Free scoping read. A written assessment of your manuscript in two business days, at no cost.

Start one
Contents

Academy · Research Publication · Case reports and EBP

The Importance of Case Report Writing in Evidence-Based Practice

The evidence hierarchy places the case report at its lowest level, and clinicians often interpret this as indicating that it contributes little. This inference is incorrect. A hierarchy ranks designs by their ability to estimate an average treatment effect, which is one question among several; for first signals, unexpected harms, rare phenotypes and single-patient response, the case report is not the weakest available design but the only one.

Pubrica AcademyWritten by the medical writing and biostatistics teams Reviewed byPubrica biostatistics team Published9 October 2026 Last checked9 October 2026 Reading time13 minutes Sources26 peer-reviewed

Evidence-based practice has never held that only randomised trials count. Sackett, Rosenberg, Gray, Haynes and Richardson, in the BMJ in 1996, defined it as the integration of best available evidence with clinical expertise and patient values, and the word “available” is essential. For many clinical questions the best available evidence is a small number of described cases, and treating that as equivalent to no evidence is a misreading with practical consequences: it suppresses the reporting of precisely the observations the literature most needs.

Learning objectives

  • Explain what the evidence hierarchy ranks, and therefore what its lowest level does and does not signify
  • Identify the four clinical questions for which a case report is the appropriate design rather than a weak substitute
  • Distinguish a case report, a case series and a single-patient trial, and determine which a given observation supports
  • Write a case report so that it functions as evidence rather than as an anecdote with references
  • Appraise a case report written by others, including the conclusions it cannot support
  • Address consent and identifiability appropriately, since a case report describes an identifiable individual

Common misconceptions

  • “Case reports are the lowest level of evidence, so they are weak.” The hierarchy ranks designs by their ability to estimate an average treatment effect under confounding. For first signals and rare phenotypes, no higher design exists, so the case report is not a weak option but the only one.
  • “A case report cannot change practice.” It can, where the observation concerns a harm. A single well-documented adverse event with a plausible mechanism and a temporal relationship has repeatedly been the first entry in a sequence that ends in a label change.
  • “Case reports are not needed now that we have large databases.” Databases detect what their coding systems represent. Splinter and colleagues, in the New England Journal of Medicine, studied patients with previously undiagnosed disease — by definition, patients no code describes.
  • “A case report and an N-of-1 trial are the same kind of thing.” They are not. Serpico and Maziarz, in the European Journal for Philosophy of Science, set out the pragmatic N-of-1 design as a method for investigating individual treatment response — a prospective, structured, often randomised design, not a retrospective description.
  • “If it is interesting, it is publishable.” Interest is not the criterion. A case report merits publication by being informative about something generalisable, such as a mechanism, a harm, a diagnostic pitfall or a boundary condition, and by being documented sufficiently well for a reader to judge the inference.

The basis of ranking in the evidence hierarchy

The evidence hierarchy is a tool for one purpose: ordering designs by how well they support a causal claim about an average treatment effect in the presence of confounding. On that criterion a well-conducted randomised trial ranks above a cohort study, which ranks above a case-control study, which ranks above a case report. The ordering is correct; the inference commonly drawn from it is not.

Bellomo and Bagshaw, in Critical Care, make the argument directly: classifying evidence from clinical trials by design alone ignores dimensions that matter — the plausibility of the mechanism, the consistency of the observation, the magnitude of the effect, the specificity of the association 12. Balshem and colleagues, in the Journal of Clinical Epidemiology, formalise part of this in the GRADE approach, where design sets a starting rating that is then adjusted up or down for other features 7. A case report is not assigned a grade at all in that system, which is sometimes read as dismissal and is better read as a statement that the system was built for a different question 7.

The history of the field is instructive. Sur and Dahm, in the Indian Journal of Urology, trace the development of evidence-based medicine; the Evidence-Based Medicine Working Group’s 1992 JAMA paper announced it as a new approach to teaching the practice of medicine; Smith’s BMJ editorial “Where is the wisdom?” captured the problem it was responding to 8,22,23. None of that work claimed that observations below the top of the hierarchy were worthless. The claim was narrower: that unsystematic clinical impression is a poor estimator of treatment effect 18,20. That claim is valid, but it is not an argument against describing clinical observations.

Clinical questions answerable only by case reports

Has this ever happened? The existence question. Before anything can be counted it has to be noticed, and the first instance of a phenomenon cannot be studied by a design that requires a cohort. Every established association has a first report.

