MPOG QI – Quality Committee Meeting Notes – Monday, May 18th, 2026
Attendance:
Abess, Alex (Dartmouth)
Goatley, Jackie (Michigan)
Pantis, Rebecca (MPOG)
Agerson, Ashley (Spectrum)
Goldblatt, Josh (Henry Ford Allegiance)
Pardo, Nichole (Corewell Grosse Pointe)
Bartels, Karsten (Michigan)
Gostic, Will (Stanford)
Paul, Jonathan (Columbia)
Baudo, Andrew (MyMichigan)
Grewal, Ashan (Maryland)
Poindexter, Amy (Holland)
Bauza, Diego (Weill Cornell)
Hall, Meredith (Bronson Battle Creek)
Qazi, Aisha (Corewell)
Berndt, Brad (Bronson)
Heiter, Jerri (Trinity Health)
Riggar, Ronnie (MPOG)
Bollini, Mara (WUSTL)
Henson, Patrick (Vanderbilt)
Ruiz, Alexandra (Trinity Health)
Bowman-Young, Cathlin (ASA)
Huntington, Michelle (Corewell West)
Ruiz, Joseph (MD Anderson)
Brown, Morgan (Boston Children's)
Hyman, Jamie (Yale)
Service, Alex (Indiana)
Buehler, Kate (MPOG)
Janda, Allison (MPOG)
Schwerin, Denise (Bronson)
Calabio, Mei (MPOG)
Johnson, Rebecca (UMHS West)
Scranton, Kathy (Trinity Health St. Mary's)
Cassidy, Ruth (MPOG)
Kaper, Jon (Corewell Trenton)
Shah, Nirav (MPOG)
Charette, Megan (MPOG)
Karamchandani, Kunal (UT Southwestern)
Shettar, Shashank (OUHSC)
Chopra, Ketan (Henry Ford - Detroit)
Kaushik, Teshi (UAB)
Smiatacz, Frances Guida (MPOG)
Clark, David (MPOG)
Khan, Meraj (Henry Ford)
Smith, Crisstina (Mid West Anes)
Claybaugh, Deborah (MyMichigan)
Kirke, Sarah (Nebraska)
Stierer, Tracey (Johns Hopkins)
Cohen, Bryan (Henry Ford - West Bloomfield)
Kunkler, Bryan (Corewell West)
Stumpf, Rachel (MPOG)
Coleman, Rob (MPOG)
Lacca, Tory (MPOG)
Szymanski-Bogart, Brooke (MPOG)
Corpus, Charity (Corewell Royal Oak)
Lalonde, Heather (Trinity Health)
Tyler, Pam (Corewell Farmington Hills)
Couture, Lindsey (Vermont)
Liu, Bin (Michigan Medicine)
Vaughn, Shelley (MPOG)
Cywinski, Jacek (Cleveland Clinic)
Lewandowski, Kristyn (Corewell Troy)
Vitale, Katherine (Trinity Health)
Delhey, Leanna (MPOG)
Lu-Boettcher, Eva (Wisconsin)
Wade, Meridith (MPOG)
Duggan, Beth (UAB)
Mathis, Mike (MPOG)
Wedeven, Chris (Holland)
Edelman, Tony (MPOG)
Mack, Patricia (Weill Cornell)
Weinberg, Aaron (Weill Cornell)
Finch, Kim (Henry Ford Detroit)
Malenfant, Tiffany (MPOG)
Wilson, Blake (MyMichigan)
Gedela, Radhika (University of Vermont)
McCaughan, Michael (Sparrow)
Woody, Nathan (UNC)
Georgiadis, Paige (University of Vermont)
McKinney, Mary (Corewell Dearborn / Taylor)
Yuan, Yuan (MPOG)
Gereboff, Avner (Cedar Sinai)
Mentz, Graciela (MPOG)
Zhao, Xinyi (Sarah) (MPOG)
Gibbons, Miranda (Maryland)
Milliken, Christopher (Sparrow)
Zhu, Shu (Columbia)
Gilliland, Samuel (Colorado Anschutz)
Mirizzi, Kam (MPOG)
Zittleman, Andrew (MPOG)
Glanding, Kimberly (UAB)
O'Dell, Diana (MPOG)
Agenda & Notes
Opening, Attendance, and Minutes:
• Meeting Start: 10:01 am
• Roll Call: Via Zoom or contact Coordinating Center (support@mpog.zendesk.com) if you were
present but not listed on Zoom.
