303-530-3888

Topics facing the Infectious Disease Subcommittee have been relatively few. There is the GBS Safety Alert (December 1996.) and the more current activities around discrepant resolution and what is an appropriate gold standard in light of new and advancing technology. The latter two are actively being pursued by working groups within HIMA. FDAÍs statistical group is preparing broadly focused guidance relative to appropriate use of statistical techniques. The HIMA working group is addressing specifically IVD products and how the issue of discrepant resolution should be handled. They have drafted guidance and are working with FDA to refine that draft guidance and to establish how it most appropriately fits with the overall statistical guidance document FDA is preparing. Dr. Kimber Richter is pursuing within FDA the question of what is an appropriate gold standard and what should be done when new technology outperforms currently accepted gold standards. Once she has gathered FDAÍs internal views she will be ready to meet with industry to discuss FDAÍs views and to obtain industries input. The outcome of this effort will hopefully allow both FDA and industry to smoothly cope with these issues.

In an effort to bring closure to the GBS issue the following recaps the events and actions taken regarding this subject and indicates the current status as best I have it.

December 1996 FDAÍs Office of Compliance issues a letter to all Manufacturers of Rapid immunoassay Test Kits for direct detection of Group B streptococcal antigen. They said they had received new scientific information about false negative and false positive results on certain clinical specimens and in certain populations. FDA also referred to cleared devices for testing specimens from infants at risk for GBS disease. They cited two publications. FDA had historically cleared devices based on comparison to blood agar culture using vaginal or cervical swabs from prenatal and intrapartum maternity patients. CDC recommended in 1996 a new standard, lim broth culture of vaginal and anorectal swabs collected at 35-37 weeks. They also recommended that a rapid detection assay have a high sensitivity when compared to selective broth cultures (lim broth). The two studies FDA cites indicate the sensitivity of currently available devices may be low compared to selective broth culture and therefore there may be performance limitations associated with these devices.

FDA convened in July 1996 an Advisory Panel to discuss performance. The intent of this meeting and the opportunity for manufacturers to contribute was not clearly communicated and resulted in, from industryÍs perspective, a decision based on incomplete facts. The panel recommended the performance be re-established against selective broth culture and that package inserts be revised to reflect limitations of the devices in distribution at that time.

FDA requested all manufacturers notify their customers of certain information relative to testing specimens from maternity patients and from infants. In addition, certain precaution statements needed to be added to the package insert.

The Office of Device Evaluation issued a letter dated December 31, 1996 to individual manufacturers. In it they cite a study that compared three devices to selective broth and found low sensitivity for GBS colonization in vaginal specimens from intrapartum women. FDA asked that manufacturers establish performance against selective broth culture and that study results be broken out between prenatal and intrapartum populations. Although most manufacturers chose instead to withdraw their products from the market, one has collected the requested data and is pursing closure with FDA.

FDA, based on the panel recommendations, stated that they feel that clinical performance of direct detection devices on infant specimens (blood, serum, CSF) has not been established and that performance using infant urine with direct detection GBS devices are uninterpretable b because of a lack of correlation with GBS disease. Most manufacturers disagreed with these positions and have individually and collectively pursued clarification of FDAÍs actions. Meetings were held to discuss FDAÍs expectations of what data might be needed. Performance against selective broth culture of maternal patients was available or obtainable however there was significant concern if the practitioners would understand the new numbers in light of how TSA, previous standard, performed versus the new standard. The initial thoughts regarding data needed to support claims for use with infant samples, especially urine, were considered prohibitively costly and would take years to collect adequate numbers of samples, if they could be obtained at all. It should be noted that there are practitioners that feel there is adequate correlation of GBS disease to direct detection results to be useful. The resolution appears to have been inclusion in the package inserts of limitations regarding assay performance with certain samples.