The edict has been handed down: All Americans 20 years and older need to stick out their arm, give up a vial of blood and have it tested for cholesterol. And, if it's too high, lower it.

The reason is quite convincing. For every 1 percent decrease in blood cholesterol, said a large federal study, there is a 2 percent decrease in the chance of coronary artery disease -- the leading cause of death in the United States.

Based on that data, the National Heart Lung and Blood Institute (NHLBI) has announced strict guidelines for all Americans: total blood cholesterol should be below 200 -- that's 200 milligrams of total cholesterol per deciliter of blood (mg/dl). If it's above that threshold, physicians must put their patients on cholesterol-lowering diets or use some of the new cholesterol-combating drugs to bring down the blood levels.

The facts of cholesterol haven't changed in the last several years. A high-fat diet remains the leading culprit. With the American diet, it is still more difficult to control blood cholesterol with advancing age. Blood cholesterol is still a greater problem for men than for women.

What has changed is the strategy to fight it. Doctors hope that the single target number -- 200 -- and a nearly universal two-step diet program will clarify what until now has been a muddle of sometimes conflicting advice. Questions of Accuracy

There are, however, some problems in carrying out the new public health guidelines. One third to half of the 6,000 blood-testing laboratories in the United States that performed about 100 million blood cholesterol tests last year are failing to make accurate measurements when compared to a national standard.

During a recent routine national survey of lab accuracy by the College of American Pathologists, more than 2,000 labs were unable to calculate accurately -- within plus or minus 5 percent -- the amount of cholesterol in a test sample. One sample that contained a cholesterol concentration of 262.2 mg/dl was measured by various labs from 101 to 524.

"The current state of reliability of blood cholesterol measurements made in the United States suggests that considerable inaccuracy in cholesterol testing exists, a situation that may handicap the national program to control heart disease," the Laboratory Standardization Panel of NHLBI's National Cholesterol Education Program reported Sept. 29.

While not considered a crisis, the inaccuracy of cholesterol testing makes it difficult to evaluate results in light of the new guidelines. "If the test is wrong, it will make a difference if you are going to decide to be treated based on laboratory results," said Dr. Gary Myers, chief of CDC's clinical chemistry standardization activity. Tests showing someone to be in a moderate risk group, for example, could turn out to be wrong, and the person may actually be in the high-risk or no-risk group.

For the most part, the problem with cholesterol testing is not that the laboratories make errors, but that there has not been any pressure to use standards to calibrate testing equipment. The country's 6,000 laboratories use machines made by more than a dozen different manufacturers employing nearly 35 different chemical and enzymatic methods for calculating blood cholesterol levels. Once calibrated, the machine should be tested by measuring the amount of cholesterol in a "standard" -- a sample containing a known amount of cholesterol.

Efforts are now under way to eliminate variations by making all the laboratories set their machines the same way.

Last month, the College of American Pathologists released a set of certified reference materials -- three frozen serum samples and three freeze-dried samples -- developed by the National Bureau of Standards to calibrate cholesterol tests. The federal Centers for Disease Control also has set up eight qualified regional reference labs around the country to help individual commercial labs.

Shifting to a national testing standard is much like setting all the clocks in the country to a standard time. Until the end of the 19th century, most of the world ran on local time. When the clock in the Washington Post Office tower struck noon, the clock in the Bromo Seltzer tower in Baltimore might have shown quarter past.

The pressure to standardize time across the nation came from the expansion of the railroads. In the 1870s, a passenger on the rails from Washington to San Francisco would have to reset his watch at more than 200 stops to account for different local times.

For the country's clinical laboratories, local standards for cholesterol testing, like local time before the railroad, were good enough. Variations between laboratories could be tolerated.

The national cholesterol guidelines are starting to force standardization of the 6,000 labs across the country, just as railroads forced uniformity on local time zones.

"When you start using national values, every laboratory has to be hitting the target all the time," said Dr. Basil Rifkind, chief of NHLBI's lipid metabolism branch. "Labs have to shift from what was previously an acceptable way of doing things to the current standard."

"We are optimistic that within a year or two, the quality of cholesterol measurement should remarkably improve," Rifkind said.

A similar problem occurred when the government started a blood pressure control program, said Michael White, associate director of the NHBLI's prevention, education and control programs. "In the early '70s when we began, the devices were not as accurate as now. The ability of the physician community to read blood pressure and know when to intervene was no better than cholesterol is now."

Until the problem gets sorted out, said Dr. William Weidman of the Mayo Clinic Medical School, "the patient has a right to ask his doctor, 'Does the lab that you use have external quality control?' If the doctor says, 'I don't know,' then say, 'Doctor, pick another lab.' "

The increase in pressure from patients and physicians means the accuracy issue "will reach a resolution much sooner than would otherwise be the case," NHBLI's White said.

The issue gets all the more complicated as doctors move away from using clinical laboratories and instead adopt rapid screening tests that use only a drop of blood. The machines, which have been around for about two years, are designed to be quick (usually taking less than 10 minutes to get a result), painless (only requiring a a finger puncture) and inexpensive (permitting use in doctors' offices and shopping malls).

