Table 1 Summary of main articles about EOS calculator included for review

Escobar et al., 2014 [25]

≥ 34 weeks’ GA

According to 2010 CDC guidelines, 11% of infants were treated with empirical antibiotics, although only 0.04% had blood culture-confirmed sepsis. Using a risk stratification scheme based on maternal and neonatal data, 4% of infants would have been treated with empirical antibiotics

Shakib et al., 2015 [26]

≥ 34 weeks’ GA well-appearing infants exposed to maternal CAM

Reduction of patients with testing/initial antibiotics by at least 80% if using the EOS calculator compared with 2010 CDC guidelines

Kuzniewicz et al., 2017 [27]

≥ 35 weeks’ GA

Reduction of blood culture use from 14.5% (2010 CDC guidelines) to 4.9% (EOS calculator). Reduction of empiric antibiotic administration in the first 24 h from 5.0% (2010 CDC guidelines) to 2.6% (EOS calculator) with subsequent decrease of antibiotic days per 100 births from 16.0 to 8.5 days

Money et al., 2017 [28]

≥ 37 weeks’ GA well-appearing infants exposed to maternal CAM

Reduction of empiric antibiotic treatment from 99.7% (2010 CDC guidelines) to 2.5% (EOS calculator). One patient with culture-positive EOS would not have received antibiotics based on the EOS calculator

Warren et al., 2017 [29]

≥ 34 weeks’ GA infants who received antibiotics at birth for suspected EOS

Reduction of empiric antibiotic treatment from 93% (2010 CDC guidelines) to 23% (EOS calculator). Both 2010 CDC guidelines and the EOS calculator recommended treatment for 7 patients with culture-negative EOS

Beavers et al., 2018 [30]

≥ 34 weeks’ GA exposed to maternal CAM

NICU admissions rates decreased from 91 to 37%, the number of blood cultures decreased from 92 to 50% and antibiotic administration rates decreased from 94 to 37% when 2010 CDC guidelines were replaced with EOS calculator recommendations

Carola et al., 2018 [31]

≥ 35 weeks’ GA infants exposed to maternal CAM

Only 0.43% of neonates born to mothers with CAM had culture-proven EOS. Empiric antibiotics would have been recommended in 23.5% of the patients according to EOS calculator (76.5% reduction in empirical antibiotic administration compared with 2010 CDC guidelines). Blood culture only was recommended for 8.9% of the neonates; treatment with antibiotics would have been recommended for 3 of the 5 neonates with positive blood culture. All 5 neonates with positive blood cultures had abnormal CBC and CRP values at 6–12 h

Dhudasia et al., 2018 [32]

≥ 36 weeks’ GA

Reduction in antibiotics administration from 6.3 to 3.7% when current CDC guidelines were compared to EOS calculator. There was also a reduction in use of laboratory tests for suspected EOS from 26.9 to 4.9%

Gievers et al., 2018 [33]

≥ 35 weeks’ GA infants exposed to maternal CAM

Compared to the 2010 CDC guidelines, EOS calculator yields a reduction of antibiotic exposure from 95 to 9%, laboratory evaluation from 96 to 22% and NICU observation from 73 to 10%

Klingaman et al., 2018 [34]

≥ 35 weeks’ GA

Compared to the 2010 CDC guidelines, EOS calculator yields a reduction in CBCs by 88%, blood cultures by 94%, and antibiotic administration by 78%

Strunk et al., 2018 [35]

≥ 35 weeks’ GA infants requiring evaluation and/or treatment for suspected EOS

Reduction of patients admitted to NICU from 24.2 to 21.2%, decrease of blood culture sampling from 15.2 to 11.1% and reduction of empiric antibiotic administration from 12.0 to 7.6% when using EOS calculator and not local guidelines based on AAP recommendations

Akangire et al., 2019 [36]

≥ 34 weeks’ GA

Compared to current CDC/AAP guidelines, the EOS calculator-based approach yields a reduction of empiric antibiotic administration from 11.0 to 5.0% and blood culture use from 14.8 to 7.6%

Arora et al., 2019 [37]

≥ 34 weeks’ GA infants admitted to NICU

Significant reduction in the rate of both antibiotic prescriptions (70.3% vs. 49.6%) and sepsis evaluations (90.9% vs. 68.8%) after implementation of the EOS calculator. 92% overlap in blood culture recommendations and 95% overlap between antibiotic recommendations when current CDC guidelines were compared to EOS calculator

Benaim et al., 2019 [11]

≥ 34 weeks’ GA

Over the period of study, antibiotic administration decreased by 38.0% with updated local EOS guidelines. Reduction of antibiotic administration would have been 31.0% (for an EOS incidence of 0.6/1000) and 1.0% (for an EOS incidence of 2/1000) with the EOS calculator

Bridges et al., 2019 [38]

