All posts by Angela Griffith

I lead comprehensive investigations by collecting and organizing all related information into a coherent record of the issue. Let me solve a problem for you!
Root Cause

Lethal injection fails to quickly kill prisoner

By ThinkReliability Staff

While the use of the death penalty remains highly controversial, there is general agreement that if it is used, it should be humane.  The execution of a prisoner in Oklahoma on April 29, 2014 did not meet those standards.  The inmate died 43 minutes after the drugs were injected.  (Typically death takes 5-6 minutes after injection.)  According to Jay Carney, the White House Spokesperson, “We have a fundamental standard in this country that even when the death penalty is justified, it must be carried out humanely – and I think everyone would recognize that this case fell short of that standard.”

The details surrounding this case can be captured in a Cause Map, or visual root cause analysis, to examine the causes and effects of the issue.

The problem being evaluated is the botched execution of an inmate in the Oklahoma State Penitentiary. The execution began at 6:23 pm on April 29, 2014.  An important difference in this execution, compared to other executions, is that it was the first time the state had used the drug midazolam as part of the three-drug injection protocol.  The protocol, when originally developed in 1977, called for sodium thiopental, followed by pancuronium bromide and potassium chloride.

The safety goal was impacted in this case because of the failed execution.  The public service goal can be considered to be impacted as the execution was called off (after all three drugs were administered; the prisoner later died of a heart attack.)  The schedule goal is impacted because all future executions have been called off.  The state planned a two-week postponement of the next execution (scheduled for later the same day) in order for a review of this investigation to be completed, but at the time of this writing, that execution has not yet been scheduled.  Executions across the country have been appealed or stayed and none have taken place since April 29th.  The labor/ time goal is also impacted due to the investigation into the execution, which has not yet been published.

These goals were impacted due to the failure of the lethal injection.  The process intended to be used for this lethal injection is detailed on the downloadable PDF.  However, from the start things didn’t go smoothly.  Instead of using two IVs, one in each arm, only one IV was able to be connected, in the patient’s groin.  Because sodium thiopental is no longer available (drug companies will no longer provide it for use in lethal injections), the drug midazolam was used instead.  However, the protocol for using that drug is disputed.  In Florida, five times the amount of midazolam is used.  In Oklahoma, midazolam is used along with hydromorphone.  Because of the debate about lethal injection, most states don’t divulge their suppliers, so the efficacy of the drugs used cannot be verified.  In addition, there is generally at least one doctor present to oversee the executions, but these doctors are not usually identified and may not participate in the actual administration of the drugs because many medical organizations ban doctors from participating on ethical grounds.

At this point, it’s unclear what will happen at future executions.  The investigative report being prepared by the state of Oklahoma may give some suggestions as to how to make lethal injections more humane in the future, or this may tilt the scales towards ending lethal injection, or executions altogether.

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

Root Cause

Two Los Angeles area nurses are stabbed the same morning at different hospitals by different attackers

By ThinkReliability Staff

The stabbing of a nurse that took place in a Los Angeles County, California hospital on April 20th, 2014, resulted in the serious injury of a nurse.  The danger of increasing violence and attacks within hospitals was demonstrated by this and an unrelated incident at another Los Angeles County hospital that happened later that same morning.  Both involved stabbings to nurses, though in the first case, the attacker used a knife after he bypassed security and in the second case, the attacker stabbed a nurse with a pencil.

By performing a root cause analysis of just one demonstrative case, solutions that can prevent similar issues (like the one that happened later that very day as well as many other recent cases of hospital violence) can be developed.  We will use Cause Mapping, a visual diagram of cause-and-effect relationships, of this case as an example of hospital violence.

The first step in the Cause Mapping process is to describe the what, when, and where of an incident, and define the impacts to an organization’s goals.  In this case, the employee safety goal is impacted by the serious injury to a nurse.  The patient safety goal is impacted by the potential for injury to a patient.  The patient services goal is impacted by the fact that a violent attacker was able to bypass a weapons screening area.  It’s unclear from the information available whether other goals were impacted in this case.  Once that is determined the “?” can be replaced with the actual impacts to the goals, or “none”.

