Spiked recovery experiment

Precision and accuracy Quantitative analysis by NMR is very precise with relative standard deviations for independent measurements usually much lower than 5%. The largest errors in NMR measurements are likely due to sample preparation, not the NMR method itself. If a good set of standards is available and all NMR measurements for the test and standard samples are performed under the same acquisition conditions, the quantitative results can be readily reproduced on different instruments operated by different analysts at different times. Therefore, good intermediate precision can also be achieved. An accurate quantitative NMR assay will require accurately prepared standards. The accuracy of an NMR assay can be assessed, for example, by measuring an independently prepared standard or an accurate reference sample with the assay. In many cases, a spike recovery experiment can also be used to demonstrate the accuracy of an NMR assay. 

In order to assess matrix effects, spike-recovery experiments were performed on samples from 10 individual human and six monkey matrix lots. Human samples were spiked with concentrations at the LLOQ and 10-fold the LLOQ. Monkey samples were spiked with concentrations at the LLOQ, two- and 400-fold the LLOQ. The results of the spike-recovery experiments are shown in Table 6.2. Some unspiked human samples showed a substantial amount of X, and their values were subtracted from the spiked sample results before calculating the spike recovery. For samples from monkey plasma samples, all samples were blank and no correc-... 

Table 6.2 Spike-recovery experiments of compound X to human and monkey multiple matrix lots.

NIR methods can be validated by the conventional protocols described in ICH and other regulatory guidelines as they are currently written however, some modifications have to be made to account for differences between spectrophotometric and chromatographic experimental parameters. For example, spiked recovery experiments are not relevant because NIR responses are sensitive to the production process. The criteria suggested for validating a NIR transmission method include ... 

For a majority of biomarker assays, standard calibrators are prepared in an analyte-free alternative matrix instead of the de facto sample matrix of patient samples. For such methods, it is crucial to demonstrate that the concentration response relationship in the sample matrix is similar to that of the alternate matrix. Spike-recovery experiments with the reference standard may be inadequate to evaluate the matrix effect, as the reference standard may not fully represent the endogenous analyte. Instead, parallelism experiments are performed through serial dilutions of a high-concentration sample with the calibrator matrix. Multiple individual matrix lots (>3 lots) should be tested to compare lot-to-lot consistency. In the instance that the limited amounts of sample are available, apooled matrix strategy can be used with caution as discussed by Lee et al. [15]. The samples can be serially diluted with the standard matrix (standard... 

The errors associated with the steps identified in Section 12.13.5.1 have been discussed and the approaches used to establish and combat them will now be summarized. For the extraction step, QA considerations mean the extraction efficiency needs to be validated and this can be done either by spike recovery experiments or by using a representative CRM. The main criticism of spike recovery experiments is that the spike is not bound in the sample matrix in the same way as the naturally occurring analyte being measured. However, if a low recovery is apparent using the method chosen, it would indicate an inadequate protocol or one that requires further developmental work. The range of CRM available with values for some of the more important chemical species is increasing, and includes various sediments, fish and shellfish tissue, and human matrices such as hair and urine. However, real samples are rarely identical to the matrix CRM available, so care should be taken when validating methods to be used for real samples. 

No method of wet digestion is ideal, and both contamination and losses will occur to a greater or lesser degree whichever method is used. The task of the analytical chemist is to choose a procedure that will minimize interference from contaminants, reduce losses as much as possible, and therefore bring errors within acceptable limits. At all times, blank reagents and spiked recovery experiments are required to establish the degree of analyte contamination and loss from the dissolution process. 

In a performance-based approach to quality assurance, a laboratory is free to use its experience to determine the best way to gather and monitor quality assessment data. The quality assessment methods remain the same (duplicate samples, blanks, standards, and spike recoveries) since they provide the necessary information about precision and bias. What the laboratory can control, however, is the frequency with which quality assessment samples are analyzed, and the conditions indicating when an analytical system is no longer in a state of statistical control. Furthermore, a performance-based approach to quality assessment allows a laboratory to determine if an analytical system is in danger of drifting out of statistical control. Corrective measures are then taken before further problems develop. 

The only recourse is to modify the recovery experiments above in the sense that the sample to be tested itself is used as a kind of blank, to which further analyte is spiked. This results in at least two measurements, namely untreated sample and spiked sample, which can then be used to establish a calibration line from which the amount of analyte in the untreated sample... 

To validate the analytical procedure recovery experiments are performed. To this end, the CRM is spiked with a known mass of the analytes at a variety of concentration levels (at least three different levels) and the concentrations measured are compared to the expected concentrations in at least three separate experiments. The extraction step has been shown to be a critical step in the analytical procedure and it may be responsible for poor recoveries. The efficiency of this step can be assessed either by repetitive extraction of the sample or by the addition of internal standards prior to the extraction step with the assumption that the latter actually represent the behavior of the analytes of interest. 

It should be noted that a recovery factor obtained using a single reference material or single spike addition experiment does not indicate the absence of systematic... 

Regular, routine sample recovery measurements can be made by using the method of standard addition. The matrix is spiked with the analytes in a small volume of solvent at a level which is 50%, 100%, 150%, and 200% above the estimated level in the sample. A number of independent replicates should be made at each level. Provided that sufficient material is available the sample can be analyzed prior to spiking. In case of limited size (e. g., small tissue samples) a number of samples may be pooled and homogenized for such recovery experiments. 

Note In recovery experiments, the spike levels were 150mg for vitamin C, 2-20mg for the other WSVs, 5.0mg of vitamin A acetate, 0.5mg of vitamin D2 and 25mg of vitamin E acetate. 