• Minutes from February 2026 Quality Committee Meeting
Upcoming Events:
2026 Meetings & Events
• ASPIRE-only Meeting: Friday, July 17 (Weber’s Hotel, Ann Arbor).
• MPOG Retreat: Friday, October 16 (San Diego).
• Sites outside Michigan are welcome at the Michigan in-state meetings — contact the
Coordinating Center for details.
Announcements:
• Featured Member (May/June): Dr. Michael Avidan (Washington University, St. Louis) —
longtime MPOG member and early supporter of the collaborative. Dr. Avidan also helped launch
the THRIVE PCORI trial examining general versus inhalational anesthetics.
• University of Vermont — Newest (Returning) MPOG Site: University of Vermont was one of the
original MPOG sites and has rejoined after a brief hiatus. Welcome back!
◦ Chair of Anesthesia and Principal Investigator: Dr. Rebecca Aslakson
◦ Anesthesia IT Champion: Dr. Brian Waldschmidt
◦ Quality Champion: Dr. Paige Georgiadis
Measure Review #1 — GLU-12 & GLU-13, Reviewer: Beth Duggan, MD (University of Alabama at
Birmingham)
Review Document: GLU-12/13 – Beth Duggan Review 05.18.26
Background & Rationale:
• Perioperative hypoglycemia is relatively low-incidence but potentially catastrophic, particularly
when unrecognized in an anesthetized patient who cannot report symptoms.
• Inpatient target blood glucose range is generally 100–180 mg/dL. Hypoglycemia is classified as
BG < 70 mg/dL; severe hypoglycemia as BG < 54 mg/dL (ADA Level 2, 3.0 mmol/L) per ADA 2026
Standards of Care.
• Evidence supports that intraoperative insulin administration is not independently associated
with hypoglycemia; rather, failure to recheck after treatment is the primary risk factor.
Treatment of hyperglycemia is safe and necessary; omission leads to worse outcomes.
• The MPOG registry trial confirmed that intraoperative hypoglycemia (BG < 60 mg/dL) was
uncommon and not associated with intraoperative insulin use, reinforcing the safety of
appropriate treatment when paired with timely rechecking.
Proposed Measure Modifications:
• GLU-12 (Hypoglycemia Treatment, Intraop): Retain the current criterion of IV dextrose-
containing solution within 30 minutes for BG < 70 mg/dL. Propose adding a new success
criterion for severe hypoglycemia (BG ≤ 54 mg/dL / 3.0 mmol/L): IV dextrose-containing solution
within 15 minutes.
• GLU-13 (Hypoglycemia Management, Periop): Separate treatment documentation from
recheck documentation into distinct criteria to allow more targeted quality improvement
interventions at the site level.
• Oral carbohydrate solutions (juice, ginger ale) are acceptable treatment in awake PACU patients;
MPOG can capture these if correctly mapped in the EHR.
Inclusions / Exclusions:
• Inclusions: Any case with an intraoperative BG measured between anesthesia start and end
time; any case with a postoperative BG in the PACU.
• Exclusions: ASA Class 6, cases without a recorded BG value, labor epidurals and non-operative
obstetric cases. OB-specific metrics will be developed separately.
Future Considerations (not voted on):
• Obstetric patients (maternal hypoglycemia carries fetal risk) — dedicated metrics forthcoming.
• Recheck measures following treatment to detect persistent or rebound hypoglycemia.
• Continuous Glucose Monitor (CGM) / Automated Insulin Delivery (AID) device integration into
perioperative metrics; >50% of Type 1 diabetics now use AID systems.