The early machines had calibration problems, Rifkind said. "That problem is being sorted out. In competent hands, these machines appear to be pretty accurate," although some machines reportedly give readings that are consistently off by up to 10 percent.

Finding competent hands, however, can be a problem. The machines are often operated not by highly trained medical technicians, but by secretaries and even volunteers during large-scale screening programs.

And keeping portable machines calibrated isn't always easy. They get bounced around in the backs of station wagons, and users sometimes purchase chemical supplies from cheaper sources, which may skew the results.

NHLBI has a trial now under way to study the use and accuracy of these machines. "Our impression generally is that if they are used carefully, they do work well," Rifkind said. "If they are not used carefully, they do not." Getting Tested

According to the National Cholesterol Education Program, all Americans older than 20 should have their total blood cholesterol measured every five years. If the test finds a level below 200, then the individual need do nothing.

If the cholesterol level is between 200 and 239, the individual is considered to be "borderline" and at moderate to high risk.

Those above 240 have "high blood cholesterol" and are considered to be in the high-risk category. "The cutpoint that defines high blood cholesterol is a value above which risk of coronary heart disease rises steeply," the report said.

Once it has been determined that a person is in a moderate- or high-risk group, another blood test should usually be done to analyze the concentrations of the different kinds of cholesterol. Low-density lipoproteins (LDL), one form in which cholesterol is carried in the blood, have been associated with an increased risk of atherosclerosis, the buildup of fatty deposits in blood vessels. High-density lipoproteins (HDL), a second major form, have been associated with a lower risk of atherosclerosis.

To many heart specialists, the ratio of HDLs to LDLs -- of "good" cholesterol vs. "bad" cholesterol -- is important in evaluating a person's risk of heart disease. The NHLBI recommends, for example, that LDLs be lower than 130 mg/dl. LDLs higher than 160 are considered to put a person at high risk. The panel also recommends that HDLs be above 35.

Those in a high-risk group should be treated either with diet or drugs to reduce the cholesterol levels, and the tests should be repeated yearly.

Before therapies begin, the blood tests should be repeated. "Nobody should be regarded as falling into a definite category on the basis of one single cholesterol test," Rifkind said. The panel even recommends, but does not require, repeating the test for those first found to be low-risk.

Given the questions about the accuracy of the tests, should everyone go out and get their cholesterol measured?

Yes, said NHLBI's Michael White. "If we wait until everything is perfect, we are going to lose a lot of people who will die from unprevented heart disease. I don't think we can wait." Getting Treated

For more than half of adult Americans, the test will show that the blood cholesterol levels they have lived with for years are now considered dangerous. What do you do?

Just a few weeks ago, the answers would have been confusing. For a blood cholesterol of 240, half of American physicians, according to an government survey, used to recommend a wait-and-see approach. "I'd monitor patients and wait until blood cholesterol reached 300 before treating them," said Dr. Harold Sadin, a Washington internist.

But at the same level, half of doctors, according to the same survey, suggested making dietary changes -- losing weight and cutting down on saturated fat and cholesterol. For those whose blood cholesterol levels continued to creep to 260, one quarter of doctors prescribed cholesterol-lowering drugs.

Under the new guidelines, doctors are given clear advice on what to recommend. The possibility that someone will need treatment -- first diet, then drugs -- starts with a total cholesterol count above 200.

If there is any evidence of heart disease -- for example, the chronic chest pain known as angina, a previous heart attack or stroke, or previous coronary bypass surgery -- you are considered at high risk even though your cholesterol levels may be in the borderline category of 200 to 239.

Just being male counts as one risk factor. Being a cigarette smoker counts as another, as does having high blood pressure, diabetes or being obese (more than 30 percent over ideal body weight). Having a parent or sibling who either suffered a heart attack or died suddenly of heart disease before age 55 also increases the risk.

People with a low level of HDL or a high level of LDL, also are in a high-risk category. People with LDL levels less than 130 will receive information about reducing their risk of heart disease -- changing their eating habits, giving up smoking, losing weight -- but won't need to have their blood cholesterol level rechecked for about five years. Those with high LDL levels -- 160 and above -- will undergo a physical examination and additional laboratory tests to determine the possible causes of their high cholesterol levels. They should be treated as those in the high-risk group and be retested in four to six weeks and again at three months. People with LDL levels of 130 to 159 are to be treated in one of two ways. If there is no evidence of heart disease and if they have less than two risk factors, they do not require treatment and should be retested one year later. If there are symptoms of heart disease or two or more risk factors, then these individuals will be treated exactly the same as those in the high risk group.

Making nutritional changes is the first -- and most important -- line of attack in combating high cholesterol. Virtually all people in the high- and moderate-risk categories will be put on a special low-fat, low-cholesterol diet. Drugs, according to the guidelines, should only be used in very severe cases or after diet therapy has failed.

"We want to be extremely prudent in using drugs," said Dr. Scott Grundy, chief of the Center for Human Nutrition at the University of Texas Health Science Center in Dallas.

The Step 1 diet, according to the new guidelines, calls for reducing fat from the American average 40 percent of total calories to less than 30 percent. In addition, the diet limits saturated fat to 10 percent or less of calories, and sets the maximum intake of cholesterol at 300 milligrams a day.

