Thread by @hmkyale

One of most important articles I’ve done… showing the noise in clinic BP measurement is large & makes it impossible to track Rx effects; almost useless in evaluating change from 2 clinic visits. Let me explain… ahajournals.org/doi/abs/10.116… @YaleMed @YaleCardiology @CircOutcomes

@YaleMed @YaleCardiology @CircOutcomes But first, kudos to the team @yuan_lu1 @shiwani_mahajan @rohan_khera @SpatzErica @GCLinderman #yuntianliu @jbmortazavi #chenxihuang - great team that persisted on this challenging project. And thanks @CircOutcomes @bnallamo for your support and reviews.

@YaleMed @YaleCardiology @CircOutcomes @yuan_lu1 @shiwani_mahajan @rohan_khera @SpatzErica @GCLinderman @jbmortazavi @bnallamo Now, we often see a patient with hypertension & change the meds and have them return to evaluate Rx effect. If everything goes right, drugs and changing doses may affect BP by ~5 mm Hg. Maybe a little less. We wondered how much natural variability there is between visits.

@YaleMed @YaleCardiology @CircOutcomes @yuan_lu1 @shiwani_mahajan @rohan_khera @SpatzErica @GCLinderman @jbmortazavi @bnallamo We studied 537,218 adults w/>7m measures @YNHH, and ~quarter had hypertension. We looked at pairs of consecutive visits & found that the difference was large; ±12 mmHg…and was consistent across subgroups, including treatment. For those w/htn, the standard deviation was ±14 mmHg.

@YaleMed @YaleCardiology @CircOutcomes @yuan_lu1 @shiwani_mahajan @rohan_khera @SpatzErica @GCLinderman @jbmortazavi @bnallamo @YNHH Why is this so imp? The noise in measurement between 2 visits is so large as to swamp the Rx signal… or the ability of the clinician to determine if there has been any ‘real’ change. In the real world, you can’t tell much from the follow-up visit. You need many more measures.

@YaleMed @YaleCardiology @CircOutcomes @yuan_lu1 @shiwani_mahajan @rohan_khera @SpatzErica @GCLinderman @jbmortazavi @bnallamo @YNHH Even for the diagnosis of hypertension, people are moving in and out of htn ranges just due to the noise in the system. And this makes sense: time of day, measurement method, device calibration, biological rhythms, stressors, etc. And the implications are profound.

@YaleMed @YaleCardiology @CircOutcomes @yuan_lu1 @shiwani_mahajan @rohan_khera @SpatzErica @GCLinderman @jbmortazavi @bnallamo @YNHH The real-world visit-to-visit variation is conspiring against rationale care… as we continually react to noise in the measurement. It would take many visits to be even 80% sure there has been a real change in response to a behavioral or medication change.

@YaleMed @YaleCardiology @CircOutcomes @yuan_lu1 @shiwani_mahajan @rohan_khera @SpatzErica @GCLinderman @jbmortazavi @bnallamo @YNHH The visit-to-visit variability in the real world builds on good work by others, primarily focusing on clinical trials. And even in those highly protocolized environments, there is substantial visit-to-visit variability. @MuntnerPaul

@YaleMed @YaleCardiology @CircOutcomes @yuan_lu1 @shiwani_mahajan @rohan_khera @SpatzErica @GCLinderman @jbmortazavi @bnallamo @YNHH @MuntnerPaul For better blood pressure care, we need more measures – and more continuous assessments. In a #digitalhealth age, there are many possibilities. But we do need to re-think our approach. @AHAScience @YaleMed #hypertension