Could this treatment cause this harm? The signal question. Harms are rare, late and heterogeneous, and trials are powered for efficacy. A single case with a clear temporal relationship, a plausible mechanism, and ideally a de-challenge and re-challenge, generates a hypothesis that nothing else in the system would generate.

What does this disease look like when it does not look like the textbook? The phenotype question. Rare presentations, rare diseases and atypical courses are the content of the case literature because no other design can accumulate them. Splinter and colleagues, in the New England Journal of Medicine, examined the effect of genetic diagnosis on patients with previously undiagnosed disease, and Ashley, Butte, Wheeler and colleagues, in the Lancet, reported a clinical assessment incorporating a personal genome — both describing patients that no existing category contained 3,10.

Why did this happen here? The mechanism question. A case can be investigated in depth that no trial participant ever is — with imaging, sequencing, pathology and longitudinal sampling that would be infeasible at scale. Shrager and Tenenbaum, in Nature Reviews Clinical Oncology, describe a rapid-learning model for precision oncology that depends on exactly this depth of individual characterisation 5.

Our pages on the uses of case reports in medical literature and case studies in medical education cover the adjacent functions, including teaching.

Which design answers which question, and where the case report is the only option

Clinical questionDesign that answers itIs a case report sufficient?
Does this treatment work better than the alternative, on average?Randomised trial, or a synthesis of themNo, and a case report should not be offered as partial evidence for it
How common is this?Cohort or population study with a denominatorNo. A case report has no denominator and cannot support a rate
Has this ever happened?Case reportYes. This is the existence question, and no other design addresses it
Could this treatment cause this harm?Case report generates the signal; pharmacovigilance and analytic designs test itYes for the signal. No for the test
What does this condition look like when it is atypical?Case report, then case seriesYes. Atypical presentations cannot be accumulated any other way
Why did this happen in this patient?Case report with deep individual characterisationYes, and the depth available in a single case is unavailable at scale
Does this treatment work in this patient?Single-patient (N-of-1) trial, where the treatment is reversible and the condition stableNo. Use the N-of-1 design if its conditions are met, and say so if they are not

Compiled from the design and appraisal literature cited in this article, including Serpico D and Maziarz M in the European Journal for Philosophy of Science on pragmatic N-of-1 designs and Bellomo R and Bagshaw SM in Critical Care on the dimensions a design-based hierarchy omits.

Limitations of the case report

An explicit statement of the limits is what makes the case for the design credible, and it is the element that most discussions of case reports handle poorly, through overclaiming.

A case report cannot estimate a rate, an incidence or a risk, because it has no denominator. It cannot establish causation, because a single observation cannot exclude coincidence. It cannot support a treatment recommendation, because it has no comparator and no control for regression to the mean, for the natural history of the condition, or for the several other things that were happening to the patient at the same time. Nor can it indicate whether the phenomenon is common or extremely rare, which is the question a reader most often wants answered.

Tonelli, in Respiratory Care, sets out the limits of evidence-based medicine from the other direction — population estimates underdetermine individual decisions — and in Academic Medicine defends expert opinion as a distinct kind of warrant 16,17. Read together, the two literatures locate the case report accurately: it is a source of hypotheses and of boundary conditions, and a legitimate input to clinical reasoning under uncertainty, but it is not a source of effect estimates. A discussion section that states this explicitly is stronger than one that merely asserts that “further studies are warranted”.

Reporting of harms and the problem of publication bias

Of all the functions of the case literature, harms detection has the strongest claim to changing practice, and it is also where the reporting system is most clearly failing — for a reason that merits examination.

DeVito and Goldacre, in BMJ Evidence-Based Medicine, set out publication bias in the catalogue of biases: the published literature is a biased sample of the conducted literature, and the direction of the bias is toward positive and expected findings 2. For case reports the bias operates differently and more adversely. An unexpected harm is, by construction, an unexpected finding — which is precisely what journals have historically been least interested in and authors least motivated to write up. The result is a systematic under-representation of exactly the observations on which the safety system depends.

The practical consequence for an individual clinician is a low threshold for writing. An observation that would have changed the clinician’s own decision, had it been reported earlier, is sufficient reason to report it, irrespective of whether it appears substantial enough to publish. A case report of a harm does not need to establish causation to be useful; it needs to be documented well enough that someone conducting a later systematic search can assess it. That requires the temporal sequence, the concurrent exposures, the de-challenge and re-challenge if they occurred, and the alternative explanations considered and the reasons they were rejected.

