Health Insurance Fraud Gets
Easier; So Should Stopping It
Electronic
medical
records are
being hailed
as a way to
save money,
but they’re
also making it
easier to com-
mit fraud; text
analytics and
sophisticated
data mining
software can
help stop it.
By
Julie Malida
Apharmacist takes a 30-day prescrip- tion and inputs it as four weekly pre- scriptions—quadrupling the Medicaid ispensing fee; a fraud ring obtains a list of Medicare patients who recently underwent hospital outpatient proce- dures and starts submitting phony claims, threatening the lives of par- ties who expose the scheme; the local news reports that five teachers who
find a pharmacy to dispense large prescriptions for narcotics cov-
ered under the school health plan subsequently sell the drugs.
What do these three stories have in common? They are all real
fraud incidents.
Analytics
are the key
to spotting
billers who are
submitting
medical records
that are fabricat-
ed or overstated
to support
erroneous or
fraudulent
claims billing.
-Julie Malida, SaS
age than the patient is), EMRs don’t
just make it easy to commit fraud;
they can lead to medical errors that
threaten the patient’s lives and cost
more precious medical dollars. It can
also cost individual patients money if
they later apply for life insurance and
their medical records are littered with
mistakes that make them seem older
or sicker.
Keeping information integrity
We know that medical office administrators are sometimes coached
to make sure office visits are coded properly for maximum reimbursement. And no one wants a doctor who spends 45 minutes
with a patient to get reimbursed as a Level 1 visit. But electronic
medical records make it easy to preset all visits at the highest reimbursement level, with the onus on the harried provider to select
a lower, more appropriate level. Pre-filled templates can lead to
“checking all the boxes” on the exam sheet to “upcode” the visit
further, generating the highest charge. And it is very easy for offices to cut and paste information from one record into another.
When coupled with demographic errors (putting in a higher
necessary technology
Advanced analytics are necessary to
find and effectively root out fraud,
waste and abuse. But the type of ana-
lytics that spits out a crude list of pos-
sible investigation targets isn’t much
use. Payers, both government and
private, need:
Text analytics. Very critical to find-
ing problems in the “notes” section
of EMRs. When every diabetic patient
over the age of 65 has the exact same
wording on their records from one
practice, it’s a tip-off that fraud or er-
rors might be occurring.
Intelligent outlier detection and
predictive modeling. A high-volume,
24-hour pharmacy is likely to have more weekly prescriptions
than a less busy pharmacy. Analytics that take into account vol-
ume and population demographics will do a better job of avoid-
ing false positives.
Social network analysis. This helps claim payers, for exam-
ple, find patterns of ownership in different billing entities. If
one entity is under investigation, others with related ownership
can face additional scrutiny. Social network analysis also helps
unearth the kind of multiparty fraud schemes that involve the
theft of patient records, abuse of prescription coverage and col-
lusion among doctors, patients and pharmacists.
The examples at the beginning of this column all were (or could
have been) discovered earlier by the appropriate advanced analytics.
Investigators will always be involved, and so will tips—but the basis of
early identification of the cases is the use of advanced analytics.
EMRs have tremendous potential to lower health bills and
improve care, but they also provide new options for fraudsters.
Advanced analytics are the key to spotting billers who are sub-
mitting medical records that are fabricated or overstated to
support erroneous or fraudulent claims billing.
Julie Malida is the principal for health care fraud in the Fraud and
Financial Crimes Practice at SAS.
For more about fraud search “Insurance Fraud Race” at XXX;JOTVSBODFOFUXPSLJOH;DPN
insurancenetworking.com
september/october 2011 insurance networking news 33
