Healthcare news this week was speckled with examples of how the healthcare system—no doubt with the best of intentions—sometimes ends up doing more harm than good.
To be sure, there are well-energized agencies, movements, systems and regulations, laws, and hundreds of thousands of well-meaning people trying to prevent unintended consequences. But as the saying goes, "stuff happens."
First, the U.S. Food and Drug Administration released its long-awaited findings on the mishaps that caused 385 patients at five hospitals in California and one in Alabama to receive harmful overdoses of radiation during brain perfusion scans intended to rule out stroke. Of the 385 patients, 260 patients received the scans during an 18-month period at Cedars Sinai Medical Center in Los Angeles.
"The FDA found that when properly used, the CT scanners (made by GE Healthcare and Toshiba America Medical Systems) did not malfunction. Instead it is likely that the improper use of the scanners resulted in these overdoses," the FDA said this week. The FDA was not forthcoming on what exactly those users did improperly to cause the overexposures; though that would be helpful to know.
Many of the overexposed patients suffered burns, rashes, and now run a higher risk of getting cataracts. Many of them worry they now have a higher risk of getting cancer, as well. But the FDA noted that it "found no evidence that the manufacturers were involved in modifying any of the hospitals' scanning protocols so as to cause overexposures."
The advisory makes clear that the FDA now says imaging professionals responsible for conducting CT procedures "must define the equipment parameters needed to provide an adequate image at the lowest dose."
It also wants facilities to take seven actions:
- Assess if any patients received excess radiation doses from CT perfusion scans.
- Review radiation dosing protocols for all CT perfusion scans to ensure correct dose.
- If more than one scan is performed on a patient, make sure any subsequent dose is adjusted.
- Make sure dosing protocols are followed every time.
- Check the display panel before each scan to make sure the dosage is appropriate.
- Make sure technologists are trained for the specific scanner and imaging protocol.
- Make sure technologists are trained on dose-saving features such as automatic exposure control. If the user activates AEC without carefully reviewing and adjusting the associated parameters, a default setting may occur that may not be appropriate for that scan, leading to an overexposure.
The FDA also made note that the scanners themselves can be improved to make sure the incidents don't reoccur. In a letter Monday to the Medical Imaging and Technology Alliance, Jeffrey Shuren, MD, Director for the FDA Center for Devices and Radiologic Health made five suggestions that he hopes the alliance will convey to manufacturers.
For instance, the manufacturers might install "Pop-up Notification at Threshold for Deterministic Injury" alerts to redesign certain aspects of the machines and fortify the education that goes along with their sale or lease.
Shuren also suggested "reorganizing or developing additional equipment operation information that users could readily access, understand, and apply to reduce the possibility of inadvertent application of unnecessary amounts of radiation."
The remedies "surfaced during our investigation and reflect issues that appeared to contribute to the possibility of unintended high exposure," Shuren wrote.
Dialysis Centers
The second story about unintended harm has to do with dialysis centers. Researchers working with a Centers for Disease Control and Prevention project found patients who are put on dialysis early on in their disease process, in an effort to deal with more minor symptoms of their disease, end up with a nearly triple rate of death in one year than patients who enter dialysis later on.
The authors suggested several possible mechanisms for what might be causing higher mortality rates, such as rapid lowering of blood pressure that could cause unanticipated cardiac events.
But as the authors noted in their report in the Archives of Internal Medicine, () the percentage of patients who were put on hemodialysis much earlier in the course of their disease increased from 20% in 1996 to 52% in 2008. "This trend has occurred despite no substantial evidence of benefit of early initiation of hemodialysis and many recent publications suggesting potential harm," wrote Steven Rosansky, principal author and senior research fellow at Dorn Research Institute in Columbia, SC.
In an accompanying editorial, Kristen Johansen MD of the San Francisco VA Medical Center, concurred, saying research to date does not support current practice of initiating dialysis—a costly and time-consuming and dangerous process—in patients who are not yet that sick.
"The bar for these symptoms has been dramatically lowered in recent years with no data to support a benefit to patients," she wrote.
Problems in dialysis quality were also highlighted in a series of investigative stories published this week by ProPublica, which promises to post center-specific quality data that may provoke long-overdue improvements in care for end-stage renal disease patients, and a more thorough examination of why disparities exist.
Prescibing Errors
The third example of how hospitals ended up doing more harm than good involves a report showing major geographic variation throughout the country in the frequency with which physicians prescribe potentially harmful medications to patients age 65 and older.
Yuting Zhang and colleagues at the University of Pittsburgh Graduate School of Public Health looked at 306 Dartmouth Atlas referral regions in the U.S. as they were in 2007. The maps show five levels of frequency of dangerous drug prescribing practices for two types of prescription pitfalls:
- prescribing drugs that are potentially dangerous for seniors in general, and
- prescribing drugs that have potentially harmful drug disease interactions.
I thought this was particularly important, because clearly there must be a reason why states like Alabama, Louisiana, Tennessee, Oklahoma and parts of north Texas show up as having the highest numbers of hazardous Medicare prescribing.
Likewise, there is probably a good reason why parts of Wisconsin, Minnesota, New York and most of New England have far lower rates of high risk drug prescribing.
These three reports are just a few examples of why I believe there's a critical need for comparative effectiveness research. We need to know first and foremost that the type of care, the timing of that care and the frequency of that care are in the best interests of the patient.
Not the dialysis centers, or the radiologists, or the CT scanners, the drug manufacturers, or the doctors. It's about the patients.
That's been the adage in medicine since the Hippocratic Oath: First do no harm.