Objective.-To determine the reliability of cholesterol and lipoprotein measurements conducted in local community laboratories.Methods.-Standardized duplicate serum aliquots at three concentrations (low, intermediate, and high) of total cholesterol, triglycerides, and high-density lipoprotein cholesterol were sent to 21 laboratories used by the physicians participating in the Cholesterol-Lowering Intervention Program. Results obtained from the laboratories were compared with values obtained from the Centers for Disease Control and Prevention-standardized Heinz Lipid Laboratory and with the means of the entire sample.Results.-The mean coefficient of variation (CV) was 1.3% or less for all three levels of total cholesterol, which demonstrates a high degree of precision. Accuracy was also high; over 80% of all laboratories were within 5% of the Heinz Laboratory low reference value, and all were within the 5% range for the medium and high samples. The CVs for triglycerides (<2.3%) and high-density lipoprotein cholesterol (<2.2%) were similar to that for total cholesterol, but up to 56% and 61% of the values fell outside the Heinz reference range for high-density lipoprotein cholesterol (intermediate concentration) and triglycerides (low concentration), respectively. As the Heinz Laboratory has a negative 2.7% bias versus the Centers for Disease Control and Prevention for total cholesterol and high-density lipoprotein cholesterol measurements, a higher percentage of laboratories fell outside the Centers for Disease Control and Prevention range. For medium and high total cholesterol samples, 16% of the laboratories were outside the 5% Centers for Disease Control and Prevention range, and the value was 58% for the low total cholesterol sample. For high-density lipoprotein cholesterol the percentages were 61%, 56%, and 39% for the low, medium, and high samples, respectively.Conclusions.-These data suggest that according to the standards set by the National Cholesterol Education Program Laboratory Standardization Panel, reliability of total cholesterol measurements in local laboratories is high. Lower levels of accuracy were noted for triglycerides and high-density lipoprotein cholesterol measurements.

BACKGROUND:A randomized study was conducted to test the feasibility of cholesterol lowering in physician office practices using the National Cholesterol Education Program Adult Treatment Panel 1 guidelines.METHODS:Twenty-two physician practices in phase 1 and 23 in phase 2 were recruited from communities in Western Pennsylvania and West Virginia. These physicians treated a total of 450 adults in phase 1 (190 men and 260 women) and 480 adults in phase 2 (184 men and 296 women) with hypercholesterolemia. Three models (Usual Care [phase 1], Office Assisted [phase 2], and Nutrition Center [phase 2]) for implementing the National Cholesterol Education Program Adult Treatment Panel 1 guidelines were tested over an 18-month period. The baseline serum cholesterol levels were as follows: 6.51 mmol/L (252 mg/dL) in the Usual Care Model; 6.80 mmol/L (262 mg/dL) in the Office Assisted Model; and 6.96 mmol/L (269 mg/dL) in the Nutrition Center Model.RESULTS:In the patients who were not taking lipid-lowering medication, the mean cholesterol response was significantly different between the 3 models (P < .01). Serum cholesterol levels declined by 0.14 mmol/L (5.4 mg/dL) in the Usual Care Model; by 0.31 mmol/L (12 mg/dL) in the Office Assisted Model; and by 0.54 mmol/L (20.9 mg/dL) in the Nutrition Center Model. Two factors-length of time to follow-up measurement and change in weight-were independently related to cholesterol response across all models. African Americans demonstrated a significantly smaller response than whites in the Usual Care Model, while men demonstrated greater declines in serum cholesterol levels than women in the Office Assisted Model. Patient satisfaction was very favorable in both enhanced conditions; however, those treated in the the Nutrition Center Model were more satisfied (P < .05) with program components.CONCLUSIONS:The impact of nutrition intervention delivered through physician practices on serum cholesterol levels is less than clinically desirable, and new approaches with more aggressive therapy should be tested and implemented.