Predictive model for bacterial late-onset neonatal sepsis in a tertiary care hospital in Thailand

Early detection of late-onset neonatal sepsis from noninvasive biosignals using deep learning: A multicenter prospective development and validation study

BACKGROUND

Neonatal sepsis is responsible for significant morbidity and mortality worldwide. Its accurate and timely diagnosis is hindered by vague symptoms and the urgent necessity for early antibiotic intervention. The gold standard for diagnosing the condition is the identification of a pathogenic organism from normally sterile sites via laboratory testing. However, this method is resource-intensive and cannot be conducted continuously.

OBJECTIVE

This study aimed to predict the onset of late-onset sepsis (LOS) with good diagnostic value as early as possible using non-invasive biosignal measurements from neonatal intensive care unit (NICU) monitors.

METHODS

In this prospective multicenter study, we developed a multimodal machine learning algorithm based on a convolutional neural network (CNN) structure that uses the power spectral density (PSD) of recorded biosignals to predict the onset of LOS. This approach aimed to discern LOS-related pathogenic spectral signatures without labor-intensive manual artifact removal.

RESULTS

The model achieved an area under the receiver operating characteristic score of 0.810 (95 % CI 0.698-0.922) on the validation dataset. With an optimal operating point, LOS detection had 83 % sensitivity and 73 % specificity. The median early detection was 44 h before clinical suspicion. The results highlighted the additive importance of electrocardiogram and respiratory impedance (RESP) signals in improving predictive accuracy. According to a more detailed analysis, the predictive power arose from the morphology of the electrocardiogram's R-wave and sudden changes in the RESP signal.

CONCLUSION

Raw biosignals from NICU monitors, in conjunction with PSD transformation, as input to the CNN, can provide state-of-the-art prediction performance for LOS without the need for artifact removal. To the knowledge of the authors, this is the first study to highlight the independent and additive predictive potential of electrocardiogram R-wave morphology and concurrent, sudden changes in the RESP waveform in predicting the onset of LOS using non-invasive biosignals.

An Overview of Antibiotic Therapy for Early- and Late-Onset Neonatal Sepsis: Current Strategies and Future Prospects

Neonatal sepsis is a clinical syndrome mainly associated with a bacterial infection leading to severe clinical manifestations that could be associated with fatal sequalae. According to the time of onset, neonatal sepsis is categorized as early- (EOS) or late-onset sepsis (LOS). Despite blood culture being the gold standard for diagnosis, it has several limitations, and early diagnosis is not immediate. Consequently, most infants who start empirical antimicrobial therapy do not have an underlying infection. Despite stewardship programs partially reduced this negative trend, in neonatology, antibiotic overuse still persists, and it is associated with several relevant problems, the first of which is the increase in antimicrobial resistance (AMR). Starting with these considerations, we performed a narrative review to summarize the main findings and the future prospects regarding antibiotics use to treat neonatal sepsis. Because of the impact on morbidity and mortality that EOS and LOS entail, it is essential to start an effective and prompt treatment as soon as possible. The use of targeted antibiotics is peremptory as soon as the pathogen in the culture is detected. Although prompt therapy is essential, it should be better assessed whether, when and how to treat neonates with antibiotics, even those at higher risk. Considering that we are certainly in the worrying era defined as the "post-antibiotic era", it is still essential and urgent to define novel strategies for the development of antibacterial compounds with new targets or mechanisms of action. A future strategy could also be to perform well-designed studies to develop innovative algorithms for improving the etiological diagnosis of infection, allowing for more personalized use of the antibiotics to treat EOS and LOS.

