Archive for category Malpractice and Errors
The U.S. healthcare system is going to change or at least be updated in the coming years. So, when congress tinkers with the system what might be good changes and what might be bad changes? That is the $3 trillion dollar question! It would be fair to say most people and most congressmen do not understand U.S. healthcare — the prevailing notion is overwhelming complexity and way too much cost. However, this blog is going to make the case the key to understanding and the key to making changes is to keep your eyes on the results.
What results? It’s not complicated, it has to do with measurements. Consumer Reports and J.D.Power know we want to buy value. And, value in this case is the reasonable cost for wellness, longevity and successful treatment of disease. That’s it, three things. Whatever changes or tinkering are contemplated we just need to know those three things will be getting better and simultaneously costing less. Politicians have a really bad habit of saying the changes they propose will do the job. Nobody can predict what will work — there are always unintended consequences — so, any proposal must include a dedication to measuring the outcomes we want — if the change does not work it needs to be discarded as soon as possible. And, discarding what does not work can’t wait for the next election and should not wait until tomorrow. Simply, we want results, and we want the data as proof. On a hopeful note, if something works, keep doing it.
The above diagram describes U.S. healthcare. It is more simple than the systems in other countries. The system is linear — people, illness and unlimited money on the left side pass to the results on the right side. This is a flow diagram of the system. The complexity can be hidden by thinking in terms of the five boxes. Later, some of the complexity will be discussed. First, consider the boxes:
- Money to pay for the system. The money people earn is paid to the health care system. Money is money — it does not matter if the money comes by way of taxes, insurance or cash. Funds that do not come from insurance come from the other sources. This is the cost of U.S. healthcare which is about $3 trillion. Don’t pay the money, you don’t get healthcare.
- The healthcare providers. Traditionally we only think of doctors, hospitals and drugs. We often overlook the other things in the box. Things we don’t like, things healthcare providers would like to see in another box. These other things are hugely expensive and fully under the control of the healthcare providers. Unnecessary treatment is perhaps one of the worst — treatment or tests that are not needed. For example, an EKG done as part of a yearly exam on a healthy person. Profit is in this category. Clearly, no profit, no healthcare system. But, profit beyond what is needed is just waste for the system — it is money that leaves the system and does not come back. Inefficiency comes in many forms. Failing to prevent diseases early, only to spend more money later is supremely inefficient. Corruption is a problem in every human endeavor. Errors turn huge amounts of money into waste. The money spent on medical liability suits is just the tip of the iceberg. Money spent to prevent errors is minuscule compared to the money spent on drug marketing.
- Who gets healthcare? Everybody. The aggregate need for healthcare is fairly stable for the system. But, for an individual the need is hugely variable — an auto accident is not predictable. And, when disease strikes most of us can not afford the cost without insurance. Statistics show 50% of Americans do not have access to $4oo for an emergency. The very people who don’t have emergency funds are the very people who do not want to purchase health insurance. Sadly, those people end up in bankruptcy while the system grudgingly provides the care. Now that more people have insurance those without may find less compassion from the providers. Many feel there are freeloaders in the system — people who do not contribute. Does a birth defect, mental illness or low IQ make people freeloaders — that’s an ethical question which is beyond the scope of this discussion.
- Waste. In monetary terms this about $1.5 trillion dollars per year with a huge death toll in the US. A hospital acquired infection is very expensive and kills many of those affected. The high profile infections from spinal injections are just the tip of the iceberg, again. Re-hospitalization for an unresolved health problem is another example. Paying $800 for a $10 epinephrine injector is another example.
- The results. We want those good results. Not just for cancer patients, not just for heart attack victims, not just for you, but for me too. We don’t want promises, we want results. In this age of smart phones and millions of apps there is no excuse for failing to have the data to prove the system is working in our hands every day. We want the results today, not after several years of scrubbing the data in some moldy university. We all must keep our eyes on the results and hold our elected officials accountable.
Complexity. Medicine is a science and by its nature is very complex. Open heart surgery is a good example — there are few people who understand the issues involved. But, the system, from the patient’s view does not need to be complex. In one country the cost of hospitalization is $400/day — the people there know exactly how much the illness will cost. In another country, the prices of office visits are posted in the waiting room — it does not matter what insurance company you might have. In another country all the providers use the same medical record system — not a big deal to move or see a consultant. We seem to tolerate the complexity of our system and think it should be as difficult to understand as heart surgery.
