Cefotaxime has traditionally been the preferred third-generation cephalosporin in the neonatal intensive care unit (NICU) when additional Gram-negative activity is needed for suspected meningitis or when relative contraindications for aminoglycosides exist.1,2 However, a shortage of cefotaxime occurred in the United States in February 2015 because of discontinuation by 2 manufacturers.3 As a result, neonatologists were forced to identify alternative agents. The American Academy of Pediatrics noted that ceftriaxone is not ideal for infants <2 months because of the potential for bilirubin displacement and risk for precipitate formation with concomitant use of calcium-containing intravenous fluids. Therefore, the American Academy of Pediatrics released a statement recommending ceftazidime as an appropriate alternative.4 Ceftazidime has similar antimicrobial activity to cefotaxime, with the additional activity against Pseudomonas species and Stenotrophomonas species.
The impact of ceftazidime use in place of cefotaxime in the NICU is unknown. Traditionally, ceftazidime use has been reserved in the NICU for patients with suspected Pseudomonas aeruginosa or Stenotrophomonas maltophilia infections. However, with the cefotaxime shortage, it is used more widely for early- and late-onset sepsis in those with suspected meningitis or impaired renal function. Previous reports have shown that empiric use of third-generation cephalosporins in place of aminoglycosides in the NICU has been implicated in emergence of antimicrobial resistance, including extended-spectrum beta-lactamase (ESBL) producing organisms.5 There are some concerns that induction of resistance with ceftazidime is more likely than cefotaxime because of the slow penetration into the periplasmic space of the outer membrane of Gram-negative bacteria.6,7 However, it is unknown if substitution of ceftazidime will result in changes to susceptibility patterns within the NICU. The purpose of this study was to assess the effect of the nationwide shortage of cefotaxime and subsequent increase in use of ceftazidime on susceptibilities of Gram-negative organisms in the NICU.
This retrospective, Institutional Review Board approved, study was conducted over a 5-year period (January 1, 2013 to December 31, 2017) in a single-center level IV NICU. The days of therapy per 1000 patient days were calculated for cefotaxime and ceftazidime on a monthly basis during the study period. This information was used to determine the pre- (January 1, 2013 to July 31, 2016) and postshortage (January 8, 2016 to December 31, 2017) periods for comparison. These time periods were selected because intermittent use of cefotaxime occurred at the institution between February 2015 and July 2016. All cultures from NICU patients with growth of Gram-negative organisms during this time period were included for initial review. Cultures were considered for inclusion if they contained the organisms of interest: Escherichia coli, P. aeruginosa, Klebsiella species, Enterobacter species, Serratia species and S. maltophilia. Cultures were excluded if minimum inhibitory concentration (MIC) values were not provided or if they were cultures from the eye or unspecified location. Duplicate isolates, defined as the same organism isolated from the same patient and location within a 90-day period, were excluded from primary analysis.
Objectives and Data Collection
The reported MIC values were used to determine if an isolate was resistant, intermediate or susceptible using the Clinical and Laboratory Standards Institute breakpoints.8,9 Of note, the MIC breakpoints for Enterobacteriaceae changed at the institution on September 15, 2014; however, the reported MICs for those organisms isolated after that time were still assessed using the old MIC breakpoints to maintain consistency in evaluation. The susceptibility rate to third- and fourth-generation cephalosporins for the organisms of interest were compared between study periods. In addition, the rate of ESBL producers confirmed by phenotype testing for E. coli and Klebsiella species were also compared. Last, data from the institution’s database regarding number of admissions/year, mean daily NICU census, mean length of stay, mean birthweight, distribution of gestational age and mortality rate were collected on an annual basis to provide a surrogate marker for acuity during the study period.
The primary objective was to compare the third- and fourth-generation cephalosporin susceptibility rates for selected Gram-negative organisms pre and post cefotaxime shortage. Other objectives included a comparison of pathogens and culture type and rates of ESBL-positive isolates between study periods.
Categorical data were summarized in contingency tables as frequency (percent) with risk ratios (RRs) (95% CIs), and group comparisons were made using χ2 or Fisher’s exact tests. Continuous variables were summarized using medians (interquartile range), and group comparisons were made using Wilcoxon tests. SAS software 9.4 (SAS Institute Inc., Cary, NC) was used for all analyses with an alpha of 0.05.
Overall, the admissions/year, mean census/day, mean birthweight, mean length of stay were consistent across the study period. The average number of admissions per year ranged from 1194 to 1295 patients, with a mean census of 70.1–81.0 patients per day across the study period. The overall hospital mortality rate ranged from 4.3% to 6.9% during the study period. The median (interquartile range) days of therapy/1000 patient days of ceftazidime increased significantly from the pre- to postcefotaxime shortage period, 18.6 (5.6–33.1) versus 52.9 (20.9–63.1), P < 0.005.
A total of 1,244 isolates of Gram-negative organisms of interest were identified. However, 37 (3.0%) were excluded because they were eye cultures or from an unspecified device. An additional 541 (43.5%) were excluded from the primary analysis because they were duplicate isolates that occurred within 90 days. The remaining 666 isolates were analyzed. Of these, 12 (1.8%) were duplicates that occurred after a 90-day period. The preshortage period included 464 (69.7%) isolates and the postshortage included 202 (30.3%). A majority of isolates were identified in tracheal aspirate (47.8 vs. 44.6%) or urine (35.3 vs. 36.6%) cultures, while the remainder were from blood (7.5 vs. 9.9%), abdominal fluid (5.0 vs. 2.5%), wound (3.9 vs. 4.5%), cerebrospinal fluid (0.4 vs. 1.0%) or pleural fluid (0 vs. 1.0%) cultures, in the pre- and postshortage period, respectively. No difference was found in isolate distributions across study periods, P = 0.1761.
