Antibiotic treatment for this case
The main facts that influence targeted therapy are:
Mr Jenkins had a (severe) urinary sepsis and we have good evidence of the etiological agent: an ESBL-producing E. coli
Although he is still febrile, the clinical course is favourable
No evidence of abscess or urinary obstruction has been found
Considering all these facts, Mr. Jenkins probably benefits from:
Continuing intravenous antibiotic therapy, at least until he is afebrile
Continuing therapy with a single antibiotic
Narrowing antimicrobial spectrum
Although probably the most active ß-lactam among the choices is meropenem, carbapenem is one of the most valuable drugs available in the antibiotic armamentarium and should be protected whenever possible.
Carbapenem-sparing strategies are needed in an era of increasing antimicrobial resistance. In this regard, as we seek to treat a urinary tract infection, which is a “low-risk” source provided there is no abscess nor obstruction and that the patient has clinically improved, amoxicillin/clavulanate (if the intravenous presentation is available) or piperacillin/tazobactam could be appropriate treatment options. Since we do not need anti-Pseudomonas activity, amoxicillin/clavulanate would be our choice if we were willing to continue therapy with a ß-lactam. In addition it has the advantage of an easy IV to PO switch.
Mr Jenkins could also be treated using non ß-lactam mono therapy with either amikacin or TMP/SMX.
Since Mr Jenkins is elderly, I would avoid amikacin if possible due to is nephron and ototoxicity.
TMP/SMX would be preferred to amikacin especially when the patient is ready for the IV to PO switch. In addition, TMP/SMX would be a better option than amikacin and even that ß-lactam if a prostate involvement (prostatitis) is suspected given its PK/PD properties.
You may find the paper from the Scottish Antimicrobial Prescribing Group (SAPG) which provides advice for optimising antimicrobial prescribing for multi-drug resistant Gram-negative bacteria useful.