Andersson 2008.

Methods	Design: prospective cohort study Study dates: March 2004 to November 2006 Malaria transmission pattern and local antimalarial drug resistance: malaria attack rate of 44% with P falciparum in another similar study at the time Adverse event monitoring: patient self‐reported questionnaire
Participants	Number enrolled: 690 soldiers sent questionnaire, 609 respondents Inclusion criteria: all Swedish soldiers deployed to Liberia within the study dates Exclusion criteria: none stated. Factors influencing drug allocation: "...mefloquine was prescribed to almost all soldiers in the first two contingents and to about two‐thirds in the last three contingents. The remaining soldiers were recommended atovaquone/ proguanil. The latter group consisted mainly of those with body weight < 70 kg and those who had already experienced adverse events with mefloquine. No other drug regimes were used". Country of recruitment: Sweden Country of malaria exposure: Liberia Duration of exposure to malaria: 6 months Type of participants: military
Interventions	1. Mefloquine* 2. Atovaquone‐proguanil* *dosing regimen not specified
Outcomes	Included in the review: 1. Adverse events; any, nausea, vomiting, abdominal pain, diarrhoea, headache, dizziness, abnormal dreams nightmares, insomnia sleep disturbance, depression 2. Serious adverse events; serious 3. Adverse events; other (concentration difficulties, mouth ulcers, fever, muscle pain) 4. Discontinuations of study drug due to adverse effects Outcomes assessed not included in the review: 5. Clinical cases of malaria 6. Overall satisfaction with the drug 7. Whether they would take the drug again 8. Measures of adherence to the drug regimen (data provided on aggregate)
Notes	Funding sources: Not stated
Risk of bias
Bias	Authors' judgement	Support for judgement
Other bias	Unclear risk	1. Confounding: moderate Information on potential confounders is not provided across prophylactic groups 2. Selection of participants into the study: moderate 609/690 (88%) response rate 3. Measurement of interventions: low All participants were issued with the study drug. 4. Departures from intended interventions: low Switches were recorded and reported 5. Missing data: serious Outcomes were reported from 3 of 5 cohorts. No information was provided for 2 remaining cohorts. 6. Measurement of outcomes: serious The outcome measure was subjective, participants and personnel were not blinded. 7. Selection of the reported results: low All outcomes prespecified in the introduction were reported. 8. Other: moderate Study sponsor not mentioned, but 2 study authors worked for GlaxoSmithKline