Welcome to Understanding Link Analysis. The purpose of my site is to discuss the methods behind leveraging visual analytics to discover answers and patterns buried within data sets.

Visual analytics provides a proactive response to threats and risks by holistically examining information. As opposed to traditional data mining, by visualizing information, patterns of activity that run contrary to normal activity surface within very few occurances.

We can dive into thousands of insurance fraud claims to discover clusters of interrelated parties involved in a staged accident ring.

We can examine months of burglary reports to find a pattern leading back to a suspect.

With the new generation of visualization software our team is developing, we can dive into massive data sets and visually find new trends, patterns and threats that would take hours or days using conventional data mining.

The eye processes information much more rapidly when information is presented as images, this has been true since children started learning to read. As our instinct develops over time so does our ability to process complex concepts through visual identification. This is the power of visual analysis that I focus on in my site.

All information and data used in articles on this site is randomly generated with no relation to actual individuals or companies.

Analyzing Medical Fraud

One of the most prevalent and costliest fraud is medical provider and billing fraud. It is also one type of insurance fraud which incorporates numerous other criminal activities into it. Through analyzing and investigating medical fraud, you will find

  • Staged Accident Rings
  • Money Laundering
  • Legal Provider Fraud
  • Drug Offenses
  • Tax Evasion
  • Wire and Mail Fraud
Medical fraud is probably one of the easiest investigations to migrate into RICO statues as the activity incorporates so many other criminal activities.

Medical fraud analysis and investigation is very complicated as it is so paper and record intensive. I can't imagine putting together a medical fraud investigation without visual analysis as there are so many records to incorporate and rationalize to detect the activity.

The best place to start is through strategic or proactive analysis of billing data. There is a wealth of information contained on HICFA medical billing forms that can be utilized to detect patterns and relationships between doctors, clinics, patients and the procedures they are billing.

Medical Billing Analysis

The first place I am going to start is by patterning what medical providers are billing me to find indicators of fraud such as over billing, billing for services not rendered and improper or illegal procedures.

This task can be accomplished in two different ways, one by utilizing desktop applications such as Microsoft access or excel and the other by leveraging visual analysis. Both of the methods assume that you have already data based your medical billing into a format which can be incorporated into your visual analysis software or brought into desktop applications.

Lets start with utilizing excel to analyze billing patterns. For this scenario I am going to download six months of medical billing from a problematic geographical area, in this example we are going to use Naples Florida. We can limit this data set down a little more by eliminating or focusing on specific specialties that are more prone to fraudulent activity. For the purpose of this exercise I am going to focus on chiropractors billing from Naples Florida to bring my data set to a manageable size.

Below is an example data set downloaded or data based from my medical billing.

By utilizing pivot tables, I am able to get an overview of what this group of providers in the area is billing for and can detect any irregularities.

Right away I spot some billing irregularities, there is a Chiropractor who is billing 99205 for several claims. For those who are not familiar with medical billing, for a person to be billed this particular code, you should be close to dead laying in a hospital, not walking into a chiropractic office.

I can also look in the pivot tables for the same person visiting multiple clinics in different claims. This person is either really unlucky or this can be an indicator of staged accident activity. Another option is to look at patterns of billing between chiropractic offices, radiology clinics and medical equipment supply companies which can be an indicator of patient brokering, particularly if the same person owns all three locations.

Through simple of use of pivot tables from one set of data, I can look at the results in several different ways without having to run any additional queries to locate targets of opportunity for further investigation.

Medical Provider and Patient Analysis

Now that I have examined all of my medical billing for Naples Florida, I have narrowed down a few providers which may be involved in medical fraud or other nefarious insurance fraud activities.

At this point I want to produce a visualization of the clinic and the patients to look for signs of fraud such as medical providers who are associated with numerous clinics and patients who are treating at them, particularly if those patients are involved in multiple medical claims.

I am going to import my medical data into my analytical software to detect these relationships. Like the other import specifications we have put together, there are unique identifiers in medical billing which we can leverage for our visualization.

For clinics in my data, I am going to utilize the tax id as the unique identifier. Next, I want to link the medical providers to the clinics. The unique identifier for the doctor is going to be the license number which is captured on the HICFA form, this will prevent two doctors with the same name from being created as one entity.

Next I am going to link the patients to either the clinic, the doctor or to both depending on the type of claim it is. For auto related medical claims, I have found it is best to link the patients to the clinics, for medicare or Medicaid fraud, linking the patients to the doctors is most effective, that is because in auto related medical fraud, normally the accidents are also fraudulent, staged by runners who work for a group of clinics.

In both scenarios I am look for the migration of doctors across numerous clinics, and the patients following them to their associated clinics which is a good indicator of fraud, such as in the visualization below

What we are seeing here is multiple doctors associated with multiple medical clinics. Along with this pattern we also see the patients following to each clinic the group of doctors are associated with.

I can then leverage visual analysis to examine the medical billing to see if the same CPT codes are being used for each patient who is treating at the group of clinics

From here I am going to incorporate in the legal providers which are representing the patients treating. Medical fraud normally requires several people filling roles in the scam:

  • The runner who works for the clinics and brings in the patients through staged accidents or solicitation
  • The lawyer who is associated with the scheme who "represents" the claimants in order to maximize the benefits from the insurance company.
  • The doctor who "treats" the patients
  • The clinic who houses the scheme and bills the insurance company.
To understand the relationships through all these moving parts, we can visualize the events from the medical provider, lawyer and clinics in a timeline to find patterns.

Ultimately, the goal of the investigation is to find the top of the pyramid, normally the person who is getting all of the money. Rarely are the claimants in charge of the scheme and are normally paid very little considering the profit. In the case of medical fraud it can be the doctor, lawyer or the person who finances the clinics and following the money is the only way to determine who is ultimately responsible.

This is where we shift from our medical fraud analysis to our financial analysis included on the site. In this article we have developed a query to make a data set of medical data for analysis, utilized pivot tables to develop targets for investigation and used visual analysis to discover the scheme.

Aside from using visual analysis to help you investigate medical fraud, it will also help your counsel, prosecutor and jury more easily understand the fraud to see it in a visual representation. While medical fraud is one of the most complicated fraud schemes to investigate, it's one of the hardest to explain to a criminal or civil jury, one of the reasons why prosecutors are reluctant to present it. The way visualization helps you understand and organize your investigation will ultimate serve to help the people you are explain it to.