A third possibility to estimate the bias component of the uncertainty is to perform a recovery experiment, where a sample is spiked with the analyte. The nncertainty of the nominal valne (i.e. the spiked amount) has to be calcnlated from the uncertainty of the volume added and the concentration of the spike solution. In the example in the slide the spiking is done using a micropipette. The uncertainty of the concentration of the spike solution is taken from the certificate of the manufacturer of the standard solution used and the uncertainty (repeatability and max. bias) of the micropipette is delivered by the respective manufacturer. Both components are combined to the uncertainty of the spike. 

Matrix effects are sometimes not obvious to recognize, which is one of the major pitfalls when using LC-MS/MS. The most practical experiment is probably a recovery experiment. A spiked matrix sample is analyzed and the result is compared to a spiked solvent sample. If no matrix effect is present, the same concentration should be found in both samples. 

The extraction recovery has been assessed for this assay as well. A very good recovery of > 98 % was found for all analytes, and for all concentration levels of the quality control samples (5,15 and 50 pg/L). The recovery experiment was done by comparing analyte peak areas that were obtained from plasma samples spiked before the extraction, with equal concentration levels spiked post-extraction. 

The recovery experiment carried out for this assay resembles somehow the sum of two potential effects The loss of compound during the sample preparation process (in this case on-line extraction) and a potential matrix effect during sample analysis. However, since an online sample clean up procedure is used here, these two potential effects cannot be separated in the usual way (One experiment would be the comparison of a processed spiked plasma sample with a blank plasma sample spiked post sample processing in order to determine the recovery of the sample preparation step. In a second experiment, again a processed blank plasma sample would be spiked post processing and the result would be compared with the response of a spiked solvent sample in order to reveal a potential matrix effect during analysis). 

For time experiments, when the initial recovery experiment is completed (when using a real drug product, not from spiked experiments), the solution that is left over is set aside for time t (usually 24 hours at a temperature condition known not to affect the inherent stability of the solution). After time t, the solution is re-shaken by hand and then re-injected to determine assay value. 

The accuracy of the method can be determined by performing recovery experiments or by comparison to another analytical method (such as titration, DSC, and PSA). Spiking experiments, where increasing amounts of an impurity are introduced into the sample and the accuracy of the result versus... 

Yield and recovery experiments for copper were carried out as described in the 8-HQ method. Spike recovery was observed to be 50-75% complete when carried out at pH 3. Precision was again measured as half the range of duplicate results, or 11 %, for the same seawater sample. 

The recovery of a known amount of analyte from a sample matrix has long been taken to demonstrate accuracy. For this to hold true, the standard material must be pure and the matrix must not interfere with the reactions. When the experimental conditions have been set, the recovery experiments involve spiking a base matrix with analyte. Recovery (R) is calculated from the signal difference between the spiked base and the base itself, and is equal to (measured concentration-base concentration)/(added concentration). The volume of the spike should be small (< 10% of the total volume), so that the matrix is essentially unchanged by the spike. The recovery (%) can be calculated using Eq. 16.30. 

Make four recovery measurements by spiking the contaminated wood sample with a known amount of the analytes. The spiking level shall be varied in each recovery experiment from 50% to 200% of the target value of the contaminated wood sample. The material has to be allowed to absorb the spiked analyte at least 24 hours. 

Chromatogram E is the sample of apple juice under scrutiny, spiked with 100 ppm of malic acid. The increased concentration of both malic and fumaric acids is due to the presence of these organic acids in the original sample, plus the addition of malic acid contaminated with fumaric acid. The verification of the purity of the standard is very important as contamination will give an erroneous result. Spiking is essential in that it 1) contributes to identify the peak of interest through comparison of retention times, 2) enables the analyst to perform a recovery experiment since the original concentration of the standard is known. From these data, it can be concluded that the sample under scrutiny meet the standards of the authentic apple juice sample. 

It is important to emphasise that the different manifold architectures originally designed for the application of SAM have also been used for spiking purposes to assess accuracy. This practice is however not recommended, as analyte addition for recovery experiments should always be done before the analytical procedure. 

Recovery experiments were performed by spiking high-purity water (Lichrosolv, Merck, Germany) with concentrations of 800 ng/L to 1500 ng/L of the respective reference compounds dissolved in methanol (see Table 1). After an equilibration time of 12 hours the extraction procedure was performed as described above and the extracts were analysed by GC/MS and GC/irmMS. Former investigations revealed recoveiy rates between 25 % and 95 % for the individual substances (Dsikowitzky et al., 2002, Dsikowitzky et al., 2004a/b). 

An important precondition for the successful determination of carbon isotope ratios is the prevention of isotopic shifts as a result of the analytical procedures applied. Therefore, five recovery experiments were performed in order to detect changes of the carbon isotope ratios during sample preparation and measurement. The compounds selected for these experiments are known riverine contaminants and comprise hexachlorobutadiene, several musk fragrances, phthalates and other plasticizers, a flame retardant and a pesticide. All recovery samples were spiked with concentrations between approx. 800 ng/L and 1500 ng/L for each compound representing a common abundance level in river systems. 

However, spiked recovery can give insights into possible matrix effects. Moreover, many commercial assay kits quote recovery experiment results from the manufacturer s internal validation procedures and hence reproducing those experiments in the assay validation exercise is often useful in demonstrating that the assay is performing as expected by comparing results with the manufacturer s quoted data. 

---