As the rest of the nation's healthcare markets grapple with HIPAA costs vs. benefits, Boston providers are finding that the case for electronic claims transaction standards is open and shut.
Who can argue with a double or triple return on investment within two to three years? That's what's being reported by some of the Boston market's dominant providers as a result of implementing the first of eight transaction standards required by the 1996 law.
And now that a regional cooperative has engineered and paid for the means to route that first transaction -- patient eligibility for treatment -- throughout the market, participants anticipate adding the others one by one at progressively less expense and with quicker payoffs.
Once it's humming, the network's ability to process more transactions at virtually no additional expense could trim the cost per transaction to less than a nickel. That pales next to the typical 35 cents to 50 cents for nonstandard electronic transactions and $5 or more using phones, faxes and forms.
At full volume, the network project is expected to collectively save participants more than $66 million per year, according to the consulting firm acting as program manager.
What's more, the whole system for exchanging patient-specific insurance details is predicated on an approach so secure that the passed information exists in cyberspace for only a moment and technically never leaves its source.
In summary, the Boston initiative is way ahead in covering all the bases of compliance with HIPAA's transaction and security standards. But while HIPAA compliance is a plus, it was not the impetus for the project, called the New England Healthcare EDI Network (NEHEN).
Area providers and payers were looking for a way out of an inefficient mess that was frittering away their financial wherewithal. They saw the standards adopted by HIPAA as a basis for building a common infrastructure and reaping a return on investment in a data network.
"HIPAA was the catalyst. (Return on investment) was the driver," says John Halamka, M.D., chairman of the NEHEN project and chief information officer of Boston-based CareGroup Healthcare System.
The situation was this: Boston had a collection of some of the nation's most highly regarded provider networks and managed-care organizations, yet high costs were running their operations into the ground. The market's healthcare premiums were 20% higher than national norms on average, but operating margins were at or near the bottom nationally.
Each payer and provider system communicated via battalions of claims-processing employees on both ends of the transaction pipeline, who completed paper forms and hashed out hitches in claims over the phone or by fax and mail.
At CareGroup, "25 cents of every healthcare dollar is spent on administration," Halamka says. That encompasses all the overhead costs of running the hospital, including the cumbersome submission and adjudication process, he says.
Between disputes over eligibility for payment and elementary mistakes in filing claims, much of the cost in that process was a consequence of not getting claims right the first time, Halamka says: "Ninety percent of the work that takes place in the (claims) department is rework."
Early organizing. Though the Boston healthcare market is fiercely competitive, leaders began kicking around the idea of forming a joint organization for coordinated exchange of electronic information as early as 1995, more than a year before passage of HIPAA.
A regional not-for-profit agency called the Massachusetts Health Data Consortium organized an initial roster of six provider organizations, eight health insurers and the state medical society into a forum for promoting electronic data interchange (EDI) and forging a consensus on ways to achieve it.
Leaders at the time agreed that a key component of such an effort was the adoption of standards for the content and technical specifications of electronic messages. And though HIPAA had not yet burst onto the scene, the Boston effort was well aware of industry standards for health insurance transactions under development by a subcommittee of the American National Standards Institute.
In fact, the first director hired by the consortium to lead its regional project, Lee Barrett, was then chairman of the subcommittee creating the standards. Those standards were incorporated into HIPAA and into the Boston effort, called the Affiliated Health Information Networks of New England.
The regional group eventually broke through competitive barriers and got most of the payer, provider and major employer organizations to work together on several technical and policy issues aimed at smoothing the way for electronic data interchange, says Greg DeBor, a Boston-based partner with the global health solutions division of Computer Sciences Corp.
But while the group succeeded in gaining common interest and outlook in areas such as insurance transactions, identification of patients and confidentiality, it was an unwieldy forum for implementation, DeBor says. The group eventually grew to 27 organizations, with three to four projects going on at a time but with limited resources and expertise, he says.
The current NEHEN effort grew out of discussions between CSC and a handful of payers and providers that wanted to jump-start the implementation process and were large enough that they "could be responsible for significant volume," DeBor says (See chart). CSC was hired in August 1997 to chart a course, and NEHEN was off the ground by early 1998. DeBor was named program manager.
Driven to collaborate. NEHEN participants by then were cognizant of the slow and stumbling claims processes and the potential for improvement. "All of us understood the value proposition quite clearly," Halamka says.
For example, in a study of one undisclosed 700-bed hospital, 26 full-time-equivalent employees were needed to handle claims involving Medicare, Medicaid and the region's three dominant payers: Harvard Pilgrim Health Care, Tufts Health Plan and Blue Cross and Blue Shield of Massachusetts.
