|Home | About | Journals | Submit | Contact Us | Français|
Staphylococcus aureus bacteremia (SAB) causes high rates of morbidity and death. Several studies in academic health settings have demonstrated that consultations from infectious diseases specialists improve the quality of care and clinical outcomes for SAB. Few data that describe the impact in resource-limited settings such as community hospitals are available. This retrospective cohort study evaluated the adherence to quality-of-care indicators and the clinical outcomes for SAB in a five-hospital community health system (range of 95 to 272 available beds per hospital), for patients with versus without infectious diseases consultation (IDC). IDC was provided if requested by the attending physician. The primary outcome was the incidence of treatment failure, defined as 30-day in-hospital death or 90-day SAB recurrence. Other outcomes included adherence to quality-of-care indicators. A total of 473 adult patients with SAB were included, with 369 (78%) receiving IDC. We identified substantial differences in baseline characteristics between the IDC group and the no-IDC group, including greater incidences of complicated bacteremia and intravenous drug users in the IDC group, with similar rates of severe illness (measured by Pitt bacteremia scores). Adherence to quality-of-care indicators was greater for patients with IDC (P < 0.001). After adjustment for other predicting variables, IDC was associated with a lower rate of treatment failure (adjusted odds ratio, 0.42 [95% confidence interval, 0.20 to 0.86]; P = 0.018). IDC provided better quality of care and better clinical outcomes for patients with SAB who were treated at small, resource-limited, community hospitals.
Staphylococcus aureus is the most common bacterial pathogen isolated from inpatient cultures and the second most common in outpatient cultures (1). S. aureus bacteremia (SAB) is associated with significant morbidity and demonstrates attributable mortality rates of 10 to 30% (2,–7). Management of SAB can be challenging, as quality-of-care indicators include identification and removal of infectious foci, collection of repeat blood cultures until resolution of bacteremia, echocardiography, differentiation of complicated and uncomplicated bacteremia, and selection of appropriate intravenous (i.v.) treatment and duration (8). Infectious diseases consultation (IDC) has been associated with greater adherence to these quality-of-care indicators (4,–7, 9, 10), with some studies also demonstrating lower mortality rates (4, 5). The vast majority of studies evaluating this impact have been conducted at large academic hospitals, with few data from resource-limited settings such as community and rural hospitals. Community hospitals may differ from academic hospitals in patient volumes, patient characteristics, available resources, and clinical outcomes for certain medical conditions (11,–14). Because substantial differences may exist, study of the impact of IDC on the outcomes of patients with SAB in community hospitals is necessary. We conducted a retrospective cohort study at a five-hospital community health system to examine the impact of IDC on adherence to quality-of-care indicators and clinical outcomes of adult patients with SAB.
(This study was presented in part at the 55th Interscience Conference on Antimicrobial Agents and Chemotherapy, San Diego, CA, 2015.)
This study included adult patients who were treated between April 2012 and June 2015 at Legacy Health hospitals in the greater Portland, Oregon, area. Legacy Health is a five-hospital health system in an urban/suburban location, with a total of 938 available beds. Hospital sizes range from 95 to 272 available beds. All hospitals provide general medical and surgical services; three provide cardiac intervention and oncological services and one provides trauma and burn services. During the study period, the health system employed two or three adult infectious diseases specialists to provide bedside consultations at four of the hospitals, with a two-member community infectious diseases group contracted to provide services at the fifth hospital. IDC is performed only upon request from the primary treating practitioner, who contacts the covering specialist via telephone. Infectious diseases specialists are available all days of the week during normal working hours but typically are not available during the night for bedside consultations. IDC is equally available at all hospitals, with specialists traveling between hospitals on most days in order to provide consultations. No system-specific clinical guidelines for treatment of Staphylococcus aureus are provided to clinicians, and treatment is at the discretion of the primary treating practitioner. A dosing protocol is utilized for i.v. treatment with vancomycin, which is dosed exclusively by clinical pharmacists to target troughs of 15 to 20 mg/liter for serious infections.
In our system, it is preferable to collect at least two initial sets of blood cultures for patients suspected of having possible bacteremia. Blood cultures from all hospitals are sent to a centralized microbiology laboratory for incubation in a BacT/Alert 3D instrument (bioMérieux, Marcy-l'Etoile, France). Blood cultures flagged as positive are subjected to Gram staining 24 h per day, and the results of Gram staining are paged to clinical staff members. Identification of bacterial species and drug susceptibility testing are performed using a Vitek 2 instrument (bioMérieux). Susceptibility is determined in accordance with Clinical and Laboratory Standards Institute guidelines (15). Susceptibility testing is performed manually for vancomycin and daptomycin.
