List of Factor affecting data quality

Factors Identified Effecting Quality of Data

This list has been generated after the presentation of the results of the Peer audit at the [[Collection Team Meeting September 29, 2010].

Please feel free to add other thoughts and ideas Template:Discussion

1. One example identified at a meeting today by CMarks was if there is no MOST, because a patient was in palliative care, and only vital signs available is from 1 week prior to acceptance to medicine service then what do you put in as a BP, HR, WBC? The guidelines state if there are no values assume normal. But what are the normal values that should be recorded?
   • Solution: Tina suggested putting not available and when data sent to Server, values will be output as a preset standard normals.
2. Another example ID'd by LBilesky was for APACHE - the exact physiological item is not recorded because APACHE score is based on selecting a value within a RANGE for points. If the points are the same why fuss about which value to select? There were never any guidelines instructing a collector which value to choose within a range.
   • Solution: Update guidelines for APACHE II collection rules.
3. PStein - ID'd that she sometimes find 2 or more sources of DATE and TIME first accepted to the Medicine Service. She uses the first date and time she finds on the chart?
   • Solution: Update guidelines so that FIRST date and time found on a chart or EPR at STB is recorded.
4. FLindell - if actual med ward arrival time is obtained the data collector from the ward log book, then a auditor of a chart would not have assess to the log book, hence the ward arrival date and time is not easily reproduced. --TOstryzniuk 17:24, 30 September 2010 (CDT)
   • Solutions: ?