The healthcare industry suffers from an inability to systematically fix its incomplete and inaccurate provider data. This problem not only prevents the industry from achieving optimal efficiency and automation, but also has larger-scale implications in terms of hindering healthcare initiatives, such as consumer-directed healthcare, pay-for-performance reimbursement and establishment of effective treatment protocols.
The healthcare industry suffers from an inability to systematically fix its incomplete and inaccurate provider data. This problem not only prevents the industry from achieving optimal efficiency and automation, but also has larger-scale implications in terms of hindering healthcare initiatives, such as consumer-directed healthcare, pay-for-performance reimbursement and establishment of effective treatment protocols.
Today, an estimated 50 percent to 70 percent of provider records contain data errors. Multiply this across the billions of annual transactions that occur using faulty information and it is easy to see how the industry loses an estimated $26 billion annually due to bad provider data. This includes costs incurred from billing errors, reissuing checks, late penalties and other fines, and additional administrative expenses resulting from fixing or otherwise manually coping with the problem.
Although statistics may be abstract, The Wall Street Journal reported in 2006 that consumers need only consult a health plan’s provider directory to experience first-hand how erroneous information can pose significant challenges when trying to find a doctor. Why hasn’t an industry as sophisticated as healthcare been able to solve something as elementary as maintaining basic information—ID number, name, phone and address—that uniquely identifies each provider in the system?
Analyzing how bad provider data has infiltrated the healthcare industry can help the industry develop a sophisticated, comprehensive cure.
Origin: No Standards to Begin With
The origin of the provider data problem is multifaceted and can be tracked to the healthcare industry’s need to address escalating costs. As the industry tried various strategies to control double-digit increases—chief among them group health insurance, managed care and consumer-directed healthcare—the healthcare system became more complex and convoluted.
In the fee-for-service market, dynamics dictated free access and skyrocketing costs. The advent of managed care introduced restricted access and added new provider obligations, such as obtaining treatment authorizations. In addition, managed care organizations negotiated with provider groups to keep costs down. Claims had to be scrutinized to ensure fee and billing compliance according to the agreed upon terms. Healthcare became big business, with shareholders and an investment community eager to see the industry diminish its high administrative overhead.
As financial pressures increased, the claims process was targeted as a key practice that could benefit from automation. At the same time, providers pressured legislatures to pass prompt-pay legislation. To streamline operations, health plans began to introduce electronic transactions in billing and claims and to rely on computerized codes to identify the health services rendered and the parties involved in transactions.
As part of these electronic transactions, health plans assigned providers unique provider identifier numbers. Since providers typically worked with several health plans, they were likely to have a different identifier number for each plan. In some cases, a single identifier was issued for multiple providers. The two initial problems were a lack of standardized provider data and no effective way of maintaining the accuracy of this information across all systems.
Propagation of the Problem
Over time, provider inaccuracies accumulated, while data management efforts failed across the industry. Examples abound: In any given year, approximately 20 percent of physicians change their addresses; 5 percent change their status (license, sanctions, death); and 30 percent change their affiliations. When handling these changes at a system level, distinguishing between new providers or variations of names and addresses for pre-existing records is complex.
Inconsistencies are propagated through the loading and dissemination of data across corporate IT infrastructures, as well as across the greater healthcare industry. As changes are reported from disparate sources, various information systems and multiple users across the enterprise, information comes together in very inconsistent ways. The result: bad provider data.
For the most part, the industry relies on physicians to report changes and updates. Although most providers can access a computer and the Internet, they mainly use these technologies for medical research, relying on antiquated tools—such as the phone and fax—to report informational changes, tools notoriously known to perpetuate delays, data errors and lost information.
The other method health plans rely on for provider data is the credentialing process with the National Committee for Quality Assurance. Although credentialing is essential to ensuring quality medical professionals, verification is tedious and expensive. Because a physician’s credentials must be verified with each plan and are valid for up to three years, the information provided often becomes outdated over time.
Some health plans call or send faxes to providers to update information, but providers are contacted to verify their data only once every year or two, and even then the process is time-consuming and costly. Although plans claim it takes two weeks to reflect a provider record change, in actuality, it may take as long as six months for the update to be made throughout the enterprise.
This problem has less to do with one factor than it has to do with the complexity of the healthcare system, combined with remnants of various legacy operations and processes that require greater automation.
Inefficiency, Higher Administrative Costs
Poor provider data has significant impact on operational efficiency and the industry’s bottom line. The problem costs an estimated $26 billion a year. A number of factors contribute to this.
Foremost is the lost opportunity in automating claims. The current rate of auto-adjudication varies across health plans, as data errors result in a higher number of pending claims, a slower claims cycle, costly manual interventions and late payment penalties. By some estimates, the cost to process a claim manually can range from $8 to $15, compared with $.25 to $.35 to automatically adjudicate a claim. Also, when payments to providers are returned because of an outdated provider address, manually researching the correct address and reissuing a new check can cost a healthcare payer up to $20 per transaction.
