Anesthesiology Performance Improvement and Reporting Exchange (ASPIRE)

Pediatric Subcommittee Meeting Minutes – March 2, 2026

Attendance:

Peter Bow, MPOG

Cathie Jones, Boston Children’s

Morgan Brown, Boston Children’s

Rahul Koka, Johns Hopkins

Robert Brustowicz, Boston Children’s

Eva Lu-Boettcher, UW Health - Wisconsin

Kate Buehler, MPOG

Frederick Mansfield, U.S. Anesthesia Partners*

Mei Calabio, MPOG

Kam Mirizzi, MPOG

Megan Charette, MPOG

Viviane Nasr, Boston Children’s

Joseph Cravero, Boston Children’s

Allison Nye, OHSU

Jurgen de Graaff, Weill Cornell

Katie O'Conor, Johns Hopkins

Lucy Everett, Mass General Brigham

Diana O'Dell, MPOG

Marla Ferschl, UCSF

RJ Ramamurthi, Lucile Packard Children’s*

Jackie Goatley, University of Michigan

Nirav Shah, MPOG

Kirsten Groody, University of Michigan

Ruchika Sharma, University of Virginia

Ruchika Gupta, University of Michigan

Brady Still, UChicago Medicine

Melanie Herren, MPOG

Meridith Wade, MPOG

John Huntington, Helen Devos Children's

Theodora Wingert, UCLA

*Denotes participant from non-active MPOG Institution

Start: 1601

Announcements & General Updates

Minutes from December 1, 2025 meeting approved - minutes and recording posted on the MPOG

website for review

Leadership update

• Dr. Eva Lu-Boettcher (UW Health - Wisconsin) was introduced as the new Pediatric

Subcommittee Vice Chair. Morgan Brown noted Eva’s prior involvement in MPOG measure

reviews and thanked her for stepping into the role.

Pediatric participation in MPOG

• Two additional pediatric hospitals have joined MPOG: Riley Children’s in Indianapolis and UCSF

Benioff Children’s, Oakland.

• The committee noted that MPOG now includes 32 pediatric hospitals, reflecting continued

growth in pediatric participation and broader national representation across both academic

and community pediatric practices.

2026 MPOG Pediatric meeting schedule

• The Pediatric Subcommittee plans to meet three times by Zoom in 2026: March, June, and

December.

• The committee also expects to hold an in-person pediatric gathering during the MPOG Retreat

at ASA, with a virtual option again available, consistent with the meeting format used in late

2025.

2026 Measures Requiring Formal Review

• Four pediatric measures are scheduled for literature review and update:

1. June Meeting

• TEMP-04 – Intraoperative Normothermia

• PAIN-01– Multimodal Analgesia

2. December Meeting

• SUS-06 – Low Fresh Gas Flow, Induction

• FLUID-02 – Minimizing Colloid Use

• Main reviewers must practice at an active MPOG site and will use an MPOG review template.

• Contributions will be recognized on the MPOG website and at the annual meeting.

• Volunteers are encouraged, especially where there is strong clinical interest.

AKI-03 Pediatric Cardiac Measure Update

• AKI-03-Peds has been released on the MPOG QI Dashboard as the first pediatric cardiac-

specific MPOG measure.

• The measure applies existing acute kidney injury logic to a cohort of pediatric patients

undergoing cardiopulmonary bypass. Key exclusions include:

o Patients not on bypass

o Patients with preexisting renal failure

o Patients without a baseline creatinine

o Patients without a postoperative creatinine

Discussion

• Morgan Brown reviewed site-level variation in AKI rates and noted that the data demonstrate

meaningful variation across pediatric cardiac centers of different sizes. The committee

discussed this as an important early step in pediatric cardiac quality measurement, especially

given that adult centers have more established AKI prevention efforts and bundles than

pediatric centers.

• AKI-03 is expected to be a future discussion topic for the pediatric cardiac subgroup.

Measure Development – Patient Blood Management (PBM)

Background

PBM and postoperative pain management were previously identified by the committee as the two

highest-priority topics for new pediatric measure development. The committee has not yet started the

postoperative pain build and devoted this meeting primarily to PBM.

Rationale

Morgan summarized the three major domains commonly described in patient blood management

literature:

1. Screening for, diagnosing, and treating anemia

2. Minimizing surgical, procedural, and iatrogenic blood loss

3. Managing coagulopathic bleeding

The subcommittee chose to begin with the anemia domain, since preoperative anemia identification

and treatment are considered foundational elements of blood management programs.

Current PBM concept under development

• The group is exploring a measure focused on identifying pediatric patients undergoing elective

major non-cardiac surgery who are at meaningful risk of transfusion and therefore may benefit

from earlier identification and treatment of preoperative anemia.

• The intended use case is a departmental or institutional quality metric that helps sites identify

potentially modifiable anemia in advance of elective surgery.

Preliminary Cohort Development

MPOG reviewed approximately one year of pediatric data to identify patients at high risk of transfusion

during elective non-cardiac surgery.

