Tag Archives: medical error

Root Cause Analysis

Study Finds Only 1 in 7 medical errors reported

By ThinkReliability Staff

A study by the Office of the Inspector General (OIG) of the Department of Health and Human Services found that hospital employees are only reporting fourteen percent of all medical errors.  Additionally, the study determined that processes are rarely changed to help prevent similar errors in the future.

The study was conducted by reviewing the medical records of Medicare patients. As a condition of participating in the Medicare program, hospitals are required by federal regulations to track medical errors that harm patients and to implement solutions to protect patients from similar mistakes in the future.  All hospitals involved in the study had a method to collect information on medical errors, but all adverse events were not being efficiently captured by the systems in place.

The study found that out of 293 cases reviewed, only 40 were reported, 28 led to investigations and only five resulted in changes in the hospital’s processes.  Additionally, the inspector general estimated that more than 130,000 Medicare beneficiaries experience at least one adverse event in a hospital in one month so there is a plenty of room for improvement in patient care.

The study found a number of reasons that adverse events and medical errors were not being reported.  Confusion over requirements was one cause of the under reporting because hospital employees did not always recognize that a particular event harmed a patient and was required to be reported.  Also, there were cases where employees assumed someone else would report the incident or they considered the incident to be so common that it didn’t need to be reported.  There was also a tendency to not report things that were considered to be isolated events that were unlikely to recur.

To help combat these causes, the OIG report recommends the development of a standard list of medical errors that should be tracked and reported.  The OIG also recommends that guidance be developed and provided for the accreditors of hospitals since they do not typically investigate adverse event collection methods.  Additionally, some consumer groups are pushing for public reporting of medical errors to help pressure hospitals to improve their policies and practices.

Click on “Download PDF” above to view a high level Cause Map of this event which is an intuitive, visual root cause analysis of this issue.

Root Cause Analysis

Young Boy Killed by Projectile During MRI

By Kim Smiley

It has been over ten years since six year old Michael Colombini died as a result of injuries that occurred during a routine MRI.  He was undergoing a post-surgery MRI after removal of a benign brain tumor when he was hit by a magnetic oxygen tank that was pulled toward the MRI machine at high speed.  His skull was fractured and he died two days later.

How did this horrible accident happen?

A Cause Map, or visual root cause analysis, can be built to help explain the causes that contributed to this death.  In this example, the patient needed an MRI because he had a brain tumor removed and he was hit by a magnetized oxygen tank while in the MRI machine.  This occurred because the oxygen tank was attracted by the huge magnet in the MRI and flew towards the machine at high speed.

MRI magnets will always attract magnetic things, even when the machine is off.  Bringing a magnetic oxygen tank into a MRI examination room is a dangerous situation.  In this example, there are several reasons why the tank was brought into the MRI area.  The tank was bought into the room by a well-meaning nurse who heard the anesthesiologist calling for oxygen.

The nurse had returned to the MRI area in order to retrieve something she had left there earlier in the day when she had accompanied a patient to the space.  She heard the anesthesiologist calling for oxygen and saw the tanks nearby so she handed one to the doctor.  Contributing to this accident is the fact that the oxygen tanks were stored near the door to the MRI exam room.

The anesthesiologist was calling for oxygen because the patient had low oxygen saturation levels and needed additional oxygen.  The patient was a six year old boy so he had been sedated for the MRI procedure.  A piped in system normally supplied oxygen for use during MRIs, but the system had malfunctioned.  Both MRI techs on duty had gone to investigate the piped in oxygen system failure so nobody trained on MRI safety was around when the nurse bought in the oxygen tank.

Some simple solutions that might have helped prevent this accident, even with the oxygen system failure include storing oxygen tanks far away from the MRI exam room and not allowing unescorted non-MRI staff into the space.  It’s also always a good idea to have an acceptable back up for important systems planned in advance.  If a second safe oxygen supply was already provided, this accident could have been prevented.

The magnets in MRI machines are 200 times stronger than a refrigerator magnet and, as this example illustrates, the potential for injuries from projectiles is very real. Like most accidents, this death was caused by a number of failures that occurred at the same time.  All of the staff involved was trying to do the right thing, but the end result was the unnecessary death of a young boy.

