Abstract
Background
The preferred ambulatory IV therapy for cellulitis is often once-daily cefazolin combined with once-daily oral probenecid (C+P). However, due to a national probenecid drug shortage in 2011, our centre developed a replacement protocol for the administration of cefazolin continuous infusion (CCI) using elastomeric infusers. Our goal was to compare treatment efficacy, duration of IV therapy, and recurrence associated with CCI and C+P using retrospective data from our centre.
Methods
We conducted a non-inferiority single-centre retrospective cohort study of emergency department medical records. Patients received either C+P (cefazolin 2 g IV once daily plus probenecid 1 g PO once daily) or CCI (cefazolin 2 g IV loading dose, followed by cefazolin 6 g IV via continuous infusion over 24 hours, via an elastomeric infuser). We compared treatment efficacy, duration of IV therapy, and recurrence rates.
Results
total of 203 patients were analyzed, with 107 included in the CCI arm and 96 in the C+P arm. Overall, CCI users and C+P users were comparable in their sociodemographic and clinical variables measured at admission. We observed increased odds of achieving successful treatment among the CCI group, however it did not reach statistical significance (odds ratio [OR] 2.25; 95% CI 0.84 to 6.07). Recurrence rates were similar between both groups (OR 1.91; 95% CI 0.32 to 11.31). The average duration of IV therapy was similar between groups (p = 0.6).
Conclusions
ith results suggesting that CCI was non-inferior to C+P, and that both approaches required similar treatment durations, CCI could represent an acceptable alternative to C+P for the ambulatory IV treatment of cellulitis.
Key words: ambulatory, cefazolin, cellulitis, continuous infusion, probenecid
Abstract
Historique
Le traitement intraveineux (IV) ambulatoire favorisé pour la cellulite prend souvent la forme d’une injection quotidienne de céfazoline associée à une dose quotidienne de probénécide par voie orale (PO) (C+P). Cependant, à cause d’une pénurie nationale de probénécide en 2011, le centre des auteurs a mis au point un autre protocole de perfusion continue de céfazoline (PCC) à l’aide de perfuseurs élastomériques. Les auteurs ont comparé l’efficacité du traitement, la durée du traitement IV et la récurrence associée à la PCC et aux C+P à partir des données rétrospectives du centre.
Méthodologie
Les auteurs ont réalisé une étude de cohorte rétrospective monocentrique de non-infériorité à partir des dossiers médicaux du département d’urgence. Les patients ont reçu soit des C+P (2 g de céfazoline IV une fois par jour associés à 1 g de probénécide PO une fois par jour), soit une PCC (dose de charge de 2 g de céfazoline IV, suivie d’une perfusion continue de 6 g de céfazoline IV par perfuseur élastomérique sur une période de 24 heures). Ils ont comparé l’efficacité thérapeutique, la durée du traitement IV et le taux de récurrence.
Résultats
Au total, les auteurs ont analysé 203 patients, dont 107 faisaient partie du volet de PCC et 96, du volet de C+P. Dans l’ensemble, les utilisateurs de la PCC et des C+P présentaient des variables sociodémographiques et cliniques semblables à l’admission. Les auteurs ont observé que le traitement du groupe sous PCC avait tendance à être plus efficace, mais cette constatation n’était pas statistiquement significative (rapport de cotes [RC] 2,25; IC à 95 %, 0,84 à 6,07). Les taux de récurrence étaient semblables entre les deux groupes (RC 1,91; IC à 95 %, 0,32 à 11,31). La durée moyenne du traitement IV était semblable entre les groupes (p = 0,6).
Conclusions
Puisque, selon les résultats, la PCC n’était pas inférieure aux C+P et que la durée du traitement était semblable dans les deux cas, il serait acceptable comme traitement de rechange acceptable de C+P pour le traitement IV ambulatoire de la cellulite.
