Archive for the ‘Repetitive Failure’ Category

When does a meeting end?

Sunday, November 30th, 2008

When, exactly does a meeting end?

Is it when everybody has left the meeting room? When you leave the room? Or, is it when the time allocated has been exceeded?

In some cases and with some people this is when they’ve lost interest in the proceedings and they tune out and think of other things.

My experience is that it typically ends when the attendees leave and return to their desks or proceed to the next meeting.

Anyone out there got any ideas they’d like to add?

Post a comment, let’s see what comes out of this!

AA Worst site in New Zealand?

Friday, October 31st, 2008

Yet again, I tried to use the AA Site. For years it has been the most useless site in the entire .nz domain. Even if you have logged in, it “forgets” who you are, depending on which page you navigate to. Today, I logged in, then went to renew my membership. It asked me for my details - even though they are all contained within the site!!! Another thing, they used to have a so-called petrol watch page. I wrote to them about a year ago and told them their prices were about two months out of date. I never receved a reply, of course. I now see that they do not have this page anymore. Or, if they do, it’s too hard to find, as I’ve just spent ten minutes searching.
If only there was some competition, like the UK. That would make them sort their ideas out. Time for a beer.

Will these changes prevent a repetition?

Sunday, June 15th, 2008

In my previous post, about the preventable perinatal death I had missed a report on the Stuff website titled “Hospital procedures to change after mother dies“.

According to this report

An internal investigation by Auckland’s District Health Board (ADHB) was launched after Ms Bayliss’ death, and several procedural changes have followed including a new process for obtaining emergency blood and checking all pregnant women for risk of postpartum haemorrhage.

Assuming this was reported correctly, can we be confident that those reported changes will prevent problems in the future and if we look retrospectively would they have resulted in a different outcome?

Firstly, a disclaimer - we’re working off newspaper reports 5 months after the event - but that’s all we have to work with. The report stated that there were “several procedure changes” but nominated only two - we don’t know what the others were, but presumably the most important ones were reported.

A new process for obtaining emergency blood

What “emergency blood” means I don’t know. If it means the putting aside of blood in case of emergency so it is easily and quickly available when required then yes that will be beneficial.

Checking all pregnant women for risk of postpartum haemorrhage

On the face of it this sounds good - but isn’t it happening already? Certainly in the case of Renee Bayliss it would make no difference as she had been checked (on December 31st) and the risk had been identified.

This last recommendation sounds like a typical shallow response to failure. The only way to really address repetitive failure is with rigorous root cause analysis. When a proper root cause analysis is conducted everything related is put on the table as a possible cause. In our experience, whatever is first considered to be the real reason for failure is never the reason.

Given the identified risk that “Her placenta was also found to be at risk of causing problems, including bleeding following birth” did the investigation include analysis of the clinical decision to choose induction instead of Caesarean? Maybe, I’m hypothesising, that was the root cause and all the rest were contributing factors - I don’t know, but to prevent the preventable all options need to be covered.

Prevention of repetitive failure requires more than “investigation” it requires application of a rigorous process - otherwise it just a …

Confident it was an isolated case

Saturday, June 14th, 2008

The Chief Medical Officer at Auckland Hospital claimed a case of preventable death had a combination of circumstances that was unusual and he was confident it was an isolated case. Sounds like code for “it won’t happen again because it’s very unlikely”. That’s not an effective way to deal with repetitive failure - and preventable hospital death in New Zealand is definitely a case of repetitive failure, the national solution to which has been to take seven years to produce an incomplete report.

Auckland City Hospital admitted that the death of Renee Bayliss on January 26 withn 4 1/2 hours of the birth of daughter, who survived, was preventable. The report in the NZ Herald stated, “A first-time mother died from severe bleeding during childbirth after blood was not put aside for her, despite her specific request.”

The most galling aspect of this case was that the risk of haemorrhage was recognised before the commencement of induction and Ms Bayliss had requested cross-matched blood be put aside. That request was made on January 24: it wasn’t needed until January 26 but when it was needed it wasn’t there.

Key issues in care were identified as:

* Why she was not quickly given a transfusion of “universal donor blood”, whose higher risks are considered justifiable, rather than having to wait for blood cross-matched because of particular antibodies that had been injected before the birth. The anti-D antibodies are given if there is a risk of the mother’s immune system attacking the fetus’ blood cells.

* Whether a hysterectomy, to stop the bleeding, should have been done sooner.

