Risk factors for blood transfusion and postoperative anaemia following total knee arthroplasty

Listen to Andrew Duckworth and Brian Chalmers discuss the paper 'Risk factors for blood transfusion and postoperative anaemia following total knee arthroplasty' published in the October 2023 issue of The Bone & Joint Journal.

Click here to read the paper.

Find out as soon as the next episode is live by following us on Twitter, Instagram, LinkedIn, Tik Tok or Facebook!

Transcript

[00:00:00] Welcome everyone to our BJJ podcast for the month of October. I am Andrew Duckworth and a warm welcome to you all from your team here at The Bone & Joint Journal. As always, we'd like to thank you all for your continued comments and support as well as a big gratitude to our many authors and colleagues who take part in the series that highlights just some of the great work published by our authors each month here at the journal.

Today, I have the pleasure of welcoming Dr Brian Chalmers, who is an orthopaedic surgeon in the Department of Orthopedic Surgery at the Hospital for Special Surgery in New York, to discuss their paper entitled 'Risk factors for blood transfusion and postoperative anaemia following total knee arthroplasty'.

Which has been published in the October edition of the BJJ. Welcome Brian, it's great to have you with us. I really appreciate you taking the time. Thank you. I think it's great to be here and thanks for having me. So Brian, the the primary aim of this paper was to investigate risk factors for blood transfusion and postoperative anaemia following total knee arthroplasty in a large retrospective patient cohort, with the secondary aim being to attempt to establish a single preoperative haemoglobin cutoff that would minimize the likelihood of blood transfusion while [00:01:00] simultaneously maximizing the number of patients at low transfusion risk.

So, Brian, maybe as a background to the study for our listeners, can you give us a brief overview of the state of the current literature on this topic as you know it, and sort of what made you look at this question in particular? That's a great question. You know, I think, as far as my background and just things that I've been interested in research you know, I, I was a resident at the Mayo Clinic and then a fellow at the Hospital for Special Surgery and then stayed on staff and kind of throughout my career

I did a lot of research on. You know, outcomes implants, you know, things with fractures and a lot of revisions and a lot of metal and and things that were very cool to me at the time. And I think, as you go on to practice, you know, you realize that yes, those are important and and very interesting things to look at.

But really practice management, you know, patient outcomes around the time of surgery are also very important and things that I think about and worry about even more than than the case sometimes. And so. Okay. This is something that, you know, has in perioperative [00:02:00] outcomes in general pain, transfusion, patient experiences in general around the time of arthroplasty just matters a lot to me.

And so I focused a lot of my research in my early career and hopefully to to come on those types of topics. And so this was something that that we've looked at extensively. And especially with the advent of of tranexamic acid looking at transfusion rates and blood loss around the time of surgery, rapid rehabilitation protocols, getting patients out of the hospital fast for really more moving towards a model of

a well patient programme than more of a sick patient programme model. And so you know, historic transfusion rates, as we talked about in the paper have been extremely high. And I think we've done a great job with expedient surgery, tranexamic acid, you know, change changing our transfusion thresholds to decrease those and get patients out of the hospital faster

and in a more safer manner. [00:03:00] So that's kind of what what inspired me to do this paper as well as a lot of the other perioperative outcome type of papers that we're that we're looking at. Yeah, that's a really nice. I mean, Brian, I think, like, I couldn't agree more in the terms of as you get more into your practice, you realize that these little nuances that we take for granted.

And actually, there's a lot of work to be done with, shall we say, the more simpler things, but actually really important things to the patient, like, like you say, and I suppose related to that, you know, I'd. You know, from from your time at HSS and maybe maybe for your knowledge of the States as well, how is the practice and maybe indications from transfusion around joint arthroplasty changed or is it adapted at all?

Or is there any sort of trends you can see? Yeah, no, that's a great question. So, you know, I definitely think that we've, you know, revolutionized our blood management protocol as far as, you know, preoperative optimization of haemoglobin levels intraoperatively doing things, you know, like I said, expediently.

We use a lot of [00:04:00] neuraxial anaesthesia, which helps decrease you know, blood pressure during the surgery, decrease transfusion rates, blood loss. Obviously the tranexamic acid has been a huge game changer in the world of blood management. And then I think a big thing is, you know, postoperatively

monitoring patients and restrictive haemoglobin transfusion thresholds have been, you know, extremely important in, in decreasing our transfusion rates, which we know can lead to, to, you know, negative consequences. Not only transfusion reactions, but periprosthetic joint infection. So I think that there's been a, a big shift in the, in the paradigm of blood management and management of the patients in, in the last ten, ten to 20 years.