Mots-clés : ambulatoire, céfazoline, cellulite, perfusion continue, probénécide
Cellulitis is a frequent cause of ambulatory visits; in 2005, there were approximately 9.6 million ambulatory visits for cellulitis in the United States (1). While the majority of mild-to-moderate cases of uncomplicated cellulitis can be successfully managed with oral antibiotics, moderate-to-severe cases will often require IV therapy (2–4). Current practice in many establishments for the ambulatory IV treatment of cellulitis is to use once-daily cefazolin combined with once-daily oral probenecid (C+P). However, due to a Canada-wide probenecid drug shortage in 2011, our centre was faced with a need to find a replacement for C+P. While many centres simply switched from C+P to ceftriaxone, concerns were raised by local stakeholders that this may result in increased local resistance to third-generation cephalosporins and could increase Clostridioides difficile rates. After evaluating potential options, it was decided to develop a protocol for the administration of cefazolin continuous infusion (CCI) using elastomeric infusers. Due to an absence of recommended continuous infusion dosing regimens, the CCI dosing regimen (2 g IV loading dose followed by 6 g IV via continuous infusion over 24 hours) was chosen using the cumulative daily dose and infusing over 24 hours. CCI has been shown to achieve adequate serum concentrations and has also been successfully used to treat bone and joint infections (5–7). Despite this, there are no clinical trials evaluating CCI for the ambulatory treatment of uncomplicated cellulitis. In this retrospective study comparing treatment efficacy, duration of IV therapy, and recurrence associated with CCI and C+P using retrospective data from our centre, we hypothesized that CCI would be non-inferior to C+P for the ambulatory treatment of uncomplicated cellulitis.
Methods
We conducted a non-inferiority single-centre retrospective cohort study. The cohort was constructed using data from the reviewed charts of ambulatory patients presenting to our centre’s emergency department (ED) between 2008 and 2012. Our centre, the Dr. Georges-L.-Dumont University Hospital Centre, is a 342-bed tertiary care hospital located in the largest urban centre in the province of New Brunswick. Charts were identified using the hospital pharmacy’s computer records, and were reviewed in ascending (for CCI) and descending (for C+P) chronological order starting in 2011 (the year when C+P was replaced by CCI). Chart review was undertaken by a pharmacy student using a standardized data collection worksheet.
The cohort inclusion criteria were as follows: 1) adults 18 years of age or older, 2) presenting with uncomplicated cellulitis requiring IV therapy (as determined by the ED physician), 3) candidates for ambulatory therapy (as determined by the ED physician), and 4) receiving either C+P (cefazolin 2 g IV once daily + probenecid 1 g PO once daily) or CCI (cefazolin 2 g IV loading dose, followed by cefazolin 6 g IV via continuous infusion over 24 hours, via an elastomeric infuser). Exclusion criteria were: 1) cellulitis associated with bite wounds, undrained abscesses, bone/joint infections, or bursitis, 2) history of methicillin-resistant Staphylococcus aureus (MRSA) infection or colonization prior to initiating therapy, 3) known bacteremia prior to initiating therapy, and 4) positive culture of a cefazolin-resistant pathogen prior to initiating therapy.
Exposure to CCI or C+P was determined through a search among medications administered to patients during their ambulatory visit and recorded in the electronic pharmacy records. The primary outcomes were treatment efficacy and recurrence. Treatment efficacy was defined as an improvement of signs and symptoms of the infection (as judged by the ED physician) permitting either discontinuation of antibiotics, or oral step-down (8). Recurrence was defined as an admission or ED visit for cellulitis within 30 days of the end of an effective treatment. The secondary outcome studied was the treatment duration, defined as the number of days of IV therapy. Total treatment duration (i.e., days of IV therapy plus days of oral therapy) was not measured, due to lack of data on the duration of oral therapy.
