Ok, here is a dataset with fake numbers. I cant share any part of the original database. I have 5 different ways of measuring the blood pressure. One is a gold standard (arterial line). One is a device, and the others are cuffs at different locations. I'm just using mean arterial pressures here. These are taken intra-operatively in a sample of patents before an event. Then repeated 5 times . Ive done Bland Altman Plots, Lin's concordance coefficient, and ran a linear regression with all the Blood Pressure parameters measured each time within one model comparing to each other. What I don't have is a way to include all the data within one model comparing the precision and bias of all the measures at each repeated time point to the gold standard. If I just take a mean of each measure at each time point, then we lose the granularity of the data at each time point. The Bland Altman tests and Lin's Con concordance coefficient do a great job of comparing MAP measures to the MAP arterial line, but only for specified times where data are measured.
... View more