Background/Purpose:

Flare is a central feature of gout and patient self-report of flare was found to be an important element in a gout flare definition for research. However, patients may disagree with the clinicians on presence of a flare. In this study, we investigated factors impacting the concordance and discordance between investigators with expertise in gout and their patients, on presence of a gout flare.

Methods:

We used gout patient data collected at 21 international sites from a gout flare definition study published in 2012 and its ongoing validation study. Information on demographics, gout flares, and anatomical location of swelling and warmth during a flare were collected during routine clinic visits. We performed Chi-square and t-test comparisons of these variables in patients who agreed with the investigator on the presence of a flare (concordant) versus those who disagreed with the investigator on the presence of a flare (discordant). We also made comparisons within the discordant group (when investigator diagnosed a flare but patient disagreed versus when patient self-reported a flare and the investigator disagreed).

Results:

Concordant and discordant flares were noted in 187 and 81 cases, respectively. There were no differences between groups in age, sex, or disease duration. Compared to the discordant group, the concordant group had higher pain scores, increased patient global assessment of disease severity, greater proportions of patients with lower extremity joint (knee, ankle, or foot joints) involvement, and lower proportion of patients with tophi (Table). Fewer patients in the discordant group had either swollen or warm joint (patient reported) compared to the patients with concordant flares (76.5% vs 97.9%, p <0.0001). Within the discordant group, 70.4% of flares were patient determined and not endorsed by the investigators. All patients with an investigator-determined flare where patients disagreed had at least one swollen joint (100%) compared to only 57.9% in flares that were patient-determined but the investigator disagreed with (p<0.0001).

Conclusion:

We identified factors associated with agreement and disagreement among patients and investigators on the presence of a gout flare. Lower extremity involvement, higher patient global assessment of disease, less tophaceous disease, higher pain scores, and presence of at least one swollen or warm joint is associated with concordance. Having a swollen joint was a determinant in all the investigator-defined flares when patients disagreed, however it was much less important in patient-determined gout flares.  

Table: comparison between patients in discordant flare group to concordant group

SD= standard deviation, PGA = patient global assessment of disease severity