It can be helpful to determine the frequency of a type of incident.  Clearly, since about seven hours passed between two stabbings of nurses within the same county in California, the frequency of these types of attacks is much too high.

Next, cause-and-effect relationships are determined by beginning with an impacted goal and asking “Why” questions.  In this case, the injury to the nurse was caused by multiple stabbings.  The stabbings resulted from the nurse encountering a violent attacker and were impacted by the response time.  (In this case, security was searching for the man after he bypassed the weapons screening and was alerted to his presence when the attacked nurse began to scream.)   It is unclear how the man was able to bypass the weapons screening station, but ideally improvements that would decrease the possibility of entrants bypassing it in the future will be implemented.

Violence within hospitals has been increasing over recent years, believed to be due to a number of factors.  In addition, nurses and other hospital personnel have noted the difficulty in determining the potential for an escalation of violence in patients and other visitors.  According to the President of the Emergency Nurses Association, Deena Brecher, R.N.,”You need to be able to recognize when things are starting to escalate.  We know our behaviors can help escalate a situation, not intentionally.”

Many nurses are calling for establishment of workplace violence plans that would provide nurses and other hospital workers tools to identify and de-escalate potentially violent behavior, as well as provide additional protections against these types of attacks.  Some hospitals have begun using a mobile distress system, such as a help button worn around the neck that allows a worker to request backup in a situation that feels unsafe.

These solutions bring up an interesting discussion about prevention and blame.  The solutions listed above all require action by the part of nurses or hospital workers.  Many organizations attempt to determine the person to “blame” for a situation, and then assign corrective actions accordingly.  Clearly, nobody is trying to imply that hospital workers are at fault for these violent attacks (blame) but are rather trying to provide tools within their sphere of control to reduce the risk of worker injury (prevention).  Preventing all people prone to violence from entering a hospital, while theoretically more effective at solving the problem, is neither practical nor possible.  Thus it is hoped that providing hospital workers additional tools will result in reduced injuries from hospital violence.

To view the Outline and Cause Map, please click “Download PDF” above.  Or view the Workplace Violence Prevention for Nurses course offered by the Centers for Disease Control and Prevention (CDC).

Root Cause

Lack of Available Treatment Leads to Fatal Heroin Overdose

By ThinkReliability Staff

The death of a young man in New Jersey on September 23, 2010 from a heroin overdose was tragic, but part of a trend becoming more and more common.  His death mirrors many of the other fatal heroin overdoses and by examining the issues that led to this fatality, solutions that could reduce the death rates from heroin overdoses across the country (and perhaps beyond) can be developed.

We will examine this particular case in depth by using a Cause Map, or visual root cause analysis. First we capture the particulars of the issue – what, when and where – as well as the impact to the goals.  The fatality is an impact to the patient safety goal, while insufficient help being available is captured as an important difference, and is also an impact to the patient services goal.

Beginning with an impacted goal (in this case, the patient safety goal), we ask why questions to determine the cause-and-effect relationships that led to the impact.  In this case, the death resulted from a heroin overdose.  Overdoses typically result from use of this specific drug, with which overdoses are not uncommon.  Though it is not clear if this played a role in this particular death, heroin overdoses can occur after a user attempts to get clean and relapses.  If the user goes back to the dose from before ending use of the drug, the body (if it has been drug free for some period of time) is unable to handle it, resulting in the overdose.

In order to overdose, heroin use has to begin.  The use of heroin is rapidly increasing, with an estimated 669,000 users by 2012.  First-time users increased from 90,000 in 2006 to 156,000 in 2012.  The reason for the increase is believed to be the comparatively inexpensive cost compared to prescription opiates.  While a gram of heroin might sell for $100, crackdowns against prescription drug “pill mills” have increased the cost of prescription opiates (like OxyContin) to $1,000 a gram.

