Table 3.

Predictive scores for the LOS with combined variables.

Reference	Country, Year	Method	Design	Population	Scoring System (Points)	Main Findings
[24]	Germany, 1982	Retrospective and prospective	Original	403 neonates: Retrospective: 83 with sepsis Prospective: 39 with sepsis, 42 with amniotic infection, 28 with post-asphyxia syndrome, 28 premature with cerebral hemorrhage, 183 controls	skin coloration (0–4) microcirculation (0–3) metabolic acidosis (0–2) muscular hypotonia (0–2) bradycardias (0–1) apneic spells (0–1) respiratory distress (0–2) liver enlargement (0–1) gastrointestinal symptoms (0–1) WBC count (0–3) Shift to the left (0–3) thrombocytopenia (0–2)	Analysis was divided into 3 phases: onset, at the beginning and at the peak of the illness. Each phase gave different results: as the illness evolved, the scores got higher. Changes in skin coloration: the most frequent sign of NS. Septic neonates performed high scores (47% at the beginning of the illness and 92% in seriously ill infants), in contrast with non-septic neonates.
[33]	USA, 2007	Prospective	Original	337 neonates: 76 episodes of proven sepsis (blood culture (+) in 63 neonates, 80 episodes of clinical sepsis (blood culture (–) in 63 neonates Age ≥ 7 days old and ≥ 7 days of HRC monitoring Out of 337 neonates: 172 were < 1500 g (VLBW)	Feeding intolerance (2) Severe apnea (2), 50% increase in the number of apneic episodes over a 24 h period in an infant stable for 3 days (2) I:T ratio > 0.2 (2) Increase in ventilatory support and FiO2 by 25% from baseline (1) Lethargy or hypotonia (1) Temperature instability (>38 °C or <36.2 °C), 2 episodes within 8 h (1) Hyperglycemia (>180 mg/dL) (1) Abnormal WBC count (>25.000 or <5.000)	Hyperglycemia and abnormal WBC count: highly associated with NS only the time of the blood culture. Hypotonia and lethargy: great association with NS only the time preceding the blood culture. Infants with sepsis had higher scores than controls. Hypotension in only 3% of infants with NS (not included in the score). HRC index and clinical score were predictive for NS in the next 24 h. Clinical tests less useful before the NS, because signs and symptoms are present less often. Infants with clinical or proven sepsis: higher scores than controls. Feeding intolerance: the most predictive clinical sign of NS. Feeding intolerance, hypotonia, lethargy and abnormal I:T ratio: the most predictive findings. I:T ratio the most robust independent predictor. Increase in the score in the 24 h before the clinical diagnosis HRC index adjunctive to clinical information proved useful.
[34]	Thailand, 2020	Retrospective	Original	208 neonates: 52 sepsis (only proven bacterial LOS), 156 controls Aged ≥ 7 days	poor feeding (2) abnormal heart rate (outside the range 100–180 x/min) (3) abnormal temperature (outside the range 36–37.9 °C) (4) abnormal O2 saturation (<92%) (1) abnormal leukocytes (outside the range of 5.000–20.000/cm) (2) abnormal pH (outside the range of range 7.27–7.45) (2)	Duration of hospitalization, intracranial hemorrhage, high-risk pregnancies and resuscitation: the most powerful risk factors. Abnormal temperature and abnormal HR: the most common sings in NS. Abnormal HRC occurred early in the course of the illness. Abnormalities were found 12–24 h before the clinical diagnosis of NS. No infant with hypothermia had LOS. Antibiotic therapy to be guided according to the score.