Distinctions between case reports, case series and single-patient trials

Three designs are routinely conflated, and the distinction determines what an author can claim.

A case report describes one patient retrospectively. It records the observation as it occurred; no comparison is built in.

A case series describes several patients sharing a feature, also retrospectively. It permits description of a range but still has no denominator and no comparator, and the selection of which patients to include is the dominant source of bias. Our pages on the difference between case series and case reports and patient numbers in case series cover where the boundary sits.

A single-patient trial is a different kind of object. Serpico and Maziarz, in the European Journal for Philosophy of Science, discuss the pragmatic N-of-1 design as a method for investigating individual treatment response, as distinct from an averaged effect 1. It is prospective, uses multiple crossover periods within the same patient, and can be randomised and blinded. It supports a causal claim about that individual, which is something no case report can do and no parallel-group trial attempts 1.

The practical rule is as follows: if the treatment is rapidly reversible, the condition is stable, and the outcome is measurable over short periods, a single-patient trial is feasible and is a stronger design than a case report for the same clinical question. If any of those three conditions is not met, the case report is the available design, and stating this in the methods is preferable to implying a design that was not used.

Three designs that are routinely conflated

DesignDirectionComparison built in?What it can support
Case reportRetrospective, one patientNoneThat the phenomenon occurred; a hypothesis about mechanism or harm
Case seriesRetrospective, several patients sharing a featureNone; selection of cases is the dominant biasThe range of a presentation; a sharper hypothesis. Still no rate
Single-patient (N-of-1) trialProspective, one patient, multiple crossover periods, can be randomised and blindedYes — the patient is their own control across periodsA causal claim about treatment response in that individual

Distinctions as set out in the design literature cited here. The N-of-1 characterisation follows Serpico D, Maziarz M. Averaged versus individualized: pragmatic N-of-1 design as a method to investigate individual treatment response. European Journal for Philosophy of Science.

Writing a case report that functions as evidence

The difference between a case report that enters the evidence base and one that does not lies almost entirely in the completeness of the description. A reader conducting a systematic search years later must be able to assess the case against inclusion criteria that have not yet been written, which means reporting more than seems necessary at the time.

The elements that matter are the chronology with actual intervals rather than “subsequently”; all concurrent exposures, including those the authors consider irrelevant; the diagnostic criteria applied and the tests that were negative as well as positive; the alternative explanations considered and the specific reason each was rejected; and the follow-up duration with its end state. The structured reporting guidance for case reports exists to enforce this — see our CARE checklist page and CARE-compliant case report support.

Two elements are frequently handled poorly by authors. First, the discussion should state what the case cannot establish, in those terms. A paragraph that says “this case cannot establish causation; the temporal relationship and the response to withdrawal are suggestive, and the alternative explanation of X was excluded by Y but Z remains possible” is considerably more useful, and considerably more likely to be cited, than one claiming the case “demonstrates” anything. Second, consent and identifiability are not a formality. A case report describes a specific identifiable person, and the requirements are substantive — see journal consent requirements for case reports, patient-identifiable consent, and our worked example on a transplant case report under GDPR.

The discussion paragraph of a case report describing a suspected drug-related harm

Overclaims, and is therefore unusable

This case demonstrates that the drug can cause this adverse reaction. To our knowledge this is the first report of such an association. Clinicians should be aware of this potentially serious complication and should consider alternative agents in similar patients. Further studies are warranted to confirm these findings.

“Demonstrates that the drug can cause” is a causal claim one observation cannot support. A reader who notices this discounts the whole report, including the parts that were well observed.

“First report” asserted without a described search. Priority claims require the search that justifies them; otherwise they suggest that no search was conducted.

A practice recommendation from a single case. “Consider alternative agents” is advice with no comparator, no denominator and no estimate of the risk being avoided.

Alternative explanations are absent. With no record of what was considered and rejected, a later reviewer cannot assess the case against their own criteria and must exclude it.

“Further studies are warranted” names no study. The sentence appears where a testable hypothesis should be.

Claims exactly what one case supports

This case cannot establish causation, and we do not claim it does. Three features support a drug-related mechanism: the reaction began within a defined interval of the first dose, stated above; it resolved after withdrawal without any other change in management; and it recurred on inadvertent re-exposure, with the same interval and the same clinical picture. Three alternative explanations were considered. An infective cause was excluded by the investigations listed in Table 1, all negative. The underlying condition itself can produce a similar picture, but had been stable for the period stated and did not fluctuate with drug exposure. A concurrent medication, started in the same week, remained unchanged throughout both the resolution and the recurrence, which argues against it without excluding an interaction. What this case therefore supports is a hypothesis, specified as follows, which could be tested against spontaneous reporting data or in a cohort with adequate exposure. We would not change practice on the basis of this report alone, and we describe it so that it can be assessed by anyone later searching for cases of this type; the reporting items in the relevant case report guideline are addressed in the supplementary file.

