Background/Purpose : Recurrent flares of inflammatory arthritis are the central clinical feature of gout.  However, methods of gout flare reporting in research settings are inconsistent and poorly defined.  The aim of this study was to map patterns of gout flare and assess the concurrent validity of different methods of flare reporting by comparison with other methods of gout disease activity.  

Methods: Flare diary entries from a randomised controlled trial of patients with gout were analysed.  Over a four-month period, participants (n=120) completed daily flare diary entries with recording of self-report of flare, pain score (Likert scale 0-10) and medications taken for flare.  The time-course domain in the 2015 ACR/EULAR gout classification criteria was used to define the time elements of a typical flare (time to maximal pain <24 hours, resolution of symptoms in ≤14 days, complete resolution between symptomatic episodes).  Pain x time plots for each individual participant were inspected and analysed for six methods of flare reporting including flare count, time to first flare, number of days with self-reported flare, number of days with self-reported flares requiring medication, number of days with Gaffo-defined flare (CART approach: pain score >3 and self-report)¹, and area under the pain-time curve (AUC pain). Concurrent validity assessment included correlation analysis of these methods of reporting with other measures of gout activity measured monthly over the same time period (area under the curve (AUC) variable-time plot analysis); patient and physician global assessments, joint counts, and C-reactive protein (CRP).  

Results: Inspection of individual plots showed wide variation in the severity, frequency and duration of flare (Figure), with only 55/120 (45.8%) experiencing predominantly typical flares over the four-month study period.  Flare counts over time could not be reliably calculated due to difficulty determining the boundaries of individual flares for some participants.  Time to first flare correlated poorly with other measures of gout activity. In contrast, all other tested methods of flare reporting significantly correlated with other measures of gout activity; AUC pain correlated most strongly with AUC CRP (r=0.43, P<0.001), and number of days of self-reported flare correlated most strongly with other measures of disease severity (AUC patient global 0.62, AUC physician global 0.62, AUC tender joint count 0.53, AUC swollen joint count 0.44, P<0.001 for all).  

Conclusion: There is wide variation in the patterns of flare over time in individuals with gout, leading to challenges for flare reporting in clinical trials.  Time-dependent reporting strategies such as the number of days in self-reported flare correlate well with other measures of gout disease severity and may provide a more accurate measure of severity of flare burden over time.    Reference: Gaffo et al A&R 2012