Early Diagnosis of Late-Onset Neonatal Sepsis Using a Sepsis Prediction Score

Sepsis represents a common cause of morbidity in the Neonatal Intensive Care Unit (NICU). Our objective was to assess the value of clinical and laboratory parameters in predicting septicemia (positive blood culture) in NICU infants. In the first part of the present study (derivation cohort) we retrospectively reviewed the clinical files of 120 neonates with symptoms of suspected sepsis and identified clinical and laboratory parameters associated with proven sepsis on the day the blood culture was taken, as well as 24 h and 48 h earlier. These parameters were combined into a sepsis prediction score (SPS). Subsequently (validation study), we prospectively validated the performance of the SPS in a cohort of 145 neonates. The identified parameters were: temperature instability, platelet count < 150,000/mm³, feeding volume decrease > 20%, changes in blood glucose > 50%, CRP > 1 mg/dL, circulatory and respiratory deterioration. In the retrospective cohort, on the day the blood culture was obtained, a SPS ≥ 3 could predict sepsis with 82.54% sensitivity, 85.96% specificity, 5.88 PLR (Positive Likelihood Ratio), 0.20 NLR (Negative Likelihood Ratio), 86.67% PPV (Positive Predictive Value), 81.67% NPV (Negative Predictive Value) and 84.17% accuracy. In the prospective cohort, on the day the blood culture was obtained, a SPS ≥ 3 could predict sepsis with 76.60% sensitivity, 72.55% specificity, 2.79 PLR, 0.32 NLR, 83.72% PPV, 62.71% NPV and 75.17% accuracy. We concluded that this combination of clinical and laboratory parameters may assist in the prediction of septicemia in NICUs.

Prevalence and determinants of early onset neonatal sepsis at two selected public referral hospitals in the Northwest Ethiopia: a cross-sectional study

INTRODUCTION

Globally, neonatal mortality is decreasing, and road maps such as the Early Newborn Action Plan set ambitious targets for 2030. Despite this, deaths in the first weeks of life continue to rise as a percentage of total child mortality. Neonatal sepsis with early onset continues to be a significant cause of death and illness. The majority of sepsis-related deaths occur in developing nations, where the prevalence and causes of newborn sepsis are yet unknown. As a result, the goal of this study was to determine the prevalence of early-onset sepsis and identify determinant factors.

METHODS

A cross-sectional study was conducted on 368 study participants in referral hospitals of East and West Gojjam Zones from March 1^st to April 30^th, 2019. Study participants were selected at random using lottery method. Face-to-face interviews with index mothers for maternal variables and neonatal record review for neonatal variables were used to collect data using a structured pretested questionnaire. Data were entered into Epidata 3.1 and then exported to STATA/SE software version 14. Finally, the logistic regression model was used for analysis. Statistical significance was declared at P < 0.05 after multivariable logistic regression.

RESULTS

A total of 368 newborns and their index mothers took part in this study. The mean age of the newborns was 4.69 days (± 1.93SD). Early-onset neonatal sepsis was seen in 34% of the babies. Nulliparity (AOR: 3.3, 95% CI: 1.1-9.5), duration of labor > 18 h after rupture of membranes (AOR: 11.3, 95% CI: 3.0-41.8), gestational age of 32-37 weeks (AOR: 3.2, 95% CI: 1.2-8.5), and neonates who require resuscitation at birth (AOR: 4, 95% CI: 1.4 -11.8) were all found to be significantly associated with early-onset neonatal sepsis.

CONCLUSION AND RECOMMENDATION

Early-onset neonatal sepsis was found to be high in this study. Early-onset neonatal sepsis was found to be associated with maternal, obstetric, and neonatal variables. Comprehensive prevention strategies that target the identified risk factors should be implemented right away.

Catheter-related bloodstream infections in infants hospitalized in neonatal intensive care units: a single center study

Sepsis in neonates carries a high morbidity and mortality rate and is among the most feared complications in the neonatal intensive care unit (NICU). Catheter-related bloodstream infections (CRBSI) are a common etiology of late-onset sepsis. The aim of this study was to compare risk factors and characteristics between patients according to the type of catheter that was utilized and according to birth weight classification. The study included 51 newborns with confirmed CRBSI, which were hospitalized in our level 3 NICU between January 2017 and December 2018. The study population was stratified according to the type of venous catheter utilized (peripherally inserted central catheter, central venous catheter (CVC), and peripheral venous catheter). Infants with low birth weight and those who required prolonged parenteral nutrition were most likely to develop CRBSI in our study group. The type of venous catheter was not associated with blood culture results. Also, infants with a birth weight of < 1500 g and > 1500 g did not differ in sepsis etiology. Further research is required to assess venous catheters relative risk of causing sepsis and if the outcome can be traced back specifically to catheter type or patient characteristics.

