Anesthesiology Performance Improvement and Reporting Exchange (ASPIRE)

Quality Committee Meeting Minutes – July 27, 2026

Attendance:

Ansari, Bilal (UHN)

Foster, Julie (MyMichigan)

Norman, J. Blake (Alabama)

Abou Nafeh, Nancy (AUB)

Georgiadis, Paige (University of Vermont)

O’Conor, Katie (Johns Hopkins)

Bartels, Karsten (Michigan)

Gibbons, Miranda (Maryland)

O’Dell, Diana (MPOG)

Bauza, Diego (Weill Cornell)

Glanding, Kimberly (UAB)

Pace, Nathan (Utah)

Bollini, Mara (WUSTL)

Goatley, Jackie (Michigan)

Pantis, Rebecca (MPOG)

Booker, Cheyenne (Henry Ford)

Gostic, Will (Stanford)

Pardo, Nichole (Corewell Health)

Bow, Peter (Michigan)

Grewal, Ashan (Maryland)

Paul, Jonathan (Columbia)

Bowman-Young, Cathlin (ASA)

Hall, Meredith (Bronson Battle Creek)

Poindexter, Amy (Holland)

Brennan, Alison (Maryland)

Heiter, Jerri (Trinity Health)

Rolfzen, Megan (Michigan)

Brown, Morgan (Boston Children’s)

Herren, Melanie (MPOG)

Roselinsky, Howard (Yale)

Brown, Sheree (Trinity Health)

Hetrick, Chris (Trinity Health)

Ruiz, Alexandra (Trinity Health)

Buihuynh, Vivien (VCU)

Huntington, Michelle (Corewell West)

Ruiz, Joseph (MD Anderson)

Calabio, Mei (MPOG)

Janda, Allison (MPOG)

Schwerin, Denise (Bronson)

Cassidy, Ruth (MPOG)

Johnson, Rebecca (UMHS West)

Scranton, Kathy (Trinity Health)

Charette, Megan (MPOG)

Karamchandani, Kunal (UT Southwestern)

Shah, Nirav (MPOG)

Chopra, Ketan (Henry Ford - Detroit)

Kaushik, Teshi (UAB)

Shettar, Shashank (OUHSC)

Clark, David (MPOG)

Khan, Meraj (Henry Ford)

Smiatacz, Frances Guida (MPOG)

Claybaugh, Deborah (MyMichigan)

Kirke, Sarah (Nebraska)

Stewart, Alvin (UAMS)

Cohen, Bryan (Henry Ford)

Kumar, Vikram (MGH)

Stumpf, Rachel (MPOG)

Coleman, Rob (MPOG)

Kunkler, Bryan (Corewell West)

Szymanski-Bogart, Brooke (MPOG)

Corpus, Charity (Corewell Health)

Lai, Emily (MD Anderson)

Tao, Jing (MSKCC)

Couture, Lindsey (Vermont)

Lalonde, Heather (Trinity Health)

Tyler, Pam (Corewell Health)

Cusick, Jordan (OHSU)

Lee, Maxine (Carilion Clinic)

Wade, Meridith (MPOG)

Cywinski, Jacek (Cleveland Clinic)

Lewandowski, Kristyn (Corewell Troy)

Wedeven, Chris (Holland)

Ellison, Pavithra (WVU Medicine)

Lopacki, Kayla (Trinity Health)

Weinberg, Aaron (Weill Cornell)

Esmail, Tariq (UHN)

Mack, Patricia (Weill Cornell)

Wissler, Richard (University of Rochester)

Duggan, Beth (UAB)

Malenfant, Tiffany (MPOG)

Zhao, Xinyi (Sarah) (MPOG)

Fermin, Lilibeth (Cleveland Clinic)

Mathis, Mike (MPOG)

Zhu, Shu (Columbia)

Finch, Kim (Henry Ford Detroit)

McKinney, Mary (Corewell Health)

Zittleman, Andrew (MPOG)

Fisher, Clark (Yale)

Munson, Kristin (Michigan)

Start: 1001

Announcements & General Updates

Minutes from May 2026 Quality Committee Meeting approved – minutes, slides and recording posted on the

MPOG website for review

Roll call

• Roll call taken via the Zoom attendance list. Participants who joined by phone, or who were present but not

listed on Zoom, were asked to contact the Coordinating Center (support@mpog.zendesk.com) so

attendance credit could be recorded.

2026 Meetings & Events

• MPOG Annual Retreat: Friday, October 16, 2026, San Diego, CA — held the day before the ASA annual

meeting. Both in-person and virtual attendance will be available.

o Registration for the MPOG Retreat is now open.

• Next Quality Committee meeting: Monday, September 28, 2026.

Featured Member – July and August

• Patrick Henson, DO, FAOCA (Vanderbilt University) was recognized as the featured member for July and

August. Dr. Henson is a long-standing quality champion and contributor to MPOG and completed the

glycemic management measure reviews earlier in 2026. His featured member profile is posted on the

MPOG website.

Associate QI Director Transitions

• Anthony Edelman, MD, MBA is transitioning off the Associate QI Director role, having accepted

appointment as Senior Associate Chair of Clinical Operations at the University of Michigan. The committee

thanked Dr. Edelman for his contributions, particularly his perspective spanning community and academic

practice settings.

• Allison Janda, MD was welcomed as the new Associate QI Director. Dr. Janda is an Assistant Professor of

Anesthesiology at the University of Michigan, Chair of the MPOG Cardiac Subcommittee, and Co-PI of

VEGA2.

New and Returning MPOG Sites

• West Virginia University joins MPOG as the newest site, bringing a broad range of practice settings of

interest for both QI and research work.

o Department Chair: Dr. Ed Nemergut; Principal Investigator: Dr. Heather Hayanga; Quality Champion:

Dr. Pavithra Ellison; Anesthesia IT Champions: Dr. Charles Barry and Dr. David Melnick.

• University of Alabama at Birmingham returns to MPOG following a transition to Talis and is again

submitting an MPOG extract.

o Chair: Dr. Dan Berkowitz; Principal Investigators: Dr. Philip McArdle and Dr. Brant Wagener; Quality

Champion: Dr. Blake Norman.

PONV-05 Revision

Background

The 5th consensus guidelines for the management of postoperative nausea and vomiting were published in

2025. Dr. TJ Gan (MD Anderson), a lead author of the guidelines, presented on PONV at the 2025 MPOG Retreat.

Drawing on the guidelines, Dr. Gan’s presentation, and prior Quality Committee discussion, the Coordinating

Center brought forward a set of planned modifications to PONV-05 together with one open question for the

committee.

Planned Modifications to PONV-05 for 2026

• Add olanzapine as a qualifying antiemetic. Olanzapine is referenced in the consensus guidelines and was

identified as an effective prophylactic antiemetic in a single-center randomized controlled trial conducted

by colleagues at Yale (Dr. Jamie Hyman and others).

• Exclude all obstetric cases. Cesarean deliveries will be addressed in a separate measure, PONV-06-OB, in

recognition of the different guidelines and protocols that apply to that population.

• Modify the definition of propofol infusion to require a duration of ≥ 5 minutes rather than 1 minute. Case

validation and cross-MPOG discussion indicated that the 1-minute definition introduced measurement

noise; the 5-minute threshold reduces that noise.

• Apply the “remained intubated” phenotype, so that PONV prophylaxis expectations apply only to patients

extubated at the end of the case and not to patients transferred directly to the ICU intubated and sedated.

Revisions to that phenotype are essentially complete.

• PONV-03, the PONV outcome measure, is scheduled for review at the September 2026 Quality Committee

meeting, with Dr. Tariq Esmail (University Health Network, Toronto) as reviewer.

Discussion – Minimum Propofol Infusion Dose

Nirav Shah (MPOG) raised the question of whether PONV-05 should specify a minimum propofol infusion dose.

