Quick BSI Score to Identify Bloodstream Infection at the Emergency Department

Abstract


Background
Bloodstream infection (BSI) is a common as well as urgent condition at the emergency department (ED) [1,2].Incidence of bloodstream infection increased to 38.1 persons per 100,000 per year in 2010, while mortality rate may be as high as 50% [3].Early diagnosis and appropriate antimicrobial therapy are a key to improve patient outcomes [4], particularly among individuals displaying either septic shock or sepsis [5,6].
Current guidelines advise on obtaining hemoculture in patients suspected of sepsis in order to diagnose BSI [5,7].Positive blood culture is an important yield in terms of the appropriate antibiotics in sepsis patients [5,8].However, the guideline fails to mention when blood cultures ought to be procured.Furthermore, the decision of whether or not to take hemoculture is based solely upon clinical judgment which could result in wastefulness resultant of inappropriately requested blood cultures [8][9][10].Notwithstanding, there are several predictors of bloodstream infection at the ED, such as blood pressure less than 60 mmHg, a high procalcitonin level of over 2 µg/L, or C-reactive protein > 10 mg/dL [11].
Shapiro et al reported clinical scoring for bloodstream infection at the ED with a decent validation of 83% [12].However, these previous reports may call for laboratory results to predict bloodstream infection at the ED.Unfortunately, this may result in delays in sepsis management on waiting for laboratory tests [6].Hence, this study aimed to develop bloodstream infection predictive scoring at the ED using only clinical factors presented on ED arrival.

Study design and Ethical Approval
This study was conducted retrospectively at the ED, Khon Kaen University Hospital, a tertiary care hospital with approximately 60000 annual ED visits.Moreover, this study formed part of an ED infection project.
Inclusion criteria were adult patients suspected of infection de ned by blood culture collection at the ED with intravenous antibiotics initiated during ED visits.Cases demonstrating cardiac arrest or trauma; those referred from other hospitals; those who had received antibiotics prior; and those missing clinical data were excluded.The study period took place between January 1st, 2016 and December 31st, 2018 with study protocol approved by the ethics committee in human research, Khon Kaen University (HE631115).Informed consents were waived.

Source Of Data And Microbiology Methods
Blood cultures at the ED comprised of two aerobic bottles.Bloodstream infections were de ned as positive blood culture with a similar pathogen to at least one sample with clinical relevance.The contaminated pathogens (e.g.coagulase-negative Staphylococci, Corynebacterium spp., Propionibacterium spp., Viridans group streptococci, Micrococcus spp.and Bacillus spp.) were considered as such if they were isolated from a patient twice or more consecutively with clinical relevance [7,13,14].Clinical data of eligible patients were retrieved from the computerized hospital database and chart records.Data were subsequently categorized as comorbid conditions, ED arrival parameters, and beyond the initial hour following ED visit.Comorbid conditions were de ned according to the Charlson Comorbidity Index (CCI) [15].ED arrival parameters were history of fever, the chills, vital signs, and sepsis scores including Systemic In ammatory Response Syndrome (SIRS), quick Sepsisrelated Organ Failure Assessment (qSOFA) and National Early Warning Score (NEWS).For parameters beyond the initial hour post ED visit, laboratory results incorporated white blood cell count and lactate level.

Statistical Analysis
In regards statistical analyses eligible patients were categorized into two groups with respect to blood culture results: positive blood culture and negative blood culture groups.Descriptive statistics were used to compare differences in studied variables between both groups.Factors associated with positive blood culture were calculated via logistic regression analysis.A univariate and multivariate logistic regression was applied to calculate unadjusted/adjusted odds ratio (95% con dence interval) of each factor.Independent positive predictors for positive blood culture were utilized to create the Quick blood stream infection score (qBSI score).Clinical factors excluding laboratory results were used for the qBSI score with the aim of identifying bloodstream infections faster minus the wait for laboratory results.Each predictor presented a clinical score based on the coe cient yielded by the nal model for positive blood culture.The qBSI score revealed summation of each predictor.Various qBSI score cutoff points were executed and reported along with diagnostic properties including sensitivity, speci city, positive/negative predictive values (PPV/ NPV), and positive/negative likelihood ratios (LR+/LR-).A receiver operating characteristic (ROC) curve of the qBSI score was computed and compared with other sepsis scores.All statistical analyses were performed using STATA software, version 10.1 (College Station, Texas, USA).

Patient Characteristic and Microbiology Data
A total of 169578 patients visited the ED during the study period as retrieved from the hospital database.Of those, 12556 (7.40%) were suspected of infection.After exclusion, 8177 individuals met the study criteria and were categorized according to blood culture results as follows: positive bloodstream infection (741 patients; 9.06%) and negative blood culture or non-pathogen bacteremia (7,436 patients; 90.94%) as shown in Fig. 1.Among the variables studied, almost all were signi cantly different between both groups (Table 1).Merely proportion of AIDS was not signi cantly different between groups (2.16% in positive blood culture group and 1.44% in negative blood culture group; p 0.125).The most com-mon Gram negative and positive pathogens were Escherichia coli (274 patients; 36.98%) and Streptococcus (76 patients; 10.26%).