Yeah, no, absolutely. And I think we've certainly seen the same here. And like you say, something we've definitely seen more recently is blood is such a precious resource and actually we don't, it's not an endless supply. And actually these, the importance of looking at these things and actually only giving patients who really need a transfusion, a transfusion is a really important thing, whether it be in elective surgery or [00:05:00] trauma.

So moving on to the, to the, the study itself, just the design of it, you know, it was a retrospective cohort study utilized your institutional database of total knee arthroplasty patients from your large academic tertiary care institution. That was from 2016 to 2020, and that had a total of 14,901 patients that underwent a primary unilateral TKA as an inpatient.

So, Brian, maybe just very briefly, just as an overview, just for who it covers, what were the inclusion exclusion criteria for the study? Who did you actually look at? That's a great question. So, HSS is unique in a way, in a couple of ways. One is that, you know, we do the highest volume of, of arthroplasty in the world.

So there's a lot of patients, a lot of data. And secondly is that, we are very traditional in how we do things. Everybody gets preoperative labs. You know, traditionally, we've had patients stay overnight and get postoperative labs on a lot of patients, we're moving away from that in the last five years to doing some more ambulatory surgery.

Like I said, treating patients weren't a well model, but but [00:06:00] traditionally, we've gotten had patients stay overnight and gotten postoperative labs on pretty much everybody. And so there's a lot of. A lot of data in there, and that's why why you're seeing such large numbers here. Specifically for this study

our inclusion criteria was unilateral primary knee replacements and patients that did stay overnight and got postoperative day one labs. So anyone that left the same day anyone that. And it was only for a diagnosed osteoarthritis. So anyone with post-traumatic arthritis or AVN or anything like that, or underwent a simultaneous bilateral total knee replacement were excluded.

So we were able to capture a large, large number of patients because of those things that I mentioned. Yeah, absolutely. Very large. And like you say, clear indication for it. And you sort of touched upon it then, but in terms of your sort of institution's perioperative protocol during the same period. Was that sort of fairly standardized?

Did it change at all over that period of time? It's fairly standardized. There are some differences between surgeon between, you know, tranexamic [00:07:00] acid protocols and drain use that that changed. And we looked at in this study, which I think is one of the highlights of this study and potential things to change, but overall, as far as the perioperative protocol, everyone gets preoperative labs.

Most people stay in the hospital overnight. Get it in this study at least. Get postoperative day one labs. And so that's what we really looked at as far as looking at the haemoglobin levels, blood loss and those types of factors. Absolutely, and in terms of the sort of indication of the paper, but that was fairly standardized across across that period of time as well.

Yes. Yeah, exactly. We changed to a more restrictive transfusion regimen. I think that most institutions around the country have changed to about 8 to 10 years ago. So that that was standardized and, the techniques obviously are, you know, a little bit variable between surgeons, but throughout this study period, and we're still doing, [00:08:00] you know, most people use a tourniquet for for for the case.

Most the vast majority more than 99 percent are done with neuraxial anaesthesia and perioperative blocks by our anaesthesia colleagues. So overall, pretty standardized besides the few variables that we looked at in this study. So. Absolutely. Absolutely. That's great. And in terms of that, the preoperative risk factors and outcomes, what were the main things you wanted to look at?

Yeah, so we wanted to look at things that you can modify, right? Like tranexamic acid use drain use. We're, we're coming to some of the big things operative time. And then some things that you, you can't modify and and how we can, you know, potentially optimize patients better age, sex, body mass index and preoperative haemoglobin as well as ASA class.

So those things were the main things that we wanted to study in this in this paper. Perfect. And in terms of, before we get onto the results, Brian, in terms of the analysis performed, just very briefly, nothing, obviously, not [00:09:00] too much detail required. What, what sort of things were you looking at? And, and any sort of, so when you, particularly when you're looking at the cutoffs, you, you, you set some criteria for that.

We wanted to look at, you know, who had preoperative postoperative anaemia after surgery, which we had defined as a cutoff of haemoglobin less than 10. And then we looked at transfusion rates. And for this kind of predictive modelling, we, we looked at a transfusion rate of less than 1%. You know, that was kind of something that we decided that would be, you know, if you have more than 1%, that was, you know, probably clinically relevant.

 So so Brian, just briefly before we move on to the results what were the analysis that you should have carried out and in particularly make the, the cutoffs you were looking at when you were looking at transfusion rates? So we had to, to put some clinical make, make some clinical decisions and put some cutoffs and we you define preoperative he or postoperative haemoglobin as a haemoglobin less than 10 and a transfusion rate of, of 1% being, you know, clinically relevant.

