Do Medical Errors Happen by Accident… or by Design?
Case # 5 - If a Wound Becomes Infected, Who Is to Blame?
OPINION
Medical errors like Case #5 highlight the need for health system redesign. This essay will examine Case #5 from a Safety Science perspective. These concepts are critical to appreciating the complexity of medical errors which will be addressed in the subsequent discussions.
Why Do We Blame Doctors for Systemic Failures?
When errors occur, the initial reaction is to blame someone. However, even single adverse events are due to the convergence of multiple factors. Because blaming an individual does not change the contributing factors, therefore, the same error will happen repeatedly.1 Improving patient safety requires altering the system conditions that lead to errors. Complex systems like healthcare can have coincidences that could not have been foreseen. As a result, they are reviewed after an error occurs. Hindsight bias occurs when things that were not obvious at the time of the event seem obvious in retrospect. Hindsight bias makes it easy to arrive at a simple solution, but difficult to determine what really went wrong.2 In Case #5, the lady with the laceration that became infected, it is easy to say the nurse should have thoroughly cleaned the wound of dirt and debris and the doctor should not have applied sutures to a wound 9 hours old.
When errors occur in most industries, a large volume of products are defective, or an otherwise healthy worker is injured. In healthcare, a third party (the patient) who may have complex medical issues suffers the injury. Also, preventable harm in healthcare occurs only to one patient at a time, not in large groups like an airplane crash. These factors make healthcare errors harder to discern.3
How Do Experts Approach Errors?
Expert analysis of the Three Mile Island nuclear incident and the Challenger space disaster contributed to advancements in safety science. James Reason defines an error as the failure of a planned sequence of mental or physical activities to achieve its intended outcome when the failure cannot be attributed to chance.4 Charles Perrow suggests that 60-80% of errors involve some human error.5 Edwards Demming believed that 94% or more of all problems, defective goods or services came from problems with the system, not from a careless worker or a defective machine.6 To determine human factors in errors, Reason suggested two types of errors: active errors and latent errors.7 Active errors occur at the level of the frontline worker (i.e. doctor or nurse} and have an immediate effect (the patient suffers a preventable adverse event).8 Latent errors are removed from the frontline (doctor, nurse and patient), but include less obvious things such as poor process design, incorrect installation of equipment, faulty maintenance, bad management decisions, and poorly structured organizations. An active error is the patient did not survive the doctor performing heart surgery; the latent error is the defibrillator used to restart the patient’s heart malfunctioned because of lack of maintenance.
Safe systems incorporate redundancies to manage errors. Highly reliable systems prioritize quality, have redundant personnel and safety measures, and foster a culture of continuous learning and adaptability.9 As we will see in subsequent essays, healthcare is almost completely the opposite of a safe system.
What If the Problem Isn’t the People, But the System?
Case #5 demonstrates how it’s easy to blame the doctor who sutured the wound 9 hours after the injury, an example of an active medical error. Let’s examine more subtle latent errors that might have contributed.
1. Hidden Challenges of Physician Scheduling
New doctors start their jobs right after Residency Training ends June 30th, often joining short-staffed hospitals during the busy 4th of July holiday weekend. To an administrator, every doctor with a license is the same. A nonclinical administrator making the schedule might not recognize the potential for problems that an inexperienced clinician unfamiliar with the hospital’s medical record system and emergency room staff might pose. A physician reviewing the schedule would know not to place a new doctor in a busy emergency room on the first day with less than veteran nursing staff. This was a latent error.
2. Risks of Scheduling Less Experienced Staff
Less experienced nurses may also be on duty to aid the new physician due to veteran staff taking vacations. Emergency room staff often work together as a team each knowing how to communicate effectively with each other. The triage nurse may have failed to note or failed to communicate to the charge nurse that the time was approaching the 6-hour window for suturing in Case #5. They should have easily recognized that the wound had not been cleaned at the previous facilities. An experienced charge nurse would have recognized the potential for errors and made sure the wound was cleaned and the patient brought back to treatment in time. They would also inform the treating physician of their concerns about the potential for infection. Less experienced staff might not function at such a high level, creating a latent error.