Hospital Acquired Condition, Root Cause Analysis

Teenager Paralyzed After Epidural Not Removed

By ThinkReliability Staff

In May 2008, a fourteen-year-old entered an English Children’s Hospital for a routine surgery to remove gallstones.  The recovery, however, was anything but routine.  The patient was given a spinal epidural to reduce pain during the operation; however, the epidural was not removed until two days later. By then, permanent damage of the spinal cord caused the patient to be paralyzed from the waist down.

The hospital has admitted liability, possibly leaving them responsible for some or all of the patient’s specialist care and support.  Because the anesthetic needle was not removed until the patient’s body until far later than it should have been – and more than a day after the patient’s first complaints of leg numbness – it begs the question whether the procedure for administering an epidural included follow-up care, including removal.  Procedures – whether they are written down or not – exist for most complex tasks, especially medical tasks that involve risks to patient safety.  If use of the procedure results in an error, it should be re-examined.  However, many procedures only include the first part of a procedure, or the administration, ignoring follow-up that must be completed to ensure the process is a complete success.  In this case, that follow-up should have included checks to ensure that the patient was recovering from the epidural (which would have noted something amiss when she continued to feel numbness in her legs) and a schedule to remove the epidural.  Because neither of these things happened, a plan for follow-up after administering epidurals must be developed and put into practice.

To view the Outline and Cause Map, please click “Download PDF” above.

Never Event

Infant Death Due to Sodium Chloride Overdose

By ThinkReliability Staff

On October 15, 2010, a 40-day old prematurely born infant died from a sodium chloride overdose at an Illinois hospital.  Because a computerized system was involved, this case has been noted as a harbinger for possible issues resulting from the use of computerized systems.  To learn more about what happened, we can look at the case in a visual root cause analysis, or Cause Map, to examine all the causes.

First we begin with the impact to the goals.  The infant’s death was an impact to the patient safety goal.  A death resulting from a medication error is a “never event“, which is an impact to the compliance goals.  There is a related wrongful death lawsuit, which is an impact to the organization’s goals.  The overdose of sodium chloride delivered to the patient is an impact to the patient services goal.

We begin the analysis with the impacts to the goals and ask why questions to fill out the Cause Map.  The infant death was caused by the sodium chloride overdose, which occurred when the infant received more than 60 times the dosage ordered by the doctor intravenously.   The infant was receiving sodium chloride intravenously to provide nutrition, as he had been born prematurely.  Although a blood test indicated abnormally high levels of sodium, it has been reported that the lab technician assumed they were inaccurate, resulting in the infant not receiving immediate care for the overdose.

When a process – in this case, the medication delivery process – does not work correctly (such as occurs when an overdose is given), it means that the checks at every level of the process were ineffective.  The final check at the patient’s bedside was ineffective because the label on the IV bag did not match the actual contents.  It’s unclear how that occurred.  The error was made at the pharmacy, when a pharmacy technician entered an incorrect number into the compounding system.  Normally entering a too-high dose would trigger an alert with an automated system, but the alerts were turned off.  Part of the reason for the error was that the pharmacy technician had to manually enter the prescription in the first place.   A  doctor enters a prescription via the automated dispensing system.  However, the automated dispensing system, and the computerized compounding system did not communicate with each other, so for orders that required compounding, a technician had to transfer the order from one system to the other, manually.

A computerized system is no better if it’s not used properly.  If parts of the system don’t communicate with each other, and safety checks are turned of, a computerized system may actually be less safe, especially if people expect the automatic checks are being performed, and so don’t perform any of their own.  Computerized systems have a lot to offer – namely, reducing the number of medication errors relating to illegible handwriting or providing automatic checks for drug interactions.  But these systems are not fail-safe and checks used to ensure that patients

Root Cause Analysis

The July Effect

By ThinkReliability Staff

No one ever looks forward to a trip to the hospital, and a new study suggests that you might be particularly wary during the coming weeks.  A new study shows a 10% spike in patient fatalities during the month of July.  Many in the medical profession have been aware of “the July Effect” anecdotally for years, but researchers in the University of California at San Diego study combed through over 62 million death certificates dating back to 1976 to prove its existence.

