Antimicrobial Prescribing Practices

Today we are meeting on a virtual platform and I'm joining from the treaty 6 in the homeland of the Métis. Please take a moment to acknowledge the importance of the land which we each call home. We do this to reaffirm our commitment and responsibility in improving relationships between nations as well as improving our own understanding of local Indigenous peoples and their cultures. 

From coast to coast to coast, we acknowledge the ancestral and unceded territory of all the Inuit, Métis and First Nations people that call this nation home. As we move forward with today's discussions, let us remember our commitment to walking a path of partnership and friendship with Indigenous Peoples and remain mindful of the importance of person- centered, culturally safe care that honors the traditions, languages, cultures, and experiences of Indigenous Peoples. Today's webinar includes information external to the agency which may not reflect the views of the Public Health Agency of Canada. 

Following today's webinar, a copy of this presentation and recording as well as a feedback form will be sent to you. This webinar will be conducted in English. Simultaneous interpretation in French can be accessed through selecting the interpretation button in the menu at the bottom of your screen. 

Your audio has been muted to reduce noise, and video is disabled for this webinar. If you have technical problems, you can reach us through the chat function in the language of your choice. If you have a question for the presenters, please submit your question using the Q&A function and we will look to answer them during the panel discussion at the end of the webinar. 

There is an option to check if you do not want your name attached in the Q&A. I will now introduce our speakers. Robyn Mitchell has completed her Master of Health Science and Epidemiology from the University of Toronto and is a graduate of the Canadian field epidemiology program. 

She is a senior epidemiologist with the Public Health Agency of Canada with a focus on antimicrobial stewardship and the surveillance of healthcare associated infections and antimicrobial resistant organisms.  

Dr. Miranda So received her Doctor of Pharmacy degree from the University of Toronto and Master of Public Health degree in the field of epidemiology from the Harvard T-Chan School of Public Health. She is the manager of the antimicrobial stewardship program at the University Health Network in Toronto. She is also a clinician investigator with the Toronto General Hospital Research Institute and an assistant professor at the University of Toronto. 

Her clinical practice and research focus includes antimicrobial stewardship in immunocompromised patients and the intersection between antimicrobial use, resistance and population health. Please join me in welcoming Robyn and Miranda.  

Over to you. 

Robyn Mitchell: Okay. Thank you. So, thanks very much for the warm welcome Jenn. 

Miranda and I are excited to be here today to kick off WAA week and we will be presenting some information for you on prescribing practices in Canadian hospitals and some data and some future directions. So the objective of today's presentations are I'll be presenting some preliminary findings from the Canadian National Antimicrobial Prescribing Survey and then Miranda will present some of the work that her and her colleagues have done regarding presenting regarding development of a consensus protocol from to evaluate appropriateness of antimicrobial prescribing and then we will highlight some next steps.  

So just a bit of background and context for the purpose of this work. As we know optimizing antimicrobial use is really critical to slowing the spread of resistant organisms. However we need to have we need to collect standardized data on prescribing practices in order to inform how to best optimize use. And so while the Public Health Agency of Canada has routinely conducted surveillance for the consumption of antimicrobials in humans existing data sources have been unable to assess the appropriateness of antimicrobial prescribing practices in Canadian healthcare settings. 

So in 2018, the Public Health Agency of Canada partnered with Sinai Health and the University Health Network to monitor prescribing practices in Canadian hospitals using an Australian based platform called the National Antimicrobial Prescribing Survey or NAPS. And so I'll just briefly touch on the NAPS methodology. it is a hospital-based antimicrobial prescribing audit tool that uses a local prescribing guidelines to assess the appropriateness of antimicrobials dispensed in patient populations. 

And this is used this is done using a point prevalence methodology. And so data were collected from Canadian hospitals from 2018 to 2024 and entered into the online Australian NAPS web-based audit tool by hospital staff. And so NAPS developed and validated an assessment framework for adjudicating appropriateness of prescribing. 

So that's the algorithm that we see here. I'm not going to go into all of the details of this assessment but just to note that appropriateness you can see at the top there is defined whether guidelines are available or not available and as you can see there's a fair amount of detail required to adjudicate appropriateness. So it is very resource intensive and this assessment is done at the hospital level. 

So the staff that were collecting this information and making that interpretation, the assessment of appropriateness was done at the hospital level and then entered onto the Australian NAPS platform. So an overview of the participating hospitals. our data set includes 69 hospitals with representation across most of Canada. 

