When Is It Necessary to Read the Original Study Rather Than a Review Citing It?

When is a review enough, and when should you open the study it cites?

You are reading a good review and find exactly the evidence you need. The authors summarize an earlier study, provide the citation, and perhaps even report the effect or conclusion. Do you need to retrieve the original paper too?

Not automatically. Reviews exist partly to synthesize research so that researchers do not have to reconstruct an entire literature from scratch every time. A rigorous systematic review may be the better source when your question concerns the overall evidence across many studies.

The answer changes when your sentence depends on what one particular study actually did or found. At that point, the review becomes an intermediary between you and the evidence, and details lost or altered during summarization may matter.

Read the original study when the study itself carries your claim

The source you need depends on what you want to know

Reviews and original studies answer different kinds of questions

An original empirical study gives you direct access to how particular research was conducted and reported. You can examine the participants, sampling, intervention or exposure, measurement instruments, analytical procedures, results, uncertainty, and limitations.

A review operates at another level. It brings multiple studies together and may compare, synthesize, critique, or statistically combine their findings. A systematic review can therefore answer questions that no individual primary study can answer well, such as whether findings are reasonably consistent across a body of research.

Neither source type is inherently superior. They occupy different positions in the evidence chain.

Read the original when you report a specific result

If you want to write that a study found a particular effect size, percentage, correlation, odds ratio, mean difference, confidence interval, or other quantitative result, check the original report whenever feasible.

A review may accurately report the number, but the original study shows what the number refers to. Was it adjusted or unadjusted? Was it the primary outcome or one of many secondary outcomes? What was the denominator? At what time point was it measured? Did the confidence interval indicate substantial uncertainty?

These details can materially alter the interpretation of a seemingly simple number.

Read the original when methodological details affect your interpretation

A review necessarily compresses methods. That is usually appropriate because a synthesis cannot reproduce every methodological detail of every included study.

But suppose your argument depends on whether participants were randomly assigned, whether an outcome was objectively measured or self-reported, how missing observations were handled, which covariates were included, or whether researchers measured actual behavior rather than behavioral intention. Those questions require the original report and, sometimes, its supplementary materials.

The same principle applies when you are adapting a procedure, instrument, intervention, or analytical method. A review's summary is rarely an adequate substitute for the detailed methodological source.

Read the original when you are attributing something to its authors

If you write “Smith and colleagues concluded...” or “The study demonstrated...,” you are making a specific attribution. Verify it against the source to which you are attributing it.

This is particularly important because citation accuracy is imperfect. A systematic review and meta-analysis of quotation accuracy in medical journals estimated a total quotation-error rate of 25.4%, although estimates varied substantially across included studies. A later methodological review using a more specific denominator and eligibility approach estimated quotation errors in original medical research articles at 14.5%, illustrating how estimates depend on definitions and methods.

Those estimates come from medical literature and should not be treated as universal rates for all disciplines. They do establish a more modest point: a secondary author's representation of a primary source should not automatically be assumed to be exact.

Read the original when the claim seems stronger than you expected

Watch for words such as “causes,” “proves,” “demonstrates,” “effective,” “established,” “always,” or similarly strong formulations. If the review uses language that carries substantial certainty, inspect the study to determine what its design and results actually permit.

A cross-sectional association, for example, does not become causal merely because a later source describes it with a causal verb. Likewise, a statistically significant result on one outcome does not necessarily justify describing an entire intervention as effective.

If the difference appears substantive, the issue may be that a secondary source has strengthened the wording of the original finding.

Read the original when the claim is central, surprising, or consequential

Verification effort should be proportional to argumentative importance. A peripheral background statement usually does not deserve the same investigative effort as a finding on which your hypothesis, theoretical framework, interpretation, or recommendation depends.

Claims that are surprising or repeatedly described as established can also deserve closer inspection. Greenberg's detailed analysis of one biomedical citation network found that citation bias, amplification, and several forms of citation-based invention helped create apparent authority around a claim. Importantly, Greenberg distinguished papers that actually contained data addressing the claim from papers that merely participated in its citation network.

This does not mean that every repeated claim is suspect. It means that the popularity of an interpretation cannot substitute for checking its evidential provenance when the stakes are high.

Use a review when your claim is genuinely about the body of evidence

Now reverse the situation. Suppose your question is whether an intervention generally improves an outcome across the available research. One original study cannot answer that question by itself.

A well-conducted systematic review or meta-analysis may be more appropriate because it identifies studies using explicit methods, evaluates the evidence across them, and may quantify their combined results. In this situation, replacing the review with one convenient primary study could actually weaken your representation of the literature.

Reading the original does not mean believing the original

Primary-source checking is sometimes misunderstood as deference to the first paper. It should be the opposite. Opening the original gives you the information necessary to evaluate it more critically.

The primary study may have weaknesses that the review identifies more clearly. Later studies may contradict it. A meta-analysis may show that an initially striking result is unusual within the wider evidence base.

Tracing an important claim to its original source establishes what was originally reported. Evaluating whether that report deserves confidence is a separate task.

Sometimes the original study is not accessible

You may encounter an older article, obscure conference paper, unavailable dissertation, or publication behind an access barrier. If you cannot obtain the original, do not silently cite it as though you read it.

Instead, follow the secondary-citation conventions appropriate to your citation style and publication venue, or find another source that you can inspect directly. What matters is preserving the actual path by which you know the information.

When opening the original paper changes what you can write

The review was not necessarily wrong. Its purpose required compression. The problem would arise if you used that compressed description to make a more specific or stronger claim than you had actually verified.

Primary studies and reviews are not competing source types

Choose your reading depth according to the claim's job in your paper

You do not need to turn every citation into a miniature systematic review. Instead, decide what information you need from the source and how much argumentative weight the claim carries.

The practical rule is simple enough: the closer your sentence gets to claiming what one study specifically did or found, the stronger the reason to read that study yourself.

Should you open the original study?

Questions about reviews and original studies

Read at the level of evidence your claim requires

How to Cite This Guide

This guide is intended to be read, shared, and used in research, teaching, and academic work. If you draw on its ideas, explanations, or other content, please acknowledge the source by citing the guide. Doing so gives appropriate credit and helps your readers locate the original resource.