Deadly Radiology Error at Tufts has an Evidence-based Cure
On August 31, 2014, the Boston Globe published a report about a medical mistake that killed a woman undergoing surgery. This particular kind of error is called “misadministration”; it is where radioactive substance (such as dye) used in imaging is injected wrongly into the patient. In short, a “misadministration” refers to administering radioactive substance with one or more of the following errors: 1) wrong patient, 2) wrong substance, or 3) wrong site.
In this particular case, the surgeon asked the nurse to bring him a dye from the pharmacy during surgery to inject into the patient’s spine. The surgeon wanted to image the spine during the surgery. The pharmacist did not have this dye, so the pharmacist provided a similar dye to the nurse. Unfortunately, this similar dye had a warning on it saying specifically not to inject it into the spine.
The surgeon clearly feels very bad about this. He immediately took responsibility and apologized to the family, and the Globe described “cognitive bias” as a possible cause.
“Cognitive bias” means the surgeon looked at the warning label, but was so used to seeing something else in that part of the label that he did not even see the warning.
There is not a lot of evidence in the scientific literature that “cognitive bias” causes misadministration. Richard Fitzgerald stated it best in his landmark paper, “Error in Radiology“:
…the experience of safety cultures in other high-risk human activities has shown that a system approach of root cause analysis is the method required to reduce error significantly.
In other words, the root cause of radiology errors is the lack of a safety culture, and that is the responsibility of leadership and management, not a surgeon, a nurse, and a pharmacist.
Here is evidence-based advice to Tufts leadership from the scientific literature about how to address errors in radiology such as misadministration:
- Create Processes Aimed at Reducing or Preventing Errors: This is recommended by Pinto et al. in their 2012 paper. For Tufts, this likely includes getting the pharmacist, nurse, and surgeon to communicate with each other about the patient and about problems when one substance is ordered but only an alternative is available.
- Develop an Effective System to Detect Errors Quickly to Reduce their Impact: According to the Globe report, “It took until the next day [after surgery] to figure out that [the patient] had received the wrong dye”. The report does not speculate on whether earlier detection would have saved this particular patient, but in general, detecting errors as soon as they happen “buys time” to reduce their impact on the patient.
- Develop a System to Manage Errors Appropriately: The surgeon immediately apologized to the adult sons of the patient who died as a result of the error, and said “a mistake was made”. Eight months after this, the sons received a letter from attorneys “denying the surgeons … and pharmacists were negligent in her death, or injured her,” so they filed a lawsuit against the hospital and 12 of its staff.
- Have an Ongoing Active Surveillance Process: This process should be supported by educational programs, “morbidity and mortality meetings“, and a “comprehensive and respected root cause analysis process” needs to be conducted on every error. Understanding the root cause allows leadership to redesign processes so as to remove these root causes.
In response to this particular error, Tufts has added “more layers of protection”, even though they stated already that “nurses and surgeons are supposed to verify they have received the correct medication”, and in this case, that already did not happen. The new “layers” include more detailed written orders except in emergencies. And hospitals including Tufts are working on ways to apologize and compensate patients who are clearly victims of a medical error, rather than forcing them through the malpractice process.
The following questions remain:
- Did Tufts really do a “comprehensive and respected root cause analysis” of what went wrong in this case? A real root cause analysis would probably not have found “cognitive bias” to be the answer, because that essentially blames the surgeon, not the system.
- Will these new “layers” help? Already one policy was not followed that helped lead to this error. It is hard to believe that adding more layers will magically result in the development of a “safety culture”.
- Has Tufts improved their surveillance system so that errors are detected in a shorter time? Nothing to this effect was mentioned in the Globe article. Do they really have a program of ongoing surveillance that detected the error, or did the clinicians just happen upon it themselves when the patient continued to be ill?
Read More about Medical Errors
Updated August 31, 2014