I made a multiparameter fittting (7 adjustment parameters, 45 experimental points). I obtained enormous confidence intervals for 95%. There is a criterion for deciding whether the adjustment is considered appropriate?

I have a more general question: in a parameter fit for a correlation, how confidence intervals should be interpreted?