Once heroin use has begun, quitting is extremely difficult.  While withdrawal symptoms are not life-threatening, they are extremely unpleasant (to use a massive understatement).  Because they are not life-threatening, emergency care is limited (the victim in this case was unable to be admitted to the hospital) and many insurance companies won’t cover treatment, which can be extremely expensive.  In 2012, only 2.5 million of the 23.1 million Americans who needed drug or alcohol treatment received aid at a special facility.

Hope for overdose victims is available in the form of naloxone.  Since 2001, the use of naloxone by emergency responders resulted in reversal of over 10,000 overdoses.  The Affordable Care Act should improve insurance coverage for treatment, though it may take years for this to be in effect and, with the treatment availability shortage, likely means that not everyone will get the help they need.

However, solutions that address the problem of heroin use itself are being developed.  According to Attorney General Eric Holder, “Confronting this crisis will require a combination of enforcement and treatment.  The Justice Department is committed to both.   Since 2011, the DEA has opened more than 4,500 investigations related to heroin.  And as a result of these aggressive enforcement efforts, the amount of heroin seized along America’s southwest border increased by more than 320 percent between 2008 and 2013.   Of course, enforcement alone won’t solve the problem.  That’s why we are enlisting a variety of partners – including doctors, educators, community leaders, and police officials – to increase our support for education, prevention, and treatment.”  With the help of the federal and local governments, as well as dedicated families of users, it is hoped that the tide of heroin use will be turned.  This will be the most effective way to stop overdose deaths.

To view the Outline and Cause Map, please click “Download PDF” above.  Or click here to read more.

Root Cause

Hundreds Affected in ‘Unprecedented’ Ebola Outbreak

By ThinkReliability Staff

The ongoing Ebola epidemic in Africa is “unprecedented” due to its high mortality rate (up to 90%), geographic spread (at least 5 countries have reported cases of the disease, which has spread to urban areas as well), and difficulty enforcing quarantines that would reduce the spread.  As with many outbreaks, the factors involved are complex and wide-ranging.

We can address the issues contributing to the outbreak by capturing them in a Cause Map, or visual root cause analysis. This intuitive method ties impacted goals to cause-and-effect relationships, allowing development of solutions to all aspects of an issue.

First we begin with the impacts to the goals.  The outbreak began in Guinea at some point in early 2014, but was reported to the World Health Organization (WHO) on March 23, 2014.  The outbreak is still ongoing and has impacted Guinea the most, but has also spread to neighboring countries.  The strain involved is the Zaire Ebola virus, which is spread by bodily fluids.

At the date of publication, the virus has killed at least 101 out of 157 infected in Guinea alone.  The infections and deaths, as well as the spread of the disease, can be considered impacts to the public safety goal.  This is the first outbreak to have impacted urban Guinea, though there have been dozens of outbreaks in Africa over the past 40 years.

“Why” questions are used to determine the cause-and-effect relationships that resulted in the impacted goals.  Death typically results from bleeding or shock, which occurs due to infection with the virus and insufficient treatment. Infection results from the initial transmission (caused by eating raw infected meat), and the spread of the disease.  The spread in this case has resulted from the unusual migratory pattern, both because of the easy and frequent travel between countries but also due to an as-yet-unknown factor.  Normal outbreaks involve a much smaller geographic area.) Victims are contagious for a long time, meaning the disease is easily spread, and it has been difficult to enforce quarantine, because of mistrust of local authorities and foreign aid workers.  According to Stéphane Hugonnet  of WHO, “The mortality rate is extremely important.  Nine out of ten patients will die.  If we look at this from the population’s perspective, why would you go to a hospital if you have almost zero chance of getting out of it.”  However, with effective care, there is a chance of surviving Ebola.

