Table 1 Effect of test sensitivity and specificity on model inputs

True Positive

Patient Septic

LOS: Full-length hospital stay

Mewes 2019 & Balk 2017 [23, 25]

Treatment length: Full-length

Voermans 2019 & Balk 2017 [24, 25]

Readmission: Average

Broyles 2017 [30]

True Negative

Patient Not Septic

LOS: Reduced by 4 days^a

Skoglund et al. 2019 [29]

Treatment Length: 1 day SOC; 0 days rapid PSP

Assumption (antibiotics initiated until lactate results are received under SOC; No antibiotics under PSP)

Readmission: None

Assumption (patient not septic)

False Positive

Patient Not Septic

LOS: Reduced by 3 days^a

Assumption (delays true negative by one day)

Treatment Length: 2 days

Assumption (treated for one day in addition to the day of treatment prior to initial lactate results before detecting false positive with second test)

Readmission: None

Assumption (patient not septic)

False Negative

Patient septic with increased risk of severe sepsis

LOS: Increased by 1 day^a

Assumption (delays true positive by one day)

Treatment Length: Full-length

Assumption (full treatment required)

Readmission: Average + 2.1%

Paoli 2018 [7]