In community hospitals, antimicrobial stewardship team notification of rapid diagnostic testing (RDT) results may not be feasible. A retrospective quasi-experimental study was conducted evaluating 252 adult inpatients with blood cultures positive for Gram-positive cocci in clusters (pre-RDT, n = 143; post-RDT, n = 109).
KEYWORDS: antimicrobial stewardship, bloodstream infections, coagulase-negative staphylococci, rapid diagnostic testing
ABSTRACT
In community hospitals, antimicrobial stewardship team notification of rapid diagnostic testing (RDT) results may not be feasible. A retrospective quasi-experimental study was conducted evaluating 252 adult inpatients with blood cultures positive for Gram-positive cocci in clusters (pre-RDT, n = 143; post-RDT, n = 109). The median time to appropriate therapy was significantly shorter in the post-RDT group (15 versus 0 h, P < 0.001), and the mean length of stay for patients with coagulase-negative staphylococcus was significantly shorter (10.5 versus 7.7 days; P = 0.015).
TEXT
Decreased time to appropriate antibiotic therapy has been demonstrated for bloodstream infections when rapid diagnostic testing (RDT) implementation is coupled with antimicrobial stewardship team (AST) intervention (1–5). However, in community hospitals with limited resources, real-time AST notification of positive blood cultures may not be feasible. The purpose of this study was to determine the impact of GeneXpert MRSA/SA BC test (Cepheid, Sunnyvale, CA) implementation without real-time AST notification on the time to appropriate antibiotic therapy for blood cultures growing Gram-positive cocci (GPC) in clusters in a community hospital.
This retrospective, quasi-experimental study was conducted at a 350-bed community teaching hospital with an on-site microbiological laboratory. The study was approved by the hospital's institutional review board with a waiver of informed consent. Adult patients (≥18 years of age) with a monomicrobial positive blood culture reported as GPC in clusters between 1 January 2016 and 30 June 2016 (pre-RDT group) and 1 January 2017 and 30 June 2017 (post-RDT group) were eligible for inclusion. Patients were excluded from the study if they had blood cultures drawn at another institution, had Staphylococcus aureus bacteremia (SAB) within the previous 90 days, were discharged prior to Gram stain results, were HIV positive, or were transferred, expired, or transitioned to comfort care within 48 h of admission.
The AST at our institution is co-led by a 1.0 full-time equivalent (FTE) infectious disease-trained pharmacist and a 0.1 FTE infectious disease (ID) physician. There are no evening or weekend AST services and no clinical decision-making support software available to assist in positive culture identification. The ID physician consult service is available daily, including evening and weekend hours, and staffed by a single ID physician. The ID consultants are not alerted to positive blood cultures by the microbiology staff. The AST methods did not differ between the pre-RDT and post-RDT groups.
In the pre-RDT and post-RDT groups, blood culture bottles were incubated using the BacT/Alert blood culture system (bioMérieux, Inc., Durham, NC), with Gram staining performed using the Previ Color Gram automated Gram stain (bioMérieux, Inc.) as soon as growth was detected. Positive blood cultures were plated onto standard media and processed using standard microbiology techniques according to the Clinical and Laboratory Standards Institute. Final organism identification and in vitro antimicrobial susceptibilities were confirmed using the Vitek 2 system (bioMérieux, Inc.). In total, this process takes approximately 72 h from blood culture draw to final organism identification and susceptibility results.
In the pre-RDT group, the Staphaurex (ThermoFisher Scientific, Lenexa, KS) rapid latex agglutination test was utilized on-demand to determine coagulase-negative staphylococci (CoNS) versus SAB, and the penicillin-binding protein 2a (PBP2a) test (Alere, Inc., Waltham, MA) was utilized to determine methicillin-resistant Staphylococcus aureus (MRSA) and methicillin-susceptible Staphylococcus aureus (MSSA) strains prior to the final automated susceptibility results. The use of these rapid chemical tests shortened the time to organism identification by approximately 24 h. In the post-RDT group, these two biochemical tests were replaced with the Xpert MRSA/SA BC test, which was implemented in October 2016. The microbiology report in the electronic medical record lists a two-line result as follows: (i) S. aureus detected (yes/no) and (ii) methicillin-resistance detected (yes/no). Prior to RDT implementation, the AST physician and pharmacist leaders provided face-to-face education to prescriber groups regarding the Xpert MRSA/SA BC test and how to interpret and tailor therapy according to the results (Fig. 1).
