Design vs Orchestration

When you design care flows, you define templates that describe all possible paths a care flow could take. It's only during orchestration that all the relevant data becomes known, enabling dynamic adaptation to a patient's specific circumstances.

Design Time Data

At design time, you have a comprehensive overview of patient data and diagnostics data that may be collected during the care flow. This includes but is not limited to:

• Patient Data: Demographics and administrative details such as name, age, gender, etc.
• Diagnostics Data: Potential responses to questionnaires and clinical scores to be computed.

In a depression screener care flow for example, you know that an intake questionnaire will be completed and a PHQ-9 score will be computed. However, you don't know the actual responses or the PHQ-9 score until a specific patient goes through the care flow.

Complex care flows can have conditional logic and dynamic content, making it difficult to track which data will be available. For instance, a care flow might use gender to determine whether to administer a PHQ-2 or a PHQ-9 assessment. At design time, you know that either score might be computed, but you won't know which until you know the patient's gender.

Orchestration Time Data

At orchestration time, you have complete and specific information about the patient, including:

• Patient Demographics and Administrative Data: Name, age, gender, etc.
• Responses to Questionnaires: Submitted by the patient and care team.
• Computed Clinical Scores: Based on the responses provided.

During orchestration, the care flow evaluates conditional logic and selects relevant paths, progressively refining the treatment plan. Essentially, the care flow starts as a plan with many possible paths (a tree with branches) and narrows down to a specific path as decisions are made.

Practical Implications

Understanding the distinction between design time and orchestration time data is crucial for effective care flow design.

Consider the following scenario for example:

A care flow designer reviews patient feedback and decides to change the depression screener to avoid requiring women with no signs of depression to fill out the lengthy PHQ-9 assessment. They plan to use the PHQ-2 score to determine if the PHQ-9 assessment is necessary. However, they miss that the PHQ-2 score is only computed for male patients, so their change doesn't impact female patients as intended.

By understanding the distinction between design time and orchestration time data, designers can avoid such mistakes, ensuring that changes to care flows are effective and improve the patient experience.