The first two posts in this series showcased how to build a basic workflow, or process as it is referenced in BPMN terminology, and how to expand upon an existing workflow. In this post we will take what we have covered and show how BPMN is used within RO Dynamics and how to get a workflow to be applied within RO Dynamics. If you missed our previous posts in this series they are linked at the end of this blog post.
There are several general Radiation Oncology workflows that automatically ship with RO Dynamics, one of which is the Treatment Planning workflow. All workflows and workflow changes are done within the Orchestration Module. Ideally there will be only a few department leads that will have access to this module due to the impact that any changes to a workflow can have on a department. Once in the Orchestration Module, all available workflows for your department will be located to the left in the Workflow list.
One of the nice features of RO Dynamics is that we store all versions of a workflow, that way you are able to migrate plans between different versions of the same workflow as well as see how many plans are currently running in each workflow version. Within RO Dynamics, all the tools that were used in the previous Shareable Clinical Pathways tutorials that were demonstrated using bpmn.io, can also be found in RO Dynamics.
In this tutorial, we are going to walk through the process of making a workflow compatible with RO Dynamics. The Treatment Planning workflow is divided into 2 lanes that consist of the Radiation Oncologist and the Planner roles. The workflow begins with a start event labeled “Start Planning” followed by a parallel gateway that will execute the ICD Dosimetry Eval workflow at the same time the treatment planning task will become available for those in the planner role. Once the Treatment Planning task is completed the MD Plan Review task is available for the Radiation Oncologist to then approve or reject the plan and give comments. If the MD Plan Review task is rejected, then the workflow reverts back to the Treatment Planning task along with the comments on what the Radiation Oncologist would like to have changed in the plan. If the MD Plan Review task is approved, then the workflow moves forward and the “Plan Approved” end event is completed.
Beneath the RO Dynamics workflow editor is where all task, activity, flow and other properties may be edited. Most workflow items need to have an ID and it is always considered best practice to give an ID name that describes the item. In our blog post, Advanced Care Team Tasks for Automation we covered all the task options that are available in RO Dynamics. The task type and assigned groups may be edited in the task properties.
In the Treatment Planning workflow, after the two tasks we hit an exclusive gateway. The direction of the workflow through an exclusive gateway is determined by previous data such as task decisions and external information systems. In this example, it is based on the answer to the MD Plan Review task. For all exclusive gateways, a default flow must be selected. Typically, this is the flow that most often the workflow will naturally go down unless there is a special case to handle. For the MD Plan Review task, the rejected flow is defaulted since you don’t want a plan to ever be marked as automatically approved without review. If the task is rejected, then the workflow is reverted back to the
Treatment Planning task for the Planner to make any necessary or requested changes and then submit it back to the Radiation Oncologist.
One of the advantages of BPMN is it visually documents your clinical workflows and is easy for everyone to understand and learn. At the same time, with more work the workflows can be automated to implement your clinical workflows and not just document them. This not only makes your departments workflows clearer but improves clinical efficiency.
Hannah Shappell is an ARRT certified radiation therapist with years of clinical experience who graduated with a Bachelor of Science in radiation therapy from Texas State University. She now currently resides in beautiful Central Oregon where you will typically find her outside trying to keep up with her two sons, crazy dog, and husband.
In the paper beneath, we present the need for shareable clinical pathways in healthcare. If you want to get involved, contact us and also join our mailing list!
Stay in the know about our new products and updates:
We will never send you spam and you can unsubscribe anytime.
Gabe strives to create innovative solutions to solve real-world problems effectively. Currently he’s focused on improving Radiation Oncology and healthcare workflow and interoperability, with previous experience in the defense industry, quantum computing, and healthcare. He holds a B.S. in Engineering Physics from Colorado School of Mines and an M.S. in Medical Physics from Oregon Health & Science University. He is also certified by the American Board of Radiology in Therapeutic Medical Physics. He resides in Bend, OR where he enjoys the beautiful outdoors hiking, climbing, and camping with his wife and kids.
Before starting our tutorials on creating shareable clinical pathways, hopefully you have already read our post on A Vision for the Future of Workflow in Healthcare. If not, you should check it out to set the context of the standards we will use in these tutorials such as Business Process Model and Notation, Decision Model and Notation, Case Management Model and Notation, and Fast Healthcare Interoperability Resources to create shareable clinical pathways. The models in these tutorials may be deployed and executed in our RO Dynamics platform, or in other commercial or opensource BPM tools. For more information visit our product page at https://quanteksystems.com.
In this series we will work with examples in the context of Radiation Oncology, which uses radiation to treat cancer. The goal is to help those involved in healthcare to understand how to read and create basic workflows using BPMN. Knowledge of radiation oncology is not required, but will make the workflows more interesting. We will start with simple workflows, and over time get more elaborate to handle the nuances of real-world workflows.
