Clinica Chimica Acta 436 (2014) 319–322
Contents lists available at ScienceDirect
Clinica Chimica Acta journal homepage: www.elsevier.com/locate/clinchim
Limitation of the delta neutrophil index for assessing bacteraemia in immunocompromised children Jong Gyun Ahn a,b, Seong Yeol Choi c, Dong Soo Kim b,c, Ki Hwan Kim b,c,⁎ a b c
Department of Pediatrics, School of Medicine, Ewha Womans University, Seoul, Republic of Korea Department of Medicine, The Graduate School of Yonsei University, Seoul, Republic of Korea Department of Pediatrics, Severance Children's Hospital, Yonsei University College of Medicine, Seoul, Republic of Korea
a r t i c l e
i n f o
Article history: Received 3 February 2014 Received in revised form 29 April 2014 Accepted 19 June 2014 Available online 27 June 2014 Keywords: Delta neutrophil index Bacteraemia Children Immunocompromised
a b s t r a c t Background: Delta neutrophil index (DNI) has been reported to be useful in assessing the prognosis of suspected bacteraemia in adults. However, data in children are limited. The purpose of this study was to investigate the correlation between the DNI and bacteraemia in children and to compare DNI values between immunocompetent and immunocompromised hosts. Methods: We retrospectively collected data on 8593 children who underwent simultaneous blood culture and DNI testing. DNI was determined with a blood cell analyser (ADVIA 120, Siemens, Inc.). The children were divided into immunocompromised (n = 664) and immunocompetent groups (n = 7929). Results: DNI was higher in the bacteraemia group than in the non-bacteraemia group (P b 0.001). According to the group, DNI was higher in immunocompromised than immunocompetent patients (P b 0.001). Within the immunocompromised group, there was no significant difference between the bacteraemia and non-bacteraemia subgroups due to a higher DNI as compared to the immunocompetent group. Conclusions: DNI could be an additional method for the early diagnosis of bacteraemia in children. However, the use of DNI for the prediction of bacteraemia in immunocompromised cases has limitations. Further studies on the usefulness of DNI according to specific diseases are needed. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Bacteraemia is one of the most serious and life-threatening childhood diseases. The frequency of fatal bacteraemia is increased in immunocompromised individuals. Although early diagnosis and treatment is crucial for better outcomes [1], bacteraemia is identified in only about 30% of patients with sepsis, depending on previous antibiotic history, and microbiologic cultures take at least 24–72 h for adequate growth [2–4]. Several biomarkers that are used to aid rapid identification of patients with bacteraemia, such as C-reactive protein (CRP), procalcitonin and various interleukins, respond to both infection and non-infectious inflammation [5]. The search for more reliable and specific biomarkers for early diagnosis of bacteraemia is therefore critically important [6,7]. In recent years, the delta neutrophil index (DNI) has been shown to be useful for early diagnosis and prognostic assessment of adult and neonate patients with sepsis [8–11]. DNI is calculated by an automatic counter (ADVIA 120; Siemens Healthcare Diagnostics) using myeloperoxidase (MPO) channels and nuclear lobularity channels in the test system [12]. The difference between the leukocyte differentials assayed in the two ⁎ Corresponding author at: Department of Pediatrics, Severance Children's Hospital, Yonsei University College of Medicine, 50 Yonsei-ro, Seodaemungu, Seoul 120–752, Republic of Korea. Tel.: +82 2 2228 2050; fax: +82 2 393 9118. E-mail address: [email protected] (K.H. Kim).
http://dx.doi.org/10.1016/j.cca.2014.06.020 0009-8981/© 2014 Elsevier B.V. All rights reserved.
channels is designated as the DNI, which correlates with the fraction of immature granulocytes in peripheral blood calculated by manual counting. DNI is strongly associated with the disseminated intravascular coagulation score, positive blood culture rate, and mortality in patients with suspected sepsis [13]. The data in children, however, are limited and little is known about the usefulness of DNI in evaluating bacteraemia in immunocompromised patients. 2. Materials and methods 2.1. Study population and specimen collection Between June 2010 and May 2011, we retrospectively reviewed medical records of 8593 patients aged b 19 y who had laboratory requests for both blood culture and DNI under suspicion of having bacteraemia at Severance Children's Hospital in Korea. The children were divided into immunocompromised (n = 664) and immunocompetent groups (n = 7929). The immunocompromised state was defined as current treatment with chemotherapy or immunosuppressive therapy, hematopoietic stem cell or solid organ transplantation, or known primary immunodeficiency. Bacteraemia was defined as laboratoryconfirmed blood stream infection using the criteria of the US Centers for Diseases Control and Prevention/National Healthcare Safety Network surveillance [14]. The criteria are as follows: (1) a recognized
320
J.G. Ahn et al. / Clinica Chimica Acta 436 (2014) 319–322
