01 · The Question
If case reports are all you have, should you use them?
You search for trials, cohort studies, or other comparative research and find almost nothing. What you do find are individual case reports and perhaps a few small case series.
It is tempting to dismiss them immediately because they sit low in conventional evidence hierarchies. The opposite temptation is equally problematic: if case reports are the only evidence available, perhaps they should simply be treated as the best evidence and used to answer the question.
Neither response is particularly useful.
A case report can contain important evidence. It can document something that has rarely or never been observed, draw attention to an unexpected adverse event, describe an unusual presentation, or show that a particular event occurred under specific circumstances. What it usually cannot tell you is how often that event occurs, whether an intervention generally works, or whether the observed exposure caused the outcome.
The relevant question is therefore not whether case reports are “good” or “bad” evidence. It is what kind of claim the design can reasonably support.
03 · What You Need to Know
A case report can answer “Did this happen?” better than “How often does this happen?”
Case reports describe individual patients, participants, events, or observations in detail. Case series describe multiple cases, usually without a comparison group. Their strength lies in observation rather than comparison.
That distinction becomes particularly important when the literature is sparse.
Case reports can reveal phenomena larger studies were not designed to detect
Some research questions concern events that are rare, unexpected, newly emerging, or difficult to capture prospectively. A large trial may contain no examples simply because the event is extremely uncommon. A case report, by contrast, may exist precisely because something unusual occurred.
Cochrane guidance on adverse effects recognizes that spontaneous case reports and case series may help signal rare and previously unknown events. Such reports can therefore alert researchers to potential harms or phenomena that deserve more systematic investigation.
The same basic logic extends beyond adverse effects. Case reports may document rare diseases, unusual clinical presentations, unexpected intervention responses, novel complications, uncommon combinations of circumstances, or emerging phenomena for which larger studies have not yet accumulated.
The value is often informational rather than estimative: the report tells you that something has been observed and describes the circumstances in which it occurred.
Case reports are particularly useful for signal generation
A signal is not the same as a conclusion.
Suppose several case reports describe the same unexpected adverse event after exposure to a new intervention. The pattern may justify further investigation. It might influence which adverse outcomes researchers monitor in subsequent studies or which hypotheses deserve testing.
But the reports alone do not tell you the probability of the event among everyone exposed. They also do not necessarily establish that the intervention caused it.
Signal
An observation or pattern suggesting that a phenomenon may warrant further investigation.
Established effect
A conclusion supported by evidence capable of evaluating the relationship with appropriate comparison, measurement, and control of alternative explanations.
This distinction prevents an important inferential leap: moving from “this occurred after the intervention” to “the intervention causes this outcome.”
The missing denominator is a major limitation
Imagine that you find twelve published case reports describing a particular complication after an intervention. Twelve may sound alarming.
But twelve out of what?
If twelve cases occurred among 200 exposed people, the situation looks very different from twelve among twenty million. Published case reports typically do not provide the denominator needed to estimate incidence, prevalence, risk, or event rates. Cochrane explicitly notes this limitation when discussing spontaneous reports.
The existence of cases may still matter. Their frequency remains unknown from those reports alone.
Case reports usually cannot determine comparative effectiveness
Suppose a patient receives an intervention and improves dramatically. That observation may be clinically interesting. It does not establish that the intervention caused the improvement.
Without an appropriate comparison, alternative explanations remain. The condition may have improved naturally. Another treatment may have contributed. Regression toward the mean may matter. The patient's characteristics may be unusual. The outcome may have been measured selectively. Publication itself may be influenced by how remarkable the outcome appeared.
This is why a collection of successful case reports should not be interpreted as though it were a trial without a control group conveniently attached.
Publication processes can make case reports unusually selective
Routine, expected, and uneventful cases are unlikely to inspire a case report. Dramatic outcomes, unusual presentations, unexpected harms, novel responses, and striking coincidences are much more likely to be written up and published.
This selectivity is part of what makes case reports useful for detecting unusual observations. It is also why counting published cases cannot reliably describe how typical those observations are.
There may also be duplicate reporting, incomplete clinical histories, variable diagnostic certainty, inconsistent follow-up, and selective presentation of details. The existence of a publication does not eliminate these limitations.
Case series add cases, but they do not automatically solve the design problem
A case series can reveal whether similar observations occurred in multiple individuals and may provide richer information about patterns across cases. Yet simply increasing the number of cases does not create a comparison group or a reliable denominator.
