Drug use evaluation (DUE) of ceftriaxone: A quality metric in a pediatric hospital

Abstract

Background

Ceftriaxone is frequently used as empiric therapy because of its broad spectrum and dosing characteristics. The purpose of this study was to evaluate the appropriateness of ceftriaxone therapy among hospitalized children using drug use evaluation (DUE) methodology.

Methods

Hospitalized patients who received one or fewer dose of intravenous ceftriaxone at Children’s Hospital of Eastern Ontario between January 1, 2018, and June 30, 2018, were identified. Duration was defined as empiric if 72 or less and definitive if more than 72 hours. Two infectious disease physicians reviewed the charts and rated appropriateness using a previously developed scale.

Results

A total of 276 ceftriaxone courses in 248 patients (mean age 6.0 y) were reviewed. Of these, 153 (55.4%) were assessed as definitively or possibly indicated. The most common reason for inappropriate empiric use was an overly broad spectrum. Of the 120 courses given empirically for which there was no indication, the three most common reasons were lower respiratory infections (51; 42.5%), head and neck infections (18; 15.0%), and intra-abdominal infections (15; 12.5%). Of the 39 (14.1%) courses of ceftriaxone that were given for more than 72 hours, 14 (35.9%) met criteria for a definitive or possible indication.

Conclusion

Ceftriaxone is still overused as empiric therapy. Although 85% of courses were discontinued after three doses, 14% were continued for longer than 72 hours, with approximately one-third ultimately meeting an indication. DUE using Canadian pediatric and local guidelines criteria is useful to identify clinical presentations for which narrower spectrum antimicrobials should be used.

Ceftriaxone, a third-generation cephalosporin, has broad-spectrum activity against common gram-negative organisms, such as Escherichia coli, Haemophilus influenzae, and Neisseria spp, and gram-positive pathogens, including Streptococcus pneumoniae. Its prolonged half-life and substantial biliary penetration results in microbiome changes that lead to colonisation with resistant Enterobacteriaceae and an increased risk of Clostridium difficile infections (1–4). A recent US survey indicates that use of third-generation cephalosporins increased by 12% between 2006 and 2012 (5). Drug use evaluations (DUEs) can potentially identify areas of suboptimal use and identify targets for antimicrobial stewardship programs (ASPs) (6,7). The purpose of this study was to analyze the appropriateness of ceftriaxone therapy in patients hospitalized at a pediatric hospital in order to target its overuse.

Methods

The Children’s Hospital of Eastern Ontario is a 92-bed tertiary care pediatric teaching hospital in Ottawa. The ASP has performed prospective audit and feedback rounds two to three times a week in the pediatric intensive care unit since January 2013 and in all pediatric inpatient medicine services since January 2015. The hospital uses ceftriaxone with patients aged older than 6 weeks; younger children receive cefotaxime.

A retrospective study was conducted among patients aged 6 weeks to 18 years who received at least one dose of ceftriaxone (dispensed in either premixed bags or vials) between January 1, 2018, and June 30, 2018. The site uses Epic’s (Verona, WI) electronic health record (EHR) system for patients and its Willow pharmacy electronic record, which interfaces with patient EHRs. Any doses of ceftriaxone given during a specified time period were obtained using a predefined logic program from the EHR for this purpose. Using patient-unique identifiers, the EHR was accessed to review notes and extract the following data: patient demographics (month of admission, date of birth, age, sex, date of admission, admitting service), clinical characteristics (admission diagnosis, underlying medical conditions, any known allergy to antibiotics or other agents), and microbiological data (including bacteriology and virology results).

Data were then exported into an Excel program (Microsoft Corp., Redmond, WA). Start and stop dates were available for each patient, from which days of therapy could be calculated (stop date minus start date + 1). Each consecutive day of treatment was considered as one episode. Separate episodes of treatment were also included if they had different admission dates or were separated by more than 72 hours of treatment.

The local research ethics board reviewed the protocol and determined that the study met criteria for a quality assurance study and did not require individual patient consent.

An appropriateness scale was developed for use of antibiotics for empiric and definitive indications (Box 1). This scale was developed using Canadian Paediatric Society guidelines for community-acquired pneumonia and urinary tract infections and the Redbook as reference points. Local guidelines based on these references are available on site and electronically (Spectrum, Spectrum Mobile Health Inc., Vancouver, BC) and inform ASP recommendations. Empiric therapy was defined as antibiotics given for 72 hours or less and definitive therapy as antibiotics given for more than 72 hours. Two infectious disease physicians independently reviewed the clinical data and assigned an appropriateness grade. Discrepancies were identified and discussed to arrive at a consensus. Descriptive statistics, including averages and proportions, were used to compare appropriateness ratings by indications (empiric or definitive).