That you would, you know, maybe want to monitor monitor [00:10:00] patients more closely, or if they had a risk of a transfusion over 1%, maybe you want to keep them in the hospital or do additional monitoring or do additional preoperative optimization prior to the prior to the arthroplasty, so we made those as as our cutoffs for the study.

 Absolutely. Yeah, no, that makes a lot of sense. If we move on to the results of the paper, so and what you found. So, like we say, there was almost 15,000 patients included. The mean age of the patients were 68 years and just about, just over 37 percent were male. About a third almost had a surgical drain during the procedure.

And that was one of the things, like you said, that was a factor you looked at. And the mean operating time for patients was just about around one and a half hours. So so I suppose the main findings, the initial findings you found, what were they in terms of the rate of blood transfusions and rate of postoperative anaemia, Brian?

Yeah, so this actually kind of surprised me. Our rate of transfusion was 3.7%, which is much higher than I expected. I honestly can't remember the last time that I [00:11:00] transfused a primary knee replacement. And so that surprised me a bit. And that's one of the reasons we, you know, looked at, looked at a lot of the analyses further.

About 15 percent of patients had a haemoglobin less than 10 on the postoperative day one labs and about 25 percent of patients who had a transfusion had more than one, one transfusion. So like I said, that that rate surprised me and the reason that we looked at this and looked at kind of the analysis further, because I think we're all a bit surprised by that, that transfusion rate.

Absolutely. I think that's a great thing to see, see, isn't it? When you look at these things, it's often, it's not always what you expect. And I think that's, that's one of the great things about them going to look at it further. And, and in terms of that, you know, what did you find in terms of the associations with regards preoperative haemoglobin levels and, and gender or sex with these outcomes?

Yeah, so what we found some, some expected findings, right? You know, increased operative time. Females just having a lower preoperative haemoglobin level, but even despite you know, controlling for [00:12:00] a lower preoperative haemoglobin level females in general had a higher transfusion rate you know, lower preoperative haemoglobin levels as we would expect those patients tended to have higher transfusion rates.

Lower doses of tranexamic acid. Like I said, in this study period, we hadn't really standardized our tranexamic acid dosing, whether it was IV topical multiple doses. And then drain use. I think those are those were the big things drain use leading to increase rates of of transfusion and postoperative anaemia were the big things that we found to be associated with with those results.

Absolutely. And they were sort of also sort of seen on the multivariate analysis that you then did after that. And and sort of finally with the results with, you know, with one of the thought there was very interesting things with regards establishing the preoperative haemoglobin cutoff. What did you find?

And also, which I thought was really interesting. What were the potential cost savings if you sort of extrapolated that out? Yeah, so we, you know, one of the goals was [00:13:00] of this study was, do we need to get lab postoperative labs on everybody? Right? I mean, that model, you know, moving to more ambulatory surgery.

You know, like I said, this model of having a well patient model, do we really need to get labs on everybody? And so we looked at, you know, what would what would it look like if we said we're accepting a less than 1 percent transfusion rate? And that was about 13 haemoglobin before surgery. So, those patients, you know, unless they're having some significant symptoms after surgery, probably don't need any postoperative labs.

And so that's, that's kind of what we found. It was a little bit higher than I would but, you know, I wouldn't think that anyone with a haemoglobin above 13 would, would need a transfusion. But obviously, in our study, there are, there were a few that did. So we found that that was the kind of significant cut off.

Yeah, absolutely. And like, like you said, if you extrapolate that out, there's quite some quite big cost savings, which are, which are interesting. Maybe we can come on to further, but in terms of. I suppose those are the key findings of the study in terms [00:14:00] of the implications of it. You know, you know, I think the strengths and the importance and the clinical evidence are really clear.

You know, it's a large, very large single centre study very well controlled environment in terms of a lot of constants throughout the study period. It's reporting on blood transfusion and postoperative anaemia following TKA. As you said, you know, the study really clearly highlights, you know, the importance that, you know, increased use of tranexamic, increased tranexamic acid dosing.

Decrease in the operating time, decreased drain usage may reduce those transfusion rates. And that single preoperative cut off of 13, like you say, may be a bit higher than you're anticipating, but obviously, aiming at less than 1%, and the implications that has for sort of postop labs, as you say. So, Brian, I suppose, how do you, how do you interpret these findings?

And I suppose, you've already alluded to it, but was it what you expected, and will it change things moving forward, do you think, for you guys? Yeah, I think it really highlighted the fact that we, you know, need to be a little bit more standardized and critical of our own, you know, blood management [00:15:00] strategies at our hospital.