Under the current specification, any dose of propofol infusion counts as an antiemetic. Data presented by Dr.

Gan (Sprung J et al., Anesthesia & Analgesia 2024;139:26–34) show a dose-dependent reduction in the need for

rescue antiemetics in the PACU, with the effect plateauing at approximately 90–100 mcg/kg/min. The

Coordinating Center proposed a minimum dose of 50 mcg/kg/min for ≥ 5 minutes as a practical floor.

• Morgan Brown (Boston Children’s) noted that a small single-center study conducted at her institution used

doses of approximately 10–50 mcg/kg/min as sub-hypnotic dosing for antiemetic purposes, and that this

dosing appeared to be effective; such cases would not meet the proposed 50 mcg/kg/min cutoff. She

reiterated the point via chat, noting that Boston Children’s doses lower than the proposed threshold and

found it effective in their single-center experience.

• Ketan Chopra (Henry Ford – Detroit) asked, both verbally and via chat, whether he should begin educating

his staff to move toward the higher dose in anticipation of the change, noting that current practice at his

site is generally 10–15 mcg/kg/min. Nirav Shah shared that he had personally stopped using sub-hypnotic

dosing following Dr. Gan’s presentation and did not think advance education was unreasonable.

• Lilibeth Fermin (Cleveland Clinic) noted, via chat and then verbally, that CMS has recently proposed

removing PONV prevention from the anesthesiology quality metrics for 2027, on the reasoning that

prophylaxis is now close to standard of care, with a comment period available for anesthesiology teams to

respond.

o Nirav Shah responded that MPOG data continue to show that PONV occurs and that meaningful

variation exists across sites, and that this data would be shared at the September 2026 review. He

noted that data sources including self-report may show less variation than MPOG’s PACU medication

and nursing note–based data.

• Pavithra Ellison (WVU Medicine) commented via chat that her children’s hospital uses low-dose propofol

infusions in most cases with an IV, at a minimum of 50 mcg/kg/min and usually 100 mcg/kg/min — a

threshold arrived at independently of the MPOG proposal.

• Amy Poindexter (Holland) asked via chat whether the impact of the change on PONV scores had been

evaluated. Meridith Wade (MPOG) responded via chat that the Coordinating Center is building a propofol

infusion phenotype with these parameters for addition to PONV-05, and will compare the score impact

before publishing the measure revisions.

• Chris Wedeven (Holland) proposed considering a lower floor of 30 mcg/kg/min for elderly patients, noting

that efficacy appears to begin around 30 mcg/kg/min and that 50 mcg/kg/min approaches an anesthetic

dose in many elderly patients. Nirav Shah agreed this was an important point, noting existing concerns

about polypharmacy and about achieving a third antiemetic in high-risk elderly patients, and committed to

reviewing the implications for that population.

Poll – Minimum Propofol Infusion Dose (QC 7.27.26 PONV-05)

Open to all attendees rather than one vote per site. 35 of 80 participants (43%) responded; 35 of 35 (100%)

answered.

Discussion – Dexmedetomidine

• Dosing across the available studies and meta-analysis was highly variable (one study assessed 1 mcg/kg

before incision), and nearly all studies were small (< 100 patients).

• Overall quality of evidence was judged to be poor.

• The 5th consensus guidelines describe dexmedetomidine in the text but do not include it in Table 1, the list

of included antiemetics — possibly because the correct dosing for PONV is not established.

Decision

• The committee supported a minimum propofol infusion dose of 50 mcg/kg/min for ≥ 5 minutes for PONV-

• Review the result with the PONV-05 measure reviewer and evaluate the implications of the dose floor for

elderly patients, including the possibility of a lower threshold in that population.

• Dexmedetomidine will not be included in PONV-05 as a qualifying antiemetic at this time. The Coordinating

Center is open to reconsidering once stronger evidence and greater expert consensus are available.

Measure Review – OPIOID (Opioid Equivalency, OME)

Reviewer: Clark Fisher, MD, PhD (Yale New Haven Health)

Review document: OME – Clark Fisher Review 7.27.26

Background

• OPIOID is an informational measure that sums all opioids administered between anesthesia start and

anesthesia end and converts them to oral morphine equivalents (OME). The measure was created and has

been maintained by Dr. Mike Burns.

• Included anesthesia CPT-based case types: Cardiac, Spine (Adult), Spine (Pediatric), Upper Abdomen, Lower

Abdomen, Hysterectomy, Knee/Popliteal, Hip, and Tonsil/Adenoid (pediatrics only).

• Attribution is assigned to the provider(s) signed in for the largest duration between anesthesia start and

anesthesia end. The reviewer found this to remain appropriate.

• Approximately 80% of the current MPOG conversion factors derive from three sources: the CDC Guideline

for Prescribing Opioids for Chronic Pain (2016), the American Pain Society Principles of Analgesic Use (7th

edition, no longer updated), and OpenAnesthesia.

Discussion

• Clark Fisher (Yale) proposed four criteria for an ideal MPOG opioid measure: supported by strong evidence;

easily understood by practicing anesthesiologists rather than strictly academic; able to handle the opioids

most commonly used in the OR (IV fentanyl, morphine, hydromorphone, remifentanil and others); and

applicable to how opioids are used in the OR. He noted that no available approach meets all four, and

therefore proposed a two-track model.

o He illustrated that “equivalent dose” is context dependent. To match 100 mcg of IV fentanyl:

approximately 15 mg of morphine matches peak effect, 10 mg matches average effect over an hour,

and 40 mg matches the effect 5 minutes after administration. Conversion factors derived in steady-

state settings such as PCA or chronic oral therapy may not hold in the operating room.

• Nathan Pace (Utah) noted via chat that Rücker G, the third author of the Dinges network meta-analysis, is

an outstanding statistician who developed many network meta-analysis methods. He subsequently asked

via chat about Eleveld models, and commented verbally that many pharmacokinetic models exist, that

some recent models have been shown to outperform others, and that he agreed with standardizing on a

single model — noting that because anesthesiologists titrate to effect, the specific choice matters less in

practice.

o Clark Fisher agreed, observing that moving from OME to pharmacokinetic modeling does not remove

the problem of arbitrary choices — it adds a choice of model on top of the choice of conversion factor

— and that the reasonable first step is to adopt a defensible existing standard (Steven Shafer’s

stanpumpR) and adjust as issues are identified.

• Michael Mathis (MPOG) supported keeping the measure informational, with no value judgment attached,

while providing awareness of variation across providers and sites. He noted that anesthesiologists

administer opioids for materially different reasons — hemodynamic stability in cardiac cases,

supplementing depth of anesthesia in MAC cases, minimizing movement with a remifentanil infusion when

muscle relaxant cannot be given, and postoperative analgesia — and that these different purposes imply

different ways of displaying the information on the dashboard.

o Focusing on the analgesia component, Dr. Mathis argued that the most informative endpoint is opioid

effect at the time the patient becomes conscious — arrival in the PACU or ICU — rather than total

exposure across the case. He noted that a milligram of fentanyl given before incision in a cardiac case

is very different from the same dose given at the end of the case, and that a patient whose

remifentanil infusion is turned off with no other opioid will have a very different PACU experience

than a patient given 5 mg of morphine at the end of the case.

• Morgan Brown (Boston Children’s) asked via chat about adding time to OME — for example, OME per

hour of case.

o Clark Fisher responded that the existing measure already presents OME normalized in several ways,

including by time and by patient weight. He proposed that the initial pharmacokinetic build

approximate the existing OME — total intraoperative mu receptor exposure — because it is a

substantial lift to stand up, but noted that once the modeling exists, questions that are impossible to

answer with OME become straightforward: opioid effect at PACU handoff, opioid effect at the first

PACU rescue dose, or opioid effect at laryngoscopy compared between providers. The subsequent

challenge becomes selecting which outputs to present.