FIG 1.
Interpretation of preliminary and final RDT results. CoNS, coagulase-negative staphylococci; GPC, Gram-positive cocci; ID, identification; MRSA, methicillin-resistant Staphylococcus aureus; MSSA, methicillin-susceptible Staphylococcus aureus; RDT, rapid diagnostic testing. *, if deemed a contaminant.
Positive blood cultures are considered critical laboratory values; therefore, in both study groups, as soon as a blood culture resulted with GPC in clusters, the patient's bedside nurse was notified by microbiology staff. In the pre-RDT group, no further communication was made by laboratory staff following the initial positive Gram stain report. In the post-RDT group, the laboratory technician contacted the bedside nurse a second time once the final organism identification was reported via the Xpert test. The bedside nurse was responsible for notifying the patient's primary inpatient care provider after each call, and the provider was then responsible for assessing the patient and evaluating antimicrobial therapy. There was no active notification of the AST positive blood culture results in either group.
Active and appropriate therapy for MRSA bacteremia was defined as intravenous (i.v.) vancomycin or daptomycin. Preferred and appropriate therapy for MSSA was defined as a susceptible i.v. β-lactam. If a severe β-lactam allergy precluded the use of these agents, MRSA-active therapy was deemed appropriate. A contaminated blood culture was defined as CoNS in one out of two blood culture sets drawn via the same order, or CoNS in two out of two blood culture sets, with each set identified as a different species. The appropriate therapy for CoNS strains that were deemed pathogenic was MRSA-active therapy. The appropriate therapy for CoNS strains that were deemed a contaminant was no antimicrobial therapy. For patients with CoNS-contaminated blood cultures receiving only antibiotics directed at another source of infection (e.g., pyelonephritis), time to appropriate therapy could not be determined, as blood cultures were not used to direct therapy, and these patients were excluded from the primary endpoint. Time to appropriate therapy was calculated as time from Gram stain result to the time of antibiotic order. Subjects were considered immunocompromised if they had a history of solid organ transplant or current steroid use equivalent to ≥15 mg of oral prednisone for at least 30 days.
All statistical analyses were performed using SPSS software, version 22 (SPSS Inc., Chicago, IL). Nominal variables were assessed using a chi-square or Fisher's exact test as appropriate. Continuous variables were assessed using Student's t test or Mann-Whitney U test as appropriate, based on the distribution of data. A P value of <0.05 was considered statistically significant for all comparisons.
A total of 373 patients had positive blood cultures with GPC in clusters over the study period, of which 252 patients met the inclusion criteria for the study (Fig. 2). The pre-RDT group included 143 patients, while the post-RDT group included 109 patients. The baseline characteristics in each group were similar (Table 1). In total, 58 (23%) patients had SAB, of whom 20 patients had MRSA and 38 patients had MSSA isolates. The remaining 194 (77%) patients had CoNS bloodstream isolates (Fig. 3), and the majority in each group were determined to be contaminants, including 107 (91.5%) patients in the pre-RDT group and 72 (93.5%) patients in the post-RDT group.
FIG 2.
Study population. staph, Staphylococcus.
TABLE 1.