Often the best way to learn something new is by doing it. Perhaps the easiest way to get started is using the free web-based modeler at http://bpmn.io. To create a new BPMN model you may use this link: https://demo.bpmn.io/new. Alternately, you may download a free modeler such as the Camunda Modeler.
Let’s now dive in to creating shareable clinical pathways. As more tutorials are added, they will be accessible beneath for easy navigation.
Stay in the know about our new products and updates:
We will never send you spam and you can unsubscribe anytime.
Gabe strives to create innovative solutions to solve real-world problems effectively. Currently he’s focused on improving Radiation Oncology and healthcare workflow and interoperability, with previous experience in the defense industry, quantum computing, and healthcare. He holds a B.S. in Engineering Physics from Colorado School of Mines and an M.S. in Medical Physics from Oregon Health & Science University. He is also certified by the American Board of Radiology in Therapeutic Medical Physics. He resides in Bend, OR where he enjoys the beautiful outdoors hiking, climbing, and camping with his wife and kids.
In this post we are building off of our Shareable Clinical Pathways Tutorial Part 1- Basic BPMN Events, Tasks, and Gateways (https://quanteksystems.com/shareable-clinical-pathways-tutorial-part-1-basic-bpmn-events-tasks-gateways/) post linked here. We will be building off of the workflow that was created in that tutorial so it would be beneficial to review that before reading this post.
In our initial introduction to BPMN and building a simple workflow the ideas of events, tasks and gateways were introduced and a workflow depicting the Radiation Oncologist consult workflow was created. Events, tasks, and gateways are connected together using sequence flows or flows, which consist of a line with an arrow indicating the direction of the sequence. In this post we will build off the Radiation Oncologist Consult Workflow and incorporate additional staff roles with the use of lanes and also covering the editing options that are available.
We will begin with the workflow that we originally created as our starting point. Which is shown below. To create the workflow yourself follow the tutorial using a BPMN modeling tool such as at bpmn.io. (create link) If you need help, watch the video beneath where we show the entire model creation.
Step 1: Add additional lanes for nursing and simulation therapists.
Lanes add an additional level of readability and organization to a workflow. You are able to add lanes above, below or divide the current lane into divisions of two or three by selecting the original lane and using the tools to the right of the lane.
Tip: Tools to edit workflows (Add formatting for tips some how?)
Symbols in BPMN are able to be relocated by dragging and dropping in many cases. When tasks are added and removed it can create a disruption in the aesthetic of the workflow. Our modeler within RO Dynamics, as well as other BPMN modelers have a “Create/Remove Space” tool that can be utilized in these cases located in the tool menu on the left.
Tip: To insert between two existing symbols, it is often easiest to delete the sequence flow(s) in between and then branch new symbols off of the previous symbol.
Step 2: Relocate the start event and add a “Pre-Consult” task
Since the first task of the consult workflow will now belong to Nursing, that is why the start event is relocated to the nursing lane along with the additional nursing Pre-Consult with Patient task. The sequence flow now needs to be corrected with a flow going from the Pre-Consult to the Consult with Patient task in the “Radiation Oncologist” lane.
Step 3: Adding a “Schedule Simulation Appointment” task for the simulation therapist
Once the decision for the “Determine if Radiation Therapy Indicated” task is determined as “Yes” then the next symbol is a signal event that will start the radiation course workflow. After this signal event but before the patient consult ends, a “Schedule Simulation Appointment” needs to be added in the simulation therapist lane. Once this is completed, then the Patient Consult workflow is completed and ends with the “Patient Consult Ended” end event.
Now we have worked through how to expand upon an existing workflow. It is evident that the use of BPMN allows you to be flexible as your department needs change over the course of time.
Hannah Shappell is an ARRT certified radiation therapist with years of clinical experience who graduated with a Bachelor of Science in radiation therapy from Texas State University. She now currently resides in beautiful Central Oregon where you will typically find her outside trying to keep up with her two sons, crazy dog, and husband.
We will first begin with modeling shareable clinical pathways using the Business Process Model and Notation (BPMN) standard. If you want to follow along, please see the introductory post to this series for links to free modeling tools you can immediately use.
In this tutorial we will model a basic radiation therapy consult workflow from the perspective of the radiation oncologist. The final end result of the workflow is shown beneath.
Basic Radiation Therapy Consult Workflow in BPMN
To create the workflow yourself follow the tutorial using a BPMN modeling tool such as at bpmn.io. If you need help, watch the video beneath where we show the entire model creation.
