Slide 1Critical appraisal means picking up a scientific paper and systematically evaluating whether you should believe what it claims. This chapter is about how to do that, and the first skill is the smallest one: working out what kind of paper you are holding.

Slide 2In 1972, a Scottish epidemiologist named Archie Cochrane published a slim book called Effectiveness and Efficiency that would revolutionize medicine. His argument was blunt: most medical practice wasn't based on evidence. Doctors did what they were taught, what their mentors did, what seemed reasonable, but rarely what had been rigorously tested. He called for medicine to ground itself in randomized controlled trials, and for someone to actually synthesize all that evidence so clinicians could use it.

Slide 3Two decades later, a group of physicians at McMaster University in Canada took that challenge seriously. Gordon Guyatt coined the term evidence-based medicine in 1991, and alongside David Sackett and others, began teaching a new approach: clinicians should learn to find, appraise, and apply research evidence to patient care. They should read the studies and judge for themselves.

Slide 4This created a problem. Most clinicians had never been taught to read research critically. They could read a paper, sure, but could they tell whether the findings were trustworthy? Whether the methods were sound? Whether the conclusions followed from the data? The evidence-based medicine movement realized that reading scientific literature was more than reading. It was critical appraisal, a skill that had to be taught.

Slide 5So that is what we are doing here. By the end of the chapter you'll know how to classify studies by type so you're asking the right questions from the start. You'll learn to read methods sections first, and to identify the features that determine whether a study's findings are credible: research design, sampling, measurement, and potential biases. You'll be able to interpret effect sizes and confidence intervals. And you'll develop strategies for reading efficiently, because nobody has time to deeply analyze every paper they encounter. Critical appraisal is reading with your eyes open, appreciating what a study can tell you while recognizing its limitations. Every study has limitations. The question is whether those limitations undermine the conclusions.

Slide 6Before you dive into any paper, take 30 seconds to figure out what species of study you're dealing with. Look at the title and the methods section just long enough to classify the study type. Why does this matter so much? Because different study types require different reading strategies. You wouldn't evaluate a qualitative study on community health worker experiences the same way you'd evaluate a randomized controlled trial of a malaria vaccine. They're answering different questions using different logic. Knowing the study type tells you which critical appraisal toolbox to reach for.

Slide 7Experimental studies randomly assign participants to conditions. The gold standard is the randomized controlled trial, where researchers actively intervene and randomize who gets what treatment. If you see words like randomized, randomly allocated, or experimental group, you're probably looking at an experimental design. Quasi-experimental studies attempt to estimate causal effects of some manipulation without random assignment. Look for designs like difference-in-differences, interrupted time series, or regression discontinuity. These are common in global health because true randomization isn't always ethical or feasible. Observational studies observe existing groups without intervention. The big three are cohort studies, which follow people over time from exposure to outcome; case-control studies, which start with outcomes and look backward at exposures; and cross-sectional studies, which capture a snapshot at one moment.

Slide 8Qualitative studies explore experiences, meanings, and processes through interviews, focus groups, or observation. Look for words like thematic analysis, phenomenology, grounded theory, or ethnography. Systematic reviews and meta-analyses synthesize findings across multiple studies. They're identified by explicit search strategies, inclusion criteria, and pooled effect estimates. There are too many research designs to list here, and the chapter names a few more. You'll learn about each one in Part III of this book.

Slide 9Here is how I classify a paper quickly. Imagine I've just downloaded a paper titled: Impact of community health workers on maternal health outcomes in rural Kenya, a mixed methods evaluation. Step one, ten seconds. Scan the title. Key words: community health workers, impact, mixed methods, evaluation. That tells me the study is assessing an intervention's effects using both quantitative and qualitative approaches. Step two, twenty seconds. Jump to the Methods section and look for the design description. I see: we conducted a quasi-experimental study comparing villages with trained community health workers to villages without, alongside in-depth interviews with 30 community health workers and 45 community members.

Slide 10Step three, fifteen seconds. Confirm the classification in the abstract. This is a quasi-experimental design, the quantitative component, plus qualitative research, the interview component. Villages weren't randomly assigned to receive community health workers. The comparison is between villages that happened to have them and villages that didn't. Step four, fifteen seconds. Note what I'll need to evaluate, and determine which critical appraisal and reporting guidelines could help me evaluate the study. Total time: sixty seconds. I now know what species of study I'm dealing with and what questions to ask.

Slide 11This rapid classification becomes automatic with practice. You're getting your bearings so you know which appraisal lens to apply. One note that holds for all six of these videos: the chapter carries more than they do, so read it for the additional details not covered here. Next, the section where critical appraisal actually begins, and the questions to ask of it.