The Pizza Party
In college, it seemed pizza could fix almost anything: hunger, stress, hangovers…But as magical as pizza is, it really cannot fix everything; certainly not laboratory defects. Let me explain…
The hospital group I was working with had a long running issue with lab defects; mainly bad blood samples. There seemed to be a lot of problems with those little test tubes the used to collect your blood: incorrect labels, unreadable labels, not enough blood in the tube, and more. All of this wasn’t new. When I got there I found a very pretty report in the back of a desk drawer. It was 8 pages of narrative, Excel? generated charts, and data about the overall defects found in the lab. I read through it all and got to the recommendations. There were two:
It was quite curious that the report was missing a few things I would expect:
But there were pretty color charts of defects by department! And a clip art example of a Pizza Party Poster. And that was the substantive part of the report, there were 5 pages of discussion and references as to why lab defects are so bad. Which could have been condensed to about 5 bullet points. And in the end it called for pizza parties to encourage attention to detail, which, evidently had not had any impact.
Shortly after joining the hospital group I taught some Lean classes, which included a section on generating improvement project ideas (As I described in USING CPI TO IMPROVE CPI). Quite a few of the students came up with projects around Lab Defects, so many that I figured it might be worthwhile exploring it further.
A group of stakeholders gathered over coffee: the Lab Director, 2 Charge Nurses, and one Provider. As we talked it was apparent that lab defects were still a plague at the hospital. The Lab tracked overall defects and had them at 3 percent. There was no data collection plan in place to futher break out the numbers, and so the range across the departments was unknown, but the stakeholders all agreed that critical high traffic departments suffered the worst: Emergency Department and Intensive Care Unit. And these were the departments that needs accuracy the most.
We started an A3 (my favorite) and realized we need more recent detailed data. We did a quick high-level map of the lab sample process. All samples came through the Lab Collection team, and as they did the Quality checks, it was the perfect point to collect data. We gave them tally sheet and within two weeks had large dataset. The main defects found in the samples included:
Now we were cooking with gas! We had problem statement and current data, and we then set a simple goal 10% reduction to start. We then dove into root cause analysis, and found some very interesting things.
Incomplete fills, tubes not properly mixed, and hemolyzed samples were all largely results of improper collection techniques. Something was going wrong when the sample was being taken. Digging deeper we were analyze which type of employees did specific samples. We found that the phlebotomists from the lab that went onto the floors had the lowest rate of defects. Registered Nurses (RN) had a distant second error rate, and the Licensed Nurse Practitioners (LPNs) had the worst. This had us develop an assumption that there was a positive correlation with training/experience had blood sample defects.
This came at a fortuitous time. Recently there had been some reorganization. The Hospital had eliminated the team that went around inserting PICCs, Ports, and Central Lines. These are the specaility access lines inserted into patients. These travelling teams often helped other tasks as well, like difficult IVs, blood draws, and they acted like a travelling training team. Providing mentoring an all things. The abolishment of this travelling IV team gutted an informal training network. When we pulled past data, and lined it up against the reorganizational timeline, we saw a correlation between the elimination of the IV teams and increase in Blood Sample defects.
Regarding the labels, we discovered root causes by observations. Following staff on the floor, watching them take blood samples, we documented their process. About mid morning, the floor staff would do morning tests. They’d review all the orders, gather up supplies, and then print the orders and the accompanying labels. These stacks of papers and labels were usually stuck along side of the plastic bin used to carry the supplies. The staff would go to a room; read an order, take the appropriate samples, labelling them as they went along. If convenient, afterwards they’d send the sample and the paper order to the lab via pneumatic tube. If the nurse didn’t have time, or not near a tube station, the sample would be stuck in the bin and they’d be off to the next room. Then a bunch of samples would be sent after all the rooms were done.
Reading the process above, you may have figured out what we discovered. The loose sheets of paper orders and labels were easily shuffled about and mixed. This is where the errors were happening. It was even worse in the high activity units such as the ED and ICU. Orders and labels were constantly being printed at the main nursing station, and left in printer tray until somebody could grab them. When there was a lot of action, various orders were getting printed and mingled.
The helpful hints on the Pizza Party Posters said that all samples should be double checked… another set of eyes looking at the labels and papers before going into the tube. But given the often short staff and chaotic environment, this didn’t happen. And so errors were made. Usually the concept of batch processing caused a waste of time, but here it was causing waste of defects.
As our little team of stakeholders worked through the steps of the A3, we got to developing countermeasures and creating action plans. And this is where things got a bit interesting.
Two big ideas were:
领英推荐
The team was excited to try things out. We contacted our equipment vendors and they agree to let us use 4 of the hand held scanners on a trial basis (they were confident we’d love them and order some).
But here things got interesting. Up to now, this project had been an offshoot of a Lean Six Sigma class. It didn’t have any formal authority beyond that.
We needed leadership approval to bring in new equipment, even for a trial basis. We were given 5 minutes at an upcoming Senior Leadership Team (SLT) meeting. The SLT consisted of the Chief Operating Officer (COO, The Chief Medical Officer (CMO), and the Chief Nursing Officer (CNO). The team gave a short compelling presentation that covered our work up to this point and the desire to improve training and try the handheld scanners.
The COO approved.
The CMO was on board.
The CNO said No.
The CNO felt the nurses were already overwhelmed and could not absorb more training. We explained it would not be so much formal training that took time away from work, but more one on one mentoring in the course of regular work. But she still said no.
The CNO did not think the nurses could learn new equipment, nor did she think we would have the budget to buy new scanners anyways. We highlighted the easy to use interface of the product, and, since it came from existing vendor, would not require any IT support to use. We also showed the relatively low cost, and the likely return by cost reductions as defects decreased. But she wasn’t convinced.
The CNO was convinced it was just paying attention to detail. If we stressed that, and rewarded that, the defects would drop. It was then that I realized she was the one who created the Pizza Parties.
Despite her admant protests, it was two against one, and we got approvals. We’d been in contact with the training division, and they’d had come up with a plan similar to ours, so that was started up right away.
Our scanner idea hit a snag. The units for the trial arrived. We had them for 4 weeks, then they had to be returned. Since they were from same brand, and same vendor as other equipment being used in the hospital, they were plug-and-play; turn them on and they’d auto set-up and be good to go. But got the units to IT to be put on the asset list, we found out that new equipment, ANY new equipment, had to be validated for use in the hospital. The equipment had to checked to see if it interfered with other equipment or posed any security risk. There were no exceptions. And with the current backlog, it would take 6 weeks…The best laid plans of mice and men.
And so time marched on…
We had 4 weeks to at least demonstrate the scanners: anyone who saw them, loved them. The training was a success and actually helped with another project we had going on to reduce Central Line Acquired Blood Stream Infections (CLABSIs), but more about that in another article.
More reorganizations hit the hospital. The CMO left, leaving a power vacuum in the Senior Leadership Team, and allowed the CNO to block the implementation of the hand held scanners. And we lost ability to fight back. The Lab Director left for another lab. I was transferred to support another Hospital.
But there was some success. Defects took a little dip with the training.?And there was still the occasional pizza party.
Ever been in an organization that did pizza parties instead of action? Would love to hear about them in the comments.