Category Archives: Quality Improvement and Patient Safety

VA Report: Carbamazepine, HLA-B*1502, and SJS/TEN!

1) There is a strong association between the presence of HLA-B1502 and carbamazepine-induced SJS/TEN in Han-Chinese, Thai, and Malaysian populations.  

2) The FDA has recommended screening patients of Asian ancestry for the HLA-B1502 allele before starting carbamazepine.

3) Overall, carbamazepine is the most common cause of SJS-TEN!

Tangamornsuksan et al. Relationship between the HLA-B*1502 allele and carbamazepine-induced Stevens-Johnson Syndrome and toxic epidermal necrolysis. JAMA Dermatology. 2013;149(9):1025-1032.


Also, take a look at this figure for a great DDx of short-lasting headaches!

Short lasting HAMatharu MS, Goadsby PJ. Trigeminal autonomic cephalgias. J Neurol Neurosurg Psychiatry. 2002;72(Suppl II):ii19-ii26.

Evernote Link!



VA Quality Improvement Project #6

On 1/12/16 after a long hiatus for the holidays our group met again to discuss our progress thus far and to plan out our next PDSA cycle.

I won’t summarize all of the learnings from the last PDSA cycle, but highlight a few key issues that we identified as major obstacles that needed addressing to maximize our potential for success.

#) Unclear Front Line Buy-in:  Critical to behavior change is involving the front-line staff in developing the plan for the intervention.  Not only do these front line workers have the expertise about the nitty gritty involved in changing processes, but they also have the most at stake and need to feel like their opinion is valued. Early in the project we had tried to get LVNs to attend our meetings and work with the team, but this was quite difficult due to scheduling restraints and likely very different working cultures.  Nevertheless, given how important this step is, we decided to redouble our efforts to meet with the LVNs, explain the rationale behind the project, and incorporate all of their feedback into the process.  We also discussed giving them as much positive feedback and recognition as possible to help them get credit for helping with the project.

#) The Devil is in the Details: Our first PDSA pointed out multiple easily fixable but very relevant flaws in the execution of the plan.  Things like where to position the paper to best fit into the work flow, how to ensure the return of pens and clipboards, and how to gather these forms from providers all created barriers to implementation.  For our next phase of the project we are working to address these and make it easy to do the desired behavior.

#) The Importance of Stories: As time marches on projects inevitably begin to feel a bit old. Everyone gradually starts to lose enthusiasm and maintaining commitment and investment in something as difficult as behavior change becomes really important.  Today our group took a moment to reflect on why we think this project is actually very important.  A number of our group members shared stories of patients who didn’t have their top concern addressed, or situations where a “door knob” question came up that caused a provider to fall behind because of the delay and then have other patients leave unseen so they could catch the shuttle.  This served to reanimate the group and remind us of the overall goal of these QI projects- to improve the quality of care we provide to our veterans here in the clinic.

We plan on launching PDSA #2 soon.  Stay tuned for details and results!

Previous QI Post

VA Quality Improvement Project #5

Last week our team took on our first iteration at a PDSA cycle.

To recap, a PDSA is where you Plan an intervention (in our case the agenda setting form), Do that intervention (try it out in the real world setting), See what happened (gather data, comments, feedback), and then Act to either adopt, adapt, or abandon the intervention based on what we learned.  Ultimately this is done with the hope of using these learnings to drive further PDSA cycles and come up with the best intervention possible.

First off, we learned a ton from just trying to operationalize this.  Building the agenda setting form was pretty straightforward, but trying to decide who would hand it out, when they would hand it out, who would collect it etc got much more complicated.  Also we would need pens- many pens that would largely disappear.

This PDSA cycle provided our group with a windfall of data that I have summarized below.

