NVD - record status May 09-May 10...are all your records available for analysis??

The graph below shows the status of AAA records on the NVD per month from May 2009 until May 2010. The White bars represent records that have the data fields fully complete, the Yellow bars represent records that have all the mandatory fields completed but are still missing some data. The Red bars represent records that have incomplete mandatory data fields. When records are Yellow and White they can be included in any data analyses that are performed, however Red records cannot be included because the mandatory fields must be complete in order for the records to be analysed.

Please ensure all your data is entered into the NVD and that your red records are updated as soon as possible. We realise that this requires significant effort by all member so the VSGBI, but with complete data entry we will be confident in our data analysis reports in addition to demonstrating a commitment to gathering data to support patient safety. Those centres collecting data in house for upload should try and upload at least 3 monthly. We will be providing these run charts on a monthly basis to highligh NVD contribution rates.

Comments

[Peter Barker]

From May 2009 to May 2010 there is a distinct downward trend in the total number of cases reported, which, if I understand the graph correctly, will not change when the partial and red cases are completed. Presumably the trend does not represent a real drop in numbers, but either a significant delay in entering data, or even worse a decreasing % of cases being reported? Is there any sound evidence to prove/disprove the reasons behind this trend?

[Helen Hindley]

Thanks for your comment Peter. There appears to be a downward trend displayed here. Although, if you take a look at the June charts, figures are noticibly improved for April/May 2010 which implies that the reason for the reducution in figures would be due to a delay in entering/uploading data. This seems to be a reoccurring event when you compare the charts produced. At the moment we have not completed a sufficient analysis that would be sound enough evidence to say that this is the exact reason for this downward display. I plan to produce some run charts shortly looking at record contribution rates over a longer time period than just 12months which may display a more representative picture of contribution rates, allowing us to look in more detail at confirmed "trends". We are conscious not to fall into the trap of what Peter Lachman highlighted on Monday, that looking at too smaller a time period can look like trends occur, but when larger time periods are considered these previously identified trends are not trends but just normal fluctuation.