Ali 2006.

Study characteristics
Methods	Single‐centre trial at Sher‐i‐Kashmir Institute of Medical Sciences (SKIMS), Srinagar, India Period of study: October 2004 to November 2005 Randomisation details not reported. Concealment of allocation: unclear Blinding of intervention: no Blinding of outcome assessment: unclear Completeness of follow‐up: yes
Participants	50 premature infants with culture‐proven sepsis (< 37 weeks' gestational age); 25 in PTX group and 25 in control group  Inclusion criteria: culture‐proven sepsis, gestation < 37 weeks, clinical signs of sepsis including cardiovascular and respiratory dysfunction, and written consent Exclusion criteria: intraventricular haemorrhage, congenital infection, and culture negativity Gestational age in both the treatment and placebo groups ranged from 32 to 37 weeks. Birthweight ranged from 950 g to 2580 g in the treatment group, and 1000 g to 2650 g in the control group. 42/50 of enroled infants had gram‐negative sepsis.
Interventions	Pentoxifylline intravenously (5 mg/kg/h for 6 hours for 3 consecutive days) in conjunction with antibiotics or control group who received antibiotics First‐line antibiotics were amoxicillin + clavulanic acid and amikacin
Outcomes	Mortality, development of NEC, length of hospital stay, duration of ventilation, and adverse effects (hypertension, irritability, deterioration of vital signs) were reported.
Notes	Although mean length of hospital stay and duration of ventilation were reported, standard deviations were not, and so not included in meta‐analyses. Funding sources and declarations of interest were not stated.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Not clear
Allocation concealment (selection bias)	Unclear risk	No details of randomisation reported.
Blinding (performance bias and detection bias) All outcomes	High risk	None reported.
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Unclear.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Outcomes for all participants reported.
Selective reporting (reporting bias)	Low risk	None noted
Other bias	Low risk	None noted