pathogen, which does not include organisms that are considered common skin contaminants, cultured from 1 or more blood cultures, and (2) at least 1 of the following signs or symptoms: fever (N 38 °C), chills, or hypotension(in case of patient ≤1 year of age, at least 1 of the following signs or symptoms: fever (N 38 °C, rectal), hypothermia (b37 °C, rectal), apnoea, or bradycardia). Patients with blood culture yielding organisms considered contaminants were excluded. In accordance with the criteria described previously, positive blood culture for nonspecified streptococci, coagulase-negative staphylococci, Bacillus sp. other than Bacillus anthracis, diphtheroids sp., Corynebacterium sp., Propionibacterium acnes, Micrococcus sp., and Clostridium perfringens, or Streptococcus mitis isolated in a single set of culture bottles were regarded as contaminants [15,16]. All blood culture negative patients were included in the non-bacteraemia group. The study was approved by the Institutional Review Board of Severance Hospital and informed consent was waived. 2.2. Blood culture, Delta neutrophil index (DNI) and other laboratory data Blood cultures were performed in more than one pair of bottles for aerobic and anaerobic cultures, respectively, and incubated in an automated blood culture system (BacT/Alert 3D; bioMérieux), capable of detecting the growth of any organism. When bacterial growth was noted, a small volume of cultured specimen was inoculated onto both blood and MacConkey agar plates that were cultured overnight at 37 °C in an incubator enriched with 5% CO2. Isolates were identified based on typical colony morphology, Gram staining, and biochemical tests, and were further verified by the Vitek Identification System (bioMérieux). When no bacterial growth was detected within five days after primary inoculation of the blood sample, the result of the blood culture was considered negative. Data of complete blood cell (CBC) counts were collected with an automated blood cell analyser (ADVIA 2120i; Siemens). DNI was calculated using the following formula that was introduced in a previous study: DN = (the leucocyte subfraction assayed in the MPO channel by cytochemical reaction) – (the leucocyte subfraction counted in the nuclear lobularity channel by the reflected light beam) [13]. Serum CRP concentration (normal range 0–8 mg/L) was measured using the nephelometric method (Beckman Coulter, Fullerton, CA, USA). Among a total of 8593 cases, 959 cases did not receive a CRP test and the platelet count of one case was missing.
test for categorical variables. Potential risk factors for sepsis were identified by univariate analysis. Significant variables were further subjected to multivariate logistic regression analysis. Receiver operating characteristic (ROC) curves of DNI between groups were drawn and compared. A p b 0.05 were considered statistically significant. 3. Results 3.1. Comparison of bacteraemia and non-bacteraemia groups During the study period, data on a total of 8593 patients were collected. Fifty-eight percent of the subjects (4947/8593) were male and the median age of children was 2.6 years (range: 0.1–18 y). Among a total of 8593 cases, 44 children were diagnosed with blood cultureproven bacteraemia. The comparison between the bacteraemia and non-bacteraemia groups is summarized in Table 1. The proportion of male patients in the bacteraemia group and the non-bacteraemia group was 75.0% and 57.5%, respectively. The median ages of the bacteraemia and non-bacteraemia patients were 2.6 (IQR, 0.4–12.8) and 2.6 (IQR, 1.2–5.3) y, respectively, with no significant difference. In a univariate analysis, the concentrations of both DNI and CRP in the bacteraemia group were higher than those in the non-bacteraemia groups (P b 0.0001, P = 0.004, respectively), while white blood cell (WBC) counts, haemoglobin concentration, and platelet counts in the bacteraemia group were lower than those in the non-bacteraemia groups (P b 0.0001 for all 3 markers). In a multivariate model, significant risk factors for bacteraemia were DNI (P = 0.001; 95% CI = 1.043–1.168), WBC counts (P = 0.047; 95% CI = 1.000–1.000), and CRP (P = 0.043; 95% CI: 1.000–1.010). 3.2. Demographic characteristics of immunocompromised and immunocompetent groups Among a total of 8593 children, 644 children were classified as belonging to the immunocompromised group and 7929 children were classified as immunocompetent. The male proportion, age and DNI value were significantly higher in the immunocompromised group than in the immunocompetent group (P b 0.0001 for all three parameters). WBC counts, haemoglobin, and platelet counts of immunocompromised patients were significantly lower than those of the immunocompetent group (P b 0.0001 for all three markers). CRP was excluded in this analysis due to significant missing data in the immunocompromised group.
2.3. Statistical analysis Statistical analysis was performed with SPSS (ver 18.0). Results were expressed as median values and interquartile range (IQR). Comparisons between groups were evaluated by the nonparametric Mann–Whitney U test for continuous variables and the Pearson χ2 test or Fisher's exact
3.3. Comparison of DNI and other laboratory markers between bacteraemia and non-bacteraemia children by group according to immune status In the immunocompromised group, all parameters including DNI were not different between the bacteraemia and non-bacteraemia
Table 1 Demographic and laboratory characteristics between bacteraemia and non-bacteraemia groups. Characteristics
Bacteraemia (n = 44)
Non-bacteraemia (n = 8549)
Univariate analysis p-value1
Multivariate analysis p-value2
OR (95%CI)
Male Age, y DNI3 WBC (×109/l) Haemoglobin (g/dl) Platelet (×109/l)4 CRP (mg/l)5
33 (75.0) 2.6 (0.4–12.8) 1.2 (0–8.7) 5.04 (0.54–9.76) 10.4 (9.4–11.6) 138 (29–253) 22.6 (8.5–89.9)
4914 (57.5) 2.6 (1.2–5.3) 0 (0–1.1) 10.18 (7.16–14.06) 12.0 (11.1–12.9) 330 (250–423) 8.2 (2.3–28.1)
0.019 0.924 b0.0001 b0.0001 b0.0001 b0.0001 0.004
NS
0.521 (0.185–1.461)
0.001 0.047 NS NS 0.043
1.104 (1.043–1.168) 1.0 (1.0–1.0) 0.902 (0.681–1.197) 1.0 (1.0–1.0) 1.005 (1.0–1.010)
Data presented as median values (interquartile range) or number of patients (percentage). 1 P-values were calculated by Mann–Whitney U test or χ2 test. 2 In a multivariate model: 7633 cases with all parameters that were significant in a univariate analysis were subjected to multivariate logistic regression analysis. 3 DNI, delta neutrophil index. 4 1 data point was missing. 5 959 data points were missing.