• Nirav Shah (MPOG) framed the work as a learning health system opportunity: variation identified through

the measure can feed back into observational analyses and potentially pragmatic clinical trials, generating

recommendations that could be shared with quality champions and providers. He emphasized that the

immediate purpose is understanding current practice rather than prescribing practice change.

• Nirav Shah (MPOG) also reported that the current OME informational measure is not displaying as

intended in the QI Reporting Tool. The measure’s display broke during the migration to the new platform,

so the measure summary screens showing opioid administration per hour equivalents are not accessible

and the values currently shown are not fully accurate. The intended display is the average opioid

equivalent, in morphine equivalents, for an average patient and an average case duration.

o Michael Mathis (MPOG) confirmed that the pharmacokinetic approach is computationally feasible,

noting that Dr. Fisher performed exactly this analysis in the MPOG cardiac surgical population, and

suggested that the pilot include at least one group in which patients are awake at the end of surgery.

Clark Fisher agreed the computation is doable, while distinguishing the resources available for a

dedicated research project from those available for one measure among many in a monthly

production process.

• Tariq Esmail (University Health Network) noted via chat that chronic pain patients with high baseline 24-

hour opioid requirements presenting for large surgeries such as spine will have substantially different

opioid amounts than opioid-naive patients undergoing the same surgery. Clark Fisher agreed this would be

valuable to break out once outpatient medications are captured in MPOG.

• Sarah Zhao (MPOG) asked via chat whether the stanpumpR package is designed specifically for OME work.

She explained that in her own study she found the package includes only approximately seven opioid

medications — fentanyl, hydromorphone, morphine, oxycodone, methadone, alfentanil, and sufentanil —

along with other non-opioid medications, and does not cover additional agents she needed (for example

hydrocodone and naloxone), which could produce inaccurate results for broader analyses.

o Clark Fisher (Yale) acknowledged the limitation and shared his email address so that the question

could be followed up directly.

Reviewer Recommendations

• 1. Update MME conversion factors to align with the NIH HEAL standard. Changes are modest — for

example oral hydromorphone from 7.50 to 6.00, oral levorphanol from 2.00 to 2.73, oral methadone from

6.00 to 6.38, oral tramadol from 300 to 150, and sublingual buprenorphine from 0.400 to 0.773.

• 1a. Fix the IV buprenorphine multiplier to 0.4. The current MPOG value of 75 appears to be a

typographical error; the reviewer confirmed 0.4 against the original cited source. Other IV conversion

factors should be retained for now: while the Dinges 2022 network meta-analysis is more evidence-based

than existing sources, it has its own limitations and has not been widely adopted, so the reviewer

recommended reassessing at the next review while remaining open to earlier adoption if the committee

preferred.

• 2. Pilot a pharmacokinetics-based measure of intraoperative opioid use. Build a metric using the same

inclusion and exclusion criteria as the existing MME opioid measure, using the models and conversion

factors in Steven Shafer’s stanpumpR (https://stanpumpR.io) to measure Fentanyl Equivalent AUC for

every case, presented alongside the MME-based measure. This would allow remifentanil effect to be

captured and provide a corrective where MME conversion factors do not fit the intraoperative use case.

• 3. Add a measure for PACU opioid use. A vote to create an MME measure for the PACU was taken at the

2023 review but never implemented. It remains valuable and could use the updated conversion tables. An

equianalgesic concentration conversion measure could also be built for the PACU, but only if judged

valuable after the intraoperative pilot.

• 4. Considerations for future reviews. Whether the neuraxial opioids currently included should remain;

whether to add an MME value for remifentanil (currently counted as zero opioid); flagging outpatient

opioid use if home medications become more integrated in MPOG; adding or altering procedure groups;

and evaluating the success of the pharmacokinetics-based measure.

Vote

QC 7.27.2026 OME Overall Vote — 21 of 72 participants (29%) responded; 21 of 21 (100%) answered.

QC 7.27.26 MME, IV, Pharmacokinetics — 20 of 68 participants (29%) responded, across four questions.

Decision

• OPIOID (OME) will be modified

o MME conversion factors will be updated to align with the NIH HEAL standard

o The IV buprenorphine multiplier will be corrected from 75 to 0.4.

o Existing IV conversion factors will otherwise be retained, with the Dinges-based conversions

reassessed at the next measure review in 2029

o A new MME-based measure of PACU opioid use will be developed

o A pharmacokinetics-based measure of intraoperative opioid use will be piloted

• Next Steps

o Dr. Fisher and the Coordinating Center will convene the outside experts who contributed to the

measure review to develop an implementation plan

o The Coordinating Center will correct the display of the OME informational measure in the QI Reporting

Tool so that the per-hour and per-weight measure summary screens function as intended.

MPOG Web Application Updates

MPOG Data Library

• The Concept Browser, Phenotype Browser, and Measure Browser have historically been three separate

browsers hosted on different platforms with unaligned user interfaces, making related data elements

difficult to find.

• These have now been consolidated into a single application, the MPOG Data Library, with a universal

search bar that returns matching phenotypes, concepts, and measures together.

• Release is expected on Monday August 10

, with the announcement distributed via the MPOG forums and

email.

• The principal change for users is that login will now be required to view the concept, phenotype, and

measure browsers, using the same credentials as the QI Reporting Tool and DataDirect. Questions can be

directed to support@mpog.zendesk.com

Adjourned: 1101

Next Meeting: Monday, September 28, 2026

Full Transcript

ASPIRE Quality Committee – July 27, 2026

Original wording preserved; filler, false starts, and speech-to-text artifacts corrected for readability. Chat comments are inserted at

the chronologically correct point and marked “via chat.” Timestamps are elapsed recording time (hh:mm).

00:14 – Nirav Shah (MPOG):

All right, I think we're going to get started now. Good morning, everyone. Happy Monday. I hope everyone is

doing well and that everyone's summer is going well. It's been a couple of months, and I'm excited to share

some updates with you and to dive into some of the great work that's been happening across MPOG. We have a

busy agenda: a couple of announcements, a couple of measure updates — which may not all happen in this

order — a measure review of the opioid equivalency measure by Dr. Clark Fisher at Yale. Clark, I see you're on;

thank you so much. And then a couple of web app updates as well. So another packed agenda, as we usually

have.

It looks like we have quorum, so I think we can move on. The last meeting was in May; the minutes for that

meeting are posted on the website. As always, roll call will be via the Zoom attendance list. If you happen to be

dialed in via phone, not signed in to Zoom, and you need credit, don't hesitate to reach out to us so that we can

mark your attendance down.

00:15 – Nirav Shah (MPOG):

The MPOG Annual Retreat, as always, is the day before the ASA — this year in San Diego, as we have done for

the last few years. It will be both in person and virtual, so I'm hoping to see as many of you there as possible. I

think we have a great agenda lined up. Dr. Karandeep Singh, who is the Chief AI Officer at UC San Diego, will be

speaking. Dr. Kayla Putti [name as transcribed] from UHN in Toronto will be talking about transfusion, and we

have a great panel, so I'm super excited about the agenda this year.

00:16 – Meridith Wade (MPOG) (via chat):

MPOG Retreat - MPOG Registration is now open!

00:16 – Nirav Shah (MPOG):

A couple of announcements. Our featured member for July and August — I'm not sure if Dr. Henson is on — is

Dr. Patrick Henson from Vanderbilt University. Dr. Henson, as many of you know, is a long-standing quality

champion, member, participant, and contributor to MPOG across multiple facets; he did the glycemic

management reviews a couple of months ago. If you want to learn a little bit more about what Dr. Henson is

thinking in terms of MPOG and quality, please go to our website, click on his featured member profile, and you'll

have a good read.

