Detection Failures: Why Teams Miss the First Signal

If you don’t sense the signal, you can’t work the problem.

Recently, we had the chance to run a group of instructors through the DR5—MCTI’s framework for helping individuals and teams operate in complex, rapidly adaptive environments like cardiac arrests and trauma resuscitations. (You can find a medically focused version, the DR5-MED, here.)

The DR5 breaks down immersive decision-making into six components: Detect, Recognize, React, Respond, Reset, and Reflect.

There’s a lot to unpack across that framework. But in our work with these instructors, one question kept surfacing: Why do some individuals and teams fail to detect the crisis at all?

This post looks at that question through the lens of signal detection and analysis. We’ll consider three clusters of detection failures, two focused at the level of the individual, and one at the level of the team.

The Signal Never Lands

Sometimes, the system misses the problem entirely. Not because it misjudged it—but because the signal never made it into the system at all.

• The detector is off. Defibrillator pads disconnected from the machine won’t pick up a rhythm, and blood tests that get lost on the way to the lab can’t give us critical information. If we don’t have our fingers on the patient’s pulse, we can’t feel when it gets thready or stops

• The detector is misaligned. Myocardial infarction in the elderly can present as fatigue, not chest pain. If we only ask about chest pain, we’ll miss the signal. Sensors only work when they’re aligned with the actual presentation.

• The detector doesn’t know what to look for. A new clinician might miss a critical torsades pattern on an ECG, even when the waveform is right in front of them. The detector is on and looking in the right direction, but there’s a knowledge gap on what signals to flag.

  
  

The Signal Gets Lost in the Noise

Other times, the signal is there, but it’s buried in interference, obscured by chaos, or overshadowed by something louder. These are signal-to-noise ratio (SNR) problems.

• The signal is too faint for the detector to find. Identifying a STEMI or VTach on an ECG is (usually) straight forward, but finding the subtle shape changes on an ECG that reflect Brugada syndrome can be substantially harder and easy to miss without dedicated effort. Weak signals require experience and focus to detect.

• Background noise overwhelms the signal. As an extremely literal example, the ambient acoustic noise of a trauma bay can prevent a physician from hearing a heart murmur. If you don’t account for the critical aortic stenosis in your patient, your resuscitation strategy can fail wildly.

• Distortion or interference gets in the way. Metal implants and jewelry scatters x-ray beams, which can prevent teams from identifying life-threatening injuries. In patients with extremely low (or extremely high) blood pressure, distinguishing arterial from venous bleeding can be much harder than expected. In both cases, teams have to adjust for the distortion to find the real signal.

The Team Doesn’t Detect as a System

Successful team detection of a problem relies on the many parts of the teams being able to communicate, compare and contrast ideas, and synergistically combine signals into a shared mental model. If communication is disrupted, signals are never combined, or contrasting ideas are suppressed, crises can easily be missed.

• Communication gaps. If the triage nurse knows about a life-threatening allergy but the team in the back doesn’t, the system has failed. Information stuck in silos is functionally invisible.

• Fragmented pictures.The "blind men and the elephant" problem: multiple people have different parts of the picture, but no one connects the dots. The team never sees the whole, and frequently, if there’s no synthesis, detection fails.

• Suppressed signals. Junior team members may notice something others miss. But if they’re ignored or discouraged from speaking up, what they add to the ability of the team as a whole to detect never enters the decision loop.

---

If a team doesn’t identify that something’s wrong, nothing else in the DR5 chain has a chance to happen. Post-event reviews often jump to what was done, and in future posts, we’ll explore what happens after that signal is spotted, but before you look at the response, ask a simpler question: When did we know there was a problem? If that answer comes late—or not at all—start by looking at how your team runs detection.