Multivariate Analysis of Klebsiella pneumoniae Antimicrobial ‎Resistance Phenotypes Isolated from Different Clinical ‎Specimens at Mila Hospital, Algeria

Aims: There are few data regarding the prevalence and trends of Klebsiella pneumoniae antibiotic resistance in Algeria. The present study was conducted to investigate the spatial distribution of K. pneumoniae antibiotic resistance phenotypes in time and according to specimen source.
This retrospective study was performed between January 2011 and December 2015 at Mila Hospital, Algeria. A total of 172 K. pneumoniae were isolated from consulting and hospitalized patients, and their antimicrobial susceptibility was tested. The Principal Component Analysis (PCA) was used to study correlations among antimicrobial resistance phenotypes observed, and Factorial Correspondence Analysis (FCA) was used to study the spatial distribution of antibiotic resistance phenotypes according to specimen source.
Findings: The specimens were obtained from urine (n=89), vagina (n=39), pus (n=33), blood (n=9) and surgery (n=2). PCA showed two principals associations of resistance phenotypes gathered in two clusters. The first profile regroups amoxicillin-clavulanic acid, cefazolin and ampicillin. The second assembles cefotaxime, nalidixic acid and sulfamethoxazole-trimethoprim. In FCA, nalidixic acid was connected with urine specimens, registering maximum resistance (52.8%) compared to the other samples. Vagina specimens were associated to sulfamethoxazole-trimethoprim and colistin phenotypes registering maximum resistances with 89.7 and 76.9%, respectively. Pus manifested a near association to cefotaxime with a maximum resistance (48.5%).
The model developed in FCA, highlights typical associations of antibiotic resistance phenotypes to specimen source and confirms the difference in resistance profile according the source of specimen in K. pneumoniae infections.