00:17 – Nirav Shah (MPOG):

A couple of transitions I wanted to announce. Dr. Tony Edelman, my colleague here at the University of

Michigan, has accepted a senior leadership role within the department; he'll be the Senior Associate Chair of

Clinical Operations. He'll be very busy with that, so he'll be transitioning away from his Associate QI Director

role. It's been really fun and great to work with him over the last couple of years. His perspective, especially

having been in both community and academic settings, has been really helpful.

However, I am super excited to announce that Dr. Allison Janda, who I think pretty much everyone here

probably knows, will be our new Associate QI Director. I think all of you know that Dr. Janda is a cardiac

anesthesiologist here at the university. She's chair of the Cardiac Subcommittee, and also co-PI of VEGA2. So,

welcome, Allison — although you are a familiar face — and I'm super excited that you'll be the new Associate QI

Director.

00:18 – Allison Janda (MPOG):

Thank you, Nirav, and thanks, everyone. I'm really excited to be working more closely with this group, and

looking forward to the work to come.

00:18 – Nirav Shah (MPOG):

Awesome, thank you. We have a couple of new sites. First of all, West Virginia University is a new MPOG site.

Great partners out in West Virginia. It took them a little while to get onboarded, but we're super excited to have

them join us. They have a really broad range of practice settings, which I think will be super interesting from

both the QI and the research perspective. Dr. Ed Nemergut is the chair at West Virginia. Great team out there,

so welcome, West Virginia colleagues, and looking forward to working with all of you.

00:19 – Nirav Shah (MPOG):

And then, welcome again to UAB, the University of Alabama at Birmingham. They have been a site for several

years. They did a transition to Epic and are now on board again. Dr. Dan Berkowitz, who's chair there, is also on

the board of MPOG. Great team out there at UAB, so welcome again. I'm glad that you made the transition to

Epic and are now on board with your MPOG extract as well.

Before we get into Dr. Fisher's review, we did have just a couple of announcements or updates — and really a

question, too, actually — regarding PONV-05, the PONV prophylaxis measure. As most of you know, the latest

consensus guidelines came out last year. Dr. Gan from MD Anderson, who was a lead author for the consensus

guidelines, spoke at the MPOG retreat last year. So between the conversation that we've had at the Quality

Committee, the discussion that we had both with the panel and Dr. Gan's talk last year, and just reading through

the 5th consensus guidelines, we have a couple of updates that we want to make, and there is actually one

question that I'm going to throw out to the group as well. I want to get a little bit of feedback.

00:20 – Mei Calabio (MPOG) (via chat):

MPOG Measure Spec - PONV-05

00:20 – Nirav Shah (MPOG):

So the first thing is, we're going to add olanzapine as a qualifying antiemetic. It's mentioned in the guidelines,

and our colleagues at Yale — Dr. Jamie Hyman and others — did a really nice single-center randomized

controlled trial identifying olanzapine as an effective prophylactic antiemetic. So we'll add olanzapine.

Per previous discussion, we'll be excluding obstetric cases, only because they're going to be in their own

measure. Cesarean deliveries will be PONV-06-OB, just to acknowledge that there are some different guidelines

and protocols for patients undergoing cesarean deliveries.

We are also modifying the definition of propofol infusion, based on case validation that we've done and some

discussions we've had across MPOG. The definition of propofol infusion currently in PONV-05 is that if you even

have, like, one minute of a propofol infusion, it counts as an antiemetic. We're getting a little bit of noise with

that. We found that if we change the duration from one minute to greater than or equal to five minutes, some

of that noise gets reduced. So we'll make that change, and we'll make the measure just a little bit more

accurate.

And then finally, we're going to apply the “remained intubated” phenotype. We've tried different ways to

capture this concept: if the patient is going directly to the ICU intubated and sedated, then some of these PONV

prophylaxis rules should not apply — they should only apply to patients who are extubated at the end of the

case. The remained intubated phenotype has undergone some revisions and updates. That's essentially

completed now, and so we're going to apply the remained intubated phenotype. Overall, I think we'll make the

measure more accurate, and so we're looking forward to that as well. So those are the changes we're planning

on making.

And then PONV-03, which is the PONV outcome measure, is due to be reviewed at the September 2026

meeting. Dr. Tariq Esmail from University Health Network in Toronto will be doing that, so we're looking forward

to that.

00:23 – Nirav Shah (MPOG):

There's one question, or at least one point of discussion, that I don't think we've talked about at this Quality

Committee meeting before — maybe we talked a little bit about it at the last review, and maybe decided to wait

for the consensus guidelines and Dr. Gan's discussion at the MPOG retreat. Currently, essentially any dose of

propofol infusion counts as an antiemetic. There's some reasonable data out there to suggest that any dose of

propofol doesn't really move the needle in terms of reducing nausea and vomiting at the end of the case.

At Dr. Gan's talk last year, he presented some data which showed this. If you look at this graph, the x-axis is

propofol infusion dosing and the y-axis is essentially the efficacy of that dose. The higher the dose you get, until

you get to about 90 to 100 mcg/kg/min, the lower the rate of PONV. There are effects at lower doses as well,

but the dose effect starts to plateau at around 90 to 100 mcg/kg/min. And so what Dr. Gan mentioned in his talk

is that you really need a decent dose of propofol to have an effect; 0 or 5 or 10 or 15 mcg is probably not it.

00:25 – Nirav Shah (MPOG):

And what we're proposing is a minimum dose of 50 mcg/kg/min, still keeping the duration at greater than or

equal to five minutes, as a kind of floor for using a propofol infusion as an antiemetic. The thought, at least at

the Coordinating Center, is that at 50 mcg/kg there's definitely an effect — if you look at the graph on the right,

there's some effect — and the greater than or equal to five minutes captures, I think, the floor, or what the

minimum duration could be. You could make an argument to make it even longer. That does then potentially

cause some other issues, especially for shorter cases. But this seems like a reasonable compromise from our

perspective.

I wanted to get some thoughts from the group as well, to see if folks have any thoughts about having a floor for

dosage for propofol infusions. Again, this is based mostly on the evidence that Dr. Gan presented and some of

the studies that have come out; I don't think the consensus guidelines speak a ton toward minimum duration or

dosage, that I remember, but I could be wrong. So, anyway, with that, I'll pause for a minute and let folks take a

look at this, and then I'm curious to hear if folks have any feedback or comment on this floor for propofol

infusion dosage. I'm going to throw out a poll as well, just to hear what folks think about this. I'll keep an eye on

the chat as well.

00:25 – Mary McKinney (Corewell Health) (via chat):

[Replying to “MPOG Retreat - MPOG Registration is now open!”] Do you have the code required to register for

virtual attendance?

00:27 – Nirav Shah (MPOG):

Okay, not hearing so much feedback on this. Let me release this poll and see what folks think. Anybody can vote.

This is not just a one-vote-per-site item, anybody who has an opinion can vote on this one.

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

We dose lower than this at BCH and found in our single center that it was effective.

00:28 – Nirav Shah (MPOG):

Okay. And the poll – 35 folks voted, of which 74% agree with the 50 mcg/kg/min for five minutes; 9%, so 3 out

of 35, thought we should go with the higher dose; and 17% chose another dose. So we'll take this information

back — also to whoever did the PONV-05 measure review, I can't remember who it was, but we'll take it back to

that person as well. And before we make any changes, we'll make sure we announce this either via the forum or

at the next Quality Committee meeting as well.

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

Nirav, we did a small paper on this many years ago. It's supposed to be a sub-hypnotic dose — that's the way it

was originally proposed as an antiemetic — and so we take anything from about 10 mcg up to 50 as being sub-

hypnotic for antiemetic purposes. For that reason, we wouldn't meet this cutoff. I was very surprised but when

we at least looked at it, it looked like it actually does work.

00:29 – Nirav Shah (MPOG):

That’s kind of what informed the original inclusion of essentially any dose, any time duration, Morgan. But it

looks — especially from this, and I wonder if it's different in the pediatric and the adult population as well — it

does look, from at least this one dosing study from Anesthesia & Analgesia 2024, Sprung et al., that to get a

decent effect there has to be, maybe not a TIVA dose, but something more than sub-sub-hypnotic. Ketan, you

have your hand up?