Initial exclusions included:

• Cardiac patients

• MRI-only or minimal procedural cases

• Endoscopy, bronchoscopy, and cystoscopy

• Lung transplant

• Short-duration cases less likely to reflect clinically meaningful intraoperative blood loss

The pediatric build used aspects of the adult major non-cardiac surgery phenotype, but the minimum

case duration threshold was increased from 30 minutes to 60 minutes.

The preliminary cohort included approximately 77,000 cases across 71 institutions, with more than

3,000 pediatric red blood cell transfusions, corresponding to an approximate transfusion rate of 4.6%.

Discussion

Clarifying the purpose of the metric

• Ruchika Gupta (University of Michigan) asked how the measure would function if the intended

intervention is something that must occur 4 to 6 weeks before surgery, such as iron

optimization.

• Morgan Brown (Boston Children’) emphasized that the metric is not intended simply to

measure anemia prevalence, but rather to identify a subgroup of patients at high enough

transfusion risk that preoperative anemia screening and treatment may be clinically

meaningful.

o Decision: The committee agreed that the metric is best understood as a process-

oriented or informational quality measure rather than a direct measure of

“preventable transfusion.”

Neonates and young infants

The committee discussed whether neonates should be excluded because they may not fit the intended

PBM model. Concerns included physiologic anemia in infancy, limited opportunity for preoperative

iron-based optimization, and the fact that neonatal transfusions are often driven by hemodynamic

instability or NICU/surgical decision-making rather than standard anemia thresholds.

• Joseph Cravero (Boston Children’s) noted that neonates may be transfused preoperatively

based on multidisciplinary decisions in ways that differ substantially from older pediatric

patients.

• Cathie Jones (Boston Children’s) suggested reviewing the procedures represented in the

youngest age groups to help determine whether those cases are appropriate for inclusion.

• Ruchika Sharma (University of Virginia) commented via chat that most surgical neonates may

be transfused more for hemodynamic instability than pure hemoglobin targets.

o Decision: Emerging consensus favored excluding or at least separately considering

patients under 30 days of age.

Emergency cases and highest-acuity patients

The committee discussed whether emergency cases and ASA 5/6 patients should be removed from the

cohort to improve actionability.

• Morgan Brown noted that many ASA 5/6 cases may already overlap with emergency cases and

may not be realistic candidates for the type of preoperative optimization this measure is

intended to support.

• Ruchika Gupta suggested via chat that the team specifically evaluate what the cohort looks like

after excluding emergency cases.

• Ruchika Sharma also asked via chat whether any of the ASA 5 cases were on ECMO, reinforcing

concern that the sickest patients may not be appropriate targets for this metric.

Which populations are modifiable?

The committee discussed whether transplant, trauma, burn, oncology, hematologic, and chronic

disease populations are too clinically complex or insufficiently modifiable for this measure.

• Morgan Brown noted that transplant cases represented one of the largest transfused groups in

the preliminary cohort and questioned whether transplant patients are realistic candidates for

anemia optimization pathways.

• Eva Lu-Boettcher raised concern that oncology and hematologic populations may skew the

data because their anemia is often driven by chronic inflammation, marrow pathology, or

protocolized inpatient care rather than a short-term modifiable deficiency.

• Eva later added via chat that less modifiable populations, including active malignancy,

hematologic disorders, oncology, and chronic disease, may need to be excluded because their

anemia is less responsive to short-term interventions such as IV iron therapy.

o Cathie Jones agreed via chat with that concern.

o Decision: The committee generally agreed that the measure should prioritize

populations in which preoperative intervention is realistic and useful.

Naming and framing the metric

• Theodora Wingert (UCLA) commented via chat that the goal seems to be “preventable,

unnecessary” RBC transfusion and suggested the measure may need to be named in a way that

makes that concept clearer. She also noted that neonates may not be good candidates for iron

infusions but may still be candidates for other PBM interventions such as PCC or coagulation

optimization.

o Morgan Brown agreed with the conceptual concern but clarified that this specific

measure is focused on anemia optimization, not the broader coagulation-management

arm of PBM.

Intraoperative vs perioperative transfusion

• Robert Brustowicz (Boston Children’s) asked whether the metric is trying to reduce only

intraoperative transfusions and whether it might simply shift transfusions to the preoperative

setting without a true net benefit.

• Morgan Brown noted that MPOG currently has its strongest data for intraoperative

transfusions, which is why the build has centered there.

• Meridith Wade added that MPOG can also capture some transfusion data in the four hours

before anesthesia start, which may help inform future refinement of the metric.

Hematocrit availability and timing

The team found that approximately 94% of patients in the preliminary cohort had a preoperative

hematocrit. The committee discussed how recent that hematocrit must be to be meaningful.

• Cathie Jones (Boston Children’s) stated that values older than one month are difficult to

interpret and likely should not be relied upon.

o The group generally supported using a recent hematocrit, likely within the prior month, as

part of the measure logic.

o Ruchika Gupta (University of Michigan) asked whether same-day labs are captured as

preoperative values; Morgan Brown confirmed that values drawn before anesthesia start

would qualify.