Predictive Scores for Late-Onset Neonatal Sepsis as an Early Diagnostic and Antimicrobial Stewardship Tool: What Have We Done So Far?

Background: Late-onset neonatal sepsis (LOS) represents a significant cause of morbidity and mortality worldwide, and early diagnosis remains a challenge. Various ‘sepsis scores’ have been developed to improve early identification. The aim of the current review is to summarize the current knowledge on the utility of predictive scores in LOS as a tool for early sepsis recognition, as well as an antimicrobial stewardship tool. Methods: The following research question was developed: Can we diagnose LOS with accuracy in neonates using a predictive score? A systematic search was performed in the PubMed database from 1982 (first predictive score published) to December 2021. Results: Some (1352) articles were identified—out of which, 16 were included in the review. Eight were original scores, five were validations of already existing scores and two were mixed. Predictive models were developed by combining a variety of clinical, laboratory and other variables. The majority were found to assist in early diagnosis, but almost all had a limited diagnostic accuracy. Conclusions: There is an increasing need worldwide for a simple and accurate score to promptly predict LOS. Combinations of the selected parameters may be helpful, but until now, a single score has not been proven to be comprehensive.

Early-onset neonatal sepsis and antibiotic use in Indonesia: a descriptive, cross-sectional study

BACKGROUND

Early diagnosis and prompt antibiotic treatment are crucial to reducing morbidity and mortality of early-onset sepsis (EOS) in neonates. However, this strategy remains challenging due to non-specific clinical findings and limited facilities. Inappropriate antibiotics use is associated with ineffective therapy and adverse outcomes. This study aims to determine the characteristics of EOS and use of antibiotics in the neonatal-intensive care units (NICUs) in Indonesia, informing efforts to drive improvements in the prevention, diagnosis, and treatment of EOS.

METHODS

A descriptive study was conducted based on pre-intervention data of the South East Asia-Using Research for Change in Hospital-acquired Infection in Neonates project. Our study population consisted of neonates admitted within 72 h of life to the three participating NICUs. Neonates who presented with three or more clinical signs or laboratory results consistent with sepsis and who received antibiotics for 5 consecutive days were considered to have EOS. Culture-proven EOS was defined as positive blood or cerebrospinal fluid culture. Type and duration of antibiotics used were also documented.

RESULTS

Of 2,509 neonates, 242 cases were suspected of having EOS (9.6%) with culture-proven sepsis in 83 cases (5.0% of neonatal admissions in hospitals with culture facilities). The causative organisms were mostly gram-negative bacteria (85/94; 90.4%). Ampicillin / amoxicillin and amikacin were the most frequently prescribed antibiotics in hospitals with culture facilities, while a third-generation cephalosporin was mostly administered in hospital without culture facilities. The median durations of antibiotic therapy were 19 and 9 days in culture-proven and culture-negative EOS groups, respectively.

CONCLUSIONS

The overall incidence of EOS and culture-proven EOS was high in Indonesia, with diverse and prolonged use of antibiotics. Prospective antibiotic surveillance and stewardship interventions are required.

Risk Factors for Mortality From Late-Onset Sepsis Among Preterm Very-Low-Birthweight Infants: A Single-Center Cohort Study From Singapore

OBJECTIVE

To determine the risk factors for mortality associated with late onset sepsis (LOS) among preterm very-low-birthweight (VLBW) infants.

STUDY DESIGN

We performed a retrospective cohort study of infants born <32 weeks gestation and <1,500 gm admitted to a Singaporean tertiary-level neonatal intensive care unit. We determined the clinical, microbial, and laboratory risk factors associated with mortality due to culture-positive LOS in this cohort.

