Outcome switching

Outcome switching or outcome swapping is when authors of a clinical trial "move the goal posts" during a trial, which may be done to achieve the desired results or may accidentally affect the trial's findings. For example, changing the primary outcome that is used to decide if a treatment is successful from the distance walked in 6 minutes to the percentage of patients who rated their symptoms as "better" or "much better" on a patient questionnaire.

The PACE trial authors, in a clinical trial of ME/CFS patients undergoing GET/CBT, employed this tactic.

Pre-specified outcomes

When a clinical trial has a protocol published in advance, it will typically include pre-specified primary outcomes, these are the main measures used to determine if an intervention or drug is effective. Secondary outcomes are considered to be less important. For example, change in average pain level measured with a questionnaire, biological test results or physical measurements of ability.

Outcome swapping is the process of changing these primary outcomes during the trial or prior to publication. For example, a secondary outcome may be changed with a primary outcome, which may give then change the results reported in the trial publication.

In the PACE trial, outcome swapping led to results showing that both CBT and GET were moderately effective, but when a secondary analysis used the original pre-specified outcomes this showed the interventions were less effective than the previously published results.

Articles on Outcome switching in clinical trials

• How Outcome Switching is Corrupting Medical Research - Psychology Today
• For my next trick... - The Economist
• Is Outcome Switching Still A Problem In Clinical Trials? - clinicalleader.com
• How researchers dupe the public with a sneaky practice called "outcome switching" - Vox

Tracking outcome switching

• compare-trials.org

PACE trial

• In 2016, Julie Rehmeyer gave the talk Bad Statistics, Bad Reporting, Bad Impact on Patients: The Story of the PACE Trial Slideshow given at the 2016 Joint Statistics Meetings (Notes may be be viewed by clicking the “Notes” icon just above the comments section in the slideshow link.)

• David Tuller spoke about and answered questions on the PACE trial and its flaws.

• Feb 28, 2016, David Tuller gave a speech with Q&As.

• Feb 27, 2016, An Interview with David Tuller, held at The Forgotten Plague Conference, discussing the flaws of the Pace trial.

• PACE trial - Release of Data
• Geraghty (2017) discussed outcome switching and other PACE trial flaws.

SMILE trial

In the SMILE trial, school attendance was swapped from a primary outcome to a secondary outcome during the trial, with all primary outcomes then based only on questionnaire results, despite the unblinded treatment which encouraged children to believe they would recover and no longer see themselves as ill. The Archives of Disease in Childhood held an investigation but failed to retract the study, instead publishing an extensive editor's note.

Learn more

• Open data
• Open access
• Peer review

See also

• Alem Matthees won a Freedom of Information Act tribunal for the release of the PACE trial data.