Evaluating direct antimicrobial susceptibility testing (DAST) of urine samples with the help of automation in a hospital setting.

OBJECTIVE

Direct antimicrobial susceptibility testing (DAST) results were compared against routine antimicrobial susceptibility testing (AST) results for seven commonly prescribed antibiotics to enable next-day reporting of positive urine cultures.

METHOD: Urine samples were set up by BD Kiestra InoqulA for routine culture and additional DAST, using 10 µL as inoculum volume. The categorical agreement (CA) between DAST and routine AST methods by Phoenix and standardised disk diffusion (SDD) was established. Statistical analysis of antimicrobial activity interference in urine was performed using Fisher's exact test.

RESULTS

Mono-microbial Enterobacterales growth was observed in 506 of 1372 urine samples. CA between Phoenix and DAST was 95.7%. Ciprofloxacin performed worst with 3.7% very major errors and 89.7% categorical agreement. Major error of 5.4% was observed for Trimethoprim-sulfamethoxazole. For a subset of 200 urine isolates, SDD was performed in addition to Phoenix. The CA between DAST and SDD exceeded 90% for all studied antibiotics. Antimicrobial activity was detected in 58 samples but demonstrated no statistically significant impact on error rates.

CONCLUSION

We propose a streamlined workflow for urine DAST to enable next-day reporting. Additional AST is only required when no or non-confluent growth on DAST plates, polymicrobial growth or resistance to cephalexin/cefuroxime which indicates presence of resistance mechanisms. This approach significantly reduces routine UTI diagnostic turnaround times whilst maintaining appropriate detection of clinically significant resistance.