Table 3.

Comparison of SPC error rates vs. traditional audit error rates on trials audited independently by DCRI internal Quality Assurance group. Comparable error rates suggest that the SPC approach is likely adequate, as is the combination of sample size and audit frequency.

	SPC error rate (per 10,000 fields)	Fields audited	QA audit error rate (per 10,000 fields)*	Fields audited
Study 1	2.5	4048	0	**
Study 2	12.2	93,677***	14.5	4398
Study 3	46	2414	24.8	4475

^*values have been scaled by the response index for the corresponding CRF;error rates are therefore calculated using the same method.

^**QA did not count the number fields since they did not find any errors. They audited 4% of the patients in database at the time.

^***Sponsor requested a larger sample size for each SPC audit.