More importantly, we've made these changes with a serious professional purpose in mind: to render the content as accessible and usable as possible. The meta-message to our readers is that Annals is less of an archive and more of a living document, doing everything it can to help readers take the information off its pages and out into practice. Vladimir Nabokov's comment that "In art, as in science, there's no delight without the detail" tells us much about what's required here, namely paying meticulous attention to the details. We have tried to do just that; some of the most important details follow.

-Type size is increased throughout. The intent here is to make all of the text more readable, but the change will be most noticeable in the Methods sections which, like those in many journals, have in the past literally been "the fine print." Not surprisingly, the implication has been that Methods sections are intended for researchers and are something most readers can skip. But verily it is said, "What the Results section giveth, the Methods section taketh away." We therefore say to our readers that the Methods section is a critical and integral part of a paper and, as such, is meant to be read, understood, and used by all. It may be harder to learn how to read the Methods than the Results or the Conclusions, but it is the Methods section that sets up the entire study, gives it strength, and, most importantly, determines its limitations. Readers ignore the Methods section at their peril.

-The type face of Tables and References is new. We're now using type that lacks serifs, the little hooks and curlicues that make larger type attractive but also create visual clutter that can interfere with reading small type in Tables and References.

-Headings, dividers, and breaks in layout have all been reworked, to make the organization of content more easily apparent.

Inevitably, changes in design will change what authors need to do in preparing their manuscripts. Details are in the full Information for Authors on pp. I-11 through I-15, but among the most important of these are the following.

-We have reduced the upper limit on manuscript length slightly for most categories. Annals articles have tended to be long, but space on the printed page is an increasingly precious commodity, given the economics of publishing. We will not skimp on publication of critical information; at the same time we can continue to publish the maximum number of important submissions only if authors tighten their prose and their logic even more carefully than they've done in the past.

-Systematic reviews that meet appropriate standards may be published in the Articles rather than the Reviews section. Systematic reviews are studies that synthesize high-quality evidence on clinical problems, based on exhaustive search of published and unpublished data sources, and using rigorous statistical methods [1, 2]. The more traditional reviews, being so-called "secondary" publications, have generally occupied a lower position on the intellectual and academic food chain. But why should clinical researchers' data sets be any less important as material for the development of meaningful evidence than, say, blood pressures or hormone levels taken directly from patients? We think the worth of a study should be determined by the nature and quality of the intellectual work and the value added to our knowledge by that work, rather than by the source and nature of the data analyzed. In our view, therefore, most systematic reviews will belong in the category of Articles along with other original papers; we will continue to publish the more traditional analytic reviews in the Reviews section.

-The earlier Diagnosis and Treatment category has been replaced by a new category of Updates, and the much-read, much-appreciated On Being a Doctor is joined by a new companion feature, On Being a Patient.

-Finally, we reaffirm the need for full disclosure of potential conflict of interest, and of the role of funding agencies in the conduct of research. All research is fraught with the potential for bias: bias from chance, bias from inappropriate experimental design, statistical bias, psychological bias, and others. In a sense, minimizing bias is largely what good research is all about, and with industry playing an increasing and increasingly important part in the support of clinical research, the need to recognize and manage potential bias from financial conflict of interest and from delegation of the control of research to funding sources grows in proportion. At the same time, we see bias from these sources as being essentially no more and no less important than inadequate randomization, or inappropriate exclusions, or the many others in the larger set of well-recognized biases that haunt all research. As such, does information about potential sources of financial and commercial bias, when it is published, therefore belong in the Methods section, rather than tucked away in foot-notes at the beginning or end of articles, as has been the more traditional practice?

As always, we invite comments and opinions from readers on this last question, and on any other aspect of Annals design or content.