Discussion
The qBSI score, to our knowledge, is the rst score to quickly identify bloodstream infection at the ED minus the wait for additional laboratory results.The advantage of this score is that the risk of bloodstream infection at the ED can be calculated speedily and easily using clinical formulae mentioned in the results.Also, it demonstrated an ample sensitivity of 92.98% with high negative predictive value.
These results may indicate that with a qBSI score of 1 or over, the likelihood of bloodstream infection stands at 92.98%.Meanwhile, a qBSI score of less than 1 presents a 95.65% chance of not having a positive blood culture.The qBSI score can be calculated within minutes at the ED in patients at risk of bloodstream infection once a complete medical history has been obtained and a physical examination has occurred.
Even though sepsis scores SIRS, qSOFA, and NEWS score exhibit high accuracy [16,17] amid the identifying of sepsis patients, these scores were not an effective indicator in terms of bloodstream infection.The qBSI score displayed superior prediction of bloodstream infection when compared with sepsis scores (Fig. 2).As the qBSI score demonstrated a greater area under the ROC curve than others, this may indicate that emergency physicians could feasibly apply the qBSI score to identify sepsis patients who may be exhibiting bloodstream infection.Despite some laboratory tests including lactate or white blood cells, they were independent factors in terms of bloodstream infection (Table 3), and they were not included in the qBSI score to save time amid the identifying of patients at risk of bloodstream infection.These results also indicate that laboratory tests for sepsis scores were ample parameters amid bloodstream infection.Yet, once again, this may delay the sepsis treatment bundle [6].
In accordance with our outcomes, employing this score could prove effective in determining cases calling for blood culture at the ED.In those suspected of infection, cases with a qBSI score of 1 or over they require blood culture which is attributable to a high chance of exhibiting bloodstream infection.Conversely, those recording a qBSI score of 0 present a low chance of bloodstream infection (NPV 95.65%).Thus, this patient group probably do not need blood cultures.A more appropriate blood culture order may be reached as an outcome of abiding by this decision rule.
The positive blood culture rate in this study was comparable with previous studies: 9.06% (up to 12.4%) [11,18].Inappropriate blood cultures obtained from patients at a low risk may yield false positives as well as antibiotic overuse [18].Besides that, the qBSI score may be a helpful tool in correctly identifying patients likely to display positive blood culture, as previously discussed.Even though this study revealed different predictors to previous studies [19,20], these factors were reported to be associated with positive blood cultures [21,25].For example, cirrhosis patients demonstrated a superior incidence of bloodstream infection over non-cirrhotic patients throughout ten instances [24].As this study included only comorbid conditions and ED factors, these may result in different predictors of positive blood culture to other studies.

Strengths And Limitations
There are some limitations in this study.Even though the study incorporated quite a large sample size, some clinical data may be missing due to the retrospective study design.Second, the qBSI score did not include laboratory results.Finally, score speci city was not high as previously reported by Shapiro et al. [12].Still, sensitivity is of superior importance in this situation and those with bloodstream infection ought not be overlooked.Additionally, the qBSI score is a speedier clinical tool than the Shapiro report, that is to say the qBSI score can be calculated within minutes following ED arrival.

Conclusions
The qBSI score had good sensitivity and negative predictive value for positive blood culture in patients presenting at the ED with suspicious of infection.The score comprised of six clinical variables without laboratory results.Using this score may facilitate in the determining of those exhibiting the need for blood culture at the ED.

Figure 1 Study
Figure 1

Figure 2 The
Figure 2

Table 1
Baseline characteristics of patients with suspected infection presenting at the emergency department categorized by blood culture results.Among three categories of studied variables, there were four signi cant factors in comorbid conditions, two factors at ED, and three factors beyond the 1st hour (Table2).The six signi cant predictors for positive blood cultures were age over 55 years, moderate to severe CKD, solid organ tumor, liver disease, history of chills, and body temperature of over 38.3oC.The scores of each parameter were shown in Table3with the total score of 9.The probability of positive blood culture calculated by age over 55 years CKD: chronic kidney disease; SBP: systolic blood pressure; MAP: mean arterial pressure; SIRS: Systemic In ammatory Response Syndrome; qSOFA: quick Sepsis-related Organ Failure Assessment; NEWS: National Early Warning Score.CKD: chronic kidney disease; SBP: systolic blood pressure; MAP: mean arterial pressure; SIRS: Systemic In ammatory Response Syndrome; qSOFA: quick Sepsis-related Organ Failure Assessment; NEWS: National Early Warning Score.Clinical factors and calculation of Quick Bloodstream Infection score (qBSI score) were 0 or 1, respectively.The sum of all parameters represented probability of positive blood culture.A cutoff point of this qBSI score of 1 or over had sensitivity, speci city, PPV, NPV, LR+, and LR-of 92.98%, 15.40%, 9.87%, 95.65%, 1.09, and 0.45, respectively (Table4).The ROC of qBSI score (65.8%) was higher than the SIRS (61.8%), qSOFA (58.6%), and NEWS score (60.8%) as shown in Fig.2.

Table 2
Factors associated with positive blood culture in patients suspected of infection presenting at the emergency department.

Table 3
Individual and adjusted odds ratio of component of the Quick Bloodstream Infection score (qBSI score) predicting positive blood culture in patients with suspected infection presenting at the emergency department.