Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Biotech / Medical : ACMI - Accumed Inc. -- Ignore unavailable to you. Want to Upgrade?

To: george who wrote (1051)	1/10/1998 12:51:00 PM
From: Sigmund	Read Replies (3) \| Respond to of 1894

You may be right but this has been said throughout industry commerce and other areas of social activity from the start of the information technology revolution. And quite frankly it really doesn't make much sense. I am not saying that the existing automated screening products are good enough. That is up to the FDA to decide. But we certainly know about the fallibility of people. They get tired, they are not all as good or motivated as each other so the patient gets the luck of the draw. Even in chess the computer ultimately beat the human. It is destined to happen in diagnostics. Whether it is here today remains to be determined but to think that it ultimately won't happen is shortsighted. The trick is to use the computer in ways that people get to do their part where they are most effective and economically justified. This often requires a reorganization of the way things are done. Thus I tend to agree with you that simply replacing people with machines is not the best solution. But you can't just keep the existing primary/secondary model and think that this is the way to assess where automation will fit in in the long run. In IT sometimes the long run is measured in months. Back to humans. My girlfriend is a lab director for a generalized lab. She used to be a hematology supervisor. She had someone working as a tech and she wondered about him. After a while she figured out that he was rather blind. He could run the tests (could read the digital displays) but screwed up on the set up and quality control or something like that. I imagine that some cytotechs have better vision than others and for all of us our vision varies from day to day and moment to moment. Try the same vision test when you have the flu. Do labs send cytotechs home when they have the flu? when they partied all night? do they give them an alcohol or drug test each day before they are allowed to read a slide? Do they have to report each day on what medications they are taking? How about if they just broke up with their significant other? Or if they are thinking about the stock market etc etc. Let's be realistic about the virtues of human beings as sensors. I don't think that ACMI depends on primary screening not being approved. I think that ACMI will do even better if primary screening is approved. The problem with AutoPap and other such devices is that they only do the test and they are very expensive. Small and medium sized labs will not be able to afford them and for the large labs, it doesn't make much sense to have such an expensive machine (even if you are paying per test) if you don't automate the entire process around the machine i.e. record keeping, reporting, facilitating the work of the techs on the slides which are not elimated by the automated screening etc. I think the future of ACMI is as part of a more comprehensive and integrated system which includes primary screening and which utilizes a smaller number of better trained and even better paid technologists to greatly improve the accuracy of Pap screening and other diagnostic tests. Progress will happen and ACMI will be part of the overall solution. It will not be ACMI in the absence of progress elsewhere in the lab.

To: george who wrote (1051)	1/11/1998 2:02:00 AM
From: Jim Armstrong	Read Replies (2) \| Respond to of 1894

The problem as I understand it is twofold, there are not enough cytologists to handle the load, and there is an unacceptably high error rate on interpretation of the slides, both false negative and false positive. There are many cytologists working over the legal limit of slides on a given day, and some even working outside the lab environment to get the slides "done". Add to that the fact that most slides are clear, and the job is repetitive (and basically boring), and you have a situation where there is an overall high error rate on the slides. Because most of the slides are clear, there is a tendency to zone out and miss some features of importance as the day grows older. The more tired the cytologist, the greater the chance for error. And if the cytologist is one of those cheating, doing 100, 200, or maybe even 300 more slides than the legal daily limit, the chance of error is even larger. The automation is tireless. Plus, its function is only to weed out the unambiguously clear slides (and there will be a lot of them). The ThinPrep process greatly improves the clarity of the slides, improving both machine and cytologist evaluations. The effect of the automated primary screening is to enrich the slide complement that will be evaluated by the cytologists. The result is a win-win. The total number of slides to be cytologist-evaluated is smaller. That brings the number of slides that need to be evaluated by cytologists more in line with the number of cytologists available to read them (legally). In addition, there is more to find in the residual slide population, so the risk of cytologist zoning-out is reduced, improving the error rate. This is truly very important, because the error rate has been pretty high, high enough that you'd really want it done twice to improve the odds of a correct reading. So it gets down to whether the machine's concept and algorithms are demonstratably capable of screening out a relatively minority of clearly clear slides with high accuracy. That's clearly vital (in every sense), and that's what the approval of the machine for primary screening is about. If the data truly proves that the machine is highly accurate in this primary screening task (the question before the FDA), then it will save lives because it will reduce errors, both false negatives (killers), and false positives (a source of anxiety at the very least, and potentially worse). I think I could live with that. I think women could live more securely with this additional screening asset. It represents a powerful asset for relieving part of the load of an overworked cytologist population, and greatly reducing existing error rates. It's now up to the FDA to see if the primary screener accuracy case is made. JimA