00:30 – Ketan Chopra (Henry Ford – Detroit):

Do you think that this is something I should start educating my staff to do now? I know it's not official, but the

general practice with all of us is that everyone runs it at 10 to 15. So just to get used to it — what are your

thoughts? Should I be telling them, let's try 50 and kind of see how things go?

00:31 – Nirav Shah (MPOG):

In my own practice I have kind of stopped doing that sub-hypnotic dose of propofol, at least once we heard from

Dr. Gan and once he shared this. I've started using higher doses of propofol as an antiemetic, and so I personally

feel that, at least based on this data, I needed to bump up my dose from an antiemetic perspective. I don't think

that's a bad idea at all.

00:31 – Lilibeth Fermin (Cleveland Clinic) (via chat):

CMS proposed recently to remove the prevention of PONV from the

00:31 – Pavithra Ellison (WVU Medicine) (via chat):

We use low dose propofol infusions in most cases with an IV at our childrens hospital - min 50, usually at 100

mcg/kg/min

00:31 – Lilibeth Fermin (Cleveland Clinic) (via chat):

Anesthesiology metrics in 2027

00:31 – Nirav Shah (MPOG):

I see a couple of items in the chat. “CMS proposed recently” — this is Lilibeth Fermin — “to remove the

prevention of PONV from” something, but then I think it got cut off. Lilibeth Fermin, are you able to unmute?

00:32 – Lilibeth Fermin (Cleveland Clinic):

Yes, there's a CMS proposal to delete that quality measure from [unclear] for 2027. I think they're thinking that

it's like the standard of care nowadays, so there's a period for the anesthesiology teams to [unclear] a rebuttal.

00:32 – Nirav Shah (MPOG):

Yeah, so, interesting, because at least from our PONV data, we definitely see that PONV is still happening —

folks are still nauseated and vomiting in the PACU — and there's variation across sites as well. We'll share some

of that data at the September meeting for the review. It's surprising to me that they're removing it as a quality

measure, because it's a very patient-focused measure. Now, it could be that in the data they have, which may

include both automated extracts and self-reported data, there's not a lot of variation in PONV. But we know

from our own data, which is based on PACU medications and nursing notes, that there is variation in

postoperative nausea and vomiting across sites.

00:32 – Pavithra Ellison (WVU Medicine):

We use low-dose propofol infusions in most cases with an IV at our children's hospital, minimum of 50. That

coincides — I think we independently came up with that 50 number.

00:33 – Amy Poindexter (Holland) (via chat):

Have you looked at how this change would effect PONV scores?

00:35 – Meridith Wade (MPOG) (via chat):

[Replying to “Have you looked at how this change would effect PONV scores?”] We are in the process of

building a propofol infusion phenotype with these parameters and plan to add to PONV-05. We will

compare the score impact prior to publishing the measure revisions

00:34 – Chris Wedeven (Holland):

I do think it's worth looking at, and there probably is some minimum dose — and I voiced this at our meeting

last week — but I think with the elderly patients, would it be a compromise, maybe? Because it looks like the

effect starts at 30, and between 30 and 100 there's some efficacy between those numbers. I wonder, for the

elderly folks, if I could propose or maybe think about 30 being the floor instead of 50, because I think 50 is

almost an anesthetic dose for a lot of elderly folks.

00:34 – Nirav Shah (MPOG):

I think that's a great point, Chris. We've already had some issues regarding polypharmacy and so on in the

elderly that have been brought up regarding PONV-05 — trying to get that third antiemetic for a high-risk

population can be challenging. We will go back and think about the implications for the elderly.

00:35 – Nirav Shah (MPOG):

Dexmedetomidine. This has come up, a couple of times: should we include dexmedetomidine? We did a brief

review of the literature, which was highly variable both in terms of efficacy and the dosing required, and the

overall quality of evidence did not seem that great. The fifth consensus guidelines do mention it in the text, but

they actually don't include it in the table of included antiemetics. And so for that reason, until there's better

evidence out and more consensus among the experts , we decided to leave it off PONV-05 as an included

antiemetic. Definitely open to reconsidering that once there's better evidence and maybe more discussion

among the experts in the field.

Okay, with that, I'm going to pass the baton to Dr. Fisher, who many of you know, from Yale.

00:37 – Clark Fisher (Yale):

I'm Clark, I'm at Yale. Thank you for inviting me. I've worked with the research arm of MPOG a fair amount in the

past; I'm happy to be here on the quality arm, which I've been a consumer of but haven't contributed to.

So this is the opioid, or OME, measure that Mike Burns created a number of years ago and has kept updated. I

think there are a lot of different ways to potentially go with this, so I'm going to try to keep this streamlined, but

I'm open to talking about a wider array of things.

Currently, this is an informational measure. It's just saying, essentially, what the total dose of opioid is across a

number of different case types, with MPOG phenotypes, which you can see here. What it does is take all the

opioids administered during these procedures between anesthesia start and end, and try to convert them to

standard oral morphine equivalents.

00:38 – Clark Fisher (Yale):

This is just a little bit of the table that Mike Burns assembled, showing both the conversion factors to oral

morphine equivalents as well as some of the sources — and you can find this table online. About 80% of the

different conversion factors come from one of three sources: either the CDC from 2016, the American Pain

Society guidelines (which they have actually stopped updating, but the most recent version), or, in a few cases,

OpenAnesthesia. And beyond those 80%, there are a few other sources for different, mostly less common,

opioid medications.

And here is the high-level view of what the output looks like. I do have some questions about some of these

numbers, which lead me to believe maybe some of these calculations are not spot-on, but this is something, I

think, to cover in the review.

00:38 – Meridith Wade (MPOG) (via chat):

Full measure review - OME - Clark Fisher Review 7.27.26 - Google Docs

00:39 – Clark Fisher (Yale):

First, just reviewing the new data since the last update three years ago. I did a combination of PubMed and

Google and some new AI tools like Consensus. The vast majority of the literature looking at putting all opioids

together into one single value —what we call equianalgesic opioid conversion — just says that it's a very difficult

problem that no one agrees upon how to do correctly. This is a sampling of that sort of literature. Pretty much

anyone who's tried to look at how different hospitals and different specialty groups compare and combine

opioids comes to the conclusion that it's largely done by expert opinion, and expert opinion varies depending on

which experts you ask.

There's a lot of diversity, which I think speaks to the importance of keeping a measure like this at MPOG,

because I think anesthesiologists have used opioids in a relatively unique way compared to a lot of other

doctors, and having a standard that anesthesiologists can reference about opioid exposure is valuable.

00:40 – Clark Fisher (Yale):

There are a few, I think, important things which systematically change how we can get toward better consensus

on these things. First, and I think most important here, is this paper, which came out in 2025. This was an

initiative within HEAL at NIH to create, essentially, a consensus equianalgesic calculator for oral morphine

equivalents. This is based off the CDC guidelines, which form the backbone of our own conversions. It does

pretty much only handle oral formulations and some things like fentanyl patches; it doesn't really include any IV

formulation, so it's limited. Another nice thing it does is that they try to grade the quality of evidence behind

each recommendation, which, as I alluded to previously, tends to not be very high — and we can get into

reasons why that is.

00:41 – Clark Fisher (Yale):

The other interesting systematic approach to this that I found actually came out just before the last review; it

wasn't covered in there. There was an attempt to take all of the best available data about some of the IV opioid

formulations, and they performed a meta-analysis looking across the different opioids and using the data not

just from head-to-head comparisons, but doing the network meta-analysis approach, to sort of squeeze all the

juice they can from these. This is 52 randomized controlled trials, all focusing on PCA patients, mostly acute pain

post-surgically, and comparing head-to-head different types of IV opioids and coming up with equivalent doses.

