Empiric antibiotic prescribing in selected primary care settings: Identifying possible areas for improvement

Abstract

Background. Two undeniable challenges facing modern medicine are the increasing problems of antimicrobial resistance and the lack of new antibiotics to combat infections. Alone these problems are serious, but the synergy of these two problems threatens to return us to the pre-antibiotic era. These problems are global and affect all levels of the health care system, from the rural family physician to the intensive care unit of tertiary hospitals. Rational antibiotic prescribing is paramount in meeting the challenges of antibacterial resistance. In Norway, approximately 90 % of all antibiotics are prescribed in primary care. The aim of this thesis is to elucidate antibiotic prescribing in selected primary care settings to identify possible areas for improvement. Methods. Papers I and II are retrospective examinations of patient records at nursing homes over a one year period to identify the infection being treated with antibiotics and whether the choice of antibiotic was in line with the national guidelines for antibiotic treatment. In paper I the extent of microbiologic diagnostics ordered prior to empiric antibiotic therapy is registered. Paper III examines positive urine cultures from patients 65 years and older living in 34 different nursing homes over a thirteen month period and compared etiology and resistance rates of uropathogens isolated from patients 65 years and older living at home. Paper IV is a non-randomized controlled cluster intervention study comparing antibiotic prescribing for cystitis and pyelonephritis prior to and after an intervention. Results. Paper I. 94 infections were treated with ciprofloxacin. Urinary tract infection (UTI) was the most common infection both on the long-term (78 %) and the short-term wards (40 %). Respiratory tract infection was almost as common as UVI on short-term wards (37 %), but was uncommon on long-term wards (4 %). Specific bacterial etiology was identified in 44 infections (47 %), 12 of these by bacteria only susceptible for ciprofloxacin. Paper II. 714 antibiotic courses were prescribed to 327 patients yielding a prevalence of patients treated with antibiotics of 6.6 %. Prescribing compliant with national guidelines was 77 % for UTI, 79 % for RTI, and 76 % for skin and soft tissue infections (SSTI). Ciprofloxacin was responsible for 63 % of noncompliant prescribing. On the short-term wards there was a higher rate of total prescribing, non-compliant prescribing, and prescribing by physicians employed at the local hospital Paper III. In both the nursing home group and in the group living at home Escherichia coli (64 % both groups) was the most commonly cultured bacteria followed by Enterococcus faecalis (10 % vs 8 % respectively). Escherichia coli was also the most commonly cultured bacteria in females (70 %) and males (39 %), but Enterococcus faecalis was significantly more common in males (18 %) than females (7 %) (p < 0.05). For males there was a significantly higher resistance rates to ciprofloxacin for Escherichia coli than for females (12 % vs 7 %; p < 0.05) and to mecillinam for Proteus mirabilis (12 % vs 3 %; p < 0.05). Paper IV. In the targeted emergency department (ED), there was a significant (p < 0.05) reduction in ciprofloxacin prescribing and a significant increase in mecillinam prescribing for cystitis (p = 0.042). In the control ED, prescribing of ciprofloxacin doubled (p < 0.05). There were no significant changes in antibiotic prescribing for pyelonephritis in either ED. Conclusion. The areas of improvement identified in this thesis include pre therapy microbiologic diagnostics, the need to consider restrictions on prescribing broad spectrum antibiotics, and the need for specific guidelines for the elderly based on gender for the treatment of UTI.