The following variables were considered as potential confounders based upon literature evaluation and the authors’ opinions on their clinical relevance: age (years), sex, diabetes (yes/no), hypertension (yes/no), fever (yes/no), previous antibiotic use (yes/no), and infection site (upper limbs, lower limbs, head-neck, or other). Other potential clinical variables were not included due to inconsistent documentation practices in the ED. There were no renal dysfunction dosing adjustments in the local protocols for CCI or C+P; therefore, patients with significant renal dysfunction were not assessed. Descriptive statistics for the characteristics of patients were calculated and compared between groups of CCI and C+P users, with independent t-tests and Chi-square tests applied to test the difference for continuous and discrete variables, respectively. The primary outcomes were analyzed in crude models using logistic regression, and subsequently in adjusted logistic regression models by adding the potentially confounding variables. We reported the crude and adjusted odds ratios (OR) with 95% confidence intervals (CIs). Poisson regression was used to model the treatment duration outcome to determine crude and adjusted differences between the groups. Results from adjusted models represent fully adjusted models for all variables (i.e., age, sex, fever, diabetes, hypertension, previous antibiotic use, and infection site).
In the pre-hoc sample size calculation using a type I error of 0.05 and 90% power, a sample size of 98 patients per group (196 patients in total) was estimated to be sufficient to detect an established maximal difference of 15% between the treatment success rates from the patient groups compared (8,9). Statistical analyses were conducted using SAS software, version 9.4 (SAS Institute Inc., Cary, NC)
This study was approved by the Research Ethics Committee of Vitalité Health Network. Consent to participate was waived, as only retrospective and de-identified data were used.
Results
In total, 226 charts of ambulatory care patients between May 2008 and September 2012 were reviewed. We excluded 23 patients based on our pre-specified inclusion/exclusion criteria. The final sample analyzed included 203 patients, with 107 included in the CCI arm and 96 in the C+P arm. Most patients (55.8%) were male, with an average age of 50.8 years. The most common site of infection in both groups was either upper or lower limb, accounting for over 75% of patients. Overall, CCI users and C+P users were comparable in their sociodemographic and clinical variables measured at admission (Table 1). Tests of differences between groups suggested balance between groups in all variables (p ≥0.05). Only one patient from the CCI arm of the study was lost to follow-up, and was excluded from the analysis of treatment success.
Table 1:
Characteristics of patients included in the analyses comparing continuous cefazolin IV users versus cefazolin plus probenecid users
| Sociodemographic and clinical variables measured at admission | Total N available | CCI users (n = 107) No. (%) or mean (SD) | C+P users (n = 96) No. (%) or mean (SD) | p-value |
|---|---|---|---|---|
| Age (years) | 203 | 52.9 ± 17.0 | 48.7 ± 15.1 | 0.066 |
| Sex (male) | 203 | 58 (54.2) | 55 (57.3) | 0.659 |
| Fever | 197 | 9 (8.7) | 9 (9.7) | 0.803 |
| Diabetes | 203 | 7 (6.5) | 8 (8.3) | 0.626 |
| Hypertension | 201 | 35 (33.0) | 22 (23.2) | 0.122 |
| Previous antibiotic use | 203 | 30 (28.0) | 29 (30.2) | 0.734 |
| Infection site | 203 | 0.341 | ||
| Upper limbs | 34 (31.8) | 41 (42.7) | ||
| Lower limbs | 48 (44.8) | 40 (41.7) | ||
| Head or neck | 14 (13.1) | 9 (9.4) | ||
| Other | 11 (10.3) | 6 (6.2) |
CCI = continuous cefazolin infusion; C+P = cefazolin plus probenecid
Primary and secondary endpoints are summarized in Table 2. In crude and adjusted models, an increased odds of achieving successful treatment was observed among the CCI group, however it did not reach statistical significance (Table 2). Recurrence rates were similar between both groups in crude and adjusted models. The average duration of IV therapy was similar between groups; however, most of the patients receiving CCI were treated for 2 days, compared with 3 days for C+P (Table 2). In the adjusted Poisson regression model, the estimated difference in success rate favoured CCI compared with C+P, but did not reach statistical significance (p = 0.634). Rates of documented adverse events (only rash and nausea were reported) were similar; 4/107 (3.7%) in the CCI group and 4/96 (4.2%) in the C+P group.
