1	Recalls Single site venue Connected programs Under construction
2	Recalls Same players Same principles Faster More nuanced
3	Class I Recalls Questions, answers, and puzzles
4	Why more Class I recalls? (9 in FY 04) High risk problems – identified in field Numbers are deceiving Not statistically significant Criteria not changed
5	Have the severity and likelihood of health hazards increased? Yes but incrementally 10 billion lab tests 80% medical decision making More stand alones Background analytical error rate (.008%)
6	How does FDA determine the population at risk? Information from manufacturer Information from field Package insert Premarket submission (if available) Medical officers (in CDRH and outside) Literature
7	How is severity and likelihood of health hazard determined? Contingent Contextual All available information No of complaints Numerator data, denominator data Understanding of practice use Analysis of events
8	Likelihood Remote Rare Occasional Frequent Continuously occurring
9	Severity None (no adverse health consequences) Limited (transient, self limiting illness or injury) Moderate (Significant impairment but temporary/reversible) Severe (Serious injury, permanent, irreversible) Life threat (death could occur)
10	Howe does FDA determine short, intermediate, and long-range? Medical and clinical assessment Usually temporal impact is obvious
11	Mathematical Formulas Decisions are not numbers based but depend on combined factors Index questions on HHE form are for purposes of standardization
12	The Role of the District Office DO serves as a conduit High octane decisions are made by CDRH/CBER with efforts to maintain consistency while taking into account unique circumstances that may exist
13	Press Releases FDA prefers these be initiated by companies FDA likes to review to ensure consistency If company does not wish to issue a press release, FDA will do so
14	Response to Class I Recalls FDA is interactive Interested in working with company to craft plans that protect public health Interested in minimizing disruption to lab community
15	Device Correction and Removals Reports on class III recalls are not required FDA does request that in doubt cases be submitted
16	View from OIVD Long history of recall participation Now faster and better feedback systems Consistent, helpful Regulatory conscience Conservative stance – no history of over-calls
17	View from OIVD Laboratory based view Recalls are sign of function quality system Worry if none Opportunity to show quality commitment J&J and the Tylenol model