Asghar 2008.

Methods	Randomised, non‐blinded, multi‐site efficacy study
Participants	Children between 2 to 59 months of age with very severe pneumonia
Interventions	Ampicillin plus gentamicin (ampicillin 200 mg/kg/d in 4 doses every 6 hours, and gentamicin 7.5 mg/kg/d as in a single daily dose) or chloramphenicol (75 mg/kg/d given in 3 doses) every 8 hours for minimum of 5 days. After that first group received oral amoxycillin (45 mg/kg/day in 3 divided doses) and the other group received oral chloramphenicol 75 mg/kg/day to complete 10 days
Outcomes	Primary outcome Treatment failure by 5 days after admission, defined as new development or persistence of at least 2 of the following: inability to drink; tachypnoea (≥ 50 breaths/minute in children aged 2 to 11 months and ≥ 40 breaths/minute in children aged 12 to 59 months) and abnormally sleepy or difficult to wake Development or diagnosis of any of the following: bacterial meningitis, empyema, septic shock, renal failure or newly diagnosed co‐morbid conditions. Serious adverse drug reaction Modification of antibiotic treatment Voluntary withdrawal or absconding Death Secondary outcomes Treatment failure as defined above at 48 to 60 hours Treatment failure as defined above plus relapse (hypoxaemic pneumonia at 10 to 12 days and 21 to 30 days, with oxygen saturations ≤ 90%, or ≤ 88% in the 2 high altitude sites in Mexico and Yemen) Death by 30 days after enrolment Bacterial pathogens isolated from blood or other sterile sites Antimicrobial susceptibility of the isolated pathogens
Notes	Exclusion criteria: wheezing, with a history of 3 or more attacks, or known asthma, known heart disease, duration of present illness more than 10 days, history of serious adverse reaction to any of the study drugs, previous enrolment in the study Admission to hospital for more than 24 hours within past 7 days Documented evidence of injectable antibiotic treatment for more than 24 hours before enrolment, stridor, known renal failure or not passed urine during past 6 hours, evidence of cerebral malaria Evidence of bacterial meningitis, clinical jaundice, residence of patient in an area where follow‐up was not possible, empyema or presence of pneumatoceles on chest radiograph
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Sequence generated by World Health Organization by using variable size blocks
Allocation concealment (selection bias)	Low risk	Separate randomisation lists were prepared for each site examination, and individual patient assignments were placed in opaque, sealed envelopes. Before opening each envelope the doctor in charge signed and dated the opening flap of the envelope
Blinding of participants and personnel (performance bias) All outcomes	High risk	Open‐label randomised controlled trial
Blinding of outcome assessment (detection bias) All outcomes	High risk	Open‐label randomised controlled trial
Selective reporting (reporting bias)	Low risk	No selective reporting
Incomplete outcome data (attrition bias) All outcomes	Low risk	Incomplete data adequately addressed
Other bias	Low risk	Funded by World Health Organization and Center of International Health and Development, Boston University and Johns Hopkins Bloomberg School of Public Health, Baltimore