In the last two posts, I’ve examined the phenomenon of “delayed diagnosis” or missed injury. I believe that there are only two fundamental reasons why this occurs:

• Insufficient diagnostic technique – A good physical exam and/or specific diagnostic techniques were not performed. Or rarely, the injury cannot be readily detected by existing techniques and technology. The former is usually the real problem and may be an issue with either the physical exam completeness and/or technique, or judgment used to obtain the appropriate diagnostic test. Example 1: a penetrating injury to the back is missed because the patient is not logrolled to examine this area. Example 2: a spine fracture is missed in an elderly patient with a fall from standing because the back pain found on physical exam is evaluated only with conventional imaging of the spine, not CT.
  
• Failure to recognize the injury – The injury was actually identified on a test but was not appreciated by the clinician. Example 1: the radiologist may not have appreciated and reported out a subtle anomaly in the cervical spine imaging. Example 2: you fail to check your patient’s lab tests and miss a sudden spike in serum amylase or lipase the day after your patient was kicked in the epigastrium by a horse.

So what can you do to avoid this potential problem? Here are some tips:

• Admit that it can really happen to you. If the missed injury rate at your center is off the low end of the bell curve (< 5%), then you are either really good or really blind. You’d better take a close look at your performance improvement process because you may be fooling yourself.
• Adopt a firm definition of “delayed diagnosis.” Basically, you need a time frame after which a new diagnosis is considered “delayed.” It should be a reasonable time interval after the patient has left the ED. If it’s too short an interval (e.g. once they leave the ED), your number will be unnecessarily high. If it’s too long (days and days later), then significant morbidity may occur that you don’t account for. Most centers have adopted 24, 36, or 48 hours after patient arrival.
• Implement a tertiary survey process. This is a complete physical re-examination followed by a review of all diagnostic studies (lab and radiology) that have been performed. This exam needs to be dated and timed to ensure that it is performed within the time frame noted above. If a new finding is discovered on the tertiary survey, it is not considered a delayed diagnosis. If found after the survey (or after the pre-determined time interval), it is and must be entered into your performance improvement process.
• Be paranoid. I hate the phrase, “maintain a high index of suspicion” because it’s meaningless. It’s like those stupid “start seeing motorcycles” bumper stickers. You can’t see what you can’t see. But you can be suspicious all the time, constantly looking for the inevitable clinical surprises of trauma care.

In my last post, I wrote about the usual reasons for delayed diagnosis: insufficient diagnostic technique or insufficient recognition. What about the time of day? An interesting paper looked at the correlation between admission time and the rate of missed injuries.

The work was done at a large teaching hospital and Level I trauma center in Australia. A large number of patients were reviewed over an 11-year period. The study was complicated slightly by implementing a dedicated trauma unit in the middle of the study period, but the authors stratified their groups to account for this.

Results were stratified by time of admission: office hours, after-hours, and weekends. Missed injuries were defined as those found after the completion of the primary and secondary surveys. The overall statistical treatment appeared to be robust.

Here are the factoids:

• A huge number of patients (53,000) were reviewed. This is a busy place!
• There were 2519 missed injuries in 1262 patients (2.4%) [low!]
• Missed injuries occurred during office hour admissions in 2.2%, after-hours in 2.6%, and on weekends 2.5% of the time
• The increased incidence of delayed diagnosis in after-hours admits was marginally significant (p = 0.048)
• Missed injuries appeared to have increased over time and were 1.34 times more likely at the end of the study period vs the beginning
• Thoracic spine and abdominal injuries were most commonly missed

Bottom line: Hmm, time of day was not on my list of reasons for missing diagnoses. What gives? If you read the article closely, the trauma service at this hospital was staffed with a higher number of trainees after hours and on weekends than during office hours. It was also noted that incomplete physical examination was thought to be a factor in many of the delays. Most likely, both of my listed reasons were in play here. Inexperienced clinicians and insufficient examination are both major factors. And what about the increase in missed injuries over time? Midway through the study, the hospital implemented a dedicated trauma unit, and a tertiary exam became routine. This identified more injuries after the primary and secondary surveys were complete. 

In the next post, I’ll review strategies to decrease the incidence of missed injury.

Reference: Office hours vs after-hours: do presentation times affect the rate of missed injuries in trauma patients? Injury 2015, in press.