EP Evaluator Corr Coef

Typical Correlation Coefficient (R) acceptable range is 0.975 - 1.000, where 1.000 is perfect. It is important to understand that R should not be used as a sole criteria for least squares (or your regression line comparison). R is really there to help assess the reliability of the linear regression calculation. It describes how much of the change in one instrument (this one is your POCT instrument)  is due to change in the other instrument (this one is the main lab or true instrument).
Correlation does not mean match...

Using Bland altman plots give better indicator for comparisons. 

We use 0.95 - 1.05 for Acceptable Correlation Coefficient. 
The more data you have, the better your correlation will be.
We use EP Evaluator to calculate our statistics.
 It has a good interpretation and explanation of the values you get. 

Your medical director should have signed off on what acceptable limits are and define your ways to prove accuracy in a method evaluation procedure. Check it out. Typical Slope acceptable criteria is 0.95- 1.050. Correlation Coefficient is different. Using Westgard as a reference, R=1 is perfect and this shows that there is perfect correlation between the two methods. Data <1 shows scatter between the two methods. When R is 0.99 or greater, then the linear regression line shows a reliable estimate of error. (reliable being the operative word here) When R is less than 0.975, it is better to use the other paired data calculations.

Again, check with your medical director and current procedure that would outline acceptable limits.

Good luck!
Erika

I agree with Erika, you cannot use Corr Coef to determine whether your correlation passes or fails, but you can use Total Allowable Error.  It is up to your Lab Director to define what is acceptable.  Please reference your policies.  If you don't have one, please write one.