Box 1: Rating scale used to measure the indication for ceftriaxone.

Definite indication Isolation of a pathogen before susceptibility results or confirmed resistant to narrower spectrum agents Empiric therapy for bacterial meningitis or presumed CNS infection or confirmed pathogen requiring ceftriaxone Empiric therapy for fever without focus in infants aged ≥6 weeks or <3 months Empiric therapy for fever in a returning traveller at risk of Salmonella typhi/paratyphi Suspected or confirmed gonococcal infection Empiric therapy for endocarditis Fever in a high-risk group (e.g., asplenia, sickle cell anemia) Suspected meningococcemia

Possible indication While awaiting culture data for serious sepsis Severe CAP prompting ICU admission

No indication No CNS infection Spectrum too broad for the indication or culture result Duration too long according to local guidelines Drug–bug mismatch (i.e., isolated pathogen not covered by ceftriaxone) No indication for antimicrobial use

CNS = Central nervous system; CAP = Community-acquired pneumonia; ICU = Intensive care unit

Results

A total of 276 courses of intravenous ceftriaxone were prescribed to 248 patients between January 1, 2018, and June 30, 2018, and 18 (7.2%) patients had separate courses during the study period. The patients’ mean age was 6.0 years, and 118 (47.6%) were male. Overall, 92 (37.1%) had an underlying medical condition, and 12 (4.8%) reported a penicillin allergy. Of the 276 empiric courses, blood or urine cultures were taken before the start of ceftriaxone for 248 (89.9%).

Of the 276 total courses, 129 (46.7%), 87 (31.5%), and 21 (7.6%) ceftriaxone courses were administered for 1, 2, and 3 days respectively, and 39 (14.1%) were given for more than 3 days. The number of courses that were given empirically for 72 hours or less was 237 (85.9%) (Figure 1). The most common indications for empiric ceftriaxone use were respiratory tract infections (RTIs) (n = 70; 25.3%), followed by suspected sepsis or bacteremia (SSB) (n = 58; 21.0%), suspected meningitis, or central nervous system infections (n = 53; 19.2%). SSB in high-risk groups (including people with sickle cell disease), fever in returning travellers, and suspected endocarditis were treated appropriately with empiric ceftriaxone 100% of the time.

Figure 1:

Distribution of appropriateness ratings by clinical indication: Duration of therapy (A) 72 hours or less and (B) more than 72 hours

CNS = Central nervous system

Overall, 103 (37.3%) of the 276 empiric courses had a definite indication, 53 (19.2%) had a possible indication, and 120 (43.5%) were deemed to have no indication. Of the 120 patients who had no indication for empiric ceftriaxone, the three most common reasons for prescribing it were RTIs (51; 42.5%), head and neck infections (18; 15.0%), and intra-abdominal infections (15; 12.5%) (Figure 1a). For 19 courses (15.8%), there was no apparent clinical rationale for the antimicrobial prescription. The most common reason for inappropriateness (n = 100; 83.3%) was related to the use of a spectrum deemed too broad according to indication or culture result for the clinical syndrome.

In total, 39 (14.1%) of courses were given for more than 72 hours, with a mean number of 7 days. Of these, 14 (35.9%) met criteria for appropriate indication (Figure 1b). Seven (17.9%) patients had underlying malignancies or sickle cell disease, with the remainder being previously healthy, and 10 (25.6%) had isolates recovered. Only 5 patients (12.8%) had a further definitive or possible indication to continue ceftriaxone (1 had Streptococcus pyogenes in blood and cerebrospinal fluid and was treated for 5 days with ceftriaxone [considered possible indication] and was then changed to penicillin G, 1 had Klebsiella oxytoca in blood, 2 had beta-lactamase–producing H. influenzae [1 in in endotracheal secretions, 1 in cerebrospinal fluid plus endotracheal secretions], and 1 had Moraxella catarrhalis in endotracheal secretions). Five others had the following isolates: Streptococcus viridans and Prevotella in a pleural effusion, methicillin-resistant Staphylococcus aureus in a wound, complicated pneumonia with Fusobacterium, penicillin-susceptible S. pneumoniae in a pleural fluid, and Streptococcus anginosus associated with a perforated appendicitis. The remainder did not have documented bacterial pathogens and were being treated for culture-negative infectious syndromes such as pneumonia, orbital cellulitis, appendicitis, fever associated with seizure, or abdominal trauma with bowel perforation.