≥ 37 weeks’ GA infants exposed to maternal CAM

Compared with 2010 CDC guidelines, 93.0% of patients were not admitted to the NICU and only 11.0% required laboratory evaluation; rates of exclusive breastfeeding increased from less than 10.0% to greater than 50.0% after implementation of the EOS calculator. The length of the NICU stay decreased from an average of 138 to 12 days with no negative consequences

Eason et al., 2019 [39]

≥ 37 weeks’ GA infants with risk factors for EOS or suspected EOS

The percentage of infants screened with a suspected infection receiving 5 days of antibiotics reduced from 31.0% with NICE guidelines to 5.0% with EOS calculator. Clinically well infants with risk factors alone receiving 36 h of antibiotics, reduced from 63.0% with NICE guidelines to 3.0% with EOS calculator

Fowler et al., 2019 [40]

≥ 34 weeks’ GA

6 patients with culture-positive EOS were identified in the study period and recommendations from the calculator were in alignment with current CDC/AAP guidelines

Goel et al., 2019 [41]

≥ 34 weeks’ GA

16% of infants were started on antibiotics as per NICE recommendations compared with 4.3% with EOS calculator. There were seven positive blood cultures (three infants were recommended antibiotics by both, three were not identified in the asymptomatic stage by either; one was a contaminant)

Gong et al., 2019 [42]

≥ 34 weeks’ GA infants exposed to maternal intrapartum fever

Compared to the CDC/AAP guidelines, the EOS calculator-based approach yields a net monetary benefit (3998 $ per infant), largely by preventing unnecessary antibiotic treatment (67.4% decrease in antibiotic use in the calculator arm)

Hershkovich-Shporen et al., 2019 [43]

≥ 35 weeks’ GA newborns with the following inclusion criteria: treated with antibiotic, born to mothers with risk factors for EOS, born to mothers with clinical CAM or that received IAP

15.0% of the patients received antibiotic treatment according to 2010 CDC recommendations; 8.0% of the patients would have received antibiotic treatment according to EOS calculator. Only 2/89 (2.25%) newborns treated for maternal clinical CAM according to 2010 CDC guidelines, had proven EOS. Three of the mothers whose newborn developed EOS, had no risk factors so there was no need for the EOS calculator

Joshi et al., 2019 [44]

≥ 34 weeks’ GA well-appearing newborns exposed to maternal CAM

Compared to the CDC/AAP guidelines, the usage of the EOS calculator yields a reduction of empirical antibiotics administration from 100% of patients to 8.9%

Leonardi et al., 2019 [45]

≥ 35 weeks’ GA newborns exposed to maternal CAM and/or intrapartum fever

228/312 (73.1%) infants did not require admission to the NICU based on their risk assessment using the EOS calculator; according to local guidelines, all infants would have been admitted to the NICU for evaluation and treatment of presumed sepsis, regardless of clinical appearance. Breastfeeding rates at discharge were 89.0% for infants remaining with their mothers in the newborn nursery, and 37.0% for infants admitted to the NICU

Stipelman et al., 2019 [46]

≥ 34 weeks’ GA infants exposed to maternal CAM

Reduction in antibiotics administration from 7.0% (according to CDC/AAP guidelines) to 1.0% after implementation of the EOS calculator. 2 missed cases of culture-positive EOS with EOS calculator

Benincasa et al., 2020 [47]

≥ 34 weeks’ GA neonates who received EOS antibiotics according to the hospital’s current practice

219/384 (57.0%) patients received antibiotics by EOS calculator and 64/384 (16.7%) by evaluation of clinical signs. All patients with positive blood culture were detected by both EOS calculator and clinical signs surveillance. Estimated costs were US$ 415.576 for EOS calculator and US$ 314.353 for evaluation of clinical signs

Morris et al., 2020 [48]

≥ 34 weeks’ GA infants with EOS confirmed on blood or cerebrospinal fluid culture

Within 4 h of birth, antibiotics were recommended for 39/70 (55.7%) infants with NICE guidelines, compared with 27/70 (38.6%) with the EOS calculator. The 12 infants advised early treatment only by NICE guidelines remained well, only one showing mild symptoms after 4 h. Another 4 babies received antibiotics by 4 h outside NICE and EOS calculator guidance. The remaining 27 infants (38.6%) received antibiotics when symptomatic after 4 h. Only one infant who was unwell from birth, died. Both NICE guidelines and EOS calculator were poor in identifying EOS within 4 h; NICE guidelines were superior to the EOS calculator in identifying asymptomatic cases

Perez et al., 2020 [49]

≥ 35 weeks’ GA

Compared to the current AAP guidelines, the usage of the EOS calculator yields 54.0% reduction in the number of infants undergoing sepsis workup evaluations and 51.0% decrease in the number of infants receiving antibiotics

van der Weijden et al., 2020 [50]

≥ 34 weeks’ GA neonates at risk for EOS

Dutch guidelines recommended antibiotic treatment for 363/890 (40.8%) neonates versus 101/890 (11.3%) with EOS calculator (p < 0.01). Antibiotic treatment was recommended by both methods for 90/890 (10.1%) neonates, including 2 patients with positive blood culture