RESULTS

A total of 1,740 infants were admitted, of which 169 (9.7%) developed LOS and 27 (16%) died. Compared to survivors, those who died had lower birth gestational age (median 24 vs. 25 weeks, p = 0.02) and earlier LOS occurrence (median 10 vs. 17 days, p = 0.007). There was no difference in the incidence of meningitis (11.1 vs. 16.9%, p = 0.3), NEC (18.5 vs. 14.8%, p = 0.6), or intestinal surgery (18.5 vs. 23.3%, p = 0.6) among infants who died compared to survivors. Gram-negative bacteria accounted for 21/27 (77.8%) LOS-associated deaths and almost all (13/14, 93%) fulminant episodes. The presence of multiorgan failure, as evidenced by the need for mechanical ventilation (100 vs. 79.0%, p = 0.008), elevated lactate (12.4 vs. 2.1 mmol/L, p < 0.001), and inotropic support (92.6 vs. 37.5%, p < 0.001), was significantly associated with mortality. Infants who died had significantly lower white blood cell (WBC) counts (median 4.2 × 109/L vs. 9.9 × 109/L, p = 0.001), lower platelet count (median 40 × 109/L vs. 62 × 109/L, p = 0.01), and higher immature to total neutrophil (I: T) ratio (0.2 vs. 0.1, p = 0.002). Inotrope requirement [AOR 22.4 (95%CI 2.9, 103.7)], WBC <4 × 109/L [AOR 4.7 (1.7, 13.2)], and I: T ratio >0.3 [AOR 3.6 (1.3, 9.7)] were independently associated with LOS mortality.

CONCLUSIONS

In a setting with predominantly Gram-negative bacterial infections, the need for inotropic support, leukopenia, and elevated I: T ratio were significantly associated with LOS mortality among preterm VLBW infants.

Performance Comparison of Infection Prediction Scores in a South African Neonatal Unit: A Retrospective Case-Control Study

BACKGROUND AND OBJECTIVES

Infection prediction scores are useful ancillary tests in determining the likelihood of neonatal hospital-acquired infection (HAI), particularly in very low birth weight (VLBW; <1,500 g) infants who are most vulnerable to HAI and have high antibiotic utilization rates. None of the existing infection prediction scores were developed for or evaluated in South African VLBW neonates.

METHODS

We identified existing infection prediction scores through literature searches and assessed each score for suitability and feasibility of use in resource-limited settings. Performance of suitable scores were compared using a retrospective dataset of VLBW infants (2016-2017) from a tertiary hospital neonatal unit in Cape Town, South Africa. Sensitivity, specificity, predictive values, and likelihood ratios were calculated for each score.

RESULTS

Eleven infection prediction scores were identified, but only five were suitable for use in resource-limited settings (NOSEP1, Singh, Rosenberg, and Bekhof scores). The five selected scores were evaluated using data from 841 episodes of HAI in 659 VLBW infants. The sensitivity for the scores ranged between 3% (NOSEP1 ≥14; proven and presumed infection), to a maximum of 74% (Singh score ≥1; proven infection). The specificity of these scores ranged from 31% (Singh score ≥1; proven and presumed infection) to 100% (NOSEP1 ≥11 and ≥14, NOSEP-NEW-1 ≥11; proven and presumed infection).

CONCLUSION

Existing infection prediction scores did not achieve comparable predictive performance in South African VLBW infants and should therefore only be used as an adjunct to clinical judgment in antimicrobial decision making. Future studies should develop infection prediction scores that have high diagnostic accuracy and are feasible to implement in resource-limited neonatal units.

Development and validation of a sepsis diagnostic scoring model for neonates with suspected sepsis

BACKGROUND

We aimed to develop and validate a diagnostic model for sepsis among neonates evaluated for suspected sepsis, by incorporating thromboelastometry parameters, maternal/neonatal risk factors, clinical signs/symptoms and laboratory results.