These hospitals were fairly diverse in size. You can see a range of small to large hospitals and that's based on hospital bed size. We also have a balanced mix of teaching and non-teaching institutions and hospitals were located across a range of small to large urban areas. 

Additionally, approximately one-third of participating hospitals indicated that they have dedicated funding or a dedicated stipend for an antimicrobial stewardship physician. And so this slide highlights some of the key indicators with respect to the quantity and quality of prescribing in Canada. So nationally we see an average use of 1.39 prescriptions per patient. 

Those admitted to a teaching hospital received an average of 1.4 prescriptions per patient compared with 1.32 in non-teaching hospitals. And similarly, we see that those admitted to a hospital which had that dedicated funding for an ASP physician received an average of 1.44 prescriptions per patient compared with 1.35 in hospitals without that dedicated funding. So there is definitely some overlap as you can imagine in these two indicators and it's likely that teaching hospitals are those with more resources and funding. 

Nationally we see that 72% of antimicrobial prescriptions were deemed appropriate, and this is based on the audit data collected from all 69 hospitals across our study period. We also see that 52% of antimicrobial prescriptions were deemed to be compliant with guidelines. So I'll go into a lot more detail about these indicators in a moment. 

I just wanted to kind of highlight the national level data because this figure summarizes the geographical variation of the of those indicators. So again we have data from nine provinces, Ontario representing the largest proportion of sites. So there was 32% of our sites were from Ontario. 

The number of prescriptions per patient is fairly consistent across the provinces. We see a range of 1.3 in BC to 1.46 in Quebec and that's really aligning with that national average of 1.39. 

So very the that variation is very tight around that that central measurement there. We do see a wider variation across provinces in the proportion of appropriate prescriptions. So that ranges from 63% in Nova Scotia to 84% in Alberta. 

Again, definitely kind of a wider range around that 72% national average. something else to keep in mind as we're interpreting these data that participation is voluntary and so there is definitely likely some selection bias in these data. So kind of key to keep in mind as we go through this. 

So first I'll just touch on some data regarding antimicrobial use. And so this chart shows the distribution of antibiotic prescriptions by the World Health Organization AWaRe categories. And so the AWaRe classification of antibiotics was developed by the WHO as a tool to support stewardship efforts. 

And antibiotics are classified into three groups as you can see here access, watch and reserve. And they're classified according to their activity as well as the risk of developing multi-drug resistance. Now for the purpose of this analysis, we did classify the reserve antibiotics according to Canada's reserve list. 

And so we have two links down there which you'll be able to access. And the main difference to the WHO list is being that Canada's list includes all carbapenems. And so what we see here is that among hospital antibiotic prescriptions, under half were antibiotics in the access category, so 44%. 

50% of antibiotics were in the watch category and close to 6% were reserve antibiotics. it is worthwhile noting that the WHO did set a target that at least 60% of total antibiotic prescribing at a country level should be access antibiotics. So based again just based on our data and with it all with all of its limitations regarding selection bias as I mentioned it does appear that we are you know falling a bit short of that target so shows room for improvement.  

And so taking a closer look at those reserve antibiotics, we see that 77% of reserve antibiotics were deemed appropriate. So which is good, just slightly higher than that overall antimicrobial average. 

You can see at the table here that carbapenems are the most commonly used reserve antibiotics with Meropenem being the highest one. And common indications for reserve antibiotics include bacteremia, bone and joint infections and skin soft tissue infections.  

Some additional key indicators looking at quality of prescribing. So we see, we found in our data that 22% of prescriptions were for indications or conditions that did not require antimicrobial therapy. We also observe that review or stop date were documented for 51 of antimicrobial prescriptions and that indication was documented for 85% of prescriptions. So this figure shows the most common indications for antimicrobial prescribing in Canadian hospitals. 

So that's along the y-axis and we can see that skin and soft tissue infections, community acquired pneumonia, surgical prophylaxis, these are the most common indications of use that came out in our data. And so then we stratified those prescriptions by their appropriateness. And so of those top ones, we can see that nearly one quarter to a third of those prescriptions were assessed to be inappropriately prescribed. 

So those are the pink bars that you see there. And then we also can see that the indications with the most inappropriate prescribing were surgical prophylaxis, cystitis and healthcare associated pneumonia. So the question becomes why are these the most inappropriately prescribed indications? 

And so we took a closer look at the data and again we've pulled those top three indications because they're the ones with the most prescribing volume. And so we looked at their reason for inappropriate use. And interesting to see for skin soft tissue infection, the most common reason for inappropriate prescribing was incorrect duration. 