Procedure-based definition of high-risk surgery

• Cathie Jones (Boston Children’s) asked how the measure will determine which surgeries are

considered high-risk for transfusion and whether sites will be able to understand what

procedures are included.

o Morgan Brown explained that the current approach uses age, ASA class, anesthesia

CPT-based logic, and case duration to define a high-risk cohort.

o She also noted that once the cohort is refined, the team can review the most common

CPT codes represented in the measure to confirm that the included cases align with

clinical expectations for major pediatric surgery.

Additional potential data elements

• Theodora Wingert (UCLA) asked via chat whether MPOG has access to first postoperative

hemoglobin values.

o Meridith Wade (MPOG) responded via chat that this was not pulled for the current

analysis but is available in MPOG data.

o This may be useful for future PBM-related exploratory work, though no decision was

made to incorporate postoperative hemoglobin into the current build.

Narrow vs broad cohort strategy

• Nirav Shah (MPOG QI Director) suggested that if the group continues to struggle with defining

a broad “major pediatric surgery” cohort, it may be easier to start with one or two narrow

service lines or procedure groups, similar to approaches previously used in other MPOG

subcommittees. Example service lines mentioned included spine surgery and major bowel

surgery.

o Morgan Brown (Boston Children’s) agreed this may be the best fallback strategy if the

current broader approach proves too heterogeneous, though she noted pediatric

sample sizes remain a challenge.

Decision

• Continue development of the PBM/anemia measure concept.

• The measure should prioritize actionability and modifiability rather than simply capturing the

highest-risk or most transfused patients.

• Patients under 30 days of age will likely need to be excluded or analyzed separately.

• A recent preoperative hematocrit should likely be required, with approximately one month

discussed as a reasonable starting threshold.

• The committee favors first using this metric as an informational or departmental dashboard

measure rather than immediately deploying it in feedback emails.

• Additional cohort refinement is needed before formal implementation.

Next Steps

Morgan Brown and the coordinating team will refine the preliminary PBM cohort using feedback from

this discussion.

The team will specifically explore the impact of excluding:

• Neonates / patients under 30 days

• Emergency cases

• ASA 5/6 patients

• Less modifiable populations where feasible

The team will further review:

• Common CPT codes represented in the cohort

• Whether case duration thresholds should be adjusted further

• Whether diagnosis- or phenotype-based exclusions are feasible

• Whether transfusions in the four hours before anesthesia start should be incorporated into

future iterations

• If the broad cohort remains difficult to define, the group may pivot to a narrower, service-line-

based pilot measure.

PBM measure refinement will return for continued discussion at a future pediatric meeting.

Adjourned: 1653

Full Transcript

00:03:49 – Morgan Brown (Boston Children’s):

All right, maybe a few more people will keep trickling in, but we might as well get started. Thanks,

everybody, for joining today’s spring MPOG Quality and Pediatrics meeng. We’ll begin with a few

announcements and updates, review AKI-03, and then spend most of our me discussing the new

measure work based on our last group discussion on pediatric anemia. I’d especially like your input on

what we’ve developed so far and how we should reﬁne it to make the measure as useful as possible.

00:04:37 – Morgan Brown (Boston Children’s):

First, if you do not already know her, this is Dr. Eva Lu-Boecher, a pediatric anesthesiologist in

Wisconsin, who has graciously agreed to serve as the new vice chair for the commiee. She has already

been very involved with MPOG, including parcipang in mulple measure reviews. Thank you again,

Eva.

00:05:08 – Eva Lu-Boecher (UW Health - Wisconsin):

Thanks a lot.

00:05:11 – Morgan Brown (Boston Children’s):

We’re also very excited that pediatric parcipaon in MPOG connues to grow. Two more pediatric

hospitals—Riley Children’s in Indianapolis and UCSF Benioﬀ Children’s—are now on board, which is

great. We’re looking forward to having new members on this commiee from both instuons.

There are now 32 pediatric hospitals in MPOG, and I think we’re starng to get a very nice

representave sample across the country. We now have parcipaon from both smaller community

pracces and many larger academic instuons as well.

00:06:04 – Morgan Brown (Boston Children’s):

We also wanted to let everyone know, so you can plan ahead a bit, that we are going to have three

Pediatric Subcommiee meengs this year, all by Zoom. This is the March meeng, then we’ll have one

in June and another in December.

As you all know, in March we also have SPA, CCAS, and the pain group meengs, so we’re not planning

anything speciﬁcally in person there. But I know I’ll be there, and Eva probably will be as well, so if you

have quesons or want to talk about MPOG or pediatric quality work, please come ﬁnd us.

Last year, we also had a really nice in-person gathering at the MPOG retreat during ASA. It overlapped

with the SPA meeng, but it was valuable to be able to meet in person and interact with people we

usually only see on Zoom. We’ll be doing that again this year. There was also a Zoom opon last year,

and people joined that way as well, so if you can’t aend in person, you should sll be able to

parcipate remotely.

00:07:33 – Morgan Brown (Boston Children’s):

As part of our regular MPOG work, we also need to make sure we stay current with measure reviews.

This year, we would like to complete four measure reviews, and we are looking for volunteers. If no one

volunteers, you’ll have to listen to a cardiac anesthesiologist like me review pain measures, which might

be a lile painful.

If there’s a topic you’re especially interested in, this is a great chance to stay current with the literature.

