Not unlike the stale piece of cheese in the fridge that no one has the courage to remove, version 9 of the clinical modification of the International Classification of Diseases codes (ICD-9-CM) long ago passed its expiration date. That might not be a problem, except that the code is the basis of the prospective payment system, upon which is built the foundation for U.S. healthcare reimbursements.
Caregivers who code incompletely leave money laying on the table. Those who do it wrong face regulatory sanctions and fines. But the task of keeping up with rapid--and often illogical--changes to ICD-9-CM while navigating often-inconsistent reporting rules has become so onerous, and the scrutiny on coded data so intense, that many qualified coders have left the field. The time has come for ICD-9-CM to go the way of the abacus, says Rita Scichilone, director of coding products and services at the American Health Information Management Association (AHIMA), Chicago.
"We do need to replace it," she says. "It doesn't meet the needs of the 21st century." ICD-9 replaced its predecessor, ICD-8, 25 years ago. Until then, medical classification systems generally were replaced once a decade because of rapidly changing medical terminologies, procedures and technologies. "But because of the investment in the reimbursement systems and data gathering in the United States, we've hung onto it," Scichilone says.
Ironically, a replacement--ICD-10-CM and the related procedure classification, ICD-10-PCS--waits in the wings. ICD-10 was adopted in 1994 as the standard by the World Health Organization and is used throughout Europe. When it is adopted by the United States, the new system will be a lot more precise in identifying conditions and diagnoses--and, unlike ICD-9-CM, it is unlikely to run out of space.
ICD-9-CM, on the other hand, is a code in chaos. Maintenance of the diagnosis classification system is handled by the National Center for Health Statistics, Hyattsville, Md., while the Centers for Medicare and Medicaid Services, Washington, D.C., maintains procedure classifications. In an attempt to keep ICD-9-CM afloat, new diagnoses and procedures often have been grouped into classifications that don't logically fit with basic underlying conditions. Some terms and classifications are outdated. "There are a lot of discrepancies within the ICD-9 coding system," says Chris Ritchie, director of health information services at St. John's Health Center, Santa Monica, Calif. "I would say the biggest problem with it is the lack of specificity."
ICD-10 will largely resolve that. Its alphanumeric structure and seven-character code length will allow for greater detail than the five-digit numeric ICD-9-CM code structure. "I don't think there's much of an argument. It needs to be replaced, and as quickly as possible," Scichilone says.
Making do
That's not going to happen for another couple of years. According to AHIMA, the federal notice of proposed ICD-10 rulemaking may occur in February or March 2005. If a final rule gets adopted, a 24-month public-comment period would follow. Since healthcare organizations can't wait to be reimbursed, they have been finding ways to cope with ICD-9-CM.
"The difficulty that I see," says Wes Rishel, an analyst at Stamford, Conn.-based Gartner, "is that it's an art form looking at all the facts associated with a case and picking the right codes. And the difference between almost the right codes and the right codes is a fair amount of money."
The past several years have seen the emergence of remote-coding application service providers (ASPs) like eWebHealth, Alpharetta, Ga., and Nauvalis Healthcare Solutions, Nixa, Mo. They don't alleviate coding problems, but they do help relieve the coder shortage by using document-imaging technology to scan medical records and progress notes for secure transmission over the Internet, allowing coders to work at home. "It allows the organization to be able to retain coders," says Beth Friedman, eWebHealth's director of marketing. "It eliminates the open positions."
Another ASP, CodeCorrect, Yakima, Wash., helps coders figure out how to navigate regulations that further complicate ICD-9-CM coding tasks. "All the hospital coders, the charge-master people or the department directors need is an Internet connection and their browser. And they can put in an ICD-9 code and get all of the regulatory information available about that code," says Kerry Martin, CodeCorrect's CEO. "And they can get [it] for whoever they're sending the bills to, or get the national information."
There is also an increasing trend toward outsourcing of coding work, either to centralized offsite--and increasingly, offshore--locations, or onsite at provider facilities. Among the companies engaged in domestic outsourcing are Precyse Solutions, King of Prussia, Pa.; KForce Inc., Tampa, Fla.; and C.H.I.S. Inc., Palm Desert, Calif.
C.H.I.S. President Mark Perlmuetter says the coder shortage has led to widespread outsourcing, following the pattern of medical transcription work and radiology. "It is more efficient, easier and less management than the recruitment and maintenance of employees to do that work," says Perlmuetter, whose company employs 15 coders. "Our coding quality coordinator has told me that if he could have 10 more people, he could have 10 more people working. The trouble is finding the qualified people."
In fact, the Chicago-based American Hospital Association reported in 2001 that the vacancy rate for coder positions stood at 18 percent, even higher than the 11 percent nursing shortage.
Code-amatic
All of these options help, but in one sense they all represent a compromise, says Rishel. "There are things you can do to make coders more efficient, there are things you can do to make the job more attractive to coders by letting them work at home," he says. "But the answer is more than just rearranging the deck chairs on the Titanic."
The best answer would be to automate the codes, he suggests, mapping them to a standardized medical nomenclature--perhaps SNOMED. Then coding could be integrated into a hospital's internal computers. "What we think is going to lead to automatic coding is systems that are set up to capture the data right during the process of giving care," says Rishel.
Vendors that are marketing electronic systems with both a computer-based patient record and accounting applications on the same platform--he mentions Cerner Corp., Kansas City, Mo.; Siemens Medical Solutions, Malvern, Pa.; and Epic Systems Corp., Madison, Wis.--are best positioned to take coding automatic.
"Then you've got more of an ability to sort of structure--in the outpatient environment, for example--the visit template. So you've collected all the right information and the code, and all those things come out automatically at the end," Rishel says.
