Editorial
Improving the Reporting of Clinical Case Series

https://doi.org/10.1016/j.ajo.2004.12.009Get rights and content

Purpose

To describe common errors in the analyses and data presentation of a clinical case series and to suggest simple solutions.

Design

Instructional examples.

Methods

Problems with commonly used data analysis and reporting techniques in clinical case series are described using both theoretical examples and those from the literature.

Results

An analysis reporting the proportion of a series of patients with variable follow-up does not adequately account for the differential follow-up among patients and is a potentially misleading way to present data. Instead, the proportion of patients at presentation (or study entry) and the rate during follow-up should be reported. Similarly, an analysis in which the final visual acuity of a series of patients with variable follow-up is reported does not adequately account for the effect of time and also may be misleading. Reporting of the rates of visual acuity events during follow-up (e.g., falling below a specified threshold, such as 20/50 or worse) is preferred. Alternatively, when there is nearly complete follow-up, reporting the distribution of visual acuity at specified time points (e.g., 1 year after study presentation) is appropriate. Small case series should not be overinterpreted because of the effects of chance, and appropriate statistical analyses should be employed.

Conclusions

Clinical case series often suffer from several potential reporting flaws. Correction of these flaws would permit the proper interpretation of the data and allow for the ability to combine data from several case series to assemble more meaningful and reliable conclusions.

Section snippets

Proportions of a population with variable follow-up

One of the more common statements made in a clinical case series is one such as “In 28% of patients with some disease (e.g., uveitis) followed for a mean of some time (e.g., 9 months, range 3 months-5 years), some complication (e.g., elevated intraocular pressure) occurred.” Although at face value, the statement seems reasonable, it actually conveys little or no information and may convey misleading information. As shown in Figure 1, two very different event curves both are consistent with this

Small studies

A third problem present in clinical case series is the overinterpretation of small studies. This problem is not restricted to uncontrolled case series but also may affect underpowered clinical trials. One example is a small clinical trial that suggested a new treatment was ineffective for preventing relapse of uveitis because there was no significant difference in the proportion relapsing between the treatment group (30%) and the placebo group (50%, P = .66).18 The problem with this study was

Conclusion

Uncommon diseases may not be easily amenable to investigation in adequately powered randomized controlled clinical trials. The disease may not be judged to be important enough for the large amount of funding required for a large multicenter clinical trial, and the sample size may not be easily achievable. No one center or small group of collaborating centers may have enough patients to conduct a large epidemiologic study. A new treatment may be being used and pilot data needed. Thus, case

References (19)

There are more references available in the full text version of this article.

Cited by (136)

View all citing articles on Scopus

Supported in part by grant EY 00405 from the National Eye Institute, National Institutes of Health, Bethesda, MD.

The Editors-in-Chief of the American Journal of Ophthalmology, Archives of Ophthalmology, and Ophthalmology have reviewed the manuscript, and with the approval of the author, have included this manuscript in the Editorship series of these three journals.

View full text