Wednesday, September 25, 2013

Could the "Good ID" be better?


On Friday, September 20th the FDA announced the release of its Final UDI rule. This marked an important step forward along what has been a long and rocky road towards supplies and medical devices having unique device identifiers. Unlike drugs which have long had a specific, unique NDC (National Drug Code) number which allows for identification and tracking, medical supplies/devices have had no such similar system - with each manufacturer assigning and using its own catalog number we have a situation where a given catalog number may be used by one manufacturer for a wound dressing, by another manufacturer for a urinary catheter, and by yet another for an implantable device...

It has long been recognized that the use of a standard naming convention and unique device identifiers would greatly benefit almost all the parties, from the supply chain (e.g. to simplify and facilitate inventory, recalls, etc.) to providers (e.g. for supporting the tracking devices implanted in patients, etc.) and eventually also patients (e.g. to look up information about the safety or effectiveness of devices).

The FDA Amendments Act of 2007 mandated that the FDA develop a system to implement a unique device identifier for medical devices. This led to a series of workshops, meeting, studies, etc. that in turn led to a proposed UDI rule being published in July 2012 and opened for comment. This eventually led to the issuing of the Final UDI rule in September 2013 (a little later than the mandated time frame of June 2013!)

So, what is the UDI? In the words of the FDA: 

"The Food and Drug Administration (FDA) has released a final rule requiring that most medical devices distributed in the United States carry a unique device identifier, or UDI. It also applies to certain combination products that contain devices and to devices licensed under the Public Health Service (PHS) Act (e.g., donor screening assays). A UDI system has the potential to improve the quality of information in medical device adverse event reports, which will help the FDA identify product problems more quickly, better target recalls and improve patient safety. In developing the proposed UDI system, the FDA worked closely with industry, the clinical community and patient and consumer groups, and conducted four pilot studies. A UDI is a unique numeric or alphanumeric code that consists of two parts:
  • A device identifier (DI), a mandatory, fixed portion of a UDI that identifies the labeler and the specific version or model of a device, and
  • A production identifier (PI), a conditional, variable portion of a UDI that identifies one or more of the following when included on the label of a device:
    • the lot or batch number within which a device was manufactured;
    • the serial number of a specific device;
    • the expiration date of a specific device;
    • the date a specific device was manufactured;
    • the distinct identification code required by §1271.290(c) for a human cell, tissue, or cellular and tissue-based product (HCT/P) regulated as a device."
A schedule with deadlines by which manufacturers have to obtain and label their devices with UDIs has been published, ranging from Class III devices which need to be labeled within a year to seven years when all Class I and unclassified devices need to be labeled with UDIs. There are many exceptions and special cases, for example Class III device makers can apply for a one year extension, Class I devices sold at retail do not require a UDI and may use their existing UPC code, etc.

The FDA's final rule allows for multiple issuing agencies - they will start off with three issuing agencies, GS1, Health Industry Business Communications Council, HIBCC, and ICCBBA, but will consider approving others that apply.


Finally, the benefits of the UDI as articulated by the FDA, include:
  • Allow more accurate reporting, reviewing and analyzing of adverse event reports so that problem devices can be identified and corrected more quickly.
  • Reduce medical errors by enabling health care professionals and others to more rapidly and precisely identify a device and obtain important information concerning the characteristics of the device.
  • Enhance analysis of devices on the market by providing a standard and clear way to document device use in electronic health records, clinical information systems, claim data sources and registries. A more robust postmarket surveillance system can also be leveraged to support premarket approval or clearance of new devices and new uses of currently marketed devices.
  • Provide a standardized identifier that will allow manufacturers, distributors and healthcare facilities to more effectively manage medical device recalls.
  • Provide a foundation for a global, secure distribution chain, helping to address counterfeiting and diversion and prepare for medical emergencies.
  • Lead to the development of a medical device identification system that is recognized around the world. 
But enough about the UDI, this post is about the part that underpins the implementation, the Global UDI Database, GUDID or "Good ID" in FDA parlance... this database will house all assigned UDIs.  The graphic below shows the possible interactions with the GUDID. Vendors will enter their devices and UDI information, either directly via web or electronically using a HL7 SPL transaction via the FDA Electronic Submissions Gateway. It is anticipated that the issuing agencies will develop electronic submission so that when vendors get a UDI (be it a GS1 GTIN or a HIBCC code, etc.) from the issuing agency the information will flow to the GUDID in an automated fashion. Providers, patients, etc. will either access the GUDID via web look-up or via Web services' search and retrieval.


Wisely the FDA is incorporating robust Master Data Management/Data Quality to ensure that the GUDID remains "clean.' The FDA has published a GUDID Draft Guidance, and will take comments through November 25th. So, perusing the draft guidance one has to ask 'Could the "Good ID" be better?'