I'll show you some of the results from this here. This is an overall plot showing the comparisons they worked

out. You can see some comparisons — they did a lot of studies to compare between morphine and tramadol,

and morphine and pethidine, interestingly; for morphine and fentanyl and others, it's more like one or two

studies. And these were the potency ratios they came up with.

I think this is a very interesting study in a field which is mostly based on expert opinion and vibes. This was a

really grounded approach to try to use only randomized trials, put them all together, pool them. It does lead to

some interesting problems. One of them is that there's a relatively high level of inconsistency that they found

within this meta-analysis, so when they pooled this data and came up with these conversion factors, they were

actually different by a fair amount from the conversion factors from each individual study.

Also, these conversion factors are pretty different from conversion factors that people generally accept readily.

Fentanyl is usually a conversion factor of 100 relative to morphine, and here it's 60, so it's a pretty big change

from what the rest of the world is using. So, just some things to consider, but I thought it was worth bringing up.

00:43 – Nathan Pace (Utah) (via chat):

The third author of that paper, Rucker G, is an outstanding statistician who developed many network meta

methods.

00:43 – Clark Fisher (Yale):

And then the last thing I'm going to mention, which is, I think, why I was invited to do this, is that there are other

approaches to looking at opioid equianalgesia beyond trying to convert a dose of one drug into a dose of

another drug. And there's a very small literature — one of them, the second one, is my paper, and this other one

from Santa Cruz Mercado — where it was trying to use pharmacokinetic models and think about drugs that are

equivalent in the concentration sense, like a bloodstream and effect site concentration sense, rather than a dose

sense. And I'll talk a little more about that, and where I think this might be helpful.

00:44 – Clark Fisher (Yale):

I want to take a step back. Going through the format of the written review, one of the questions to ask is, is this

still a valuable measure? I think it is. Since the last review, four studies have come out of MPOG looking at opioid

use, I still think it's something people are interested in. I still think it should only be informational, because I

don't think there's consensus about how much opioid is the right amount of opioid.

I think the question comes down to: are we pooling all of our opioids together in the right way? And just at a

very high level, I think an ideal opioid measure for MPOG would be supported by strong evidence. It would be

easy to understand, because our primary audience here is anesthesiologists — it shouldn't be a strictly academic

measure; it should be something that makes intuitive sense and that people can talk about and, if they choose

to, change their practice based on. And it should apply specifically to our context, which is the operating room,

and therefore handle the drugs that we use in the operating room.

So, just to spoil this: I don't think I have a measure that meets these criteria. There isn't one supported by strong

evidence in any case, and with some of the other ones I think there are trade-offs. I've proposed a two-track

model here.

00:46 – Clark Fisher (Yale):

Going back to talk about one of the issues with morphine equivalents in our field: it raises the question of what

an equivalent dose of opioid is. And where we run into trouble in anesthesia is that we're not dealing with

patients who are in a steady state of analgesia. We're not trying to maintain some state from pre-induction to

outside of the operating room in most cases, as opposed to people who are dealing with either long-term oral

opioid therapy, or even PCAs, where there are relatively steady levels over a course of hours.

This is a demonstration highlighting this problem. When you give a dose of opioid, the effect follows a specific

time course, and the time course depends on the opioid. This is what the effect of an IV bolus of fentanyl looks

like over a period of one hour, and this is the time course of IV morphine effect over the time period of an hour.

If you're asking, what's the equivalent dose of 100 mcg of fentanyl in morphine? If you're talking about peak

effect, and you don't care about when it is, maybe 15 milligrams of morphine is right. If you're talking instead

about average effect, maybe it's only 10 milligrams of morphine. If what you're interested in is intubating the

patient five minutes after you gave your bolus dose, you really need 40 milligrams of morphine. And these are

all reasonable definitions of equivalent dose in our context — which points to equivalent dose not really being a

specific thing.

There's a further problem, which is that when you talk about equivalent doses in terms of oral morphine

equivalents, those are all measured in a specific context, be it a patient who's in a steady state on a PCA or a

patient who's taking oral morphine, and those doses might not hold up under a different context like we see in

the operating room.

00:47 – Clark Fisher (Yale):

To get a little more into the weeds here about this other way you can deal with this, which has its own issues:

one thing that's much more steady in terms of opioid than dose is effect site concentration. At a given effect site

concentration — or, to make it simpler, think about plasma concentration — opioids have relatively steady,

comparable strengths. And this is just showing one way of measuring this, which is EEG spectral edge at

different concentrations of different opioids. For fentanyl, we have a number of lines which show different

studies, but they all sort of cluster together. What this is doing is separating out the pharmacokinetics — how

the drug moves through the body — from the pharmacodynamics, the effect the drug has once it's acting at the

receptors.

00:48 – Clark Fisher (Yale):

I'm going to lean on my own past work to talk about how we could use this to do other measures of opioid

exposure. This is one sample case here. You gave a bolus of 100 mcg of fentanyl at 11 minutes, you turned on a

remifentanil infusion at 12 minutes from start, and you turned it off at 98 minutes. We can use pharmacokinetic

models to see what the effect site concentration of these things is. Fentanyl — we already saw this graph —

peaks and comes down after the bolus, and remifentanil sort of follows this time course with the infusion.

Based on that last graph I showed you, we can put fentanyl and remifentanil on a standard plot. Remifentanil is

less potent than fentanyl when you measure it in terms of concentrations, so we can put it in terms of fentanyl

equivalents here, and it sort of shrinks. And then, we can add these things together, since they're all acting at

the mu receptor, and do whatever we want with it. We can measure the peak effect at a given time. I, in my

study, measured the area under the curve here to look at total opioid exposure.

00:49 – Clark Fisher (Yale):

These are sort of all the background; so now let's get to recommendations. Given that one of the goals is to be

able to have a measure that's easily understandable, and the standard everywhere is oral morphine equivalents,

I think it makes total sense to continue to use and update our oral morphine equivalent measure, and make it

the best oral morphine equivalent measure we can.

The first recommendation I have is that we update our conversion factors to agree with the NIH HEAL standard

that was just published. And these are relatively modest updates. I've highlighted the changes in the multipliers

here, and you can see, like, for hydromorphone ours is 7.50, HEAL suggests 6.00; for methadone, ours suggests

6.00, HEAL suggests 6.38. So these are relatively modest changes that I don't think will drastically upset anyone's

expectations, and they put us in line with, I think, a good consensus statement.

00:50 – Clark Fisher (Yale):

The trickier thing comes with the mass of opioids that we use, which are IV, so these have a much bigger impact.

And the question is, do we stick with what we have, or do we change to — I think the only real systematic

approach has been that network meta-analysis I showed. I alluded to the fact that the numbers are fairly

different, and this shows them side by side here. And it was published now four years ago, and there hasn't

been a lot of uptake in the wider community. My own approach would be to hold off on this for right now,

maybe reassess this in three years. And there has been some discussion at the end of the HEAL consensus about

including IV drugs in the future, and I think we keep an eye on that.

The one IV drug I think we should change is our IV buprenorphine. I think it was probably just a typo that we

have a multiplier of 75 as opposed to 0.4, and I traced that back to the original reference in our old guidelines to

confirm that's true. But I would suggest largely leaving these alone and reassessing in the future, otherwise.

00:52 – Clark Fisher (Yale):

And then, I think, in parallel — and this depends on resources and interest at the coordinating site — I think it

would be valuable to pilot this sort of alternative measure of using pharmacokinetics and effect site

concentration conversion as a different measure of intraoperative opioid use.

And because this is fairly involved, I would try to do this first in the simplest way possible, which is to use all the

same inclusion criteria. There are a lot of choices along this path in terms of what specific pharmacokinetic

models you use, and who you choose for your specific conversion factors. And to start with, I would say that we

lean on the work of other people. Steven Shafer of Stanford, and in general a leader in anesthesia, continues to

maintain stanpumpR, which has his own choices for these, so I would say we start with those for right now.

