1This video walks through one real systematic review from the first step to the last, using the steps we covered in the last video.

2The review is a Cochrane review on drugs for preventing malaria in pregnant women in endemic areas. It exemplifies rigorous systematic review methods, and it illustrates how the concepts translate into practice.

3Step one, their question. The review team formulated a clear question using the PICO framework. Population: pregnant women of any gravidity living in malaria-endemic areas. Intervention: any antimalarial drug chemoprevention regimen. Comparison: placebo or no intervention. And outcomes: maternal outcomes including mortality, severe anaemia, and parasitaemia, and infant outcomes including mortality, birthweight, and placental parasitaemia.

4Written out, the question was this. In malaria-endemic areas, what are the effects of malaria chemoprevention compared to no chemoprevention on maternal and infant health outcomes? Notice how they kept the question broad enough to capture all relevant evidence while being specific enough to be answerable. They didn't restrict to a single drug or outcome. They wanted comprehensive evidence on chemoprevention as a strategy.

5Step two, their search strategy. The team searched multiple databases using terms for each concept in the question. Their search covered a Cochrane specialized register, the Cochrane Central Register of Controlled Trials, then MEDLINE from 1966 to June 2014, then EMBASE from 1974 to 2012, then LILACS from 1982 to 2012, and finally the reference lists of identified trials. They also contacted researchers working in the field for unpublished data, confidential reports, and raw data.

6Step three, their criteria, specified before searching. Include: randomized controlled trials and quasi-randomized trials of any antimalarial drug regimen for preventing malaria, compared to placebo or no intervention. Exclude: studies comparing different active drug regimens, which went into a separate review.

7They also included regimens that were known to be effective against malaria parasites at the time, even if those drugs are now obsolete because of resistance. This illustrates an important principle. Those trials still inform our understanding of chemoprevention as a strategy. The question wasn't whether one particular drug works. It was whether chemoprevention works.

8Step four, the protocol. As a Cochrane review, the protocol was registered and published before the review was conducted. This pre-specification of methods protects against bias. The team couldn't secretly change their analysis approach after seeing the results.

9Steps five and six, collecting and screening. Running their search across databases, the team retrieved 181 records through database searching, plus 2 additional records from reference lists and expert contacts. Removing 2 duplicates left 179 unique records to screen. Two reviewers independently applied the inclusion criteria. Title and abstract screening on the 179 records excluded 126 as clearly not relevant, leaving 53 full-text articles to retrieve for detailed assessment.

10Full-text screening then excluded 31 articles with documented reasons, leaving 17 trials included, described in 22 separate publications. Each of the 31 excluded articles had a documented reason for exclusion. That transparency is essential for reproducibility, and it lets readers assess whether the exclusions were appropriate.

11Step seven, extraction. Using standardized forms, reviewers independently extracted study characteristics: authors, year, country, setting, and malaria transmission intensity. Population: sample size, parity, and HIV status where reported. Intervention details: the drug, and the dosing schedule, whether daily, weekly, or intermittent. And the outcomes: maternal mortality, severe anaemia, any anaemia, parasitaemia, then infant mortality, birthweight, and placental parasitaemia.

12The 17 included trials enrolled 14,481 pregnant women across eight countries, and they were conducted between 1957 and 2008. Three trials each came from Nigeria, from The Gambia, and from Kenya. Two each from Mozambique, from Uganda, and from Thailand. And one each from Cameroon and from Burkina Faso.

13Step eight, appraisal. Using Cochrane's risk of bias tool, the team assessed each trial, and their findings reveal a common pattern in older literature. 6 trials had adequate allocation concealment, meaning low risk of selection bias. 4 trials were quasi-randomized, meaning high risk of selection bias. And 7 trials had unclear risk. So only 6 of 17 trials met modern standards for preventing selection bias. Excluding the weaker trials didn't change the conclusions, which strengthens confidence in the findings.

14Step nine, the meta-analysis. With multiple trials measuring similar outcomes, meta-analysis was appropriate, and the review used GRADE to rate the quality of the evidence. For women in their first or second pregnancy, the results pointed the same way for mothers and for infants. Mothers were far less likely to develop severe anaemia, and far less likely to have malaria parasites detected during pregnancy, both rated high quality evidence. Infants were heavier at birth, less likely to be born with low birthweight, and less likely to show infection in the placenta, rated moderate to high quality. Every outcome the review examined moved in the same direction, which is what gives the finding its weight.

15Step ten, the write-up. The flow diagram documented the journey from 181 records to 17 included trials, and the Summary of Findings tables presented the key results in accessible format for decision-makers. The team acknowledged limitations honestly: most evidence came from Africa, the trials spanned five decades of changing drug resistance and healthcare contexts, and mortality outcomes were underpowered. They noted that benefits might be attenuated in settings with declining malaria transmission and improving antenatal care. Their conclusion was appropriately cautious yet actionable.

16We've covered a lot of ground. From choosing your review type, to building Boolean searches, to avoiding predatory journals, searching the literature involves many skills that improve with practice. But I want to close with something that transcends technique. Searching the literature is fundamentally an act of connecting with the collective knowledge of your field. When you search, you're joining a conversation that stretches back decades, sometimes centuries, and spans the globe. You're learning what questions others have asked, what they've found, and where they've hit dead ends. This is humbling. For any question you have, someone has probably thought about it before. Your contribution isn't to reinvent everything from scratch but to build on what's known, filling gaps and advancing understanding incrementally.

17Searching the literature also reveals how a field thinks. The concepts that have MeSH terms, the journals that exist, the research questions that get funded: these reflect assumptions, priorities, and power structures. As global health researchers, we should notice whose voices appear in the literature and whose are absent. The gaps in the literature are as informative as the papers you find. They reveal where evidence is needed, where voices are missing, and where your contribution might be most valuable. I encourage you to embrace searching as a core scholarly practice to develop. The better you get at finding evidence, the better your research questions, the stronger your arguments, and the more impactful your contributions.