Laboratory practice session 02

PHARMACEUTICAL ANALYSIS DATA TREATMENT. VALIDATION OF ANALYTICAL PROCEDURES

I. Questions and tasks for discussion:

1. The terms «error in analysis» and «uncertainty of measurement». Classification and control of errors.
2. Statistical treatment and reporting of results. Confidence interval.
3. What are the two main validation parameters? What types of errors do they characterize?
4. How many measurements must be performed to obtain statistically reliable data during the validation process?
5. What are the main ways for achieving the accuracy?
6. Give your definition of «repeatability» and «reproducibility».

II. Project «Validation of Analytical Procedure».

Description

Pharmaceutical company is planning to sell generic. To confirm that the manufactured product meets quality requirements, the company orders a testing laboratory to develop an analytical procedure of quantitative determination of acetaminophen (assay).

For the research laboratory the possibility to get this order depends on the reliability and on validity of its analytical procedures.

So, if the first task for you is to select and evaluate regulation validation parameters in terms of ICH, the second task will be to assess the ability of testing laboratory to provide competent performance validation.

When the Lab receives from the Customer a list with approved ICH validation parameters required for the HPLC determination of acetaminophen, Lab should make sequence (order) of their evaluation. Along the way, set the appropriate limits for specification parameters (Table 1).

Table 1 Necessary validation parameters & specifications

Validation parameter	Acceptance criteria (norms)
Specificity	Homogeneity of the peaks, resolution coefficient (Rs) >1,5%
Linearity	Correlation coefficient (r) > 0,998
Range	Homogeneity of dispersion confirmed by Fischer method
Accuracy	Recovery is 98–102% (spiked placebo, standard addition of analyte to placebo)
Repeatability	RSD < 1,5%

Assessment of relevant validation parameters is performed according to the scheme shown in Figure 1. Selected validation parameters will be determined after the necessary research is conducted.

 

Figure \(\PageIndex{1}\): Sequence of validation parameters determination

1. Precision study (repeatability conditions).

HPLC separation and determination of acetaminophen in testing solutions was performed six times in repeatability conditions. The mass concentration of acetaminophen in testing solutions equals 30 mg/l. Peak areas are shown in the table 2.

Table 2 HPLS assay results: chromatographic peak areas

Run	Peak area, u
1	63 120
2	62 601
3	63 887
4	62 714
5	62 986
6	63 407

Task

Using relevant software or calculator estimate the Standard Deviation and RSD, % (coefficient of variation).

2. Specificity Study.

The expected value of mass concentration of acetaminophen in the sample was 30 mg/l. The Figure 2 shows the HPLC chromatogram obtained with the assay of sample. It can be seen that except the main peak of acetaminophen, there is a peak of relative component. Homogeneity of the peaks was performed by the means of chromatography-mass spectrometry to assess the specificity. The results showed that the peaks belong to pure substances, the homogeneity of the peaks in this case was confirmed.

Figure \(\PageIndex{2}\): Chromatogram of standard sample with known acetaminophen concentration

Task

Calculate resolution (R_s), which defines the degree of separation. Use chromatographic parameters (Fig. 2.8) and equation:

[equation]

where Δt – distance between the peaks;

w – width of each peak at the base as shown in Fig. 2.8.

3. Range Study.

Six calibration solutions with acetaminophen concentration from 20 to 45 mg/l were prepared to test the range for homogeneity dispersion. The measured peak areas are shown in Table 3.

Table 3 HPLS results for calibration solutions

Peak area, u (cacetaminophen = 20 mg/l)	Peak area, u (cacetaminophen = 45 mg/l)
45 530	99 472
45 209	98 912
44 889	99 130
45 022	99 899
45 410	99 451
44 706	98 971

Homogeneity of dispersion at the end of the calibration curve was performed by Fisher method.

Task

Calculate Fisher statistics (ζ) using equation:

[equation]

where S1, S2 – dispersions of the two series of chromatographic data, S₁² > S₂².

If the confidence level P=95% and the number of degrees

[equation]

critical value equals 5,05. Make a conclusion about the homogeneity of dispersions of the two series of chromatographic data.

4. Linearity Study.

To estimate the concentration (mass) of acetaminophen in the test solution calibration solutions were prepared and the area of the peaks was measured using relevant software. The HPLS results are presented in Table 4.

Table 4 HPLC analysis of the calibration solutions of acetaminophen

Cacetaminophen, mg/l	Peak area, u
20	45 450
25	55 290
30	65 165
35	76 254
40	88 019
45	99 309

Task

Using data shown in Table 4 and Microsoft Excel construct a calibration linear curve «C_{acetaminophen} – peak area» and obtain linear regression equation. Calculate correlation coefficient (r).

5. Accuracy Study.

To determine the recovery, % three placebo samples were spiked by 24, 30 and 36 mg of paracetamol respectively («3 × 3 method»). Concentrations of paracetamol obtained by HPLC analysis are shown in the Table 5.

Table 5 HPLC analysis of the calibration solutions of acetaminophen

Run	Added amount of acetaminophen, mg	Obtained amount of acetaminophen, mg			Mean, mg
1	24	20,2	21,6	21,5	21,1
2	30	26,8	27,5	26,6	27,0
3	36	32,5	32,6	31,6	32,2

Task

Using data shown in Table 6 calculate recovery, % by the equation:

[equation]

6. Results and reporting.

Summarize the results of validation studies in the table 6 Check the acceptance criteria: are the established norms fulfilled? Fill in the last column of the table (Yes/No).

Answer the question: Would you trust this laboratory to perform an assay?

Table 6 Validation results

Validation parameter	Acceptance criteria	Result	Acceptance criteria met (Yes/No)
Linearity	correlation coefficient r > 0,998	correlation coefficient r =
Range	Fischer method (P=95%, n=6) statistics over the table value	Statistics value = Critical value =
Specificity	homogeneity of the peaks is observed when resolution Rs >1,5%	homogeneity of the peaks is observed when Rs =
Accuracy	Recovery = 98–102% (spiked placebo method, standard addition of analyte to placebo)	Recovery =
Repeatability	RSD < 1,5%	RSD =

III. Self-test:

1. Metrological certification of analytical procedure is – ...
   1. its qualification.
   2. its validation.
   3. laboratory verification of the its suitability.
   4. within-laboratory quality control.
2. Ways to identify and eliminate systematic errors in analysis …
   1. varying the sample size.
   2. use standard additions of analytes.
   3. use of reference materials.
   4. application of all mentioned ways.
3. Validation of analytical procedure is performed by evaluation the following characteristics (parameters):
   1. trueness
   2. repeatability
   3. robustness
   4. selectivity (specificity)
   5. Linearity
   6. detection limit
   7. quantitation limit
4. Random errors …
   1. characterize the precision of analysis.
   2. can be estimated by statistical treatment methods.
   3. can not be estimated by statistical treatment methods.
   4. have causes that can not be established definitely.
5. Systematic errors …
   1. characterize the accuracy of analysis.
   2. characterize statistically significant difference between the average and the true value.
   3. can be identified using reference standards, certified metrological procedures or by standard additions method.
6. Precision evaluation includes assessing:
   1. trueness
   2. repeatability
   3. accuracy
   4. limit of detection
   5. reproducibility