During the study period, there was no infectious diseases fellow or infectious diseases pharmacist. A multidisciplinary antimicrobial stewardship oversight committee was formed in July 2014 but did not perform patient care interventions. One of the first efforts of this committee was to justify mandatory IDC for patients with SAB. There was concern that physician groups would not support the concept of mandatory IDC, particularly as limited data demonstrating the benefits of IDC for SAB outcomes outside academic health settings are available. Therefore, we sought to determine the impact of IDC on the clinical outcomes of patients with SAB within our community health system. This study was granted approval by the Legacy Health institutional review board.
Adult patients with S. aureus isolated from at least one blood culture were screened for inclusion. Patients who were discharged or placed on palliative care within 48 h after the positive blood culture result, were transferred from an outside facility for SAB management, had experienced SAB in the previous 90 days, or had polymicrobial bacteremia were excluded.
Polymicrobial bacteremia was defined as isolation of an additional organism from a blood culture within 72 h after the original culture, with the exception of common skin contaminants such as coagulase-negative staphylococci, Bacillus species, and Corynebacterium species (16). We defined uncomplicated bacteremia as in the Infectious Diseases Society of America guidelines for treatment of methicillin-resistant S. aureus (MRSA) (8), i.e., exclusion of endocarditis and other metastatic sites of infection, the absence of implanted prostheses, clearance of bacteremia within 4 days for patients with repeat blood cultures, and defervescence within 72 h after the initiation of effective therapy. Complicated bacteremia was defined as cases not meeting the criteria for uncomplicated bacteremia. Pitt bacteremia scores were calculated as described previously (17). Definitive therapy was defined as the antibiotic regimen selected for the completion of therapy after susceptibility results were known. Optimal therapy was defined as cefazolin or nafcillin for methicillin-susceptible S. aureus (MSSA) and vancomycin or daptomycin for MRSA. Antimicrobial duration was deemed appropriate if planned for ≥14 days for uncomplicated bacteremia, ≥28 days for complicated bacteremia, or ≥42 days for infective endocarditis. Length of stay was calculated as days from the time of the original culture to discharge, and infection-related length of stay was calculated as days until discharge or discontinuation of therapy directed at SAB. Recurrence was defined as repeat SAB within 90 days after the original culture, following the resolution of bacteremia. Death was defined as in-hospital, all-cause death within 30 days after the original culture.
The primary outcome was treatment failure, defined as meeting the criteria for death or recurrence (10, 18). Secondary outcomes included 60-day readmission with a possible complication related to SAB and hospital, intensive care unit (ICU), and infection-related lengths of stay. In addition, we examined adherence to previously defined quality-of-care indicators, including collection of repeat blood cultures within 2 to 4 days after the original culture, echocardiography, appropriate duration of i.v. antibiotic therapy, and selection of optimal definitive therapy (3,–7).
It was determined that a sample size of 454 patients was required to achieve 80% power with an alpha of 0.05, assuming failure rates of 20% and 10% in the control and IDC groups, respectively, with 70% of the cohort receiving IDC. Baseline characteristics and outcomes were compared between patients with IDC and those without IDC by Fisher's exact test, the chi-square test, Student's t test, and the Wilcoxon rank sum test, as appropriate. Variables identified as potential predictors of treatment failure with P values of <0.2 in univariate analyses were included in a multivariate logistic regression model. All reported confidence intervals (CIs) are 2-sided 95% intervals, and P values of <0.05 were considered significant. All analyses were performed with Stata version 13 (StataCorp, College Station, TX).
Of 616 patients screened, 473 were included in the final analysis, with 369 (78.0%) receiving IDC (Fig. 1). Baseline characteristics are presented in Table 1. Substantial differences between patients who received IDC and patients who did not receive IDC were identified (Table 1). In general, patients who received IDC were younger, had better baseline renal function, were more likely to have complicated bacteremia, and were more likely to be i.v. drug users. However, illness severity levels appeared to be similar between the groups, with similar incidences of intensive care unit (ICU) admission and Pitt bacteremia scores of ≥4. The median time from the availability of Gram staining results to the initial IDC was 1.4 days, and times varied from 0.8 to 1.9 days among hospitals (P = 0.016). Patients who received IDC were more likely to receive treatment adherent to quality-of-care indictors (Fig. 2). Hospital and infection-related lengths of stay were slightly greater for patients with IDC (Table 2).