The problem with these “key issues” is that they deal with conditions that happened as a consequence of earlier treatment - they do need to be addressed - but the real issues, in my view, are:

  • Number one issue - why was cross-matched blood not delivered until 6:17am on the 26th? (over a 32-hour interval)
  • Why did the patient have to request for blood to be put aside? Isn’t that what doctors are for?
  • On December 31st, “Her placenta was also found to be at risk of causing problems, including bleeding following birth.” Why did this information, of identified higher risk, not ensure that blood was ready and available?
  • Given the known risk of haemorrhage, why was induction started before securing the matched blood? More than three weeks had elapsed since original hospital diagnosis, so it shouldn’t have been an urgent induction.
  • “Although a registrar had asked at 4.15am for cross-matched blood to be provided, the request was not carried out.” Why not?
  • “The blood bank was not told a cross-match of blood was required until 5.20am.” What is happening here? Is there a communication problem with the blood bank? Not forgetting that they were advised of this same need more than 24 hours before.

The hospital has commissioned an external review. Hopefully that will arrive at uncovering the root causes of failure (tough to do after this much delay) - otherwise we will see this kind of preventable death recurring. An external review is better than nothing - but much better would be to have that kind of review happen automatically within hours of the event to ensure that all relevant facts can be gathered while they are available and memories are fresh.

Hospitals should have automatic review and root cause analysis processes built in when serious incidents occur - otherwise combinations of unusual circumstances and “isolated” cases of preventable death will continue to happen.

The very least we can do is to ensure that a preventable death prevents another.

When does a meeting start?

Wednesday, May 21st, 2008

Not a silly question, really - especially if you’re footing the bill.

When does a meeting start? Is it when everyone has finally arrived in the meeting room? Or perhaps when the group decide to start proceedings, even though one or more of the participants haven’t arrived yet?

The Action Meetings approach to more effective meetings is that a meeting starts when the need for it is identified. From that point on time and money are expended. There is much to do:

  • Who will be there and will add value (not just make up the numbers)?
  • What is the meetings purpose - in other words what outcomes must be achieved?
  • How long should it take? The default time seems to be an hour irrespective of whether an hour is the right length of time, too long or too short!
  • What about the venue?
  • What pre-meeting preparation must occur in order to achieve the desired outcomes?

The list above is not exhaustive - you may well be able to add to it - but it illustrates that a meeting starts well before the participants to get to eyeball one another.

Action Meetings is a method for designing and delivering effective meetings - the end-to-end meeting process not just the time spent in the room together.

Police attack the cause of the problem

Tuesday, May 20th, 2008

Unlike the scatter gun approach to “speed” being the cause of many accidents, it is welcome to see the Police in Christchurch taking a specific remedy to a specifically defined problem and cause - tailgaters causing crashes.

Sergeant John Hamilton of Christchurch Police is quoted:

“We are going where crashes are and the motorway is a grade-two crash area (grade one being the worst).

“We’re doing it as a lead-up to wet weather. When we get rain, that’s when crashes occur.”

As regular readers of this blog will know, repetitive failure is effectively addressed when the specific cause of failure is identified and addressed. Anything else may look like you’re fixing it but it’s really a WOMBAT.

Australian drivers put schoolchildren at risk

Monday, May 19th, 2008

Recent article in The Wellingtonian (May 15) bemoans the speed at which kiwi drivers go past schools. The New Zealand Police “Speed Kills Kids” campaign has stated that if you go more than 5km over the limit within 250m of a school you’ll be ticketed. This year they have caught over 9000 motorists going over 54km/h. The article cites some schools where the rate is 13 vehicles per 100 and 8 vehicles per 100 over the limit.

Across the ditch, Australian drivers legally drive past their schools at 60km/h because that is the standard urban speed limit. If the local publicity and statistics used to base this campaign are to be believed, it must be carnage over there.

If these campaigns are to be really effective they have to be credible. A person of close acquaintance was ticketed for driving more than 5km over the limit going past a school. The rural school was on a road with a 70km/h limit (how dangerous is that?) and the ticket was given in mid-January (school holidays).

Waste Management manage to waste my time

Wednesday, April 23rd, 2008

Ordering a rubbish skip should be a simple no-brainer. Unless of course the company lures you in by offering a special deal and then fails to deliver on its promise.

The ad on the radio sounded promising - order a skip for Anzac weekend and save $27 on usual price with Voucher Mate. I needed one anyway, so it was perfect (I thought). First step go to the website and order it there - enter all my details and go to the pricing page - no discount or facility to get my voucher discount. No worries, I’ll call them directly and it’ll all be sweet.

