Improving medical protocols by formal methods

Abstract

Objectives During the last decade, evidence-based medicine has given rise to an increasing number of medical practice guidelines and protocols. However, the work done on developing and distributing protocols outweighs the efforts on guaranteeing their quality. Indeed, anomalies like ambiguity and incompleteness are frequent in medical protocols. Recent efforts have tried to address the problem of protocol improvement, but they are not sufficient since they rely on informal processes and notations. Our objective is to improve the quality of medical protocols. Results Our main results are: a consolidated formal language to model medical practice protocols; two protocols, each both modelled and formalised; a list of properties that medical protocols should satisfy; verification proofs for these protocols and properties; and perspectives of the potentials of this approach. Our results have been evaluated by a panel of medical experts, who judged that the problems we detected in the protocols with the help of formal methods were serious and should be avoided. ConclusionsWe have succeeded in demonstrating the feasibility of formal methods for improving medical protocols.

Using preference learning for detecting inconsistencies in clinical practice guidelines: Methods and application to antibiotherapy

“…The consistency of recommendations can be verified by formal methods [13]. The recommendations are first represented using an explicit and non-ambiguous model in a formal language.…”

“…They allow detecting ambiguity, incompleteness, inconsistency or redundancy within CPGs [6,15,13,3,16,17]. For example, some authors [18,19] state that, if narrative guidelines are encoded into logical language ("if... then..." rules), then the generation of all possible variable combinations allows the detection of incompleteness (i.e.…”

“…drug properties such as contraindications) and the identification of the medical principles underlying the recommendations of CPGs. However, formalizing the knowledge and the reasoning principles is a complex task [13]. For example, in oncology, a medical domain where multiple drugs are often prescribed, the adverse events can be limited by checking the known adverse effects [20].…”

A Hybrid Approach to the Verification of Computer Interpretable Guidelines

“…As discussed in G.1 many approaches have focused on the development of guideline representation formalisms, and/or systems to acquire, store and execute CIGs. However, the effort in defining and disseminating CIGs has not always been coupled by a parallel effort in guaranteeing their "quality" [2]: despite the fact that CPGs and/or CIGs are issued by recognized experts' committees, they might be ambiguous or incomplete [3], or even inconsistent. The need for guideline quality verification is thus clearly emerging.…”

Engineering Medical Processes to Improve Their Safety

“…For example, the Protocure II project [15] has goals that are quite similar to ours, but uses a rather different, AI-based, linguistic paradigm for defining processes. Noumeir has also pursued similar goals, but using a notation like UML to define processes [16].…”