Over the past several months I’ve been involved with a committee tasked with performing a failure modes and effects analysis (FMEA) on our bedside scanning initiative. An FMEA is a procedure for analyzing potential failure modes within a system and classifying those potential failures by frequency and severity. The failure modes can be actual or potential. It’s a way to plan for holes in the system before they actually develop, and can be quite useful in creating possible solutions for future problems. Being proactive is always easier than being reactive, I think.
The committee consisted of individuals from quality assurance, pharmacy, internal audit, nursing administration and nurses currently practicing on various units within our facility.
The medication administration process was broken down into 16 very distinct steps. Each step was scrutinized for potential weaknesses. In total, we described 100 potential failure modes associated with the medication administration process along with their causes and effects. We used current literature sources to describe known events.
Each failure mode was assigned a Risk Priority Number (RPN) based on the product of the occurrence likelihood, detection likelihood and severity (completely subjective). Higher RPN values represented failure modes with greater potential for harm. Failure modes with an RPN greater than 100 were dissected further and recommended actions were developed to mitigate the effects of the failure mode (n = 19). Would you be surprised to learn that almost all the failure modes with an RPN greater than 100 were tied directly to some form of human behavior, i.e. willfully choosing to circumvent the system? I was.
Below are some of the FMEA committee’s observations and recommendations.
CONTROLLING THE HUMAN ELEMENT
Although the implementation of bar code medication administration (BCMA) serves to automate the medication administration process, there remain human elements to the process, namely the nurse verifying the patient’s identity (visual of armband, asking patient’s name, etc.), verifying the medications (reading label on packaging, reading the order, etc.) and giving the medication to the patient. It is difficult to put failsafe automatic controls around these processes to prevent the willful disregard of process steps.
– Of the 19 failure modes with an RPN >100 which we identified as high risk, 12 out of 19, or 63%, are associated with nurse verification of the patient’s identity and nurse verification of the medication.
– Underlying the needed updates to the medication administration policy, a thoughtful and comprehensive medication administration procedure covering all 16 process steps is imperative for a successful BCMA implementation. A consistent process that can be repeated will yield predictable, reliable, and measurable results.
– Nursing Administration and HR should determine appropriate and adequate disciplinary action for staff that willfully disregard and fail to follow established medication administration policies and procedures and ensure these disciplinary actions are applied consistently.
– Adopt the Kaiser MedRite Program for use during the medication administration at our facility. This program helps to minimize interruptions during medication administration to further enhance patient safety and provide needed controls around the human elements of the process.
ADEQUACY OF PREPARATION
The representatives we met with at other facilities stressed the importance of adequate preparation for BCMA implementation and pointed to it as the one area they would have liked a “do overâ€. Preparation includes user training as well as having the appropriate resources in place to ensure a smooth implementation as well as adequate support in not only the weeks immediately following “go live”, but in ongoing system management.
– Adequate time for nurse training on the BCMA system is needed prior to “go liveâ€, with 8 hours per nurse being optimal.
– The training should include hands-on simulation training using all procedures and equipment under various patient scenarios. This helps ensure that the majority of the nurses are able to “hit the ground running†immediately upon “go liveâ€.
– Strong consideration should be given to the designation of a BCMA Nurse Administrator during the preparation phases to serve as the point person for the nurses for all phases of the project. This position ideally is a full time position to allow issues to be fully addressed and to provide consistency throughout the different phases of the project.
– As noted above, the two hospitals which have a similar number of total beds to our facility employ 2.5 IT Pharmacists to manage their systems which are also similar in size and scope to ours. Consideration should be given to reviewing the adequacy of our current resources in this area and filling any needs prior to system implementation.
ADDRESSING SYSTEM ISSUES
No system however well designed will satisfy all users or be adequate for all situations. However with proper planning and continual open and transparent communications, many of the potential failures due to system issues can be addressed so that a majority of the users are satisfied immediately upon implementation.
