Table 2 Representative quotations about interpreting the data

Reactions to the data

Satisfaction: case 2C participant

“One of my thoughts was … whatever the number is, 5 or 6 people that fall into whichever category. And so now it makes me wonder what did I do there? Cause that 2 from an individual perspective is a very small number. Why did I prescribe and not discontinue? Why did I discontinue and not re-prescribe? Why did I prescribe these drugs rather than those drugs in this particular case? When the absolute number for me as an individual is so small, and I scratch my head about that. So that, it’s fascinating to see where I hit with the rest of everyone else.”

Skepticism: case 1 participant

SPEAKER X: “And I’ll just mention all my 10 patients said not recorded and I never do an IV induction and I’m not sure how that data was…”

Understanding and questioning

Case 2A participant: “I have a question about that last chart. I am just trying to tease out in my mind where the hospice patients are. If they would pass away post-discharge, so how does that, would they drop from that? I am just wondering, I am trying to sort out…”

Case 2B: “How did you screen out the group of patients where medications were ordered in your absence, like when the psychiatrist or _____ the orders comes under your name?”

Case 3 participant: “Is this all procedures, is it, ER?” Moderator: “Yes. So again, … All general surgery Yes, so all of the general surgery, so the total number of patients is 716, so this data, … represents 72% of all procedures because those are the cases that were given Ketorolac and then in the chart below, that’s looking at the 512.” Participant: “How do you tease out what’s a general anesthetic and what’s a spinal?...or a regional…”

Justifying/contextualizing

Findings that physicians attributed to the behavior of others

Case 1 participant (referring to differences in surgical techniques and surgeons’ practices): “And 25 minutes versus just short of 4 minutes for some of them, as well.” Moderator: “And as well as prescribing when they are home. Some [surgeons] never give morphine. Some will only give, like, Tylenol or something. Some will give oral morphine. So, it’s not surprising pain scores vary.”

Findings that result from external factors (caring for hospice patients in this example): case 2A

Moderator: “So again, on the other hand you can see this person out here is probably a better example, saw lots of patients, 35, and half of them got either an antipsychotic or a sedative...” Speaker 1: “... So that’s probably where my midazolam comes from. Because we use tons of midazolam, Haldol, and stuff on hospice.”

Findings that physicians attributed to personal practices

Case 2A participant: (discussing why order a zopiclone on admission): “So, cause one of the things … I try to anticipate the calls I am going to have when I am the ward doctor, and try to prevent those. One of those is zopiclone.”

Case 3 participant (discussing use of ketamine for pain control intra-operatively): “I was using it for a while just to, my numbers are actually very low cause I used it for a while and then I got to know the surgeons and realized that for the most part, I know which ones are really good at [inaudible] filtration and the patients are waking up very comfortable, they are actually not requiring a lot of opioids, so then I did not feel like they needed the Ketamine.”

Findings that physicians attributed to patient factors

Case 2A participant: “Let us say when I look at my own it’s the same thing. I see drugs on mine that I am like, I can use that once a year. Like chlorpromazine, I guarantee you it’s because somebody had hiccups from something and I gave them chlorpromazine. I never use that drug.”

Case 2B participant: “There is a subtly different patient population as well. [Hospital D] is quite a bit older than [Hospital B].”

Case 2C participant: “I do not know, I do not know if it would, if the numbers would be large enough to have separate graphs for antipsychotics and sedatives. Sedatives like specifically zopiclone, just because probably the populations that would use those 2 medications are different. You know, just for your sleep, difficulty sleeping in hospital, versus dementia with behavioural issues for antipsychotics.”

Findings that physicians attributed to system factors:

Case 2A

Speaker 1 participant “The fall of 2013, was that our Ativan shortage too? Because that will tie in….Because that’ll factor into this, too, because we were ordering goofy things because we could not get the stuff we usually would use. Because that lovely plant in Quebec was down for so long. …Yeah, but midazolam is only IV. So, I used it to substitute for [IV] Ativan in a few people at times.”

Case 2D participant: “Maybe not that the patients that we see here are different, but our prescribing practices might be influence by like the consultants that we work with in this hospital, or the particular group that we work within.”

Reflecting

Case 1: Participants sharing their practices (discussing the use of nitrous in anesthesia in children): Speaker 1 “[NAME] tried it, without nitrous. I made a switch. Your induction. If that’s your induction try doing it without. Because traditionally we did it when this apoptosis came out I decided to delete nitrous from my practice. So, I have zero for induction. And I haven’t really noticed that longer an induction”. Speaker 2 “I don’t use it either.” Speaker 1 “In fact, I don’t notice a longer induction time actually. So, you might be saving 30 s or something. I haven’t documented it. But it’s something you could try to avoid having to push a button and remembering one more button. Just use air oxygen.”