However, providing that care is another challenge.  There is no cure for Ebola, possibly because financial incentives to develop a cure for a rare disease that primarily strikes poor African villages isn’t there. Care essentially involves keeping a person alive long enough for their body to be able to fight back, difficult in a country that has 0.1 physicians for every 1,000 people fighting a disease that rapidly replicates and – through an unknown mechanism – disables the immune system.

So what’s being done to end this outbreak?  Medical teams from Doctors without Borders (or Médecins Sans Frontières) and WHO have been dispatched to the area.   These medical teams may include anthropologists, to better address local concerns regarding the disease.  WHO has also recommended limiting personal contact and a on raw bush meat.  Meanwhile, researchers are working on a vaccine to prevent  transmission of Ebola.  It is hoped that these steps together will end this outbreak – and prevent future outbreaks as well.

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

Root Cause

Failure to Read Back Physician Order Causes Patient Death

By ThinkReliability Staff

A patient suffering from pneumonia required a bedside bronchoscopy in a California hospital.  In order to provide sedation for the procedure, the physician performing the procedure requested a dose of Versed.  Although the actual dosage requested was not recorded, the nurse gave the patient 2 milligrams via IV and, a minute later, another 2 milligrams.  The maximum published dose for Versed is 1.5 milligrams over no less than 2 minutes.

Because of the bedside scenario and the verbal order for medication, the nurse was required by hospital policy to repeat back the order.  He did not, so there was no opportunity for the physician to realize the error.  Within a few minutes, the patient stopped breathing and was administered CPR.  However, the patient never regained consciousness and died nine days later.

We can look at this issue within a Cause Map, a visual root cause analysis that addresses all the cause-and-effect relationships that resulted in the issue being investigated.  The analysis begins with the impacted goals.  In this case, the patient safety goal is impacted due to the patient death.  The failure to follow hospital policy regarding repeat back of verbal orders is an impact to the compliance goal.  The patient services goal is impacted by the overdose that was administered.  The overdose resulted in extra care required for the patient, an impact to the labor goal.  As a result of the issue, the hospital was fined $50,000 by the California Department of Public Health.  (Click here to read the report, which was used to create this blog.)

Beginning with an impacted goal and asking “Why” questions adds more detail to the analysis.  In this case, the overdose occurred due to the need for Versed and the larger than ordered dose.  The larger than ordered dose resulted from a miscommunication between the physician, who ordered the Versed, and the nurse, who administered it.  The nurse did not repeat back the order as required, and the physician did not request a repeat back.  Although the requirement was apparently for the person receiving the order to repeat back, patient safety is everyone’s responsibility.  Pausing the procedure to ask for a repeat back would have likely saved the life of this patient.

Not mentioned in the analysis was the conditions under which the order and procedure were performed.  Clearly ability to hear was a concern.  A study published in May of 2013 determined that background noise in the operating room can result in difficulty in communication between team members, not only by affecting team members’ ability to hear each other, but could also impair an individual’s ability to process auditory information.  Other studies have found that other environmental factors can impact medical errors.  Specifically, one study found that most medication errors were more likely to occur when the previous 30 minutes were hectic and involved staff member distraction.  It is unclear how much of a role the environment played in this case.

The hospital involved in the issue focused efforts on ensuring hospital policies were re-emphasized.  While this is a typical response in this type of situation, the training efforts must ensure that the importance of the policies is emphasized, possibly by using lessons learned from actual cases to demonstrate the risk of these policies not being followed.  Additionally, all staff must take responsibility for patient safety.  Even though the policy required repeat back by the nurse, other staff members involved with the procedure should have played a role in ensuring that the communication between members was adequate to ensure patient protection.

Want to learn more? See our webpage about medication errors in medical facilities or watch the video.

 

Root Cause

Stroke Treatment Inadequate in Many Cases

By ThinkReliability Staff

Research presented at the American Stroke Association’s International Stroke Conference 2014 showed that although 81% of people in the United States lives within an hour’s drive of a hospital equipped to treat acute stroke, only 4% received tPA, a drug which can reduce disability if given within 3-4 hours of the first stroke symptom, and the only drug approved by the FDA to treat stroke.

