Diagnostic Errors and Root Cause Analysis
September 25, 2013
Many patients seek to health care to treat their medical problems and they expect to receive high quality of care. Unfortunately, for some patient, instead of receiving high quality of cares, they receive cares that harm them actually or are potentially harmful to them. As in the story of Lewis Blackman, a series of error had happened such as diagnostic error, communication error, transparency error, medication error, and negligence. However,Diagnostic error has become the most concerned for all health care providers and cost billions of dollars. Identifying and reducing diagnostic errors is a major priority and challenge. The purpose of this paper is to investigate how diagnostic errors happen and how to prevent the diagnostic errors. Background
Lewis Blackman, who was a healthy 15 years old, underwent elective surgery what believed to be a low risk surgical procedure for the defect called pectus excavatum (a crease in chest cavity) at Medical University of South Carolina Children’s Hospital and died four days after the surgery without knowing the cause of death. His death is learned only after the autopsy is done. The autopsy shows that his abdomen is filled with almost three litters of blood and digestive fluid. He is announced dead after thirty hours of showing sign and symptom of internal bleeding while health care providers, who were taking care of him, underestimated and missed signs of internal bleeding (Helen Haskell, 2009). Errors Identification
Diagnosing a condition and making safe and effective patient management decisions involves a critical thinking that related to individual clinical judgment and skill over time. To investigate and prevent the diagnostic errors, answering the following questions is necessary: what happened, why it happened,how it happened, and what solution can be developed to solve problems. According to Norman & Eva (2010), there is variety of mechanisms that might cause diagnostic errors and one of those mechanisms is cognitive skills. Cognitive skills. Cognitive skill errors (processing biases) are more common than other errors. Under cognitive skill errors, there is variety of biases: availability, representativeness, confirmation bias, and premature closure. Availability is tendency to judge diagnoses as if they recall the diagnosis from memory. Representativenessis the tendency to diagnose based on the similarity of disease.Confirmation biasis the tendency to look for data to confirm the hypothesis. Prematureclosure means to stop gathering data or considering other possibilities after having an initial diagnosis(Norman & Eva, 2010). To have better understanding about these biases, let use the examples from Lewis’s story, thedoctor diagnosed blocked intestine in a patient with abdominal pain, which turns out to be caused by internal bleeding. Lewis has been taking full adult dose of Ketorolac for three days and showing sign and symptom of hypovolemic shock. The significant side effect of Ketorolac is gastrointestinal bleeding. The nurse fails to assess Lewis completely for sign and symptom of internal bleeding and she jumps into conclusion that pain is caused by gas pain. The chief resident also diagnoses probable ileus (blocked intestine) which is not related to fever and tachycardia.The chief resident might be influent by the bias of availability because he may have other patients with blocked
intestine previously or being affected by representativeness because patient may show sign and symptom similar with blocked intestine. Confirmation bias is another bias that affects the decision of doctor because he tends to seek data consistent with his hypothesis of blocked intestine and neglect data coherent with internal bleeding. Lastly, premature closure is another bias that affects the process of making the right diagnoses because the doctor did not gather all the necessary data prior making the decision. The doctor might come into consideration of internal bleeding if he ordered complete blood count. Complete blood count is a routine blood test and it would show that Lewis is bleeding internally. Computed tomography scanmight also help rule in or out of blocked intestine. On the fifth day, Lewis’ abdominal pain stops, the nurse concludes immediately that is good sign without assessing Lewis. The aid is unable to take vital signs and unable to detect blood pressure. The resident and nurse believe the blood pressure devices are broken instead of consulting with veteran doctor (Helen Haskell, 2009) Quality Measures
According to Institute of Medicine who defined six quality and safety measures, the health care providers who were taking care of Lewis have violated these measures. In the Institute of Medicine, there are six different measures: safety, patient-centered care, teamwork and collaboration, informatics, timely, and continuous quality improvement; and each of them addresses a different aspect of safety and quality of health care system and aims to improve health care system (Cherry & Jacob, 2011).
Patient-centered carefocuses on needs and rights of each individual patient (Cherry & Jacob, 2011). Cognitive skill errors may lead to the ignorance of patient’s needs and rights. Considering patient and family members as part of the treatment plan is necessary because they know how they are doing better than anyone else. For example, when Helen expresses her concern of Lewis’ pain, nurse told Helen that the pain is gas and Lewis needs to move around although Lewis is having slight fever, cold to touch, and pain is five on scale of five (Helen Haskell, 2009). This premature closure has led to fail to treat Lewis properly. Another example is when Helen calls nurse many times to complain about Lewis’ pain, she tries to convince Helen that
Lewis is not walking enough to relief the gas (Helen Haskell, 2009).Availability bias and responsiveness bias may also affect nurse’s belief on the cause of pain and lead to wrong diagnosis. Nurse may have seen many patients who have gas pain prior to Lewis and they use their past experience of gas pain to diagnose Lewis’ problem without considering other possibilities. The nurse also seeks data relevant with gas pain while she ignores the data relate to internal bleeding. The other failure of patient-center care in Lewis story is no one notices the deadly side effect of Ketorolac although it is stated clearly about the risk.