Baseline characteristicsa
| Characteristic | Pre-RDT (n = 143) | Post-RDT (n = 109) | P value |
|---|---|---|---|
| Male | 79 (55.2) | 60 (55) | 0.975 |
| Age (yr) (mean [SD]) | 64 (15.2) | 63 (16.2) | 0.534 |
| Penicillin allergy | 24 (16.8) | 23 (21.1) | 0.383 |
| Cephalosporin allergy | 6 (4.2) | 3 (2.8) | 1.0 |
| Vancomycin allergy | 3 (2.1) | 0 (0) | 0.261 |
| Neutropenic | 5 (3.5) | 3 (2.8) | 1.0 |
| Immunocompromised | 4 (2.8) | 2 (1.8) | 0.701 |
| Source of infection | 0.265 | ||
| Skin and soft tissue infection | 8 (5.6) | 4 (3.7) | |
| Urinary | 0 (0) | 1 (0.9) | |
| Respiratory | 0 (0) | 2 (1.8) | |
| Endocarditis | 1 (0.7) | 2 (1.8) | |
| Vascular catheter | 12 (8.4) | 8 (7.3) | |
| Osteomyelitis or septic arthritis | 6 (4.2) | 5 (4.6) | |
| Other | 2 (1.4) | 4 (3.7) | |
| Unknown | 7 (4.9) | 11 (10.1) | |
| No infection | 107 (74.8) | 72 (66.1) | |
| Infectious disease consultation | 73 (51) | 64 (58.7) | 0.226 |
| Prior to Gram stain result | 35 (47.9) | 20 (31.3) | 0.047b |
| After Gram stain result | 38 (52.1) | 44 (68.8) | |
Results are reported as number (%), unless otherwise indicated. RDT, rapid diagnostic testing.
This P value compares prior to and after Gram stain results.
FIG 3.
Staphylococcus species isolated.
Of patients who were eligible for transition to appropriate therapy (pre-RDT, n = 75 [52.4%]; post-RDT, n = 53 [48.6%]), the median time to appropriate therapy was significantly shorter in the post-RDT group than in the pre-RDT group (pre-RDT, 15.0 h [interquartile range {IQR}, 0 to 27.2 h] versus post-RDT, 0 h [IQR 0 to 2.9 h]; P < 0.001). There was an 18-h decrease in the time to appropriate therapy for patients with CoNS-contaminated blood cultures not receiving antibiotics for another infection (n = 39 [36.4%] in pre-RDT group; n = 16 [22.2%] in post-RDT group) (Table 2). Furthermore, significantly fewer MRSA-active doses were administered per patient after the Gram stain result (pre-RDT, 1 dose [IQR 0 to 3 doses], versus post-RDT, 0 doses [IQR 0 to 2 doses]; P = 0.003). A total of 106 (54.6%) patients with CoNS isolated were not receiving empirical MRSA-active therapy prior to Gram staining, including 65 (55.5%) patients in the pre-RDT and 41 (53.2%) patients in the post-RDT group. Following the Gram stain results, 47.7% of these patients in the pre-RDT group versus 24.4% in the post-RDT group were initiated on MRSA-active therapy (P = 0.042). There were no differences detected in the times to MRSA-active or MSSA-preferred therapy between groups (Table 2). Overall, there was no difference in mean length of hospital stays (LOS) between groups; however, for patients with CoNS isolated, the mean LOS was significantly shorter in the post-RDT group than in the pre-RDT group (10.5 versus 7.7 days, respectively; P = 0.015). Optimization of antimicrobial therapy was initiated by a pharmacist in approximately half of patients in both the pre-RDT and post-RDT groups (pre-RDT, 50% versus post-RDT, 44.4%; P = 0.746). Additional secondary outcomes are found in Table 3.
TABLE 2.
RDT impact on antimicrobial therapya
| Variable by cohort | Pre-RDT | Post-RDT | P value |
|---|---|---|---|
| Total cohort (n) | 75 | 53 | |
| Time to appropriate therapyb (h) | 15.0 (0–27.2) | 0 (0–2.9) | <0.001 |
| CoNSc (n) | 39 | 16 | |
| Time to appropriate therapy (h) | 26.4 (20–36.3) | 8.3 (1.5–26.7) | 0.006 |
| MSSA (n) | 20 | 18 | |
| Time to MSSA-preferred therapy (h) | 0 (0–0) | 0 (0–0) | 0.654 |
| No. of MRSA-active doses following Gram stain result | 1 (0.25–2) | 1 (0–2) | 0.409 |
| MRSA (n) | 6 | 14 | |
| Time to MRSA-active therapy (h) | 0 (0–8.7) | 0 (0–0) | 0.274 |
Data are presented as median (interquartile range [IQR]), unless otherwise indicated. CoNS, coagulase-negative staphylococci; MRSA, methicillin-resistant Staphylococcus aureus; MSSA, methicillin-susceptible Staphylococcus aureus; RDT, rapid diagnostic testing.