The causal limit is stated first, which makes everything after it credible rather than defensive.

The three supporting features are specific and of the kind a pharmacovigilance assessor weighs, including the inadvertent re-exposure.

Each alternative explanation is named with the reason it was rejected, and the one that cannot be excluded is acknowledged rather than dismissed.

The output is a specified hypothesis with the data source that could test it, rather than a call for unspecified further study.

The authors explicitly decline to recommend a practice change, which is what a single case warrants and what makes the report usable by a later reviewer.

A case report earns its place in the evidence base by being assessable, not by being conclusive. The second version will be found, included and weighed by someone conducting a systematic search; the first will be excluded for insufficient reporting even though the underlying observation may have been the better one.

Critical appraisal of published case reports

Reading case reports well is a separate skill from reading trials well, and the questions are different. The task is not to assess internal validity against a counterfactual but to assess whether the observation is credible and whether the inference the authors draw is proportionate.

Five questions do most of the work. Is the chronology specified with intervals, or is it narrated? Are the concurrent exposures listed, or only the one the authors are interested in? Were alternative explanations considered and dismissed with stated reasons, or not mentioned? Is the follow-up long enough that the outcome is settled? And does the discussion claim more than a single observation can support?

A report that fails the first four is not usable as evidence, whatever the phenomenon. A report that passes them and fails the fifth is usable — the reader discounts the authors’ conclusion and retains the observation. That distinction should be made explicitly when citing such a report: “a case described by X and colleagues, in which the authors propose a causal role we consider unproven” is a legitimate and precise use of the literature.

Ilic and Maloney, in Medical Education, reviewed methods of teaching evidence-based medicine to trainees; Maggio, Tannery, Chen, ten Cate and O’Brien, in Academic Medicine, reviewed EBM training in undergraduate medical education; Meats, Heneghan, Crilly and Glasziou, in Medical Teacher, surveyed UK medical school teaching 4,6,11. The common gap across this literature is that appraisal teaching concentrates on trials and reviews, so most clinicians are never taught to appraise the design they encounter most often.

The position of the case report in the translational pathway

A case report is the first link in a chain that runs, when it works, from observation to series to analytic study to synthesis to practice change. Being the first link is a position of consequence, because nothing downstream happens without it, and because a poorly documented first link cannot be built on.

How long the chain takes is less well established than the literature implies. Morris, Wooding and Grant, in the Journal of the Royal Society of Medicine, reviewed 23 papers that quantified translation time lags and found the frequently stated average of 17 years for research evidence to reach clinical practice rests on only a handful of estimates 9. The underlying studies define start and end points incompatibly — discovery, publication or first trial at one end; first clinical use, guideline inclusion or majority adoption at the other — so the figure should not be quoted as a measurement. The defensible version of the point is qualitative: the interval is long, variable and poorly characterised, and shortening it is the subject of its own literature. Our page on translating research into practice interventions covers that work.

Grol and Grimshaw, in the Lancet, describe the organisational side of moving from best evidence to best practice; Eddy, in Health Affairs, argued for a unified approach to evidence; Wennberg’s Science paper on small-area variation and Chassin, Kosecoff, Solomon and Brook’s JAMA study of angiography use document the variation that motivates the whole enterprise 15,24,26. The case report’s contribution to this chain is narrow and irreplaceable: it supplies the observations that everything else is built from. Our page on how EBP supports informed decisions covers how such evidence is weighted at the bedside.

Key points

Key points

The evidence hierarchy ranks designs by their ability to estimate an average treatment effect under confounding, so the position of the case report at the lowest level indicates what it cannot do rather than its value. Four questions have no better design: whether a phenomenon has ever occurred, whether a treatment could cause a particular harm, how a condition presents atypically, and why an event occurred in one thoroughly characterised patient. The limits should be stated with corresponding rigour: without a denominator there is no rate, without a comparator no treatment recommendation, and with one observation no causal claim. These limits should be stated in the discussion, because a report that states its limits is one that a later systematic search can use. Where the treatment is reversible, the condition stable and the outcome measurable over short periods, a single-patient trial is a stronger design for the same question and is distinct from a case report. The chronology should be reported with actual intervals, together with all concurrent exposures, the negative tests, and the alternative explanations with the reason each was rejected. The 17-year translation figure should not be quoted: Morris, Wooding and Grant found that it rests on a small number of estimates with incompatible definitions across the 23 papers they reviewed.

