Table 1.
Study (year); setting, country | Cohort | Design | Quality assessment | CPM or PF study* | Sample size | Population | Outcome | Outcome prevalence (n/N) | ||
Risk of bias | Applicability concern | Inclusion criteria | Exclusion criteria | |||||||
Outcomes including death, organ dysfunction, organ support and PICU admission | ||||||||||
Scott (2020)40; Secondary and tertiary care hospitals, USA | Hospital OPD/ED | Retrospective cohort | High | Low | CPM | 2464 | Age 60 days to 18 years; Clinician-suspected sepsis | Hypotensive septic shock on arrival;† transfer to another centre; leaving ED before formal evaluation; incorrect registration | Hypotensive septic shock‡≤24 hour | 11.4% (282/2464) |
Walia (2016)39; Tertiary care hospital, India | Hospitalised§ | Prospective cohort | High | High | CPM | 100 | Age 3–36 months; Axillary temperature >36.9°C (early morning) or >37.4°C | Non-infectious cause of fever; immunisation ≤2 days; immunodeficiency, autoimmune disorder | In-hospital mortality; Mechanical ventilation | 11.0% (11/100); 17.0% (17/100) |
Aramburo (2018)26; Secondary and tertiary care hospitals, Kenya, Tanzania and Uganda | Hospitalised§ | Randomised controlled trial | Moderate | High | PF | 3008 | Age 60 days to 12 years; history of fever or axillary temperature ≥37.5°C or <36°C; severe febrile illness¶ |
Non-infectious cause of illness; SAM, gastroenteritis, burns, chronic kidney disease, pulmonary oedema, intoxication, surgical conditions, receipt of isotonic fluids during the same illness | In-hospital mortality (72 hours) | 10.3% (309/3008) |
George (2015)31; Secondary and tertiary care hospitals, Kenya, Tanzania and Uganda | Hospitalised§ | Randomised controlled trial | High | High | CPM | 3121 | Age 60 days to 12 years; history of fever or axillary temperature ≥37.5°C or <36°C; severe febrile illness¶ |
Non-infectious cause of illness; SAM, gastroenteritis, burns, chronic kidney disease, pulmonary oedema, intoxication, surgical conditions, receipt of isotonic fluids during the same illness | In-hospital mortality (48 hours) | 9.8% (306/3121) |
Scott (2012)37; Tertiary care hospital, USA | Hospital OPD/ED | Prospective cohort | High | High | PF | 239 | Age <19 years; temperature >38.5°C or <36°C and heart rate >2 SD above normal for age; underwent phlebotomy as part of usual care | Transfer from another health facility; known inborn errors of metabolism; receipt of >15 min of intravenous therapy | 24 hours organ dysfunction | 5.4% (13/239) |
Scott (2014)36; Tertiary care hospital, USA | Hospital OPD/ED | Prospective cohort | High | High | PF | 239 | Age <19 years; temperature >38.5°C or <36°C and heart rate >2 SD above normal for age; undergoing phlebotomy as part of routine care | Transfer from another health facility; known inborn errors of metabolism; receipt of >15 min of intravenous therapy | 24 hours organ dysfunction | 5.4% (13/239) |
Nadjm (2013)34; Secondary care hospital, Tanzania | Hospitalised§ | Prospective cohort | Moderate | High | PF | 3319 | Age 2 months to 5 years; history of fever in last 48 hours or axillary temperature ≥37.5°C | Chronic illness (excluding HIV and malnutrition); trauma; surgical conditions | In-hospital mortality | 5.1% (170/3319) |
Mtove (2011)32; Secondary care hospital, Tanzania | Hospitalised§ | Prospective cohort | Moderate | High | PF | 3248 | Age 2 months to 13 years; history of fever in last 48 hours or axillary temperature ≥37.5°C | Chronic illness (excluding HIV and malnutrition); trauma; surgical conditions | In-hospital mortality | 5.0% (164/3248) |
Lowlaavar (2016)42; Secondary and tertiary care hospitals, Uganda | Hospitalised§ | Prospective cohort | High | High | CPM | 1307 | Age 6–60 months; admitted during study working hours or within 8 hours of study shift with a proven or suspected infection | Previous enrolment; residence outside study catchment area | In-hospital mortality | 5.0% (65/1307) |
Conroy (2015)27; Tertiary care hospital, Uganda | Hospitalised§ | Prospective cohort | High | High | CPM | 2502 | Age 2 months to 5 years; history of fever in last 48 hours or axillary temperature >37.5°C | None reported | In-hospital mortality | 4.7% (99/2089) |
van Nassau (2018)38; Secondary care hospital, The Netherlands | Hospitalised§ | Retrospective cohort | High | High | CPM | 864 | Age <18 years; suspected bacterial infection** | Surgical conditions | PICU transfer and/or in-hospital mortality | 2.7% (24/864) |