Includes all S. aureus strains, CoNS pathogens, and CoNS contaminants with antibiotics directed at bacteremia and no additional infectious source.
Contaminants only.
TABLE 3.
Secondary outcomesa
| Variable by cohort | Pre-RDT | Post-RDT | P value |
|---|---|---|---|
| Total cohort (n) | 143 | 109 | |
| Length of stay (mean [SD]) (days) | 10.8 (9.8) | 8.8 (7.5) | 0.074 |
| 30-day hospital readmission (no. [%]) | |||
| All-cause | 25 (17.5) | 17 (15.6) | 0.691 |
| Infection-related | 10 (7.0) | 8 (7.3) | 0.650 |
| In-hospital all-cause mortality (no. [%]) | 9 (6.3) | 6 (5.5) | 0.793 |
| Hospital-onset Clostridioides difficile incidence (no. [%]) | 1 (0.7) | 1 (0.9) | 0.847 |
| Staphylococcus aureus (MRSA and MSSA) (n) | 26 | 32 | |
| Length of stay (mean [SD]) (days) | 12.1 (7.5) | 11.7 (10.1) | 0.847 |
| SAB recurrence within 90 days (no. [%]) | 0 (0) | 1 (3.1) | 0.251 |
| ID consulted (no. [%]) | 26 (100) | 32 (100) | 1.0 |
| Time to ID consult (median [IQR]) (h) | 6.5 (2.2–25) | 5.8 (1.6–10.5) | 0.508 |
| CoNS, all patients (n) | 117 | 77 | |
| Length of stay (mean [SD]) (days) | 10.5 (10.2) | 7.7 (5.9) | 0.015 |
| CoNS contaminants (n) | 107 | 72 | |
| Length of stay (mean [SD]) (days) | 10.1 (10.2) | 7.5 (6.0) | 0.036 |
| CoNS pathogens (n) | 10 | 5 | |
| Length of stay (mean [SD]) (days) | 14.9 (9.8) | 9.8 (2.1) | 0.279 |
| ID consulted (no. [%]) | 10 (100) | 5 (100) | 1.0 |
CoNS, coagulase-negative staphylococci; MRSA, methicillin-resistant Staphylococcus aureus; MSSA, methicillin-susceptible Staphylococcus aureus; RDT, rapid diagnostic testing.
Our study demonstrated that implementing RDT with the GeneXpert MRSA/SA BC test for Gram-positive blood cultures in a community teaching hospital significantly improved the time to appropriate therapy for patients with staphylococcal blood cultures. Furthermore, the RDT decreased MRSA-active therapy exposure and LOS for patients with CoNS contaminants. This impact was demonstrated despite the absence of real-time AST notification by the microbiology staff and rather relied upon the house-wide educational effort by the AST leaders. With only two results reported for interpretation by clinicians, the Xpert MRSA/SA BC test may be a starting point for community hospitals looking to implement RDT with limited resources. This assay has the potential to help demonstrate improved outcomes and associated cost savings, which could be used to gain further AST resources to implement different types of RDT or to increase lab support for AST notification.
This study further supports the findings of two previous investigations evaluating RDT implementation using the Xpert MRSA/SA BC test without real-time AST notification (6, 7). Similar to findings from our study, Parta and colleagues found that significantly fewer patients with CoNS blood cultures were treated for SAB following RDT implementation in a Veterans Affairs medical center (76% versus 55%, respectively; P < 0.001) (6). The duration of MRSA-active treatment in patients with CoNS was also significantly decreased in the XPert MRSA/SA group (44 versus 31.8 h, respectively; P = 0.03) (6). Turner and colleagues compared the Xpert MRSA/SA BC test to the standard microbiological identification of SAB in a large multisite health system. The RDT significantly decreased the time to optimal therapy for the subgroup of patients with MSSA by 11.7 h (P = 0.0011) (7). In our study, the implementation of the RDT did not demonstrate a significant difference in time to appropriate therapy for SAB, which was likely due to most patients empirically receiving active therapy; additionally, as only 58 patients had SAB, we may have lacked power to detect a difference between groups. This finding is important to note, as community hospitals may see more benefit in implementing RDTs for the detection of CoNS blood culture contaminants, as the incidence of SAB may be low.