Application to your study

Application to your study

  1. Adopt a lower threshold for reporting a harm. An observation that would have altered a clinical decision had it been known is sufficient reason to report it. Unexpected findings are those the publication system under-represents.
  2. Determine whether an N-of-1 trial is feasible. Where the treatment is reversible, the condition stable and the outcome measurable over a short horizon, a single-patient trial answers the question and a case report does not.
  3. Report the chronology with intervals rather than with terms such as “subsequently”. Intervals should be stated in hours, days or weeks, since a later reviewer assessing temporality requires the numbers.
  4. List every concurrent exposure, including those considered irrelevant. The author’s judgement of relevance is not available to the reader, and an omitted drug is likely to be the one the reader asks about.
  5. Report the negative investigations. The conditions excluded, and the tests by which they were excluded, allow a reader to judge the diagnosis and distinguish a report from an assertion.
  6. State the limitations explicitly. State “This case cannot establish causation”, followed by what the case does support and the specific study that would test it.
  7. Treat consent as a substantive rather than a procedural matter. A case report describes an identifiable person; the journal’s requirements and the applicable data protection position should be checked before writing rather than at submission.

Frequently asked questions

Because the hierarchy ranks designs by one criterion — how well they estimate an average treatment effect in the presence of confounding — and several important clinical questions are not of that kind. For the existence question (has this ever happened), the harms signal question (could this treatment cause this), the phenotype question (what does this look like when atypical) and the mechanism question (why did this happen in this patient), no higher design is available, so the case report is not a weak option but the only one. Bellomo and Bagshaw argue in Critical Care that design-based classification omits dimensions that matter clinically, including mechanism plausibility and specificity of association.

A case report describes one patient retrospectively with no built-in comparison. A case series describes several patients sharing a feature, also retrospectively, and permits description of a range but still has no denominator and no comparator — with case selection as the dominant bias. A single-patient or N-of-1 trial is a different object entirely: prospective, using multiple crossover periods within one patient, and able to be randomised and blinded. Serpico and Maziarz, in the European Journal for Philosophy of Science, describe it as a method for investigating individual treatment response rather than an averaged effect. It supports a causal claim about that individual, which no case report can do.

Where the observation is a harm, yes — and that is its strongest claim. A single well-documented adverse event with a specified temporal relationship, a plausible mechanism, and ideally resolution on withdrawal and recurrence on re-exposure, generates a signal that nothing else in the system generates, since trials are powered for efficacy rather than for rare and late harms. For efficacy claims the answer is no: a case report has no comparator and cannot control for natural history, regression to the mean or concurrent changes in management. A discussion section that recommends a practice change on the strength of one case will be read as overclaiming, which discredits the observation itself.

More detailed than seems necessary, because the audience includes someone conducting a systematic search years later against inclusion criteria that have not yet been written. The elements that determine usability are the chronology with actual intervals rather than narrative connectives, all concurrent exposures including those the authors consider irrelevant, the diagnostic criteria applied with the negative investigations as well as the positive, the alternative explanations considered with the specific reason each was rejected, and the follow-up duration with its end state. Structured reporting guidance for case reports exists precisely to enforce this, and a report that omits these items is commonly excluded from later syntheses regardless of how interesting the phenomenon was.

Because publication bias acts on them with particular force. DeVito and Goldacre, in BMJ Evidence-Based Medicine, describe publication bias as a systematic skew toward positive and expected findings. An unexpected harm is by definition an unexpected finding, which has historically been the kind of report journals were least inclined to publish and authors least motivated to write. The result is that the observations on which the safety system most depends are the ones least likely to reach the literature. The practical remedy lies with individual authors: such a harm should be reported even when the observation appears insubstantial, and documented well enough to be assessable rather than merely persuasive.

Five questions should be asked. Is the chronology given with intervals or merely narrated? Are all concurrent exposures listed, or only the one the authors are interested in? Were alternative explanations considered and rejected with stated reasons? Is the follow-up long enough for the outcome to be settled? Does the discussion claim more than one observation can support? A report failing the first four is not usable as evidence whatever the phenomenon. A report that passes them and fails the fifth is usable — the authors’ conclusion should be discounted and the observation retained, and this should be stated when citing it, which is a precise and legitimate use of the literature.