GUDID "scorecard"
Click to enlarge:

Thinking forward to when all devices have UDIs and providers, patients, etc. are routinely using the GUDID, one would hope that the 'Good ID' would allow you to do certain things, including for example:
  • The ability to see all packaging units of measure in which the base item is available e.g. EA, BX/5, CA/50 (each, box of 5, case of 500, etc.) - this appears to be supported by the GUDID.
  • The ability to get all necessary item attributes from one source of truth. The GUDID is initially mandating a limited number of item attributes (see chart above), for example the presence of latex or natural rubber in a product, its MRI safety status, etc. There are so many other possible attributes that would be useful, it would be nice if the FDA could start off with a much larger initial set, understanding that going back and adding them later will overall end up to be more onerous than including them from the outset. Examples of other useful attributes could include the presence of DEHP, BPA, etc,: safety-related attributes (e.g. hazardous status, flammability, active or passive safety system if it includes 'sharps.' etc.); usage-associated attributes (e.g. the maximum number of permitted reuses for a 'reposable', etc.); environmental-related attributes (e.g. special disposal requirements, if recyclable, if green or containing recycled product, etc.); and so on...
  • The ability to 'trace' predecessor or successor items. The GUDID requires that a device get a new UDI if  the product is changed or if certain attributes change (e.g. if the item goes from containing latex to not, if an item goes from non-sterile to sterile, etc., since such changes  essentially result in a new product). When a new UDI is obtained and the 'old' product is no longer manufactured, the old UDI is "retired" - the 'package Discontinue Date' is entered and the 'Package Status' field value is changed from 'In Commercial Distribution' to 'Not in Commercial Distribution.'  However, if the item is still being manufactured in a new version with a new UDI it does not appear that you can pull up the discontinued item in the GUDID and see the successor (new) product UDI number... Conversely, you apparently cannot go back and see predecessor version UDI.  With changes and improvements occurring at a fairly rapid pace for some devices this would be an important ability, as a provider could have items on its shelf past a 'Discontinue Date'.  Also related to this is the fact that some device manufacturers have a habit of changing their catalog numbers... This blogger is not aware if a good (i.e. "real") business reason exists for such a practice, and like others has wondered if this is being done uniquely to impede the ability of providers or third parties to develop cross-references between 'equivalent' products made by different manufacturers! At the 2013 UDI Conference one manufacturer attendee asked if they were allowed to change the UDIs on items even if they were not modified or new (the answer was 'yes')! 'Tracer' functionality in the GUDID would go a long way to help in ensuring transparency.
  • The ability to either see or link to the device's recall history. The GUDID is designed to only contain Device Identifier information and not Production Identifier information. Given that one of the major benefits of  UDI implementation is to make recalls easier, one would hope that as vendors inform the FDA of recalls (which would include both DI and PI information) the corresponding historical recall PI information can be pulled up via the DI in the GUDID when researching an item. This would be invaluable in multiple instances, for example when a provider is choosing between a number of items (UDIs)  to carry one input could be the items' recall histories; or if a patient is researching potential  implants; etc. It is not clear that this will be possible with the GUDID as currently envisaged.
  • The ability to look up the UDIs of all the items present in a custom pack or convenience kit (for when they are individually available commercially that is...). For convenience kits or custom packs where the constituent items are also in commercial distribution, it would be most beneficial to pull up a kit DI and see the DIs of the constituent items!
From a perusal of the draft guidance it appears that only one of these five is fully supported, while one is partially supported and three are not.  

Additionally, also not clear to this blogger is the robustness of the GUDID. It needs to be robust, massively scalable, and always available. At the 2013 FDA UDI Conference mention was made of the GUDID being down for maintenance on the weekend. This may be OK in its initial phases, but once up and running with providers using it and 'pinging' it on a regular basis, this sounds problematic. A good model to follow could be one conceptually similar to that of the DNS system - if the DNS can support all the addressing needs of the web, with distributed servers and replication of records,  and with the necessary scalability, it would be a good model (note: I'm sure the DNS is a lot more complex than this simplification, and I am no expert on the DNS, but am suggesting an analogous system be considered).

The bottom line? The answer to 'Could the "Good ID" be better?' is 'yes.' While it is ambitious and represents a breakthrough compared to current state, it could and should be even more ambitious and include these and other capabilities!

What do you think? Feel free to leave a comment...

UDI Resources:
FDA UDI page
GS1 and UDI 
HIDA UDI page
Premier UDI page
AHRMM UDI page
AdvaMed UDI page 
A simple guide to UDI
HIBCC UDI Resource page
The Healthcare Hub  (GHX) UDI page
Brookings Roadmap for UDI Implementation

Note; All slides (except the "GUDID scorecard") courtesy of FDA, GS1, HIBCC  handouts from the 2013 FDA UDI Conference.

Sunday, September 15, 2013

Random charts - firearm costs



"In 2010, there were 36,341 ED visits (11.8 visits per 100,000 people) and 25,024 hospitalizations (8.1 visits per 100,000 people) due to firearm assault injuries..."

Monday, September 9, 2013

A stretch...


President Obama, looking for reasons for military action in Syria engages in somewhat of a stretch to tie military action against the Syrian regime for chemical weapons use in Syria to the safety and security of the U.S.

