Interval type-2 fuzzy logic for encoding clinical practice guidelines

In the last years, the advent of Decision Support Systems (DSSs) embedding Clinical Practice Guidelines (CPGs) has created the premise for improving quality of care and patient safety. However, CPGs, typically encoded in the form of if-then rules, are still not completely suitable for computer implementation, due to different kinds of uncertainty affecting them. In order to face this issue, this paper proposes a novel approach for automatically encoding CPGs by means of if-then rules based on interval type-2 fuzzy sets, with the final aim of dealing with two different kinds of uncertainty, namely intra-guideline uncertainty and inter-guideline uncertainty. The approach is structured into four sequential steps: (i) the encoding of multiple and different CPGs concerning a same problem as if-then rules built on the top of crisp sets; (ii) the mapping of these crisp sets first into possibility distributions and, then, into type-1 fuzzy sets; (iii) the construction of final interval type 2 fuzzy sets; and (iv) the specification of fuzzy rules on the top of the interval type 2 fuzzy sets produced. As a proof of concept, the approach is employed to deal with some CPGs pertaining the hypertension treatment, showing its feasibility and also suggesting that its application could simply and proficiently aid the embedding of CPGs into clinical DSSs.