Firstly, the call quality was rubbish (appropriate), which didn’t help - I could barely hear them. The words “voucher mate” were either comprehended as Kitumbatu Swahili or, perhaps, familiar Kiwi “voucher - mate”. Whatever - they had no clue what I was talking about - so put me on hold and then dropped the call. I tried again - more time on hold - “could they take my details and ring me back tomorrow?” By that stage I’d had enough- called Skip Bins instead - in 2 minutes all sorted, no fuss and only eleven bucks more than “the deal”.

When a company’s processes aren’t coordinated between its marketing arms and customer channels it winds up creating customer dissatisfaction. I’d been a Skip Bins customer before and was willing to switch on “the deal” - but they won’t get me again.

Finally, did anyone think of testing? An employee could have acted as a “mystery shopper” and tried to order a bin with a voucher discount. That would have cost all of five minutes to ensure the promotion was working.

Accidental hospital deaths plummet from estimated 750 to 40 actual

Monday, February 25th, 2008

This is either a remarkable improvement or the figures are understated. To say they are understated is itself an understatement: we’re talking about more than a tenfold understatement. National party health spokesman Tony Ryall said “the committee’s figures vastly understated the size of the problem”. He’s not wrong.
The NZ Herald article states,

“Committee member Dr Mary Seddon, the clinical director of the Counties Manukau board’s quality improvement unit, said there were more preventable incidents than had been included in the report.”

I’d expect that “more” to be maybe 10-20% more. What was the real number? Will we ever find out?

A May 2007 report into accidental hospital death (The Press, Christchurch) put the figure at an estimated 750 each year (see our original blog post) but the recent report released stated only 40 deaths in the past financial year.

40 preventable deaths from 834,000 admissions is a fantastically good result - a pity it’s not the real number.

How about we take a look at the report? Where do we start - Ministry of Health or the Health and Disability Commissioner?

Let’s try the HDC first - go to Publications - then Other Reports where we find the Safety of Patients in New Zealand Hospitals: A Progress Report known here as the “plethora of policies” report. OK - that was October 2007 - nothing since then.

Must be on the Ministry of Health website - go to Publications and Resources and Find by Date. Not there! OK - so since we’re there let’s have a look at Quality Improvement since it’s a hot topic. Quality Improvement - Publications - nothing since 2004 (could this have anything to do with “slow and patchy” progress?) However, under Toward Clinical Excellence is a page Toward Clinical Excellence: Learning from Experience. This page tells us that in September 2001 -

“The Sentinel Events Project Working Party members have been brought together from throughout the health sector to make recommendations to the Director-General of Health on the feasibility of implementing a mandatory event reporting system for health and disability services and related matters.”

The page links to their report that includes a number of recommendations. Here is Recommendation 3:

“Implement a national system (to be called the Sentinel Events Reporting System) that requires health and disability services to report a defined list of events (to be called Sentinel Events) to the Ministry of Health for review.”

The recommendation was made in September 2001. The initial report was published in February 2008. The data it published is neither accurate nor standard. Meantime, preventable death and harm occur. That’s a WOMBAT.

Seven years to get an incomplete report - how long to get some effective interventions?

And where do you find the Commentary On Sentinel & Serious Events Reported By District Health Boards - 2006/07 report? On the site of the Quality Improvement Committee - a sub-site within the Ministry of Health. Don’t bother trying to find it easily from the Ministry of Health website (it’s buried in there somewhere) - it also needs some quality improvement.

“Slow and patchy” progress on the $1.8bn hospital death problem

Tuesday, February 19th, 2008

750 people unnecessarily die each year in New Zealand hospitals - that’s almost twice the annual road toll. Fifteen people per week are dying and using the “statistical value of life” of $2.5m per person that’s a $1.8bn problem - as we noted last May.

The road toll gets huge focus and attention with advertising and drink-drive campaigns to bring this horrendous figure down - but meantime this problem of almost twice the proportion seems to be mired in bureaucracy. The finger is being pointed at regional inconsistency and a lack of information sharing and the notion of safety “league tables” has been mooted.

“League tables” are appealing as they are intended to provide some competition and aspiration to lift levels of performance. However, what often happens is that they lead to divisive competition and “burying” of data that will cast a bad light. There is often an uncomfortable paradox in that we do want to embrace accountability but at the same time be careful not to produce or develop blame cultures as these are ones in which morale degrades and no quality improvement is possible.

The CATS process that we subscribe to and practise is only successful in an open environment where blame is not appointed to individuals and points of failure are assigned to missing or failing processes. A complete repair to a process results in a permanent change that removes that point of failure. Getting rid of an individual just leaves the replacement exposed to the same risk of failure and blame the next time.

Accountability goes to the process owner - it is up to them to be managing robust processes.

We don’t know the real reason why progress is slow and patchy but we do know that each unnecessary death could teach us how to prevent another.