Stratified CU results calculation as per ASTM E2709_E2810.pdf

This presentation describes stratified CU calculation and their interpretation as per ASTM E2709 and E2810. Examples are in this present to show step by step calculation.

1. STRATIFIED SAMPLES CONTENT
UNIFORMITYRESULTS CALCULATION
AND
ACCEPTANCE LIMITS DETERMINATION
AS PER ASTM E2079/E2810
SAMPLING PLAN 2

2. INTRODUCTION
FDAwithdrewitsdraftguidanceforindustryonPowderBlendsand
FinishedDosageUnits—StratifiedIn-ProcessDosageUnitSamplingand
Assessment.
FDA’smajorconcernwasthatsectionsVandVIIofthewithdrawndraft
guidancenolongerrepresentedtheAgency’scurrentthinking.
SectionV(Exhibit/ValidationBatchPowderMixHomogeneity)
recommendedthatatleast3replicatesamplesbetakenfromatleast10
locationsinthepowderblender,butthatonly1ofthe3replicatesbe
evaluatedtoassesspowderblenduniformity.TheAgencycurrently
recommendsthatallreplicatesamplestakenfromvariouslocationsin
theblenderbeevaluatedtoperformastatisticallyvalidanalysis.This
analysiscandemonstratethatvariabilityattributabletosamplelocationis
notsignificantandthatthepowderblendishomogenous.
SectionVII(RoutineManufacturingBatchTestingMethods)acceptance
criteriadesignatedtotheStandardCriteriaMethodandtheMarginal
CriteriaMethodwerebaseduponthelimitspublishedintheUnited
StatesPharmacopeia(USP)GeneralChapter<905>Uniformityof
DosageUnits.

3. INTRODUCTION
However,theproceduresandacceptancecriteriainGeneral
Chapter<905>arenotastatisticalsamplingplanandsotheresults
oftheproceduresshouldnotbeextrapolatedtolargerpopulations.
Therefore,becausetheprocedureandacceptancecriteria
prescribedinsectionVIIprovidedonlylimitedstatisticalassurance
thatbatchesofdrugproductsmetappropriatespecificationsand
statisticalqualitycontrolcriteria,FDAnolongersupportstheiruse
forbatchrelease.
Stratifiedsamplingisrecommendedtobeusedwhenthe
populationisknowntohaveseveralsubdivisions(i.e.,locations),
whichmaygivedifferentresultsforthequalitycharacteristics
measured.TheAgencyexpectsthatnosignificantdifferences
shouldexistbetweenin-processlocationsthatcouldaffectfinished
productquality.Between-andwithin-locationvariabilityisacritical
componentoffinishedproductqualityandthereforeshouldbe
evaluated.Forfurtherguidanceonestablishingacceptancecriteria
forastratifiedsamplingplan,ASTME2709andASTME2810tobe
referred.
Inthispresentation,stratifiedtestresultscalculationand
calculationlimits(ULandLL)determinationasperASTMhas
beendiscussed.

4. CALCULATION AS PER ASTM : EXAMPLE 1
Three units (tablets) from 20 locations were tested and results
are in Table-1:
Table-1 (Assay results in %)
Variance (V)Std Dev (SD)Mean (M)Unit-3Unit-2Unit-1Location
0.3030.551100.6100.6100.1101.2
1
2.6431.626101.6103.4101.3100.2
2
0.1630.404101.9101.5102.0102.3
3
3.4631.861102.8104.8101.1102.6
4
0.0430.208101.2101.3101.0101.4
5
1.1231.060101.1102.2100.9100.1
6
0.6030.777101.0101.6100.1101.2
7
1.711.308101.9101.3103.4101.0
8
7.332.707100.1101.397.0102.0
9
10.873.29799.0100.795.2101.1
10
0.2530.503102.5102.4103.0102.0
11
0.310.557101.1101.0100.6101.7
12
10.4933.239103.9102.2107.6101.8
13
0.2130.46299.999.6100.499.6
14
0.390.624101.2101.4101.7100.5
15
0.280.52999.499.2100.099.0
16
0.4630.681100.2101.099.7100.0
17
0.4430.66699.6100.298.999.8
18
0.3030.55199.999.9100.499.3
19
0.6930.83399.7100.0100.498.8
20

5. CALCULATION AS PER ASTM : EXAMPLE 1
Calculation:
Calculatemean(M)ofeach20locations.ThencalculateStd.deviationof
meanofall20locations.Thisvalueisstd.deviationoflocationmeanor
betweenlocationstd.deviation.
Calculatevariance(V)ofeach20locations.Fromvarianceofeachlocation,
calculateaverageofvariance.[Variancecanalsobecalculatedfrom
standarddeviation.Variance=(SD)*(SD)]
Calculate the value from Table-1,
Std. deviation of location mean or between location std. deviation = Std.
deviation of mean of all 20 locations
= Std. deviation of
(100.6+101.6+101.9+102.8+101.2+101.1+101.0+101.9+100.1+99.0+102.5+
101.1+103.9+99.9+101.2+99.4+100.2+99.6+99.9+99.7)= 1.26
Average of variance = Average of (0.303+2.643+0.163+3.463+0.043
+1.123+0.603+1.710+7.330+10.870+0.253+0.310+10.493
+0.213+0.390+0.280+0.463+0.443+0.303+0.693) = 2.1
SE/Within location std. deviation = square root of avg. of variance = √ 2.1 =1.45