METHODS

This retrospective cohort study included 291 neonates with presumed sepsis, hospitalized in a NICU, from 07/2014 to 07/2021. Laboratory tests were obtained on disease onset and prior to initiating antibiotic therapy. Τhromboelastometry extrinsically activated (EXTEM) assay was performed simultaneously and Tοllner and nSOFA scores were calculated. Sepsis diagnosis was the outcome variable. A 10-fold cross-validation least absolute shrinkage and selection operator logit regression procedure was applied to derive the final multivariable score. Clinical utility was evaluated by decision curve analysis.

RESULTS

Gestational age, CRP, considerable skin discoloration, liver enlargement, neutrophil left shift, and EXTEM A10, were identified as the strongest predictors and included in the Neonatal Sepsis Diagnostic (NeoSeD) model. NeoSeD score demonstrated excellent discrimination capacity for sepsis and septic shock with an AUC: 0.918 (95% CI, 0.884-0.952) and 0.974 (95% CI, 0.958-0.989) respectively, which was significantly higher compared to Töllner and nSOFA scores.

CONCLUSIONS

The NeoSeD score is simple, accurate, practical, and may contribute to a timely diagnosis of sepsis in neonates with suspected sepsis. External validation in multinational cohorts is necessary before clinical application.

The diagnostic and prognostic role of MiRNA 15b and MiRNA 378a in neonatal sepsis

BACKGROUND AND OBJECTIVES

Sepsis is one of the major factors for both term and preterm babies with morbidity and mortality. On the basis of recent clinical trials, altered circulating micro-RNAs (miRNAs) may serve as possible biomarkers in sepsis for diagnosis and prognosis. The aim of this research is to assess the diagnostic and prognostic biomarkers of miRNA 15b and miRNA 378a for neonatal sepsis.

SUBJECTS & METHODS

This study was carried out 25 neonates with sepsis admitted to neonatal ICU of Menoufia University Hospital and 25 healthy controls from February 2019 to May 2020. The relative quantification (RQ) of miRNA-15b and miRNA-378a expression was assessed using real time PCR technique. Results: Our results demonstrated that patients with sepsis had significantly higher level of MiRNA-15b than the healthy volunteers. On the other hand, patients with sepsis had significantly lower level of MiRNA-378a than the healthy volunteers. The ROC curve showed that the serum MiRNA-15b was a significant discriminator of sepsis with a combined sensitivity and specificity of 76% and 88% with cutoff point of 3.24. In addition, serum MiRNA-378a was a significant discriminator of sepsis with a combined sensitivity and specificity of 60% and 88% with cutoff point of 0.361. The miRNA-15b expression significantly correlated positive with respiratory rate (r =0.415,p =0.039), WBCs (r = 0.408, p =0.043), and CRP (r =0.407, p=0.043). Likewise, miRNA-378a expression significantly correlated negative with respiratory rate (r =-0.415p =0.024), WBCs (r =- 0.442, p =0.027), and CRP (r =- 0.459, p=0.021).

CONCLUSION

Both MiRNA 15b and 378a are promising biomarker for neonatal sepsis.

Microbial Patterns and Antibiotic Susceptibility in Blood Culture Isolates of Septicemia Suspected Children in the Pediatrics Ward of a Tertiary Care Hospital

Objectives  This study aims to determine microbial patterns and antibiotic susceptibility to alert clinicians to the emerging pathogens that may pose a threat to the community, especially children.

Materials and Methods  This retrospective cross-sectional study was conducted in the Department of Pathology, Holy Family Hospital, Rawalpindi, Pakistan, from July 2019 to December 2019. Two samples were taken from two different sites of each patient at the time of fever (> 100° F) to avoid chances of contamination. A blood sample of 5 to 10 mL was drawn from each site. In a 50 mL brain heart infusion broth (BHIB), 5 to 10% blood was injected in two different bottles and incubated for 48 hours at 37°C. After 48 hours, subculture was done by streaking the drops of blood samples taken from BHIB on blood and MacConkey agar. Cultures were incubated in aerobic conditions at 37°C for 24 to 48 hours. For testing antibiotic susceptibility, criteria defined by the Clinical and Laboratory Standards Institute (CLSI) were followed. Microbes were identified under a microscope by observing their morphological characteristics after gram staining and applying biochemical tests. Antibiotic sensitivity test was performed using standard aseptic methods.

