Table 2.
Additional studies that did not meet full inclusion criteria but examined components of early sepsis management included in the review.
| Study | Design | Bias | Country | Participants | Population | Study Summary |
|---|---|---|---|---|---|---|
| Reddy, 2010 [25] | Systematic review | African continent | 58,296 (adults and children) |
Patients with ≥1 blood culture | Reports bacterial pathogens isolated from blood cultures: S. enterica, S. pneumoniae, S. aureus and E. Coli. Mycobacterium tuberculosis frequently isolated (166/539) where appropriate techniques used. | |
| Moore, 2017 [26] | Systematic review | Sub-Saharan Africa | 5573 | Adults admitted to hospital | Pooled data from 13 cohort studies to derive a ‘universal vital assessment’ score to predict in-hospital mortality based on physiological parameters. Clinical variables include: temperature, heart rate, respiratory rate, systolic blood pressure (BP), SpO2, Glasgow coma scale (GCS) and HIV-infection status. Score AUCROC for mortality 0.77 (0.75–0.79). | |
| Gupta-Wright, 2018 [27] | RCT | 5/5 | Malawi and South Africa | 4788 | Adult inpatients with HIV-infection | Urinary lipoarabinomannan guided therapy did not reduce overall mortality (adjusted risk reduction [aRD]—2.8%, CI—5.8 to 0.3; p = 0.074) but did reduce mortality in pre-defined subgroups with CD4 count <100, severe anaemia and clinically suspected tuberculosis (TB). |
| Peter, 2016 [28] | RCT | 4/5 | Sub-Saharan Africa | 2659 | Adult admissions with HIV-infection and TB symptoms | Urinary lipoarabinomannan guided anti-tuberculosis treatment reduced initiation time (median 0 day [IQR 0–2] vs. 1 day [IQR 0–3), p < 0.0001) and eight-week mortality (21% vs. 25%, ARR 4% C.I. 1–7%). |
| Jacob, 2009 [29] | Prospective Observational | 6/9 | Uganda | 382 | A & E with severe sepsis [23] | Management and outcomes of patients with severe sepsis: including poor fluid resuscitation (median 500 mL within 6 h) and antibiotic administration (61% patients received within 6 h) following sepsis diagnosis. Thirty-day mortality 43.0%. |
| Belle, 2010 [30] | Cross-sectional | 6/9 | Multiple (African continent) | 231 hospitals and health centres | Healthcare facilities | Snapshot survey assessed oxygen supply and infrastructure in 12 African countries. Only 43·8% of facilities had uninterrupted access to an oxygen source and 24·6% had a fully functioning oxygen concentrator. Electricity fully available at 35·1% of facilities |
Risk of bias assessments: Newcastle–Ottowa scale and Jadad scale for cohort studies and randomised controlled trials, respectively. AUCROC: Area Under Receiver Operator Curve; ARR: Absolute risk reduction.