For community acquired pneumonia, it was prescribed too broad and then for surgical prophylaxis it was again a duration issue. So surgical prophylaxis was greater than 24 hours. So this really provides some targeted actionable information that we can use. 

And then similarly this figure shows the most common antimicrobials prescribed in Canadian hospitals again based on our data. So you can see here that on the y-axis ceftriaxone, cefazolin and piperacillin-tazobactam were the most commonly prescribed antimicrobials. And then we stratified again by appropriateness. 

And so similar to the previous slide if we focus on those most commonly prescribed antimicrobials. So where we see the largest volume of prescribing approximately 20 to 25% of prescriptions were deemed inappropriate. And so again why is that? And so is what we did we pulled out those top antimicrobials and just focused on the top watch antimicrobials and as you can see here for so it's ceftriaxone, piperacillin-tazobactam and Vancomycin and there really is a common theme here that the main reason for inappropriate use among all three of these antimicrobials is spectrum too broad. And this figure just summarizes overall for all antimicrobials and all indications of use, what are the main reasons for inappropriate use. 

And so we see spectrum too broad again, incorrect duration, incorrect dose or frequency and surgical prophylaxis greater than 24 hours. And so we also looked at the appropriateness of prescriptions by patient age group, and we found that as you can see here appropriateness was highest among pediatric patients. So those less than 18 years. 

And so among peds patients, 88% of prescriptions were appropriately prescribed. Now we see that appropriateness is lowest for those patients 80 plus. And again always keeping in mind looking at the volume of prescribing for our more elderly patients versus pediatric patients in our data set. 

So keeping that in mind when we're interpreting. Also the key indications for the inappropriate prescribing among the patients 80 plus was cystitis, skin soft tissue infection and community acquired pneumonia. And so another useful and helpful way to look at the data is appropriateness by specialty. And so specialty is defined as the specialty under which the patient was admitted on the day of the audit. 

And so looking at this table here, we can see that you know peds are doing pretty well when it comes to appropriate prescribing, and that links back to the data in the previous slide.  

Again, keeping in mind the low number of prescriptions in peds patients in our data set compared to the adults population. We see that appropriateness is high for ID consults and for immunocompromised patients so this includes hematology, oncology, transplant and so we start to see lower appropriateness when we look at those under the specialties of critical care, surgery and medicine again keeping in mind you can see that there's really a high volume of prescribing in those in those specialties and then again sort of thinking this through and looking at the inappropriate prescribing and where we're seeing that inappropriate prescribing within those specialties. 

So within the critical care specialty we see that the most common indications for inappropriate prescribing are pneumonia, so that's community and healthcare acquired, surgical prophylaxis, COPD, sepsis. for the surgical specialty as we would expect the most common indication for inappropriate prescribing in that specialty is a surgical prophylaxis. 

And then for the patients under the medicine specialty, we see the most common indications for inappropriate prescribing are skin soft tissue infections, community acquired pneumonia and cystitis. So really some reoccurring messages throughout our data to help inform us. And so one of the last indicators which we'll be reviewing is compliance with guidelines. And so this is really key because as we saw in that NAPS assessment framework, compliance and availability of guidelines is one of the main criteria for determining appropriateness. 

And so what we see here is that 52% of prescriptions were compliant with guidelines. 19% were non-compliant with guidelines. And for nearly one-third of prescriptions, there was no guidelines available. 

And you can see here on the other side, the NAPS definitions for compliance, non-compliance, and then no guideline available. And now this figure shows the proportion of prescriptions by compliance with guidelines and then we've stratified that by teaching hospital status and is what we can see that is that among teaching hospitals which is the dark purple bar 58% of prescriptions were compliant with guidelines whereas only 44% of prescriptions giving at non-teaching hospitals were determined to be compliant with guidelines. And so what is really interesting here I think is the difference we see between teaching and non-teaching hospitals with respect to the proportion of prescriptions where guidelines were not available. 

So guidelines were not available for 35% of prescriptions in non-teaching hospitals, which is significantly higher than the 24% of prescriptions at teaching hospitals. And so this is another figure. It looks very similar but this figure again is showing the proportion of prescriptions by compliance with guidelines, but we've stratified it by whether a hospital indicated if they have that dedicated funding or that dedicated stipend for an ASP physician or not. And so similarly we see the guidelines were not available for 35% of prescriptions among hospitals without a dedicated that dedicated funding and resource which is much higher than those hospitals who have that dedicated funding. So I don't think it's overly surprising that hospitals you know with fewer resources may have reduced access to treatment guidance. 

