Activity data-quality requirements

The data-quality requirements for QSAR models relate to several aspects of the experimental procedure, data transformation and the selection of the appropriate test compounds. Only if the input data of a QSAR meet the highest quality standards may a sound model be derived. Because the accuracy of predictions can never be better than the variability of the respective measurements (usually 20% and more), validity assessment of the activity and effects data is crucial in QSAR derivations. The data should be generated by tests that are methodologically and mechanistically defined. The latter is not trivial for parameters such as biodegradability, soil sorption and ecotoxicity. With regard to the considerable variability of measurements, inter- and also intra-laboratory, the test results, especially when collected from different literature sources, should be critically evaluated with respect to ... 

Assurance of acceptable data quality requires a system which ensures that all activities associated with the sampling program be well defined and supported by accepted and standardized practices. 

Many scientists confuse the terms QC and QA. In terms of clinical trials, there is a very real difference. QC is the operational techniques and activities undertaken by all participants to verify that the quality requirements of the clinical trial have been fulfilled whereas QA verifies that the QC has satisfied these requirements. In other words, QC is where the data recorded is checked with source documents, and that measurements and procedures followed are those described in SOPs and the protocol. QA is where independent individuals establish that QC is in place and report any deficiencies without bias. 

The business activity of the organization dictates quality requirements for the LIMS. Security and regulatory requirements for LIMS data define the level of effort expended to validate a LIMS and the data being stored. In addition, the quality of the hardware and software used to implement the LIMS both play a role in determining overall system quality. 

At the top of the data collection pyramid shown in Figure 5.1 is assessment. By the time the data collection process enters assessment, its third and final phase, the provisions made in the planning phase have been already implemented in the field and at the laboratory as the requirements for sampling, analysis, and QA/QC activities. In the assessment phase, by conducting Task 6—Data Evaluation and Task 7—Data Quality Assessment, we will establish whether the collected data are valid, relevant, and may be used for project decisions. 

In general, the greatest resolution can be obtained in estimates of current or recent historical (within the last 10 to 15 years) emissions. This is because reliable data on fuel use (both quality and quantity) and other activity levels are available, and good estimates of emission coefficients and control efficiencies are available. As one goes further back in time, the data needed for detailed emission estimates are either not available or are less reliable. Recently, SOp and nitrogen oxide emissions were estimated for EPA for the period 1900 to 1980 at the state level by fuel and source sector (4J1). It was particularly difficult to obtain reliable estimates of pre-1940 fuel use and quality, control efficiency, and emission coefficients. Obviously, the less data that are available, the simpler the methodology that must be used. A discussion of a data set required for detailed analysis of emissions and deposition is beyond the scope of this paper, but is available elsewhere (6). 

All sampling and laboratory activities are aimed at one target the production of quality data that is reliable and has a minimum of errors. Further, reliable data must be produced consistently. To achieve this, an appropriate program of quality control (QC) is needed. Quality control includes the operational techniques and activities that are used to satisfy the quality requirements or data quality objectives (DQOs) (FAO, 1998). 

For each herbal preparation, a comprehensive specification is required. This should be established on the basis of recent scientific data and should give particulars of the characteristics, identification tests, and purity tests. Appropriate chromatographic methods should be used. If deemed necessary by analysis of the starting material, tests on microbiological quality, residues of pesticides, fumigation agents, solvents, and toxic metals should be performed. Radioactivity should be tested if there are reasons for concern. A quantitative determination (assay) of markers or of substances with known therapeutic activity is also required. The content should be indicated with the lowest possible tolerance (the narrowest possible tolerance with both upper and lower limits stated). The test methods should be described in detail. 

In Fig. 22.2 the elements of QSAR or QSPR studies are depicted. On one side high quality and relevant biological data are required, while on the other relevant chemical descriptors should be defined. A further critical element is the proper choice of a model to investigate relationships between these data. If the right prerequisites are met, relevant information may be extracted from the data, which can be used to get better understanding of the molecular structures and possibly the mode of action at the molecular level. This information may then be used to predict the properties and activities of new compounds. 

Active control implies that the "paper on glass" documents are no longer static. Data may be collected automatically so that only a few of the data entries are manually entered, most of them are entered through automatic data capture, possibly with a border value check and some kind of compliance enforcement. This requires a certain level of integration with physical data equipment, such as control systems, instruments, bar code readers and so on, which may prevent common errors from manual data entry and additionally provide enforcement of basic quality requirements. Active control requires the paperless system to be a true application that actively assists the user s data entry with automatic data collection and possibly interfaces to other systems. 

The aforementioned measurement proved that the system reliability had been compromised, requiring action to be taken, such as the execution of new studies on energy consumption by the W SN, in order to ascertain if some of the defect sensors could be recovered. If this action failed to yield the desired effect, a localized investigation of the defective sensors would be necessary, when they could be repaired or even replaced. Considering that is possible to take advantage of WSN in terms of reductions of eneigy, if compared with another centralized solutions, these benefits just only is possible if the data quality is reliable. For this purpose, is necessary to establish a set of dimension oriented to the specific issue. For example, the computation of completeness dimension should be more than 95%, that means, 95% of sensors should be active. 

The second document is the QA Project Plan , a technical document that specifies the QA and QC requirements for each project. The plan specifies any QA/QC activities required to achieve the data quality goals of the project. It describes how all data are assessed for precision, accuracy, completeness, comparability, and compatibility (components of data quality criteria and objectives). The QA Project Plan further requires that all data generated be thoroughly documented, and be in sufficient detail to permit unambiguous evaluation of project results. 

---