• Type 1 vs. Type 2 diabetes distinction may warrant separate metric pathways.
Discussion Highlights:
• Dr. Xan Abess (Dartmouth): Raised concern about measure fidelity when adding multiple
granular sub-metrics. Noted the risk that providers under a 15-minute documentation window
may back-document compliance rather than change behavior. Questioned whether increasing
measure complexity risks loss of departmental buy-in.
• Dr. Tony Edelman (MPOG): Acknowledged the debate on timing. Group consensus was that BG
< 54 mg/dL represents a critical value deserving a distinct, shorter response window —
analogous to MAP thresholds for intraoperative hypotension.
• Dr. Beth Duggan (UAB): Emphasized that the goal is behavioral: distinguishing a BG of 40 from a
BG of 69 in terms of urgency. Affirmed that treatment-related hypoglycemia is uncommon and
should not deter providers from administering insulin when clinically indicated.
• Dr. Josh Goldblatt (Henry Ford – Allegiance): Strongly supported separating treatment from
recheck metrics. Henry Ford currently treats BG < 70 with 0.5 amp and BG < 50 with 1 amp of
dextrose; policy threshold is being updated from 40 to 50 mg/dL to align more closely with ADA
guidelines.
• Dr. Tariq Esmail (Toronto) via chat: Noted that BG ≤ 54 mg/dL = 3.0 mmol/L; requested SI units
be included in measure specifications. Raised operational concern about timing origin — draw
time vs. result time — noting that point-of-care lab transport and processing can add 10–15
minutes before a result appears in Epic, yet the timestamp reflects draw time. Concerned that a
15-minute window may generate unavoidable flags.
• Dr. Nirav Shah (MPOG) via chat: Confirmed that SI unit equivalents will be incorporated into
measure specifications.
• Blake Wilson (MyMichigan) via chat: Echoed the draw-time vs. result-time concern. Supported
60 minutes as a reasonable threshold if result time is used as the start point.
• Dr. Patrick Henson (Vanderbilt): Endorsed tracking BG < 54 mg/dL incidence data separately,
even if the treatment window debate continues. Supported separating treatment from recheck
for targeted QI.
• Kate Buehler (MPOG) via chat: Clarified that MPOG typically captures collection (draw) time
from Epic’s standard extract; some variation across sites exists.
• Kim Finch (Henry Ford) via chat: Asked whether MPOG can capture oral glucose administration
(e.g., juice). Meridith Wade (MPOG) confirmed oral glucose is captured if mapped correctly in
the EHR. Kim noted that juice specifically has been difficult to map at their site.
• Xan Abess (Dartmouth) via chat: Noted that some providers may administer juice, ginger ale, or
other oral carbohydrate solutions as treatment, particularly in awake PACU patients.
• Morgan Brown (Boston Children’s) via chat: Raised scenario of a diluted blood gas sample
prompting a repeat draw. If the repeat BG is > 70 mg/dL, the case should pass. Meridith Wade
(MPOG) confirmed the current success criteria include treatment or recheck within 30 minutes,
so a second lab draw would currently result in a pass; the group noted this criterion could be
refined to require the recheck to confirm BG > 70.
Vote & Decision:
Decision:
• GLU-12: Modify — Add a 15-minute treatment requirement for severe hypoglycemia (BG < 54
mg/dL / 3.0 mmol/L) while retaining the 30-minute criterion for BG < 70 mg/dL.
• GLU-13: Modify — Add a 15-minute treatment requirement for severe hypoglycemia (BG < 54
mg/dL / 3.0 mmol/L) while retaining the 30-minute criterion for BG < 70 mg/dL.
• Coordinating Center will evaluate draw-time vs. result-time implications and assess metric
fidelity before finalizing specifications.
