Patterns for modeling process variability in a healthcare context

Abstract

Purpose Medical diagnosis and treatment processes exhibit a high degree of variability, as during the process execution, healthcare professionals can decide on additional steps, change the execution order or skip a task. Process models can help to document and to discuss such processes. However, depicting variability in graphical process models using standardized languages, such as Business Process Model and Notation (BPMN), can lead to large and complicated diagrams that medical staff who do not have formal training in modeling languages have difficulty understanding. This study proposes a pattern-based process visualization that medical doctors can understand without extensive training. The process descriptions using this pattern-based visualization can later be transformed into formal business process models in languages such as BPMN. Design/methodology/approach The authors derived patterns for expressing variability in healthcare processes from the literature and medical guidelines. Then, the authors evaluated and revised these patterns based on interviews with physicians in a Danish hospital. Findings A set of business process variability patterns was proposed to express situations with variability in hospital treatment and diagnosis processes. The interviewed medical doctors could translate the patterns into their daily work practice, and the patterns were used to model a hospital process. Practical implications When communicating with medical personnel, the patterns can be used as building blocks for documenting and discussing variable processes. Originality/value The patterns can reduce complexity in process visualization. This study provides the first validation of these patterns in a hospital.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 8 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.