Table 1.

Primary outcome: number and proportion of contaminated samples, device failures, and RRs of contamination for intention-to-treat and per protocol analysis

Urine resultPeezy UCDWhiz Midstream UCDStandardised verbal instructions
Intention-to-treat analysis, N a395398405
Mixed growth, n (%)b101 (26.5)110 (28.2)117 (29.0)
Not ‘mixed growth’, n (%)280 (73.5)280 (71.8)286 (71.0)
Device failures, n100350
Missing, n1482
Adjusted RR of contamination0.910.98Ref
  95% CI0.76 to 1.090.97 to 1.20
  P-value0.3150.820
Per-protocol, N c295363405
Mixed growth, n (%)b77 (26.3)104 (29.1)117 (29.0)
Not ‘mixed growth’, n (%)216 (73.7)254 (70.9)286 (71.0)
Missing, n252
Adjusted RR of contamination0.911.01Ref
  95% CI0.73 to 1.140.82 to 1.24
  P-value0.4050.955
  • Log-Poisson generalised linear mixed effect model of lab culture urine result modelled against intervention arm as fixed effects, and GP practice as a random effect. Level of significance = 0.025.

  • a All participants with a urine culture result available regardless of whether the allocated collection method was used.

  • b Mixed growth = contamination.

  • c All participants with a urine culture result available from the allocated collection method instructions. RR = relative risk. UCD = urine collection device