Advertisement

Doctors Fear Patients Will Suffer Ills of the Millennium Bug

Share
TIMES STAFF WRITER

Not long ago, Beverly Hills cardiologist Debra Judelson began entering the medical history of an elderly patient into her new computer system, holding her breath as she tapped in the woman’s birth date: 1895.

“I wondered whether the computer would tell me she was 3 years old,” she recalled. But she was pleasantly surprised. “A 103-year-old female?” it asked. Judelson exhaled and happily typed in: “Yes.”

But Judelson acknowledges that her feeling of relief may prove premature. Like many of the nation’s other 700,000 doctors, she is waiting anxiously to see how the millennium, and its computer bug, will affect her patients and her practice.

Advertisement

Most computers made before 1997 use two digits instead of four to keep track of years. As a result, they cannot distinguish between the years 2000 and 1900.

And when the clock moves from 11:59 p.m. Dec. 31, 1999, to 12 a.m. Jan. 1, 2000, many in the medical community fear that computers may misplace prescriptions, refuse medication, produce faulty test results or deny insurance coverage.

No one suggests that the patient records kept by doctors are suddenly going to evaporate in the crossover. The concern, instead, is that they could be corrupted by their interaction with software and hardware that helps a computer keep track of time.

The problem is exacerbated by the growing number of doctors using computers to transfer and retrieve patient information to and from insurance companies, pharmacies and hospitals. Computers at any of these locations might mishandle dates, thus posing one of the biggest risks from what is known as the Y2K computer bug.

The American Medical Assn., in a survey conducted last year, found that almost 90% of the nation’s doctors are using computers in their practices and 40% already have logged in their patients’ medical histories. The numbers appear to be growing, in part because of the burgeoning demands of the managed-care era, with referrals, drug prescriptions and insurance claims relying increasingly on the sharing of information among computers.

“Computers add and compare, and if a computer is not getting what it’s expecting, it does what it wants to do,” said Frederick Kohun, an expert on Y2K computer issues at Robert Morris College in Pennsylvania. “What concerns me--really frightens me more than anything else--is the possibility of bad data, or data being changed or corrupted, that misinformation could get spliced in, resulting in bad medical decisions.”

Advertisement

Dr. Donald J. Palmisano, a New Orleans surgeon and AMA official who has been studying the problem, agreed.

“If the Y2K problem causes these systems to fail, critical data could be lost, or worse, unintentionally and incorrectly modified,” he said. “Even an inability to access critical data when needed can seriously jeopardize patient safety.”

What could happen?

Plenty, experts said, citing some potential scenarios:

* A patient on critical medication--insulin or immune-suppressing anti-rejection drugs, for example--could walk into a pharmacy for a prescription and the pharmacists’ computer could refuse to fill it, insisting that the customer already has enough medication. Or the dosage could pop up changed--larger or smaller--which could have serious consequences.

Or the pharmacy itself--whose inventory is tied into the computer’s internal clock--could fail to reorder the medication, causing inadequate stocks.

“If I were someone on a life-sustaining prescription, I would stock up before the end of the year,” Kohun said, adding that insurance companies “are beginning to break down their own policies and rules on this, knowing who will be liable--them--if somebody dies because they cannot get their drugs. They are starting to allow people to stockpile.”

* Test results could be inaccurately transmitted among different computers. For example, a cholesterol reading of 193 could pop up as 300 because of incompatible computer codes, resulting in confused or inappropriate treatment decisions. Or a laboratory machine could make a similar mistake, putting a blur on a CAT scan reading, perhaps resulting in an unnecessary exploratory operation.

Advertisement

Or, if a Y2K-induced error causes a piece of laboratory equipment to skip a function or perform a function twice, “a patient could get the lab results of the patient who preceded or succeeded him or her, with potentially adverse consequences,” Dr. Kenneth W. Kizer, Veterans Affairs undersecretary for health, recently told a hearing of the Senate Special Committee on the Year 2000.

In the old days, most doctors just used paper to keep track of their patients. Computer record-keeping was rare. When doctors began using computers, programs were designed to accommodate century lines--because patients’ lives so often cross them. It was common for doctors to treat patients who were born in the previous century as well as others expected to live into the next.

But today, many physicians and medical facilities use computers to communicate referrals, transmit prescriptions and create databases aimed at improving patient care and increasing efficiency. The result is that personal medical histories have been integrated with other systems--and these latter systems were not necessarily created with the same foresight about patients.

“The problems came when they integrated with financial records,” Kohun said. “When you start interfacing finances with clinical stuff, that creates a basis for problems. The billing systems are at fault because they are based on a two-digit date field.”

Kohun predicted that problems will arise from internal computer clocks--where date functions are embedded--and in various computer software programs that manage the patient data that doctors store.

“The real problems will come when different insurance companies start talking to different doctors and hospitals and different systems,” he added. “The best thing that could happen in some cases would be if the whole system goes down--that is, the system just won’t work. This would be better than the transmission of bad data.”

Advertisement

Patient record programs were designed by health care professionals “who understood health care needs,” while financial systems “were designed by computer nerds who only wanted to minimize costs and get the job done,” Kohun added.

“My fear is this potential for erroneous data passing” by a computer and the possibility that it won’t be caught in time by a human, he said. “What if nobody’s paying attention? A cholesterol reading of a ‘million’ is not realistic. I know that. You know that. And a doc knows that--but a machine doesn’t.”

Frank Reilly, a consultant to the Senate Special Committee, said that not much is currently being done to address the potential crisis because most insurance companies and other third-party payers “are devoting all their time trying to solve their own internal problems” by bringing their own hardware and software into compliance--a necessary first step before “they can test what happens with outside systems.”

“Doctors’ offices are the unknown,” he said. “Nobody knows yet what will be done--it’s a big question--but they can’t do anything until they bring their own systems up to date. They need to get their computers and their software programs compliant first, which is what they are doing now, before they can start dealing with the outside world.”

Times staff writer Jube Shiver Jr. contributed to this story.

Advertisement