Tophi and Frequent Gout Flares Are Associated With Impairments to Quality of Life, Productivity, and Increased Healthcare Resource Use

Results From a Cross-Sectional Survey

Puja P Khanna; George Nuki; Thomas Bardin; Anne-Kathrin Tausche; Anna Forsythe; Amir Goren; Jeffrey Vietri; Dinesh Khanna


Health and Quality of Life Outcomes. 2012;10(117) 

In This Article


Data Source

Self-reported data were obtained from patients identified through the US and EU versions of the 2010 National Health and Wellness Survey (NHWS; Kantar Health, New York, NY, USA) and the Lightspeed Research (LSR; New York, NY, USA) ailment panel (used to supplement respondents not available through NHWS). The NHWS is a demographically representative, annual cross-sectional survey of adult respondents (18 years of age and over), providing self-reported information on treatment, healthcare attitudes and behaviors, patient disease and demographic characteristics, and health-related outcomes. NHWS and LSR ailment panel members are both sourced from a more general LSR panel, whose members are recruited through opt-in emails, co-registration with panel partners, e-newsletter campaigns, online banner placements, and both internal and external affiliate networks. All panelists explicitly agreed to become panel members, registered through unique email addresses, and completed in-depth demographic registration profiles.

A stratified random sample procedure is implemented for NHWS so that the final sample mimics the demographic composition of the country in which it is administered, in order to achieve better representativeness. The US sample is stratified by age, gender, and ethnicity, and the EU sample is stratified by age and gender. Comparisons between the NHWS sample, the US census, and other national surveys have been made elsewhere.[25,26]

All respondents, from both NHWS and the current study, gave informed consent, and the study was approved by the Essex Institutional Review Board (Lebanon, NJ, USA).

Study Population

Panel members reporting a physician diagnosis of gout were invited to participate via email in an online, self-administered survey. Out of 1936 patients reporting a physician diagnosis of gout invited to participate, 747 responded (a 39% response rate), and 620 patients completed the survey (563 via NHWS), including 338 (54.5%) from the US, 181 (29.2%) from the UK, 85 (13.7%) from Germany, and 16 from France (2.6%).


Participants completed a web-based questionnaire that included questions regarding the patient's gout and several validated scales.

Gout Characteristics. Information was collected concerning the patient's gout including their most recent serum urate level (sUA: <6 mg/dL/360 μmol/L, 6 - 8 mg/dL/360-480 μmol/L, or >8 mg/dL/480 μmol/L), the number of flares in the past 12 months (don't recall, 0, 1–2, 3, 4–5, or 6+), the presence of tophi (not sure, 0, 1, or 2+), and the number of years since they were diagnosed with gout.

Health-related Quality of Life. HRQOL was assessed using the Medical Outcomes Short Form 12 (SF-12v2) questionnaire.[27] This instrument allows for the calculation of physical (PCS) and mental (MCS) component summary scores. Scores for the PCS and MCS are normed to the US population (Mean = 50, SD = 10), with higher scores indicating greater HRQOL. SF-6D health utilities were also calculated from responses to the SF-12v2.[28] Scores for the SF-6D range from 0.29 (extremely poor health) to 1 (perfect health). Differences in PCS and MCS exceeding 3 points are considered minimally important differences (MIDs),[29] and 0.03 is the MID for the SF-6D.[30]

Resource Use. The number of visits to a traditional healthcare provider in the past six months was assessed.

Work Productivity and Activity Impairment. Work productivity impairments and impairment in daily activities were assessed using the validated Work Productivity and Activity Impairment (WPAI) questionnaire.[31] Four subscales (absenteeism, presenteeism, overall work impairment, and activity impairment) were generated in the form of percentages, with higher values indicating greater impairment. Absenteeism represents the percentage of work time missed due to health in the past seven days. Presenteeism represents the percentage of impairment while at work due to health in the past seven days. Overall work impairment (OWI) is the total percentage of work time missed in the last 7 days due to either absenteeism or presenteeism. Activity impairment represents the percentage of impairment during daily activities. Only employed respondents provided data on absenteeism, presenteeism, and overall work impairment, but all respondents provided data on activity impairment.

Statistical Analyses

Descriptive statistics and frequency distributions were undertaken for all variables. Bivariate comparisons on continuous variables and scales were conducted using t-tests for comparisons between two groups and ANOVA for comparisons with three groups or more, followed by Tukey HSD post hoc tests. Chi-square tests were conducted in the case of categorical variables. Because symptom levels were collected as categories rather than exact numbers, correlations between levels of flares, tophi, and sUA were conducted using Spearman's rho.

For the purposes of analysis, patients who did not know whether they had experienced flares were combined into the zero flares group, the assumption being that any flare that they might have experienced was not serious enough to have been identified by a physician. Moreover, analysis of the prevalence of flares revealed that there were only eight patients (1%) who did not recall whether they had flares. Nearly one-third (n = 195, 31%) of respondents were unsure of whether they had tophi, possibly indicating that they suspected a tophus but were not certain given the description in the questionnaire ("Tophi are deposits of crystallized uric acid that can appear as moveable lumps or whitish nodules"). Given the substantial number of such patients, it was deemed prudent to examine them as a separate group.

Multivariable models predicting healthcare resource use, work productivity, and HRQOL included frequency of flares in the last 12 months (1–2, 3, 4–5, or 6+, relative to 0/don't recall flares), the presence of tophi (1+, or not sure, relative to 0 tophi), age, gender, and length of gout diagnosis (in years) as predictors. HRQOL analyses (MCS, PCS, and SF-6D) were conducted using maximum likelihood multiple regression to adjust the MCS and PCS scores and SF-6D health utilities for the effects of age, gender, and length of illness. Distributions of scores were examined and were normally distributed, so no transformation was applied. Because work productivity, impairment, and resource utilization are often highly skewed, we used a generalized linear model (GLM) approach specifying a negative binomial distribution. Adjusted logarithmically transformed counts were modeled in the GLM. This approach tested whether the adjusted transformed counts differed among the groups while accounting for covariates. In order to make the results more interpretable, we calculated the antilog of regression estimates, which yields rate ratio (RR) values. The rate ratio indicates the number of times greater impairment was for the given group compared with the reference group (e.g. a rate ratio value of 1.5 would indicate that the mean level of absenteeism for the ≥ 3 flares/year group is 1.5 times that of no flares).

The gout questionnaire did not assess respondents' comorbid status. However, given the likely contribution of higher comorbidities to poorer health outcomes among respondents, comorbid status may be confounded with gout flares and tophi in the analyses. Therefore, a supplementary set of analyses was conducted on the subsample of respondents re-contacted via the NHWS (n = 563), using the same predictors and outcomes as described above, but adjusting additionally for relevant variables available from the NHWS, in which respondents had participated previously. In particular, the supplementary models controlled for self-reported diagnosis with diabetes mellitus, hypertension, or chronic kidney disease, as well as BMI category (overweight, obese, or missing BMI, vs. normal/underweight as reference). Given the loss of statistical power associated with the reduced sample, additional covariates, and noise introduced by non-contemporaneous measures assessed at different lags over the course of the previous year, the supplementary results are reported only as a footnote to the main analyses, for the purpose of helping rule out obvious confounds.

Statistical significance was set at p < 0.05 for all hypotheses tested. Analyses were performed using SPSS version 19.0.