1. Introduction
Neonatal sepsis is a common cause of morbidity and mortality among newborns [
1]. It presents nonspecific clinical signs and symptoms, such as respiratory distress, hypotension, apnoea, patent duct arteriosus, and necrotizing enterocolitis; thus, it is difficult to diagnose [
2,
3]. Sepsis is caused by the extreme response of the host to infection. Furthermore, neonatal sepsis has been defined as an infection of a sterile site (e.g., blood, urine, or cerebrospinal fluid) accompanied by a systemic inflammatory response that can manifest within the first 72 h of life (early-onset sepsis). Symptom onset in late-onset sepsis occurs at >72 h of life [
4].
One study performed in Saudi Arabia demonstrated that the main cause of neonatal sepsis in prematurity and caesarean operation cases was Gram-negative bacteria, which was also the main cause of deaths in early-onset neonatal sepsis (EONS) and late-onset neonatal sepsis (LONS). Moreover, leucocytosis, higher C-reactive protein (CRP), and thrombocytopenia were reportedly significant sepsis markers, especially in LONS [
5]. Meanwhile, two studies investigated the colonization rate of group B streptococcus in pregnant women and women in labour [
6,
7]. Several types of pathogens are incriminated in pathogenic sepsis. Although bacterial infection is still the main cause of pathogenic sepsis, viral and fungal infections represent an important percentage of sepsis aetiologies, especially in immunocompromised patients [
8,
9]. The World Health Organisation classified sepsis as a global health priority in 2021 [
10]. Moreover, each hour of delay in treatment raises the mortality burden of sepsis between 7% and 10% [
11]. Thus, early and rapid diagnosis of sepsis is crucial. However, the diagnosis of patients at risk of sepsis is challenging because it is a dynamic disease with multiple inflammatory responses based on different pathogens and immune statuses [
12]. Several studies have been conducted to detect reliable biomarkers for predicting outcomes and evaluating treatment responses; however, not all are available for routine clinical use [
13]. Nevertheless, the most common biomarkers employed in sepsis diagnosis are the CRP and procalcitonin (PCT) [
3].
In addition to the traditional complete blood count (CBC) parameters, a new haematological parameter analyser has emerged, which is called cell population data (CPD). The monocyte distribution width (MDW) is a CPD parameter that reflects the morphological and functional characteristics of monocytes (MO) [
14]. Moreover, the Food and Drug Administration recently approved the MDW as an early sepsis indicator [
15]. Since MOs are one of the first responders against infection, the MDW was proposed to be a novel biomarker of sepsis in the ED under the definition of Sepsis 2 criteria [
15]. Several studies have been conducted to prove the diagnostic performance of the MDW in predicting sepsis, especially in the ED and intensive care unit [
16,
17,
18], and they have shown good results. Furthermore, the diagnostic performance of the MDW, CRP, and PCT as predictors of sepsis have been studied in association and individually, and the results have shown that the diagnostic ability of the MDW was not lower than those of the CRP and PCT in terms of the areas under the curve (AUCs) [
16]. However, the CRP is insensitive and non-specific for LONS diagnosis [
19]. By contrast, the PCT’s discriminative value plays a role in detecting the severity and mortality of LONS. This study aimed to identify more sensitive and applicable peripheral haematological biomarkers in neonatal infection.
2. Materials and Methods
2.1. Study Design, Setting, and Population
A cross-sectional study was conducted at the King Saud University Medical City (KSUMC) hospital neonatal intensive care unit (NICU) in Riyadh, Saudi Arabia from September 2019 to August 2020. The study included 147 Saudi Arabian neonates (72 males and 75 females) with a mean age of 3.17 days at the admission time to NICU. Neonatal blood samples were categorized into the following: the control group, which included neonates with normal blood values and who were free from health complications, and the study group, which included neonates who were admitted with signs of sepsis and abnormal blood values. In the study group, the samples were subdivided according to their susceptibility to developing sepsis: the sepsis group and the other health complication (OHC) group.
The OHC is defined in the current study as comprising cases with negative results for blood culture and other developed health problems, such as respiratory distress syndrome, jaundice, decreased activity, low birth weight, transient tachypnoea of the newborn, hyperglycaemia, and hypoglycaemia. At study initiation, every taken blood sample was first subjected to a blood culture assay to be confirmed as whether it is sepsis, along with suggestive clinical signs. Septic microorganisms were isolated from blood culture assay using the BD Bactec analyser (blood culture system). Among the 147 newborns, 77 blood samples from neonates (39 males and 38 females) born at KSUMC had clinical symptoms and sepsis as determined by a blood culture assay; they were therefore included in the study group. In the control group, a total of 70 blood samples from healthy newborns (36 males and 34 females) born at KSUMC were collected (
Table 1).
2.2. Biomarkers
Two-millilitre samples of venous blood from each neonate from each group were collected in ethylenediaminetetraacetic acid microtainer tubes for CBC assay and leucocyte differential count (white blood cell count (WBC), red blood cell count (RBCs), haemoglobin (Hgb), haematocrit (Hct), mean cell volume (MCV), mean corpuscular haemoglobin (MCH), red cell distribution width (RDW), platelet (PLT), mean platelet volume (MPV), MDW, neutrophil (NE), lymphocyte (LY), eosinophil (EO), MO, and basophil (BA)). Not Applicable (N/A), CBC and leucocyte differential count were performed and analysed on a full automated haematological analyser (Unicel DxH800 analyzer) (Beckman coulter Inc., Miami, FL, USA) Version 3.0 software (Beckman Coulter 2009, Miami, FL, USA).
