However we are just beginning to realize why they don't work when it seemed, in thepast, so obvious that they must work.
We now know that unexpected death on the hospital general floor usually proceedsby one of three basic patterns over time. These have been called the patternsof evolving death(PED) and conventional hospital monitors cannot identify ordetect them.
In just one recent example, a 63 year old sleep doctor colleague of mine diedabout 2 months ago. He had a elective surgery for colon polyp removal. He wenthome and then was developed with a fever then he developed shortness of breath.They thought he was OK and didn't need ICU, but then his condition worsened onthe general floor, still they thought he was OK. Finally, he was urgentlytransferred to ICU and died shortly after due to septic shock.
This is a type 3 pattern of evolving death. A simple threshold pulse oximetry alarmsystem provides a false sense of security with this pattern and I have seen somany delays in treatment due to this I had a annual lecture at my hospitalwarning of this pitfall of threshold pulse oximetry as an alarm system on thegeneral floor.
Since it is not possible to know in advance which PED (if any) will occur in anygiven patient, the simple threshold monitor may provide early warning with onepattern of death or may provide a false sense of security (delaying action)with another pattern of death so this is why these devices have provenineffective when studied prospective. The monitors are simply not smart enoughto tell the nurse that a potential pattern of death is evolving and which typeit is.
It is ironic that a physician may die in his own hospital connected to a pulseoximeter with an utterly simplistic 20+ year old threshold alarm processorwhich cannot identify a single pattern of evolving death and, as he dies, hehas an iphone in his pocket which can identify a song just by listening to it.
We need a new vision and new standards for hospital monitoring which require thatmonitors used on the general floor have a minimum ability to detect, andidentify quantify and track the various patterns of evolving death.
This need is analogous to the past need for other standards (such as in theautomotive and aircraft design) which simulate the real world adverse events(like a dynamic front end skid) and then design systems which timely detect,identify, and quantify various adverse events (like dynamic skids) to allowtimely correction.
Universal skid identification, quantification, and correction is estimated to save over10,000 lives annually. Universal timely Pattern of Evolving Deathidentification and quantification might save 10 times that many annually.
There is no need to mandate how the relational skid is detected, only that this beaccomplished with a minimum sensitivity and specificity by say 2015. As in theauto industry this begins the safety race and the public benefits much soonerthan 2015.
This is a wonderful opportunity and we are standing at a time similar to that forthe early mandating process of automotive safety.
How many young children and babies are alive because those people stepped up againstindustry pressure and against the scientists who were comfortable with thestatus quo and made it happen.
We need the public to know that the space age looking monitors used their familieson the general floor today are the same simple "fire alarm" conceptfrom the 1970s before the PC was introduced.
A 2009 meta analysis published in Oct. indicates (after review of 22,000patients) that the long tried 30 year old idea of threshold alarm does notwork.
Its time for the public to say that since these simplistic devices (designed beforethe PC) do not provide enough information to health care workers, we demandinnovation.
Dr.Larry Lynn- CEO Of Lyntek Medical Technologies, Columbus, Ohio