Table 5.

Multivariable GEE model estimating the odds of receiving a plasma glucose test within an eligible period, for the RA cohort.

Predictive VariableaOR (95% CI)*p
Male vs female0.91 (0.87–0.96)0.0003
Rural vs urban0.99 (0.93–1.05)0.7689
Northern, Coastal, and Interior vs Island and Fraser Health Authorities0.93 (0.88–0.97)0.0022
SES index, highest quartile vs others0.88 (0.84–0.93)< 0.0001
Romano comorbidity score, ≥ 1 vs 00.94 (0.90–0.99)0.0099
Having a hospitalization in the prior year1.06 (1.01–1.11)0.0278
Seeing a rheumatologist in the prior 5 yrs**0.78 (0.75–0.82)< 0.0001
Having no physician visits in the prior year0.53 (0.45–0.62)< 0.0001
Timing of periods, starting < 2001 vs ≥ 20010.49 (0.47–0.51)0.0002
GC use within the eligibility period, yes vs no1.12 (1.05–1.19)< 0.0001
  • * Adjusted OR from GEE model estimating the odds of receiving a PG test for each categorical variable, adjusted for all variables listed and for age and age-squared.

  • ** Measured in the 5 years prior to the start of period.

  • Measured in the year prior to the start of period. RA: rheumatoid arthritis; SES: socioeconomic status; GC: gluco-corticosteroids; PG: plasma glucose; GEE: generalized estimating equation.