And so this really speaks to the importance of accessible and relevant guidance and also lends itself to the work that AMMI Canada is doing in partnership with PHAC and they have just recently released some of their national guidance providing empiric treatment recommendations. So I think hopefully this information that we're collecting here will help inform kind of future work for guidance development and hopefully we can see some improvements in these indicators here. And so again finally this this figure shows the most common indications for antimicrobial prescribing in Canada. 

And then we have stratified it by compliance with guidelines and is what we see is that the indications with the highest proportion of guideline non-compliance. So again that's the pink bar were surgical prophylaxis, cystitis and healthcare associated pneumonia. Important to note that these were the indications that had the highest proportion of inappropriate use. 

So really suggesting that more work needs to be done kind of to support prescribers in these clinical areas. And I think these data are also quite valuable because they highlight which for which indication guidelines were not available. So we can see skin soft tissue infection, sepsis, intra-abdominal abscesses and for bone and joint infections. 

Those were the common those were the most the indications where guidelines were not frequently available. Ao giving us some again some targeted action items there. And so I'll just summarize some of the data that's been presented. 

Overall we see that 72% of antimicrobial prescriptions were deemed appropriate. The indications with the highest proportion of inappropriate prescribing were cystitis, surgical prophylaxis and healthcare associated pneumonia. The main reasons for inappropriate prescribing were reported to be incorrect duration and spectrum too broad and the lack of availability of guidelines and low compliance with existing guidelines were very much identified. 

So I think you know we've really just scratched the surface with these data. But I think it really does provide us with a direction of where to focus our efforts on improving prescribing practices and really you know supporting prescribers to help out with that. So with that I will pass it along to Miranda. 

Miranda So: Thank you.  

Thank you so much Robyn and thank you for the opportunity to present this information. I really appreciate this opportunity to talk about a new approach to develop a Canadian protocol. 

So I'm presenting on behalf of the project team about the development of Canadian consensus protocol to evaluate appropriateness of antimicrobial prescribing in hospitals for local quality improvements and national benchmarking. We're trying to get to the best of both worlds through an implementation science framework. Next slide please. 

So for some background about this project. Robyn had already provided a very nice summary about the NAPS project and we know that antimicrobial stewardship programs are established in hospitals under quality and safety framework and with a programmatic interventions as per accreditation Canada but currently there's no grassroot consensus among Canadian hospitals antimicrobial stewardship programs on a standardized approach to evaluate and report on appropriateness of antimicrobial use or quality of prescribing. A lot of work has gone into reporting on the quantity and cost of antimicrobial prescribing. 

But the next frontier is about the quality of use while accounting for hospital characteristics, local epidemiology and resources. And then of course this also includes the type of patient populations that are under the care of that hospital site. You heard Robyn mention several times that although we had good representation from coast to coast to coast in the NAPS initiative, it doesn't capture all of the Canadian hospitals. 

And so we need to be able to scale up the reporting of appropriateness across the hospital across Canada to better support these hospital-based antimicrobial stewardship programs. Next slide please. So the objective of this project was to develop consensus protocol with appropriateness definition, assessment framework and output design for evaluating and reporting the quality of prescribing tailored to the Canadian context in hospital-based antimicrobial stewardship programs. 

The intention here is that it's a bilateral relationship. Hospital antimicrobial stewardship program would be able to conduct appropriateness assessments or quality of prescribing to facilitate local quality improvement initiatives. At the same time, contribute data to national benchmarking so that the Public Health Agency of Canada can collect this information to better inform Canada's antimicrobial stewardship strategies, identify gaps, disparities and raise awareness and in turn be better able to support to provide supporting resources for antimicrobial stewardship efforts. Next slide please. So in conducting this project we applied the consolidated framework for implementation research or CFIR as kind of the backbone for the for the project. 

The innovation here is the Canadian protocol that's being designed. So that's kind of you know what we are trying to get to and CFIR as an implementation research methodology helps us to delineate who does what, when, where and how and we want to be able to understand who are the actors, where and what are the barriers and the facilitators, any factors related to the environmental settings and the conditions as well the individuals that are carrying out the intervention or the innovation itself and this is an established implementation science methodology. Next slide please. 

Thank you. So we carried out this project in several phases. Phase one started with a literature review to identify published assessment tools, appropriateness definition and the format of such output reports. 

