Human Factors and Safety and Quality (?)

In 1986 Briggs Phillips published a small, probably near forgotten, but important study in laboratory safety (see in Laboratory Safety: Principles and Practices.  ASM Press 1986).  The author was able to identify two groups of people: one group of 33 who had experienced at least one laboratory accident in the last 2 years, and another group of 33 who had been accident free during the same 2 years.   The two groups were matched by job classification.    

The two groups were interviewed and surveyed for additional information.   

While not matched for the following characteristics, all including age, years of education, years on the job, weight, height and gender and marriage status were identical.  In addition, wearing glasses, medication use, showed no differences.  Also of interest, accidents outside the laboratory, and driving records were also equal.  

The factors that had significant differences were (A) smoking  (B) the number of accidents and injuries and infections acquired in the 2 years prior to the current 2 years (C) personal attitude towards safety (D) opinions of the safety awareness of their supervisors and co-workers, and (E) importance of training and communication versus personal experience.  In all these, the accident involved were more likely smokers (remember this was 25 years ago), had a long history of accidents,  believed that technique was more important than attitude, but thought that personal experience was more important than training.   

The accident free group were significantly more critical of the supervisors and significantly more critical of their co-workers safety consciousness, and believed that safety attitude and awareness were more important than having techniques and equipment.  

[If you are safety aware you will avoid getting into the high risk behaviour of rushing or allowing distraction.  If you are not safety aware, you are more likely to assume that equipment will allow you to go faster and increase risk.]

The message of the study points to importance of attitude and awareness in the avoidance of having a laboratory accident or injury.  

So why am I bringing up 25 year old studies?  Two reasons.

First, a recent study in ASQ’s Quality Management Journal V18, N:1, 2011 on factors associated with continuous  improvement in patient safety  by E. Naveh et al came to a similar finding of the relative importance of opinions on attitude and priority.

Secondly, and more importantly, what made the first study possible was that records were maintained on who had been involved in a laboratory accident or injury.  In many laboratories today, we have virtually no records of Quality accidents, and where we do have records,  the names of the people are commonly deleted.  So in many (most?) laboratories if we wanted to study the human characteristics associated with increased error, it would be impossible because we have no place to even begin.

Deming wisely called to “Drive out Fear” and I agree with that.  He also said that most error was systemic, and I agree with that too.  But Quality in the medical laboratory is a new field, and it is important that we study why errors occur.  Confidentiality is important, and so if avoidance of a sense of retribution.  But let’s not “throw the baby out with the bathwater”.  Continuous improvement requires having data to study.

That’s why they call it PDSA.

Quality and Obama Care.

Welcome to 2011.  

I suspect that most folks missed an interesting story that was published on the Times wire in US newspapers just before Christmas.  The article described how new regulation set to take effect in January will direct a substantial portion of insurance premium dollars away from executive pay packages and overhead and instead will require increased support for medical care and health care quality improvement.  Apparently insurance companies now can retain 25 percent of premium revenue for non-medical costs, but under the new rules they will be able to retain only 15 percent.  

Well that sounds good, especially the part about increased spending on quality improvement. 

I suspect that the way this is will play out will be that insurance companies will not reduce the actual amount of money for salaries and bonuses, but rather will just increase the premiums.  If I do the arithmetic correctly, that will require an increase in premium rates by about 66%.  Again if my arithmetic is right, this will not go over very well.  A massive raise in health insurance rates may not be perceived as a totally positive step forward.  But we will leave that battle to someone else.  

But I think there is an important question here, and that is, is there is a way in which money collected through the private sector can be allocated to quality improvement with a reasonable expectation of success.  My guess is that this will be a major challenge.

Let’s see if we can break this down.
1: If money collected by individual insurers remains in the hands of the insurer to do good things, there is little  reason to believe that each insurance company would develop an effective distribution process to move funds into the appropriate sectors to improve clinical and laboratory care.  Further, development of distribution offices within each insurance company would require massive costs and result in huge redundancy.  As US insurers do not cross state borders, this would result in a huge amount of geographic mal-distribution of funds and resources.

2: If each insurer was required to transfer the funds to a central government office, it would likely be perceived as a new tax, and would probably get tied up in a series of court cases ending at the Supreme Court.  That would take years and cost a fortune resulting in quality improvement money mainly support legal staff.

3: If each insurer was required to pass funds to a private sector central Quality Improvement Office operated by the consortium of insurers, the costs of redundancy would go down, but process and distribution costs would go up.  In addition there would still need to be an active oversight agency to establish and monitor rules, and that too would create even more costs.

So while the idea sounds great, my cynical self sees more opportunities for failure than success.  It will be interesting to watch to see what happens.  But I am not holding my breath.

In the meantime, in Canada and other countries where this is a public health funding model, there is every reason that agencies be designed for Quality Initiative Allocations with independent oversight.  In Canada we would still end up with redundant spending because of our obsolete structure makes health a provincial responsibility with trivial federal funding support.

The reality is quality initiatives programs would create net savings not net losses, so there should still be an incentive to implement these agencies.  So far the answer is that they do not exist.

With so much to gain, one wonders why.

m

Patience and Patients Safety.