Please note that in BPMN the term “process” is used, rather than workflow. We may use the terms interchangeably throughout this series, but in the BPMN terminology we use processes to implement the workflows.
One of the great aspects of BPMN is it makes workflows transparent, and easily understandable by even those without training in BPMN. If you have no experience in BPMN, you can probably understand the workflow in the diagram. Even better, is the workflow isn’t hard-coded and embedded in software code that is not available to the user, and would be difficult to understand even if it were available.
In this process we have a single lane denoting the radiation oncologist as the participant. We could have other participants with their own lanes, such as the patient, and other clinical staff such as nurses. For now we will stick with one participant.
Step 1: Create a lane and name it “Radiation Oncologist”.
This process starts with a New Patient Consult Started event. In BPMN events are depicted by a circle. In this case we have a start event, denoting the start of the process. In a process being executed by software, the start event could be triggered manually through a user interface, or automatically when a patient checks in for their consult. Once started, the workflow executes.
Step 2: In the lane create a start event and label it “New Patient Consult Started”.
Once the process starts the first step is a consult between the radiation oncologist and the patient. This is modeled as a user task in BPMN. Tasks of various kinds all have a rounded rectangle. The person icon indicates this is a user task, to be performed by a person. The process waits until this task is complete before proceeding. In software this could be completed manually from a task list, or it could be completed automatically when the patient checks out from the consult. These are implementation details beyond the current scope.
Step 3: Create a new user task, and label it “Consult with Patient”. Make sure the start event is connected to this task.
Once the consult task is complete, the radiation oncologist must determine if radiation therapy is indicated. In the present example we will assume that if radiation therapy is indicated the patient also decides to undergo radiation therapy. In a more realistic model we would also handle the decision from the patient. The decision if radiation therapy is indicated would be selected by the radiation oncologist in a user interface as a simple Yes/No choice.
Step 4: Create a new user task and label it “Determine if Radiation Therapy Indicated”. Make sure the consult task is connected to this new task.
Once the decision is made regarding if radiation therapy is indicated, there are two different flows. BPMN uses gateways with a diamond shape to direct flows within a process. Decisions are not made in gateways, but must happen before the gateway. In this example an exclusive gateway, also know as XOR is used, which only allows one path to be executed out of the gateway.
Step 5: Create a new XOR gateway, and connect the previous task to it. Set the label for the gateway to “Radiation Therapy Indicated?”.
In the current example if radiation therapy is not indicated, we end the consult and for the sake of this example, assume there is nothing further to do in this workflow. If radiation therapy is indicated, a Start Radiation Therapy event will be generated, using a BPMN signal event depicted by two circles with a solid triangle in the middle. Signal events broadcast to other processes to react to. In this case a separate process could be triggered from this signal to start a new radiation course workflow automatically. Finally, once the signal is generated, the process ends with an end event.
Step 6: Create an end event beneath the gateway and label it “Patient Consult Ended”. Connect the gateway to the event, and mark the flow as the default. Finally, label the flow “No”.
Step 7: Create a signal event to the right of the gateway and label it “Start Radiation Therapy Course”. Finally, create an end event and label it “Patient Consult Ended”.
Your workflow should now look like our diagram, and you have created a basic clinical workflow. To make it executable, one would normally add a few more details, such as the candidate group for the task (radiation oncologist), as well as create the implementation details for the start new radiation course signal.
In this tutorial you have been introduced to the following BPMN symbols: lanes, start/signal/end events, user tasks, and XOR gateways. To learn more, follow our blog and we also encourage you to sign up for our mailing list.
Stay in the know about our new products and updates:
We will never send you spam and you can unsubscribe anytime.
Gabe strives to create innovative solutions to solve real-world problems effectively. Currently he’s focused on improving Radiation Oncology and healthcare workflow and interoperability, with previous experience in the defense industry, quantum computing, and healthcare. He holds a B.S. in Engineering Physics from Colorado School of Mines and an M.S. in Medical Physics from Oregon Health & Science University. He is also certified by the American Board of Radiology in Therapeutic Medical Physics. He resides in Bend, OR where he enjoys the beautiful outdoors hiking, climbing, and camping with his wife and kids.
Radiation Therapy Treatment Planning Workflow in BPMN
Prior to focusing full-time on developing our digital whiteboard and workflow orchestration platform RO Dynamics, I worked as a clinical physicist in radiation oncology. I enjoyed working in radiation therapy as it allowed me to directly apply my background in physics and engineering. I was able to be part of a team helping to improve and save the lives of those who have been diagnosed with cancer in the region that I live. Cancer directly affects the lives of many of those around us, and indirectly affects many more. While there has been marked improvement in treatment, there is further to go.