The limit should be stated first, then what the case supports, then the study that would test it. A paragraph of the form “this case cannot establish causation; the following three features support the proposed mechanism; the following three alternatives were considered and here is why each was rejected or could not be; what this supports is the following specific hypothesis, testable against this data source” is both more useful and more citable than one asserting that the case demonstrates something. The phrase “further studies are warranted”, which names no study, should be avoided, as should priority claims such as “the first report of” unless the search that justifies them is described.

Treat it as substantive rather than procedural. A case report describes a specific identifiable individual, and identifiability is not removed by omitting a name — a rare condition, a stated age and an institution can be sufficient together. Journal requirements vary and data protection obligations apply independently of them, so the position should be checked before writing rather than at submission, when a missing consent can make an otherwise publishable report unpublishable. This is particularly acute for the reports that are most worth publishing, since rarity is both what makes a case informative and what makes the patient identifiable.

References

  1. Serpico D, Maziarz M. Averaged versus individualized: pragmatic N-of-1 design as a method to investigate individual treatment response. European Journal for Philosophy of Science. 2023;13(4):59. doi:10.1007/s13194-023-00559-0 read 9 October 2026. doi.org/10.1007/s13194-023-00559-0
  2. DeVito NJ, Goldacre B. Catalogue of bias: publication bias. BMJ Evidence-Based Medicine. 2019;24(2):53-54. doi:10.1136/bmjebm-2018-111107 read 9 October 2026. doi.org/10.1136/bmjebm-2018-111107
  3. Splinter K, Adams DR, Bacino CA, et al. Effect of genetic diagnosis on patients with previously undiagnosed disease. New England Journal of Medicine. 2018;379(22):2131-2139. doi:10.1056/NEJMoa1714458 read 9 October 2026. doi.org/10.1056/NEJMoa1714458
  4. Ilic D, Maloney S. Methods of teaching medical trainees evidence-based medicine: a systematic review. Medical Education. 2014;48(2):124-135. doi:10.1111/medu.12288 read 9 October 2026. doi.org/10.1111/medu.12288
  5. Shrager J, Tenenbaum JM. Rapid learning for precision oncology. Nature Reviews Clinical Oncology. 2014;11(2):109-118. doi:10.1038/nrclinonc.2013.244 read 9 October 2026. doi.org/10.1038/nrclinonc.2013.244
  6. Maggio LA, Tannery NH, Chen HC, ten Cate O, O'Brien B. Evidence-based medicine training in undergraduate medical education. Academic Medicine. 2013;88(7):1022-1028. doi:10.1097/ACM.0b013e3182951959 read 9 October 2026. doi.org/10.1097/ACM.0b013e3182951959
  7. Balshem H, Helfand M, Schunemann HJ, et al. GRADE guidelines: 3. Rating the quality of evidence. Journal of Clinical Epidemiology. 2011;64(4):401-406. doi:10.1016/j.jclinepi.2010.07.015 read 9 October 2026. doi.org/10.1016/j.jclinepi.2010.07.015
  8. Sur RL, Dahm P. History of evidence-based medicine. Indian Journal of Urology. 2011;27(4):487-489. doi:10.4103/0970-1591.91438 read 9 October 2026. doi.org/10.4103/0970-1591.91438
  9. Morris ZS, Wooding S, Grant J. The answer is 17 years, what is the question: understanding time lags in translational research. Journal of the Royal Society of Medicine. 2011;104(12):510-520. doi:10.1258/jrsm.2011.110180 read 9 October 2026. doi.org/10.1258/jrsm.2011.110180
  10. Ashley EA, Butte AJ, Wheeler MT, et al. Clinical assessment incorporating a personal genome. Lancet. 2010;375(9725):1525-1535. doi:10.1016/s0140-6736(10)60452-7 read 9 October 2026. doi.org/10.1016/s0140-6736(10)60452-7
  11. Meats E, Heneghan C, Crilly M, Glasziou P. Evidence-based medicine teaching in UK medical schools. Medical Teacher. 2009;31(4):332-337. doi:10.1080/01421590802572791 read 9 October 2026. doi.org/10.1080/01421590802572791
  12. Bellomo R, Bagshaw SM. Evidence-based medicine: classifying the evidence from clinical trials - the need to consider other dimensions. Critical Care. 2006;10(5):232. doi:10.1186/cc5045 read 9 October 2026. doi.org/10.1186/cc5045
