In May 2016, the Industrial Designers Society of America (IDSA) approved the formation of the Patient Safety Taskforce, commissioning it to develop a strategic plan to apply design skills to help healthcare providers minimize medical errors in diagnosis or treatment that experts believe have become the third-leading cause of death in the United States. The group has a specific goal to reduce adverse events 33% in 5 years in collaboration with the healthcare community. In short, they are trying to figure out if industrial designers can fix patient safety.
Dr. Richard Griffith, D.Eng. Sci. E.E., M.D., serves as the clinical liaison for the group and also led the charge in forming the Taskforce. A retired Associate Professor of Anesthesiology at Albany Medical Center in New York, Griffith is also the former World-Wide Hypodermic Medical Director at Becton Dickinson. He answered some quesions from Medical Design & Outsourcing via email to discuss the state of patient safety and the role that industrial design might play.
MDO: What are the causes of errors in heath care?
Griffith: Dr. Marjorie Stiegler on the anesthesia faculty at the University of North Carolina has said that 70% of medical errors come from cognitive mistakes. We define cognitive mistakes as ones that the individual easily recognizes as errors, retrospectively. She says that most errors in healthcare do not come from a lack of knowledge, but from distraction or failed perception, or perhaps failure to gather all the available information before acting (jumping to a conclusion). But we need to make it clear that healthcare folks are well intentioned and well meaning. Errors arise inherently from the way our neurophysiology processes operate.
Sean Hägen, who leads the IDSA Taskforce, researched a number of taxonomies for errors. Cataloging mistakes can get very complicated, but I think we need to keep it simple. In my mind we have three headings: diagnostic, treatment, access. Diagnostic errors simply mean the patient’s treatment does not address the real condition. Treatment errors include medication errors and procedures that go astray. Access includes the failures in healthcare that come from the patient never getting into the system or not getting the therapy because the patient’s social setting blocks the therapy in some manner. We do not normally include the access element in patient safety but legislators like Elizabeth Warren have pushed for doing so because, in reality, it injures and kills people also.
MDO: Can you explain further what you mean by diagnostic errors, and perhaps offer some solutions?
Griffith: The scope of mal-diagnosis looms large. Numerous publications suggest that in something like 20 percent of autopsies the pathologist makes a diagnosis that does not match the pre mortal one. There is plenty of evidence of the problem. In “The Digital Doctor,” Robert Wachter has written an elegant paragraph about conflicting medical adages: one says the diagnostician should seek a single cause for all symptoms while another says the patient can have as many diseases as he or she likes.
Renowned internist, Dr. Larry Weed, complained 50 years ago that his residents stopped taking a history as soon as they could guess a diagnosis and set about then to confirm their guess. He was right and he proved it by having nurses record the patient history and create a problem list without any attempt to reach a diagnosis. They consistently defined two to three times more problems than his residents. Better definition of the problems fosters better care and ultimately more accurate diagnoses.
A design opportunity lies in elevating primary care into a diagnostic mega-engine. To a large degree we have allowed primary care to slide into the role of deciding only if the patient has real disease or not. When a patient does have real disease they go to a specialist who makes the final diagnosis and treats or refers for treatment. To elevate primary care we need to arm those practitioners with new tools.
Scientists have created the ability to electronically perform hundreds of antigen-antibody tests on a drop of blood (this could happen right inside the primary care office) for perhaps less than $10 using a chip that is read electronically. One can imagine a library of these chips designed for various symptoms or complaints, but in fact we cannot do that because our patent office has allowed people to patent molecules they discovered but did not invent. We face a legal quagmire. Congress or the courts could change that to make it easier for primary care offices to get the right diagnosis quickly for patients.
Also designers could focus on new tools to make the physical examination more reliable. Why can’t a smart stethoscope precisely time and grade a murmur or consistently find pneumonia? Primary care offices would need more powerful digital logic tools to sort information and guide them; and these tools really need to be visually or orally interfaced, not keyboard oriented, to make them efficient and robust in a clinical setting. As Dr. Weed suggested, getting the right diagnosis should not require a medical “oracle.”
MDO: How can designers offer safer options to prevent treatment errors?
Griffith: We have two major categories in treatment misadventures: medications and procedures. The big challenge in preventing medication errors is the fact that no single mind can grasp the enormity of all the medications used today to treat illness and promote health. Healthcare practitioners know well, only the drugs specific to their practice specialty. Years ago some author joked that physicians really only know about 10 medications. I recall reading that and telling my wife that, in my case, five of my ten were morphine. Seriously, though, most healthcare professionals will confirm deficiencies in their knowledge of their patient’s medications. Hospital nurses routinely administer medications without any notion of the purpose of many of them, so they can easily give an anticoagulant to a patient admitted to the hospital for bleeding.
Hospitals have a requirement to complete an accurate list of all medications used by every patient they admit and they must update that list before discharge. However that list need only contain a name (not all names) of the medication, and the dosing. If that list contained better information, then perhaps everybody on the care team would have a chance to prevent silly mistakes. Off the top of my head, the list should include all the names for the medication, an indication as to whether the dose is low, average, or high, why the patient is taking the drug, a therapeutic index that measures safety, when dosage was started, by whom, the mechanism of action, and specific dangers/warnings (at least if there are or aren’t any). We have the technology to make that happen; we need the will.
Brent James at Intermountain Healthcare in Utah has found a path that reduces errors in hospital treatment. He sold the effort to Intermountain on the basis of cost reductions, but it saves money by reducing complications and improving outcomes. His program also circumvented the physician autonomy issue by being optional and by stressing the role of the physician as the adjuster of the protocols to fit individual circumstances. We do not need to reinvent James’s methods, but we may need to promote their use as a public policy and figure out how to encourage patient’s to insist their local hospitals work that way.
In the realm of procedures, I do not see any magic bullets. Medical device designers currently work effectively in this domain creating better tools and methods.
Several years before I retired from Albany Medical Center they bought a thromboelastogram, a machine that measures clotting of blood. The machine is a great tool, but the hospital bought it without requiring that the manufacturer include a computer that converted the findings into a therapeutic recommendation. Instead they set about to train practitioners to figure out the appropriate therapy from the shape of a curve on graph paper. This approach seemed error prone since many health professionals would only rarely need the results of this analysis, and when they did need it they were in the midst of a crisis. I would guess no industrial designer ever had any input to the manufacturer of this tool or to any hospital looking to add it to its facility.
MDO: How can IDSA change the narrative of patient safety?
Griffith: Clearly all of us involved with the IDSA Patient Safety initiative have been struggling with not just what to do, but what would make an IDSA sponsored initiative unique and powerful.
This effort has gone further than I ever imagined and that has happened because of Sean Hägen and the people he has pulled into the effort. I want to further that progress but not get in the way. I am constantly in awe of the accomplishments of industrial designers and I sincerely believe they hold the key to making a genuine difference in the patient safety agenda; all others have been rearranging the deck chairs.
IDSA might play the role of the impartial friend of society to lobby for free diagnostic use of patented molecules and more Watsons-like assistance for the primary care offices, along with calling for a new generation of physical examination tools.
In the treatment realm, IDSA might play a role in convincing practitioners and hospitals that they would profit from having a few industrial designers on their payroll helping their quality effort identify the best tools and communicating the need for better tools to other designers in the field. Designers inside hospitals could also apply their skills to processes around surgery and other procedures to build cues and forcing functions for safety into the work process. Medical people generally seem not to be able to do this for themselves. Industrial designers do have these valuable skills.