The study revealed that 90% of the FTEs' time was spent reworking claims that had problems when initially submitted. Or in other words, it took 23 FTEs to fix the problems, which, at a salary of about $30,000 per position, cost the hospital about $700,000 a year.
In addition, the hospital employed five FTEs just to handle eligibility checks and authorizations for treatment. Though most queries are routine, workers spent from 15 minutes to 45 minutes on the phone per verification, most of the time on hold.
One plan had an average wait of 44 minutes, and providers were limited to two questions per phone call. If they had more, they had to hang up and start over, DeBor says.
The NEHEN participants figured they could save millions of dollars in administrative costs by reducing the time spent manually verifying eligibility and by increasing the accuracy of claims, he says.
By pooling resources, member organizations could develop the necessary infrastructure and share the attendant intellectual property at a lower cost than each could do alone. Payers and providers alike would benefit by sharing standardized administrative data electronically.
And by engineering the network themselves instead of relying on outside suppliers of EDI capacity, participants eventually could trim network costs to the minimum necessary to operate it instead of paying per-transaction fees indefinitely to a for-profit contractor, DeBor says.
Reducing the denial rate. NEHEN decisionmakers chose eligibility as the first HIPAA-related transaction because of its crucial influence in attaining a clean claim the first time around, says John Glaser, vice president and CIO of Partners HealthCare System, which took the lead in implementing the network's technology.
Concurrently, technologists developed techniques to nail down the identity of patients no matter where they enter the network or how many provider organizations have registered them as patients.
Those moves were important because the problems associated with patient identification and eligibility accounted for more than half the claims denials by payers.
Incorrect identification numbers or inconsistent patient names are typically responsible for a third or more of claims denials, according to research by NEHEN. The problem may be as nitpicking as an omitted middle initial, Halamka says. Eligibility problems are the root of about one in five denials -- a procedure isn't covered, or services are provided after the cancellation of a patient's coverage.
In the case of the 700-bed hospital, more than 950 claims per month were being rejected for reasons directly related to eligibility verification. The value of those claims average $368,000 per month, or $4.4 million a year. Besides the delays and costs of correcting and resubmitting claims, the hospital ended up writing off 1% to 3% of those claims per year.
The hospital's investment in the hardware, labor and communications network expenses to participate in NEHEN amounted to $250,000, plus an annual $72,000 in program management fees. For that investment, the hospital gained access to a network that could reduce waiting time per verification to fewer than three minutes and reduce the number of reworked claims by 40%. NEHEN's payer participation guaranteed that at least 70% of the hospital's eligibility transactions could be done online.
NEHEN calculated that the hospital would net savings of $164,000 the first year, for a return on investment of 150% (See chart). That includes 40% savings on salaries in a claims department with 40% less work to do, plus the costs of capital tied up in accounts.
With only minimal implementation expenses in succeeding years, the return on investment escalates to 300% in the third year, with cumulative net savings of nearly $1 million.
Precedents for the calculations come from an early test of the eligibility transaction network in the outpatient department of a hospital within the Partners system. In that pilot, claims rejections were reduced by 50% in the first four months, Glaser says.
The Partners pilot demonstrated that the electronic verification process worked, but that's not all. It also made eligibility checking cost-effective in a setting where workers were performing the checks only rarely.
Because outpatient cases generate far less in charges than inpatient admissions, the cost of checking eligibility manually wasn't worth the overhead. Consequently fewer than 5% of the outpatient cases underwent an eligibility check before the electronic system was introduced, Glaser says.
The electronic transaction process, instituted at the patient registration or check-in point, runs the check behind the scenes while clerks finish the routine questions and other administrative details of registering someone for a procedure, DeBor says.
Volume goes up, cost goes down. NEHEN's eligibility transaction costs initially were high because of low volume and an initial investment of $750,000 to plug into the equation. In the first three months of operation during the summer of 1998, transactions sent to the five major payers totaled fewer than 5,000 a month. Against amortized development and implementation costs of about $12,500 a month, the cost per transaction reached $4 or more.
By comparison, manual costs for the same volume would have been $3.75 or less per transaction. Transactions routed through a clearinghouse -- which first edits each request for the different requirements of each payer -- would have cost 50 cents each at that low volume.
But as volume increased, NEHEN transaction costs plummeted while equivalent manual costs rose and clearinghouse fees dipped only slightly (See chart, p. 31).
By March 1999, per-transaction costs in the NEHEN network had dropped to 45 cents, identical to a clearinghouse's cost, as volume rose to 41,000 transactions. Meanwhile, the equivalent manual cost had risen to $4.22.
