Premium
Reproducibility of peripheral arterial tonometry measurements in male cardiovascular patients
Author(s) -
Nil Markus,
Schäfer Daniela,
Radtke Thomas,
Saner Hugo,
Wilhelm Matthias,
Eser Prisca
Publication year - 2014
Publication title -
european journal of clinical investigation
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.164
H-Index - 107
eISSN - 1365-2362
pISSN - 0014-2972
DOI - 10.1111/eci.12341
Subject(s) - reproducibility , intraclass correlation , medicine , cardiology , coefficient of variation , coronary artery disease , framingham heart study , peripheral , population , framingham risk score , mathematics , disease , statistics , environmental health
Background Assessment of endothelial function of the microvasculature by peripheral arterial tonometry (Endo PAT ® ) has gained increasing popularity in patients with cardiovascular risk factors. Only limited knowledge about its reproducibility in patients with coronary artery disease ( CAD ) is available. We therefore aimed to quantify reproducibility of Endo PAT ® parameters in patients with stable CAD . Design Endo PAT ® measurements were performed repeatedly in 78 male patients (age 66 ± 8 years) with CAD on stable medication. We calculated overall mean, standard deviation ( SD ), coefficient of variation ( CV ) and intraclass correlation coefficient ( ICC ) of the following parameters: reactive hyperemic index ( RHI ), PAT ratio of the postocclusion period 90–150 s as used for calculation of the RHI ( PAT ratio 90–150 s ) and 90–120 s ( PAT ratio 90–120 s ) as used for the often employed Framingham RHI (F‐ RHI ), as well as PAT ratio of the peak hyperemic response ( PAT ratio peak response ). Additionally, least significant changes ( LSC ) for individual subjects and minimum sample sizes for parallel and cross‐over design studies were calculated. Results Mean RHI was 1·84 ( SD 0·36). For RHI , PAT ratio 90–150 s , PAT ratio 90–120 s , and PAT ratio peak response the CV s were 17·0%, 25·4%, 26·1%, and 25·0%, respectively. The ICC s were 0·45, 0·49, 0·48 and 0·51, respectively, and LSC for RHI was 47·2%. Conclusions CV of RHI in our population was moderate; however, we consider this precision insufficient to monitor changes in individual patients, as they would need to exceed 47% to show a significant change. Further, the poor ICC s reflect the difficulty of detecting treatment effects in homogenous populations, such as patients with stable CAD .