Pseudomonas aeruginosa and antimicrobial resistance : Mono-therapies vs Combinations, a big data perspective

Big data analytics in Indian context is a new and emerging tool for surveillance of critical health issues such as antimicrobial resistance. World bank report has stated that AMR may cost 4-5 % of wold GDP by 2050 and push millions of people in poverty. In this article i am discussing how big data analytics can help us in not only long term policy formulation but guiding medical fraternity in taking evidence based decisions in management of their patients. Antimicrobial resistance is a dynamic phenomenon which shows a great spatio-temporal variation and a right candidate for a keen and tight surveillance.

Community as well nosocomial Pseudomonas infections are one of the most dreaded infections and associated with high mortality and incur huge cost to health care system. Though lab based surveillance system has been established by Indian council of medical research but still a large gap exist. Involvement of large number of private partners, who cater a major chunk of patients and have a great amount of information with them, can add valuable insights.

An analysis was conducted on diagnostic lab data (Private sector) of AMR with regard to Pseudomonas aeruginosa. Major findings are described here. Important antimicrobial agents used in treating Pseudomonas includes aminoglycosides (Tobramycin, amikacin & gentamicin), carbapenems (imipenem, meropenem), cephalosporins (cefepime, ceftazidime), fluoroquinolones (levofloxacin, ciprofloxacin), penicillin (Ticarcillin-clavulanic acid, Pipercillin-tazobactem), monobactem (aztreonam), polymyxins (Colistin, polymyxin B) and fosfomycin. Analysis was conducted on positive culture results of 6379 subjects. It includes positive samples of blood, urine, pus, stool and sputum for Pseudomonas aeruginosa with variable number of sensitivity tests for different antimicrobial agents. In our analysis we included 20 individual and 27 different combinations of antimicrobial agents commonly used against Pseudomonal infections. samples were taken between 2007-2018 from various diagnostic labs across India.

Important findings of the study are summarized here. Among individual antibiotics Colistin and polymyxin B were found to be with highest coverage (>90%), while fosfomycin was modestly good (84% sensitivity). Other drugs showed lesser coverage like aminoglycosides (Amikacin, tobramycin, gentamicin) close to 70% or less. Carbapenems (Imipenem, meropenem) were also close to 70%. Among cephalosporins cefepime had better coverage than ceftazidime (68% vs 61%). Sensitivity to ciprofloxacin and levofloxacin was in the range of 60%. Surprisingly Pseudomonas showed very less sensitivity to Tigecycline (6.7%). Coverage of piperacillin-tazobactem and aztreonam was also poor.

Among combinations highest sensitivity was for combinations of tobramycin with imipenem, doripenem, meropenem and piperacillin-tazobactem. Coverage of these combinations were in the range of 85-90%. Combination of these antimicrobial agents with other aminoglycosides (gentamicin, amikacin) has coverage less than 80%. Combinations of cefepime, ceftazidime, aztreonam with aminoglycosides or floroquinolones showed coverage in range of 70-77%.

In conclusion colistin and polymyxin B are far superior in coverage for Pseudomonas in comparison to any other individual antimicrobial agent or any combination of these. Theses type of big data based insight can guide medical practitioners in making evidence based decisions during empirical management.