Featured Post

Healthcare Customer Satisfaction: More Talk AND More Action

Healthcare Customer Satisfaction: More Talk AND More Action Customer satisfaction (Voice of the customer) is a recurrent th...

Saturday, September 25, 2010

Medical Laboratory Quality and Risk

From the beginning of classification, there have always been lumpers and splitters.  There are some who think that Quality Management, Risk Management, Financial Management, Safety Management, Personnel Management (Human Resources),  are all the same, just Management. 
I don't. 
But I do try to find the interfaces.


Laboratorians get Quality Management.  We may not all do it, but we understand why it exists, and work towards it, each in our own fashion.  What Qualitologists try to do is smooth that process out.  
And I think that laboratorians have a sense of Risk Management, although I think we see risk in three lumps: risk tolerance, risk avoidance and risk aversion.  What we tend not to see is where and how managing risk fits into the the process of managing quality.


There is a small, poorly known, ISO document entitled ISO/TS 22367:2008,Medical laboratories -- Reduction of error through risk management and continual improvement.  The TS in the title means it is a "technical specification" rather than a standard.  (Think of TS as "guideline").  But it is very helpful, because it links quality management and risk management together at the point of preventive action.  The point is that avoiding risk is something that can and should be done before the bad thing happens.  And using Risk Management tools like Failure Mode Effects Analysis can be useful.  The most common situation in the medical laboratory where FMEA comes into play should be before you purchase the new equipment, before you make an overhaul of management structure, before you introduce that LEAN restructuring.  Doing a swap out can be costly of TIME, EFFORT, ENERGY  and MONEY, so before you expend it ask "if this goes wrong, why is it likely to happen, and what can I change now to prevent it".  It is stunning how often we see changes that go in, and later after all sorts of problems have been discovered, it has to get pulled out.  And I know that you have seen it too.  

Now we all know the Rumsfeld classification of known problems, known unknowns, and unknown unknowns, and the unknown unknowns will always bite you in the neck.  But there are so many problems we can prevent by a little risk management FMEA before hand.

But there is another Risk Tool that I was introduced to a few years ago by an excellent Project Manager - the likelihood-outcome table.  It is a valuable part of risk and equally part of priorities planning.  By sitting back and looking objectively at a series of choices or problems to address, as yourself, "when these go south, what happens?  How often do they occur and how bad is the outcome?"  Those things that occur rarely but with a fatal outcome have to come before the event that occurs commonly but causes some maybe some annoyance or inconvenience.   

Below is a theoretical table, in which I structured likelihood of occurrence from remote to frequent and outcome from trivial to critical.  We can put a value on those from experience, or literature, or estimation.  In this table any L/O product greater  than 0.2 has to have high priority.  Any with an L/O less than 0.1 can have a pass.  And those in between need to be addressed, but after the high priority ones are taken care of.  
From personal experience, it is a very helpful way of looking at projects and getting things done.



There are a number of courses on the web addressing risk management.  From what I have seen, they all cover similar concepts, but look at the one on www.westgard.com.  As far as I can see, it is the only one that includes ISO/TS 22367:2008. 



1 comment:

  1. I love the concept of the L/O table. It's very similar to most methods of scoring failure modes in FMEA. I developed a web-based FMEA system that hospitals in the US and Canada are using with success. It uses a 3-part method to assign scores to failure modes: Likelihood (1-5); Severity (1-5); and Detectability (1-4). This way, the maximum score any failure mode can have is equal to "100". If anyone is interested in seeing the system, let me know...
    Tom

    ReplyDelete