If it’s something you haven’t worked on before, it’s also a good opportunity to learn something new.

We’re hoping to review at least one measure at the June meeng, and hopefully two.

00:08:22 – Morgan Brown (Boston Children’s):

I also wanted to menon that AKI-03 has now been released. It came out on Basecamp, and this is the

ﬁrst measure we’ve developed speciﬁcally for pediatric cardiac paents.

We started by building a cohort of children undergoing cardiopulmonary bypass and then applied the

exisng acute kidney injury logic to that group. There are many exclusions, most notably paents who

are not on bypass, paents who came in with renal failure, and paents who did not have either a

baseline creanine or a postoperave creanine. Unless there is a site-level data issue, most pediatric

cardiac paents would be expected to have those data elements.

Meridith pulled data to show site variaon, and I think this is a good example of a metric with

substanal variaon. The orange-yellow dots indicate center volume, and there is a wide range of

pediatric cardiac volume represented in MPOG. The most common range in this dataset seems to be

around 100 to 300 cases, which probably reﬂects many instuons. There are also some very high-

volume centers represented.

You can see there is considerable variaon in AKI rates, and that makes this a very interesng measure.

We now have a pediatric cardiac subgroup, and although we have not yet announced dates for those

meengs, I expect this will be an important discussion topic. Adult centers have focused on AKI

prevenon for a long me and oen use AKI bundles, but I don’t think that has fully worked its way into

pediatric pracce yet.

If anyone has quesons, please feel free to put them in the chat, interrupt me, or raise your hand.

00:11:00 – Morgan Brown (Boston Children’s):

What we really wanted to spend most of our me on today is the development of new pediatric QI

measures. Aer a lot of discussion over the last year, we idenﬁed two topics that seemed to be of

greatest interest to the group: paent blood management and postoperave pain management.

We have not yet started the postoperave pain build, so we will likely discuss that either at the next

meeng or at the MPOG retreat in October. Today, we want to walk through what we’ve done so far on

paent blood management and get your input.

Blood management is obviously a large topic. Susan Goobie previously joined us and talked about the

principles of paent blood management and what she thought would be useful to instuons working

on this topic. When you look across the literature, there are three recurring themes: screening for,

diagnosing, and appropriately treang anemia; minimizing surgical, procedural, and iatrogenic blood

loss; and managing coagulopathic bleeding to improve outcomes.

Diﬀerent authors phrase those ideas diﬀerently, but conceptually I think of paent blood management

as something like ERAS for transfusion. It’s a more holisc way of organizing principles we already know

are important. Each component maers, and focusing on just one may not be enough. But instuons

sll need a praccal way to act on those ideas, so we decided to focus ﬁrst on the anemia component,

since screening for and diagnosing anemia is really one of the cornerstones.

00:13:17 – Morgan Brown (Boston Children’s):

When you look more closely at the guidelines, they recommend diagnosing and treang anemia at

least three to six weeks before elecve surgery so there is me to intervene meaningfully. They also

discuss postponing elecve major surgery in high-risk paents in order to opmize them ﬁrst.

The problem is that there is not much clarity on exactly what “high risk” means in pediatrics. Some

authors just say elecve surgery, while others say elecve major surgery. So before developing a metric,

we felt we ﬁrst had to deﬁne which pediatric paents are actually high risk enough that preoperave

anemia opmizaon would make sense.

00:14:01 – Morgan Brown (Boston Children’s):

What we did was try to idenfy paents at high risk of blood transfusion during elecve non-cardiac

surgery. The idea was that if we can idenfy which children actually receive transfusions during surgery,

we can deﬁne a cohort of paents at meaningful transfusion risk and therefore potenally amenable to

intervenon.

00:14:42 – Morgan Brown (Boston Children’s):

At the start of this work, there were 33 million cases across 78 instuons, so we added several

inclusion and exclusion criteria to make the analysis more manageable. We limited the review to about

one year of data. We excluded cardiac paents because that populaon gets complicated quickly,

parcularly with cyanosis and diﬀerent transfusion thresholds.

We also excluded paents whose transfusions were unlikely to relate to what happened during the

procedure itself—for example, MRI-only cases, tonsillectomy paents who only got intubated, block-

only cases, endoscopy, bronchoscopy, cystoscopy, and lung transplant. We based some of that on the

adult major non-cardiac surgery phenotype, which we were trying to emulate.

00:24:26 – Ruchik Sharma (University of Virginia) (via chat):

Did this include transplant anesthesia, neuroanesthesia?

00:25:00 – Meridith Wade (MPOG) (via chat):

Yes, it included both.

00:14:42 – Morgan Brown (Boston Children’s):

The adult phenotype requires a case duraon of at least 30 minutes, but we changed that to 60

minutes because if the case is shorter than an hour, it seems less likely that a transfusion would be due

to something happening during the case and more likely due to some unusual circumstance.

That le us with about 77,000 cases across 71 instuons. Within that cohort, there were over 3,000

pediatric cases with red cell transfusion, which is a rate of about 4.6%. In other words, about 1 in 20

paents in this group were transfused, which is actually fairly high given how infrequently pediatric

paents are generally transfused.