The upsides of this, are that I think it gracefully handles bolus dosing, rapid changes in titration, and things like

context-sensitive half-time, which are not at all addressed by MME. And it lets us answer interesting questions,

like what's the overall opioid level at the time of PACU handoff, in the future. Downsides are that it's not a

widely understood measure, and it's computationally a lot more involved. My proposal is that we give this a

shot, see if it's reasonable to do, and then continue to follow it up the next time.

One more recommendation, which is that at the last review, three years ago, it was agreed that we should

create another measure which looks at OME given in the PACU. And I think it was just a matter of priorities that

this hasn't been done, but I think it would still continue to be valuable, and let people know, once their patients

are dropped off, how much additional opioid is being given.

A note for the future: three years from now, what are possible things to think about? One lack that I ran into

with our current measure is that remifentanil is counted as no opioid whatsoever. With the pharmacokinetic-

driven model, remifentanil is handled very elegantly and it's not a problem, but I think it's also reasonable to put

in an MME value for remifentanil, and different people have different ones. That network meta-analysis that I

referenced presents one. I think that's something to discuss in the future.

On the other hand, while we have no MME value for remifentanil, we do have MME values for neuraxial opioids,

which are mostly referenced to OpenAnesthesia articles, and I don't think there's a lot of consensus around that,

so I'm not sure how valuable that is. I'm interested in people's thoughts on that in the future.

00:54 – Nathan Pace (Utah) (via chat):

Eleveld models?

00:55 – Clark Fisher (Yale):

One thing that came up during the last review, I know, is that tracking outpatient medications in MPOG has been

a long-time headache and a long-time goal, and we're making progress towards it. If that comes to reality, I think

it's valuable to know if people are taking opioids when they come in for their surgery, and it can help us

understand the opioid they get during their surgery.

Another thing that's moving along within MPOG more widely is that there's a lot more standardization of

phenotypes for different procedure groups, I think we have an opportunity here to look at which procedure

groups are reporting out — ones like, perhaps, transplant — that could benefit from this sort of measure. And

then, as I said earlier, assuming that we get some traction with this new pharmacokinetic-derived model, I think

the next review will be a time to really see what works, what doesn't, and what we can do in the future there.

00:56 – Nirav Shah (MPOG):

Awesome. Thank you. That was fantastic. I do have a couple of questions, but I did want to open it up to the

floor. Nathan, you just put in the chat, “Eleveld models.” Do you want to just share a little bit about what that is?

00:56 – Nathan Pace (Utah):

There are many, many models. This is some recent ones. Some studies show they're superior to some other

models. There are so many choices — I totally agree with the idea of standardizing on some model. At a certain

point, since we titrate to effect, it doesn't matter.

00:56 – Clark Fisher (Yale):

If we move away from OME — the problem with OME is there are so many choices for conversion factors, and

moving to pharmacokinetic models actually just means more choices, because you have your choice of models

and then your choice of conversion factors. It doesn't get rid of that. But I agree entirely, and I think the first

step would just be choosing something reasonable, and letting Steve Shafer choose for us was my proposal. And

then adjusting as we recognize what are undoubtedly issues with the initial choices.

00:57 – Michael Mathis (MPOG):

Clark, I think this is fascinating, and I really applaud your efforts to take a deep dive into this. I do agree that this

should be an informational measure — there's not any value judgment on this — but it's useful to have an

awareness of providers and sites that have variation in how much opioid they're giving.

I think one of the challenges with designing this measure is that anesthesiologists use opioids for different

reasons, and you have to think about the reason you're giving the opioid to have the optimal measure design. In

a cardiac case, you might give a ton of opioids not for postoperative analgesia, but maybe just for blood

pressure stability during the case. Or you might want to give it during a MAC case to supplement the depth of

anesthesia. Or you might have remifentanil running to minimize patient movement if you can't give muscle

relaxant, for example.

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

Thoughts about adding in time to OME? I.e. OME per hour of case?

00:58 – Michael Mathis (MPOG):

Or, closer to the original reason for designing OME, you might be giving opioids genuinely to have a patient who

has good analgesia in the PACU. And so those different reasons imply different ways of displaying the

information to anesthesiologists on our dashboard.

If we're putting the blood pressure thing and the adjuvant-to-sedatives kind of measurements aside, and just

focusing on the analgesia part of this, I would argue that you probably most care about analgesia when the

patient's awake, which is at the end of the case as they're going to the PACU or ICU. It might be very interesting

— in that sense, in a cardiac case, it's very different to give a milligram of fentanyl right at the beginning of the

case before incision versus a milligram of fentanyl, or 10 milligrams of morphine, at the end of the case as you

go into the ICU. And similarly, in a case where you're running a remifentanil infusion and then turning it off, that

opioid exposure is very different from an opioid exposure given for intubation, versus analgesia in the PACU.

I love the idea of being ambitious and keeping a simple OME measure that's useful and easily understood by

everyday practicing anesthesiologists, but then also just being ambitious and doubling down on this OME

measure. What I think would be most interesting for the OME measure is not looking at the total exposure over

the course of the case, but actually, what's the effect site concentration when the patients go into the PACU?

Because that's when the patient has a conscious experience and is actually experiencing pain — for at least for

cases under general anesthesia. And that might be very interesting: seeing, again, a patient that got

remifentanil, and then that was turned off and got no other opioid — that patient's going to have a very

different PACU experience from a patient that got, you know, 5 milligrams of morphine at the end of the case.

And so—I'll stop there.

01:00 – Tariq Esmail (UHN) (via chat):

Great points Mike. Also Chronic Pain Patients who have high 24 h requirements at baseline when they come for

large surgeries (eg. spine etc) Opioid amounts are going to vary significantly to a patient with the same surgery

but opioid naive.

I have to log off, sorry. Really appreciate the review Clark!!

01:00 – Nirav Shah (MPOG):

Thanks, Mike. So, Clark, were you envisioning, in the pilot component of this, that the number you're going to

share with the provider or the quality champion would be the concentration — like the median concentration,

the average, into the case? What were you thinking?

01:01 – Clark Fisher (Yale):

Mike — and I want to combine my response to you with Morgan Brown's question about thoughts on adding in

a time-normalized OME, which actually, I think, the current measure does present OME in a number of different

ways, depending on the — it can be normalized by time, normalized by patient weight, that sort of thing.

Yes — Mike, I agree entirely that one of the challenges of this is that people talk about opioid dose, and their

goals for opioid dose are about very different things in different cases. It depends on how they're using it, which,

as you said, is for very different reasons, even maybe within the same surgery.

My initial thought, being cognizant of the fact that it's a fairly big lift just to get it off the ground, was to try to

come up with as close to the equivalent of our existing OME as possible when designing this PK/PD measure for

the first time. And to me, that means opioid exposure during the case. There's no exact equivalent, but it's just

overall, how much mu receptor activation has this person gotten?

But I think the benefit is that once we've done that work, it's really easy to answer other questions about opioid

during the case which are impossible to answer with OME. There's no way with OME to answer how much

opioid effect this person has when they're dropped off at the PACU, or how much opioid effect they have when

they get their first rescue dose in the PACU, which is another interesting question, right? There's no way with

OME to answer how much opioid effect am I giving to my patient before I do laryngoscopy compared to my

colleague?

And these are all — once you've sort of made these graphs and done the modeling, it's just picking different

points on the graphs, or if you want to tell how much per hour, you take the average as opposed to the end of

the curve. The math becomes very simple once you've done the modeling. And then I think if we get the

modeling right, the challenge will be choosing which ones to present and how to present them, because it's just

an incredibly rich amount of data. But yeah, my hope would be that if this works, we can go down all these

pathways and figure out what's most useful and the best practice for everyone.