Teamwork and collaboration are the ability to work effectively within nursing and inter-professional teams to achieve the expect outcome (Cherry & Jacob, 2011).Lack of teamwork and collaboration might cause diagnostic error, delay in treatment, and medication error easily.Communicating data and information with each other will help health care providers aware of patient health status and prevent from delay in treatment and diagnostic error. For example,in Lewis’ story, resident, chief resident, and nurses do not communicate and teamwork with each other. The nurse jumps into conclusion of gas pain without considering other possibilities, gathering additional data, and consulting with veteran doctor. When the aid unable to detect the blood pressure, the resident and nurse believe the blood pressure devices are broken without assessing Lewis (Helen Haskell, 2009). The resident and nurse should consult with veteran doctor instead of spending almost two hours to detect undetectable blood pressure without notify veteran doctor. If the resident and nurse consult with veteran physician about the incident, the veteran physician might know how to act on these crisis symptoms. Teamwork and collaboration is very important in preventing diagnostic error and improving patient outcome.
Evidence-based practice: “integrate best current evidence with clinical expertise and patient family preferences and values for delivery of optimal health care” (Cherry & Jacob, 2011, pg. 472). Diagnostic errors occur as the result of the consequences of not following evidence-based practice. Diagnostic errors correlate with incomplete history taking or physical examination, inadequate data and information, failure to consider correct
diagnosis, and bias toward a single explanation (Norman & Eva, 2010). Lewis has shown signs and symptoms of possible perforation and internal bleeding and health care providers have failed to pay attention to those signs and symptoms. Health care providers tend to seek data relevant to the diagnosis they made and ignore important data that indicates considering of other possibilities. Nurses and residents who take care of Lewis tend to use their past experience to diagnose Lewis’ problem. Based on their experience, patient is at high risk for constipation if he/she is taking pain medication and not ambulating. Failure to assess and gathering additional data will lead to diagnostic error and delay in treatment.
Continuous quality improvement: “use the data to monitor the outcomes of care processes and use improvement method to design and test changes to continually improve the quality and safety of health care systems” (Cherry & Jacob, 2011, pg. 472).Gathering additional information and data, considering other possibilities, and consulting with attending physician prior to make final decisions will avoid unnecessary harm, improve quality care, and help prevent diagnostic errors also. Lewis’s problem will be treated correctly and his outcome may have been improving instead of declining if health care providers get more data and consider other possibilities.The resident and nurse should take into consideration when patient is taking Ketorolac without urine output. Decreasing urine output will lead to accumulate the concentration of Ketorolac in the body and increase the side effect. To avoid diagnostic error and improve the accuracy of diagnosis, health care providers must look at data carefully, gather additional information if needed, consider other possible diagnosis, and consult with others in health care team prior to make the final decisions.
Safety means to avoid any of unintended harm for patients (Cherry & Jacob, 2011). Premature closure may harm patient and put patient at risk for injuries as diagnostic errors occur. Doctor will make incorrect diagnosis if the data is not fully collected and lead to delayed or wrong treatment. This put patient at risk for injuries potentially.Health care providers havefailed to provide safety care for Lewis. In Lewis’ case, Lewis’ belly is distended and hard, temperature drops, eyes are sunken, skin grows pale, and
experience great pain (Helen Haskell, 2009). These are signs and symptoms of possible intestinal perforation and internal leakage instead of blocked intestine and the resident and nurse should assume the worst and gather more data either to rule it in or out. Nurse who took care of Lewis has failed to assess Lewis and jumped to conclusion that Lewis is not walking enough to relief gas pain. The resident should order computed tomography scan to rule out intestinal bleeding and confirm his diagnosis.If the nurse, residents, and veteran doctor communicate with each other, Lewis’ problem will not get worse and worse instead Lewis might be rescued.