This means for someone eating 2,000 calories a day that fat must be limited to 600 calories or less. Saturated fat should not exceed 200 calories.

It is a diet, experts say, from which most Americans could benefit. For those with high cholesterol, however, it could be life-saving.

To many people, the idea of counting calories is intimidating. But adhering to the diet can result in the desired calorie goals without meticulous counting, and it need not mean radical changes in eating habits.

What the diet requires is making substitutions, and choice of meat is an important place to start. What's in? Lean anything. That means eating more fish, poultry without the skin, shellfish and lean cuts of red meat. What's out? Fat cuts of beef, lamb, pork, spare ribs, organ meats, sausage, cold cuts, hot dogs and bacon.

The Step One diet also recommends drinking skim or 1 percent milk instead of whole milk or even 2 percent. Cream is out. So are half-and-half, nondairy creamers and whipped toppings. Sherbet and sorbet are okay, but ice cream is not. Low-fat cheeses (with 2 to 6 grams of fat per ounce) are recommended, but the more popular varieties, such as cheddar, Swiss, Camembert and blue cheese, need to be decreased. Low-fat yogurt is okay.

Eggs -- particularly the whites -- are okay, but no more than three yolks per week. Consuming any kind of vegetable or fruit is encouraged -- as long as it hasn't been prepared in high-fat butter, cream or other sauces.

Unsaturated vegetable oils -- such as corn, olive, canola, safflower and peanut -- are preferable to butter, lard, chicken or bacon fat. The tropical oils -- coconut and palm -- are largely saturated fat and are discouraged.

Baked goods -- generally the homemade variety prepared with unsaturated oils -- are okay. Angel food cake is good. So are low-fat crackers and cookies. But eating most commercially prepared baked goods like pies, cakes, doughnuts, croissants, pastries, muffins, biscuits, high-fat crackers and cookies is discouraged. In place of chocolate, which is high in fat, the diet recommends using baking cocoa.

The diet also recommends choosing rice and pasta instead of egg noodles, and eating whole grain breads and cereals that don't have eggs or saturated fats as ingredients.

If after three months, cholesterol levels have been reduced sufficiently, the guidelines call for long-term monitoring to go into effect. Under this regimen, blood cholesterol is retested four times during the first year and twice each year thereafter.

Should the Step 1 diet not be successful, doctors will refer patients to a registered dietitian for additional treatment. For many patients, that may be another three-month retrial on the Step 1 diet.

If that fails, treatment moves to the more stringent Step 2 diet. This eating plan, which for many will require radical changes of habit, limits saturated fat to 7 percent of total calories and cholesterol to less than 200 milligrams day. Fried foods and many meats, such as hot dogs and sausages, are virtually eliminated. Egg yolks are limited to two per week.

If cholesterol levels remain elevated over six months, despite intensive diet therapy, cholesterol-lowering drugs can be added to the treatment.

But since drug treatment usually commits an individual to long-term therapy for years -- and sometimes for life -- the University of Texas' Grundy and others advise great caution in prescribing drugs.

When drugs are necessary, the first three to be considered are cholestyramine and colestipol -- two drugs that are bile acid sequestrants -- and the vitamin nicotinic acid. All three have proven long-term safety. They also reduce the risk of coronary heart disease and lower LDL blood cholesterol levels. Cholestyramine and colestipol work by pulling LDL out of the blood and shunting it to the bile acids, where it is excreted through the intestine. In addition, these two drugs also lower LDL levels by causing the liver to produce more LDL receptors and thus take LDL out of the blood.

Nicotinic acid's action is less well understood, but studies suggest that it blocks the mobilization of fatty acids from fat cells and also interferes with the liver's production of lipoproteins.

But these three drugs also have drawbacks. Cholestyramine and colestipol may interfere with the absorption of other drugs. They can also increase blood levels of triglycerides, or blood fat, which is itself a risk factor for heart attacks. Nicotinic acid, too, has side effects, including gout, elevated levels of blood sugar and interference with liver function.

The second phase of drug treatment is the newly approved lovastatin, first of a new class of drugs called "HMG Co A reductase inhibitors." These drugs both increase the removal of LDL from the body and reduce the production of LDL by interfering with the liver's cholesterol synthesis. Lovastatin and other drugs of this type are more effective in lowering LDL cholesterol than cholestyramine and colestipol are but have yet to have their long-term safety established.

Only 5 to 6 percent of Americans with high blood cholesterol will require drug therapy, estimates Dr. DeWitt Goodman, chairman of the NHLBI panel that drafted the new cholesterol guidelines. For most Americans, dietary changes will suffice. The main message of the new guidelines is "Know Thy Cholesterol."

Changing eating habits to reduce blood cholesterol -- and the risk of heart attacks -- "is no longer hypothesis or theory, but very sound science," says Goodman. "I think that if people really made an effort with diet, we ought to be able to lower cholesterol levels by 15 to 20 percent." And that could mean at least 300,000 fewer heart attacks per year, and 100,000 fewer deaths.

"You can see," said Goodman, "what a big impact this could have."