Decision rules for predicting future lipid values in screening for a cholesterol reduction clinical trial

Area of Science:

• Clinical trial methodology
• Biostatistics
• Health economics

Background:

• Large clinical trials often require serial measurements for patient screening, increasing costs and potentially excluding eligible patients early.
• Efficient execution of clinical trials is crucial due to rising healthcare expenditures.
• Current screening protocols may be resource-intensive, necessitating optimized approaches.

Purpose of the Study:

• To propose and evaluate a novel procedure for predicting average patient values from a limited number of serial measurements during screening.
• To enhance the efficiency of large clinical trials by reducing the number of required screening measurements for potentially ineligible patients.
• To apply this predictive procedure to lipid level screening in clinical trials and population health initiatives.

Main Methods:

• Development of a statistical procedure to predict an average value based on 'm' out of 'n' serially obtained measurements (m < n).
• Application of the procedure in a large clinical trial using low-density lipoprotein (LDL) cholesterol, total cholesterol, and triglycerides as entry criteria.
• Validation using data from a postinfarction population to assess lipid level screening thresholds.

Main Results:

• The proposed procedure effectively predicts average values using fewer measurements, offering a more efficient screening approach.
• In population screening for lipid levels, a single LDL cholesterol measurement above 146 mg/dl indicates a >95% probability that the average of two measurements will exceed the 130 mg/dl treatment threshold.
• This predictive capability can help identify patients likely to be ineligible earlier in the screening process.

Conclusions:

• The developed predictive procedure can significantly improve the efficiency and reduce the costs of large clinical trials.
• This method provides a valuable tool for optimizing patient screening for entry criteria and population health management, particularly for lipid levels.
• Early identification of potentially ineligible patients based on predictive analytics can streamline trial recruitment and resource allocation.