00:16:41 – Morgan Brown (Boston Children’s):

We then tried to reﬁne the cohort further so that it would beer represent a truly high-risk populaon.

One queson is what to do with babies younger than 30 days. Do people think we can realiscally

intervene on anemia in that group, or would including them just dilute the metric? Ideally, in a

preoperave anemia clinic, treatment would involve oral or IV iron. We are not talking about

transfusing children before they come to the OR.

Does anyone have thoughts on that?

00:18:06 – Ruchika Gupta (University of Michigan):

What about the physiologic anemia we also see in infants? That drop is expected and happens around

the ﬁrst few months of life.

00:18:16 – Morgan Brown (Boston Children’s):

Yes, that’s a good point. Are you suggesng excluding babies younger than around 3 months because

there may not be much we can do? Or do you think some would sll be candidates for treatment with

iron?

00:19:03 – Joseph Cravero (Boston Children’s):

I think the neonatal group is diﬀerent. A lot of those children may be transfused to a certain

hemoglobin or hematocrit before surgery based on decisions made by the NICU team and the

surgeons. That is a very diﬀerent situaon from older children.

So I sll think the queson is interesng, but I wonder whether neonates should be considered

separately because the mechanism is so diﬀerent. We are usually not transfusing a 2-year-old before

surgery based on anemia alone, but a former 28-week infant may very well receive a transfusion before

surgery based on the judgment of the care team and surgeon.

00:20:44 – Morgan Brown (Boston Children’s):

Yes, exactly. That is why we wanted to talk this through as a group. We are trying to develop a metric

that helps prevent transfusion through earlier idenﬁcaon and treatment of anemia. If a populaon is

already likely to be transfused preoperavely, that does not really ﬁt the concept we are trying to

measure.

What we are imagining is a paent blood management approach where a child is seen in clinic, gets a

CBC, is found to be anemic, and then the team decides to delay an elecve surgery in order to treat

that anemia. I’m not sure there are many 2-week-olds in that category. Most surgeries at that age

probably need to happen.

00:21:43 – Cathie Jones (Boston Children’s):

My thought is that excluding children under one month seems reasonable. But I also wonder whether it

would help to look at what surgeries are actually being done in that one-month-to-one-year age range.

Seeing those procedures might help answer whether physiologic anemia really maers for this metric.

00:22:11 – Morgan Brown (Boston Children’s):

We could deﬁnitely try to look at that. The challenge is that when you have this much data, looking at

procedures becomes diﬃcult because there are so many CPT code combinaons. It makes it hard to

develop isolated exclusions. But yes, we could look at the cases under one month to beer understand

them. Even so, our inial thought was to exclude those paents.

00:22:51 – Ruchika Gupta (University of Michigan):

I think I’m also trying to get clearer on exactly what this metric is trying to achieve. It sounds like we are

looking for paents who should have had some intervenon before the day of surgery—iron or some

other treatment—based on informaon available weeks earlier.

00:24:26 – Ruchik Sharma (University of Virginia) (via chat):

Did this include transplant anesthesia, neuroanesthesia?

00:25:00 – Meridith Wade (MPOG) (via chat):

Yes, it included both.

00:22:51 – Ruchika Gupta (University of Michigan):

So would the metric ulmately be showing that we knew six weeks ago that a child had a low

hemoglobin and then sll ended up transfusing them during surgery? Is the goal to idenfy missed

opportunies for preoperave opmizaon and push pre-op clinics or surgical teams to change

pracce?

00:23:51 – Morgan Brown (Boston Children’s):

Yes, that’s essenally the idea. Simply measuring how much anemia exists in the populaon does not

mean much on its own. We are trying to idenfy a paent populaon at meaningful risk of transfusion,

with the idea that if you corrected anemia in that group beforehand, you might avoid some

transfusions.

That is how I interpret what the paent blood management literature is geng at. In adults, this has

existed for a long me. Paents coming for something like a hysterectomy may be found to be

signiﬁcantly anemic in pre-op clinic, receive IV iron, and then avoid transfusion. We are trying to

idenfy which pediatric populaon might ﬁt that same model.

Looking at anemia in children coming for ear tubes would not be useful, because transfusion would

almost never be relevant in that seng. We need a populaon where transfusion is a plausible

perioperave risk.

00:25:27 – Morgan Brown (Boston Children’s):

Another queson is what to do with the highest-acuity paents. When we look at ASA class, the

paents being transfused oen fall into ASA 5 or 6. Do people think those are realisc candidates for

intervenon? My hope is that many of those paents would already drop out if we exclude

emergencies, but I’m sll wondering whether it makes sense just to remove ASA 5 and 6 altogether.

Any thoughts?

00:32:22 – Ruchik Sharma (University of Virginia) (via chat):

I feel most surgical neonates get transfused more for hemodynamic instability than pure Hb targets…

pung myself out there!

00:26:25 – Morgan Brown (Boston Children’s):

We also started looking at surgical speciales. The largest group receiving transfusion in this non-

cardiac cohort was transplant—not cardiac or lung transplant, but kidney, small bowel, and liver.

That raises another queson: do people think anemia opmizaon is relevant for transplant paents?

Would those paents realiscally be brought in and treated in advance for anemia? Or is that too

complex a populaon for this type of metric?

