Basic patient characteristics predict antimicrobial resistance in E. coli from urinary tract specimens: a retrospective cohort analysis of 5246 urine samples.

BACKGROUND

Antimicrobial resistance data from surveillance networks are frequently do not accurately predict resistance patterns of urinary tract infections at the bedside.

OJECTIVE: To determine simple patient- and institution-related risk factors affecting antimicrobial resistance patterns of Escherichia coli urine isolates.

METHODS

From January 2012 to May 2015 all consecutive urine samples with significant growth of E. coli (≥103 CFU/ml) obtained from a tertiary care hospital were analysed for antimicrobial susceptibility and related to basic clinical data such a patient age, ward, sample type (catheter vs non-catheter urine).

RESULTS

Antimicrobial susceptibility testing was available for 5246 E. coli urine isolates from 4870 patients. E. coli was most commonly resistant to amoxicillin (43.1%), cotrimoxazole (24.5%) and ciprofloxacin (17.4%). Resistance rates were low for meropenem (0.0%), fosfomycin (0.9%) and nitrofurantoin (1.5%). Significantly higher rates of resistance to ciprofloxacin (32.8 vs 15.8%) and cotrimoxazole (30.6 vs 23.9%) were found in urological patients compared with patients on other wards (p <0.01). In multivariable analysis, predictors for E. coli resistance against ciprofloxacin and cotrimoxazole were: treatment in the urological unit (odds ratio [OR] 2.04, 95% confidence interval [CI] 1.63-2.54; p <0.001 and OR 1.33, 95% CI 1.07-1.64; p = 0.010, respectively), male sex (OR 1.93, 95% CI 1.630-2.29; p <0.001 and OR 1.22, 95% CI 1.22-1.04; p = 0.015), and only to a lesser extent urine samples obtained from indwelling catheters (OR 1.30, 95% CI 1.05-1.61; p = 0.014 and OR 1.26, 95% CI 1.04-1.53; p = 0.020). Age ≥65 years was associated with higher resistance to ciprofloxacin (OR 1.42, 95% CI 1.21-1.67; p <0.001), but lower resistance to cotrimoxazole (OR 0.76, 95% CI 0.67-0.86; p <0.001).

CONCLUSIONS

Simple bedside patient data such as age, sex and treating hospital unit help to predict antimicrobial resistance and can improve the empirical treatment of urinary tract infections.