalized FROI (First Report of Injury), and OSHA (Occupational Safety and Health Administration) reports. Each lives within its own business domain that does not communicate with the others. Yet, each contains critical claim information. Essential to analytics is integrating data, however to date, appetite for data integration is tepid.

Another reason for slow adoption of analytics, particularly relating to the medical portion of claims is those most knowledgeable are not computer savvy or involved in IT. Medical professionals do not understand what is missing and the possibilities for enlightenment. They cannot imagine possibilities related to using data as a work-in-process tool. They have no experience to rely on. So they do not ask and they are not sought for input.

Nevertheless, the barriers to analytics in Workers’ Comp are easy to overcome. The industry needs to step up to analytics, particularly medical analytics, the most unenlightened portion of claims management. There is little reason to continue swimming in water we know little of.

Please visit MedMetrics for more information.

You Might be Helping Doctors Defraud the System

Medical fraud in Workers’ Comp comes in many forms. Determined abuse of the system, perpetrated by the most callous of providers is the most destructive form. Fraudulent doctors and other providers know the system and how to manipulate it to their financial benefit.

Fraudulent providers use tactics such as increasing the frequency and duration of medical services, billing at the highest levels regardless of state fee schedules, and billing repeatedly to generate duplicate payments. Even more subversive are those who add multiple diagnoses so their exaggerated billing to avoid exposure by bill review systems. Such perpetrators also shrewdly submit bills using slightly altered names and addresses so their maneuverings are not easily noticed by electronic systems.

Modifying names and addresses is an easy and effective way to obfuscate data. Computer systems are literal, meaning they accept the data as it is. Consequently, adding a comma, reversing first and last names as they appear in one field, and adding or omitting a suite number, are all common ways to cause multiple records. Each iteration of the information is treated as unique by the computer system so that each becomes a separate record representing the same person or entity. While providers are perpetrators of these data deceptions, payers often contribute to the problem.

Data quality is a people problem
Data quality in provider records is critical to evaluating provider performance. How can individual provider performance be evaluated when multiple records representing the same person are present in the data? How can individual providers be identified when several hide under the same TaxID number? How can we differentiate the good and the bad from the ugly?

Accurate data entry is critical to data quality, yet little attention is paid to the process. A policy requiring names and addresses be pulled from a drop-down list of provider records would prevent creating multiple entries caused by misspellings and similar errors. This is basic software design. For those unable to create a hard-coded list from which the data entry person can select, a copy and paste policy should be established. Manually typing the information for each bill guarantees error, record duplication, and confusion. Process management is needed to resolve the data entry problem.

Developing software interpretive rules to correct and combine multiple records is fraught with uncertainties. For instance, a software rule could be written to interpret name reversals by looking for a comma indicating the last name is first. However, the comma is often not present, so even more confusion is created. Commas and periods, present or not, in names and address are a common issue of data quality and very difficult to correct programmatically. It’s a people problem.

Unique identifier
Still, the best way to resolve the problem, whether it results from provider billing practices or data entry at the payer level, is to require unique provider identifiers such as NPI or state license numbers. NPI (National Provider Identifier) is a system required by CMS (Centers for Medicare and Medicaid Services). Individual providers must have an NPI number to be reimbursed by Medicare. Workers’ Compensation payers should require the number, as well.

Most medical providers currently have NPI numbers because they want to be reimbursed by CMS for non-Workers’ Comp services. NPI numbers in the bill would eliminate the disguise offered by deliberate or unintended data duplication and confusion.

Fighting medical fraud
Fighting medical fraud is frustrating and elusive. But it isn’t only providers who contribute to the problem. Clean and complete provider records where the data are entered exactly the same way for every bill received from a provider will go a long way to correcting the problem.

Evaluating provider performance and rating providers analytically depends on correct individual identification. Multiple records in the data for the same provider generated by sloppy data entry practices simply perpetuate and exaggerate the problem.

Learn more about Provider Performance management.