Similarly, the second major group that stands out is trauma and burn paents. Do people think those

are populaons where meaningful intervenon is possible, or are they also poor candidates for this

kind of measure?

00:34:08 – Ruchik Sharma (University of Virginia) (via chat):

Were any of the ASA 5s on ECMO?

00:28:09 – Morgan Brown (Boston Children’s):

I see that Theodora put in the chat, “I understand the goal—preventable unnecessary RBC transfusion.

It may be helpful to name it in a way that gets the idea across. Neonates may not be candidates for iron

infusions, but they may be candidates for PCC and other agents to opmize coagulaon.”

I completely agree with that point. The only challenge is that coagulaon opmizaon really belongs to

the third pillar of paent blood management, whereas this parcular metric is focused on the anemia

component. So yes, neonates may be candidates for other types of blood management intervenons,

but that would be a diﬀerent metric concept.

What we are really trying to do here is deﬁne a populaon that instuons can idenfy and potenally

opmize, not necessarily prove that every transfusion was preventable.

00:35:12 – Theodora Wingert (UCLA) (via chat):

I understand the goal to be “preventable, unnecessary” RBC transfusion. Maybe it would be helpful,

semancally, to name it in a way that gets this idea across. Neonates may not be candidates for iron

infusions, but would be or are candidates for PCC and other agents to opmize coagulaon.

00:29:22 – Theodora Wingert (UCLA):

Yes, that makes sense. I just think that many people looking at the data may not interpret it that way. It

is hard to classify transfusions as preventable.

00:29:58 – Morgan Brown (Boston Children’s):

I agree. It is very hard to know which transfusions are truly preventable. If a surgeon causes enough

bleeding, transfusion may be unavoidable. But the idea is that if the paent begins in a beer place,

then maybe a smaller amount of blood loss would not push them to transfusion.

So yes, this measure depends on accepng the basic paent blood management premise that treang

anemia is probably good for the paent and may also decrease transfusions. In cardiac surgery, for

example, there has been debate because bypass itself causes so much hemodiluon that some argue

preoperave anemia maers less. I’m not sure I completely believe that, but those arguments do exist.

00:31:13 – Morgan Brown (Boston Children’s):

As we connued looking at the data, a few other paerns emerged that seemed sensible. Paents who

were transfused tended to have longer surgeries. They also tended to be smaller children. And, not

surprisingly, they had lower hemoglobin and hematocrit levels than children who were not transfused.

00:32:01 – Robert Brustowicz (Boston Children’s):

Are we trying to eliminate only intraoperave transfusions? Could this just shi transfusions to the

preoperave seng without producing any real net gain?

00:32:19 – Morgan Brown (Boston Children’s):

That’s a very good queson. Unfortunately, because of how MPOG is currently built, we have detailed

data about intraoperave transfusion, but not the same level of detail about preoperave or

postoperave transfusions. So from a measurement standpoint, we are mainly limited to intraoperave

transfusions.

That said, the rates we are seeing are very similar to what has been reported in the literature, which

suggests the data are representave. Conceptually, paent blood management is about avoiding

transfusion overall, but the measurable endpoint here would be intraoperave transfusion.

00:33:14 – Meridith Wade (MPOG):

We do have some transfusion data in the four hours before anesthesia starts, so for some cases we can

capture transfusions that overlap into the OR period. If the group wants to limit the metric strictly to

intraoperave transfusions, we can do that, or we could consider including that perioperave window

as well.

00:33:38 – Robert Brustowicz (Boston Children’s):

That would maer, because some ASA 5 paents may be transfused in ancipaon of the OR.

00:33:45 – Morgan Brown (Boston Children’s):

Yes, exactly. That is one reason I was considering excluding ASA 5 paents altogether.

00:34:07 – Morgan Brown (Boston Children’s):

We also looked at how many paents in this preliminary cohort had a preoperave hematocrit,

because that was an important queson for the feasibility of the measure. About 94% actually did have

a preoperave hematocrit, which was encouraging.

That led to another queson: how recent does that hematocrit need to be in order for us to believe it

reﬂects the paent’s preoperave status? If it was drawn more than a month ago—or certainly six

months ago—I’m not sure how useful it is. But if it was within the last month and showed anemia, I

think that is probably meaningful.

Does anyone have thoughts on that?

00:36:08 – Cathie Jones (Boston Children’s):

I think if it was not within the past month, it would be hard to judge. I would not put much trust in an

older value.

00:36:23 – Ruchika Gupta (University of Michigan):

I had a queson about the same-day hematocrit values. At our instuon, lile kids oen come in, get

an IV, and then we send a hematocrit before surgery. Are those deﬁnitely being counted as

preoperave values?

00:36:41 – Ruchika Gupta (University of Michigan) (via chat):

Can we eliminate emergency cases to see what the numbers look like? That might help with decision

making.

00:36:46 – Morgan Brown (Boston Children’s):

Yes, those should all be values obtained before anesthesia starts.

00:36:52 – Ruchika Gupta (University of Michigan):

Okay.

00:36:58 – Morgan Brown (Boston Children’s):

So there must be many instuons where that is standard workﬂow. We do not do it that oen here,

but clearly a lot of paents in the dataset do have those labs.

