Consider what a physician needs to know and do to make diagnoses and treatment decisions during a routine day.
Presented with a patient and a complaint, the doctor has to ask pointed questions from memory to narrow the possibilities from a field of hundreds.
That means getting the patient to volunteer crucial facts about the medical ailment as well as drawing on all the knowledge crammed into the physician's head in medical school and continually supplemented by new research.
Under flat-fee healthcare, the time it takes physicians to collect diagnostic clues and scribble down findings is expensive time taken away from medical examining and problem-solving.
It's also costly to have to absorb the unnecessary treatments-or worse-involved in a wrong or delayed initial diagnosis when a physician overlooks or never considers a less obvious cause buried in a medical journal.
But a Burlington, Vt.-based medical software maker is trying to hit those problems head-on with a computerized inquiry process designed for the front lines of healthcare.
No stone unturned.
By entering a patient's unique symptoms and medical history, the software user takes a general complaint such as dizziness and triggers a range of diagnostic possibilities. Each alternative calls for more information to either rule it out or delve deeper into it.
The engine of this process is a database of medical literature from which current knowledge regarding a particular problem can be summoned and cross-matched. Once a diagnosis is known, the database can be tapped for the latest line on medical management.
The database is updated quarterly and "coupled" with structured lines of inquiry into each of 49 problem areas covered to date. The result is a "coupler" between patient-specific information and the body of medical knowledge available to physicians trying to diagnose and remedy the problem.
The company takes its name, PKC Corp., from its term for these inquiry tools-Problem Knowledge Couplers.
How it works.
A monthly newsletter called Medical Software Reviews in February gave the software a 5, the highest rating, on the extent to which it performed its stated functions.
The review follows a 1994 case study in a book called Evaluating Health Care Information Systems, Methods and Applications, which said the decision-support system "embodies an entirely new philosophy of, and puts forth a thoroughly comprehensive approach to, medical care and education."
The premise is that all known causes and treatment options for a patient's ailment should be considered, not just what a doctor has committed to memory.
"The unaided human mind cannot reliably recall all the known causes and management options, nor can it process all the information collected to thoroughly consider these causes and options," said Robert R. Weaver, author of the case study. "The computer, however, accomplishes this with ease."
One advantage is a physician isn't tempted to develop an early conclusion based on a first impression and prior experience, which may bias subsequent searches for the cause and overlook a more likely one, Weaver said.
A coupler on "dizziness," for example, identifies 65 causes and gathers data for nearly 170 clinical findings.
Getting at a cause.
The structured questioning presented on a computer screen also gets the patient more involved in communicating pertinent complaints, he said.
At a 5,000-patient family practice clinic in suburban Burlington, a physician used the coupler on vertigo and dizziness to uncover the source of a patient's problem that no one had been able to diagnose during the previous three months, according to PKC.
After the physician entered findings in the computer and launched into a review of the "cause" index, the patient piped up when she saw "migraine." She said she'd been getting migraine headaches ever since receiving an electric shock. The doctor prescribed a medication that helped her recover, PKC said.
If someone had considered that cause sooner, the patient may have been able to avoid the cost of several specialist visits as well as $875 for a magnetic resonance imaging scan, PKC said.
The computerized inquiry process is aimed at the clinical world of physicians. But now that efficiency is becoming important to cost-effective operations, PKC is emphasizing the potential for couplers tosave money in pitches to large managed-care organizations, said Layton Davis, vice president for marketing and sales.
Key to this strategy is the nonphysician in the practice setting. Davis said the coupler programs can be used by nurses or other trained workers to determine enough about a medical complaint on the telephone to decide if the patient should come in, he said.
If so, a "triage" program can direct the patient to the right provider and setting, perhaps substituting an office visit for an emergency visit or a primary-care clinician for a specialist, said Davis.
Once inside the practice, trained nonphysicians can use the structured couplers to draw out the preliminary findings from patients and rule out some of the possible causes, he said.
At the point where diagnosis and treatment options come into play, the physician takes over-armed with the computer file of paths already explored.
In his Medical Software Reviews evaluation, author Keith E. Davis, M.D.-no relation to PKC's Layton Davis-said PKC's intent to market the couplers as "intake tools" in clinics might be a bit ambitious.
Keith Davis said it "would be a steep learning curve for many front office staffs to become familiar with the vocabulary involved in some of the couplers in order to ask questions of the patient."
Other than that, he said, the program's thoroughness and efficiency are major strengths. "Approximately five to 10 minutes of clinician time is needed to complete a coupler. Such a time investment can be well within the scope of an office or clinic visit," said Davis, a family physician in Shoshone, Idaho.
Through a baseline study conducted at a participating family clinic, PKC found primary-care physicians spent 60% of their time on nonclinical functions, Layton Davis said. He declined to disclose the name of the study site.
The doctors spent about 25% of their time taking patient histories and 25% documenting findings during or after a patient visit. Davis said the couplers can reduce that load by entering the information automatically as a byproduct of the process, either beforehand by a nurse or during the doctor's evaluation.
"All that time now becomes available for more high-level care functions," he said. For health networks, that can translate to more thorough problem-solving but less time spent per patient, allowing doctors to log more visits.
PKC figures the efficiencies add up to a savings of at least $1.50 to $2 per member per month for a managed-care organization serving a defined population. The company plans to charge 9 cents to 20 cents per member per month as an ongoing fee, in addition to up-front fees for installation, training and customization for the practice's routines and clinical makeup, Davis said.
The ongoing fee pays for quarterly updates of couplers based on new information from 84 routinely reviewed medical journals. A half-dozen medical librarians work full time to do the digging, and consulting physicians review the results for accuracy, he said.
Also included in the fees are new couplers, which are being created at a rate of about 20 a year, Davis said.
For providers not organized for per-member capitation, an alternate fee formula is based on the number of providers using the system.
Symptoms `couplers' connect to diagnoses
The computer program's starting points for inquiring about patient problems can be specific or vague, just like the feedback physicians get from their patients. Here are some of the problem areas covered by PKC Corp.'s "couplers," which connect symptoms to known diagnosis and treatment.
Abnormal heart sounds
Depression, fatigue, apathy
Psychotic or bizarre behavior,
Vertigo and dizziness
Vomiting not easily explained
in an adult or older child
Source: Medical Software Reviews, February 1995