"... Well, what’s happened has been heartbreaking, but when you start talking about chemical weapons in a country that has the largest stockpile of chemical weapons in the world, where over time, their control over chemical weapons may erode, where they’re allied to known terrorist organizations that, in the past, have targeted the United States, then there is a prospect, a possibility, in which chemical weapons that can have devastating effects could be directed at us. And we want to make sure that that does not happen.  

There is a reason why there is an international norm against chemical weapons. There’s a reason why consistently, you know, the rules of war have suggested that the use of chemical weapons violates Geneva Protocols. So they’re different, and we want to make that they are not loose in a way that ultimately, could affect our security.

And if, in fact, we can take limited, tailored approaches, not getting drawn into a long conflict, not a repetition of, you know, Iraq, which I know a lot of people are worried about – but if we are saying in a clear and decisive but very limited way, we send a shot across the bow saying, stop doing this, that can have a positive impact on our national security over the long term, and may have a positive impact on our national security over the long term and may have a positive impact in the sense that chemical weapons are not used again on innocent civilians..."

So,  in effect, to sum up President Obama's position:
  • Syria has the largest chemical weapons stockpile in the world (bad).
  • They have used chemical weapons against civilians (bad).
  • Their control over the stockpile might erode (bad).
  • They're allied with known terrorist organizations that have targeted the U.S. in the past (bad).
  • We want to make sure that the chemical weapons don't get loose, since that could effect our security (bad)
and therefore we should bomb the evil regime (good).

Say what? Yes, this might prevent the further use on chemical weapons on innocent civilians (good), but how exactly does degrading the regime's capabilities make the U.S. safer? Does it not increase the risk that their control over the stockpile will erode (bad); that the chemical weapons might get loose (bad); and that they might get acquired by known terrorist organizations that have targeted the U.S. in the past (bad)

Eh?

Oh well, at least we are OK because the president has "... looked at all the evidence, and we do not believe the opposition possessed nuclear weapons..."

Hmm...


OK, so there has been a lot of talk about 'proof' that Assad gave the order to use sarin on civilians... This blogger observes that much of it seems to be of the 'well it's obvious that only Assad (or the Syrian army) could have done this' type statements, and that the administration seems to labor under the misapprehension that the continued repetition of statements such as this by multiple administration officials in multiple venues adds to the "proof."

However, let's look closely at what has been reported on this topic by segments of the press based on purported statements by German officials. First reports surfaced that the Bundesnachrichtendienst (BND, or German intelligence) had wiretap evidence! As the story goes the Germans had intercepted communications between an operative of Assad ally Hezbollah and Iran, and that "... Assad was desperate to halt the rebel advances in the suburbs of Syria's capital after a mortar attack on the president's motorcade as he traveled to a mosque during Ramadan." As a result Assad  "lost his nerve" and "made a big mistake" by ordering the attack..

Wow, sure sounds definitive! Also, this would seem to answer a question raised by many, asking why Assad would choose to openly use chemical weapons just as U.N. chemical weapons inspectors were in the country... and that a mere 10 minutes drive away from their hotel room! He "lost his nerve" after a mortar attack on his motorcade!

Umm, one small problem with this time line - Ramadan ended on August 7th and the Ghouta attack occurred on August 21st. Not exactly indicative of a "lost his nerve," panicked response... It was, however, a nice touch attributing the "proof" to Hezbollah!

Then soon thereafter different reports were that the BND said that the chemical weapons attacks “might not have been sanctioned by Assad.” 

And this blogger's absolute favorite - an article that has both versions in the exact same article! "... President Bashar al-Assad did not personally order last month's chemical weapons attack near Damascus that has triggered calls for US military intervention, and blocked numerous requests from his military commanders to use chemical weapons against regime opponents in recent months, a German newspaper has reported , citing unidentified, high-level national security sources....  

Schindler said that BND had intercepted a telephone call in which a high-ranking member of Hezbollah in Lebanon told the Iranian embassy in Damascus that Assad had made a big mistake when he gave the order to use the chemicals, the magazine Der Spiegel said."

Wow, just wow.

Syria Crisis: Hezbollah Wiretaps Reveal Assad Ordered Sarin Attacks after 'Losing his Nerve'
Gas Attack: Germany Offers Clue in Search for Truth in Syria
German intelligence concludes sarin gas used on Assad’s orders – reports


Syria Chemical Weapons Attack ‘Not Ordered’ by Assad – Report

Assad did not order Syria chemical weapons attack, says German press

Wednesday, September 4, 2013

The cabal?


A recent NYT oped "How a Cabal Keeps Generics Scarce" again blames drug shortages on GPOs (Group Purchasing Organizations)... "The economic root cause is simple: the purchasing organizations have squeezed manufacturers’ operating margins to razor-thin levels. By awarding select suppliers exclusive contracts in return for exorbitant (and undisclosed) “administrative,” marketing and other fees, they have reduced the number of suppliers to just one or two for many generics..."

An earlier blog entry had poured cold water on the position of PADS (Physicians Against Drug Shortages) on this topic. One has to wonder if the NYT so hard up for content that they need to provide a platform for these conspiracy theorists?  Faugh! Replace "cabal" with "shibboleth", because that's what we have here - a common saying or belief with little current meaning or truth.