The US pays about twice what other countries do for similar or better care. There is enough money in our system now. Our problem seems to be in the area of wasted money and effort. It seems unlikely that just reducing payments to providers will reduce errors and wasted money — this supply-side economics does not get to the real problem. More than likely, lower payment to providers will only result in lower income for them and perhaps more errors and unnecessary services. But, if it works, do it.
Back to the initial warning. Keep you eyes on the results of the system and the cost. Whether any economic hypothesis proves correct is irrelevant. What matters is the system must move in the right direction, always.
There is a lot to recommend the quality improvement method called “Plan – Do – Study -Act” or PDSA. The idea is to plan a change to a system of care, do the plan, make measurements to study the results then act to change the system to get better results. This is an ongoing process. Congress seems to be mired in a system of management which is one hundred years out of date — if anything, that’s what needs to change first.
According to a recent report in the BMJ medical errors are the third leading cause of death in the United States. The chart below shows where the estimated number of deaths from medical error fit into known mortality data. The article makes the remarkable observation “medical error” is not allowed as the cause of death to be recorded on death certificates. Indeed, when “medical error” is entered into the ICD 10 look up site for CMS here is the result: There are no ICD-10 Codes that match that fragment.
To be fair, the rule for death certificates is the cause of death must be the final cause. For instance, even though smoking causes heart disease which leads to a myocardial infarction (heart attack), the allowed cause of death is myocardial infarction rather than smoking, which is truly the root cause. That’s just how the system has worked for decades. So if a medical error causes a myocardial infarction the doctor must list myocardial infarction as the cause of death — in most states the qualifier “due to …” is allowed and smoking could be entered by the doctor. Interestingly, if “medical error” is listed the death would become an “accidental death” requiring review by the medical examiner (not something most physicians or families want).
251,000 deaths in the United States are caused by medical errors. This fact has been hidden, like many other underlying causes of death. The effect is that research into medical errors is lacking, the funding for research is lacking and the problem is ignored. Why should heart disease research be funded when some “cardiac” deaths are actually due to a nurse giving the wrong medication to the patient. The research really needed is to find how to prevent the medication error.
Although the graph above is bad enough, consider that the size of the bars for heart disease, cancer and COPD should all be smaller — because the real cause, in many cases, is medical error.
The number of medical errors is staggering. The errors that cause death are just the tip of the iceberg. Lots and lots of medical errors only have minor consequences. Every physician makes several errors each day. Every review of a medical chart reveals numerous errors. Care providers are astounded by this news. However, simple things like not seeing a lab report with a low blood count until the following day is an error even if nothing bad happened as a result. Of course, it would be malpractice if something bad did happen as a result.
The prevailing notion in quality improvement circles is, “don’t waste your time unless there was a bad outcome.” This notion comes from a lack of staff, and intense criticism when a provider is involved. “What do you mean there was an error, my care was excellent, nothing bad happened, you should be fired.”
Although there is a real tendency to want to blame an individual provider, both small and large errors are the result of poor clinical processes. Even errors due to provider fatigue and lack of sleep are truly process problems — a provider who worked during the night should not be working the following day. A provider sick with a cold should not be working, but so often they are.
What needs to be done:
- Just like alcoholism, the first step is to admit there is a problem.
- See medical care as a process. Strive to make sure the same problem gets the same treatment every time in every location. OR change the process.
- Stop seeing medical care as art. What work of art ever killed a person?
- Separate the compensation for medical errors from improving medical processes. Injured patients need quick compensation and medical care needs quick changes to improve. Lawsuits seem to have little impact on process of care — if lawsuits prevent errors why are we in this mess?
- Establish a non-profit foundation to advocate for reducing medical errors — something the public might be very willing to support. Sorry, Heart Association, Lung Association, Cancer Association — some money needs to go toward reducing errors. After all, if a wonderful heart medication is given to the wrong patient what good is that?
Saying “sorry” is the human thing to do. Doctors and nurses should say it when they feel it.
Saying “sorry” seems to have two meanings: 1) something bad happened and I understand your emotions 2) something bad happened and I had some connection with the event for which I feel partly responsible. Bad things do happen in health care but “sorry” is a very uncommon utterance for health care providers.
Dr. Abigail Zuger writing in the New York Times 7/14/14 “Saying Sorry, but for What?” compared how she felt about a plumber who broke a valve in her house with medical personnel who broke other things — neither said “sorry.” Sorry truly does not fix anything; but, the absence of “sorry” is infuriating.
The problem is ego. Ego infuses some health care providers with the notion bad things are an act of God but good things are an act of ME. Absence of “sorry” is a sure sign of defense (a defense of self). Perhaps the health care provider was spanked as a child or yelled at by teachers. Who knows … ego has gone wild.