Why the spike?  Sociologist Dr. David Phillips, who conducted the study, believes it is because new doctors begin their residencies in July each year.  The phenomenon is limited to fatal medical errors, and is not evident in surgical or “general” error rates.  Consistent with the study’s “New Resident Hypothesis”, fatalities are even higher in counties with higher concentrations of teaching hospitals, in which there would be more resident doctors.  It is clear there is a link between higher rates of medication errors and the presence of brand new doctors.

The study is one of the first to demonstrate the linkage though.  Multiple smaller studies have failed to show any correlation between time of year and death rates.  Researchers point out that the new study focused on a much longer time range and broader geographic area than any previous study however.

Although the study raises some interesting questions, it stops short of providing solutions.  Doctors already face a rigorous course of study to prepare for their residencies, which of course are designed to provide the experience needed.  New doctors are also generally well supervised.  And to some extent there will always be risk associated with inexperience when it comes to teaching hospitals.

A Cause Map can illuminate areas that might benefit from further research.  The study narrowed down one of the contributing factors to medication administration.  Why just in that area though?  Are new residents better supervised in the OR?  Do new doctors have the capability of prescribing and administering medication during their first month?  What types of errors do they make when doing this?  Do they prescribe the wrong medication completely?  The wrong dosage?  Or do they overlook adverse interactions with other medications?

More research is needed to accurately determine why the July Effect occurs, but patients can be prepared.  Experts agree that patients should ask plenty of questions and bring along an advocate for support.  For more information, the study, “A July Spike in Fatal Medication Errors: A Possible Effect of New Medical Residents”, is available here.

Root Cause Analysis

Hungover Surgeons More Likely to Err

By ThinkReliability Staff

The headline probably isn’t shocking to anyone who’s woken up the next morning with a pounding headache and dry mouth.  Clearly one’s performance at work is going to be impacted by a night of unabated drinking.  However a recent Irish study, published this month in the Archives of Surgery, show surprising results regarding the lingering effect of alcohol consumption.  Their findings show that well into the day surgeons are more likely to make mistakes.

Modern surgical techniques, including laparoscopic surgery, require great manual dexterity and control as well as sustained mental focus.  It is common knowledge that both of these skills are impaired while intoxicated.  What is unknown is how these skills are impaired after one is no longer intoxicated, but obviously still affected.  In all but one test subject, their blood alcohol content (BAC) had returned to 0.00%.  Initial testing done in the morning showed no significant difference between test and control subjects, however later in the day there was a perceptible decline.  While the study was only a preliminary one, it indicates that more research is needed in this area.

A Cause Map can be especially helpful in a research environment because it helps define causal relationships.  In this case, the researchers focused on the effects of drinking the night previous.  But perhaps there are other reasons at play, such as fatigue, which contribute to the effect.  When searching for causes it is important not to focus in on one aspect, ignoring others, since all causes are required to produce an effect.

It is expected that surgeons wouldn’t actually drink while at work.  However, there are surprisingly no guidelines about when they should stop drinking beforehand.  Pilots are federally mandated not to drink at least 8 hours prior to flying or fly with a blood alcohol content (BAC) of .04% or greater.  Perhaps this study will generate an overdue discussion on the need for abstention prior to surgery.  Potential solutions, such as training or regulations, can be displayed directly on the Cause Map above the appropriate cause.

Root Cause Analysis

Aging Surgeons

By ThinkReliability Staff

Over 20% of today’s physicians are over the age of 65.  Should this be cause for concern?  After all, we rely on our doctors to take care of us when we are often at our most vulnerable.  While increased age means increased experience, there are also down sides.  Age can bring with it a decrease in physical and mental capabilities, as well as a reluctance to adopt newer technologies.  At least this is what multiple studies have hinted at over the past few years.

The problem is that such a “decrease in capabilities” is highly subjective and difficult to measure.  Surgeons rely on a variety of cognitive and tactile skills in their craft – steady hands, learning new techniques, composure under stress, communication skills, and so on.  As highly trained professionals, it is sometimes difficult to decide when it is time to call it quits.

