In addition to Max' comments, there might exist situations in SLE patients (other than infections) where CRP is elevated. One recent study found crp to be elevated in lupus pleuritis but not effusion alone http://www.ncbi.nlm.nih.gov/pubmed/25318967.
I also think there is no good explanation for the ESR/CRP dissociation. The main clinical problem is the distinction between flare and infection. Some have looked at procalcitonin levels. One recent review article discussed procalcitonin in rheumatic diseases, in most instances (with the exception of gpa and aosd) procalcitonin should be negative in the absence of infection. Of note, viral infections suppress the generation of pct.