Plastic surgery resources for med students
Research is the single biggest lever for matching plastics, and the hardest thing to start from zero. Here's how to think about which studies carry weight, and where to begin.
Not all research is equal. Aim as high on the pyramid as your time and mentor allow, but a case report you finish beats a systematic review you don't.
A systematic review answers one focused question by searching the databases systematically, screening every hit against criteria set in advance (ideally by two independent reviewers), and appraising each included study for risk of bias. A meta-analysis then pools their results statistically. It sits at the top because it summarizes the whole literature rather than a single sample, but it inherits the quality of what it includes: a rigorous review of weak studies is still weak evidence.
Participants are randomly assigned to treatment or control, which balances confounders you know about and, crucially, the ones you do not. No observational design can do that, which is why one well-run trial outranks any single non-randomized study. Trials are uncommon in plastic surgery: you often cannot blind an operation, sham surgery raises ethical problems, and recruiting enough patients for an uncommon reconstruction is genuinely hard.
Defines groups by their exposure now and follows them forward to see who develops the outcome. Because exposure is recorded before the outcome occurs, it establishes temporality and avoids recall bias, the two things backward-looking designs struggle to guarantee. Strong for questions of prognosis and harm. The cost is time and money, since you wait for outcomes to happen.
Looks backward at data that already exists, either institutional charts or a registry such as NSQIP, TriNetX, or the NCDB. It is fast, inexpensive, and by far the most realistic way for a student to complete a real analysis, which is why so much of the plastic surgery literature looks like this. The trade-off is that you are limited to the variables someone else chose to record, and the design is open to selection bias, coding error, and unmeasured confounding.
Starts from the outcome rather than the exposure: assemble people who have the condition and a comparable group who do not, then look backward for differences in what they were exposed to. Efficient for rare outcomes, where a cohort study would need thousands of patients to see enough events. The hard parts are choosing a control group that is genuinely comparable and limiting recall bias.
Describes a consecutive group of patients who all received the same treatment, with no comparison group. It is a legitimate way to report a technique's outcomes and complication profile, and it is often the first evidence a new operation generates. Because nothing is being compared, you cannot attribute the result to the treatment itself.
A single patient. It is the lowest study-design tier, but still genuinely useful for rare presentations, unexpected complications, and novel techniques. For a student it is often the most realistic first publication, and it teaches the entire pipeline: consent, imaging, a literature search, writing to a journal's format, submission, and revision.
Editorials, invited commentary, un-systematic reviews, and bench work in animals or in vitro. This material is valuable for context, teaching, and seeing where a field is heading, but what gets included reflects the author's judgment rather than a reproducible search, so it cannot be graded as clinical evidence.
Hierarchy perOCEBM Levels of Evidence 2 (Oxford CEBM, 2011)Burns, Rohrich & Chung. Levels of evidence and their role in EBM. PRS 2011Sullivan et al. The level of evidence pyramid. PRS 2011
Four concrete next steps. Pick the one that fits where you are.
Educational guidance only. Adapt it to your own situation and follow your institution's policies.