6. CALCULATION AS PER ASTM : EXAMPLE 1
Calculate the value from Table-1.
Minimum assay value in % (from all 60 assay values) 95.2
Maximum assay value in % (from all 60 assay values) 107.6
Overall Mean of assay values in %(Mean of all 60 assay values) 100.9
Std. deviation of location mean between location std. deviation 1.26
Average of variance 2.1
SE/Within location std. deviation (square root of avg. of variance) 1.45
Using value of Std deviation of location mean (1.26 ) and SE/within location
standard deviation (1.45), check LL (Lower limit) and UL (Upper limit) from
acceptance limits for content uniformity sampling plan 2 , 20*3.
From Table-2, LL and UL value is 89.9 and 110.1respectively. It means overall
mean of assay value should be in the range of 89.9 and 110.1. Observed mean
result is 100.9 which is between LL and UL value. Hence, it passes the ASTM
acceptance criteria.
Note: To check acceptance criteria of LL and UL from content uniformity sampling
plan 2 , 20*3, Std. deviation of location mean is 1.26, use value of 1.3 because
1.26 is more than1.2. SE value is 1.45, use value of 1.5 because 1.45 is more
than1.4.

7. CALCULATION AS PER ASTM : EXAMPLE 1
Calculation: Table -2

8. CALCULATION AS PER ASTM : EXAMPLE 2
Three units (tablets) from 20 locations were tested and results
are in Table-3:
Table-1 (Assay results in %)
Variance (V)Std Dev (SD)Mean (M)Unit-3Unit-2Unit-1Location
0.2100.45895.596.095.495.1
1
0.8400.91793.294.292.493.0
2
8.0432.83693.897.192.092.4
3
9.2133.03595.799.093.095.2
4
3.0101.73594.596.094.992.6
5
0.3430.58695.294.595.495.6
6
0.2530.50395.595.095.496.0
7
1.2131.10296.096.594.796.7
8
1.1201.05896.497.296.895.2
9
0.7600.87297.798.398.196.7
10
1.6901.30098.998.1100.498.2
11
5.8432.417101.8102.199.2104.0
12
1.6231.274101.0102.5100.2100.4
13
9.0233.004101.2103.0102.897.7
14
4.2632.065102.4104.8101.0101.5
15
3.6701.916102.0104.2100.7101.1
16
0.0100.100102.9103.0102.9102.8
17
0.0030.05899.899.999.899.8
18
19.0634.366102.5105.7104.297.5
19
14.5433.814105.3101.1106.4108.5
20

9. CALCULATION AS PER ASTM : EXAMPLE 2
Calculation:
Calculatemean(M)ofeach20locations.ThencalculateStd.deviationof
meanofall20locations.
Calculatevariance(V)ofeach20locations.Fromvarianceofeachlocation,
calculateaverageofvariance.
Calculate the value from Table-3,
Std. deviation of location mean or between location std. deviation = Std.
deviation of mean of all 20 locations
= Std. deviation of (
95.5+93.2+93.8+95.7+94.5+95.2+95.5+96.0+96.4+97.7+98.9+
101.8+101.0+101.2+102.4+102.0+102.9+99.8+102.5+105.3)
= 3.63
Average of variance = Average of (
0.210+0.840+8.043+9.213+3.010+0.343+
0.253+1.213+1.120
+
0.760+1.690+5.843+1.623+9.023+4.263+3.670+0.010
+
0.003
+
19.063+14.543
) = 4.24
SE/Within location std. deviation = square root of avg. of variance = √ 4.24 =2.06

10. CALCULATION AS PER ASTM : EXAMPLE 2
Calculate the value from Table-3.
Minimum assay value in % (from all 60 assay values) 92.0
Maximum assay value in % (from all 60 assay values) 108.5
Overall Mean of assay values in %(Mean of all 60 assay values) 98.6
Std. deviation of location mean 3.63
Average of variance 4.24
SE/Within location std. deviation (square root of avg. of variance) 2.06
Using value of Std deviation of location mean (3.63) and SE/within location
std. deviation (2.06), check LL (Lower limit) and UL (Upper limit) from
acceptance limits for content uniformity sampling plan 2, 20*3.
From Table-4, LL and UL value is 99.2 and 100.8respectively. It means overall
mean of assay value should be in the range of 99.2 and 100.8. Observed
mean result is 98.6 which is not between LL and UL value. Hence, it does not
pass the ASTM acceptance criteria.
Note: To check acceptance criteria of LL and UL from content uniformity
sampling plan 2 , 20*3, use value of 3.7 for Std. deviation of location mean
because 3.63 is more than 3.6 and use value of 2.1 for SE because 2.06 is
more than 2.0.

11. CALCULATION AS PER ASTM : EXAMPLE 2
Calculation: Table -4

12. CALCULATION AS PER ASTM : EXAMPLE 2
Ifresultsof20*3unitsdonotcomplywithacceptancelimitsforcontent
uniformitysamplingplan2,20*3.Performanalysisoffurther20locationsand
resultsofall40locationstobecalculatedasperexample2.
UsingvalueofStddeviationoflocationmeanandSE/withinlocationstd.
deviationofall40locations,checkLL(Lowerlimit)andUL(Upperlimit)from
acceptancelimitsforcontentuniformitysamplingplan2,40*3.