As Dr. Fronduti surmises, the statement that screening is not recommended for persons 75 years of age and older applies only to primary prevention. The recommendation to do lipid analysis in all persons with known coronary heart disease or other vascular disease applies to men and women of all ages.

In many practice settings, as Dr. Goldstein observes, the key issue is what to do with knowledge of a cholesterol level that may have been obtained incidentally as part of a blood chemistry panel. Recognizing that incidental abnormalities discovered in this way may be falsely positive or otherwise unhelpful, many institutions (including our own) no longer report the results of dozens of tests when the physician had only intended to order one or a few. Until this practice is more widespread, we suggest interpreting the cholesterol level by using the information given in the background paper [1], especially Tables 3 and 4, to quantify the potential benefits (in terms of mortality from coronary heart disease) of treatment for men and women of different ages and underlying risks. This information can be combined with the general treatment strategy proposed by the National Cholesterol Education Program (NCEP), which sets cut-points for further diagnostic evaluation and therapy on the basis of risk for coronary disease.

Drs. Altus, Grundy, and Pearson object to the American College of Physicians guidelines because these guidelines differ from the recommendations of NCEP. The background paper [1] addresses their objections, but several points deserve clarification. First, Grundy and Pearson misleadingly state that "numerous other professional and national bodies recommend cholesterol screening for all adults 20 years of age and older." Yet (see Table 2 of the College's guidelines), only one other published set of major North American guidelines concurs with NCEP in recommending universal cholesterol screening for adults. The U.S. Preventive Services Task Force; the Canadian Task Force on the Periodic Health Examination; the provinces of Ontario [2], Quebec [3], and Saskatchewan [4]; and many provider groups (such as Kaiser-Permanente) have issued screening recommendations similar to those of the College.

In stating that the College recommends "limiting screening for lipid abnormalities to men aged 35 to 65 years and women aged 45 to 65 years," Grundy and Pearson misrepresent the content of the guidelines, overlook the rationale behind them, and exaggerate the differences between the College's guidelines and those of NCEP. Although the College's guidelines recommend universal screening only in this age group, they also recommend screening in individual persons of other ages who are likely to benefit from the reduction of an elevated cholesterol level. Altus and Pearson and Grundy cite two major published trials of statins, the Scandinavian Simvastatin Survival Study (4S) and the West of Scotland Study, as evidence that cholesterol reduction has been proven beneficial in persons who would be excluded from screening according to the College's guidelines. Yet all of the persons included in these trials would be screened under these guidelines. It is unlikely that the low-risk populations excluded from these trials would enjoy benefits from treatment similar to those seen in the included participants. For example, the 5-year coronary heart disease mortality rate was about 100 times as great in the control group of the West of Scotland study and more than 400 times as great in 4S as it is for young women with high blood cholesterol levels. Adverse effects that were delayed or that occurred too infrequently to be detected in these trials could easily overshadow any benefits in young persons.

The College did not make a recommendation for or against screening in asymptomatic adults between the ages of 65 and 75 years because, as was documented in the background paper, there was too little clinical evidence on which to base a recommendation. Grundy and Pearson claim that 4S and the West of Scotland Study included "many persons older than 65 years of age, showing the efficacy of cholesterol reduction in decreasing mortality and improving quality of life in persons for whom the College now recommends no cholesterol testing." Because the College recommends measurement of cholesterol levels in all adults with symptomatic coronary disease, this would include all of the participants of 4S, which was a secondary prevention study. The College's guidelines differ from those of NCEP only with regard to primary prevention, but neither the West of Scotland study, with a maximum age at enrollment of 64 years, nor any other published primary prevention study has addressed the question of whether cholesterol reduction is beneficial after the age of 65 years. Furthermore, both the West of Scotland study and 4S (which enrolled participants as old as 70 years of age) reported that the relative reduction in the risk for death resulting from treatment decreased with age. Thus, as the discussion of these trials in the background paper indicated, their results support the College's recommendations.

Dr. Kryshak suggests that a patient's knowledge of his or her own cholesterol level can help to promote healthy behaviors. However, studies cited in the background paper [1] showed that any such effects are small at best. A more recently published randomized trial [5] confirmed the previous results, leading its authors to conclude that "cholesterol measurement should not be used as a health promotion tool to motivate change in behavior."

Brief guidelines cannot give detailed recommendations for every clinical situation that might confront the physician. The College's guidelines emphasize the principle that screening should be targeted toward persons who, because they are at elevated short-term risk for coronary heart disease, are likely to benefit from treatment if they are found to have an elevated cholesterol level. In listing some of the best-accepted characteristics that place persons at elevated short-term risk, the language of the guidelines-that risk factors included age, high blood pressure, diabetes mellitus, and so on-was meant to suggest that other risk factors should also be considered. Dr. Thacker is correct in noting that premature menopause is a risk factor for coronary heart disease, and we believe that physicians should take this information into account in deciding whom to screen for high cholesterol levels. Yet Dr. Thacker's example also illustrates the need to place cholesterol into the overall perspective of each patient's well-being. The first question most physicians would ask about the young woman with premature menopause is not "What is the cholesterol level?" but "Is there a contraindication to hormone replacement therapy?" Although no randomized trials have evaluated the effectiveness of either cholesterol reduction or hormonal therapy in such a situation, observational studies suggest that hormonal therapy not only improves symptoms but may also reduce all-cause mortality by as much as 50%. As the guidelines state, screening and treatment recommendations are likely to change as further data become available about treatments specifically for women, such as postmenopausal estrogen replacement therapy.

Any set of guidelines written for physicians should reflect trust in physicians' judgment and should be sufficiently specific and evidence-based to be useful. That is why the College's guidelines were written as an aid to physicians, not as a set of edicts. It is also why, in contrast to the NCEP, the College did not recommend mandatory cholesterol screening. Physicians should exercise discretion in ordering cholesterol screening tests, modifying their practices according to patient preferences and other individual factors. At the same time, when they follow the College's recommendations, physicians can feel confident that they will be measuring blood cholesterol levels in all adults for whom existing evidence indicates that the benefits of treatment exceed its risks.