Cynthia Mulrow and her coauthors are to be congratulated for their attempt to translate the large body of evidence about elevated blood pressure into information that “meet[s] the needs of hurried and harried clinicians.” Intended to be carried and used in the clinical setting, the book is organized by means of clinically relevant questions, patient scenarios, the presentation of graded evidence and summary “bottom lines.” The authors describe their task in writing this book as Herculean. That is certainly true for the task of converting the evidence into bottom lines and perhaps explains why the different chapters succeed only to varying degrees.

The presentation of evidence in the last chapter, “What if my hypertensive patient becomes pregnant?”, is by far the most successful. Here, the evidence from available randomized controlled trials is presented in three figures as a meta-analysis. The reader can see at a glance that a large benefit or harm from the treatment of hypertension has not been established for either pregnant women or their infants. This might come as a surprise for the reader, as it did for me, given that most guidelines recommend treatment for women with relatively mild elevation of blood pressure during pregnancy. I also appreciated the way the evidence in this chapter was summed up. For example, for blood pressures of > 140/90 mm Hg and < 160/100 mm Hg in pregnant women, “there is little known benefit of antihypertensive treatment,” and such patients should be informed of this fact. The decision as to what to do in an individual case is left up to the patient and the physician.

In contrast, none of the other chapters uses figures to present meta- analysis evidence. Rather, evidence is presented variously as risk ratios, absolute risk reduction or in verbal summaries. The approach used in converting this evidence into bottom-line messages also varies. Some of the authors make recommendations in the absence of trial evidence. For example, in Chapter 4, the authors recommend “slow gradual weight loss in patients who are overweight,” despite the fact that this is based on consensus opinion, not evidence. As a result, the reader cannot accept the bottom-line statements in the various chapters as consistently derived from clinical trials.

The grading of evidence was not helpful, as it frequently could not be verified from the text and supporting references. For example, in Chapter 7, the recommendation of target blood pressures of < 75–80 mm Hg for diabetic patients with hypertension is graded as being based on Level 1 evidence. In fact, no trial has tested the effect of lowering blood pressure to < 75 mm Hg in any population, and the only evidence for < 80 mm Hg was derived, on the basis of a very small number of events, from a subgroup analysis of diabetic patients in the HOT study. This hardly represents Level 1 evidence, in my view.

Also disturbing are various errors in the text, as for example in the Framingham Study table for women (Appendix 1, Chapter 3). The tables for women are mistakenly the same as for men and, if used to calculate Ms. Athletic's risk on page 56, yield an incorrect number and estimation of risk. As another example of inaccuracy, on page 87 nisoldipine rather than nitrendipine is incorrectly named as the drug used in the Syst-Eur trial.

On the positive side, this book is an excellent and comprehensive source of references for the major randomized controlled trials and systematic reviews published on elevated blood pressure. Anyone working in this area can learn a great deal from it. However, although this text provides ample evidence that elevated blood pressure simply represents a graded risk factor for adverse cardiovascular events and thus is not a disease entity, I was disappointed that some chapters still perpetuate the myth that hypertension is a definable disease based on a measurement (pages 19 and 39).

The main challenge in disseminating evidence-based information lies in finding a consistent method of presenting the evidence and converting that evidence into appropriate clinical messages. This text, while failing to establish a consistent method, demonstrates the difficulty of that task, and as such makes a valuable contribution.