J.G. Ahn et al. / Clinica Chimica Acta 436 (2014) 319–322
321
Table 2 Demographic and laboratory characteristics of bacteraemia and non-bacteraemia subgroups according to immune status. Characteristics
Male Age, y DNI WBC (×109/L) Haemoglobin (g/dL) Platelet (×109/L)
P-valuea
Immunocompromised (n = 664) Non-bacteraemia (n = 647)
Bacteraemia (n = 17)
423 (65.4) 6.7 (2.6–12.0) 1.4 (0–5.8) 2.82 (0.42–6.80) 8.8 (7.5–10.6) 70 (26–219)
13 (76.5) 9.2 (1.2–14.7) 1.3 (0–10.8) 0.60 (0.28–5.96) 9.2 (6.8–11.8) 52 (14–150)
0.443 0.72 0.893 0.447 0.929 0.067
P-valuea
Immunocompetent (n = 7929) Non-bacteraemia (n = 7902)
Bacteraemia (n = 27)
4491 (56.8) 2.4 (1.1–4.8) 0 (0–0.9) 10.57 (7.63–14.33) 12.1 (11.3–12.9) 341 (263–430)
20 (74.1) 1.4(0.3–3.6) 1.2(0–9.1) 7.64 (2.76–11.55) 10.5 (10.0–11.6) 209 (58–403)
0.071 0.118 b0.0001 0.004 b0.0001 b0.0001
Abbreviations: DNI, delta neutrophil index; WBC, white blood cell; y, years. Data presented as median values (interquartile range) or number of patients (percentage). a. P-values were calculated by Mann–Whitney U test or Chi-square test or Fisher's exact test.
subgroups, while in the immunocompetent group DNI was significantly higher in the bacteraemia subgroup than in the non-bacteraemia subgroup (P b 0.0001). In addition, WBC counts, haemoglobin, and platelet counts in the immunocompetent group were significantly lower in the bacteraemia subgroup than in the non-bacteraemia subgroup (P b 0.004, P b 0.0001, and P b 0.0001, respectively) (Table 2). Fig. 1 shows the ROC curves for DNI for bacteraemia in the 2 groups. The area under the ROC curve (AUC) in the immunocompetent group was 0.70 (95% CI = 0.58–0.82, P b 0.0001), while the AUC in the immunocompromised group was 0.51 (95% CI = 0.36–0.66, P = 0.897). In the immunocompetent group, the best cutoff value for DNI to predict bacteraemia was 4.4 with sensitivity and specificity of 44.4% (95% CI = 25.5–64.7%) and 94.7% (95% CI = 94.2–95.2%), respectively. 4. Discussion This study demonstrated that DNI could be helpful in predicting bacteraemia in immunocompetent children, but was not useful as a diagnostic tool for bacteraemia in immunocompromised children. Previous DNI studies, in which most immunocompromised patients were excluded, already reported it to be a useful parameter for diagnosing sepsis in elderly patients and neonates [8–11] and for differentiating true bacteraemia from contamination in patients with a positive blood culture [16]. In accordance with previous studies, our results in the immunocompetent group showed that higher DNI values were found in the bacteraemia subgroup when compared to the non-bacteraemia
subgroup. In addition, the AUC of DNI values in the immunocompetent patients were similar to those reported in previous studies [11,16]. DNI reflects the fraction of immature granulocytes in peripheral blood. Under conditions of bacteraemia, immature forms of neutrophils enter the circulation to participate in innate defence against bacteria. The left shift of the myeloid series, which is defined as an elevated immature/total granulocyte ratio, has been used in defining sepsis as one of the inclusion criteria of systemic inflammatory response syndrome (SIRS) [17–19]. When considering that DNI is automatically analysed without any additional time or cost in clinical settings and has already been shown to be well correlated with other parameters such as CRP and procalcitonin (PCT) [8,9,16], DNI analysis may contribute to the prediction of the development of sepsis in the clinical field. Despite its usefulness, there are limitations to DNI for the prediction of bacteraemia in immunocompromised cases. Immunocompromised conditions may influence various hematologic parameters including immature granulocytes [8,17]. In our study, DNI values were significantly higher in the immunocompromised group than in the immunocompetent group. As a result there were no significant differences in DNI values between the bacteraemia and non-bacteraemia subgroups in the immunocompromised group. This finding suggests that under most immunocompromised conditions immature granulocytes can increase due to stimuli from the bone marrow, and DNI analysis for paediatric bacteraemia in such situations is not useful. In addition to patients in the immunocompromised state, neonates, pregnant women, and patients with other hematologic or bone
Fig. 1. ROC curves demonstrating the predictive value of DNI (solid line) together with reference line (dash line) for bacteraemia. (A) ROC curve of DNI for discriminating bacteraemia patients (n = 17) from non-bacteraemia patients (n = 647) in the immunocompromised group. The AUC was 0.51 (95% CI: 0.36–0.66, P = 0.897). (B) ROC curve of DNI for discriminating bacteraemia patients (n = 27) from non-bacteraemia patients (n = 7902) in the immunocompetent group. The AUC was 0.70 (95% CI: 0.58–0.82, P b 0.0001).