Mary Stanfill, AHIMA's professional practice manager, prefers the term "computer-assisted coding" to "automated coding," but she agrees it is likely the wave of whatever is left of ICD-9-CM's future. "Folks think of automated coding as the computer just doing it all, and there is no human intervention," she says. "I don't believe that we have tools that can do that to ICD-9."
However, tools like structured data inputs and natural language processing can go a long way toward making the ICD-9-CM coder's job easier. But for computer-assisted coding to reach fruition, the pieces of the medical record to which coders must assign code must themselves be electronic. And there's the rub. "The problem," says Rishel, "is that's beyond the reach of a lot of practices."
Kevin Featherly is news editor of Healthcare Informatics.
E-COMMUNICATION
Electronic consults coming, but will they pay off?
Like it or not, electronic communication between physicians and patients is on its way, driven by high consumer demand and spurred by the still-elusive promise of getting paid. "I believe that ultimately all physicians will be using this in their practice," says Daniel Sands, a Harvard medical instructor, practicing Boston physician and chief medical officer at secure messaging company ZixCorp, Dallas. "To some extent," he asserts, "resistance is futile."
Many doctors are resistant, notes Mark Bard, president of Manhattan Research, New York. Many fear opening an online physician-access channel to patients when doctors already are seeing 20 patients a day. More palatable are Web-based communications that allow patients to be "interviewed" by checking off boxes on a questionnaire, resulting in machine-readable data that can be fed directly into a practice management system or an electronic medical record (EMR), he notes. In either case, adoption remains low.
"It's not the physicians saying they never want to use it," Bard says. "But if they open up [electronic] communication, they want some kind of value coming back to them."
The survey says...
Statistics from both Manhattan Research and New York-based Jupiter Research show doctors are slow to jump aboard the e-communications train. Manhattan Research indicates only 8 percent of physicians routinely use email to communicate with patients. Three-quarters of physicians say they would be willing--if they get paid.
Jupiter's numbers are similar. A survey published in October 2003 showed that only 9 percent of physicians email patients more than five times a week. Their reluctance, according to an earlier survey (see Fig. 1: not shown), stems from concerns about medical liability, compensation, privacy and other factors. A follow-up survey in January showed only 3 percent of consumers engaged in online clinical consultations, though more (17 percent) used the Internet to renew prescriptions or seek general drug and disease information.
The numbers mask pent-up demand, according to Jupiter analyst Monique Levy. In 2002, 65 percent of consumers indicated they were interested in communicating online with physicians. The bottleneck is money; few payers are reimbursing doctors for online consults, while patients tell Jupiter researchers they wouldn't pay more than $10 (see Fig. 2: not shown). Though she doesn't expect much movement over the next 18 months, Levy says payers eventually must pick up the slack. "It would ultimately have to be a co-pay," she says, "because the doctors are not going to be willing to take $10 as a fee for service."
Emerging models
At least one company is having some success attracting payers. RelayHealth Corp., Emeryville, Calif., has partnered with eight health plans in six states, including Blue Shield of California; Blue Cross Blue Shield of Massachusetts; Blue Cross Blue Shield of Tennessee; Anthem Blue Cross Blue Shield of Colorado; Group Health Inc. of New York; and two others that Eric Zimmerman, RelayHealth's executive vice president, says can't yet be named publicly.
"The doctor has to be compensated by making the Web visit a covered benefit," he says. But payers have no interest in paying for unstructured, text-based email consultations. "Our perspective is that email is wrong for medicine. It's a very blunt instrument," Zimmerman says. Instead, the company's service pivots on secure, Web-based clinical questionnaires. RelayHealth provides 140 such branched, algorithm-driven clinical interviews; it is this service that the eight payers are piloting.
"It provides a concise clinical summary that supports a clinical decision about how best to treat that patient," Zimmerman says. "The goal here is to make the doctor more clinically productive."
In addition to RelayHealth, the other two major players in the online medical consultation space are ZixCorp, which markets the secure-messaging portal MyDocOnline, and the AMA-backed Medem Inc., San Francisco. Levy expects players from the EMR and practice-management spaces eventually also will compete.
In some ways, Medem represents the converse of RelayHealth's approach, basing its model on the belief that patients will pay to communicate online. The company markets directly to practices, setting up free Web and email services and allowing practices to charge patients whatever they wish for consultations--while keeping a $2.50 cut for itself. If doctors choose not to charge anything for the consultation, says CEO Ed Fotsch, Medem receives nothing.
Fotsch disagrees that email consultations are a financial dud. He says many doctors charge patients $20 to $25 for Medem-linked email consultations. Patients largely are willing to pay for genuine medical advice online, he says, though most doctors don't charge for simple emailed questions about prescription refills and the like. "I don't think any serious, educated person would argue against the efficiency of email," Fotsch says. "It's a model that works."
Digital conversion
Count among the e-consult converts Lyle Berkowitz, a physician and medical director of clinical information systems at Chicago's Northwestern Memorial Physicians Group. Many in his group use both email and a self-built Web portal to communicate with patients, some of whom happily pay for the convenience: "A lot of our patients are busy," he says. "They would gladly pay [for e-consults] out of their own pocket."
The Northwestern Memorial Physicians Group may soon beef up its online capabilities, either using RelayHealth or IQHealth, a secure patient communications system from Kansas City, Mo.-based Cerner Corp.
"I'm often very critical and pessimistic about these new technologies that come along," Berkowitz says. But not this one: "I think it's just going to continue to grow and get bigger. It's a very easy win."
Sands, who has used email in his practice for many years, agrees with Berkowitz. "I think that eventually doctors are going to start to use this technology because we're going to continue to do a better job educating them about the benefits. There is going to be patient demand and doctors who start to use this technology are going to be at a competitive advantage."
Adds Sands, "My message to doctors is: If it hurts, you're doing it wrong."
