CAMBRIDGE INTEGRATED SERVICES
Wider Data Search Can Raise Red Flags For Workers’ Comp Claim Investigators
Recent technological advancements
in predictive modeling that
tap some unique alternate data
sources are helping to significantly boost
the claims evaluation process for workers’
compensation far beyond what a claim adjuster’s
gut instinct and overall knowledge
can achieve.
These new techniques help claim
managers expand the degree of information
gathered, accelerate the timeliness
of data collection and validate the instinctive
assessment of field personnel
against the likely outcomes associated
with their findings.
Moreover, while a number of factors
that ultimately impact the claim are not
known, predictive modeling can tap additional
information from other publicly
available alternate data sources to more
precisely predict the claim’s outcome—
which can considerably increase the likelihood
of an employee’s return to work.
There is a wealth of information gathered
during the handling of a workers’
comp claim—including the date and time
of the injury, where the injured worker
lives and receives treatment, how the injury
occurred, as well as relevant medical
history, prior claims, or information surrounding
the domestic setting.
However, collecting data from alternative
sources allows claim managers to
evaluate the claim from a different, if not
more comprehensive, perspective.
These additional data sources could include
information pertaining to prior or
concurrent employment, litigation, criminal records, financial information, census
data, dependent information and more.
Predictive modeling begins with building
a model based on historical data that
forecasts likely outcomes.
For example, data may show that 90
percent of all claims
where the distance between
the employee’s
workplace and home
is more than 40 miles
results in a greater than
average number of lost
work days.
Data could also
show that 90 percent
of injured workers who
live more than 40 miles
from the workplace and
who have worked with
their company more
than five years, experience
fewer than average
lost work days.
When applied correctly,
these data combinations can predict
the future outcome of the claim with a
higher degree of probability as compared
with the industry’s traditional red flags.
Predictive models allow claim departments
great flexibility in creating efficient
work flows and deploying the appropriate
internal and external resources to the right
claims at the right time.
In the workers’ comp arena, this enables
claim departments to move away from
the traditional medical-only and lost-time
model, as well as treat all claims equally
until the predicted outcome indicates the next probable step in the process—whether
that step requires no interaction with the
involved parties or increased communication
and action.
Predictive modeling eliminates claim
reporting and verification redundancies
while enabling the adjuster
to focus investigative
efforts on the more
complex claims, creating
a so-called educated
workload.
What happens to
those claims that do not
require investigation? The
decision made by claim
departments and administrators
is influenced by the
level of comfort with data
collection, data integrity
and the overall predictive
modeling concept.
A high level of confidence
indicates a certain
number of claims could be
processed through auto-adjudication or
low-level claim-handling units. Modeling
also allows for the establishment of triggers
in conjunction with more serious and
complex claims.
Triggers can be developed to indicate
auto-referral to telephonic and field case
management or, on the other end of the
spectrum, to a special investigation unit.
The same triggers can be applied to alert
claim supervisors and managers to potential
losses.
A fully developed model not only
redistributes the workload of claims between varying adjuster skill sets—from
auto-adjudication to the highly trained
specialists—but has a positive impact on
the resources needed to supervise and
manage a claim organization.
Despite the focus predictive modeling
places on investigation, data collection
and resource deployment, some of the
most insightful and predictive
data results come
from activity that occurs
after the claim has been
filed and investigated.
Missed physician or
therapy appointments,
physician changes and
other traditional events are
only part of determining
when a claim is about to
change course, and timeliness again plays
a critical role.
Repeating alternate data-mining activities
throughout the life of a claim is imperative
to obtaining successful outcomes.
For example, consider how an unexpected bankruptcy after back surgery could
impact the injured worker’s ability to return
to work. What may have appeared
to be a standard recovery could now be
in jeopardy as a result of this previously
unforeseen event.
By once again inputting the claim
through the predictive model with any additional
data related to these new circumstances,
the adjusting team may discover a
need to deploy additional resources to aid
the injured worker in a successful return
to work.
Overall, advanced predictive modeling tools provide adjusters with a valuable
means of assessing claims through a vast
array of publicly available data.
These tools enable claim departments
and organizations to significantly improve
claim handling efficiency and resource
deployment.
Predictive modeling can also help employers
lower claim-related costs by eliminating
process redundancies and help insurers
reduce overall claim durations and severity.
Furthermore, predictive modeling can
help ensure injured workers are receiving
the appropriate and care when they need
it most.
By assigning the right resources to an
injured worker’s case at the right time,
claim handlers can not only boost an employee’s
chances of recovery but expedite
their return to work—a key outcome for
all parties.
Click here to download the pdf.
Kevin Turner is executive vice president of business
development at Cambridge Integrated Services
Group in Greenwich, Conn. He may be reached at
Kevin.Turner@cambridge-na.com.
Get the latest news from
Cambridge delivered right to your desktop
