Issue link: https://www.e-digitaleditions.com/i/1450659
16 Pharmaceutical Technology Europe BIO/PHARMA OUTSOURCING INNOVATION eBOOK 2022 P h a r mTe c h . c o m they set limits excessively wide. Normal, Gamma, Weibull, Lognormal, Normal Mixture are preferred and should all work well to set limits. • Use Table I to determine the risk and associated toler- ance interval and margin. • Use a K Factor calculator (6) to determine K for the interval (sample size, confidence level and population) for any tolerance interval. • Set Spec using K sigma to define the limits. • Examine results and document limits, update any standard operating procedures or calculators with the defined limits. Conclusion and summary Using the correct methods for evaluating systems suitability and acceptance criteria is critical to ensuring that the ana- lytical methods is fit for use. Setting specification limits as described is statistically rigorous, scientifically sound, and defendable upon regulatory review. Table II is a summary table of all criteria and recommended limits. References 1. ICH, Q2(R1) Validation of Analytical Procedures: Text and Methodology, Step 4 version (November 2005). 2. ICH, Q6B Specifications: Test Procedures and Acceptance Crite- ria for Biotechnological/Biological Products, Step 4 version (March 1999). 3. USP, USP General Chapter <1032>, "Design and Development of Biological Assays," (Rockville, Md., 2015) 4. USP, USP General Chapter <1225>, "Validation of Compendial Procedures," (Rockville, Md., 2015). 5. USP, USP General Chapter <1033>, "Validation of Biological Assays," (Rockville, Md., 2010). 6. NIST, "Tolerance Intervals for a Normal Distribution," in NIST/SEMATECH e-Handbook of Statistical Methods, www.itl. nist.gov, April 2012. PTE Table I. Probability and risk table for setting limits. OOS is out of specification. PPM is parts per million. Cpk is process capability index. Limits z Score OOS PPM Failure rate % Quantiles (probability) Cpk Sigma quality K sigma Tolerance interval confidence/ population Risk K sigma design margin Upper 2 22,750 2.2750132% 0.9772498681 0.67 3.5 2 95/93 Very High -1 Lower -2 22,750 2.2750132% 0.0227501319 0.67 3.5 2 95/93 Very High -1 Upper 3 1,350 0.1349898% 0.9986501020 1.00 4.5 3 95/99 High 0 Lower -3 1,350 0.1349898% 0.0013498980 1.00 4.5 3 95/99 High 0 Upper 4 31.7 0.0031671% 0.9999683288 1.33 5.5 4 95/999 Medium 1 Lower -4 31.7 0.0031671% 0.0000316712 1.33 5.5 4 95/999 Medium 1 Upper 4.5 3.40 0.0003398% 0.9999966023 1.50 6 4.5 95/99974 Medium 1.5 Lower -4.5 3.40 0.0003398% 0.0000033977 1.50 6 4.5 95/99974 Medium 1.5 Upper 5 0.29 0.0000287% 0.9999997133 1.67 6.5 5 95/99995 Low 2 Lower -5 0.29 0.0000287% 0.0000002867 1.67 6.5 5 95/99995 Low 2 Upper 5.5 0.02 0.0000019% 0.9999999810 1.83 7 5.5 95/99999 Low 2.5 Lower -5.5 0.02 0.0000019% 0.0000000190 1.83 7 5.5 95/99999 Low 2.5 Upper 6 0.00 0.0000001% 0.9999999990 2.00 7.5 6 95/9999985 Very High 3 Lower -6 0.00 0.0000001% 0.0000000010 2.00 7.5 6 95/9999985 Very High 3 Table II. Summary table showing criteria and recommended limits. UAR is upper asymptote ratio. 4PL is four parameter logistic regression. Criteria Attribute Assay type Limits Systems suitability Positive control (curve or positive control) All assays Distribution analysis Signal control Curve depth All assays 50% of the baseline Standard curve Signal control Dose test All assays p-value <= 0.05 Parallelism UAR for 4PL curves Bioassay 0.5 to 1.5 Parallelism Slope ratio for linear curves Bioassay 0.5 to 1.5 Parallelism Relative potency (RP) difference of constrained–unconstrained RP Bioassay <= 20% Linearity Linearity ratio Bioassay <= 20% of the linear response Repeatability Confidence interval range % of tolerance All assays <= 30% of the tolerance R 2 Coefficient of determination All assays Report only Analytics