Measure Review #2 — GLU-09, Reviewer: Patrick Henson, MD (Vanderbilt University)
Review Document: GLU-09 – Patrick Henson Review 05.18.26
Measure Review #3 — GLU-10, Reviewer: Patrick Henson, MD (Vanderbilt University)
Review Document: GLU-10 – Patrick Henson Review 05.18.26
Measure Review #4 — GLU-11, Reviewer: Patrick Henson, MD (Vanderbilt University)
Review Document: GLU-11 – Patrick Henson Review 05.18.26
Measure Summaries:
• GLU-09 — Hyperglycemia Management, Intraop: Timing: anesthesia start to end. Inclusion: BG
> 180 mg/dL. Success: treatment or recheck within 90 minutes. Includes patients ≥ 12 years.
• GLU-10 — Hyperglycemia Management, Periop: Timing: preop start to PACU end. Inclusion: BG
> 180 mg/dL. Success: treatment or recheck within 90 minutes. Includes patients ≥ 12 years.
• GLU-11 — Hyperglycemia Treatment, Periop: Timing: preop start to PACU end. Inclusion: BG >
180 mg/dL. Success: treatment within 90 minutes (recheck not required). Includes patients ≥ 18
years.
• All measures exclude: ASA Class 5 and 6, cases < 30 minutes, outpatient cases < 4 hours, non-
labor/cesarean OB procedures.
Background & Rationale:
• Acute hyperglycemia in perioperative patients is associated with impaired wound healing,
surgical site infection, AKI, prolonged length of stay, and mortality. Glycemic control across the
preoperative, intraoperative, and postoperative periods all influences outcomes.
• NICE-SUGAR (2009) established BG < 180 mg/dL as a safer target than tight control (81–108
mg/dL) due to excess hypoglycemia and worse outcomes. Subsequent literature supports BG
100–180 mg/dL as the perioperative target.
• Recent evidence includes: (1) smaller RCTs reinforcing BG < 180 mg/dL to reduce complications;
(2) a large Vanderbilt pragmatic trial finding that an insulin-dosing reminder alone did not
improve postoperative hyperglycemia — supporting retention of management measures (GLU-
09/10) alongside treatment-only measures (GLU-11); (3) evidence that non-diabetic patients
with stress-induced hyperglycemia (~30% of surgical patients) have worse outcomes, supporting
inclusion of all patients regardless of diabetes diagnosis.
Reviewer Recommendations:
• 90-minute treatment window is somewhat relaxed but remains appropriate given practical
clinical constraints.
• Inclusion of all patients with BG > 180 mg/dL is appropriate; measures should not be restricted
to diabetics only.
• 180 mg/dL (10 mmol/L) is an appropriate threshold consistent with national society guidelines
(ADA, Endocrine Society, SCCM).
• Excluding obstetric deliveries is appropriate given unclear data on optimal treatment thresholds
for hyperglycemic parturients and risk to the fetus from rapid glucose lowering.
• Open questions for group discussion: (1) How are early preop BG values (> 4 hours before case
start) handled? (2) Nearly half of cases at some institutions are excluded due to the outpatient <
4 hours criterion — ambulatory measures are in development. (3) Age conflict: GLU-09/10
include ages ≥ 12, GLU-11 ages ≥ 18 — is the adolescent inclusion in GLU-09/10 intentional? (4)
ASA 5 exclusion — a small but potentially high-benefit cohort for glucose monitoring.
Discussion Highlights:
• Dr. Tony Edelman (MPOG): Clarified the adolescent age split: the Pediatric Subcommittee
reviewed glucose measures ~18 months ago and recommended retaining ages ≥ 12 in GLU-
09/10. In the pediatric population, providers may recheck rather than treat; GLU-11 (treatment
only) therefore applies to adults ≥ 18. The pre-op window starts at documented pre-op present;
the 4-hour fallback applies only when that document is absent.
• Dr. Nirav Shah (MPOG): Noted that MPOG has BG data from up to 4 hours before pre-op, but it
is not currently used in measure calculations. This represents a potential system-level
opportunity for hyperglycemia management improvement. The Coordinating Center needs to
determine how to present disaggregated sub-component data back to sites so quality
champions can distinguish between cases flagged for failure to treat versus failure to recheck.