322
J.G. Ahn et al. / Clinica Chimica Acta 436 (2014) 319–322
marrow alterations may show abnormal immature granulocytosis [8, 17]. Under these conditions, DNI should be interpreted with caution, compared to other biomarkers such as CRP and PCT. We note that our study had several limitations. First, we did not evaluate the comparative advantage of using DNI over CRP or PCT for predicting bacteraemia. Second, this was a single-centre study and the sample size was relatively small. Third, our study was limited by the use of a specific analyser. However, there are additional studies regarding other analysers that measure immature granulocytes automatically [20–22], suggesting that it is possible to accurately and universally count immature granulocytes. Despite these shortcomings, our data should aid in the understanding of the usefulness of DNI in children with bacteraemia and the difference between the DNI values of immunocompromised and immunocompetent hosts. In conclusion, DNI values appear to be an additional parameter for the early diagnosis of bacteraemia in children. However, because DNI is one of the granular leukocyte differentiation-related markers, there may be a lack of sensitivity or specificity for DNI as a predictive marker of bacteraemia in various conditions where immature granulocytes are elevated, such as in the immunocompromised state. Further studies are needed to evaluate the usefulness of DNI according to specific diseases. References [1] Santschi M, Leclerc F. Management of children with sepsis and septic shock: a survey among pediatric intensivists of the Reseau Mere-Enfant de la Francophonie. Anaesth Intensive Care 2013;3:7. [2] Bates DW, Cook EF, Goldman L, Lee TH. Predicting bacteremia in hospitalized patients. A prospectively validated model. Ann Intern Med 1990;113:495–500. [3] Bates DW, Sands K, Miller E, et al. Predicting bacteremia in patients with sepsis syndrome. Academic Medical Center Consortium Sepsis Project Working Group. J Infect Dis 1997;176:1538–51. [4] Marshall JC. Sepsis: rethinking the approach to clinical research. J Leukoc Biol 2008;83:471–82. [5] Standage SW, Wong HR. Biomarkers for pediatric sepsis and septic shock. Expert Rev Anti Infect Ther 2011;9:71–9.
[6] Gonsalves MD, Sakr Y. Early identification of sepsis. Curr Infect Dis Rep 2010;12:329–35. [7] Kibe S, Adams K, Barlow G. Diagnostic and prognostic biomarkers of sepsis in critical care. J Antimicrob Chemother 2011;66(Suppl. 2):ii33–40. [8] Park BH, Kang YA, Park MS, et al. Delta neutrophil index as an early marker of disease severity in critically ill patients with sepsis. BMC Infect Dis 2011;11:299. [9] Seok Y, Choi JR, Kim J, et al. Delta neutrophil index: a promising diagnostic and prognostic marker for sepsis. Shock 2012;37:242–6. [10] Kim HW, Ku S, Jeong SJ, et al. Delta neutrophil index: could it predict mortality in patients with bacteraemia? Scand J Infect Dis 2012;44:475–80. [11] Lee SM, Eun HS, Namgung R, Park MS, Park KI, Lee C. Usefulness of the delta neutrophil index for assessing neonatal sepsis. Acta Paediatr 2013;102:e13–6. [12] Kratz A, Maloum K, O'Malley C, et al. Enumeration of nucleated red blood cells with the ADVIA 2120 Hematology System: an International Multicenter Clinical Trial. Lab Hematol 2006;12:63–70. [13] Nahm CH, Choi JW, Lee J. Delta neutrophil index in automated immature granulocyte counts for assessing disease severity of patients with sepsis. Ann Clin Lab Sci 2008;38:241–6. [14] Horan TC, Andrus M, Dudeck MA. CDC/NHSN surveillance definition of health careassociated infection and criteria for specific types of infections in the acute care setting. Am J Infect Control 2008;36:309–32. [15] Hall KK, Lyman JA. Updated review of blood culture contamination. Clin Microbiol Rev 2006;19:788–802. [16] Lee CH, Kim J, Park Y, et al. Delta neutrophil index discriminates true bacteremia from blood culture contamination. Clin Chim Acta 2014;427:11–4. [17] Cornbleet PJ. Clinical utility of the band count. Clin Lab Med 2002;22:101–36. [18] Herzum I, Renz H. Inflammatory markers in SIRS, sepsis and septic shock. Curr Med Chem 2008;15:581–7. [19] Taneja R, Sharma AP, Hallett MB, Findlay GP, Morris MR. Immature circulating neutrophils in sepsis have impaired phagocytosis and calcium signaling. Shock 2008;30:618–22. [20] Briggs C, Kunka S, Fujimoto H, Hamaguchi Y, Davis BH, Machin SJ. Evaluation of immature granulocyte counts by the XE-IG master: upgraded software for the XE-2100 automated hematology analyzer. Lab Hematol 2003;9:117–24. [21] Fujimoto H, Sakata T, Hamaguchi Y, et al. Flow cytometric method for enumeration and classification of reactive immature granulocyte populations. Cytometry 2000;42:371–8. [22] Ansari-Lari MA, Kickler TS, Borowitz MJ. Immature granulocyte measurement using the Sysmex XE-2100. Relationship to infection and sepsis. Am J Clin Pathol 2003;120:795–9.