  13. Eddy DM. Evidence-based medicine: a unified approach. Health Affairs. 2005;24(1):9-17. doi:10.1377/hlthaff.24.1.9 read 9 October 2026. doi.org/10.1377/hlthaff.24.1.9
  14. Dawes M, Summerskill W, Glasziou P, et al. Sicily statement on evidence-based practice. BMC Medical Education. 2005;5(1):1. doi:10.1186/1472-6920-5-1 read 9 October 2026. doi.org/10.1186/1472-6920-5-1
  15. Grol R, Grimshaw J. From best evidence to best practice: effective implementation of change in patients' care. Lancet. 2003;362:1225-1230. read 9 October 2026. epistemonikos.org
  16. Tonelli MR. The limits of evidence-based medicine. Respiratory Care. 2001;46(12):1435-1440. doi:10.1177/002013240104601207 read 9 October 2026. doi.org/10.1177/002013240104601207
  17. Tonelli MR. In defense of expert opinion. Academic Medicine. 1999;74(11):1187-1192. doi:10.1097/00001888-199911000-00010 read 9 October 2026. doi.org/10.1097/00001888-199911000-00010
  18. Sackett DL, Rosenberg WMC, Gray JAM, Haynes RB, Richardson WS. Evidence based medicine: what it is and what it isn't. BMJ. 1996;312(7023):71-72. doi:10.1136/bmj.312.7023.71 read 9 October 2026. doi.org/10.1136/bmj.312.7023.71
  19. Rosenberg W, Donald A. Evidence based medicine: an approach to clinical problem-solving. BMJ. 1995;310(6987):1122-1126. doi:10.1136/bmj.310.6987.1122 read 9 October 2026. doi.org/10.1136/bmj.310.6987.1122
  20. Sackett DL, Rosenberg WMC. The need for evidence-based medicine. Journal of the Royal Society of Medicine. 1995;88(11):620-624. doi:10.1177/014107689508801105 read 9 October 2026. doi.org/10.1177/014107689508801105
  21. Richardson WS, Wilson MC, Nishikawa J, Hayward RS. The well-built clinical question: a key to evidence-based decisions. ACP Journal Club. 1995;123(3):A12-A13. doi:10.7326/ACPJC-1995-123-3-A12 read 9 October 2026. doi.org/10.7326/ACPJC-1995-123-3-A12
  22. Evidence-Based Medicine Working Group. Evidence-based medicine. A new approach to teaching the practice of medicine. JAMA. 1992;268(17):2420-2425. doi:10.1001/JAMA.1992.03490170092032 read 9 October 2026. doi.org/10.1001/JAMA.1992.03490170092032
  23. Smith R. Where is the wisdom? BMJ. 1991;303(6806):798-799. doi:10.1136/bmj.303.6806.798 read 9 October 2026. doi.org/10.1136/bmj.303.6806.798
  24. Chassin MR, Kosecoff J, Solomon DH, Brook RH. How coronary angiography is used. Clinical determinants of appropriateness. JAMA. 1987;258(18):2543-2547. doi:10.1001/jama.258.18.2543 read 9 October 2026. doi.org/10.1001/jama.258.18.2543
  25. Eddy DM. Variations in physician practice: the role of uncertainty. Health Affairs. 1984;3(2):74-89. doi:10.1377/hlthaff.3.2.74 read 9 October 2026. doi.org/10.1377/hlthaff.3.2.74
  26. Wennberg J. Small area variations in health care delivery. Science. 1973;182(4117):1102-1108. doi:10.1126/science.182.4117.1102 read 9 October 2026. doi.org/10.1126/science.182.4117.1102

Every source above is a peer-reviewed article, a publisher’s record of one, or a standard issued by the body that owns it. Each carries the date it was last read. Where a standard moves, this article is revised rather than left up with a note.

How to cite this article

Pick your referencing style and copy it. Your university or target journal may vary the punctuation — check its guide if it has one.

Pubrica Academy. (2026, October 9). The Importance of Case Report Writing in Evidence-Based Practice. Pubrica. https://pubrica.com/academy/medical-writing/the-importance-of-case-report-writing-in-evidence-based-practice/

This article is written to be learned from. Cite it if you are referring to what it says; if you are using the evidence behind it, cite the sources listed above instead.

Case report completeness and claim-limits worksheet

This worksheet should be completed before drafting. The chronology and alternative-explanation sections determine whether the report can be used by a later systematic search; the claims section determines whether a reader trusts the rest.

