Comparisons of referral criteria for public screening of blood cholesterol levels.

Total and high-density lipoprotein (HDL) cholesterol levels of 2,387 adults were screened at a worksite and a bloodbank. Hypothetical referral decisions were made according to three sets of guidelines: the 1984 National Institutes of Health Consensus Conference guidelines (NIHCC), a single referral cutpoint of 5.2 millimoles per liter (mmol per L), and the current National Cholesterol Education Program (NCEP) guidelines for screening in physicians' office. Under the NIHCC guidelines, 31 percent of the participants would have been referred to their physicians, 32 percent under the NCEP guidelines, and 56 percent would have been referred had the 5.2 mmol per L cutpoint been used. Twenty-four percent of the participants would have been referred under both the NIHCC and NCEP guidelines; 7 percent would have been referred under the NIHCC guidelines, but not the NCEP's. Eight percent would have been referred under the NCEP guidelines, but not the NIHCC's. Those who would have been referred were older, and more likely to be male and to have low levels of HDL cholesterol than the 7 percent who would have been referred under NIHCC guidelines only. All of the 8 percent had coronary heart disease, or two or more other coronary risk factors, whereas none of the 7 percent did. If low HDL had been used as a risk factor under NCEP guidelines, the number of persons referred would have increased slightly (to 34 percent) and low HDL levels would have become one of the most prevalent risk factors. The researchers concluded that public cholesterol screening programs should use the NCEP guidelines (with or without HDL), rather than the NIHCC guidelines, or a single 5.2 mmol per L cutpoint.