Furthermore, in the United States, age-based discrimination is outlawed in most industries except where regulated.  For instance, airline pilots and air traffic controllers are both subject to earlier-than-average retirements due to public safety concerns.  Many federal and state public workers, such as corrections officers and firefighters, are similarly limited. It’s difficult to argue that some physicians don’t make similar split-second, life-and-death decisions – especially surgeons.

The associated Cause Map visually lays out the dilemma.  Surgeons who aren’t performing adequately do so for two reasons.  First, they have a medical condition precluding them from performing to standards.  (Note that to keep this Cause Map simple, other issues such as mental health problems, addiction, and failure to maintain their continuing education were not examined.)  Second, they are allowed to continue practicing.

Such physicians continue after their abilities are impacted for a number of reasons.  Some might be unaware of their condition or unwilling to accept it, both stemming from a belief that they are still competent to practice.  Additionally, current processes at most hospital are slow to identify such physicians.  Most hospitals rely on co-workers to identify such doctors, clearly a highly subjective and ethically complex system.  Age-based screening is not common at many hospitals, partly because of resistance from hospital staff.  In fact, only 5-10% of hospitals have directly addressed this issue.  Labeling doctors as “unfit to practice” isn’t necessarily a bad thing.  If such doctors are identified early, patient safety is enhanced.  Additionally, early identification can sometimes allow those doctors to continue practicing in a controlled and safe environment.

Now that the problem has been laid out, the next step is to look for possible solutions.  It’s clear that little can be done about age-related deterioration.  So the focus moves to the other branch of the Cause Map.  Here there are a number of possibilities.  While age-based screening is certainly an option, it’s not the only one.  For instance, including hospital staff in making decisions might also help decrease resistance to identifying deficient physicians.  Additional training on the impacts of age might make co-workers more willing to discuss their concerns.  Or doctors might be more willing to adapt to their limitations if regular screening can identify possible health problems.

While more research is needed to determine how extensive this issue is, it is clear that at most hospitals current procedures to identify deficient physicians are lacking.

Root Cause Analysis

Kidney Transplant Mix-up

By ThinkReliability Staff

On January 29, 2011, a kidney was transplanted into the wrong patient.  No one was injured, but this was known as a “near miss” – had things gone slightly differently, it could have resulted in severe consequences.  Namely, the patient who received the incorrect kidney could have been killed or seriously injured, had the kidney not happened to be compatible with that patient also.  (The kidney donor had Type O blood, known as the universal donor, which aided in the compatibility.)  The patient who was supposed to receive the kidney could have had a long wait back  on the transplant list.  Luckily, a new donor was found for the second kidney and a new kidney was found for the second donor fairly quickly.  Although there were no injuries, the high potential for injury results in an impact to the patient safety goal.

To try and help figure out what went wrong, we begin with the impacted goal and ask “Why” questions to fill out the analysis. We discover that there were two kidneys that arrived at the hospital simultaneously.  In order for the kidneys to be switched, the kidneys must have been mislabeled, or miss-identified once at the hospital.  The coordinating agency for transplants states that the packaging and labeling of the organs was correct.  We then turn our focus to the identification steps of the organ once at the transplant center.

To aid in determining where process improvements can be made, first we need to define the process.  We can do this with a process map – a step by step instruction of how a process is performed.  In this case, the steps for transplants have been developed by an outside agency – the United Network for Organ Sharing (UNOS).  We can outline these steps in our Process Map.  Because of the high risk for consequence should an error occur, the process is well-defined and consists of checks to ensure that mismatches do not occur.  The last highly publicized incident of a transplant error was in 2003 (see more about that incident here).

The hospital involved has not released details about what might have occurred in the process; however, it’s certain that they’re looking at the process with a fine-tooth comb and trying to implement improvements.  The transplant program has closed down while they’re doing so.

Root Cause Analysis

Wrong Surgery Performed on Patient (Part 2)

By ThinkReliability Staff

This week, we will continue our discussion of an incident where the wrong surgery was performed on a patient.  Last week, we looked at the timeline of events and a process map of the universal protocol developed to reduce the incidence of surgical errors.  This week, we’ll perform a root cause analysis of the issue.

