Comparing Care Pathways Between COVID-19 Pandemic Waves Using Electronic Health Records: A Process Mining Case Study

Abstract

The COVID-19 pandemic caused rapid shifts in the workflow of many health services, but evidence of how this affected multidisciplinary care settings is limited. In this data study, we propose a process mining approach that utilises timestamped data from electronic health records to compare care provider patterns across pandemic waves. To investigate healthcare patterns during the pandemic, we collected routine events from Scottish hospital episodes in adults with COVID-19 status, generating treatment logs based on care provider input. Conformance checking metrics were used to select the Inductive Miner infrequent (IMi) algorithm for downstream analysis. Visual diagrams from the discovered Petri Nets indicated interactions on provider- and activity-level data subsets. Measures of "cross-log conformance checking" and graph edit distance (GED) further quantified variation in care complexity in adverse subgroups. Our baseline cohort included 1153 patients with COVID-19 linked to 55,212 relevant care provider events. At the conformance checking stage, the IMi model achieved good log fitness ( $\overline{\mathrm{LF}}$ $=0.95$ ) and generalisation ( $\overline{G}$ $=0.89$ ), but limited precision ( $\overline{\mathrm{PR}}$ $=0.27$ ) across all granularity levels. More structured care procedures were present in Wave 1, compared to limited multidisciplinary involvement in Wave 2. Care activities differed in patients with extended stay ( $GED=348,\overline{\mathrm{PR}}$ $=0.231$ vs $GED=197,\overline{\mathrm{PR}}$ $=0.429$ in shorter stays). We demonstrated that process mining can be incorporated to investigate differential complexity in patients with COVID-19 and derive fine-grained evidence on shifts in healthcare practice. Future process-driven studies could use clinical oversight to understand operational adherence and driving factors behind service changes during pressured periods.

Supplementary information: The online version contains supplementary material available at 10.1007/s41666-024-00181-6.

Keywords: COVID-19; Care pathways; Conformance checking; Electronic health records; Health services; Process mining.

Fig. 1

Partition of a Petri Net describing a process flow for a single patient case at an intermediate sequence $P_s$ and its follow-up markings at $P_{s+1}$. The enabled markings are highlighted by the respective tokens in each state (black dots) and the set of enabled transitions (black rectangles). The definitions of the coded activities are: Speech and Language Therapy – Eating, Drinking and Swallow assessment (SLT-EDS-A), physiotherapy assessment (PT-A), Physiotherapy – Mobility intervention (PT-MOBILITY-INT), Respiratory intervention (RESP-INT), Nutritional support – Enteral Tube Feeding (NUTR-ETF-INT), General Medical Doctor visit (GMD-INPUT)

Fig. 2

Conceptual flowchart describing the steps performed to transform the routinely collected data for process modelling and downstream analysis of the two COVID-19 cohorts

Fig. 3

BPMN flow diagrams discovered by the IMi model on the “provider-level event logs”. Inputs generated from the discovered Petri Nets using ProM

Fig. 4

BPMN flow diagrams discovered by the IMi model on the “activity-level event logs”. Inputs generated from the discovered Petri Nets using ProM

Fig. 5

BPMN flow diagrams discovered by the IMi model on the “activity-level event logs” in patients older than 75 years

Fig. 6

BPMN flow diagrams discovered by the IMi model on the “activity-level event logs” in patients with intensive therapy