Clean pulp versus sterile plastic for mid-stream urine collection: a paired equivalence study comparing the microbiological performance of a novel low carbon collection device with the standard of care

Objectives To determine whether a novel urine collection device (the ‘Pee-in-Pot (PiP)’) produces the same rates of reportable urine culture results as standard of care (SOC) urine collection. To determine whether the PiP produces comparable microscopy results to SOC urine collection. To estimate the carbon footprint of the PiP compared to SOC urine collection. Design A prospectively designed, single-centre, paired comparison study. Setting A district general hospital in Southwest England, including antenatal clinical, accident and emergency, medical and surgical ward environments. Participants Adults aged 18 or over. Interventions Urine passed through the PiP device before being decanted into a 10 mL boric acid tube for microscopy and culture, compared with the same urine contained only in a sterile plastic vessel before being decanted into a boric acid tube for microscopy and culture. Primary outcome measure The proportion of positive urine culture results. Secondary outcome measures The proportion of heavy mixed growth culture results. Comparison of particle counts: all small particles, bacteria, red blood cells and white blood cells. Results Microscopy was performed for 1353 paired samples, of which 808 paired samples both underwent culture. Overall, urine cultures were positive in 9.3% (75/808) and 10.0% (81/808) of PiP and control cases, respectively. Overall matching between PiP and control arms for reportable positive culture results was 98.5% (796/808), with a Cohen’s Kappa test coefficient (κ) of 0.9149 (almost perfect agreement). There was no significant difference in the rate of positive urine culture results between testing arms for any organisms (margin of non-inferiority prospectively defined as ±2.5% for Escherichia coli positive cultures). For microscopy, there was agreement in meeting culture thresholds for 1308 of 1353 paired samples with a difference in culturing rates of 0.00517 (95% CI −0.0045 to 0.015, ie, high level of agreement). The estimated base case carbon footprint of PiP testing was 95g CO2e compared to 270g CO2e for SOC testing. Conclusions This study found the PiP to be non-inferior for routine urine microscopy and culture testing and to have a lower carbon footprint compared with SOC urine testing.