Discussion

This DUE for ceftriaxone demonstrates that two-thirds of the courses were used in children who had no underlying medical conditions, and just more than half of the 276 courses met the criteria for definitive or possible appropriate empiric use according to guidelines. Duration of therapy was generally short, with only 14.1% of courses given for 72 hours or more. This DUE for ceftriaxone identified patterns of inappropriate use that were not related to days of therapy and thus was an important quality metric (6,8,9).

Ceftriaxone is perceived as a low-risk solution to management of suspected infection because it has few immediate side effects. Prescribers often do not appreciate the ensuing risk of resistance for both individuals and society, although it has long been recognized (4,10–14). Quality metrics for antimicrobials include not only duration but also selection of antimicrobials and avoidance of overly broad-spectrum antimicrobials in particular clinical circumstances. In this study, appropriateness grading was based on published principles of stewardship, specific properties of parenteral third-generation cephalosporins, and local and international guidelines because adherence to local or facility guidelines is recommended when assessing stewardship initiatives (7,15–17).

A prior ceftriaxone DUE in adults reported 65% appropriateness of use for more than 5 days (18). In another study at a community hospital, ceftriaxone was continued in 68% of patients despite negative cultures (19). Our estimate of appropriate or possibly appropriate empiric therapy of just more than 56% was likely lower because of stricter local and national pediatric guidelines. A recent meta-analysis estimated the rate of general inappropriate antimicrobial use at 30% but found considerable variability, and no official benchmarks for appropriate ceftriaxone prescribing exist (20). In our study, the majority (237; 85.9%) of empiric therapy courses were discontinued within 72 hours, but 39 (14.1%) were extended beyond 72 hours, of whom just more than one-third met criteria for continuing therapy. This suggests that the use of a third-generation cephalosporin should be carefully considered not only at the start of therapy but also at 48–72 hours.

Current guidelines for pediatric community-acquired pneumonia recommend use of primarily ampicillin unless the child is unstable and requires intensive care because the likelihood of a penicillin-resistant pathogen is low, even with empyema, and ceftriaxone is not optimal for S. aureus (21,22). Use of penicillin, ampicillin, or amoxicillin with or without empyema is also recommended in European and US pediatric guidelines (23,24). This DUE revealed that the majority of inappropriate use occurs in empiric treatment of RTIs (27% were deemed appropriate or possibly appropriate). In this study, head and neck infections (complicated sinusitis or mastoiditis) were also a common reason for inappropriate empiric use. A large study comparing outcomes of 30,000 children treated for otitis, sinusitis, and pharyngitis concluded that broad-spectrum antimicrobials were not associated with better clinical outcomes, albeit in an outpatient setting (25). Guidelines for management of intra-abdominal infections have been published but are not proscriptive for children, with choices based on severity of infection and prevalence of resistance (26). Surgeons at the Children’s Hospital of Eastern Ontario have principally used ampicillin, aminoglycoside, and metronidazole as empiric therapy on the basis of local recommendations, with modifications only if resistant pathogens are detected and the patient is not responding to source control (27). Guidelines based on best practices are needed to assess what proportion of broad-spectrum empiric therapy is acceptable in health care institutions (28).

Limitations

Limitations of this study include its retrospective and monocentric design. The assessment of antibiotic appropriateness is influenced by local guidelines, which may not be applicable to other types of health care settings or adult patients. The study is a snapshot of a 6-month period and therefore reduces the potential to record any seasonality. However, its single-centre nature provides a solid platform for future multicentre prospective studies that can replicate our methods. The classification scheme could facilitate future efforts to comprehensively measure ceftriaxone appropriateness.

Conclusion

In summary, ceftriaxone DUE was useful and revealed that although 85% of empiric courses were discontinued at 72 hours or less, approximately one in two empiric courses did not meet adherence to current Canadian and local guidelines. Further efforts to decrease empiric starts of ceftriaxone could potentially reduce courses by 50% in this pediatric setting. More data on appropriateness for empiric and longer courses of ceftriaxone would inform benchmarking criteria.

Ethics Approval:

The study protocol was approved by an ethics committee and the ethics certificate information is available from the authors upon request.

Funding:

No funding was received for this work.

Disclosures:

The authors have nothing to disclose.

Peer Review:

This manuscript has been peer reviewed.

Animal Studies:

N/A.