the use of process mining in hospitals

UHB Use Case: Using Process Mining in Hospitals

Every transaction in an IT system leaves digital footprints – also in hospitals. Whether a doctor records a diagnosis, a nurse sends a request, or the laboratory determines the blood level – every activity is logged with a time stamp in the database.

Process Mining extracts these digital traces within the hospitals' IT systems to visually reconstruct the as-is-processes in real-time and detects deviations to the to-be processes.

Example: Surgery Punctuality

Initial position:

The surgery manager of the client knows that his hospital has problems with the punctuality of surgeries. On a regular basis, patients are not placed in the operations room on time, but he does not know from where the bottleneck is coming from.

Doctors hold the transportation service provider accountable, which, in return, blames the station, because they would forget regularly to prepare the patient in time. The surgery manager himself suspects that his surgery teams is responsible for the problem, as they would forget to call for patients in time.

Aim of the project:

The aim of the project is to analyze the severity of the perceived problem and to detect inefficiencies in the process with Process Mining.


To answer these questions the patient delivery process is visualized, starting from the call for the patient in the station, via the order acceptance of the transport service provider, through to the reaching of the lock. Therefore, the data was extracted from the KIS and the smartphone application of the transportation service provider.


  • The surgery team and the transport service provider are responsible for the delay in the processes.

  • Pattern in the call for patients: In the morning reliable, in the afternoon unreliable.

  • Pattern in the pick-up of the transportation service provider: Significant time difference on Wednesdays and Fridays compared to the rest of the days in the week.

  • Root cause: On Wednesdays and Fridays there is a higher workload per employee.

  • Consequences:

  • The relevant surgery teams were addressed.

  • Reorganization of the transport service provider by considering the workload per employee.

Example emergency room:

Initial position:

Emergency rooms have limited capacities and suffer from a rising, heterogeneous patient occurrence. Therefore, it is necessary to standardize the ‘emergency room process’.

Aim of the project:

Visualization of the as-is processes in Celonis to get a holistic and realistic picture of the operations in an emergency room to derive organizatory and technical measures.


Visualization of the processes from the arriving of the patient via triage, diagnostics, doctor’s visit, medical dismissal, etc. based on the time stamps recognized in the KIS database.


  • Cycle times, especially the ones until triage, are improvable

  • Quality of triage is good but could be improved as well

  • To-be-process is not always conforming


  • Discuss cycle times and process deviations with employees on a regular basis to demonstrate the importance of a standardized, efficient way of working

  • Retrain employees responsible for triage

  • Technical changes in the process of patient registration and classification

dr. daniel koch celonis uhb process mining in hospitals blog author
Dr. Daniel Koch
Process Analytics, uhb consulting AG
Dear visitor, you're using an outdated browser. Parts of this website will not work correctly. For a better experience, update or change your browser.