Health Care Efficiency: No Miracle Cure

Health care reform ideas are everywhere these days. As a radiologist, I'm all for controlling costs—but I challenge the assumption that national electronic medical records (EMRs) will magically make providers more informed and automatically decrease unnecessary tests. It sounds great, but the idea and, at least, its partial implementation are nothing new. We know that most US hospitals have already used electronic records for years in their own systems—yet useful information is ignored every day, costing us all. An extreme, but genuine, example involves a woman who recently came to a hospital complaining of abdominal pain. Her urinalysis showed blood. The emergency-department physician asked no further questions and reflexively ordered a CT scan, questioning whether there was a kidney stone. Had he, his secretary, or the CT technologist looked at the computerized records, one of them would have noticed that the patient had undergone the same study, for the same symptoms, only two weeks before—and it was normal. So were multiple scans earlier this year and in preceding years, all listed in the information system and on the PACS. All told, the patient had an astounding 35 CT scans in five years, all essentially normal. When I pointed this out to the emergency-department physician, he was embarrassed over the waste of resources. The number of normal scans did, at least, help him confirm the diagnosis that he was already suspecting: She was a narcotics addict, coming back for more painkillers. Her 35 patient encounters, all with CT scans, mean that well over 100 employees over the past few years could have questioned this woman who kept getting expensive tests—but get them she did. Tainting urine with blood is an old trick to an experienced abuser of narcotics, and many tests, like her CT scans, are painless—at least until the bill comes due. One CT scan, of course, might cost hundreds of dollars. Add hospital charges, laboratory tests, and medications, and her total charges over the past few years are scary. Don’t forget her huge cumulative radiation dose. She doesn’t have cancer now, but she might someday, and who’s to say what might have caused it—or who will pay for its treatment. An Institutional Dilemma What about most patients, who aren’t faking? Many still get unnecessary tests because ordering physicians and other health care workers fail to consult records, and because clinicians might not understand what a test can or cannot do. It happens all the time. It might be an abdominal ultrasound to rule out gallstones, even though there were none last month. We average a couple pointless renal ultrasound studies a day for renal insufficiency to rule out obstruction, in patients with recent unenhanced CT or MRI studies that already show no urinary dilation. One colleague recently read a negative leg Doppler study and gave his impression that there was no deep-vein thrombosis (DVT). When the study was repeated only a few days later for persistent leg swelling, and was again negative, his impression read: still no DVT. Occasionally, physicians even order tests to evaluate organs removed years ago, unaware that their patient no longer has an appendix, uterus, gallbladder, or spleen. Such information would seem obvious to note in a patient's history, but sometimes that gets overlooked. Patients might consult multiple physicians, including specialists focused on their area of expertise. Information sometimes simply slips through the cracks, or is not looked for—even when it’s only a few clicks away from the user. The general public might understandably think this behavior careless, or even lazy. Health care workers might blame busy work environments and the fear of malpractice suits if no test is ordered. Don’t forget the patients: Why aren’t they always more forthcoming? Many sick or demented patients don’t remember or can’t communicate. Not all physicians speak understandably. Some patients are passive or don’t know any better. Reasons for poor communication are complex, but the result is the same: Unnecessary, expensive tests are performed every day, despite internal electronic records most places. Radiologists are not entirely blameless here. We rely a great deal on the clinical history provided for a study. We also know that such histories are often incomplete, occasionally not quite accurate, and sometimes woefully inadequate. EMRs have been greatly beneficial to those of us who look up patients' laboratory results and their surgical and medical histories as a part of image interpretation. It takes time, but might transform a limited history to something much more informed and specific. Knowing a patient's blood-test abnormalities and his or her surgery/oncology history makes our interpretations more complete, accurate, and meaningful. The same applies to old-study comparison. Nationally linked records would be great for seeing studies done far away, yet integrating myriad systems, with the complexities of access and passwords, seems daunting. Many practices already incorporate both hospital and outpatient facilities, but their current PACS are not always linked. The patient of the outpatient oncology service might later show up at the hospital emergency department, or the recently discharged inpatient might present for outpatient follow-up care. Readily available remote access might already mean the difference between having no prior studies available and a more informed and useful reading, provided we log on to check. Education and Communication How do we improve the situation? Continued efforts to educate referring clinicians on the strengths and limitations of imaging tests are key, time consuming as this might sometimes be. Ideally, these efforts foster productive, meaningful interactions among colleagues jointly working to optimize patient care. Electronic-ordering programs with automatic choices for clinical history might be useful for billing, but limited with regard to information. Abdominal pain might make an acceptable billing code, but is a far cry from a note indicating pain, anemia, and fever after colon-cancer resection, for example. Radiologists and technologists might clearly prefer that relevant information be provided by the ordering clinician. When this does not happen, but the information is available, we might have to make some effort to retrieve it. Ensuring that people use available data is a tougher challenge. Most of us are inherently motivated to be accurate, if not thorough. Whenever hospital or outpatient practices implement or update electronic record systems, physicians should be made aware and provided with instruction. Record systems are, of course, best when consistent and easy to use. Access should ideally be quick and simple, with a minimum of ever-changing passwords and remote access codes (while recognizing the need for patients’ privacy). Linking PACS and information systems would be ideal, so that looking up a white–blood-cell count, biopsy, or surgical report is as effortless as looking at a prior radiology study or report. Then what, for physicians who still avoid their patient homework? Suddenly, the exercise is even more challenging; the prospects, less enjoyable. Without straying into territory with harsher tones and consequences, most of us might agree on this: If we do not regulate ourselves more by optimizing our efficiency with the tools available to us, someone else might volunteer to try to do that for us—radiology benefit management companies being an obvious example. Empowering technologists might also serve well. When clinicians order a test that is not particularly indicated—or a study that was recently done, and might not need to be repeated—we can encourage technologists to ask us, in advance, if performing the study seems reasonable. To get closer to the root of the problem, we can maintain a presence with our referring clinicians. Physicians who know us will generally feel more comfortable asking us how best to image a patient in a particular clinical scenario, and will be more receptive if we suggest an alternative study, or perhaps none at all. Continued education of our referrers, individually and in conferences, will help build trust and confidence in our imaging guidance and in our mutually shared goal of optimal patient care. We can also work with IT specialists so that maximal clinical information is available to us with a minimum of computer obstacles. National medical records could indeed make health care more efficient, for those who take the initiative. Ensuring that people consult available data is the bigger challenge. It will take considerable effort to find that miracle cure.Cullen Ruff, MD, is a radiologist in private practice in Fairfax, Virginia; a professor of radiology at Virginia Commonwealth University; and author of the forthcoming book Looking Within.