I think you know, since this study, and since in the last couple of years, we've really standardized our, our tranexamic acid dosing and optimizing patients and, you know, everyone gets a gram of intravenous tranexamic acid before surgery. And we have almost everyone gets a second dose of intravenous afterwards, which I think, you know, significantly reduces transfusion rates.

We've standardized that there's very few patients and few criteria that that patients don't get a second dose. Our drain utilization at the hospital has significantly diminished over time. And obviously, you know, preoperatively looking at patients that we can optimize the haemoglobin levels.

And we've definitely gone to doing more ambulatory surgery and not getting preoperative labs, even if they do stay overnight or too many postoperative labs, even if they do stay overnight. With, like you said, with cost savings with just, you know, optimizing patients' experience in the hospital, not having to get a blood draw, not having to get labs [00:16:00] and really kind of standardizing our approach and and kind of modernizing everything as, as we've discussed.

Absolutely. And I think like, as you say in the study, in terms of that cut off, if you did use a cut off of hemoglobin, it's about two thirds of the patients undergoing a TKA would fit that sort of criteria. So that's, that's a huge number of patients. And in terms of, you know, how it fits with the current literature, I think you discussed that very well, but is there any sort of limitations to the study that you know, or hindsight with it that you sort of would think would be important for the listeners to know?

Yeah, I mean, obviously it's a single institution. There are variabilities that we, we can account for that, that that can lead to to variability in the study of especially you know, surgeons kind of do things differently. Postoperatively, you know, we don't didn't, you know, we captured ASA class, but there are certain groups of patients that are on, you know, postoperative or preoperative anticoagulation, that have kidney disease and other factors that that can lead to higher rates of bleeding, [00:17:00] higher rates of transfusion and, and higher blood loss.

Obviously, every patient after surgery has different cut off, you know, like I said, we use a restrictive transfusion protocol, but, you know, patients with cardiac disease or history have have higher transfusion threshold. So obviously not not generalizable to to every single patient, but I think within the literature make, you know, make sense on and what we found you know, females and drain use, preoperative haemoglobin level,

increased operative times, you're going to increase your rate of of bleeding, postoperative anaemia and requiring postoperative transfusions. And then, you know, like, I said, really, tranexamic acid has has been a big game changer in our, in arthroplasty and even in, and other fields, general surgery, trauma surgery and decreasing blood loss and decreasing transfusion rates.

So it's been found to be safe in several studies that we've done at the institution and that's been done throughout the world to not leading to [00:18:00] increased perioperative morbidity. You know, has not been shown to increase DVT or PE rates. And it has been shown to be safe. So I think you know, institutions should continue to standardize and basically ubiquitously use that to help our patients in the perioperative period.

Yeah, no, I totally agree, Brian. That's very clear and very fair assessment of it. And maybe just finally, you know, in terms of this area, maybe, you know, you talked about the beginning of the perioperative area. What do you see the future being? What's next with this? Or is there a next area that you feel that needs to be looked at?

Yeah, so I think, you know, we're going to continue this work and we're actually looking at our total hip population as well to see if there's similar trends. You know, there's obviously, you know, minimal patients with drains and total hip replacements. And so looking at different factors and different variables there.

And I think continuing to, to, to improve and move the needle. You know, tranexamic acid came out ten, 15 years ago. Is there another thing that, you know, another medication, another [00:19:00] technique, you know, something that we can do to continue to decrease where that that rate of transfusion, you know, not only 3 percent drops to 0.3%. Right? And so you know, a healthy patient undergoing a hip or knee arthroplasty, you know, we hopefully in the next ten years, it should be, you know, a, you know, no question of whether they're not going to get a transfusion. And, you know, we. You know, there's no thought that we're going to even need postoperative labs on that patient.

So, you know, just continuing to improve our techniques and and whether it's a another advance like tranexamic acid or, or just continuing to improve. I think that's what we're all looking at. No, I totally agree, Brian. I think it's a really nice way to sum it up. And actually, like you say, all these little things add to that ability to make it, like you say sort of well patient surgery and ambulatory surgery and getting people through the system in an efficient and, and, and, and safe way.

You know, I totally agree. Well, Brian, I'm afraid that's all we have time for. Congratulations to you and your team on a great study. It really added [00:20:00] to the literature in this area without doubt. And, and thank you so much for taking time to join us. It was great to have you with us. Thanks a lot, Andrew.

Thanks for having me. And to our listeners, we do hope you've enjoyed joining us. And we encourage you to share your thoughts and comments through social media, like, feel free to tweet or post about anything we have chatted about here today. And thanks again for joining us. Take care of everyone.