Thread by @DrJenniferReed

For all those who ❤️for a “behind the scenes” #tweetorial on our BEST study: frontiersin.org/articles/10.33…
@CHeartsTrees @J_HarrisPT @BruceyMd @ProfJohnBuckley @lisacotie @DrAndrewPipe

The @HeartInstitute houses 1 of the largest #cardiac #rehabilitation programs in 🇨🇦. Even with our resources and infrastructure, we cannot accommodate cardiopulmonary exercise testing (CPET) to measure VO2peak for each and every patient before their 1st cardiac rehab class.

Even if we had sufficient personnel, training, equipment and resources, are expensive CPETs for the purpose of determining VO2peak and subsequent exercise prescription necessary?

The BEST study: We investigated the use of submaximal exercise testing as a potential alternative valid, safe and practical approach to assessing changes in VO2peak following a #cardiac #rehabilitation program.

We selected 3 different modes (i.e. treadmill 🚶‍♀️, cycle ergometer 🚲, step 📥) to be inclusive of what #cardiac #rehabilitation programs of all shapes and sizes might have at their disposal. Which would be the “BEST” ❓

I felt as though I was Jane Fonda searching for the perfect “step” when shopping for this piece of equipment.

It took us nearly 2 years to recruit N=50 to complete 1 CPET and 3 submaximal tests at baseline and following cardiac rehab (a total of 8 exercise tests). Our former Research Coordinator @CHeartsTrees worked tirelessly to encourage clinicians to refer patients to our study 🙏🙏

Treadmill: We selected a modified Bruce treadmill test. People were the most comfortable with this modality. Our biggest challenge was ensuring people did not use the handrails for assistance. A few did, which led to higher VO2peak values/spurious data points.

Cycle ergometer: We selected an Astrand-Ryhming cycle test. At the outset we thought this would be the “winner”, but we discovered that many patients could not cycle continuously for 6 minutes at an intensity to elicit exercise HRs within the range of 120–170 bpm. Our centre ⬇️

continued ... (the HR range required to predict VO2peak from the Astrand-Ryhming nomogram). Our mistake was not foreseeing this challenge given our exercise training experience and the known effects of B-blockade medications. An important lesson learned!

Step: We selected a Chester step test following our discussions with our collaborator and co-author @ProfJohnBuckley 🇨🇦 + 🇬🇧 collaboration.

Duke Activity Status Index (DASI): Given the known discrepancies b/w self-reported & device measured physical activity levels, I doubted the DASI would be the winner. But, in an effort to test a possible practical, low cost option, we included the DASI.

Most patients successfully completed the step tests (baseline: 68%; follow-up: 80%), while few patients successfully completed the cycle tests (baseline: 38%; follow-up: 46%).

Significant improvements from baseline to follow-up in VO2peak were observed with CPET (+3.6 mL/kg/min) and predicted by the DASI (+2.3 mL/kg/min), yet NONE of the submaximal exercise tests.

The Bland-Altman plots revealed proportional bias between the change in VO2peak values measured by CPET and those predicted by the treadmill test. What does this mean❓ There was greater disagreement between the CPET and treadmill tests with greater VO2peak values.

The cycle test was not a winner given how few patients successfully completed this test.

The BEST tests: The step test and DASI.

*⃣ Of note *⃣ : A modest increase of 1.75 to 3.5 mL/kg/min (i.e. 0.5 to 1 METs) following cardiac rehabilitation is considered a minimal clinically important difference (MCID).

Step test = appropriate for estimating group-level changes in VO2peak following cardiac rehabilitation as the mean bias was <1.75 mL/kg/min (the MCID).

Step test continued ... step test is not appropriate for estimating individual-level changes as only 38% of the changes in VO2peak values fell within ±1.75 mL/kg/min of the mean bias, & we observed wide limits of agreement (15, −12 mL/kg/min).

DASI = appropriate for estimating group-level changes in VO2peak following cardiac rehabilitation as the mean bias was <1.75 mL/kg/min (the MCID).

DASI continued ... the DASI is not appropriate for estimating individual-level changes as only 58% of the changes in VO2peak values fell within ±1.75 mL/kg/min of the mean bias, & we observed wide limits of agreement (13.8, −11.8 mL/kg/min),

Take home: In situations where CPET is not safe, practical or feasible, the Chester step tests and DASI appear to be a valid and safe submaximal tools for predicting mean, not individual-level changes in VO2peak following cardiac rehabilitation.

@TasukuTerada apologies for not initially tagging you! I blame my tiredness at 1 am when positing.

@CHeartsTrees @J_HarrisPT @BruceyMd @ProfJohnBuckley @lisacotie @DrAndrewPipe @threadreaderapp