Lipoprotein (a) testing- insights from the laboratory

Results: All calculated methods demonstrated a systematic negative bias relative to direct measurement (Figure 1).Passing-Bablok regression confirmed both constant and proportional negative bias for all equations: Friedewald (y = 0.966x -0.231), Martin-Hopkins (y = 0.979x -0.188), Sampson-NIH (y = 0.971x -0.201), Modified Martin-Hopkins (y = 0.981x -0.175), and Modified Sampson-NIH (y = 0.988x -0.120).The Modified Sampson-NIH equation showed the most favourable performance, with the smallest mean bias (-0.12 mmol/L), narrowest proportional bias (-1.2%), and robust agreement across the widest triglyceride range (up to 11.3 mmol/L).A reflex strategy using this equation would reduce direct measurement by 99.8%, with only 149 of 71,934 samples (0.2%) requiring direct testing, yielding estimated annualised cost savings of approximately £238,000 (Table 1).Conclusion: The Modified Sampson-NIH equation provides analytical performance comparable to direct measurement and offers a robust, cost-effective alternative for routine LDL-C reporting.Adopting a calculated-first strategy aligns with national guidance, improves accuracy at contemporary therapeutic targets, and yields significant resource savings, supporting its implementation in routine clinical practice.