Delayed or missed diagnoses happen. It’s a reflection on the state of technology and our own diagnostic acumen. Unfortunately, a few cases of delayed diagnosis result in morbidity, potential lawsuits, and rarely, death.
How often does delayed diagnosis occur? A few spot check type articles were published about 15 years ago, but little has been done to slice and dice the data. And as usual, the old data ranged widely in its assessment of the incidence of this problem (1-18% !). However, I managed to find a (somewhat) more recent one that gives a little clearer picture of this issue.
A single pediatric hospital in Indiana reported its experience from 1997 to 2006. This interval included the time that it was verified as a Level II Trauma Center (2000 onwards). They included children 0-14 who had sustained “major trauma.” This was defined as multiple system injuries, high-energy impacts, and gunshots. In this study, delayed diagnosis was defined as one found after a stable patient was admitted to their room. In patients taken directly to OR, it was one found after the patient left the recovery room.
Here are the factoids:
1100 patients met study criteria. 98% were blunt trauma.
Only 44 patients had delayed diagnoses of 47 injuries
Average time to diagnosis was 4 days (range 8 hours to 28 days)
34% of diagnoses were made within 24 hours
3 diagnoses were made at a followup visit, all for upper extremity/should fractures
80% of delayed diagnoses required a change in therapy, most commonly a sling or cast. 15% required surgery.
The long-term delayed diagnosis rate was 4%
Bottom line: Delayed diagnosis remains an issue in patient of all ages. The reported 4% rate subjectively seems about right to me. The most important lesson from this study is the extremely high percentage of delayed diagnoses that required further therapy. This is why it is so important to implement a specific system (the tertiary survey) to seek out these diagnoses.
A tertiary survey is a repeat head-to-toe physical exam and a review of all radiographic imaging performed to date. The trauma center should define the time interval from admission, and I recommend no more than 24-48 hours. We do not count any diagnoses found during this exam as being delayed. However, if a tertiary exam was not performed, or injuries are found after it was completed, we do consider it delayed an run it through our performance improvement process.
A set of guidelines for management of blunt solid organ injury in children developed by the American Pediatric Surgical Association was originally published in 1999. One of the elements of the guideline was to place the child on bedrest for a period of time after the injury. Arbitrarily, this period was defined as the injury grade plus one day. So for a grade 3 spleen injury, the child would have to stay in bed for 4 days (!).
Recent work looked at the impact of shortening this time interval. Over a 6 year period, all pediatric liver and spleen injuries from blunt trauma were identified and an abbreviated bedrest protocol was implemented. For low grade injuries (grade 1-2), children were kept in bed for 1 day, and for higher grade injuries this was extended to 2 days.
Here are the factoids:
249 patients were enrolled (about 40 per year) with an average age of 10. “Bedrest was applicable for 199 patients, 80%.” Huh? Does that mean that 50 patients were excluded due to surgeon preference?
The organ injured was about 50:50 for spleen vs liver. Twelve children injured both.
Mean injury grade was 2.7, which is fairly high
Mean bedrest was 1.6 days, and mean hospital stay was 2.5
Bedrest was the limiting factor for hospital stay in 62% of cases
There were no delayed complications of the injury
Bottom line: Come on! Most centers don’t keep adult patients at bedrest this long, and we learned about solid organ injury management from kids! Children almost never fail nonop management, so why treat them more restrictively than adults? And have you ever tried to keep a child at bedrest? Impossible! This study is too underpowered to give real statistically valid results, but it certainly paints a good picture of what works. We’ve been keeping both adults and children at bedrest only overnight, and our average length of stay for isolated solid organs is about 1.5 days. But really, who says that staying in bed for any period of time avoids complications? There are lots of other evil things that can happen!
The July Trauma MedEd Newsletter will be released to subscribers this weekend. I’ll be continuing to cover that topic no one wants to think about but everyone wants to know more about: Malpractice and trauma professionals. This issue is Part 2.
Does family presence in resuscitation affect malpractice?
Does admitting an error increase malpractice risk?
Once again, the trauma world is a little smaller. Yesterday, another great trauma professional passed away, Dr. Norman McSwain. I have known the man for decades, and literally grew up reading about his advancements and accomplishments. It’s interesting that one never truly appreciates the magnitude of those achievements until the person is gone.
Norm was a skilled surgeon and teacher, but his achievements were felt far outside his home in Louisiana. He was an early member of the ACS Committee on Trauma, and was very involved in the development of the Advanced Trauma Life Support and Prehospital Trauma Life Support courses. He is credited with developing the original EMS programs in both Kansas, where he took his first faculty position out or residency, and in New Orleans, his home for the remainder of his life. He spent his career at the Charity Hospital there, weathering multiple political storms over the years, as well as the big one, Hurricane Katrina. He was instrumental in achieving Level I Trauma Center status for its replacement, Interim LSU Hospital.
Norm’s accomplishments are, as many of his contemporaries who have left us, too numerous to count. I certainly won’t try to recount them here. But it was his charm, his love for his charges, and his willingness to teach every trauma professional that will always be remembered.
I’ll leave you with his 18 rules of patient care. They are timeless, and will serve you well regardless of your degree and level of medical training.
Most trauma professionals worry about over-prescribing pain medication. But under-prescribing can create problems as well. Uncontrolled pain is a huge patient dissatisfier, and can lead to unwelcome complications as well (think pneumonia after rib fractures). Always do the math and make sure you are sending the right drug in the right amount home with your patient. If the patient’s needs are outside the usual range, work with their primary provider or a pain clinic to help optimize their care.
#8. Improper care during an emergency
This situation can occur in the emergency department when the emergency physician calls a specialist to assist with management. If the specialist insists on the emergency physician providing care because they do not want to come to the hospital, the specialist opens themselves up to major problems if any actual or perceived problem occurs afterwards. The emergency physician should be sure to convey their concerns very clearly, tell the specialist that the conversation will be documented carefully, and then do so. Specialists, make sure you understand the emergency physician’s concerns and clearly explain why you think you don’t need to see the patient in person. And if there is any doubt, always go see the patient.
#9. Failure to get informed consent
In emergency situations, this is generally not an issue. Attempts should be made to communicate with the patient or their surrogate to explain what needs to happen. However, life or limb saving procedures must not be delayed if informed consent cannot be obtained. Be sure to fill out a consent as soon as practical, and document any attempts that were made to obtain it. In urgent or elective situations, always discuss the procedure completely, and provide realistic information on expected outcomes and possible complications. Make sure all is documented well on the consent or in the EHR. And realize that if you utilize your surrogates to get the consent (midlevel providers, residents), you are increasing the likelihood that some of the information has not been conveyed as you would like.
#10. Letting noncompliant patients take charge
Some patients are noncompliant by nature, some are noncompliant because they are not competent (intoxicated, head injured). You must use your judgment to discern the difference between the two. Always try to act in the best interest of your patient. Document your decisions thoroughly, and don’t hesitate to involve your legal / psych / social work teams.