Informatics means to use technology and information to ensure patient safety, communicate, manage knowledge, and support decision-making (Cherry & Jacob, 2011). Using technology and information to support decisions can prevent diagnostic error because health care providers have more data and information prior to make the decision. Technology and information also help improve the accuracy of diagnosis. In Lewis’ story, the resident and nurse have missed many chances to prevent the error.The resident might have figured out Lewis’ actual problem if they used computed tomography (CT) scan to confirm the diagnosis of blocked intestine. The rescue might have been successful if CT scan has ordered to rule out internal bleeding. The resident and nurse may consult with veteran doctor about Lewis’ situation. Strategies for Diagnostic Error Prevention and Barriers to Error Prevention
The most important goal is to reduce the diagnostic error incidents because these errors may harm patient and cause potential injury. To develop strategies for diagnostic error prevention, understanding the causes of diagnostic errors (see Appendix A) is necessary and the causes have been addressed throughout the paper. Strategies Number 1: Using simple and transparent strategies, and interaction with specific knowledge in memory-‘Thinking of the first thing that comes to mind’ ‘think of other possibilities’. This will help to prevent premature closure, availability bias, and representative bias. However,the barrier for this strategy is that it results in poor performance because it places a heavy load on working memorywhich has limitations in speed and size (Norman & Eva, 2010). Strategies Number 2:Gathering additional data, ordering the appropriate
tests, reviewing patient history, and completing physical assessment prior to jump to final decision may prevent the occurrences of diagnostic error. They should not gather data relevant with their diagnosis and ignore data consistent with other diagnoses. They must consider other possibilities and reflect their own thinkingprior to diagnose. Health care providers may use tests to confirm their hypothesis prior to make diagnosis. The doctors and nurses should not make assumption based on patient’s presentation because patients might have different medical problem although they have similar presentation. However, the barrier of the strategies isgathering additional data may help to prevent diagnostic errors but there may also be danger in seeking additional information because health care providers perceive the information they get as more accurate and have greater confidence in their subsequent decision although the information are identical. It will also increase the cost of laboratory investigation as they quest more information to confirm their diagnosis(Norman & Eva, 2010). Strategies Number 3:The residents and nurses have sufficient medical knowledge and that the errors reflect the inappropriate cognitive skills. Usingsimulation and case studies may help improve health care providers’ clinical reasoning skills. The clinicians would be able to avoid error if they have oriented to the errors (Graber, Franklin, & Gordon, 2005).The barrier of this strategy is the collaboration of health care members. Are they willing to join the simulation? They might feel that they already know this, so they do not have to join. It also requires time and financial support. Are they willing to spend time to join the simulation? Who will be responsible for financial supportif they are paid to join the simulation? Conclusion
Diagnostic error is major concern in health care system and it is difficult to define. However, the causes of diagnostic error have been identified throughout the paper. Six qualities measures related to Lewis’ story have been defined. Diagnostic error is preventable because human causes the error. It, however, will taketime and need financial support and collaboration among health care providers. Strategies for prevention and barriers of prevention have also mentioned.
Norman, G., & Eva, K. (2010). Diagnostic error and clinical reasoning. Medical Education, 44(1), 94-100. doi:10.1111/j.1365-2923.2009.03507.x. Cherry, B., & Jacob, S. (2013). Contemporary nursing, (6th ed., pp. 374-392). St. Louis, Missouri: Elsevier Mosby Inc. Haskell, H. (Performer) (2009). The faces of medical errors…from tears to transparency [DVD] Graber, M., Franklin, N., & Gordon, R. (2005). Diagnostic error in internal medicine. Arch Intern Med, vol 165.
Appendix A: causes of diagnostic error.
Confirmation BiasPremature Closure
—-Failure to recognize significant —– Failure to order appropriate tests Signs and symptoms
—-Incomplete physical examination
—–tend to seek data consistent With initial diagnosis —-failure to gather additional data
—-Failure to consider other —-Atypical patient possibility presentation
—-Bias by experienced with cases –Failure to consider other possibilities
in the past
Lewis Blackman, who was a healthy 15 years old, came into South Carolina Hospital for an elective surgery to fix his pectusexcavatum. He died four
days after surgery without knowing the cause.
A series of error has happened including diagnostic error, medication error, transparency error, communication error, and negligence. However, the most important cause of Lewis’ death was diagnostic error. Background
Procedure was supposed to be low risk surgical procedure.
Procedure went longer than expected
Patient should be able to walk and urinate shortly after surgery. Pain has not been reduced after several injections of Ketorolac doses to relief pain. Assessment
Patient awoke and stated pain at 3 out of 10.
Patient had not urinated since post-operative.
Patient was given Ketorolac at 30mg/ 6hrs intravenously to relieve his pain. A black warning box of side effect was missed by nurses and residents. First day post-operative, Lewis had slight fever, high blood pressure, high heart rate level and abdominal pain. Lewis’s mother was told that it was constipation or gas in bowel. A common side effect of narcotic. The following days, Lewis’s pain was 5 out of 5. Abdomen was hard, firm and distended. His vital signs were unstable. He felt nauseous and had vomiting. Lewis looked paled, sunken eyes and unimaginable abdominal pain. No further assessment or study was done. The residents said that it was a probable ileus. Aide, nurses and residents could not detect Lewis’s blood pressure. Lewis was extremely fatigue and gasped right after blood sample was drawn. A code was run
Lewis died after an hour coding.
Using simple and transparent strategies and interaction with specific knowledge in memory-‘Thinking of the first thing that comes to mind’ ‘thinks of other possibilities’. This will help to prevent premature closure, availability bias, and representative bias. Gathering additional data, ordering the appropriate tests, reviewing patient history, and completing physical assessment prior to jump to final decision may prevent the occurrences of diagnostic error The residents and nurses have sufficient medical knowledge and that the errors reflect the inappropriate cognitive