00:37:10 – Morgan Brown (Boston Children’s):

So aer looking through all of this, the basic idea would be to create a metric that measures the rate of

anemia in paents older than 30 days who are undergoing elecve major non-cardiac surgery, who

have a transfusion risk of at least 5%, and who also have a recent hematocrit available.

The concept is that these should be paents idenﬁed through preoperave screening, with the

opportunity to intervene before they come for their elecve surgery.

Aer all the earlier meengs and discussion that led us here, how do people feel about that? Do you

think that would be useful?

00:38:13 – Cathie Jones (Boston Children’s):

Would sites be the ones choosing that 5% transfusion-risk threshold, or would that be built into the

measure?

00:38:21 – Morgan Brown (Boston Children’s):

That threshold would be built into the measure. We were trying to think praccally about what it

means to ask surgeons or instuons to delay or cancel surgery in order to opmize anemia. We

wanted a transfusion risk high enough that doing so would feel jusﬁed. A 5% risk, which is about 1 in

20 paents, seemed like a meaningful threshold.

We could lower it, but if we do that, we will include many more anemic paents for whom delaying

surgery may feel much less appropriate.

00:39:00 – Cathie Jones (Boston Children’s):

That makes sense. I think my queson is more how the metric would determine which procedures have

a transfusion risk greater than 5%. Is that based on instuonal history, MPOG-wide data, CPT codes, or

something else?

00:39:24 – Morgan Brown (Boston Children’s):

It would be based on the criteria we discussed: age, ASA class, speciﬁc anesthesia CPT codes, and case

duraon. Those factors together would deﬁne a high-risk cohort based on exisng MPOG data.

It is also broadly consistent with the recent paper from Susan Goobie and David Faraoni in BJA,

published in January, where they looked at NSQIP data and found similar transfusion risk factors.

00:40:14 – Cathie Jones (Boston Children’s):

Do we actually have a list of those CPT codes, though? Because people may want to know what

procedures are being included.

00:40:20 – Morgan Brown (Boston Children’s):

That is where things get tricky. Procedure coding gets complicated very quickly. For example, you can

have mulple diﬀerent CPT variaons for what feels like the same general surgery, and then the

queson becomes how to group them in a meaningful way.

We do have reasonably good data on surgical CPT codes, though not every instuon submits them

equally rigorously. Once we get closer to a ﬁnal cohort, we can deﬁnitely idenfy the most common

surgical CPT codes represented so that people have a clearer sense of what procedures are driving the

metric.

But in general, MPOG usually builds metrics by starng with a broad cohort and removing paents

through exclusions rather than by starng from a short list of approved procedure codes.

00:41:03 – Theodora Wingert (UCLA) (via chat):

Do we have ﬁrst postoperave hemoglobin?

00:41:34 – Cathie Jones (Boston Children’s):

Yes, I understand that. I’m just thinking about how people may react if they feel they are being judged

on cases where they would not have expected a transfusion and therefore would not reasonably have

delayed surgery.

00:41:46 – Meridith Wade (MPOG) (via chat):

I didn’t pull that data for this, but we deﬁnitely have it.

00:41:57 – Theodora Wingert (UCLA) (via chat):

Reacted to “I didn’t pull that data for this, but we deﬁnitely have it” with 󹰎󹰏󹰐󹰑

00:42:03 – Morgan Brown (Boston Children’s):

Exactly. That’s why I think this metric is more useful at the instuonal or departmental level. It is less

about second-guessing a single case and more about making sure a department is monitoring how

much anemia exists among children coming to major surgeries.

00:42:13 – Theodora Wingert (UCLA) (via chat):

Thanks, Meridith!

00:42:52 – Meridith Wade (MPOG):

I think we may want to start by placing this measure on the dashboard as an informaonal

departmental measure only, rather than including it immediately in feedback emails. That would allow

us to vet it over me and look at departmental performance before deciding whether to use it more

broadly.

00:43:19 – Morgan Brown (Boston Children’s):

Eva also put in the chat that we might consider removing paents with acve malignancy or

hematologic disorders, which is a fair point. We would just need to see how feasible that is. We do have

some comorbidity phenotypes available, but I am not sure whether we can reliably determine from the

data whether certain surgeries are associated with acve cancer, for example.

00:44:07 – Eva Lu-Boecher (UW Health - Wisconsin):

Yes, I think those populaons may skew the data. If you think about oncology paents, for example,

they oen undergo smaller procedures like line placements but are transfused relavely more

frequently than the general pediatric populaon. At the same me, the kinds of intervenons we are

talking about may not be especially modiﬁable in that group.

A lot of those children are transfused on the ﬂoor, so they may not even be anemic when they come to

the OR. They may also have diﬀerent transfusion criteria that are not dictated by anesthesiology. And

many of their procedures are not emergent, so they would not necessarily be excluded on urgency

alone, and they are not always ASA 4 or 5 either. So I think they are a parcularly tricky populaon for

this metric.