The specific steps identified that didn’t go well, or weren’t performed, from the process map now become causes on our Cause Map.  Instead of the causes or errors being grouped chronologically or by type (as they are on a fishbone diagram), the causes are grouped by their contribution to the incident.  The Cause Map reads from left to right by asking “Why” questions, beginning with the impacts to the goals.

For example, the patient safety goal was impacted because a patient received the wrong surgery.  Why?  Because the physician performed the wrong type of surgery. Why? Because the surgical site was not clearly marked.  Why? It was marked on the correct arm, though not the correct site (the wrong surgery was performed on the correct hand) and the mark was washed off during patient preparations.  These are both issues identified in the process map that did not follow the universal protocol for surgical preparations.  Both of these issues contributed to the wrong surgery.  In addition, the surgeon was thinking about carpal tunnel surgery, since most of his day, especially just prior to the surgery, had been spent on carpal tunnel surgery, either performing it, or doing pre- or post-surgery briefs with other patients.

Neither the patient nor the operating room staff stopped the surgeon from performing the incorrect surgery.  The patient spoke only Spanish, which may have contributed to her not speaking up.  The operating room staff did not include the nurse that had done the patient assessment, due to a last-minute operating room and staff switch due to other delays.  There was no time-out prior to the procedure, which may have alerted the staff about the wrong  procedure, or may have helped the surgeon switch from thinking about carpal tunnel surgery.

Once the analysis is complete, possible solutions are identified on the Cause Map.  Many of the solutions in this case are to ensure that the universal protocol procedures are being followed.  Had they been followed in this case, the risk of performing the wrong surgery would have been reduced.  Many facilities are already using the universal protocol; however, this case study shouldn’t be ignored by them.  The operating surgeon made this case public and added the following comment: “I hope that none of you ever have to go through what my patient and I went through. I no longer see these protocols as a burden. That is the lesson.”

This surgeon has learned his lesson and will likely be more diligent about following these protocols in the future.  However, there’s no need to wait until you, or your staff members, have their own incidents to learn from.  Use this case study to emphasize the needs for these protocols, in hopes that your facility can reduce its own risk.

(Details of this incident were recently published in the New England Journal of Medicine.)

Root Cause Analysis

Wrong Surgery Performed on Patient (Part 1)

By ThinkReliability Staff

A case study of an incident where the wrong surgery was performed on a patient was recently published in the New England Journal of Medicine.  Surprisingly, the study was published by the surgeon who performed the operation, because, in his words, ” hope that none of you ever have to go through what my patient and I went through.”  The surgeon also provided full disclosure to the patient – who requested that he also perform the correct surgery.

We will be analyzing this issue in two parts.  This week, we’ll be looking at the timeline of events and a process map of the universal protocol developed to reduce the incidence of surgical errors.  (The timeline and process map can be seen by clicking “Download PDF” above.)  Next week, we’ll perform a root cause analysis of the issue.

The timeline of events shows a harried day where the surgeon in question performed a carpal tunnel release surgery with a patient who became upset about the use of anesthetic, then briefed the patient who would later receive the wrong surgery, then performed another carpal tunnel release surgery on a second patient.  Then the first patient became very agitated, resulting in an emotional conversation for the surgeon.  Delays resulted in a change of operating room and operating staff for the third patient, so the nurse who had performed the pre-procedure assessment was no longer participating in the procedure.

The  procedure was further delayed when the circulating nurse had to leave to find a tourniquet, since there wasn’t one in the operating room.  The surgeon spoke to the patient in Spanish (she did not speak English), which the nurse took as the time-out, so a real surgical time-out did not occur.  As per hospital protocol, the patient’s arm, but not the specific surgical site, was marked, but it washed off while her arm was being prepped for surgery.

It’s easy to see how this sets the scene for mistakes. Unfortunately, these kind of things happen, and so it is important that there are procedures in place to minimize errors.  The procedures here are the universal protocol, which are shown on the PDF.  Additionally, the parts of the process that were not performed, or were performed improperly, are noted in red.