A series of fifty carefully described cases can be considerably more informative than one isolated report for some descriptive questions. It still does not automatically establish comparative effectiveness or quantify risk.
The evidentiary value depends on the claim
Question
What case reports may contribute
Major limitation
Can this phenomenon occur?
Potentially important direct observation
Verification and alternative explanations still matter
What can an unusual presentation look like?
Detailed descriptive information
May not represent typical presentations
Could this be an unexpected harm?
Signal generation, especially for rare events
Usually cannot estimate risk or prove causality
How common is the event?
Usually little direct help without a denominator
Cannot reliably estimate frequency from published cases alone
Does the intervention work better than an alternative?
May generate a hypothesis
No appropriate comparison for estimating comparative effect
What happens in a rare condition with little other research?
May provide valuable descriptive evidence
Generalization beyond reported cases remains uncertain
Do not force case reports into a question they were not designed to answer
If higher-level direct evidence is genuinely scarce , case reports may become relatively more important because they contain information unavailable elsewhere. Their methodological limitations do not disappear as the literature becomes smaller.
This principle is easy to overlook. “Best available evidence” is a comparative description. It does not transform the capabilities of a study design.
Case reports can therefore be the best available evidence and still leave the main question substantially unresolved.
Watch Out
Do not convert absence of stronger evidence into permission to make stronger claims from case reports. Scarcity changes how valuable limited evidence may be, not what that evidence can logically establish.
04 · A Practical Example
What can six unusual case reports actually tell you?
Hypothetical Example
An unexpected neurological event after a new intervention
Suppose six published case reports describe a similar neurological event occurring shortly after patients received a relatively new intervention. No randomized trial was designed to investigate this particular event, and large observational safety studies are not yet available.
Observation
The reports establish that clinicians have observed and documented a similar event after exposure in several individuals.
Signal
The temporal relationship and recurring pattern may justify treating the event as a potential safety signal worthy of targeted investigation.
What remains unknown
The reports do not provide a reliable denominator, so you cannot calculate how frequently the event occurs among exposed people.
Causal uncertainty
You examine alternative explanations, underlying conditions, concurrent treatments, diagnostic certainty, timing, and whether a biologically plausible mechanism exists. The reports may strengthen or weaken suspicion, but they do not by themselves establish causality.
Appropriate conclusion
You report that a potential rare adverse event has been described and warrants further investigation rather than concluding that the intervention has been shown to cause the event at a particular rate.
The six reports matter. The mistake would be asking them to provide information they do not contain.
06 · What This Means for You
Match the case evidence to the narrowest defensible claim
If case reports are among the strongest evidence available, start by identifying the specific information you need from them. Do not begin with their position in an evidence hierarchy.
A simple decision framework
If your question concerns a rare, novel, or unexpected phenomenon
Case reports may provide important descriptive evidence and generate a signal for further investigation.
If you need to characterize an unusual presentation or clinical course
Detailed case descriptions may be useful, while recognizing that reported cases may not represent typical cases.
If you need to estimate incidence, prevalence, or risk
Do not derive the estimate from case reports without an appropriate denominator and ascertainment framework.
If you need to determine comparative effectiveness
Treat case reports as hypothesis-generating or supporting observations rather than substitutes for an appropriate comparative design.
If case reports are all that exist
Use them for the claims they can support and make the remaining evidence gap explicit.
Depending on the broader question, other forms of evidence may also deserve attention. A sparse literature may contain small qualitative studies that answer different but important questions , or relevant information may exist outside conventional journals and require a more extensive search of grey literature .
None of these sources needs to impersonate a randomized trial to be useful. Their value comes from asking them the questions they are equipped to answer.
07 · A Quick Checklist
Before drawing conclusions from case reports
When case reports are part of a sparse evidence base, check:
Identify the precise question the case report can help answer before interpreting it.
Verify the diagnosis, event, exposure, intervention, and temporal sequence as far as the report allows.
Look for alternative explanations, concurrent exposures, underlying conditions, and other plausible causes.
Check whether multiple reports may describe overlapping or duplicate cases.
Do not calculate frequency or risk from published case counts unless a valid denominator and ascertainment process are available.
Distinguish an observed temporal association from evidence establishing causation.
Consider whether selective publication of unusual cases could distort your impression of how typical the phenomenon is.
State clearly whether the evidence documents a phenomenon, generates a signal, supports a hypothesis, or contributes to another narrower conclusion.
11 · Cite this Guide
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.
Recommended (Field Guide)
APA
MLA
Chicago
Copy Citation