General feedback points:
-More surveys tended to be delivered in the AM than in the PM.  Some felt a bit of “survey fatigue” or “initiative fatigue.”
-Some forms were delivered by LVNs in the door with the other pre-visit forms, others were brought in by the patients themselves

Comments on the form itself:
-The boxes are a bit too small to write out an entire statement
-The number 1 is written twice on the form, which can be confusing
-The form doesn’t actually say if you should save it and give it to your doctor, or to the LVN, or to the front desk
-It looks nice and official

Feedback from LVNs:
Overall, quite polarizing. One out of three surveyed thought that the forms were not helpful and that we should not do this again.  They felt like patients get so many forms and that they were tired of filling out surveys.  Notably, even this person stated that it didn’t take too long to hand out.  The other two LVNs actually liked the forms quite a bit and thought that they would help patients and were open to continuing to trial the intervention. One of the two LVNs who was supportive of the initiative was particularly enthusiastic and engaged.

Feedback from providers (about half were members of the QI project, the others were not):
Providers across the board seemed to think that the forms were helpful to patients and to themselves. They noted that only a minority of their patients arrived with these new forms, but that for those who did, it was easier to set an agenda and it streamlined the visit.  Others also noted that it was nice to have a “to do” reminder at the end of the visit to remember an eye clinic referral or refills if the patient also wanted something quick. None of the six providers we surveyed thought that the form actually made the visit take longer and all of the providers thought we should continue to hand out the form. Indeed, several non-group members were really excited about the project.

The Data:
The number of patients from the trial day was a bit smaller because of vacation days and illness, but out of a total of 11 in-person patients among our group, 7 had documented agendas, for a rate of 64%, much higher than our baseline rate and higher than our goal!  Notably, out of these 7 patients with documented agendas, 4/7 (57%) were from our intervention (the agenda setting forms), the remainder were from the appointment making system and pre-visit phone calls that already existed.  The documented agenda matched the actual agenda in all 7 cases.

Actually just doing this trial run taught us many places where the intervention could fail, which is exactly the point behind doing this micro-tests and trying to learn as much as possible upfront before making a major investment of time and effort in a project that has big flaws.

Our big steps for our next meeting will be analyzing this data to talk about what we learned and what we should adapt, adopt, or abandon and plan our next PDSA incorporating what we learned.  Hopefully after a few more small PDSA cycles we can roll out our intervention clinic-wide and improve the care we provide to the veterans we serve.

More updates to come!

Previous QI Project Post

VA Primary Care Quality Improvement Project #4

Today I attended another meeting for our VA primary care quality improvement projects group. Since the last update, the group has made tremendous strides in advancing their project to improve agenda setting in clinic.

After analyzing the process by which patient agendas are set in our last meeting, the group decided on an intervention using a worksheet for patients to fill out while in the waiting room before their visit.

Referring again to the IHI model for improvement, they planned out their small test of change. This involved creating a worksheet that elicits the patient’s agenda goals (Agenda Setting Questionnaire), making it available in the rooms where patients have their triage vitals taken, and getting buy-in from the front-line staff who could distribute these forms.

To test this change on a small scale, our group of four providers will gather some additional pre-intervention data only on their small group of patients today and implement their planned intervention next week. They will then use what they learn from their first small trial to adopt, adapt, or abandon parts of the intervention to come up with an improved intervention.

Stay tuned for pre and post intervention data and what we learned from this first test of change!

Previous QI Project Post

VA Primary Care Quality Improvement Project- Update #3

Yesterday our group at the VA took the next steps in moving forward our clinic quality improvement project.

We reviewed some initial data that we had gathered around how we document the patient agenda prior to the visit. Our current process involves an LVN calling the patient 2-3 days ahead of time to both remind them of their visit and to solicit any particular items they wanted to discuss with their provider. We found that out of a total of 24 patient-encounters that we reviewed, 3 were not called due to last minute appointment scheduling or time constraints, 8 calls went to voicemail so no agenda could be obtained, and 13 had “no urgent issues” or “will discuss with provider.” None of these 24 calls actually led to a documented agenda item in the pre-visit call note. Out of the 13 that had “no urgent issues” or “will discuss with provider,” there were 2 cases in which the patient was found to truly have no issues to discuss. Thus, only 8% of all of these patients had an accurate agenda documented prior to their visit. While this was a quick back of the envelope calculation, it certainly suggests a quality gap and an opportunity to improve how we care for patients in clinic.