There are additional limitations to our study that must be considered. First, as with all retrospective investigations, there is the risk for selection bias as well as the reliance on appropriate documentation. Standard physician documentation did not change between study arms, and each antibiotic order was required to have an indication selected. Additionally, while real-time AST notification was not provided, the AST was providing daily audit and feedback to providers during both study periods. Furthermore, this was a single-center investigation in a community hospital; therefore, the external validity may be limited, as larger academic centers may have a more difficult time providing house-wide education regarding new diagnostic tests. Finally, while we did observe a significant reduction in MRSA-active therapy and LOS for patients with CoNS blood cultures, we did not compare costs between groups. Despite these limitations, our study provides important evidence that real-time AST notification of RDT results for Gram-positive bloodstream isolates is not essential to demonstrate an impact on time to appropriate therapy or patient outcomes in community hospitals.
In conclusion, the implementation of the GeneXpert MRSA/SA BC RDT without AST notification significantly improved time to appropriate therapy for blood cultures with GPC in clusters at a community hospital. Additionally, LOS was significantly shorter for patients with CoNS. Community hospitals with limited AST resources may benefit from implementing RDT technologies with easy-to-interpret results, allowing for AST resources to be utilized elsewhere.
ACKNOWLEDGMENTS
We acknowledge Cora Manby, microbiology lead technician, and her staff for their contributions to this study.
REFERENCES
- 1.Sango A, McCarter YS, Johnson D, Ferreira J, Guzman N, Jankowski CA. 2013. Stewardship approach for optimizing antimicrobial therapy through use of a rapid microarray assay on blood cultures positive for Enterococcus species. J Clin Microbiol 51:4008–4011. doi: 10.1128/JCM.01951-13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Perez KK, Olsen RJ, Musick WL, Cernoch PL, Davis JR, Land GA, Peterson LE, Musser JM. 2013. Integrating rapid pathogen identification and antimicrobial stewardship significantly decreases hospital costs. Arch Pathol Lab Med 137:1247–1254. doi: 10.5858/arpa.2012-0651-OA. [DOI] [PubMed] [Google Scholar]
- 3.Perez KK, Olsen RJ, Musick WL, Cernoch PL, Davis JR, Peterson LE, Musser JM. 2014. Integrating rapid diagnostics and antimicrobial stewardship improves outcomes in patients with antibiotic-resistant Gram-negative bacteremia. J Infect 69:216–225. doi: 10.1016/j.jinf.2014.05.005. [DOI] [PubMed] [Google Scholar]
- 4.Bauer KA, West JE, Balada-Llasat JM, Pancholi P, Stevenson KB, Goff DA. 2010. An antimicrobial stewardship program's impact with rapid polymerase chain reaction methicillin-resistant Staphylococcus aureus/S. aureus blood culture test in patients with S. aureus bacteremia. Clin Infect Dis 51:1074–1080. doi: 10.1086/656623. [DOI] [PubMed] [Google Scholar]
- 5.Huang AM, Newton D, Kunapuli A, Gandhi TN, Washer LL, Isip J, Collins CD, Nagel JL. 2013. Impact of rapid organism identification via matrix-assisted laser desorption/ionization time-of-flight combined with antimicrobial stewardship team intervention in adult patients with bacteremia and candidemia. Clin Infect Dis 57:1237–1245. doi: 10.1093/cid/cit498. [DOI] [PubMed] [Google Scholar]
- 6.Parta M, Goebel M, Thomas J, Matloobi M, Stager C, Musher DM. 2010. Impact of an assay that enables rapid determination of Staphylococcus species and their drug susceptibility on the treatment of patients with positive blood culture results. Infect Control Hosp Epidemiol 31:1043–1048. doi: 10.1086/656248. [DOI] [PubMed] [Google Scholar]
- 7.Turner RB, Lalikian K, Fry M, Schwartz J, Chan D, Won R. 2017. Impact of rapid identification of Staphylococcus aureus bloodstream infection without antimicrobial stewardship intervention on antibiotic optimization and clinical outcomes. Diagn Microbiol Infect Dis 89:125–130. doi: 10.1016/j.diagmicrobio.2017.07.003. [DOI] [PubMed] [Google Scholar]