Dunne 2005a

Methods	Trial design: Double blind, randomized, controlled trial with two arms. Follow‐up: Hospitalized, eight‐hourly blood smears. Medical history and exam at baseline. Discharged when two consecutive blood smears negative. Follow up and blood smears on day 7, 14, and 28. Adverse event monitoring: Not reported. G6PD test at baseline. Adverse event data for those analysed.
Participants	Number enrolled: 200 Number available for analysis: 199 Inclusion criteria:  18 to 65 years, Plasmodium vivax infection with parasite density < 100,000 parasites/µl in blood smear, Parasight F test negative for Plasmodium falciparum, a history of fever in the previous 48 hours, written informed consent. Exclusion criteria: Impaired consciousness, jaundice, respiratory distress; haematuria (self‐report); treated with an antimalarial or antibiotic with antimalarial activity in the previous 15 days; laboratory evidence or history of significant cardiovascular, liver or renal functional abnormality that in the opinion of the investigator would place them at increased risk;  a serum glucose level less than the lower limit of normal; a history of allergy to or hypersensitivity to study drugs; a blood transfusion in the previous 28 days; any situation that would prevent follow up visits. Age range: Not reported. Mean age (SD):  31.7 (11.6) years in arm 1, 30.0 (11.8) in arm 2. Proportion male: 159/200 (79.5%) Pregnant women: Excluded
Interventions	arm 1) Azithromycin 1 gram per day for three days plus chloroquine placebo arm 2) Chloroquine 600 mg per day for two days, 300 mg on day three plus azithromycin placebo Supervision of intake: Not reported, but participants initially hospitalised. Other medication:  Primaquine provided from day 7 to day 20. No report on symptomatic treatment.
Outcomes	Outcomes used in this review:  Total failure by day 28 (secondary endpoint for this study) Parasite clearance time Presence of gametocytes on day 7 Adverse events (Fever clearance time mentioned as an outcome, but results not reported)
Notes	Location:  India, six sites (New Delhi, Baroda, Berhampur, Karamsad, Jabalpur and Guwhati) Date:  July 1998 to October 2001 Malaria endemicity:  Stable throughout the year, increased incidence in rainy season (Dunne 2005b) Drug resistance: Not reported Method of determining parasite density:  Giemsa stained thin and thick smear, otherwise no details. Baseline data: Baseline data for gender, age, days of illness, mean P. vivax parasite density, reported for those randomized. Look similar. Drug source: Not reported Source of funding: Pfizer Global Research and Development, Inc. Other notes:  RII and RIII failures on day three were not a routine reason to treat with alternative therapies. The investigators report: “Patients who the investigator felt were not responding to therapy were given alternative therapies at their discretion, including quinine and mefloquine.” Slower resolution of parasitaemia and fever in azithromycin group may have biased findings to earlier labelling of treatment failure.
Risk of bias
Bias	Authors' judgement	Support for judgement
Adequate sequence generation?	Unclear risk	"Study drug was provided to sites in blocks of four."
Allocation concealment?	Unclear risk	"Sealed envelopes were available at each site to be opened only in case of emergencies. The integrity of these envelopes was monitored periodically at each site."
Blinding? All outcomes	Low risk	All parties remained blinded to treatment assignment
Incomplete outcome data addressed? All outcomes	Low risk	Low loss to follow up: 98 and 102 randomised in respective groups, 97 (99.0%) and 102 (100%) results at day 28. Not clear why one participant was excluded.
Free of selective reporting?	High risk	Definition of treatment failure potentially subjective. Slower resolution of parasitaemia and fever in azithromycin group may have biased findings to earlier labelling of treatment failure. Fever clearance time results not reported. No follow‐up visit at day 21.
Free of other bias?	Unclear risk	Not clear how much study sponsor was involved in reporting of results and adverse events