00:44:56 – Morgan Brown (Boston Children’s):

Yes, that makes sense. Bone marrow procedures are another example of something that occurs

frequently in those populaons. Part of the reason we chose an anesthesia duraon threshold of one

hour was to eliminate some of those smaller procedures without having to rely too heavily on

comorbidity exclusions.

Instead of excluding all malignancy, it may make more sense to increase the required case duraon,

because the goal is really to focus on procedures where there is enough physiologic change or blood

loss that the transfusion risk reﬂects the situaon we are actually trying to prevent.

00:45:55 – Eva Lu-Boecher (UW Health - Wisconsin):

Right. And that is what is interesng about the paper you menoned from Susan. The highest-

morbidity, highest-mortality populaons are oen also the least modiﬁable. The paents we can

probably intervene on most eﬀecvely are the anemic paents who are not necessarily the ones

undergoing the most complicated, highest-risk surgeries.

So part of the queson becomes whether we want this metric to focus on the populaon with the

highest morbidity and mortality impact, or the populaon in which we can realiscally make a

diﬀerence from an anesthesia or perioperave care standpoint.

00:46:50 – Morgan Brown (Boston Children’s):

Exactly. Paents who receive transfusions certainly have higher morbidity and mortality, although it is

hard to separate whether that reﬂects the transfusion itself or simply the fact that they are sicker.

People who are very focused on blood conservaon would argue the transfusion maers, but it is

diﬃcult to untangle.

Sll, we can look more closely for things like port placements and line placements to make sure they

are not slipping into the cohort. We can also increase the anesthesia duraon threshold if needed. That

may be the simplest way to eliminate many of these short, protocol-driven cases without having to rely

on diagnosis-based exclusions.

00:48:00 – Morgan Brown (Boston Children’s):

Any other thoughts? I hope this ends up being useful. It is deﬁnitely a more theorecal and

complicated metric than some of our others. It is really a process metric focused on anemia, where the

implied intervenon happens before surgery. That makes it harder to deﬁne cleanly, and you can see

why the adult group may not have tackled it yet either.

00:48:46 – Cathie Jones (Boston Children’s):

The only other thing I would add is that in oncology paents, a number of smaller procedures oen get

bundled together, so what looks like several minor procedures can become a much longer case and sll

involve transfusion.

00:48:58 – Morgan Brown (Boston Children’s):

Yes, that is a very good point, and we should look at that as well.

00:49:11 – Nirav Shah (MPOG):

One thing you might consider, if you connue having trouble geng to the right cohort, is doing

something that we have done in the cardiac and obese subcommiees. Instead of trying to capture a

broad populaon, you could build the measure around one or two speciﬁc service lines or procedure

groups—something like spine surgery or major bowel surgery.

Somemes it is much easier to build a measure around a very narrow set of procedures where the

concept clearly applies. That can become a kind of proof of concept. If you are doing the right thing for

one service line, you can expand to others later. That may be more praccal than trying to create a

complicated algorithm that sll never fully captures the intended paent populaon.

00:50:26 – Eva Lu-Boecher (UW Health - Wisconsin) (via chat):

The other thing is to consider excluding less modiﬁable populaons—cohorts with acve malignancy

and hematologic disorders, including oncology or chronic disease paents. Their anemia is related to

chronic inﬂammaon or marrow pathology, making it less responsive to short-term intervenons such

as IV iron therapy.

00:50:38 – Cathie Jones (Boston Children’s) (via chat):

Reacted to Eva Lu-Boecher’s comment with 󷶡󷶢

00:50:38 – Morgan Brown (Boston Children’s):

Yes, we could certainly try that. One of the challenges in pediatrics is that our overall numbers are

smaller to begin with, and the number of transfused paents is also relavely low. But if this broader

strategy does not work, that may ulmately be the approach we need to take—start with a narrower

populaon and build outward.

We would sll need to deﬁne what constutes major surgery in pediatrics, which has not really been

established, but I do think that narrow-to-broad approach makes sense as an alternave strategy.

00:51:30 – Nirav Shah (MPOG):

Yes, exactly. It may simply help you get to an inial workable populaon more quickly.

00:51:41 – Morgan Brown (Boston Children’s):

I agree. I will say that our ﬁrst pass felt encouraging because we did get to roughly a 5% transfusion

rate, which seemed meaningful. I think the next step will be to test some of the changes we discussed

today and see how they aﬀect the cohort. We can also review the CPT codes represented in that

populaon, as Cathie suggested, so we beer understand what types of procedures are included.

When we looked at the CPT codes in some earlier iteraons, they generally did make sense for major

surgery, so I think there may simply be more than one reasonable way to approach this.

00:59:53 – Ruchika Gupta (University of Michigan) (via chat):

If we are looking at something that is modiﬁable four to six weeks before surgery with iron, how will

the metric work to help us achieve this?

00:52:35 – Morgan Brown (Boston Children’s):

Thank you again to everyone for your input and for spending me thinking this through. I know it is a

complicated topic. We look forward to seeing you all in June, and hopefully some of you at the MPOG

meeng in a few weeks. If you have quesons, please reach out to Meridith, and please connue

spreading the word at your site about this meeng. We look forward to another good discussion in

June.

00:53:13 – Morgan Brown (Boston Children’s):

All right, thanks, everybody.