Contents lists available at ScienceDirect
Clinica Chimica Acta journal homepage: www.elsevier.com/locate/clinchim
Limitation of the delta neutrophil index for assessing bacteraemia in immunocompromised children Jong Gyun Ahn a,b, Seong Yeol Choi c, Dong Soo Kim b,c, Ki Hwan Kim b,c,⁎ a b c
Department of Pediatrics, School of Medicine, Ewha Womans University, Seoul, Republic of Korea Department of Medicine, The Graduate School of Yonsei University, Seoul, Republic of Korea Department of Pediatrics, Severance Children's Hospital, Yonsei University College of Medicine, Seoul, Republic of Korea
a r t i c l e
i n f o
Article history: Received 3 February 2014 Received in revised form 29 April 2014 Accepted 19 June 2014 Available online 27 June 2014 Keywords: Delta neutrophil index Bacteraemia Children Immunocompromised
a b s t r a c t Background: Delta neutrophil index (DNI) has been reported to be useful in assessing the prognosis of suspected bacteraemia in adults. However, data in children are limited. The purpose of this study was to investigate the correlation between the DNI and bacteraemia in children and to compare DNI values between immunocompetent and immunocompromised hosts. Methods: We retrospectively collected data on 8593 children who underwent simultaneous blood culture and DNI testing. DNI was determined with a blood cell analyser (ADVIA 120, Siemens, Inc.). The children were divided into immunocompromised (n = 664) and immunocompetent groups (n = 7929). Results: DNI was higher in the bacteraemia group than in the non-bacteraemia group (P b 0.001). According to the group, DNI was higher in immunocompromised than immunocompetent patients (P b 0.001). Within the immunocompromised group, there was no significant difference between the bacteraemia and non-bacteraemia subgroups due to a higher DNI as compared to the immunocompetent group. Conclusions: DNI could be an additional method for the early diagnosis of bacteraemia in children. However, the use of DNI for the prediction of bacteraemia in immunocompromised cases has limitations. Further studies on the usefulness of DNI according to specific diseases are needed. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Bacteraemia is one of the most serious and life-threatening childhood diseases. The frequency of fatal bacteraemia is increased in immunocompromised individuals. Although early diagnosis and treatment is crucial for better outcomes [1], bacteraemia is identified in only about 30% of patients with sepsis, depending on previous antibiotic history, and microbiologic cultures take at least 24–72 h for adequate growth [2–4]. Several biomarkers that are used to aid rapid identification of patients with bacteraemia, such as C-reactive protein (CRP), procalcitonin and various interleukins, respond to both infection and non-infectious inflammation [5]. The search for more reliable and specific biomarkers for early diagnosis of bacteraemia is therefore critically important [6,7]. In recent years, the delta neutrophil index (DNI) has been shown to be useful for early diagnosis and prognostic assessment of adult and neonate patients with sepsis [8–11]. DNI is calculated by an automatic counter (ADVIA 120; Siemens Healthcare Diagnostics) using myeloperoxidase (MPO) channels and nuclear lobularity channels in the test system [12]. The difference between the leukocyte differentials assayed in the two ⁎ Corresponding author at: Department of Pediatrics, Severance Children's Hospital, Yonsei University College of Medicine, 50 Yonsei-ro, Seodaemungu, Seoul 120–752, Republic of Korea. Tel.: +82 2 2228 2050; fax: +82 2 393 9118. E-mail address: [email protected] (K.H. Kim).
http://dx.doi.org/10.1016/j.cca.2014.06.020 0009-8981/© 2014 Elsevier B.V. All rights reserved.
channels is designated as the DNI, which correlates with the fraction of immature granulocytes in peripheral blood calculated by manual counting. DNI is strongly associated with the disseminated intravascular coagulation score, positive blood culture rate, and mortality in patients with suspected sepsis [13]. The data in children, however, are limited and little is known about the usefulness of DNI in evaluating bacteraemia in immunocompromised patients. 2. Materials and methods 2.1. Study population and specimen collection Between June 2010 and May 2011, we retrospectively reviewed medical records of 8593 patients aged b 19 y who had laboratory requests for both blood culture and DNI under suspicion of having bacteraemia at Severance Children's Hospital in Korea. The children were divided into immunocompromised (n = 664) and immunocompetent groups (n = 7929). The immunocompromised state was defined as current treatment with chemotherapy or immunosuppressive therapy, hematopoietic stem cell or solid organ transplantation, or known primary immunodeficiency. Bacteraemia was defined as laboratoryconfirmed blood stream infection using the criteria of the US Centers for Diseases Control and Prevention/National Healthcare Safety Network surveillance [14]. The criteria are as follows: (1) a recognized
320
J.G. Ahn et al. / Clinica Chimica Acta 436 (2014) 319–322