Quality health care depends on people believing errors are due to system failures. When providers fail to embrace that philosophy they fail to correct problems. No failure, no correction.
A fall in the hospital can be deadly. Recently, a family member fell in a room while no nurse was present and they died. The nurse did not say “sorry.” There was no acknowledgement of responsibility. No acknowledgement the system was at fault, no realization there was a better way, and no reason to prevent future deaths. The simple statement “sorry, I wish I had been there to stop the fall, we will investigate this to help others” would be the right thing to say, and believe.
Lawyers are not the cause of excessive health care ego. However, lawyers with the threat of suit are a convenient excuse. When bad things happen honesty and caring are much more likely to assuage the displeasure of a family than stonewalling.
It sounds like a paradox: science studying itself. But, that is exactly what is happening in medicine. Basic research has led to applications of the research and the applications are studied for effects, benefits and cost. For example: invent robotic surgery and apply it to patients, then set it up as a program in an operating room and try to improve the technique and patient selection, and finally evaluate the program to see if it meets stated goals of quality and cost and decide if it should continue and under what conditions.
This huge simplification helps with terms doctors and hospitals often talk about:
- Discover and apply — called research.
- Try to improve — called quality improvement (QI).
- Continue the effort? — called program evaluation (PE).
Patients can be subjects of research. But, participation in research requires explicit permission since the outcome is not known and it could be bad.
If we knew what it was we were doing, it would not be called research, would it? (Albert Einstein)
Patients are hopefully impacted by quality improvement since the purpose is to make things better and thus no patient permission is required. As part of QI a hospital may try to make sure antibiotics are given before surgery because there is research evidence the practice reduces infection. Quality improvement focuses on a cycle of planning, doing, study and revision. QI has become a huge area of study with numerous books and journals on the subject. Virtually every hospital has a quality manager who is charged with improving the care at a hospital.
Patients are only indirectly affected by program evaluation. Clinics and hospitals constantly evaluate programs for positive or negative effects. Whether programs continue depend on such studies. People may read about evaluation of medical programs like care at VA hospitals and may be impacted by decisions of policy makers based on such evaluations. PE is likewise an important and growing discipline.
The concepts of research, quality improvement and program evaluation do tend to overlap. One could imagine using QI techniques to improve the quality of research. And, one could imagine research to find the fastest way to do program evaluation. However, research is mainly for the purpose the researcher decides. Whereas QI and PE are mainly for patient care, business or institutional purposes.
Quality healthcare depends on QI and PE. Patients often don’t see these efforts in action. But, ineffective QI and PE are hazardous to your health. Although doctors and hospitals don’t like the idea: law suits are a warning flag of inadequate QI and PE.
Jennifer Levits reported in the Wall Street Journal 12/18/13 “Doctors Eye Cancer Risk in Uterine Procedure“. She recounted the story of Dr. Amy Reed who had a hysterectomy. The uterus contained fibroids and the fibroids contained cancer. The procedure was done with an instrument, the morcelator. In kitchen terms it is a combination blender and vacuum cleaner. It is used during laproscopic surgery to chop up things (like a uterus with fibroids) and remove them through a small incision in the abdomen.
The problem is the morcelator does not remove all the tissue. A few cells escape the vacuum and they are left behind in the abdomen. If those cells contain cancer the cancer is then planted in the abdomen later to grow and likely kill the patient. Dr. Reed developed the seeding of cancer and claims other procedures would be better. Traditional surgery removes the uterus and fibroids intact with less chance of spreading any unsuspected cancer.
Here is what the package insert that comes with the morcelator says:
CAUTION: … use of the … Morcellator may lead to dissemination of malignant tissue.
So what are the statistics?
- 20% – 40% of women will develop fibroids
- 1 in 1000 cases of fibroids contain cancer
- intact removal of fibroids with malignancy failed to stop the malignancy 19% of the time
- morcelator removal of fibroids with malignancy failed to stop the malignancy 44% of the time
The big question is: should a morcelator be used if a woman has fibroids because it may double the risk of spreading an unsuspected cancer?
The simple answer is NO, because there are other surgical options. But, will women accept that answer? The laprosocopic procedure has less pain and quicker recovery, so the answer turns out not to be so simple. There are many forces at work on the decision to continue to use the morcelator. The analysis of these forces is called force-field analysis which was originally described by social scientist Kurt Lewin in the 1940’s. The following is such an analysis (the rating of force vectors is by the author of this blog):
So, what will happen? It seems at this point the morcelator will continue to be used. But, the risk remains. Law suits will continue. Perhaps a safer device will be developed. Perhaps a high risk of litigation will be perceived by gynecologists and the malpractice insurance companies as being too great. Such risk will lower the forces from doctors and perhaps tip the balance. Time will tell.
