CLAIMS PROCESSING
Among the many challenges that all providers of automated data capture solutions
have faced are the unique problems that are encountered in the
design and development of effective claims processing systems.
While significant success has been achieved in environments where
the physical design of the forms that must be processed can be
controlled, such as a Tax or Revenue agency, a review of a representative
sample of claims received by most organizations for processing
will typically show that there is no such thing as a “standard
form”.
Impression Technology encountered this situation when it was
awarded a contract to design, develop and implement the data capture
system for the Commonwealth of Pennsylvania Medical Management
Information System, MMIS. MMIS, which is operated by EDS Corporation
for the Department of Public Welfare of the State, began production
operations in late 2003 and it is designed to process up to 25,000
Medicare Claims and other documents submitted by both medical
providers and the residents of the Commonwealth.
The iCapture solution within MMIS is designed to maximize automated
data capture from the 13 types of forms that are received daily.
These include HCFA-1500, UB-92, dental claims, multiple types
of prior authorizations, multiple types of invoices and adjustment
documents. The analysis of the actual documents being received
revealed that even when the information was submitted on a “standard
form”, such as a HCFA-1500, there were significant enough
variations on the design and actual printing of these forms to
degrade automated ICR/OCR recognition when traditional data capture
techniques were employed. Further compounding the problem, the
amount of information on an invoice might vary from one line to
many lines on more than one page, therefore there was clearly
no “template” for the identification of areas where
automated recognition of data was desired.
Impression addressed these and other related issues through the
development of an ‘Auto-Template’ capability within
iCapture. While other systems use a ‘best fit’ approach
by maintaining a library of commonly seen templates, the iCapture
solution uses advanced image pattern recognition techniques to
identify where individual characters are on a document, to “grow”
these characters into fields, and to identify the location of
these fields for subsequent ICR/OCR recognition. This Auto-Template
process is performed in real time and the resulting electronic
form definition is much more accurate than traditional methods,
therefore yielding higher recognition accuracy than what can be
achieved previously.
In addition to the Data Perfection mode (ICR/OCR corrections),
the solution also takes advantage of the two other modes of editing,
Key-from-Paper and Key-from-Image, allowing a Data Entry Operator
to select the most effective mode for the type of batch of documents
being processed.
Another unique feature of the MMIS iCapture system is the ability
to display color images to the Data Entry Operator as the system
takes full advantage of the “dual streaming” features
of the Kodak Series 800 Document Scanners. These scanners are
cabable of simultaneously producing both a color image for display
and editing functions and a bi-tonal dropout image that is used
for OCR/ICR processing. The color image is converted to a TIFF,
CCITT Group 4 compressed file in the MMIS Document Management
System. The ability to display the color image, an exact replica
of the original form, has proven to be a definite advantage in
the data entry and editing process while the TIFF image storage
minimizes on-going file storage requirements.
For additional information on this project and Impression Technology’s
Claim Processing solutions, contact Bruce Lechner at 925.280.0010,
ext. 20 or sales@impression-technology.com.
|