pathogen, which does not include organisms that are considered common skin contaminants, cultured from 1 or more blood cultures, and (2) at least 1 of the following signs or symptoms: fever (N 38 °C), chills, or hypotension(in case of patient ≤1 year of age, at least 1 of the following signs or symptoms: fever (N 38 °C, rectal), hypothermia (b37 °C, rectal), apnoea, or bradycardia). Patients with blood culture yielding organisms considered contaminants were excluded. In accordance with the criteria described previously, positive blood culture for nonspecified streptococci, coagulase-negative staphylococci, Bacillus sp. other than Bacillus anthracis, diphtheroids sp., Corynebacterium sp., Propionibacterium acnes, Micrococcus sp., and Clostridium perfringens, or Streptococcus mitis isolated in a single set of culture bottles were regarded as contaminants [15,16]. All blood culture negative patients were included in the non-bacteraemia group. The study was approved by the Institutional Review Board of Severance Hospital and informed consent was waived. 2.2. Blood culture, Delta neutrophil index (DNI) and other laboratory data Blood cultures were performed in more than one pair of bottles for aerobic and anaerobic cultures, respectively, and incubated in an automated blood culture system (BacT/Alert 3D; bioMérieux), capable of detecting the growth of any organism. When bacterial growth was noted, a small volume of cultured specimen was inoculated onto both blood and MacConkey agar plates that were cultured overnight at 37 °C in an incubator enriched with 5% CO2. Isolates were identified based on typical colony morphology, Gram staining, and biochemical tests, and were further verified by the Vitek Identification System (bioMérieux). When no bacterial growth was detected within five days after primary inoculation of the blood sample, the result of the blood culture was considered negative. Data of complete blood cell (CBC) counts were collected with an automated blood cell analyser (ADVIA 2120i; Siemens). DNI was calculated using the following formula that was introduced in a previous study: DN = (the leucocyte subfraction assayed in the MPO channel by cytochemical reaction) – (the leucocyte subfraction counted in the nuclear lobularity channel by the reflected light beam) [13]. Serum CRP concentration (normal range 0–8 mg/L) was measured using the nephelometric method (Beckman Coulter, Fullerton, CA, USA). Among a total of 8593 cases, 959 cases did not receive a CRP test and the platelet count of one case was missing.
test for categorical variables. Potential risk factors for sepsis were identified by univariate analysis. Significant variables were further subjected to multivariate logistic regression analysis. Receiver operating characteristic (ROC) curves of DNI between groups were drawn and compared. A p b 0.05 were considered statistically significant. 3. Results 3.1. Comparison of bacteraemia and non-bacteraemia groups During the study period, data on a total of 8593 patients were collected. Fifty-eight percent of the subjects (4947/8593) were male and the median age of children was 2.6 years (range: 0.1–18 y). Among a total of 8593 cases, 44 children were diagnosed with blood cultureproven bacteraemia. The comparison between the bacteraemia and non-bacteraemia groups is summarized in Table 1. The proportion of male patients in the bacteraemia group and the non-bacteraemia group was 75.0% and 57.5%, respectively. The median ages of the bacteraemia and non-bacteraemia patients were 2.6 (IQR, 0.4–12.8) and 2.6 (IQR, 1.2–5.3) y, respectively, with no significant difference. In a univariate analysis, the concentrations of both DNI and CRP in the bacteraemia group were higher than those in the non-bacteraemia groups (P b 0.0001, P = 0.004, respectively), while white blood cell (WBC) counts, haemoglobin concentration, and platelet counts in the bacteraemia group were lower than those in the non-bacteraemia groups (P b 0.0001 for all 3 markers). In a multivariate model, significant risk factors for bacteraemia were DNI (P = 0.001; 95% CI = 1.043–1.168), WBC counts (P = 0.047; 95% CI = 1.000–1.000), and CRP (P = 0.043; 95% CI: 1.000–1.010). 3.2. Demographic characteristics of immunocompromised and immunocompetent groups Among a total of 8593 children, 644 children were classified as belonging to the immunocompromised group and 7929 children were classified as immunocompetent. The male proportion, age and DNI value were significantly higher in the immunocompromised group than in the immunocompetent group (P b 0.0001 for all three parameters). WBC counts, haemoglobin, and platelet counts of immunocompromised patients were significantly lower than those of the immunocompetent group (P b 0.0001 for all three markers). CRP was excluded in this analysis due to significant missing data in the immunocompromised group.
2.3. Statistical analysis Statistical analysis was performed with SPSS (ver 18.0). Results were expressed as median values and interquartile range (IQR). Comparisons between groups were evaluated by the nonparametric Mann–Whitney U test for continuous variables and the Pearson χ2 test or Fisher's exact
3.3. Comparison of DNI and other laboratory markers between bacteraemia and non-bacteraemia children by group according to immune status In the immunocompromised group, all parameters including DNI were not different between the bacteraemia and non-bacteraemia
Table 1 Demographic and laboratory characteristics between bacteraemia and non-bacteraemia groups. Characteristics
Bacteraemia (n = 44)
Non-bacteraemia (n = 8549)
Univariate analysis p-value1
Multivariate analysis p-value2
OR (95%CI)
Male Age, y DNI3 WBC (×109/l) Haemoglobin (g/dl) Platelet (×109/l)4 CRP (mg/l)5
33 (75.0) 2.6 (0.4–12.8) 1.2 (0–8.7) 5.04 (0.54–9.76) 10.4 (9.4–11.6) 138 (29–253) 22.6 (8.5–89.9)
4914 (57.5) 2.6 (1.2–5.3) 0 (0–1.1) 10.18 (7.16–14.06) 12.0 (11.1–12.9) 330 (250–423) 8.2 (2.3–28.1)
0.019 0.924 b0.0001 b0.0001 b0.0001 b0.0001 0.004
NS
0.521 (0.185–1.461)
0.001 0.047 NS NS 0.043
1.104 (1.043–1.168) 1.0 (1.0–1.0) 0.902 (0.681–1.197) 1.0 (1.0–1.0) 1.005 (1.0–1.010)
Data presented as median values (interquartile range) or number of patients (percentage). 1 P-values were calculated by Mann–Whitney U test or χ2 test. 2 In a multivariate model: 7633 cases with all parameters that were significant in a univariate analysis were subjected to multivariate logistic regression analysis. 3 DNI, delta neutrophil index. 4 1 data point was missing. 5 959 data points were missing.
