Monthly Archives: April 2010

Root Cause Analysis

Donated Blood Contaminated By Yellow Fever Vaccines

By ThinkReliability Staff

On March 27, 2009, 89 active duty military trainees donated blood at a local blood center.  Unfortunately, the trainees had received yellow fever vaccines (which contain live virus) only four days earlier.  The Red Cross blood donation eligibility criteria requires a two-week wait after a yellow fever vaccine.  The Red Cross does not test donated blood for yellow fever virus antibodies and the ineligible donations were not realized until the blood bank was preparing for another blood drive several weeks later.  The contaminated blood was immediately recalled, tracked down and destroyed, but only after six units had been transfused into five patients.

One of the five patients who received a transfusion died, but as he was in hospice care for terminal cancer, it’s unclear whether the contaminated transfusion was responsible.  Three of the four remaining patients had yellow fever antibodies in their system, but have not demonstrated any symptoms of yellow fever.  The potential transmission of yellow fever to these patients is an impact to the patient safety goal.  Additionally, the transfusion of contaminated blood impacts the compliance, organization, and patient services goals.   The cost of the disposal of 83 contaminated units of blood is an impact to the materials goal, and the required investigation is an impact to the labor goal.

The contaminated transfusions occurred because of the medical need for blood, and the use of contaminated donor blood.  The donor blood was contaminated because the recall did not occur for several weeks (although it happened immediately after the contamination became known) and because the blood contained yellow fever antibodies.  The yellow fever antibodies were present due to donors who had recently received vaccines, and donated, though ineligible.  Because of the large number of trainees who donated blood despite being ineligible, it’s apparent that there was a disconnect in providing the information of a required two-week deferral from donation after vaccination.

Although the Red Cross does make its eligibility criteria known, it’s clear from this incident that this was insufficient in this case.  The military has agreed to provide vaccination records for its members to ensure that blood is not donated during the deferral time after live virus vaccines.  Additionally, the Red Cross has added specific, individual questioning about recent vaccinations.  (The previous process used for screening with respect to recent vaccines was not discussed.)  A recommendation is being made to have healthcare providers discuss eligibility for blood donation after vaccines are provided, though this is not currently being specifically required.  These solutions should reduce the risk of providing contaminated blood.

To view the one-page PDF containing the outline, Cause Map, timeline and action items, please click “Download PDF” above.  The information for this investigation was provided by the Centers for Disease Control and Prevention (CDC)’s Morbidity and Mortality Weekly Report (MMWR).

Root Cause

Errors in Translated Medication Instructions

By ThinkReliability Staff

It’s well known that instructions on medication (both prescription and otherwise) can be confusing and lead to potentially lethal consequences.  (See our previous blog on the topic.)  Now imagine how much more danger there is if you don’t speak the language in which the instructions are printed.

A recent study published in the American Academy of Pediatrics Journal “evaluated the accuracy of translated Spanish-language medicine labels among pharmacies in a borough with a large Spanish-speaking population. ”  The study found significant issues with label accuracy with a popular language in an area with a large population of speakers of that language.  You can imagine how these results could get even worse for an area that had a smaller number of Spanish speakers, or for patients who speak a less common language.

One of the most striking examples was a man who received heart medication that was to be taken once daily.  The instructions were only partially translated and “once” (which means 11 in Spanish) was left on the instructions.  The patient took 11 pills (instead of 1) a day.

The study found an overall error rate of the prescription instructions that had been translated into Spanish by computer of 50%.  (86% of the pharmacies surveyed translated their prescriptions with a computer program.)  It is likely that patients with the incorrectly translated prescription instructions took the medicine incorrectly, resulting in the potential for serious harm, or even death.  This is an impact to the patient safety goal.  The rate of errors made by the computer means more work for pharmacists and translators due to the corrections that must be (or should be) made.  (Obviously this is not always happening.)  Patients receiving instructions they do not understand can be considered an impact to the patient services, compliance, and organizational goals.  (The study was performed in the Bronx, New York.  It is a law in New York City for pharmacy chains to provide translated labels for the top seven foreign languages in the area.)

Patients do not understand the directions because the patients do not speak English and the instructions are either not translated, or are translated incorrectly.  The instructions may be translated incorrectly because the computer program translates them incorrectly and there is an inadequate verification of the computer translation, because the pharmacist does not speak the language and/or there is no translator available (likely due to lack of funds or an uncommon language).   The instructions may not be translated if the pharmacy has no translating capabilities, also likely due to cost or an uncommon language.