Researchers reviewed the records of 370,000 Medicare stroke claims from 2011.  (Annually in the United States, 800,000 people suffer from stroke.)  The low percentage of patients receiving the recommended (and only) treatment for stroke is a significant impact to both the patient safety goal (because of the disability that could be avoided with proper treatment) and the patient services goal, because so many patients are not getting adequate treatment after a stroke.

There are many challenges involved in administering tPA within the recommended time frame.  Administration is ideally done within 3-4 hours of the first stroke symptoms, but faster is better.  As the study‘s lead author, Dr. Opeolu Adeoye, M.D., M.S. states, “Every 15-minute delay in getting treatment increases the odds of that patient not being able to go home.”

Although 4 out of 5 patients live within an hour’s drive of a designated stroke center, a hospital which is equipped to treat stroke (and not all hospitals are), that still leaves almost 20% who aren’t.  The drug tPA can’t be administered before imaging confirms the stroke and that it is a non-bleeding-type stroke as administering tPA to patients suffering from a bleeding-type stroke can cause harm.  Even after a patient arrives at a stroke center, delays in imaging and treatment could increase the time before tPA is administered to outside the window.  The study also found that 60% of hospitals did not administer tPA to stroke victims, though it is the only recommended treatment for acute stroke.

Many potential solutions are being studied and implemented to reduce the risk of stroke after disability.

For patients: the best way to reduce the risk of disability from stroke is to prevent having a stroke in the first place.  About 80% of strokes are preventable and due to risk factors, such as smoking or obesity.  Maintaining a healthy lifestyle can reduce the risk of stroke.  If you or a loved one suffers from a stroke, contact an ambulance, as the ambulance will be able to direct you to a designated stroke center.  If you are driving yourself, it’s worth it to go to a designated stroke center, even if that means “bypassing another hospital that isn’t set up to deliver the necessary therapy, ” says Dr. Adeoye.

For paramedics: because paramedics tend to see most stroke patients before they get to a medical center, the search is on for a drug that could be administered prior to imaging (unlike tPA) so that stroke treatment could begin prior to arriving at a medical center.  A study looked at administering intravenous magnesium in the ambulance, but found that it did not improve stroke-related disability.  However, if another drug can be found that would, researchers are hopeful, as the median time for receiving the stroke treatment by ambulance was 45 minutes after symptoms began, and 74% of patients began treatment within an hour.

For medical centers:  The American Heart Association/ American Stroke Association’s national quality initiative Target: Stroke℠ aims to reduce the time between stroke symptoms and treatment.  Since its initiation in 2010, the percentage of patients treated within 60 minutes or less from hospital arrival has increased from less than one-third to more than one-half.  Additionally, the average time from arrival to treatment dropped from 74 to 59 minutes. These faster treatment times have reduced the percentage of stroke patients who died in the hospital from 9.9% to 8.3%.  The improvements were seen in patients regardless of age, sex, or race.

Ideally, the implementation of these solutions – and many more that are in the works – will continue to reduce the risk of, and from, stroke.

To view the Outline, Cause Map and solutions related to inadequate treatment for stroke patients, please click “Download PDF” above.  Or click here to read more.

Root Cause

Man Found Dead After Waiting 8 Hours for Emergency Treatment

By ThinkReliability Staff

A man seeking treatment for a rash at a Bronx hospital emergency room (ER) was found dead eight hours later, still in the waiting room, of as-yet unknown cause.  The incident is currently under investigation by the New York State Department of Health and the cause of death will be determined by a medical examiner.

