| Handling of out of trends results during stability |
1.0 OBJECTIVE To lay down the procedure for handling of out of trends (OOT) within specification results for stability studies. 2.0 SCOPE This procedure is applicable for Calibration of Handling of out of trends (OOT) (within specification) in quality control department. 3.0 RESPONSIBILITY QC – Officers/Executive/Senior Executive/Asst. Manager and Head Quality 4.0 ACCOUNTABILITY Head-QC 5.0 PROCEDURE 5.1 This procedure should cover handling of out-of-trend results observed during stability studies. 5.2 An out of trend (OOT) result can be due to: 5.2.1 Assignable cause. 5.2.2 Non-assignable cause. 5.2.3 Assignable cause for Out of trend (OOT) results can be due to: 5.2.3.1 Laboratory error 5.2.3.2 Error in sample packaging or storage 5.3 Analyst should check the data for compliance with expected trend before discarding test preparations. 5.4 On obtaining an OOT result, the analyst who had conducted the test should report to his section In charge and retain all the test preparations, glassware etc. 5.5 The analyst and the section In charge should carry out an investigation as per the check list as mentioned in format number XXX. 5.6 Laboratory Error: 5.6.1 If investigation reveals that the OOT result is due to laboratory error, then a repeat analysis should be performed on a portion of the same larger sample previously collected for analysis. 5.6.2 Repeat analysis should be done by the same analyst after taking suitable corrective actions. 5.6.3 If on repeat analysis the results obtained are as per expected trend the original test result should be voided and the repeat test results should be reported. 5.6.4 If on repeat analysis the results obtained are still out of trend, then the repeat result should be reported. The out of trend values should be shared with the concerned departments. 5.2.3.3 If on repeat analysis the results obtained are out of trend as well as out of specifications, then a reanalysis should be performed on a portion of the same larger sample previously collected for analysis as below: 5.2.3.4 Re-analysis should be done in duplicate each by the original analyst (say A) and another analyst (say B) 5.2.3.5 The results should be reported as below. 5.2.3.6 All individual results are out of trend (Report average of average) 5.2.3.7 If one analyst gets an out of trend and the other within trend, but difference in results are within an overall RSD of 2.0 % for assay and 10 % for related substance / impurities (Report failing result or average of failing results). 5.7 Laboratory error not identified: 5.7.1 When a laboratory error is not identified, then also a repeat analysis should be performed by the analyst (# 5.6.2 onward). However, in this case if the repeat result again is on OOT or within trend, then average of both the original and repeat analysis should be reported. 5.8 What constitutes out of trend: 5.8.1 Out-of-trend evaluation should be made on asset of data for which number of data points including the “suspect” is not less than three. 5.8.2 Historical data (for older products) should also be evaluated/ taken into account before drawing any conclusion. 5.8.3 Increase or decrease in “process related impurity” only at any station, by more than 100 % (for values from 0.1 to 0.2), and 50 % (for values more than 0.2) of the originally reported (TO) results should be investigated for out-of-trend. The process related impurity if reported “not detected” at TO but observed at the levels of more than 0.05 units should also be investigated. If process related impurity is also a degradation impurity, then limits as in #5.8.4 below should be applied. 5.8.4 A decrease in assay or increase in related substances / impurities is expected on storage. However, an investigation for OOT should be performed, if there is : 5.2.3.8 A decrease in assay by more than 5.0 % (if the previous station shows an assay drop of less than 2.0 % or no drop from earlier station) 5.2.3.9 An increase in RS / impurities (individual or total) by more than 100 %, for values between 0.2 and 1 (if the previous station shows an increase of less than or equal to 10 % only) 5.2.3.10 An increase in RS / impurities (individual or total) by more than 50 %, for values more than 1 (if the previous station shows an increase of less than or equal to 10 % only) 5.2.3.11 A cyclical change (decrease, increase, decrease…) should be investigated for out-of-trend, if and only if: 5.2.3.12 Change in assay at any station (let us say n) is more than 2 % from the last station (say n-1) and is also more than 3 % from the second last station, (say n-2) 5.2.3.13 Change in individual or total related substance at any station (say n) is more than 100 % (for values between 0.2 to 1) and more than 50 % (for more than 1) from both (n-1) and (n-2) stations. 5.2.3.14 If for the same product the real time data trend contradicts the accelerated data then it should be investigated for out of trend (i.e. A decrease in assay and increase in related substances content for real time study and a constant value in accelerated study). 5.2.3.15 Abnormal increase or decrease in pH, water / LOD should be investigated for out-of-trend. 5.8.5 Fill all the entries in investigating form as per format number XXX, whenever there is an OOT. |