An obvious suggestion is to improve the accuracy of the computer programs that do the translating, perhaps standardizing the translations among the different programs that do the job.  Pharmacists could also be provided with a guidebook of translations for standard pharmacy terms (such as take orally).  Additionally, translation software could be added to the computer programs currently used by pharmacists.

I have a simpler suggestion that I borrowed from the aviation industry.  I noticed the last time I flew that instead of having translations of the safety instructions in a dozen different languages, there were practically no words at all.  Instead, the airline used picture instructions.  I suggest doing something similar with medications.  (See my example of a picture for “take orally” on the PDF.  View the root cause analysis investigation and my picture by clicking “Download PDF” above.)

Because of the risk involved, it’s clear something needs to be done.  Prescription instructions are hard enough to understand in English, much less poorly translated into another language.  I’m sure suggestions will keep coming in, and surely some smart folks out there will come up with a way to reduce the potential for confusion and injury.

Root Cause

Therapy Equipment Delivers Radiation Overdoses for Years

By ThinkReliability Staff

In September of last year, a physicist at a healthcare facility was trained on use of the BrainLAB stereotactic radiation therapy system.  During this training, the physicist realized that the system had been incorrectly calibrated, as the wrong chamber had been inserted into the machine.  The facility realized that the chamber had been incorrectly inserted at installation in 2004, and that patients who used the portion of the machine calibrated by that chamber had received radiation overdoses over those five years.

The facility is working through the impacts of these errors, the causes of the error, and what needs to be done to keep an issue like this from ever happening again.

First let’s examine the impacts to the goals resulting from this error.  There’s an impact to the patient safety goal due to potential for deaths and injuries. (Because these patients  were already sick – sometimes very sick – the facility is still determining what impact the overdoses may have had.)  There has not yet been mention of an employee impact – the physicist who set up the machine is no longer at the facility – so we’ll just put a “?” after Employee Impact.  The event was reported to The Joint Commission (no reports were required by law), which can be considered an impact to the compliance goal.   The organizational goal was impacted due to potential lawsuits against the hospital.   The patient services goal was impacted because 76 patients received an average overdose of 50% (other patients received overdoses that were considered within the acceptable range for treatment).  Because radiation was involved, there is the potential for an environmental impact.  However, there is no evidence that any radiation leaked to the environment, we’ll put a “?” by the environmental goal as well.  Lastly, the property and labor/time goals were impacted because of the additional follow-up exams, testing, support, and treatment, which the facility will provide for all those affected by the issue.

Once we’ve determined the impact of the event, we can begin an analysis of how it happened.  Or, what were the causes?  The goals were impacted due to the overdose to several patients.  The overdose occurred because the radiation therapy machine was miscalibrated and the miscalibration was not discovered for five years.  The machine was miscalibrated because the incorrect chamber was installed and the chamber installation was not verified.  The physicist chose the wrong chamber and the equipment representative (who was on hand for the installation) did not notice the error.  At this point, it’s unclear why the physicist chose the wrong chamber and why the equipment representative did not notice the error.

The miscalibration was not noticed for five years because any re-calibration of the machine depended on the chamber which was incorrectly installed.  So although the machine was not delivering the correct amount of  radiation, the problem was with the calibration itself, resulting in a propagating error.  According to the facility, none of the patients showed any unusual side effects that would indicate they were getting too much radiation. However, some of the symptoms may take years to develop.  Additionally, no other staff members were trained on the equipment for five years.  It was a second staff member who was trained on the equipment who finally noticed the error.

Even though there are some questions still remaining in our Cause Map, we can develop some solutions, as the facility in question (as well as other stakeholders) is doing.   One suggestion is to do an external calibration of the machine – i.e., use a calibration method that is completely separate from the machine to determine if the correct amount of radiation is being delivered.  Also, have an independent verification that each piece of the equipment was installed correctly.   Require the equipment representative to sign off on the installation.  Last but not least, train other staff members to operate the equipment as backup.   The facility is working with the FDA to assist in its efforts to increase the safety of radiation use in healthcare settings.  (See our previous blog about this topic.)

Step 4 to avoid radiation therapy errors: verify HOW MUCH – how much radiation therapy is required, and how much is the patient actually getting.