– Request BCMA system programmers to re-program the audible signal heard from use of the bar coding device in order to distinguish an errant scan from an appropriate scan. This reduces the need for a constant visual on the machine to catch an error and allows increased focus on the patient.
– Ensure representatives from the majority of nursing units have input into the selection of system hardware and tools to increase the sense of ownership and satisfaction among the population that will be using the system the most.
SYSTEM MANAGEMENT POST “GO LIVEâ€
Besides increasing patient safety, BCMA also will provide a wealth of data that can be used to ensure that the processes put in place continue on as expected or are flagged in a timely manner when changes may be needed. Ongoing management of the system and use of the data in meaningful ways are needed to ensure success and acceptance of the system.
– Data management and the production of meaningful reports post “go-live†are extremely important for identifying patient safety issues as well as providing one way to identify willful disregard of procedures. One report that is especially needed is a “nurse override†report. Staff solely responsible for the tracking, trending, and reporting of these metrics should be made available for these functions post “go-liveâ€. Consideration should be given to tasking the BCMA Nurse Administrator recommended above with responsibility for this area to ensure consistency and accuracy in the reports and adequate follow up on the results in the form of education and process changes as needed.
– Consideration should be given to the use of independent verification of the results of auditing and reporting implemented. Various groups in our current system, some of which had representatives participating on the FMEA Team, can assist in this area, including Performance Improvement, Internal Audit, Compliance and Risk Management.
The FMEA process was very interesting and I learned a lot about a great many things. If taken seriously by upper level administration, the information provided by the FMEA will benefit our efforts to implement bedside scanning. Overall I think the FMEA is a valuable tool that I can see using again in the future.
Comments
5 responses to “A failure modes and effects analysis on bar code medication administration”
Jerry, This is very nice analysis.
Something is confusing to me in that the Kaiser Medrite program does not mention BCMA. Am I missing something?
Hi John – No, you aren’t missing anything. The Kaiser Medrite Program was the product of one of the failure modes that wasn’t directly related to BCMA and didn’t make the RPN cutoff; nursing getting interrupted during a med pass. Nursing felt strongly that this “failure mode” should be addressed in the final report to the med exec committee. In theory, a med pass free of interruption would decrease opportunity for error. With that said, BCMA would potentially catch these errors anyway. This wasn’t the only questionable failure mode that was discussed, but it is the only one that made the cutoff. Thanks for stopping by and leaving the comment/question.
Thanks Jerry, Let me stretch this out a bit….. I question (and get a lot of heat) with the assumption that BCMA would catch the errors. There is better science around that workarounds are prevalent (below reference), than BCMA accurately catches and prevents the errors. I know it makes sense that it MIGHT, but to date there has been very very little good science and a lot of junk science.
Workarounds to Barcode Medication Administration Systems: Their Occurrences, Causes, and Threats to Patient Safety
Ross Koppel, Tosha Wetterneck, Joel Leon Telles, and Ben-Tzion Karsh
J. Am. Med. Inform. Assoc. 2008;15(4):408-423. PrePrint published July 1, 2008; doi:10.1197/jamia.M2616
Ah yes, the work around article. I hear what you’re saying and agree that people will find ways around the system no matter what obstacles you put in their way. With that said, I also believe that the majority of nurses are “law-abiding”, and for those nurses BCMA can be helpful. I can tell you that the barcoded storage system in our pharmacy has prevented some mistakes since implementation, but I can also say that barcoding creates blinders, i.e. pharmacists believe it will catch everything, but we know that simply isn’t the case. I would like to see studies pre- and post-BCMA implementation, but there are several obstacles to performing these studies. For one, how do you get observational data for comparison? I’m not saying BCMA is the savior of patients, but I do believe it has something to offer.
[…] been involved in an FMEA, and I’ve personally found them to be powerful tools. We did an FMEA prior to implementation of our BCMA system and came up with what I thought was a pretty good list […]