Pharmacy Review of Culture and Sensitivity With Prompting of Physicians to Reduce Antibiotic Pressure

To the Editor: Antibiotic resistance is a problem in nursing homes.1 Infections are treated empirically with broad spectrum antibiotics that select resistant organisms, but do not provide coverage for the resistant organisms selected.2,3 Whenever possible, antibiotic choices should be directed by culture and sensitivity to identify resistant organisms that would not respond to empiric antibiotics and to allow substitution of narrow spectrum for broad spectrum antibiotics. We report a pharmacy-based program that reviews culture results and makes suggestions to focus antibiotic therapy. The Wisconsin Veterans Home is a 721-bed skilled nursing facility (SNF). The point prevalence of antibiotic use is 6.3%. The home has an on-site bacteriology laboratory. Culture and sensitivity results and current antibiotic therapy were reviewed Monday through Friday. Pharmacists do not call if the resident is not on an antibiotic for fear of encouraging antibiotic treatment of “colonization.” If the existing antibiotic does not provide coverage, the physician is contacted. In addition, the pharmacist looked for opportunities to reduce antibiotic pressure. Between May 1, 1999, and October 31, 2000, 372 reports were reviewed, with 43 calls to physicians. The antibiotic did not cover organisms such as Enterococcus, methicillin-resistant Staphylococcus aureus (MRSA), or Pseudomonas, and the call recommended a switch to an effective antibiotic. Table 1 presents the 23 cases currently treated with a quinolone or cephalosporin for a single organism also sensitive to an antibiotic with a narrower spectrum. The physician elected to make the switch 14 times, with four cases associated with late communication. In five, the physician declined to make the switch. The rationale was not recorded. A catheterized patient was treated for Citrobacter freundii urinary tract infection with ciprofloxacin. The culture and sensitivity returned 2 days later, revealing sensitivity to ciprofloxacin, trimethoprin (TMP)-sulfa, tetracycline, and nitrofurantoin. The physician declined to make a switch to TMP-sulfa. A follow-up culture 20 days later revealed a heavy growth of MRSA (for the first time in that individual) resistant to ciprofloxacin, but sensitive to TMP-sulfa, tetracycline, and nitrofurantoin. Programs to reduce antibiotic resistance should include a diversity of antibiotics and emphasize prevention of resistance rather than cost savings.4 We have had a modest initial impact and will improve timeliness. We will determine why physicians decline a suggested switch. Possible reasons include a need to calculate a dose reduction (TMP-sulfa, nitrofurantoin); desire to treat with an antibiotic that penetrates prostatic tissue (e.g., quinolones, TMP-sulfa); lack of confidence in sputum or wound cultures; concern about infection in a second, noncultured site; or a desire not to “rock the boat” when the initial antibiotic lead to a good result. We do not know how much impact a fully effective program may have on antibiotic resistance or cost of therapy. However, a switch to narrow spectrum antibiotics after the return of culture and sensitivity is universally recommended to decrease the emergence of resistance. Quinolone resistance is common among MRSA strains.5 It follows that quinolone pressure would favor the emergence of MRSA. In the catheterized resident with Citrobacter urinary tract infection, (Table 1) one might speculate that if the switch from ciprofloxacin to TMP-sulfa had been made, MRSA would not have emerged. Bjork et al. reported that 70% of 63 antibiotic courses in catheterized residents were followed by a resistant organism.6 Our recommendations to switch primarily involved quinolones in urinary tract infections. Quinolones are also suited for empiric treatment of lower respiratory tract infections.7 A recent Society for Healthcare Epidemiology of America position paper stated, “These agents are used because they allow oral therapy with an agent with good bioavailability, are easily administered, … and have a wide spectrum of activity. With intense quinolone use in LTCFs, resistance has increased.”1 Chen et al. reported increasing pneumococcal resistance to fluoroquinolones, highest in older patients associated with higher rates of use. No pneumococci with reduced susceptibility were noted in children, where quinolone use was negligible.2 Lee et al. reported low but increasing rates of colonization with ciprofloxacin-resistant Gram-negative bacilli in a skilled nursing facility where quinolones were the most commonly prescribed antibiotics.8 Withdrawal of antibiotic pressure has been suggested to reduce resistance. Examples of successful hospital-based cephalosporin intervention programs have been reported.3,4 During a course of antibiotics, the selection of resistant bacteria is inevitable. Sensitivity is likely to be restored if the resident bacteria are given time to reclaim their territory on an even playing field. Unfortunately, resistant bacteria are constantly being selected in the 6% to 8% of nursing home residents being treated with antibiotics.1 Direct transmission probably occurs continuously between socializing, hygiene-challenged residents or on the hands of staff racing to meet the needs of dependent residents.9 The emergence of antibiotic resistance seems inevitable. We anticipate that every switch from an empiric quinolone or cephalosporin to a narrow spectrum agent will extend the efficacy of these useful agents as empiric therapy.