Analytical data should be forwarded for reporting only after you have full confidence on the results. At times the reason to doubt your results becomes obvious and the results need to be reinvestigated. Some reasons for reinvestigation are already covered in the earlier article – Honest reporting of laboratory incidents.
In addition to the obvious reasons covered in the article the present article examines some additional factors that call for validation through re- analysis.
Samples are reported failed when values of control parameters lie outside the specified range. There can even be a dispute between the laboratory and the owner of samples that the samples cannot fail on reported grounds. In such cases analysis is conducted again by another set of trained analysts and the mean of results is reported. In case if dispute still persists the sample is retested in one or two other certified laboratories for re-verification of results.
Loss of Sample
Sample loss can take place due to several reasons such spillage, container leakages, breakage of sample tubes, exposure of sensitive samples to light or high temperatures, loss of data files and folders from the system, etc. In such cases fresh sample is requested and analysis of retrieved sample is considered for final reporting.
Deviations in Analysis Conditions
Deviations in analytical method adopted from the prescribed standard warrants repeat analysis. At times such deviations result from adoption of unauthorized changes. Such changes include spectroscopic measurements at different wavelengths, use of unspecified reference absorbance standards, different flow rates, chromatographic columns, operational temperatures and pressures in chromatographic separations.
Determination in outer Limits of Calibration Curve
It is a standard practice to make observations in the linear range of the calibration plot. Above certain concentrations the calibration curve tends to deviate from linearity. In such cases if measurements are made in non-linear portion it is advisable to first prepare a calibration curve after appropriate dilutions of standards and dilute the sample accordingly and carry out fresh determinations in the linear calibration range
Errors in Sample Processing
Errors can also result from deviations in sample processing. These errors include dilution errors, incomplete extraction of analyte due to incomplete digestion, loss of material due to incomplete filtration or centrifugation or uncontrolled heating, etc. In such cases when such faults are identified re- analysis becomes necessary.
The result of any analysis is validated by making use of working standards. Such standards need to be validated periodically to ascertain their control parameters. However, errors can arise in analysis due to use of expired standards. In such situations the analysis becomes essential after making use of valid working standards.
In short it can be said that the results of analysis can be influenced by several variables which require a strict control by adopting validated analysis procedures. However, whenever such variations become apparent re- analysis should be carried out for generation of authentic analytical reports.