Table 6.

Regression model — mean absolute errora

Estimate95% CIP-value
Intercept0.620.37 to 1.040.051
Length of time in service
Years in service (each additional year)1.081.04 to 1.12<0.001
Frequency of use
  >10 times per month0.940.66 to 1.380.76
  6–10 times per month0.820.61 to 1.120.19
  1–5 times per month1reference
  6–11 times per year0.900.68 to 1.200.47
  1–5 times per year0.940.66 to 1.360.74
  Less frequently0.890.61 to 1.320.53
  Not known0.800.38 to 2.070.61
Cost of device
  Cost (every £1 increase)1.000.99 to 1.010.75
Tested (reference: no)
  Yes1.210.80 to 1.910.38
  Not known1.100.87 to 1.410.44
Make (reference: Boots)
  Lloyds1.310.90 to 1.900.13
  Omron1.010.75 to 1.370.94
  Other1.230.80 to 1.920.34
Validation status (reference: not validated/not applicable)
  Validated0.770.59 to 1.020.044
  • a The proportion in MAE change is the unit of predictor unless otherwise stated. MAE = mean absolute error.