Table 2

Absolute rates of gluten-free prescribing per 1000 patients, stratified by various factors, followed by incidence rate ratios from a univariable Poisson model, then from a multivariable mixed-effects Poisson model

Quintile boundariesMedian gluten-free prescriptions per 1000 patientsUnivariable incidence rate ratio*95% CIMultivariable incidence rate ratio†95% CI
Composite measure score (lower percentile is better)<39%13.44ReferenceReference
39% to 44%16.181.081.06 to 1.101.081.06 to 1.09
44% to 48%18.391.151.13 to 1.171.091.07 to 1.10
48% to 53%19.291.181.16 to 1.201.091.07 to 1.11
>53%21.681.271.25 to 1.291.131.10 to 1.15
Quality Outcomes Framework score (maximum=559)<52214.13ReferenceReference
522 to 54116.921.171.15 to 1.191.101.08 to 1.12
541 to 55018.131.201.18 to 1.221.101.08 to 1.12
550 to 55619.451.261.24 to 1.281.091.07 to 1.11
>55620.351.291.27 to 1.311.111.09 to 1.13
Index of Multiple DeprivationLeast deprived19.46ReferenceReference
19.671.000.98 to 1.010.960.94 to 0.98
17.360.930.91 to 0.940.950.93 to 0.97
14.930.820.81 to 0.830.870.85 to 0.89
Most deprived17.620.940.92 to 0.950.890.87 to 0.91
Practice list size<378412.35ReferenceReference
3785 to 579617.491.161.14 to 1.171.181.16 to 1.20
5798 to 801818.501.151.13 to 1.171.181.16 to 1.20
8020 to 11 16520.211.231.21 to 1.251.191.17 to 1.21
>11 16519.111.161.14 to 1.181.171.15 to 1.19
Dispensing practice?No17.57ReferenceReference
Yes19.441.061.05 to 1.081.071.06 to 1.09
% of patients over 65<10.8%10.29ReferenceReference
10.8% to 15.4%16.291.281.25 to 1.301.091.07 to 1.11
15.4% to 18.8%19.311.471.44 to 1.491.221.20 to 1.25
18.8% to 22.4%22.391.651.63 to 1.681.341.31 to 1.37
>22.4%22.831.681.65 to 1.701.461.42 to 1.50
% of patients with a long-term health condition<47.0%12.82ReferenceReference
47.0% to 51.5%17.341.191.17 to 1.200.990.97 to 1.00
51.5% to 55.3%17.871.261.24 to 1.281.000.98 to 1.01
55.4% to 59.7%21.261.361.34 to 1.380.990.97 to 1.01
>59.7%21.041.341.32 to 1.360.980.96 to 1.00
  • *Poisson regression.

  • †Mixed-effects  Poisson regression, adjusted for all other variables in table as fixed effects, plus Clinical Commissioning Group as a random effect.