Table 4.

Example results following comprehensive validation assessment.

Data element	Complete-ness	Range as expected?	Random sample reconciled against source system?	Random sample reconciled against separate GSTT dataset	Reconciled against external dataset	Conclusions	RAG
COMORBIDITY	100	Yes	Not applicable	Yes. 97% match	Not applicable	Our investigations suggest that this data is of good quality.	G
INPATIENT UTILISATION DATA	100	Yes	Yes. 100% match	Not applicable	Not applicable	Our investigations suggest that this data is of good quality.	G
NYHA SYMPTOM CLASS	96	Yes	Yes, 50% match	Not applicable	Not applicable	Clinical interpretation led to a low correlation (consistent with low inter-observer reliability)	R
DATE OF DEATH	100	Yes	Yes 100% match	Not applicable	Yes, 85% match.	Delays occur updating GSTT source. Decision made to use external data as source (ONS)	G

NYHA: New York Heart Association.