2.3. Inclusion and Exclusion Criteria
All the inclusion and exclusion markers are summarized in
Table 2.
2.4. Statistical Analyses
Statistical analysis was performed using IBM SPSS Statistics for Windows, version 22 (IBM Corp., Armonk, NY, USA) and GraphPad Prism version 8.0 (GraphPad, Boston, MA, USA) for windows. Continuous variables are reported as mean ± standard error of the mean (SEM) according to their group. Categorical variables with corresponding percentages were identified as frequencies. An unpaired
t-test (independent sample
t-test) was used to detect differences in the categorical variables between the groups. The diagnostic ability of a haematological parameter was assessed based on sensitivity, specificity, predictive positive value (PPV), and predictive negative value (PNV), which were computed as follows:
Moreover, the chi-square test depicting the correlation of sepsis with the haematological aspects was performed in this study. Additionally, the AUC for haematological parameters and established cut point for MDW was determined through receiver operating characteristic (ROC) curve analysis. A p-value of <0.05 was considered significant for all statistical parameters.
4. Discussion
Sepsis is a life-threatening disease with widely variable symptoms, ranging from silent signs during the early stages [
19] to severe and sudden [
20]. In this study, baseline characteristics that may shape the risk factors of neonatal sepsis occurrence [
21] were considered.
Bacterial infection remains a substantial cause of neonatal sepsis. In the current study, Gram-positive organisms (62.5%) were the predominant pathogens, while Gram-negative organisms (37.5%) were less frequent. This is consistent with a study that included eight Arabic countries and reported that Gram-positive organisms were the primary cause in Saudi Arabia, Bahrain and Kuwait, and the United Arab of Emirates. Contrarily, a Gram-negative organism was the main cause in Egypt, Iraq, Jordan, and Libya [
22]. Therefore, we aimed to identify the more sensitive and applicable peripheral haematological biomarkers in the neonatal infection.
Blood culture has long been the gold standard, although it is time consuming in the diagnosis of sepsis. Moreover, many clinicians often start antibiotic treatment, although the culture results are undecided. Additionally, acute reactant protein (CRP or PCT) is usually a useful marker for septic disease diagnosis, but its short half-life is a drawback. Therefore, many studies have been performed to determine reliable sepsis biomarkers.
Hundreds of biomarkers for predicting sepsis are currently being discussed [
23]. However, many limitations remain, such as sensitivity, specificity, diagnostic performance, elevated cost, and time required. Moreover, these biomarkers may have less significant results due to some inflammatory conditions in the absence of sepsis [
18,
24,
25].
Generally, our results showed that most of the haematological parameters (RBC count, Hgb, HCT, MCV, MCH, PLT count, WBC count, MDW, NE%) were significantly correlated with sepsis. However, our results emphasised the greatness of MDW as a pre-diagnostic tool like a haematological parameter. This was similar to other findings [
15,
26]. Nonetheless, this finding supports the use of MDW as a screening haematological parameter [
18]. Combining WBC count with MDW was an effective predictor of sepsis. This confirms the idea that there is no one parameter that combines the required sensitivity and specificity to precisely diagnose sepsis [
27,
28]. Furthermore, almost half of the confirmed sepsis cases had decreased WBC counts (leukopenia). Christoph et al. [
29] demonstrated a statistically significant association between WBC count and sepsis. In the same context, leukopenia is more indicative than leucocytosis; however, blood samples should be taken within 4 to 6 h of stimulation because the number of WBCs increases in the late stage of sepsis, which is one of the drawbacks of WBCs [
30,
31,
32].
Overall, the cut-off MDW value was 23, lower than the estimated value (26.63). The MDW value might vary according to the clinical department where the study was conducted, as well as the degree of severity of the cases [
33]. Severe thrombocytopenia was detected in the dominant sepsis cases, whereas moderate and mild thrombocytopenia occurred in the lowest and average sepsis cases, respectively. Thus, PLT may be a decent marker in sepsis, which is consistent with other findings [
34].
Furthermore, reduced RBC count, Hgb concentration, MCV, and MCH values were clearly significant in more than half of the sepsis cases, in agreement with previous outcomes [
35]. In the same context, CBC is considered an acceptable tool in case of sepsis as it is routinely ordered as the first diagnostic test in any clinic department. Moreover, it is feasible, not expensive, and requires less time [
18]. By contrast, CBCs were considered poor diagnostic markers because of their poor sensitivity [
29].
The noteworthy drop in MDW values after the antibiotic treatment is significant. MDW represents a significant association between disease severity, mortality, and treatment outcome among sepsis survivors in comparison with non-survivors, highlighting the association between MDW and mortality [
33]. Despite its modest sensitivity and low false-positive rate, this study confirmed that MDW dramatically improves diagnostic performance by enhancing its specificity without compromising its sensitivity [
18].
The key peripheral blood parameters as newborn sepsis indicators between the healthy group, sepsis, and OHC groups are comprehensively analysed in the current study, which is its strength. However, because we only included data from one institution, our sample size was limited. To more thoroughly analyse the variations between early-onset and late-onset sepsis, a larger sample size would be helpful.