When performing an investigation of a case like this one, it’s important to focus on the goals that were impacted by the incident and determine all the causes that resulted in the goals being impacted, not just finding one “root” cause.  In this case, the impact to the patient safety goal has clearly been impacted because of the death of a patient within the hospital itself waiting to be seen.  The patient service goal was clearly impacted because the patient did not see a doctor in the six hours – or more – between arrival and his death.  The schedule goal is impacted by the significantly higher-than-average wait at this particular ER.  Lastly, the labor goal appears to be impacted by insufficient staffing levels.

According to the hospital, the cause of the patient’s death is simple.  Per the Hospital spokesperson: “His name was called several times on several occasions, and he did not respond… People have personal responsibility when your name is called, you have to get up and see the doctor.”  The hospital says that all guidelines were met and, even goes so far to add that “probably this scenario in this shape and form has happened in any big hospital in New York City.”

Many don’t find that answer acceptable. Although hospital guidelines may have been followed, there’s no discussion of whether the guidelines were adequate.  It is apparent that the hospital guidelines do not include any sort of care or supervisor for patients prior to being called in to the waiting room.  However, there’s no discussion of whether that meets the standard of care expected of these hospitals.  As this lack of oversight resulted in the death of a patient going unnoticed – potentially for hours – in a hospital waiting room.

In addition, the incident has brought up questions about the impact of the long wait time.  The wait at the emergency room for this hospital is an average of 306 minutes – more than 5 hours.  The national average is 137 minutes and the average in the state of New York is 155 minutes.  When the patient was called, starting at 2 and a half hours after entry, he may have well been asleep, given that the next interaction he had was with a security guard who woke everyone in the waiting room at 2 a.m.  This periodic waking of people in the waiting room – meant to ensure that nobody was using the waiting room as a shelter – next happened at 6:40 AM, and is when the patient was found dead.  Unofficial reports suggest the patient may have been dead for hours.  The patient was last seen moving at 3:45 AM on security cameras.

These questions demonstrate the fallacy of the one “root” cause approach.  The hospital’s assessment begins – and ends – with placing blame on the patient for not responding to a call in the ER.  But this expectation may not be appropriate in all cases.  Although a shorter ER wait time may or may not have saved the life of the patient in this case, it would certainly ease the strain of an ER visit for most patients and potentially save a life.  There have been several publicized cases of deaths or significant disabilities resulting from waiting too long in the ER.  Certainly an incident like this occurring at a hospital merits a review of policies that allowed a man to die unnoticed by staff.

To view the Outline and Cause Map, please click “Download PDF” above.  Or click here and here to read more.

Root Cause

Inappropriate Antibiotic Risk Not Decreasing for Adults

by ThinkReliability Staff

Infections caused by bacteria (such as sinusitis and tonsillitis) respond to antibiotics; those caused by viruses (such as bronchitis and influenza) do not.  Prescribing antibiotics for viral infections will not treat the infection and contributes to the rise of antibiotic-resistant bacteria. This is known as inappropriate antibiotic use.   A recent study showed that efforts to reduce inappropriate antibiotic use have been effective in pediatric, but not adult, patients with acute respiratory tract infections.

To thoroughly understand the issue, we consider both the effects and causes of inappropriate antibiotic use.  A cause-and-effect diagram, or Cause Map, visually lays out these cause-and-effect relationships.

The effects of the issue are captured in a problem outline.  Effects are captured with respect to an organization’s goals.  In this case, the impacted goals are wide-ranging, so we look at them from a general health industry perspective.  Unnecessary antibiotic use can impact the person to whom they are prescribed, which impacts the patient safety goal. Unnecessary antibiotic use also increases antibiotic resistance, a growing public health problem with no easy answers.  This can be considered an impact to the public safety goal.  (For more information, please see our previous blog about antibiotic resistant bacteria and fungus.)

Besides patient and public health safety concerns, unnecessary use of antibiotics can result in unnecessary cost.  A program at a University of Maryland hospital that monitored antibiotic use resulted in $3 million in annual savings with no impact to care quality.  However, when the program ended, so did the savings.