Radiation therapy as a field has made significant progress since its first inception around 100 years ago. We’re still learning more every day about cancer, radiobiology, and better methods to treat cancer with radiation. The technological improvements in both the machines and software of the last 20 years has been amazing. However, one of the areas that does not receive as much attention is clinical workflow.
One of the challenges that I faced in the clinic, as do many others, is how to coordinate patient care as a team from consult to the end of treatment. Generally, this is not particularly challenging when the clinic has a single linear accelerator and doesn’t have a high volume of patients as one can keep quite a bit of information in their head or use personal methods to track what needs to be done. However, there are a variety of circumstances that can significantly compound the challenges in knowing what needs to be done, when it needs to be done, and knowing key information to get it done right. Some of the contributing circumstances include:
High patient volumes
Emergent treatments and changes to treatment plans
Patient specific information such as a pacemaker or previous radiation
Remote coverage across clinics within a healthcare system
Special Procedures (SRS, SBRT, IORT, LDR, HDR)
Projects to install, accept, commission, upgrade, and maintain equipment, machines, and software
Lack of easy to understand policies and procedures or adherence to them
Staff in and out of the department due to meetings, specialty procedures, vacation, or illness
Too many communication methods – whiteboards, excel sheets, sticky notes, email, text messages, health information systems
These circumstances can make it challenging for staff to be notified or discover important patient and treatment plan information such as:
Prescription changes
Scheduling changes
Plan Changes
Plans failing quality assurance
Cardiac implanted electronic devices
Previous radiation
Technical mistakes
Cancelled or partially delivered treatments
Necessary Insurance pre-authorizations
The unintended consequences of these challenges vary, but may include:
Delays or missed task completion
Down stream staff having to rush to complete tasks
Increased probability of making mistakes ranging from insignificant to decreased tumor control, or adverse clinical outcomes
Lack of useful data to analyze processes for improvement
Billing mistakes
In order to minimize these consequences and streamline department workflow and communication, clinicians need standardized, transparent workflows that can be customized for their departments needs and automated as much as possible by software. Even though policies and procedures to document workflow typically exist, they are often buried in a document somewhere, out of date, inconvenient to locate and reference, and often in a textual form that makes the overall high-level workflow hard to understand. The result is evolving workflows that are inconsistently executed and poorly understood. Even worse, these are mostly manually executed, and not automatable by software. Where automation does exist, it is done in proprietary workflow methods with limited functionality in oncology or health information systems. The latest configuration and historic changes of such mechanisms is typically not documented clearly in procedures and often out of date. Furthermore, being proprietary, these automation mechanisms are not transferable if a department switches software vendor, having to start from scratch to rebuild workflows.
Is there a better way to document and implement workflows without reinventing the wheel? I believe there is through existing standards used widely outside of healthcare. Look no further than the business world where standards have been widely adopted and implemented. Perhaps the most widely adopted standards in the business process world that can be applied to healthcare are Business Process Model and Notation (BPMN), and Decision Model & Notation (DMN). These standards provide standardized models and graphical symbols to define business processes that are readily understandable by all stakeholders, whether the end user, manager, or technical developer. Rather than processes being buried in textual descriptions, processes are transparently understood by visual diagrams. Even better, these models are potentially executable and automated by software in a vendor-neutral manner. These standards can be applied to many areas within healthcare.
In order to solve workflow challenges in Radiation Oncology, and broader healthcare, I believe we need to adopt existing relevant standards that enable customizable, shareable, vendor-neutral workflows and clinical pathways which is why our digital whiteboard platform, RO Dynamics, incorporates both BPMN and DMN. I am not alone in this view. In fact, the organization BPM+ Health (https://www.bpm-plus.org/) was created to make this a reality, and is composed by a community of clinicians, hospitals, universities, government agencies, and vendors working together with a common vision for the future of workflow in healthcare. Let’s work together to build a better future for clinicians, and ultimately patients.
To follow our journey at Quantek Systems to provide vendor neutral shareable workflows and clinical pathways with our software platform RO Dynamics, follow our blog. If you are interested in getting involved, please reach out to us!
Stay in the know about our new products and updates:
We will never send you spam and you can unsubscribe anytime.
Gabe strives to create innovative solutions to solve real-world problems effectively. Currently he’s focused on improving Radiation Oncology and healthcare workflow and interoperability, with previous experience in the defense industry, quantum computing, and healthcare. He holds a B.S. in Engineering Physics from Colorado School of Mines and an M.S. in Medical Physics from Oregon Health & Science University. He is also certified by the American Board of Radiology in Therapeutic Medical Physics. He resides in Bend, OR where he enjoys the beautiful outdoors hiking, climbing, and camping with his wife and kids.
