Partial reform invites docs to game the system

“Dr. Brown, the new computer order system won't accept your history of 'fever' for a pulmonary CTA on Mrs. Stern. You now have to have a reimbursable indication, like SOB, or chest pain.”

“So enter 'chest pain.'”

“I'm not going to lie. That's why Dr. White is now doing prison physicals.”

“Fine. I'll take care of it.”

“Take the medical student with you.”

“Kid, I am going to show you about real medicine, the stuff they don't teach you in medical school.”

“Mrs. Stern, you've been having chest pain, haven't you?”

“No.”

“It must hurt when you cough?”

“No.”

“You gotta have a little twinge, a slight sense of chest pressure?

“No.”

“What about when I push on your chest like this?”

“No.”

“What if I take my knuckle and grind it into your sternum….”

“OUCH!!! That hurts!”

“See, kid? Chest pain. Son, don't let any pinhead bureaucrat tell you how to practice medicine. Now I'll show you how to get an annoying patient out of the ER without getting sued. Put on that elbow-length rubber glove……”

A probing study in the February 2008 issue of Annals of Emergency Medicine, “The Law of Unintended Consequences: The Joint Commission Regulations and the Digital Rectal Examination,” reported how the commission had set up new regulations concerning digital rectal exams to improve patient care and the detection of cancer. Unfortunately, the unintended consequence was a decrease in testing for occult malignancy. It seems that doctors came to view the test as having a “cost” to perform, which was not adequately compensated.

Such has been the complicated relationship of fee-for-service and Medicare as documented in study after study. When Medicare effectively reduces fees, doctors look to compensate for the lost income by increasing utilization elsewhere. For every dollar in fee reduction, doctors recoup 40¢ through increased ordering of other tests. Doctors are smart enough to know how to game the system; so much for cost savings.

Computerized physician order entry (CPOE) has been heavily promoted during the current healthcare debate as a way to better control errors and costs. The ability to link physician ordering with decision-support software, billing requirement software, and safety features is supposed to create numerous cost-saving benefits. I don't say it won't, but we are likely to again see unintended consequences.

A University of Pennsylvania study published in the Journal of the American Medical Association in 2005 found that CPOE in-creased the risk of 22 forms of medication errors. Patient mortality in the pediatric ICU at the Children's Hospital of Pittsburgh actually increased after the implementation of CPOE. Complex new computer systems resulted in physicians responding with shortcuts, misleading data entries, and default settings that bypassed recommendations/controls in order to get what they wanted. Doctors quickly learned how to game the programs. Added to that, frequent alerts from safety programs also promote alert fatigue, resulting in the alerts being ignored or overridden, resulting in patient injury.

In a survey published in the July issue of Emergency Physicians Monthly, the variable felt to most negatively affect quality healthcare was the overwhelming threat of malpractice that ER physicians face. This is one of the prime drivers of escalating costs via overutilization, and yet President Obama has already announced that his administration and the Congress will not be addressing malpractice tort reform as part of the healthcare overhaul. This is unbelievably foolish because in that same survey, the underlying impact of malpractice cases is clearly demonstrated by the fact that 65% of respondents felt that utilization of diagnostic testing should not be considered by third-party payers when compensating ER physicians. Seventy percent felt that insurers should not increase compensation for doctors who use fewer diagnostic tests. And 59% felt physicians should not be compensated based on quality of care. And why? Because they simply don't want to be sued for missing something.

I am quite sympathetic to the ER physicians because they get sued so often for taking care of critically ill patients who they know nothing about, who may or may not speak the same language, and who may or may not even be conscious. And we act surprised when they order a CTA for fever. If we are not going to enact tort reform, then maybe someone should develop software that calculates the likelihood of being sued and the malpractice award for not having ordered a test. Now that would be a useful computer alert!

The data I feel would most help reduce my risk of malpractice and make me a more effective radiologist would be simple, accurate clinical histories. Most of the requisitions I get say either “pain” or “R/O pathology.” It is odd, but no one seems to think it worth mentioning that a patient with “abdominal pain,” as noted on the requisition, a liver transplant precipitated by metastatic colon cancer-the result of longstanding ulcerative colitis-and is on chemotherapy.

In a paper, “Radiologist use of and perceived need for patient data access,” published in the August issue of the Journal of Digital Imaging, the authors report that 72% of radiologists felt they received inadequate clinical information and 87% felt that additional clinical data would actually change or modify their radiology report. Fifty-three percent of radiologists don't seek out this information because it is too time-consuming. They argue there is a pressing need to develop systems that can collect these data in an easily accessible format for radiologists.

In the same issue of JDI, a paper presents a Google-based search engine for data collection, which might be a start. I'm just afraid that a history of “pelvic pain” would result in a porn storm each time.

If a camel is a horse designed by committee, then I am pretty sure that healthcare reform created by politicians will be a stillborn mutant aardvark. So I am not hopeful it will help me. Software for clinical data collection? It is being developed, but is not readily available unless your hospital has an up-to-date electronic medical record system that is easily accessible.

So until then, I would be satisfied if someone simply provided me with enough accurate history to know whether the patient hurts on the right or the left side. Just don't lie to get the test and tell me he has chest pain if he only has a fever.