Only 12.3% of all patients met the criteria for treatment failure, with a mortality rate of 7.8% and a recurrence rate of 4.4%. The unadjusted treatment failure rate was lower for patients with IDC; however, this difference did not reach statistical significance (Table 2). Multivariate logistic regression analysis revealed the development of renal dysfunction (defined by the RIFLE [risk, injury, failure, loss of kidney function, and end-stage kidney disease] criteria ), the presence of metastatic foci, and Pitt bacteremia scores of ≥4 to be statistically significant predictors of treatment failure. After adjustment for these and other potential predictors, IDC was associated with a lower risk of treatment failure (Table 3). IDC was associated with a lower risk of death when these same predictors were used in a multivariate logistic regression model; however, this did not reach statistical significance (adjusted odds ratio [OR], 0.45 [95% CI, 0.18 to 1.09]; P = 0.076).
In this five-hospital, single-health system study, IDC was associated with a lower incidence of treatment failure and a lower risk of death. IDC was also associated with greater adherence to quality-of-care indicators. We identified slightly longer lengths of stay for patients with IDC, which could be attributed to more frequent diagnosis of complicated bacteremia requiring further diagnostic testing and longer periods of i.v. therapy. This is one of the first reports to describe the impact of IDC on SAB outcomes in community hospitals. In 2014, Borde and colleagues evaluated the implementation of a mandatory bacteremia service by a single infectious diseases physician at a 200-bed community hospital (9). In their study of 59 patients with SAB, they identified higher levels of adherence to a stewardship bundle after the intervention and they noted a lower mortality rate. They hypothesized that the lower mortality rate might be attributable to factors other than their intervention, however, and the study was limited by the small number of patients included and the inability to control for confounding variables. Bai and colleagues identified a lower mortality rate for patients with SAB who received IDC at six academic or community hospitals; however, they did not differentiate the results by academic and community hospitals, and information pertaining solely to community hospitals cannot be extracted (4).
While clinical outcomes for various medical conditions, including pneumonia, have been shown to be worse at low-volume hospitals (11,–14), this has yet to be evaluated for SAB. Overall mortality rates in all five of our low-volume hospitals were similar, if not lower, than those reported previously for SAB (2,–7). We did not identify differences in treatment failure rates among the five hospitals in our system, with incidence rates ranging from 11.4% to 16.9% (P = 0.79). Outcomes were similar regardless of hospital, even after controlling for predictors of treatment failure. Isolated to each individual hospital, the multivariate model (outlined in Table 3) continued to find, or to suggest, that IDC resulted in more favorable outcomes across all hospitals.
Substantial differences in baseline characteristics between patients who received IDC and patients who did not receive IDC were identified (Table 1). The reasons for these differences are not entirely known. Primary practitioners may have a stronger propensity to obtain IDC for patients with certain characteristics that make complicated bacteremia easier to identify or more likely (such as patients with prosthetic material or i.v. drug users). In addition, infectious diseases practitioners are thought to be more likely to identify certain patient characteristics (such as metastatic foci) that may contribute to some of these differences. Interestingly, primary practitioners more frequently obtained IDC when SAB was part of the presenting indication than when SAB developed during the admission.
Several limitations exist in this study. First, this was a retrospective cohort study and, although we attempted to control for potential predictors of treatment failure, residual confounding may be present. Second, while this was a multihospital study, the main intervention was performed by only four or five infectious diseases specialists, and outcomes at other community hospitals may differ. In contrast to our setting, not all community hospitals have IDC available or they may have access only to telephone consultation, which was shown previously to be suboptimal (20). While most of the data, including the primary outcomes, were extracted objectively from electronic health records, other information (such as Pitt bacteremia scores and findings regarding complicated versus uncomplicated bacteremia) had to be collected manually. Reviewers had access to the IDC status of patients during manual chart reviews, and these data may be at risk of inadvertent bias. Finally, geographical and epidemiological differences may exist between this and other community hospitals, as all of our facilities are located in the same geographical region.
We conclude that IDC provided treatment more adherent to quality-of-care indicators and was associated with better clinical outcomes for patients with SAB at our relatively small, resource-limited, community hospitals. Further studies are required to validate these findings in other facilities and to establish care bundles and treatment strategies specifically addressing the challenges encountered by community hospitals.
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
The authors declare no conflicts of interest and no transparency issues.
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.