Table 4

Current and past antibiotic prescribing in predicting rate of reconsultation with cough (from 1 month to 1 year after seeing the doctor).

Adjusted IRRa (95% CI)P-value
Prior antibiotic versus no prior antibiotic (antibiotic for cough in previous 2 years)2.55 (1.62 to 4.01)<0.001
Leaflet versus no leaflet1.27 (0.86 to 1.87)0.229
Randomised prescribing strategies (when no antibiotic for cough in previous 2 years)
Immediate antibiotic1.00
No antibiotic1.23 (0.79 to 1.92)0.237
Delayed antibiotic1.19 (0.78 to 1.80)0.426
Randomised prescribing strategies (when antibiotic prescribed for cough in previous 2 years)b
Immediate antibiotic1.00
No antibiotic0.66 (0.30 to 1.44)0.295
Delayed antibiotic0.22 (0.10 to 0.49)<0.001
  • a The estimates for each variable controls for the effects of the other variables listed, for example, the incidence rate ratio for no antibiotic is 0.66, which means when no antibiotic is prescribed there is a 34% reduction in reconsultation rates when controlling for all the other variables listed.

  • b These estimates are the interaction terms between prior antibiotic and randomised strategies.