Choosing and Specifying Endpoints and Outcomes
Section 1
Introduction
For an explanatory trial, investigators can specify any outcome or endpoint, define the endpoint, and then measure it. The term outcome usually refers to the measured variable (e.g., peak volume of oxygen or PROMIS Fatigue score), whereas an endpoint refers to the analyzed parameter (e.g., change from baseline at 6 weeks in mean PROMIS Fatigue score). Even after a specific outcome is selected, it may be challenging to determine the best way to measure the effect of an intervention in terms of an analyzable endpoint, especially with pragmatic research where data are collected as part of routine care.
With pragmatic research, the endpoints and outcomes need to be available as part of routine care. Although the research question regarding the relative risks, benefits, and burdens of a specific intervention or activity will drive the selection of endpoints and outcomes, in a PCT, the selection must be balanced with an understanding of what is available in the electronic health record (EHR) or claims data and what additional resources will be needed to capture information not found in these sources.
Watch the video module: What do Endpoints and Outcomes Look Like in Pragmatic Clinical Trials
Some conditions can be objectively defined with a lab test, are straightforward to diagnose, and/or have International Classification of Diseases (ICD) codes. Some conditions, such as a broken leg, almost certainly require medical intervention and are likely to be captured in an EHR. Other conditions, however, are more ambiguous or less severe, and patients might not go to providers for treatment. Events or conditions that are not medically attended are unlikely to be captured in an EHR.
Defining endpoints and outcomes for some health phenomena is relatively easy for things like
- acute myocardial infarction
- broken bone
- hospitalization
However, many outcomes are not routinely recorded as part of healthcare delivery. For example:
- suicide attempt
- gout flare
- silent myocardial infarction
- early miscarriage
To detect these types of outcomes in pragmatic research, some additional work may be necessary, making the trial "less pragmatic"; however, this does not mean a pragmatic trial is not feasible. Key questions include:
- What challenges do you anticipate in trying to ascertain the endpoint?
- How might you address the challenges?
In this chapter, we will discuss endpoints and outcomes in pragmatic trials.
- Meaningful endpoints
- Outcomes measured via the electronic health record
- Outcomes measured via direct patient report
- Mobile health technology