Signal Detection Theory in Healthcare


Diagnosing Cancer of the Lung

One of the most obvious applications of signal detection theory in healthcare is when it is applied to diagnosing disease. In a talk given by Dr. Cosmic during my ENP-109, Medical Technology Development with Michael Wiklund, Cosmic described a PET scan procedure that was used to find cancer in patients.
If the cancer was suspected with the PET scan, a biopsy would be performed. This was not a novel experience for the patient. These biospsies were quite invasive, requiring the patient to undergo general anesthesia. I asked Dr. Cosmic about this, how often false-positives were perceived by the PET scan, and why that was. The answer was surprising: see the video to learn more.

Nuses as Signal Detection Machines

An interesting paper was published in JAN (Jounral of Advanced Nursing) proposing a theoretical framework for detection of patient risk by nurses. The ability for nurses to detect risk in their patients is one of their most important roles. As the papaer states, the patient risk detection theory synthesizes concepts of signal detection theory and high reliability theory. Signal detection theory explains the decision-making processes of nurses as they scan for signals of potential patient harm. High reliability theory explains how nurses' signal detection capacities are facilitated when healthcare settings operate as high reliability organizations making patient safety the top priority.
It concluded that the patient risk detection theory facilitates understanding of both individual and organizational factors that influence nurses' ability to detect risk in complex healthcare settings. It can be used to guide research on interventions to enhance signal detection by nurses and increase patient safety in today's complex care environments. The theory can also be used to guide design of training programmes that permit nurses to develop practical skills in signal detection.