First of all we are very pleased with the data quality for (SPOT)light. We expect around 2-3 errors per record (on average) and the average is around 0.5 (or 1 every two records). However, if you have a large number of records then this still translates into a large number of checks and a lot of work.

So to justify this then here are a few of the most common DVR checks with an explanation to justify why we are asking for your help:

• Unusual spelling for first names: We will be unable to match to death certificates with the Office of National Statistics if the name spelling is odd. We expect some false positives with unusual names but the system should learn as we go along. Unfortunately a decent number of records across the entire study have first and last name reversed, and so despite the false positives this is still an important check.
• Follow-up or recommended level of care missing. We’ll use this to classify patients  - especially important is the case where treatment limits are in place and we don’t want to consider that patient’s subsequent death outside of ICU as unanticipated.

I’ll add other explanations as they come up.

Please do get in contact if you have any queries.

Steve