Table 2:
Summary of primary and secondary endpoints and duration of treatment
| (CCI) | (C+P) | Crude OR (95% CI) | Adjusted OR (95% CI)* | |
|---|---|---|---|---|
| Number of patients | 106 | 96 | – | – |
| Treatment success (%) | 98 (92.5%) | 82 (85.4%) | 2.09 (0.84, 5.23) | 2.25 (0.84, 6.07) |
| Recurrence (%) | 4 (3.7%) | 2 (2.1%) | 1.83 (0.33, 10.20) | 1.91 (0.32, 11.31) |
| Treatment duration (days) | ||||
| Mean (SD) | 3.6 ± 2.7 | 3.6 ± 1.4 | – | – |
| Mode | 2 | 3 | – | – |
* Models adjusted for age, sex, fever, diabetes, hypertension, previous antibiotic use, and infection site
CCI = continuous cefazolin infusion; C+P = cefazolin plus probenecid; OR = odds ratio
Interpretation
In this study of ambulatory patients with uncomplicated cellulitis, we observed that the use of CCI was non-inferior to C+P for the ambulatory IV treatment of cellulitis. Moreover, the duration of treatment was comparable between the groups, with most of the patients on CCI achieving treatment success on the second day of treatment compared with the third day on C + P.
Both C+P and CCI patients presented daily at the ED to receive their treatment (C+P would receive a single dose of each and return home, while CCI patients were sent home with a new infuser each day). Therefore, the elastomeric infusers that were used for CCI in our centre permitted us to continue offering outpatient IV antimicrobial therapy to our ambulatory cellulitis patients during a probenecid backorder, without requiring a switch to ceftriaxone. However, the increased cost of CCI compared with C+P potentially limits its widespread use. In theory, continuous infusion maximizes pharmacodynamic properties of cefazolin, and could be expected to potentially outperform C+P (10,11).
Strengths of our study include an adequate sample size, adjustments for potential confounders, only one of the cases lost to follow-up, and the fact that this study reflects real-world practice.
The results should nevertheless be interpreted with consideration of the following limitations. Due to the retrospective nature of our study, documentation was not standardized; resulting in missing demographic data for some patients. We relied on the ED physician’s diagnosis and evaluation to determine if the patient’s cellulitis was severe enough to require IV therapy, but not too severe to prevent ambulatory IV therapy. We were only able to evaluate for adverse events if they were documented by the ED physician; which potentially underestimated occurrence of adverse events. Given the relatively low rates of recurrence in both groups, CIs were quite wide; future studies in larger samples are recommended.
Conclusions
To conclude, our study suggests that the use of CCI (cefazolin 6 g/24 hours via continuous infusion) was non-inferior to the more common regimen of C+P (cefazolin 2 g IV once daily + probenecid 1 g PO once daily) for the ambulatory treatment of cellulitis. Further studies are warranted to determine CCI’s place in therapy. Supplementary data used to support the findings of this study are available from the corresponding author upon request.
Acknowledgements:
The authors would like to acknowledge Mrs. Anissa Belkaid for her assistance in preparing the initial research proposal, and Vitalité Health Network’s Office of Research Services for their support.
Funding Statement
This work was supported by Baxter (research grant number 17G1211).
Competing Interests:
DLL reports funding to cover travel expenses from Baxter outside the submitted work. LPJ reports travel grants from Baxter Canada outside the submitted work. SE, GG, JC, and MB have nothing to disclose
Ethics Approval:
This study was approved by the Research Ethics Committee of Vitalite Health Network. Consent to participate was waived, as only retrospective and de-identified data were used.