01:03 – Nirav Shah (MPOG):

I mean, I think what you're proposing, Clark — and this is what this review is demonstrating — is that it's

interesting for this Quality Committee to think about this measure. This is not, like, the practical “how am I going

to implement this in my practice”; this is almost like putting your learning health sciences hat on, where we take

what we learned from this, the variation that we're seeing across sites, and plug that back into the research arm,

both for additional observational analyses as well as maybe some pragmatic clinical trials here, to then be able

to figure out what the actual recommendation is — or what we're seeing that we can share with quality

champions and providers across MPOG that could potentially change their practice. Because what we're really

talking about is not how you are going to change your practice, but really just, like, enlightening us about what

our practices are, and I think that's the hat that we should think about this with.

You know, Morgan, your statement in the chat, which is thoughts about adding in time to OME — like, OME per

hour of case — what you've actually uncovered is the fact that this current informational measure, the way it's

displayed in the QI Reporting Tool, is actually not working as intended. When we moved the QI Reporting Tool to

the new platform, the way that this measure works kind of broke, which is why what you're seeing now is not

completely accurate, and you're not actually able to dive into the measure summary screens to look at the

opioid administration per hour equivalents that we developed. What you're supposed to be seeing currently is

essentially, for an average person and an average duration case, what the average opioid equivalence was in

morphine equivalents — and you're not quite seeing that.

I think part of what we would like this Quality Committee to comment on via the poll is, first, do we actually

want to continue sharing opioid equivalents in the way that Dr. Fisher described for what you have right now,

which is these procedure groups — spine, cardiac, abdomen, hip, knees? And then, number two, should we

think about this pilot where we look at opioid concentration? And from number two, I absolutely think we

should do this. We can make the pilot somewhat restricted by a very limited amount of maybe even, like, one or

two procedure types. And we can use maybe just, like, fentanyl and morphine — limit the number of

medications. We also have to see what's available. And maybe we could just look at that and see if this is

computationally feasible for MPOG, and if the answer is yes, and if it's something that we can then share on the

dashboard, then we can come back —we don't have to wait three years; we can even come back sooner than

that, before the next measure review, and just share some details on that and decide if folks want to further

explore that. And not to mention, if there's a great research use case for it, then that's another pathway as well.

Anyway, so I'll stop there. Other comments from folks?

01:07 – Sarah Zhao (MPOG) (via chat):

Is “stanpumpR.io” package specifically designed for OME study?

01:08 – Sarah Zhao (MPOG):

I'm asking this question because I tried this package before in my study, and I found that in this package

there's only — it only includes about seven medications which are OME medications. And it also

includes other medications which are not OME. I found that in this package it only included fentanyl,

hydromorphone, morphine, oxycodone, and another two — methadone, alfentanil, sufentanil —

included in this package. But in my study, I included some other medications also. Let me give you some

examples. So, [unclear], opium, hydrocodone, and naloxone. So some of that is not covered in this

package, which means that if I want to do some study including all these opioid medications, this

package might not work, or it will give me inaccurate results. So do you have the same issue, or...?

01:07 – Michael Mathis (MPOG):

You know, I'm sold on that. I like the idea of just getting a foothold, as Clark, you're alluding to, about this effect

site concentration pharmacokinetic model. In terms of computational feasibility, it's feasible. Clark did the study

in the cardiac surgical patient group within MPOG, and so that might be a good subgroup. You probably want

some group where the patient's awake at the end of the surgery also to be part of this kind of preliminary

piloting. Clark, you have the code, and you've done it, and—

01:07 – Clark Fisher (Yale):

Yes, it is doable. It's more within the workflow and the resources devoted to it. I mean, it is one measure among

many, as opposed to the paper I was spending 60% of my time working on, right?

01:08 – Clark Fisher (Yale):

It's a different context. The math is doable; computers can do it.

01:08 – Nirav Shah (MPOG):

Allison, you may have been thinking this as well — I meant from a monthly processing basis.

01:08 – Nirav Shah (MPOG):

I know we only have a few minutes. Any other final comments, before I launch the poll? There actually are a

couple of polls related to this.

01:08 – Clark Fisher (Yale):

Tariq signed off, but was mentioning chronic pain patients who have different requirements, and yes, I think if

we do have outpatient medicines figured out, that will be valuable to break out in the future. And Sarah, are you

able to provide more context for your question? I can talk a little bit about the stanpumpR software, but I'm not

sure I understand the question.

01:10 – Nirav Shah (MPOG):

Sarah, thank you. Just in the interest of time, I want to launch the poll.

The first poll is — folks should be seeing it. Yeah, great. Do we modify it? Do we keep it as is and just fix it? And

then, do we retire it? I'm going to end these polls relatively early, so I want folks to answer as fast as they

possibly can, in the interest of time. [unclear] in five seconds or so. This one is one vote per site, the usual.

We have a quorum here, so I'm going to share the results. This is overwhelmingly voting to modify it, which is

awesome.

I do want to go into a little bit more detail here. This is the last poll that we have. It's a four-question poll, and

this, Clark, gets into your individual questions. And really, I think the first two can be lumped together, one and

two, if folks don't mind. And then the third question, which is, do we extend it into the PACU — which folks

wanted to do originally, but since we have the benefit of another review and still haven't been able to do it, for

some good reasons, in the last few years, I wanted to make sure that folks still found that PACU measure

valuable. And then the fourth question is just around this pilot: are folks conceptually interested in the pilot?

And then, Clark, if the answer is yes, which I hope it will be, then we can circle back afterward and maybe talk

about some of the implementation details.

01:12 – Nirav Shah (MPOG):

And you had also shared this measure review with a couple of other experts around the country — so maybe we

can pull some of those folks together and come up with a plan to implement, and what a pilot could actually

look like.

01:13 – Clark Fisher (Yale):

Absolutely.

01:13 – Nirav Shah (MPOG):

I'm going to end the poll here and share the results. What are we seeing here? Yes to use the HEAL standard. Yes

to reassess with the Dinges conversions at the next review. Yes, people are still interested in the PACU-based

measure. And then yes to create the pilot measure based on the PK models. Okay, so thank you, everyone.

I know we're at 11 o'clock. There were a couple of things that we were planning on going through that we will

wait till next time, but I did want to mention them. One is that the Glycemic Management Workgroup is meeting

in a couple of weeks. One purpose is to follow up on some of the work from Dr. Duggan and Dr. Henson at the

May meeting, and the second is to review a proposal to streamline our entire glycemic management bundle,

which is maybe a little bit confusing, to make it a little bit more understandable. So if folks are interested in

joining that, I believe that meeting is August 13th. Please reach out to me or anyone at the Coordinating Center,

and we'll make sure you're included on that.

And then the final thing is that we have made some updates to the concept browser, phenotype browser, and

the measure browser, where you see all the measures, on the website. We have consolidated that into a single

application. That's going to be released in a couple of weeks. We will share that message via the forums and

email. The main difference that folks need to know about is that, in order to view it, you'll have to log in — the

same way that you log into the QI Reporting Tool and DataDirect, you'll have to log into this to view the concept

browser, the phenotype browser, and the measures. If folks have any questions on that, don't hesitate to reach

out to me, or anyone at the Coordinating Center, and we can walk you through it, but those instructions should

be pretty straightforward, and that announcement should be coming out in a couple of weeks as well.

So with that, I will conclude. We're a minute over. Thank you, Clark. Thanks to everyone at the Coordinating

Center for putting all this together. Everyone have a happy Monday and a great rest of the week. Take care,

everyone. Bye.

Clark, thank you so much. That was really, really good. I appreciate it.

01:15 – Clark Fisher (Yale):

I sent Sarah my email, so hopefully she can follow up with me. Thanks a lot, Nirav.

01:15 – Nirav Shah (MPOG):

If anyone reaches out, I will forward them to you as well. Perfect. Thank you. Take care. Bye.