Table 1 includes a breakdown of the number of isolates in the pre- and postshortage periods. No significant differences in susceptibility rates between study periods were noted (Table 1). There was no difference in the rates of ESBL-positive E. coli strains in the pre- and postshortage periods [4/104 (3.8%) vs. 2/41 (4.9%), P = 1.000]. No ESBL-positive Klebsiella species were recovered.
Because of the conservative and arbitrary definition of duplicate isolate for the study, a post-hoc analysis was performed on the 386 preshortage and 155 postshortage duplicate isolates that were excluded from the primary analysis. A majority of the isolates were identified in tracheal aspirate (89.4 vs. 86.5%) or urine (8.0 versus 9.7%) cultures, in the pre- and postshortage period, respectively.
In this analysis of duplicate isolates, the susceptibility rate of P. aeruginosa for ceftazidime decreased from 86% to 50% in the pre- to postshortage period (RR 0.58; 95% CI: 0.43–0.79). Cochran-Mantel-Haenszel statistics were used to produce general associations between period and susceptibility after controlling for initial or repeat culture. Using all initial and repeat isolates, the overall susceptibility decreased from 89% to 75% and the Mantel-Haenszel common RR was 0.74 (95% CI: 0.62–0.88). In addition, the susceptibility rate of P. aeruginosa for cefepime decreased from 90% to 60% (RR 0.66, 95% CI: 0.51–0.86). Using all isolates, the overall susceptibility decreased from 92% to 75% and the Mantel-Haenszel common RR was 0.81 (95% CI: 0.70–0.93).
This is the first study to evaluate the impact of the cefotaxime shortage, and subsequent increased use of ceftazidime, on susceptibility rates of Gram-negative organisms in the NICU. Overall, when excluding duplicate isolates, no significant differences were noted in the susceptibility rates for third- and fourth-generation cephalosporins after ceftazidime usage increased by nearly 3-fold in the 1.5 year postshortage period. In addition, no increase in rates of ESBL-producing bacteria were observed. P. aeruginosa and S. maltophilia, 2 organisms for which ceftazidime is frequently used in the NICU, demonstrated a nonsignificant decrease in susceptibility in the postshortage period when excluding duplicate isolates. However, the post-hoc analysis including duplicate isolates demonstrated a statistically significant decrease in susceptibility rates for P. aeruginosa for ceftazidime and cefepime. This particular finding may indicate inducible resistance with ceftazidime use. Close monitoring of P. aeruginosa and S. maltophilia susceptibility patterns are warranted as ceftazidime use continues.
Limitations of this study include a single institution, short postshortage period and inability to control for all confounding factors. Inclusion of a single NICU may limit external validity because of differences in acuity, prescribing patterns and baseline susceptibility rates. Of note, at our institution, cefepime is a restricted antibiotic that requires Infectious Diseases service approval; therefore, use during the study period was minimal. In addition, the postshortage period was shorter than initially planned because of intermittent use of cefotaxime between February 2015 and July 2016 when small backorder shipments were received. A 1.5 year postshortage period is not long enough to determine changes in susceptibility rates. Resolution of the cefotaxime shortage does not seem imminent, so we plan to continue monitoring and follow-up with additional long-term data. Last, general institution demographic data and mortality rates were used as surrogate markers for acuity. There may be other factors contributing to acuity that were not considered when analyzing the results.
This study suggests that short-term substitution of ceftazidime in the NICU does not create major changes in the susceptibility rates to third-generation cephalosporins for E. coli, Klebsiella species, Enterobacter species, Serratia species, and S. maltophilia. However, the post-hoc analysis including duplicate isolates of P. aeruginosa for ceftazidime and cefepime are concerning and may indicate inducible resistance. Longer term monitoring is needed, specifically for P. aeruginosa and S. maltophilia.
1. Polin RA; Committee on Fetus and Newborn. Management of neonates with suspected or proven early-onset bacterial sepsis. Pediatrics. 2012;129:1006–1015.
2. Shane AL, Stoll BJ. Recent developments and current issues in the epidemiology, diagnosis, and management of bacterial and fungal neonatal sepsis. Am J Perinatol. 2013;30:131–141.
5. Banerjee T, Bhattacharjee A, Upadhyay S, et al. Long-term outbreak of Klebsiella pneumoniae
& third generation cephalosporin use in a neonatal intensive care unit in north India. Indian J Med Res. 2016;144:622–629.
6. Lister PD, Sanders WE Jr, Sanders CC. Cefepime-aztreonam: a unique double beta-lactam combination for Pseudomonas aeruginosa
. Antimicrob Agents Chemother. 1998;42:1610–1619.
8. Clinical Laboratory Standards Institute. Performance Standard For Antimicrobial Susceptibility Testing: 20th Information Supplement, CLSI Document M100-S20. 2010.Wayne, PA: Clinical and Laboratory Standards Institute.
9. Clinical Laboratory Standards Institute. Performance Standard For Antimicrobial Susceptibility Testing: 24th Information Supplement, CLSI Document M100-S24. 2014.Wayne, PA: Clinical and Laboratory Standards Institute.