J.G. Ahn et al. / Clinica Chimica Acta 436 (2014) 319–322
321
Table 2 Demographic and laboratory characteristics of bacteraemia and non-bacteraemia subgroups according to immune status. Characteristics
Male Age, y DNI WBC (×109/L) Haemoglobin (g/dL) Platelet (×109/L)
P-valuea
Immunocompromised (n = 664) Non-bacteraemia (n = 647)
Bacteraemia (n = 17)
423 (65.4) 6.7 (2.6–12.0) 1.4 (0–5.8) 2.82 (0.42–6.80) 8.8 (7.5–10.6) 70 (26–219)
13 (76.5) 9.2 (1.2–14.7) 1.3 (0–10.8) 0.60 (0.28–5.96) 9.2 (6.8–11.8) 52 (14–150)
0.443 0.72 0.893 0.447 0.929 0.067
P-valuea
Immunocompetent (n = 7929) Non-bacteraemia (n = 7902)
Bacteraemia (n = 27)
4491 (56.8) 2.4 (1.1–4.8) 0 (0–0.9) 10.57 (7.63–14.33) 12.1 (11.3–12.9) 341 (263–430)
20 (74.1) 1.4(0.3–3.6) 1.2(0–9.1) 7.64 (2.76–11.55) 10.5 (10.0–11.6) 209 (58–403)
0.071 0.118 b0.0001 0.004 b0.0001 b0.0001
Abbreviations: DNI, delta neutrophil index; WBC, white blood cell; y, years. Data presented as median values (interquartile range) or number of patients (percentage). a. P-values were calculated by Mann–Whitney U test or Chi-square test or Fisher's exact test.
subgroups, while in the immunocompetent group DNI was significantly higher in the bacteraemia subgroup than in the non-bacteraemia subgroup (P b 0.0001). In addition, WBC counts, haemoglobin, and platelet counts in the immunocompetent group were significantly lower in the bacteraemia subgroup than in the non-bacteraemia subgroup (P b 0.004, P b 0.0001, and P b 0.0001, respectively) (Table 2). Fig. 1 shows the ROC curves for DNI for bacteraemia in the 2 groups. The area under the ROC curve (AUC) in the immunocompetent group was 0.70 (95% CI = 0.58–0.82, P b 0.0001), while the AUC in the immunocompromised group was 0.51 (95% CI = 0.36–0.66, P = 0.897). In the immunocompetent group, the best cutoff value for DNI to predict bacteraemia was 4.4 with sensitivity and specificity of 44.4% (95% CI = 25.5–64.7%) and 94.7% (95% CI = 94.2–95.2%), respectively. 4. Discussion This study demonstrated that DNI could be helpful in predicting bacteraemia in immunocompetent children, but was not useful as a diagnostic tool for bacteraemia in immunocompromised children. Previous DNI studies, in which most immunocompromised patients were excluded, already reported it to be a useful parameter for diagnosing sepsis in elderly patients and neonates [8–11] and for differentiating true bacteraemia from contamination in patients with a positive blood culture [16]. In accordance with previous studies, our results in the immunocompetent group showed that higher DNI values were found in the bacteraemia subgroup when compared to the non-bacteraemia
subgroup. In addition, the AUC of DNI values in the immunocompetent patients were similar to those reported in previous studies [11,16]. DNI reflects the fraction of immature granulocytes in peripheral blood. Under conditions of bacteraemia, immature forms of neutrophils enter the circulation to participate in innate defence against bacteria. The left shift of the myeloid series, which is defined as an elevated immature/total granulocyte ratio, has been used in defining sepsis as one of the inclusion criteria of systemic inflammatory response syndrome (SIRS) [17–19]. When considering that DNI is automatically analysed without any additional time or cost in clinical settings and has already been shown to be well correlated with other parameters such as CRP and procalcitonin (PCT) [8,9,16], DNI analysis may contribute to the prediction of the development of sepsis in the clinical field. Despite its usefulness, there are limitations to DNI for the prediction of bacteraemia in immunocompromised cases. Immunocompromised conditions may influence various hematologic parameters including immature granulocytes [8,17]. In our study, DNI values were significantly higher in the immunocompromised group than in the immunocompetent group. As a result there were no significant differences in DNI values between the bacteraemia and non-bacteraemia subgroups in the immunocompromised group. This finding suggests that under most immunocompromised conditions immature granulocytes can increase due to stimuli from the bone marrow, and DNI analysis for paediatric bacteraemia in such situations is not useful. In addition to patients in the immunocompromised state, neonates, pregnant women, and patients with other hematologic or bone
Fig. 1. ROC curves demonstrating the predictive value of DNI (solid line) together with reference line (dash line) for bacteraemia. (A) ROC curve of DNI for discriminating bacteraemia patients (n = 17) from non-bacteraemia patients (n = 647) in the immunocompromised group. The AUC was 0.51 (95% CI: 0.36–0.66, P = 0.897). (B) ROC curve of DNI for discriminating bacteraemia patients (n = 27) from non-bacteraemia patients (n = 7902) in the immunocompetent group. The AUC was 0.70 (95% CI: 0.58–0.82, P b 0.0001).