Clearinghouse costs decrease to reflect volume discounts, but the NEHEN decrease was more dramatic:
* Per-transaction costs of a clearinghouse dipped to 40 cents by April 1999, while the NEHEN cost per transaction fell to 36 cents that month and dipped as low as 28 cents at times during 1999. Volume reached 58,000 in January 2000.
* When the clearinghouse cost fell to 35 cents in February to account for a doubling of volume to 122,000, the NEHEN cost per transaction tumbled to 16 cents.
Much of NEHEN's current costs reflect development costs still being amortized monthly through April 2003, DeBor says. Without those adjustments in the equation, the actual cost per transaction would fall to about two cents, he says.
Either electronic method beats manual measures by far, according to NEHEN research. Manual procedures not only cost more than electronic methods per transaction, but the total cost doubles as volume doubles, Halamka says. For automatic methods, "the more transactions, the cheaper it gets," he says.
For the first million transactions, the cost for clearinghouses is $1.9 million, compared with $1.4 million for the NEHEN network. Both electronic methods are eclipsed by the $5.2 million cost of 1 million manual transactions, according to NEHEN studies.
The difference is even more dramatic when volume doubles to 2 million:
* The fixed labor costs of manually handling the volume result in a doubling of total cost to $10.5 million.
* The total cost of using clearinghouses increases by 20% to $2.3 million. Only about $50,000 in implementation expenses is required to handle the doubling of volume, but service charges per transaction add another $350,000.
* The NEHEN network's cost increases by just 3%, barely more than $1.4 million. The network does not charge its members the per-transaction "clicks" that would accompany electronic services from clearinghouses, DeBor says. So the only added cost is about $50,000 for additional implementation and some staff time.
Rolling out transactions, reeling in return. A HIPAA-standard transaction set for referral certifications and authorizations began in February on a trial basis and went into production in midsummer. Like eligibility verification, the transactions bring information to bear on the appropriateness and accuracy of claims before they are sent for payment, Glaser says.
Once full attention is paid to clean claims, the NEHEN effort will begin adding transactions directly involving the claims adjudication process. The first of those -- for claims status inquiries -- is scheduled for rollout before year-end.
A combination of increasing volume from NEHEN participants and additional types of transactions will boost monthly transaction volume to nearly 400,000 by next June, according to NEHEN projections (See graph, p. 31).
The way the network is set up, it's able to keep track of only the transactions sent to payers. The eligibility and claims status transactions, however, have companion standards for payers to complete the loop back to providers. That means the actual transaction volume is nearly double what's being counted, DeBor says.
The gradual spread of transaction capability to the multiple facilities within participating NEHEN organizations is expected to further decrease costs and increase return on investment, DeBor says. For example, the 700-bed hospital that anticipates $1 million savings within three years expects triple that return when other parts of the healthcare system begin using the NEHEN system at little or no additional investment beyond the initial $250,000 expense, DeBor says.
Glaser says the electronic eligibility process, when installed through the Partners network, will capture an extra $15 million a year in revenue. That does not include any cost savings from efficiency gains, such as fewer staff, and it also doesn't include the value contribution of additional transactions now being added, he says.
Halamka says CareGroup is racking up similar benefits. "Our (return) is already three times our investment," he says.
Security structure. NEHEN's program managers must rely on polls of individual member organizations to produce statistics on benefits, because the network's security structure includes no central database and no way to "meter" the system, DeBor says.
Though the network produces thousands of communications daily between providers and payers, "the data never sits anywhere," Halamka says. At the conclusion of a transaction, "all we did over a leased line is exchange a question and an answer."
The arrangement resolves concerns about aggregating, storing and retrieving patient-identifiable information in a central location, he says. Those concerns helped trip up efforts in the mid-1990s to form community health information networks. There also were the practical problems of investing huge sums in database development and becoming inextricably dependent on the network, DeBor says.
All patient information carried by NEHEN is transitory in nature, and the data remain with the sources participating in the momentary transactions, he says. The network employs Internet technology in its operation but within a privately run communications network instead of the Internet.
A layer of software called "middleware" executes the transactions through links with existing physician practice management and hospital information systems, Halamka says. That avoids labor costs and accuracy problems caused by re-keying data into another information system, and it presents minimal intrusion into the information-technology strategies and architecture of each participating healthcare organization, he says.
The Web-based and standardized data can be meshed into members' computerization efforts however they choose. For example, managed-care data from NEHEN have been merged with a Web-based information network within CareGroup that also pulls clinical and administrative information from various healthcare software applications and assembles reports for physicians at the point of care (MODERN HEALTHCARE, Aug. 14, p. 56).