Scott (2017)35; Tertiary care hospital, USA | Hospital OPD/ED | Retrospective cohort | Low | High | PF | 1299 | Age 60 days to 18 years; suspected sepsis††; measurement of venous lactate as part of routine care within 8 hours of ED arrival | Transfer from another health facility | 30-day mortality | 1.9% (25/1299) |
SEAIDCRN (2017)12; Tertiary care hospitals, Indonesia, Thailand and Vietnam | Hospitalised§ | Prospective cohort | High | High | PF | 763 | Age 30 days to 18 years; modified SIRS criteria‡‡ | Suspicion of hospital-acquired infection; admission to hospital within previous 30 days; transfer from another health facility after >72 hours admission; weight <3 kg; enrolment in another clinical study | 28-day mortality | 1.9% (14/731) |
Costa de Santana (2017)28; Tertiary care hospital, Brazil | Hospital OPD/ED | Retrospective cohort | High | High | PF | 254 | Age <13 years; axillary temperature >38.5°C; measurement of respiratory rate and heart rate on three occasions in absence of fever; measurement of leucocyte count as part of routine care | Congenital malformations; bronchopulmonary dysplasia; medullary aplasia; cardiac, renal or hepatic insufficiency | In-hospital mortality | 1.6% (4/254) |
Kwizera (2019)41; Secondary care hospital, Rwanda | Hospitalised§ | Prospective cohort | High | High | CPM | 949 | Age 28 days to 18 years; confirmed acute infectious disease; symptom onset <14 days prior to hospital admission | Allergy to antimicrobials to treat sepsis (antibiotics, artesunate, artemether-lumefantrine); terminal disease | In-hospital mortality | 1.5% (14/949) |
Outcomes including length of stay and persistence of symptoms | ||||||||||
Freyne (2013)30; Secondary care hospital, Ireland | Hospitalised§ | Prospective cohort | High | High | PF | 46 | Age 6–36 months; axillary temperature >38.1°C | Chronic illness; immunisation ≤2 days, antipyretic use ≤2 hours | Length of stay >96 hours | 26.1% (12/46) |
van Nassau (2018)38; Secondary care hospital, The Netherlands | Hospitalised§ | Retrospective cohort | High | High | CPM | 864 | Age <18 years; suspected bacterial infection** | Surgical conditions | Length of stay ≥7 days | 22.2% (179/806) |
Elshout (2015)29; General Practice (out of hours), The Netherlands | Primary care | Prospective cohort | High | High | PF | 480 | Age 3 months to 6 years; history of fever | Communication in Dutch not possible; enrolment in last 2 weeks; direct referral to hospital required | Persistent fever at D3 | 13.1% (63/480) |
Mwandama (2016)33; Community Health Workers, Malawi | Primary care | Prospective cohort | High | High | PF | 285 | Age 2–59 months; history of fever in last 48 hours or temperature ≥37.5°C; negative malaria rapid diagnostic test | Receipt of antimalarial in last 2 weeks; presence of danger signs§§ | Persistent symptoms at D7 | 10.4% (19/182) |
Studies are grouped according to the type of outcome they used: ‘hard’ (death, organ dysfunction, organ support, PICU admission) or ‘soft’ (length of stay, persistence of symptoms).
*Studies evaluating both PFs and CPMs were categorised on the basis of their primary analysis to facilitate review using the appropriate quality assessment tool.
†Hypotensive systolic blood pressure on arrival with receipt of a fluid bolus or vasoactive agent within 30 min.
‡Hypotension plus receipt of ≥30 mL/kg isotonic crystalloids or vasoactive medication.
§Only children the treating physician decided to admit were eligible but recruitment occurred at the time of admission to the health facility.
¶Respiratory distress (increased work of breathing or deep breathing) and/or impaired consciousness (coma or prostration) AND evidence of poor peripheral perfusion (capillary refill time >2 s or lower limb temperature gradient or weak radial pulse or severe tachycardia).
**Initiation of antibiotics within 24 hours of arrival in the emergency department.
††Decreased mental status or perfusion in the setting of suspected infection.
‡‡Rectal temperature >38.5°C or <35°C (or equivalent) AND heart rate >2 SD above normal for age (unless hypothermic) AND respiratory rate >2 SD above normal for age AND altered mental status OR systolic blood pressure <2 SD below normal for age OR pulse pressure <20 mm Hg OR capillary refill time >2 s OR SpO2 <95% OR leucocyte count >12×103 cells/µL or <5×103 cells/µL.
§§Convulsions, repeated vomiting, lethargy, severe anaemia or loss of consciousness.
CPM, clinical prediction model; ED, emergency department; n, number of outcomes; n, number of cases; OPD, outpatient department; PF, prognostic factor; PICU, paediatric intensive care unit; SAM, severe acute malnutrition; SEAIDCRN, Southeast Asia Infectious Disease Clinical Research Network; SIRS, systemic inflammatory response syndrome.;