• Dr. Kunal Karamchandani (UT Southwestern): Raised cost-effectiveness concern about
universal BG checking in non-diabetics; asked for data on the proportion of non-diabetic surgical
patients with BG > 180 mg/dL. Noted institutional resistance to blanket glucose checks.
• Dr. Patrick Henson (Vanderbilt): Clarified that the measures do not prescribe checking — they
specify what to do once a BG value is documented. Non-diabetic inclusion is appropriate given
evidence of stress hyperglycemia. Vanderbilt has incorporated glucose checks into enhanced
recovery protocols as a structured approach rather than universal testing.
• Kim Finch (Henry Ford) via chat: Shared that Henry Ford transitioned from testing only diabetics
to testing all patients and found 13–15% of patients presenting as non-diabetic had elevated BG
values and had never been diagnosed.
• Dr. Josh Goldblatt (Henry Ford – Allegiance): Supported reducing the 90-minute treatment
window. Henry Ford policy requires insulin administration within 30 minutes of a known high
glucose value. Acknowledged that logistical factors (lab result delay, point-of-care vs. sendout)
make strict 30-minute compliance at a measure level challenging.
• Blake Wilson (MyMichigan) via chat: Supported 60 minutes if result time (rather than draw
time) is used as the measurement start point.
• Dr. Patrick Henson (Vanderbilt): Agreed that 60–120 minutes is a reasonable range. At 89
minutes post-result, a provider should at minimum recheck to confirm the glucose trend.
Supported shortening the window for GLU-11.
• Jonathan Paul (Columbia) via chat: Agreed with Dr. Goldblatt’s position on the treatment
window.
• Dr. Tariq Esmail (Toronto) via chat: BG 180 mg/dL = 10 mmol/L; requested SI equivalents in
measure specifications.
Vote & Decision:
Decision:
• GLU-09 & GLU-10: Keep as is. Age language will be reviewed and clarified (adolescent inclusion
in GLU-09/10 is intentional per Pediatric Subcommittee recommendation). ASA 5 exclusion
retained pending broader collaborative discussion.
• GLU-11: Modify — reduce treatment window from 90 to 60 minutes.
• Action Item: Per Dr. Henson’s recommendation (confirmed by Dr. Shah): the Coordinating
Center should evaluate whether treatment window timing should be harmonized across GLU-
09, GLU-10, and GLU-11, with explicit differentiation between “treatment or recheck” and
“treatment alone” success criteria. Consider how changes to GLU-11 should be aligned with the
corresponding treatment windows in other glycemic management measures.
Meeting Adjourned: 11:05 am
Next Meeting: Monday, July 27, 2026
Appendix A — Full Transcript
ASPIRE Quality Committee — May 18, 2026
(Original wording preserved; sentence structure, punctuation, and flow corrected for clarity.)
Opening Remarks & Announcements
Anthony L Edelman [MPOG]: It's 10:01 and we've got a packed agenda with over 50 participants, so let's get
started. Thanks everyone for joining on this lovely Monday morning. I don't know where you are, but here in
southeastern Michigan it is sunny and a beautiful spring morning — hopefully similar where you're at.
Today's agenda: just a couple of announcements, and then high-quality in-depth reviews — first, Glucose 12 and
13 by Dr. Duggan out of UAB, and then Glucose 9, 10, and 11 by Dr. Henson out of Vanderbilt. If we have time,
we'll proceed with the Best Practice Exchange with Dr. Chopra out of Henry Ford in Detroit, and if time also
permits, a couple of updates from Dr. Shah. As usual, roll call will be via Zoom. If you're on a cell phone or your
name isn't listed, please reach out and let us know.
Upcoming events: Friday, July 17th is our ASPIRE Collaborative meeting in Ann Arbor, held at Weber's Hotel.