CASE REPORT WORKSHEET

0. WHY THIS CASE IS WORTH REPORTING  (tick at least one)
   [ ] existence - this has not been described before
   [ ] harm signal - possible treatment-related adverse event
   [ ] atypical phenotype or course
   [ ] mechanism - deep individual characterisation available
   [ ] diagnostic pitfall others would repeat
   [ ] teaching value (state what is generalisable): __________
   If none of these apply, the case may be interesting without
   being informative. Interest is not a publication criterion.

0b. IS AN N-OF-1 TRIAL AVAILABLE INSTEAD?
   [ ] treatment is rapidly reversible
   [ ] condition is stable
   [ ] outcome is measurable over short periods
   All three ticked -> an N-of-1 design answers the question
   better. None or some -> proceed, and say in the methods
   why a prospective design was not possible.

1. CHRONOLOGY  (intervals, not "subsequently")
   Presentation ...................: day ____
   First dose / exposure ..........: day ____
   Onset of the finding ...........: day ____  interval: ____
   Withdrawal / de-challenge ......: day ____
   Resolution .....................: day ____  interval: ____
   Re-exposure, if any ............: day ____
   Recurrence .....................: day ____  interval: ____
   Follow-up ended ................: day ____  end state: ____

2. CONCURRENT EXPOSURES  (list ALL, including irrelevant ones)
   ______________________  started ____  changed? ____
   ______________________  started ____  changed? ____
   ______________________  started ____  changed? ____
   [ ] Any exposure that did NOT change across resolution and
       recurrence is noted as such

3. INVESTIGATIONS
   Positive findings ..............: _________________________
   NEGATIVE findings (report these): _________________________
   Diagnostic criteria applied ....: _________________________
   Imaging / pathology / genomics .: _________________________

4. ALTERNATIVE EXPLANATIONS  (one line each, with the reason)
   Alternative 1 ..: ____________  rejected because: ________
   Alternative 2 ..: ____________  rejected because: ________
   Alternative 3 ..: ____________  CANNOT be excluded: ______
   [ ] At least one alternative is left standing if it should be

5. WHAT THIS CASE SUPPORTS  (write the sentences)
   "This case cannot establish ______________________________"
   "The features supporting ______ are: _____________________"
   "The specific testable hypothesis is: ____________________"
   "It could be tested using: _______________________________"
   [ ] No practice recommendation is made from this single case
   [ ] No priority claim made, OR the search supporting it is
       described: ___________________________________________

6. CONSENT AND IDENTIFIABILITY
   Written consent obtained ......: [ ] yes  date: ______
   Could the patient be identified from the combination of
     condition rarity, age, dates and institution?  [ ] yes
     If yes, what has been removed or generalised: __________
   Journal's stated requirement checked: [ ] yes
   Applicable data protection position: ______________________

7. REPORTING GUIDELINE
   Guideline used ................: ______________________
   Completed checklist location ..: ______________________

PHRASES TO DELETE IF PRESENT
  [ ] "this case demonstrates that X causes Y"
  [ ] "to our knowledge the first report" with no search described
  [ ] "clinicians should consider" as a recommendation from one case
  [ ] "further studies are warranted"
  [ ] "research takes 17 years to reach practice"

Pubrica AcademyWritten and maintained by Pubrica’s subject matter experts. Articles are revised when the underlying standards change, and each article states the date on which it was last reviewed.

How Pubrica can help

Pubrica prepares case reports against the structured reporting items that journals apply, which in practice means supplying the chronology intervals, negative investigations and rejected alternatives that first drafts frequently omit. Pubrica also revises the discussion so that it claims only what a single case supports, and checks the consent and identifiability position before the manuscript is prepared, because this is the issue that most often renders a worthwhile case report unpublishable.

Find something else

Search the Pubrica Academy

Access is free and requires no account. Enter a topic, or open the full library to filter by category, sub-topic or resource type.

Pubrica can assess a clinical case for reportability, identify omissions in the record, and advise on the claims the discussion can legitimately make.

A scoping assessment is free and takes two business days. It provides a written assessment of what the work requires and which specialist would undertake it.

Get a free quote

Regulations

Check your project against this rule

  1. 1Your work
  2. 2Your details
  3. 3Send
What do you need?

Material you send is held under signed confidentiality, with access limited to the assigned team and encrypted transfer. It is not used to train anything and is not passed to third-party tools. Ask for the agreement before you send anything you would rather cover first.