Poor data has created additional strains on the provider-payer relationship. Inaccuracies hamper efficient verification of eligibility, enrollment of providers, plan management and provider contracting. Member relations can also be strained when searching for a doctor or specialist on a PPO directory, only to discover that the provider moved months or years before, all because of outdated provider records.
In addition, health plans lose millions of dollars in revenue, as they receive fewer PPO hits and consequently less PPO revenue. They also experience increased losses due to fraud and abuse. Automated fraud detection is severely limited when bad data exists, and savvy criminals can leverage the prevalence of inaccuracies to channel funds to phony provider addresses. Such scams can remain undiscovered for years.
Complications in compliance may arise, resulting in late-payment penalty fees, inaccurate 1099s for providers leading to IRS fines, and penalties from CMS for paying sanctioned doctors, whose status was not reflected in their databases.
Within the grander scope, the entire healthcare community loses because it is unable to optimize provider data to understand performance, quality care and clinical outcomes. With more accurate information, payer/provider relations could improve, and the principles of consumer-directed care and pay-for-performance reimbursement could advance. Consumers would also then have the information necessary to select appropriate physicians. Likewise, pharmaceutical companies would be able to select the right providers to participate in clinical trials. And, the industry could reap the benefit of analyzing clinical outcomes for improved treatment protocols.
A Silver Lining, Not a Silver Bullet
The provider data challenge, however, is not all doom-and-gloom. There are some optimistic trends that help to alleviate provider misinformation.
The growing trend of consumer-directed healthcare is having a beneficial effect. Consumers, having experienced the frustrations of confirming eligibility and processing claims when missing or erroneous provider data exists, now demand improved provider information. Such economic market forces are leading to new and stronger accountability.
Consumer-directed healthcare, however, is in its infancy, and in the middle of a crucial dilemma—what comes first, the chicken or the egg? Consumers may influence supply and demand, but how can they make correct decisions without accurate provider information? Incorrect information will continue to hinder this dynamic until more systematic changes are made to data standards and data management.
The other critical factor bringing about positive change is the “administrative simplification” provision under HIPAA, which requires healthcare transactions to use a standard, unique 10-digit identifier for providers—the National Provider Identifier (NPI). The NPI will ensure that each provider has one unique identifier to be used in all electronic transactions with all health plans.
Physicians and health plans must comply with the new NPI requirement by May 23, 2007. This means providers must obtain their NPIs, while health plans and clearinghouses must be able to accept NPIs in connection with electronic transactions covered by HIPAA. The challenge for all parties is ensuring that the underlying legacy provider data is accurate and up-to-date in order for an NPI to be as effective as its creators intended.
Concerns, Federal and Beyond
While this is a great plan to address data standardization, it is not without its potential problems. For instance, federal health plans are supposed to participate in the initial NPI set-up and coordination. This might present issues within Medicare’s already overburdened administration. Another concern is how the industry will manage the transition from multiple legacy identifiers to a single identifier environment.
As if that wasn’t enough to worry about, many wonder if CMS will be able to coordinate large-scale dissemination of NPIs in a timely manner. Complicating the issue of preparing for NPI compliance even more is the fact that, as of this writing, CMS has still not published the NPI dissemination rules. Ongoing delays could create a domino effect in the industry’s inability to comply.
Although the deadline is less than a year away, the same compliance date applies to both providers and health plans. Health plans will need to assess and update legacy information systems, administrative processes, reference files and forms to comply with the NPI, and to put into place some form of continuity between old provider identifiers and new NPIs. In other words, they need to get a jumpstart. But how can they get ahead of the pack, when everyone’s charted to go the same speed?
Providers will have their own IT preparations to make. For instance, practice-management software and hospital systems may require tweaking or significant re-engineering to accommodate the new NPI standard.
Past these logistical and system issues, NPIs only address half of the provider data problem: standardization. It leaves the other half of the issue—a process to continually manage data changes to ensure accuracy—largely unresolved. Even after NPI standards are in place, physicians will continue to move and make other informational changes, which presents an ongoing challenge for healthcare payers.
Preventing Future Propagation
On the surface, incorrect provider data may seem rather trivial, especially when compared to more pressing healthcare challenges—such as the growing number of uninsured, the aging of America or the rise of chronic medical conditions.
To use a computer analogy, we might compare the provider data problem to the ever-present threat of computer viruses. This comparison is effective on several levels. First, a computer virus in itself has no cognizance. Yet, when it is introduced to a complex and computerized community like healthcare, with its thousands of providers and payers all using proprietary systems, a hypothetical provider data “virus” could easily propagate itself throughout the entire system.
Contaminated healthcare provider data requires an ongoing data management approach similar to today’s virus-protection software, i.e., continual updates to ward against future corruption. Within the financial services industry, a concerted effort at customer data integration has yielded dramatic improvements and efficiencies. It may not be too much to imagine the same improvements and benefits for the healthcare industry. However, defining this type of ongoing strategy will require continued diligence and cooperation throughout the healthcare industry, even beyond the impending NPI standards.
[email protected].
