Process Mining for Healthcare | QPR

QPR Blog

You are here

25 October 2017

Process Mining for Healthcare

The roll-out of information systems has been especially prevalent in the field of healthcare over the past decade. As patients go to through the healthcare process, their status is logged in different healthcare systems, and the patient process is constantly updated to the system. Even with the amount of data being produced, the healthcare domain is trying to find the answer to meet quality of care metrics. However, not many realize that the answer for improving their processes can be found in their data. With the aid of process mining and QPR ProcessAnalyzer, the dormant data produced by these information systems can be used for improving the healthcare experience and bringing the quality of care up to the expected level of service.

Using QPR ProcessAnalyzer, companies can use their own data to find the bottlenecks that occur in their key business processes. QPR ProcessAnalyzer can be used for any process that produces data. As a lot of companies have introduced different information systems that log users’ process actions, the amount of data about processes has increased. QPR ProcessAnalyzer turns this data into knowledge that can be used to improve your operations, creating more efficient results.

An excellent application area in the healthcare domain is the patient journey process. A patient journey is the complete treatment process that begins from the first visitation to a General Practitioner to seek consultation, and ends when the patient has had the necessary treatment performed and the case is considered closed. As many patients are treated in specialized units, it can be considered a best practice to prepare a separate process model for each care unit. This approach gives you more accurate and precise analysis results, as you can use as much specialized data as you want to relate to the unit you are analyzing.

Process visualization: A patient journey can take many different steps

 

Our example model represents a head trauma patient journey in a large hospital. Being a specialized field with a lot of different paths a patient journey could branch to, it represents an ideal process mining target. The process also contains many factors that could cause the process to not meet the quality of care that is expected. The basic data needed for the analysis is simple, as you only need three data items:

 - A unique identifier that represents a single patient case

 - A description of the process step taken

 - A timestamp for the process step

The basic data gives us a good outline of the kind of flows a patient normally goes through when being treated. However, as in many other process mining areas, the data also reveals a lot of different variations and loops. This gives you more insight about how the process actually unfolds in the real world and if your quality of care levels is achieved. By looking at the all possible flows, we can quickly realize how complicated and multi-faceted our can be. Here we can see the different revisits and process variations a patient might face when they receive treatment.

A typical QPR ProcessAnalyzer model also features different case attributes for further analysis. These case attributes are additional information such as patient diagnosis. For example, we can easily profile certain types of diagnoses and create a filter excluding other treatment events. Thus, we can easily see the patient journey that occurs to only patients with our selected diagnosis. If we see a deviation from the quality of care target metric, we can easily drill-down and see which case attributes are the root causes for the deviation in our process. These findings provide important information for improving the patient journey, as you can easily see what causes a problem in your process and give you further information on its root cause.

Screenshot: QPR ProcessAnalyzer enables deeper analysis to find out root causes in process deviations

 

What benefits can be gained with QPR ProcessAnalyzer?

Loading the data and performing analysis using QPR ProcessAnalyzer can provide you important information about your patient journey process.

Here are some quick wins you are able to achieve with QPR ProcessAnalyzer:

 - Patient journey process can be instantly visualized. Using existing event data, you can easily see which steps the patient journey takes and if there are any differing paths from the ideal process. This gives you pointers on how to improve the patient treatment process to meet set quality metrics

 - Root causes for detrimental delays in the processes can be discovered. Found a bottleneck in your process? With QPR ProcessAnalyzer, you can drill-down to the details and find out the common attributes for your process deviation. This can be used to improve the patient treatment efficiency.

 - Benchmarking the process to see pain points for different types of cases. As patient diagnosis can vary, you can see which process steps each different diagnosis go through. This enables you to optimize the performance for certain cases as you can see the problems for specific cases and bring up the process to the set quality of care metrics.

These quick wins give you pointers to rally out instant improvements and it gets even better with continuous use. By using up-to-date data, you can easily follow-up on these improvements and see if there are further needs for improvement. Continuous monitoring provides your organization with important analytical data that can be used to follow-up and maintain your improvement efforts. Process mining enabled with QPR ProcessAnalyzer can improve the patient journey to adhere to the quality of care principle by enabling you to see your process as it happens in the information systems.

Would you like to know more? For a further look on how QPR ProcessAnalyzer can be used in the healthcare industry, check out our webinar below.

Watch on-demand

If you would like a more personal presentation,  book a demo with us to hear more about our approach to process mining in healthcare.

Riku Mikkonen
Product Marketing Manager