This was followed by phase two, which was an electronic survey on the redcap platform targeting a purposive sample of Canadian antimicrobial stewardship and infectious diseases experts to understand their current activities at their sites and the limitations of appropriateness assessments. We reported their descriptive statistics for this part of the project. This was followed by phase three,  which was a series of semi-structured interviews of self-identified participants plus leaders and clinical experts recruited through known networks in antimicrobial stewardship, infectious diseases, public health. 

And the findings underwent thematic analysis with a constant comparison method and mapped to the CFIR domains. After that, abstracting from the information collected through a survey in phase two and the semi-structured interviews in phase three, a draft protocol with assessment framework, process, appropriateness definition and the format of the output reports was created. This was a preliminary output of this project. 

Next slide, please. So let's start off with the survey. So we received responses from 58 individuals over a period of eight six weeks. 

The majority were antimicrobial stewardship pharmacists followed by physicians. We had representation from multi-site health systems, single-site hospitals as well as regional or provincial level antimicrobial stewardship programs. The majority of the respondents, actually 98% of them, mentioned that they will use a Canadian specific protocol for antimicrobial stewardship program to help them with program planning, program development and 21% of the respondents said that they would use this information for guideline development. 

In terms of barriers and facilitators, the respondents provided some insight into what they would prefer as the kind of the benchmark or the yard stick for appropriateness. Local guidelines was preferred over the application of antimicrobial stewardship principles, followed by national guidelines. In terms of facilitators, the most common one was compatibility with the hospitals or hospital systems electronic health record system, the EHR. 

Followed by an informative output report. So this speaks to the format in which the information is generated after the appropriateness assessment was completed or the point prevalence auditing was completed. With respect to barriers, most cited a lack of human resources, a lack of time, and a lack of local expertise being the most common barriers. 

Next slide, please. In terms of the semi-structured interviews, we interviewed 28 participants representing infectious diseases and antimicrobial stewardship experts as well as leadership from six provinces. The two key qualities of the intended protocol was, or this innovation, were efficiency and flexibility because of the barriers that had been identified earlier in terms of human resources, time and local expertise. 

Efficiency was deemed very important to make this a success and flexibility to accommodate local quality improvement as well as contributing to national benchmarking would also increase the uptake of the of the use. So addressing both validity and utility of the appropriateness evaluation report. In terms of what data to collect the data elements and the data collection method. As mentioned earlier, the flexibility needs to be there to accommodate two sets of scopes to support local quality initiatives and this could be towards specific wards, for example the intensive care unit or it could be syndromic-based assessment, such as community-acquired pneumonia, or it could be towards antimicrobial specific evaluation for example the Canadian AWaRE categories or at the time the WHO AWaRE categories. It was noted that because the data collection method tend to be laborious and also presents as a significant barrier efforts should be made towards creating a seamless process that can leverage technology, accommodating all hospital types and resources. 

And that would be crucial for national adoption. If we want more sites to participate, we have to make sure it's easy for participating sites to do the right thing. And so this was this message was also very clear. 

Next slide please. So for the conclusion and next step. So a preliminary protocol was created in draft to accommodate local QI initiatives and national benchmarking based on the lessons that were learned from the NAPS project and the consensus protocol development. 

The next steps would be to create to develop and implement a Canadian-based tool to collect and monitor prescribing practices across Canadian hospital sites. And to support that we need to have a national working group that can help to represent different types of hospitals, urban, rural, large and smaller hospitals, medium-sized hospitals, teaching and community-based hospitals to validate the protocol through pilot testing before it could be applied across the country for implementation. And that concludes my part. 

So we would like to acknowledge several groups of individuals on the Canadian NAPS team we'd like to thank Dr. Andrew Morris, Yoshiko Nakamachi and other members of the team as well as all the contributing sites who had provided data to the NAPS initiative. 

We would also like to thank Caroline Chen, Karen Thurki, Rodney James and their team at Melbourne Health Guidance Group at the National Center for Antimicrobial Stewardship in Australia. For the consensus protocol, like I mentioned at the start I'm presenting on behalf of the project team and that includes Deborah Somanader from the antimicrobial stewardship program at Sinai Health and Sarah Tony from University Health Network for expertise in qualitative research. We thank all participants of the survey and interviews for sharing their insights and expertise and for dedicating their time to help us out with that project and partners at Public Health Agency of Canada Dr. Kanchana Amaratunga, Robyn Mitchell, Philip Pelletier, Shaghig Reynolds and Aboubakar Mounchili. Thank you.