In an article¹ in this issue of American Family Physician, Dr. Zamorski and I discuss the conclusions and recommendations of the most recent National Heart, Lung, and Blood Institute (NHLBI) Working Group Report on High Blood Pressure in Pregnancy.² That report reached conclusions that differ in some important ways from previous NHLBI reports and from other groups' recommendations. The substance of the differences and the process that led to them are presented here.

The 2000 NHLBI report recommendations are the only recommendations for hypertensive disorders of pregnancy that explicitly adopt an evidence-based rather than expert-consensus methodology. The evidence-grading scheme used was the same as that used in the Sixth Joint National Committee Report on Hypertension, which was also produced by the NHLBI.³ Adherence to an evidence-based process led to the NHLBI panel's decision not to use the criterion of a 30 mm rise in systolic blood pressure or a 15 mm rise in diastolic blood pressure as an indicator of preeclampsia. Data from a cohort study⁴ of 1,496 primigravid women showed no complications in women whose systolic blood pressure increased by 30 mm or more or whose diastolic blood pressure increased by 15 mm or more when those pressures did not exceed 140 mm systolic or 90 mm diastolic.

The introduction of an evidence-based process is a significant step forward in clinical guidelines for preeclampsia. For the busy physician who may not have the time to critically review large numbers of articles, a sound, evidence-based guideline process is the best indicator of reliable, unbiased, useable clinical information. The working group strove to make recommendations to guide practice based on POEMS (patient-oriented evidence that matters) rather than on subjective expert opinion or physiologic studies. Unfortunately, with a small number of exceptions such as cited above, POEMs were very difficult to come by.

The working group would have preferred to make yea or nay statements based on good outcomes evidence about many clinical questions, not only the degree of rise in blood pressure. The questions being debated were those that family physicians address daily in their practices: what are the positive and negative predictive values of the findings looked for in the assessment of patients suspected of having preeclampsia? Which symptoms, physical findings and laboratory values are truly good predictors, which are weak, and which are lacking in predictive value? In terms of outcome, which patients benefit, and how much, from magnesium sulfate therapy? Unfortunately, the POEMs most needed by family physicians were the most difficult to obtain or were simply unavailable.

Despite the lack of basic clinical POEMs, this working group report is a valuable undertaking for family medicine for two reasons. First, it highlighted some POEMs physicians can depend on to change their practices as detailed in the article. Second, the effort to produce evidence-based guidelines inevitably highlights the missing POEMs and helps focus research agendas on the knowledge needed by practicing physicians and their patients. Perhaps by the time the next working group convenes, we will have sound answers to the questions that physicians regularly face in their practice.