In addition to capturing the impact to the goals in the problem outline, we can capture general information about the issue being analyzed, including important differences.  These differences can provide valuable information about potential causes to be evaluated.  An interesting difference noted in the study is that efforts to reduce unnecessary antibiotic use were effective for pediatric patients but not adults.  So far, the reason for the difference in pediatric and adult use has not been determined, but a decrease in inappropriate antibiotic use for children is a positive step forward.  (And not just because of antibiotic resistance.  A 2012 study found that antibiotic use in infants can lead to obesity.  Click here to learn more.)

After the effects of an issue are determined, cause-and-effect relationships that will lead to the causes of an issue can be developed by asking “why” questions.  In this case, several possible causes for inappropriate antibiotic use have been suggested.  Identifying causes allows more opportunities for solutions to address these causes.

Perceived pressure from patients to receive an antibiotic when presenting to the emergency room for an acute respiratory infection and difficulty making a definitive diagnosis to determine whether the infection is viral or bacterial are two of the reasons given for the continued inappropriate use of antibiotics.  Patient education can help.  A review of 89 studies in 19 countries found that prescriber access to education and advice or restrictions on prescribing antibiotics have been effective in reducing inappropriate antibiotic use.

A surprising increase in the use of antibiotics appears to be due to a reduced out-of-pocket cost borne by patients.  After Medicare Part D went into effect, reducing drug costs for some patients, a study found increases in antibiotic use for acute respiratory infections.  The study suggested that changes in patient cost-sharing may be effective in reducing unnecessary antibiotic use.

It’s likely that a combination of causes will be needed in order to reduce the prescribing of unnecessary antibiotics to a minimal level that can aid in the fight against antibiotic resistance.  Ideally, further studies will be able to develop lessons learned from the successful pediatric programs that have reduced inappropriate antibiotic use so they can be implemented for adult patients as well.

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

Root Cause

Patient Dumping Has Dangerous Results

By ThinkReliability Staff

“Patient dumping”, when hospital patients are improperly discharged, sometimes to dangerous areas, or even out of state, and sometimes without proper instructions for care, is a serious risk to patient safety and health and can result in serious costs for the hospitals and people involved.  In a recent case, a California hospital made a settlement for $250,000 in civil penalties and legal fees for leaving a patient at skid row without making any arrangements for her.

This case – and others like it – lead to obvious concerns for the health and safety of these patients.  That’s led city attorneys and homeless advocates to crack down.  Operators of homeless shelters and rescue missions in the area have installed “dump cams”, which allow them to identify cases where patients are being “dumped”.  In Los Angeles, the police department has stated they will arrest anyone who leaves patients outside a shelter.  And Mike Feuer, a city attorney, says, “Patient dumping is intolerable to me. I do have it in my mind to send a message to other hospitals that this won’t be tolerated.”

Although patient dumping appears to have lessened in recent years, it’s still a real problem.  In other newsworthy cases, another Los Angeles hospital settled a group of charges in 2011 when it discharged a disoriented patient – still in her hospital gown – by taxi and she was left in the street.  Yet another area hospital was sued for negligence in 2012 when it left a patient being treated for schizophrenia outside a rehabilitation center without notifying the patient’s family.  In 2013, the city of San Francisco filed suit against the state of Nevada, saying that a psychiatric hospital had issued bus tickets to California cities for mentally ill patients without making arrangements for them.

Even though the risks to patients are apparent (and financial costs to hospitals are possible), these problems continue to occur.  In cases where organizations don’t seem successful at ensuring the safety of its patients (or employees), the government will step in.  In this case, Los Angeles in particular has implemented a “patient safety zone” which encompasses most of the city’s downtown, where it is illegal to leave patients unless they are in the care of a family member.  Additionally, hospitals must obtain written consent from patients to take them to a place other than their home.

This of course can be tricky when dealing with homeless, mentally ill, or patients without relatives living nearby.  Although patient resources when dealing with these cases are limited – making proper discharge difficult in some cases – leaving a patient alone in an unfamiliar, dangerous area is never the right answer.