I was reading my local newspaper and found a columnist discussing an article on Patient Safety and Hospital Error.  We can no longer be surprised that the things we do have become front and center in the cross-hairs of the media.  We have brought this on ourselves.  

The column reported on an article in a recent edition of the New England Journal of Medicine (how scary is it that a local paper columnist is reading articles in the NEJM!), which looked at the rates of “preventable harms” in 10 hospitals after 5 years of active process (2002 – 2007).  While there was a reduction (approximately 1% per 1000 patient days) the reduction was not statistically different, nor (my opinion) clinically significant, even if it was real.  This was true for each category from inconvenience to death.  This was despite active engagement in national and state patient safety training programs. The analysis was done by thorough and fair methodology with internal and external reviewers and appropriate reviewer comparisons.

The authors commented that harm from medical care remains high despite all the programs and the money being spent.  

 The study confirms reports from across the US and Europe about how little progress is being made.  The amount of evidence-based error reduction practices implemented so far has been modest.  All the things that can be done, including substantial improvement on handwashing (!), have not be implemented with any consistency or success.

A few thoughts. 

Systemic and personal behavior does not appear to be easy to change.  While there are folks within health care who are motivated for change, they are not having a lot of success, when measured by outcome.  Most of the money being spent on patient safety teams, training seminars, poster displays, and conferences is being wasted (unless your personal income comes from being on a team or charging for putting on training seminars).  Change is not happening.

My concern is that if healthcare management and healthcare unions wanted change, then change would happen.  I could suggest that there are other vested interests at play, but that would be unfair.  But if neither altruism nor fear of malpractice litigation is insufficient to drive change, then what does it take?

At some point the media will care enough and the public will care enough and the politicians will care enough and we will end up with a healthcare version of the Transport Security Administration and we will have our own version of an intrusive airport pat-down.  And folks will say, well it may not be fun, but it makes our lives more secure.  

For example, the regulators, and legislators always have lots of options, like perhaps a variation on pay-for-performance.  Rather than institutions getting a reimbursement bonus for good deeds, every year that they miss their goals, they lose 1%.  And insurance providers can start jacking up institution protection rates.  

“They” say that carrots work better than sticks, but sticks work.

From my vantage point, hopefully medical laboratories might be having some more success in error reduction because as much as we are a complex distributed activity, our activities are more focussed than hospital admissions.  Some centers have demonstrated some levels of reduced error, at least on a short term basis.  But we need to have an institution perform the longer term year-over-year study like the one above to observe if trends can be seen as improving.  I suspect that most are afraid to look.   

The media already cares, and the public is become more aware.  

Strike 2.

 m

PS.: The article is “Temporal Trends in Rates of Patient Harm Resulting from Medical Care” by CP Landrigan et al.  N Engl J Med 2010;363:2124-34.

Heresy?

 American Society for Quality has a number of  journals, and some excellent, and others pretty good.  One of the latter category, at least in my opinion, is the Journal for Quality and Participation.  Without wanting to be harsh, I tend to find articles, while generally interesting,  tending to be more  opinion than fact.  (The irony of me, an opinion oriented blog writer making this distinction is not lost!).

Nonetheless there were three interesting articles in the October 2010 edition that I received today.  One was “Improving Project Performance with Three Essential Pieces of Information” by Portnoy, and another “Creating a Self-Confident Workforce” by Denton.   The last one is  "Training on Trial” by Kirkpatrick and Kirkpatrick.  

I don’t intend to go into any (the journal is available at www.asq.org/pub/jap) in detail other than to say that the first article made the point that when designing a project briefing (the deliverable) it is useful to make it brief and unambiguous.  The point was that the more the number of words and the more jargon included, the more variable is the document's interpretation.  The training article was making the point that training can have a lot of challenges demonstrating that it actually provides a service that will address significant business results.  And the third article made the point that workers are more self-confident with less stress when they are empowered to make certain decisions on their own.

I can support all those points.

Which brings me to the heresy. 

At a laboratory where I was working, I often found myself in conversations that suggested that some of the standard operating procedures (SOPs) were so detailed and so “precise” that in my opinion, they were largely unfollowable.  Even with an adjoining process map they were unfollowable.  They were better when pictures were added in,  but especially better when pictures were used in place of words.  In the process of trying to make SOPs that were all encompassing, I felt we were laying the foundations for error.

And so I started to think that maybe it is not so important to tell microbiology technologists how they have to hold and streak a petri dish, but to let that happen on its own.  And defining precise colony counting methods was so rigid that it likely wasn’t followed anyways.  And trying to define all the combinations and permutations of bacterial growth was confusing.  

Now there are many procedures involving many pieces of highly precise equipment that do need precise instruction (I understand that) but if we pollute those instructions that are challenging to follow, I think we run the risk of some documents  that need attention and clarity getting lost in the shuffle.  

So my point is that SOP writers and trainers and supervisors need to take a closer look at the procedures that they create to make sure that they actually are useful for training, and more importantly allow the business purpose of the procedures to come through.  Its not only about the value stream, but also creating documents that give technologists the professional autonomy.  Smaller and selective documents, it  seems to me, make the procedure and much of the decision making process both more efficient and more effective.

And how heretical is that?

m