Update (8/1/14): since the original post the FDA issued a warning about morcelators. Yesterday, Johnson & Johnson, the major supplier, stopped selling morcelators. Looks like the forces to abandon the morcelator have been joined by the FDA and the manufacturer.
Update (2/27/15): yesterday the Wall Street Journal reported United Health (insurance) requires surgeons to obtain permission for procedures that might use the morcelator — “another blow” to the device. At this point any surgeon using the device must feel like they are walking a tightrope without a net.
Making a diagnosis is difficult. And, doctors sometimes get it wrong. “Wrong” is often harmless, usually expensive, and sometimes deadly.
An article about incorrect diagnosis appeared this month in the British Medical Journal Quality and Safety which has been widely reported, including by the Wall Street Journal. Dr. Tehrani and his co-authors correlated health insurance claims (diagnosis) with malpractice suits. They found “diagnostic errors appear to be the most common, most costly and most dangerous of medical mistakes.”
One might think the errors happen because the underlying problem is very rare. On the contrary, the bulk of errors happen with common conditions.
Another article this month in JAMA Internal Medicine by Dr. Singh and co-workers reported on common types of diagnostic errors — many of which were common in primary care: (italics are blog examples)
- no chest x-ray for cough and high fever
- no chest x-ray for cough and high fever
- Decompensated congestive heart failure
- no BNpeptide checked
- Acute renal failure
- no check of basic metabolic panel for fatigue
- ignoring Mammogram findings or blood in sputum
- Urinary tract infections
- not checking urinalysis or treating soon enough
The flaw in the process that contributed to the wrong diagnosis included:
- Inadequate patient encounter (too short or not focused on problem)
- Not seeking referral when needed (like not getting a cardiology consult for chest pain)
- Patient related factors (not returning for follow-up)
- Not taking risk factors into account (like family history of colon cancer)
- Losing track of test results (urinalysis report filed but not viewed)
- Not getting the right test (not getting a chest x-ray for shortness of breath)
Problems at the time of patient encounter are a major contributor:
- Poor history taking (provider did not listen or ask questions)
- Inadequate examination (provider did not examine problem area — like a breast nodule)
- Inadequate testing (not considering a colonoscopy for blood in the stool)
When a person has a health problem the whole idea is to connect the dots …problem…diagnosis…treatment. If the diagnosis is not correct then good treatment is disconnected.
Providers often do not consider enough possible causes for abnormal findings. Those possibilities are called the “differential diagnosis”. There are books and several free sites on the Internet that provide such lists. One such site is DiagnosisPro. If you like other sites leave a comment please. Some electronic record applications include a differential diagnosis automatically — nice feature which should always be installed.
So, what is the solution? Most experts agree, the quality of the provider-patient interaction must improve. Providers need to follow known guidelines plus use differential diagnosis aids. Patients need to look out for themselves by using the Internet or books to understand symptoms and test results. The best solution is a stronger partnership between patients and providers. See earlier posts in this blog about shared decision-making and patient centered care.
Can all errors be prevented? NO. To err is human. The point is to minimize the errors, and there is obviously a lot of room for improvement.
Diagram of the Multi-Team System (MTS) for patient care is from the AHRQ web site. This is an idealized concept of what should happen that often does not happen.
A recent article in the New England Journal of Medicine (NEJM) described an unfortunate but all too common situation in hospitals. In this article a patient was very sick in the intensive care unit with respiratory failure (on a ventilator) and with an unusual skin rash. 40 doctors and far more nurses were involved in the patient’s care. So many people, in fact, that nobody knew who was in charge and except for ordering more and more tests nobody did anything. The NEJM article sites the “Bystander Effect” which is the tendency for everyone in a big group to assume someone else will act. Finally, the patient was saved by an acute problem which forced a doctor on the spot to actually do something.
Quality care is doing the right thing at the right time. On both counts the NEJM case represents low quality. Other factors beside the “Bystander Effect” may have been at work. Perhaps the “Silo Effect” where all the care givers were in their own silo without regard to the big picture. Perhaps it was the “Swiss Cheese Effect” where errors on several levels lined up and the patient fell through. But, most likely, the low quality was due to poor communication — the usual suspect. If the care team does not talk the sense of urgency and the sense of danger are lost. The patient was in grave danger!
