I came across this fascinating paper from 2002 on using a framework known as Healthcare Failure Mode And Effect Analysis to proactively evaluate healthcare processes so as to reduce patient risk. Analyzing events that have already happened & ‘close call’ events can help us fix vulnerabilities, but proactive failure mode analysis helps us prevent harm. Proactive and retrospective approaches should both be used in the quest for quality improvement
FMEA is something common in engineering, and i. It seems like a great fit for digital health tools as well. Curious for the community’s thoughts.
FMEA shouldn’t be used just in engineering, but rather for all processes. It’s not just limited to how a machine works or the steps a computer program takes. In the paper, they use the example of HFMEA for how a medical center would discharge a patient:
The processes are broken down into steps, and each step further broken down into subprocesses (eg, “set up clinic” would be a step, and “procure equipment”, “hire staff”, “rent building” or so on would be subprocesses).
We then try to figure out the failure mode of each subprocess: procure equipment might fail if: equipment is not available, equipment is damaged. Each failure mode has potential causes analyzed, and these causes are ranked by severity and probability, and given a resulting hazard score. If the hazard score is low, we don’t have to worry about it, but if it’s high, we then go through steps to figure out how to best mitigate the problem.
Anyway, I’d be happy to facilitate an FMEA brainstorming session if anyone is interested.
Looking at the recent literature, it seems that the general application of FMEA is for healthcare procedures - for improving blood transfusions, radiotherapy procedures, and ICU management, but it seems to me that these principles could just as easily be applied to things within the community health setting.
Seems like a very relevant method! While it’s less formal than FMEA, I’ve participated in “pre-mortem” brainstorming sessions, in which a group proactively tries to imagine and understand the kinds of failure scenarios that would lead to a post-mortem. I found it really helpful and would like to do this more. Would be great to see this as part of the design process for a specific implementation to keep it focused and concrete.