322
J.G. Ahn et al. / Clinica Chimica Acta 436 (2014) 319–322
marrow alterations may show abnormal immature granulocytosis [8, 17]. Under these conditions, DNI should be interpreted with caution, compared to other biomarkers such as CRP and PCT. We note that our study had several limitations. First, we did not evaluate the comparative advantage of using DNI over CRP or PCT for predicting bacteraemia. Second, this was a single-centre study and the sample size was relatively small. Third, our study was limited by the use of a specific analyser. However, there are additional studies regarding other analysers that measure immature granulocytes automatically [20–22], suggesting that it is possible to accurately and universally count immature granulocytes. Despite these shortcomings, our data should aid in the understanding of the usefulness of DNI in children with bacteraemia and the difference between the DNI values of immunocompromised and immunocompetent hosts. In conclusion, DNI values appear to be an additional parameter for the early diagnosis of bacteraemia in children. However, because DNI is one of the granular leukocyte differentiation-related markers, there may be a lack of sensitivity or specificity for DNI as a predictive marker of bacteraemia in various conditions where immature granulocytes are elevated, such as in the immunocompromised state. Further studies are needed to evaluate the usefulness of DNI according to specific diseases. References [1] Santschi M, Leclerc F. Management of children with sepsis and septic shock: a survey among pediatric intensivists of the Reseau Mere-Enfant de la Francophonie. Anaesth Intensive Care 2013;3:7. [2] Bates DW, Cook EF, Goldman L, Lee TH. Predicting bacteremia in hospitalized patients. A prospectively validated model. Ann Intern Med 1990;113:495–500. [3] Bates DW, Sands K, Miller E, et al. Predicting bacteremia in patients with sepsis syndrome. Academic Medical Center Consortium Sepsis Project Working Group. J Infect Dis 1997;176:1538–51. [4] Marshall JC. Sepsis: rethinking the approach to clinical research. J Leukoc Biol 2008;83:471–82. [5] Standage SW, Wong HR. Biomarkers for pediatric sepsis and septic shock. Expert Rev Anti Infect Ther 2011;9:71–9.
[6] Gonsalves MD, Sakr Y. Early identification of sepsis. Curr Infect Dis Rep 2010;12:329–35. [7] Kibe S, Adams K, Barlow G. Diagnostic and prognostic biomarkers of sepsis in critical care. J Antimicrob Chemother 2011;66(Suppl. 2):ii33–40. [8] Park BH, Kang YA, Park MS, et al. Delta neutrophil index as an early marker of disease severity in critically ill patients with sepsis. BMC Infect Dis 2011;11:299. [9] Seok Y, Choi JR, Kim J, et al. Delta neutrophil index: a promising diagnostic and prognostic marker for sepsis. Shock 2012;37:242–6. [10] Kim HW, Ku S, Jeong SJ, et al. Delta neutrophil index: could it predict mortality in patients with bacteraemia? Scand J Infect Dis 2012;44:475–80. [11] Lee SM, Eun HS, Namgung R, Park MS, Park KI, Lee C. Usefulness of the delta neutrophil index for assessing neonatal sepsis. Acta Paediatr 2013;102:e13–6. [12] Kratz A, Maloum K, O'Malley C, et al. Enumeration of nucleated red blood cells with the ADVIA 2120 Hematology System: an International Multicenter Clinical Trial. Lab Hematol 2006;12:63–70. [13] Nahm CH, Choi JW, Lee J. Delta neutrophil index in automated immature granulocyte counts for assessing disease severity of patients with sepsis. Ann Clin Lab Sci 2008;38:241–6. [14] Horan TC, Andrus M, Dudeck MA. CDC/NHSN surveillance definition of health careassociated infection and criteria for specific types of infections in the acute care setting. Am J Infect Control 2008;36:309–32. [15] Hall KK, Lyman JA. Updated review of blood culture contamination. Clin Microbiol Rev 2006;19:788–802. [16] Lee CH, Kim J, Park Y, et al. Delta neutrophil index discriminates true bacteremia from blood culture contamination. Clin Chim Acta 2014;427:11–4. [17] Cornbleet PJ. Clinical utility of the band count. Clin Lab Med 2002;22:101–36. [18] Herzum I, Renz H. Inflammatory markers in SIRS, sepsis and septic shock. Curr Med Chem 2008;15:581–7. [19] Taneja R, Sharma AP, Hallett MB, Findlay GP, Morris MR. Immature circulating neutrophils in sepsis have impaired phagocytosis and calcium signaling. Shock 2008;30:618–22. [20] Briggs C, Kunka S, Fujimoto H, Hamaguchi Y, Davis BH, Machin SJ. Evaluation of immature granulocyte counts by the XE-IG master: upgraded software for the XE-2100 automated hematology analyzer. Lab Hematol 2003;9:117–24. [21] Fujimoto H, Sakata T, Hamaguchi Y, et al. Flow cytometric method for enumeration and classification of reactive immature granulocyte populations. Cytometry 2000;42:371–8. [22] Ansari-Lari MA, Kickler TS, Borowitz MJ. Immature granulocyte measurement using the Sysmex XE-2100. Relationship to infection and sepsis. Am J Clin Pathol 2003;120:795–9.