Dr. Don Berwick, who founded the Institute for Healthcare Improvement (IHI) has said that “some is not a number, and soon is not a time” suggesting that to truly create change you have to have a specific goal with metrics and a goal date. To adhere to this principle we then created an aim statement focused around creating a SMART goal (specific, measureable, attainable, relevant, and time bound). Our aim is that for our subgroup of patients who present for follow up appointments in the Medical Practice clinic we will improve the percentage of patients with a documented agenda prior to the visit from ~10% to >25% by 11/17/15.

Next we reviewed a process map (see attached image) where we looked at all of the individual steps that a patient goes through to get to their appointment. By doing this (and in many cases actually going to the place where the work happens and observing the process in the wild) the group learned all of the small individual actions that have to work perfectly for a patient to actually have an agenda documented. We were then able to focus on specific steps that can easily lead to errors and brainstorm some potential counter measures to improve these steps and in turn improve the overall process.

At the end of our review, we identified two major areas where we thought we could improve the process of agenda setting:

  • When the LVN calls the patient by improving the dialogue between the LVN and the patient by changing the script.
  • By providing the patient with a sheet prompting them to list their agenda while they are sitting in the waiting room.

Our work for next week is to begin the final step of the IHI model for improvement in which the group will perform small tests of change as PDSA cycles and see if these ideas actually work.


Previous QI Project Post

VA Primary Care Quality Improvement Project- Update #2

I went to my second meeting for one of the VA primary care quality improvement projects yesterday. Over the past month or so the team thought about some issues in clinic that they think are particularly troublesome for the patients and for the providers and identified improving agenda setting in clinic as a major priority.

They felt that often our veterans travel for several hours to see their primary care doctor and often leave without having had the chance to discuss the most important issue or issues to them. Additionally, they worried that poor agenda setting led to misused time during the visit and actually slowed down their overall clinic day, making them late for other patients and adding to an already overburdened day.   Indeed, the literature supports both of these suspicions and outlines the importance of eliciting and prioritizing patient agendas to improve patient care and communication (Epstein, AAFP

Using the IHI’s Model for Improvement (, the team defined their goal as “improving agenda setting in clinic” and selected an initial metric of “percentage of patients with a documented agenda prior to starting the visit” to actually know if that change is an improvement. They then spent the remainder of the meeting creating a process map in which they outlined all of the various steps that go in to a veteran having an appointment with their provider with the goal of identifying certain steps that could be improved upon to create better agenda setting.

The next steps are 1) Improving upon the team goal by making it “SMART” and 2) analyzing the process map that was created to identify a few possible small tests of change for a plan, do, see, act (PDSA) cycle.

The team is meeting again next week to continue to build upon this work and hopes to start trialing at least one small change then. Stay tuned for updates from the clinic!

Previous QI Project Post

VA Primary Care Quality Improvement Project

Today the first group of residents from the VA primary care clinic embarked on their quality improvement project and I’m fortunate enough to get to mentor the team on Tuesday/Block A at the SFVAMC.

These projects are an annual tradition at the VA in which our housestaff at all VA sites team up with other health care trainees (nurse practitioners, psychologists, pharmacists, nutritionists) to work together to improve the quality of care for their patients and how the clinic operates.

While the projects are not without their flaws, they offer a really unique opportunity to get to work in multidisciplinary teams, try to change something that you care about, and learn QI. I think the greatest benefit in doing these projects comes from the actual experiential component of not just hearing about QI concepts and techniques, but being forced to apply them to solve a real problem in the clinic.

For the VA primary care project, we teach the Institute for Healthcare Improvement’s Model for Improvement framework to approach QI, which really revolves around three simple questions and rapid cycle improvement.   However, the first step in any good improvement project is asking the front line staff what needs to be done better. While the residents might have limited QI experience, they really are experts in the day to day practice of healthcare delivery in the clinic and know the most about the pain points and opportunities to improve.

Today after much discussion, our group settled on doing something around a pre-visit patient form to gather info on what the most important issue to the patient for discussion is that day, what meds they are taking, etc. We have starting asking additional stakeholders for their input, doing a brief lit review, and are going to think about directions to take the project and metrics before our next meeting. More to come as the project develops!