3. Results of Inadequate Training and Supervision
Emergency Department nursing staff receive training in wound cleaning. Standard procedure is 5 minutes to wash out a dirty wound with saline or even clean water to remove all debris. Nurses who work in the emergency room have all had some experience with suturing to know that after 6 hours the rate of wound infection skyrockets. The nurse that prepared the patient for the physician was not trained properly. Inadequate training and supervision is a latent error in Case #5 that contributed to the infection.
4. Poorly Designed Documentation and Support Systems
Electronic medical record systems (EHR) are mainly designed for charge capture rather than supporting patient care. Patient information is typically collected by nursing assistants who may have as little as 4 weeks’ training. The patient history often includes only the data needed to complete the billing form, known as a template in an EHR. Sometimes the documentation is simply copied from what the patient said at the last visit to create higher billing. The EHR created a latent error of incomplete information.
The treating physician may not have realized the clinical information was collected by someone with almost no medical training. If the physician in Case #5 relied on staff in a hectic setting, they would have sutured the wound without realizing it was 9-hours old, another latent error.
Finally, the physician was exposed to the hospital’s electronic health record for only a few days before the adverse patient event occurred. This would be the equivalent of asking a Cessna pilot to operate a Boeing 737 with 4 days training – yet another latent error.
5. Inadequate Personnel Management
Administrative staff often see Emergency Department personnel needs as constant year-round and set a uniform budget accordingly. Anyone who has worked in an ER has observed increased visits during the summer months, attributed to higher rates of outdoor activities and social interactions. Longer daylight hours and warmer weather also contribute to a higher incidence of injuries.
Clinicians know early spring is heart attack time because people who were inactive all winter overexert themselves as the weather improves. Fourth of July is probably the busiest of all because everyone is off work. The 4-hour wait for laceration treatment in Case #5 indicates inadequate personnel were assigned to handle the workload. Seeing patients in a timely fashion is not a priority because the customer of the hospital is the insurance company not the patient as described in Why Do You Have to Wait So Long for Care? Inadequate staffing is one more latent error because the patient should not have waited the additional 4 hours with a deep laceration that had not been cleaned.
ANECDOTE
When I oversaw a busy emergency department, I worked with the same 30 nurses every day. I knew each staff member’s strengths and weaknesses. Two nurses I taught to repair simple lacerations. Some I taught to elicit and record detailed patient histories. I trained others to clean and apply dressings. Some had limited skills, so they brought patients back to the treatment area and prepared equipment. I asked radiology technicians to help me read X-rays. I encouraged everyone to help and alert me to potential errors. Not all doctors have that attitude and may take suggestions as criticism. We worked efficiently as a team, minimizing patient wait time. I find the scenario of Case #5 hard to imagine in a modern facility but confirms my fears of patient care operated for profits instead of outcomes.
OPINION
Healthcare is a complex field that requires knowledge and careful consideration to comprehend how errors occur and devise strategies to prevent them. Healthcare systems need better safety measures. Blaming doctors alone is insufficient. Improving healthcare safety requires more transparency and access to de-identified data for research. Foundation models (AI) offer an opportunity that should be embraced, but a cultural shift to allow this type of evaluation will be needed.
Preview
The next essay describes the resolution of Case #5. Follow up essays will discusses malpractice litigation and its effect on quality issues. We will attempt to explain why our understanding of medical errors can often be difficult.
Kohn, L., Corrigan, J., Donaldson, M., Editors, To Err Is Human: Building a Safer Health System, Institute of Medicine, 1999, National Academy Press, Washington, DC, pg. 49
Idem pg. 53
Idem pg. 53
Reason, James, Human Error, Cambridge: Cambridge University Press, 1990
Perrow, Charles, Normal Accidents, New York: Basic Books, 1984
Reason, 1990
Cook, Richard, Woods, David, Human Error in Medicine, ed. Marilyn Sue Bogner, Hillsdale, NJ, Lawrence Erlbaum Associates, 1994
Sagan, Scott, The Limits of Safety, Princeton, NJ: Princeton University Press, 1993