Exploring the feasibility of an artificial intelligence based clinical decision support system for cutaneous melanoma detection in primary care – a mixed method study

Figures & data

Figure 1. Flowchart of near-live simulations with think aloud, semi structured interviews and qualitative analysis.

Table 1. Themes and subthemes generated from the qualitative analysis of the near-live simulations and individual interviews.

Theme	Sub-theme
Trust	The importance of trust
	Factors governing trust
Usability and User Experience	The Working Process
	Easiness to Use
	Integration with Electronic Health Record System
	Improvements and Additional Features
	Instructions and Education
Clinical Context	Responsibility and Relation to Guidelines
	Value for Clinical Practice
	Comparison With Ordinary Workflow

Table 2. Results from the PCPs reader study (diagnostic assessment on the basis of image interpretation) without and with the use of AI. n = number of clinical assessments.

	Sensitivity (%) [95% CI]	Specificity (%) [95% CI]	PPV (%) [95% CI]	NPV (%) [95% CI]	Accuracy (%) [95% CI]
PCP + no AI (n = 189)	54.8 [38.7 − 70.2]	41.5 [33.4 − 49.9]	21.1 [13.9 − 30]	76.3 [65.4 − 85]	44.4 [37.2 − 51.8]
PCP + AI  (n = 200)	91.1 [78.8 − 97.5]	48.4 [40.3 − 56.5]	33.9 [25.5 − 43.1]	94.9 [87.5 − 98.6]	58.0 [50.8 − 64.9]