Informed Consent:
N/A
Registry and the Registration No. of the Study/Trial:
N/A
Animal Studies:
N/A
Funding:
This work was supported by Baxter (research grant number 17G1211).
Peer Review:
This article has been peer reviewed.
References
- 1. Hersh AL, Chambers HF, Maselli JH, Gonzales R. National trends in ambulatory visits and antibiotic prescribing for skin and soft-tissue infections. Arch Intern Med. 2008;168(14):1585–91. 10.1001/archinte.168.14.1585. Medline: [DOI] [PubMed] [Google Scholar]
- 2. Raff AB, Kroshinsky D. Cellulitis: a review. JAMA. 2016;316(3):325–37. 10.1001/jama.2016.8825. Medline: [DOI] [PubMed] [Google Scholar]
- 3. NB Provincial Health Authorities Anti-Infective Stewardship Committee. Treatment of cellulitis/skin infection—September 2017 [Internet]. Bathurst, NB: Vitalité Health Network; [2017. Sep; cited 2018 Mar 1]. Available from: http://www.vitalitenb.ca/en/doctors/treatment-guidelines/antimicrobial-treatment-guidelines-common-infections-0. [Google Scholar]
- 4. Stevens DL, Bisno AL, Chambers HF, et al. Practice guidelines for the diagnosis and management of skin and soft tissue infections: 2014 update by the Infectious Diseases Society of America. Clin Infect Dis. 2014;59(2):147–59. 10.1093/cid/ciu444. Medline: [DOI] [PubMed] [Google Scholar]
- 5. Thys JP, Vanderkelen B, Klastersky J. Pharmacological study of cefazolin during intermittent and continuous infusion: a crossover investigation in humans. Antimicrob Agents Chemother. 1976;10(3):395–8. 10.1128/aac.10.3.395. Medline: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Howard GW, Begg EJ, Chambers ST, Vella Brincat J, Zhang M, Kirkpatrick CM. Free and total cefazolin plasma and interstitial fluid concentrations at steady state during continuous infusion. J Antimicrob Chemother. 2002;50(3):429–32. 10.1093/jac/dkf129. Medline: [DOI] [PubMed] [Google Scholar]
- 7. Zeller V, Durand F, Kitzis MD, et al. Continuous cefazolin infusion to treat bone and joint infections: clinical efficacy, feasibility, safety, and serum and bone concentrations. Antimicrob Agents Chemother. 2009;53(3):883–7. 10.1128/AAC.00389-08. Medline: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Grayson ML, McDonald M, Gibson K, et al. Once-daily intravenous cefazolin plus oral probenecid is equivalent to once-daily intravenous ceftriaxone plus oral placebo for the treatment of moderate-to-severe cellulitis in adults. Clin Infect Dis. 2002;34(11):1440–8. 10.1086/340056. Medline: [DOI] [PubMed] [Google Scholar]
- 9. Spellberg B, Talbot GH, Boucher HW, et al; Antimicrobial Availability Task Force of the Infectious Diseases Society of America. Antimicrobial agents for complicated skin and skin-structure infections: justification of noninferiority margins in the absence of placebo-controlled trials. Clin Infect Dis. 2009;49(3):383–91. 10.1086/600296. Medline: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Abdul-Aziz MH, Staatz CE, Kirkpatrick CMJ, Lipman J, Roberts JA. Continuous infusion vs. bolus dosing: implications for beta-lactam antibiotics. Minerva Anestesiol. 2012;78(1):94–104. Medline: [PubMed] [Google Scholar]
- 11.Roberts JA, Abdul-Aziz MH, Davis JS, et al. Continuous versus intermittent β-lactam infusion in severe sepsis a meta-analysis of individual patient data from randomized trials. Am J Respir Crit Care Med. 2016;194(6):681–91. 10.1164/rccm.201601-0024OC. Medline: [DOI] [PubMed] [Google Scholar]