Tuesday, 18 November 2008
| Why not simply use Excel? |
The most popular method of data storage of samples at this moment is with Excel spreadsheets. Spreadsheets were never build for this purpose, for collecting and long term storage of crucial data. A lot of things can and will go wrong, for instance:  Getting the right sample data in a spreadsheetIt is difficult to get factual right data in spreadsheets. Directly typing data in spreadsheets generates an error rate of at least 0,5%. This is not acceptable because samples contain multiple data elements and errors add up. A laboratory that values its data must make special input and validation software to catch the errors of the users. You must always use the same protocol It is difficult to create a reliable protocol for sample storage. It is even more difficult to guarantee that the same protocol is al  ways used. A good system will also check that the data is inserted using the right protocols and checks. But how do you guarantee that the nice temporary worker really follows the right protocol and on the right way. You must never use a backdoor The precise following of protocol also often fails, because there is always a backdoor: directly editing the data in the spreadsheet. This is easy for the users but a disaster from a point of view of data integrity recovery. The spreadsheet gives users a lot of freedom, but from a protocol control point of view, far too much freedom. You must never use spreads for long term storage. The well known problems with spreadsheets are: Spreadsheet files can be renamed, lost or be corrupted.  Rows in the spreadsheets can be accidently deleted Data can be accidently copied to other data. The problem is that this is often only determined when it is too late, when a good backup is no longer available. The only solution is to destroy the associated samples, because a sample is worthless, if you do not know what the sample contains. The Sample Navigator solves all those well known spreadsheets problems: Getting the right data in the sample is done by rigorous input checking and validation, using automatic scanners whenever possible. Always using the same protocol is guaranteed, because data can be stored only by following the work protocol. The Sample Navigator contains no backdoor for data: the only way of storage and editing of data is by strictly following the work protocol. The data is safeguarded for long storage, by using an advanced relational database: MSQL 2005, that permanently checks the integrity of the data. |