So, you think this would never happen to you or happen at your local hospital? Think again. It happens all the time when more than one doctor is involved (including on-call doctors). Are there solutions? Yes. The most desperate need is always to designate who is in charge for every minute and every hour and every shift — including doctors and nurses. In-charge is not a title, it means willing and able to act. There should be a sign in each patient’s room with the name of the in-charge doctor and in-charge nurse. Also, there should be a sign on the intensive care door: “All consultants who enter must talk to the doctor in-charge before leaving”.
On a more hopeful note, research has some useful ideas for the teamwork-challenged hospital. Here are some pre- and post-shift check lists from the STEPPS program:
During the brief, the team should address the following questions:
___ Who is on the team?
___ All members understand and agree upon goals?
___ Roles and responsibilities are understood?
___ What is our plan of care?
___ Staff and provider’s availability throughout the shift?
___ Workload among team members?
___ Availability of resources?
The team should address the following questions during a debrief:
___ Communication clear?
___ Roles and responsibilities understood?
___ Situation awareness maintained?
___ Workload distribution equitable?
___ Task assistance requested or offered?
___ Were errors made or avoided? Availability of resources?
___ What went well, what should change, what should improve?
The question of “who is in charge” is critical for hospital care. Trauma surgeons seem to have this issue mastered (they are in charge) but other doctors are in a quandry when more than one is involved. Patient safety demands US hospitals do better!
Finally, a comment about cost. The lack of someone in-charge leads to high cost. The NEJM article itself failed to mention the cost of 40 doctors working on the case. In this time of rising health care cost the nation can not afford such lavish use of resources.
What causes medical malpractice suits for health care providers and hospitals? The simple answer is ERROR. The most serious errors happen in hospitals but the most frequent errors happen in outpatient clinics.
James Reason (“Human error: models and management.” Bmj 320.7237 (2000): 768-770) is well known for his study of the causes of human error. He breaks error into 3 main categories:
- Skill based error (not paying attention to the right things)
- Rule based error (not following the right rule or following an incorrect rule)
- Knowledge based error (not knowing or incorrectly concluding)
Health care providers make human errors all the time. When the error actually causes injury the injured party will sometimes seek compensation through the legal system.
If humans “do the best they can” performing a task the error rate is about 10%. For example, if a nurse is to administer a medication to a given patient there will be an error 1 out of 10 times. Wrong patient, wrong medication, wrong dose, or wrong time just to name a few possibilities. The goal in error reduction is an error rate in the 1 out of a million range. Humans alone can not do that! Prevention of errors absolutely requires systems — sometimes as simple as a checklist or sometimes as complex as a robot that packages medications with a barcode and later scanning of the code before drug administration to a patient.
Medical malpractice suits (excluding the presence of legal malpractice) arise out of a failure to have adequate systems to control error. But, malpractice suits are only the tip of the error iceberg. All interventions in health care must have active quality monitoring and error prevention systems. Health care providers, to their patient’s detriment, yearn for simpler times — just the good old doctor-patient relationship in a small office with low overhead costs and no malpractice lawyers. Without a focus on quality and standards the good old days were really the bad old days of medical care.
The cost of medical malpractice to the US healthcare system is estimated at $55 billion or 2.4% of the overall system cost. However, the statistic misses the point. How much is spent on preventing errors? A wonderful trade-off would be to spend that much money to prevent errors in the first place rather than paying compensation for errors after they happen. Lawyers say they see the same errors over and over. One patient, one error, one trial then repeat with no intent to stop the cycle. This is a major flaw in the US justice system. Each settlement and each trial should result in some change to the system where the error happened. The usual outcome is “it’s your fault”, “try harder” and “do the best you can” — which are clearly failed quality strategies.
Prevention of errors is a costly endeavor (although well worth the cost). The following are examples of systems to reduce errors:
- Electronic medical records
- Electronic prescribing (computer checks for mistakes)
- Checklists for surgeons
- A strong quality management program both inpatient and outpatient
- Standardized orders in the hospital
- Standardized protocols for outpatient treatment
- Mandatory involvement of health care providers in quality improvement
- Development of a culture of safety for health care providers
- Formalized hand-off when changing shifts or going off call
- Barcoded medication administration
- Robotic packaging of medications
- Hourly nurse rounding to prevent falls
- Include the patient in the effort to improve safety
The above list is seriously incomplete. Health care is complex and changing. The systems to prevent error need to be adapted to the circumstances and to our changing understanding of treatment.