[35]	Belgium, 2000	Prospective and retrospective	Original (NOSEP score)	119 neonates: 154 episodes: Derivation cohort: 104 episodes of presumed NS in 80 neonates (43 proven sepsis)Validation cohort: 50 episodes of proven NS in 39 neonates >48 h in NICU	NOSEP-1 score: Fever > 38.2 °C (5) CRP ≥ 14 mg/L (5) Thrombocytopenia < 150 × 109/L (5) Neutrophil fraction > 50% (3) Total parenteral nutrition (TPN) ≥ 14 days (6) NOSEP-2 score: NOSEP-1 score + culture results	Score for nosocomial NS. BW, GA, presence of CVC, prolonged hospital stay and exposure to TPN (especially lipid emulsions) > 14 days: strongly associated with NS. TPN as the only independently associated factor. Fever and neutrophil fraction as powerful signs for prediction. Adding catheter cultures improves the diagnostic power of the score. NOSEP score as accurate as a continuous computerized scoring system. Only 2 variables do not rely on lab results. Waiting for the results for assessment.
[36]	Belgium, 2002	Prospective	Validation of NOSEP score (by Mahieu et al.) and new score	128 neonates: 155 episodes:Internal validation: 62 episodes of presumed NS in 49 neonates (20 proven NS) External validation: 93 episodes of presumed NS in 79 neonates (51 proven NS) >48 h in NICU	NOSEP-1 score: Fever > 38.2 °C (5) CRP≥ 14 mg/L (5) Thrombocytopenia < 150 × 109/L (5) Neutrophil fraction > 50% (3) Total parenteral nutrition (TPN) ≥ 14 days (6) NOSEP-2 score: NOSEP-1 score + culture results NOSEP-NEW-I: Fever > 38.1 °C (5) CRP ≥ 30 mg/L (5) Thrombocytopenia < 190 × 109/L (5) Neutrophil fraction > 63% (3) Total parenteral nutrition (TPN) ≥ 15 days (6) NOSEP-NEW-II: NOSEP-NEW-I + recent surgery, maternal hypertension and ventilation at time of sepsis work up.	Score for nosocomial NS. External validation was set in multiple NICUs. Score was higher in septic neonates in both internal and external validations. Internal validation was better than the external. Score suggested as a tool for detection of NS and for reduction of unnecessary use of antibiotics in NICUs.
[37]	Thailand, 2005	Retrospective	Original	173 neonates: Derivation phase: 100 neonates (17 NS), 40% premature and 18% LBW Validation phase: 73 neonates (25 NS), 69% premature and 49% LBW Hospitalized for >72 h after birth	Hypotension (4) Abnormal body temperature (>38 °C or <36.5 °C or temperature instability) (3) Respiratory insufficiency (apnea/bradycardia, tachypnea, cyanosis, increased oxygen requirement or ventilator settings) (2) Neutrophil Band form fraction ≥ 1% (2) Thrombocytopenia (<150 × 103/μL) (2) Umbilical venous catheterization: 1–7 days (2), >7 days (4)	First bedside score for neonates hospitalized > 72 h. Hypotension and abnormal body temperature had the strongest association with NS. Risk variables: LBW, prematurity and TPN: no significant association with LOS, while UVC usage independently associated. Combination of clinical, laboratory and management variables: suspicion of LOS without waiting for the lab results. Score based mostly on clinical sings. Risk groups: stratification of risk for LOS (low, intermediate, high risk) and benefit for decision-making.
[38]	Canada, 2007	Prospective	Validation (of Okascharoen et al.)	105 neonates: 35 NS Aged 2–90 days >48 h in NICU	Hypotension (4) Abnormal body temperature (>38 °C or <36.5 °C or temperature instability) (3) Respiratory insufficiency (apnea/bradycardia, tachypnea, cyanosis, increased oxygen requirement or ventilator settings) (2) Neutrophil band form fraction ≥ 1% (2) Thrombocytopenia (<150 × 103/μL) (2) Umbilical venous catheterization: 1–7 days (2), >7 days (4)	No significant difference in GA, BW, utilization of CVC and duration of TPN between septic and non septic children. Only utilization of UVC proved to make a difference. External validation performed similar accuracy with the internal validation. From low to intermediate risk: Se, Sp ↓ Clinicians predict LOS as strongly as the scoring system, but tend to overestimate the possibility of LOS: score performed better in prediction comparing to clinicians viewpoint. When the neonatal population consists only of proven LOS records, NS was underestimated, while when suspected LOS episodes are present, LOS tended to be overestimated.