Statistical Analysis  Bacterial isolates and their susceptibility patterns were represented using frequencies and percentage charts.

Results  Out of 423 blood cultures, growth was recorded in 92 (21.75%) of the cultures with female to male ratio 2.1:1. The gram-positive bacteria accounted for 43.48% ( n = 40), whereas gram-negative bacteria covered the majority 54.36% ( n = 50). Among isolates, Staphylococcus aureus (42.39%) was the most common, followed by Acinetobacter spp. (17.39%) and Pseudomonas aeruginosa (14.13%). Acinetobacter spp. showed 0% susceptibility to amikacin and cefotaxime. All the isolates were 100% resistant to amoxicillin-clavulanic acid. S. aureus showed lower sensitivity for ceftazidime (0%), clindamycin (66.67%), ciprofloxacin (0%), clarithromycin (11.76%), and ceftriaxone (0%). Tigecycline showed 100% sensitivity for all isolates tested.

Conclusion  Gram-negative bacteria form the majority of isolates in our setup, with Acinetobacter as the most common species among them. The resistance against cephalosporins, penicillin, and fluoroquinolones shown by Acinetobacter , Pseudomonas , Salmonella, and Klebsiella is of grave concern. Among gram-positive bacteria, S. aureus has established resistance against multiple drugs. Limited and objective use of antibiotic therapy is a much-needed strategy under new guidelines.

Diagnostic Accuracy of Clinical Tool 'STOPS' and Serum Procalcitonin for Optimizing Antibiotic Therapy in Neonates Born at ≥ 28 Weeks of Gestation with Neonatal Sepsis

Background

Antibiotic therapy is initiated in neonates on suspicion of sepsis. Optimizing therapy is a felt need of clinicians as prolonged injudicious use increases mortality and morbidity risk.

Objective

To evaluate the diagnostic accuracy of clinical tool 'STOPS' and serum procalcitonin (PCT) for identifying neonates with early-onset neonatal sepsis (EONS) or late-onset neonatal sepsis (LONS) and early discontinuation in those with no sepsis.

Methods

The study had a prospective analytical design conducted at a tertiary care hospital. Consecutively admitted neonates with suspected EONS or LONS were enrolled. The 'STOPS' tool comprising sensorium, temperature, oxygenation, perfusion, skin color, and blood sugar was applied at 6 and 12 hours of enrollment. Serum PCT was sent at 12 hr. The sensitivity, specificity, positive and negative predictive value (PPV and NPV), positive and negative likelihood ratio (PLR and NLR) were estimated.

Results

The study enrolled 380 neonates, of which 330 were given antibiotics for EONS and 50 for LONS. Temperature disturbance in the EONS group at 12 hr showed a PPV of 100% and a PLR of 9.1 (7.7 – 18). Perfusion assessment at 12 hr had a PPV of 77% and PLR of 8.25 (2.3 – 29). Skin color assessment at 12 hr had a PPV of 100% and PLR of 13.5 (9.7 – 27). The diagnostic accuracy of PCT in the EONS group was unremarkable. In the LONS group, skin color at 12 hr had a PPV of 100% and PLR of 11.2 (8.6 – 19.5). The diagnostic accuracy of PCT in the LONS group showed a PPV of 82% and PLR of 7 (1.7 – 29).

Conclusion

Identifying abnormal STOPS parameters was superior to PCT alone in EONS and as good as PCT in LONS. The 'STOPS' tool allows early identification of neonates with no sepsis, thereby optimizing antibiotic use.