The impacted goals resulting from patient dumping, some potential causes, and the solutions that have been implemented by the city of Los Angeles are shown in a Cause Map, or visual root cause analysis.  To view the Outline and Cause Map, please click “Download PDF” above.  Or click here to read more.

Root Cause

Pregnant Patient Dies After Wrong Organ is Removed

By ThinkReliability Staff

A series of errors resulted of the death of a young mother in Romford of the United Kingdom on November 11, 2011.  Details of the patient’s condition and care provided by a  local hospital during a bout of appendicitis were recently released.  We can look at the causes that led to her death – and the death of her unborn baby – in a Cause Map, or visual root cause analysis.

With a complex issue taking place over several days like this one, it can be helpful to develop a timeline to aid in understanding.  In October, 2011, the 5-months pregnant patient entered the hospital and was diagnosed with appendicitis.  Surgery to remove her appendix occurred on October 23rd.  On the 29th, the patient was discharged from the hospital.  The pathology results became available on October 31st. These tests indicated that it was not the appendix that had been removed, but an ovary.  However, the results were not read by any hospital staff at this time.

The patient returned to the hospital on November 7, still in pain.  On the 9th, she suffered a miscarriage, at which point the pathology tests were read.  The patient underwent surgery to remove septic fluid from the diseased appendix, which had not been removed.  Two days later, on the 11th, the patient underwent a second surgery to remove her appendix, and died during the operation.

Before beginning an analysis it’s important to determine which organizational goals were impacted as a result of any issue being analyzed.  In this case, the patient death and miscarriage are both impacts to the patient safety goal.  (Both the mom and baby can be considered patients.)  As a result of the issues related to the patient’s death, eight hospital staff are being investigated, an impact on the hospital’s employees.  The death of a patient related to the wrong procedure being performed – in this case, the wrong organ was removed during her appendectomy – is a “Never event”, which is an impact to the compliance goal.  The Hospital Trust has accepted liability for her death, an impact to the organization.  The wrong organ being removed is an impact to the patient services goal. Additional required surgeries are an impact to the labor goal.

To perform our root cause analysis, we begin with an impacted goal and ask “Why” questions.  In this case, the patient death was due to multiple organ failure.  The multiple organ failure occurred because the patient had sepsis, and the sepsis was not immediately recognized.  (Although it appears that nothing was done to deal with sepsis until two days after the patient returned to the hospital, details on what was done have not been released.)  The sepsis resulted from the patient having appendicitis, and the appendix not being removed for 19 days.  Why was the appendix not removed for 19 days?  Instead of removing the appendix during surgery, the patient’s ovary was removed.  The results of the pathology report (which would have identified that the organ sent was not an appendix) was not read when available.  It is also not clear what the process was for reading these reports at the hospital, and how that process is being fixed.  It is known that the pathologist did not do any special reporting of the adverse results.

Now we get to the question, why was the wrong organ removed in the first place?  The surgeons were attempting to remove the appendix, which was inflamed as the patient was suffering from appendicitis.  Because they were performing open surgery, rather than laparoscopic, and the uterus was in the way of the appendix (due to the pregnancy), the surgery was being performed by feel, rather than sight.  (As you can imagine, this makes the surgery more difficult.)  During the surgery by feel, the ovary was mistaken for the appendix.  The ovary was possibly inflamed, due to the pregnancy, but another important issue is that the surgery was performed with overall inadequate expertise – specifically by trainees with no senior medical staff present.  (Senior medical staff were not required to be present, but due to the admitted difficulty of this type of surgery, that may have been a good move.)

As with many medical mishaps, any number of staff members could have improved the patient’s outcome.  Specifically, though the pathologist was only tangentially involved in the patient’s case, had she or he called the patient’s team immediately upon noticing that what was labeled an appendix was actually an ovary, the patient’s (and baby’s) life would likely have been saved.  Patient safety depends on everyone.

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