Friday, October 16th is the MPOG Retreat, associated with the ASA meeting in San Diego. Please come join us —
hopefully we'll see lots of friendly faces in person.
Our featured member for May and June is Dr. Avidan out of Washington University in St. Louis — a longtime
member and one of the originators. Nirav, if you'd like to add any color commentary, feel free.
Nirav J Shah [MPOG]: Dr. Avidan has just been a long-time supporter of MPOG and also helped launch THRIVE, the
large PCORI trial looking at general versus inhalational anesthetics. A big shout out to Dr. Avidan.
Anthony L Edelman [MPOG]: What's old is new again — we have the University of Vermont back in the fold. The
Chair is Dr. Aslakson, our IT Champion is Dr. Waldschmidt, and the Quality Champion is Dr. Georgiadis. I apologize
for my pronunciation of those names.
Nirav J Shah [MPOG]: For those who may not know, the University of Vermont was one of our original MPOG sites
— about 18 years ago — and one of the early adopters of Centricity Anesthesia, which was the original cohort of
MPOG sites. They were not active for a couple of years and have now rejoined. Welcome back, folks.
Anthony L Edelman [MPOG]: We'll jump right in. This is Dr. Duggan from the University of Alabama at Birmingham,
with the measure review for Glucose 12 and 13. Dr. Duggan, please take the floor.
Measure Review — GLU-12 & GLU-13
Beth Duggan, MD, MA [University of Alabama at Birmingham (UAB)]: This is the first time I've done this for MPOG,
so I'll try to keep it moving, but feel free to just interrupt — I'm not very good at watching for hands while talking
and reading at the same time.
I was asked to review the metrics for hypoglycemia. This is very specific to hospital hypoglycemia, which is slightly
different from outpatient management. Most of you know this, but the target range is generally 100 to 180.
Hypoglycemia is classified in the hospital as a blood glucose less than 70; severe hypoglycemia as less than 40. I
differentiate those because it will come up later.
It's a multifactorial issue — a lot to think about in the operating room. We know that both in cardiac and non-
cardiac surgery it leads to poor outcomes, and particularly when left untreated, patients can have severe,
debilitating consequences. Our role as anesthesiologists is to continue to test and treat for both hyperglycemia and
hypoglycemia. The incidence of hypoglycemia is relatively low, but it is really catastrophic if it goes unrecognized.
Our goal here is to uphold best practices. I highlighted some past literature — the NICE-SUGAR study, which you
know: if you are targeting tight control at 80 to 110, you will cause hypoglycemia and likely have worse outcomes.
The Lazar study on aggressive blood glucose management during CABG found limited outcome benefit and
increased hypoglycemia risk. These studies and the subsequent retrospective reviews are well established.
Importantly, the MPOG registry trial looked at the incidence of intraoperative hypoglycemia in adult patients using
a threshold of less than 60 mg/dL. The key finding is that intraoperative insulin was not independently associated
with a low blood sugar. A lot of fear exists around using insulin perioperatively because of concern for
hypoglycemia, and it has been well-documented that withholding insulin when needed not only exacerbates
hyperglycemia but very infrequently actually causes hypoglycemia. That said, there are retrospective reviews
showing that if you give insulin and do not repeat the blood sugar, you will have a problem. So: treating is okay,
but it must come with retesting.
Reviewing definitions: the ADA Standards of Care 2026 classifies hypoglycemia as Level 1 — blood glucose less
than 70 mg/dL; Level 2 — blood glucose less than 54 mg/dL (3.0 mmol/L), which we will gravitate toward in this set
of metrics; and Level 3 — symptomatic severe hypoglycemia, which does not apply to the anesthetized patient. So
for our purposes, we focus on numerical Levels 1 and 2.
The current metric is: provide a dextrose-containing solution, or recheck glucose, within 30 minutes of a blood
glucose less than 70. A proposed new metric would require IV dextrose treatment within 30 minutes for blood
glucose less than 70 — similar to current, but removing the recheck-only option. Additionally, new to this proposal:
for severe hypoglycemia (blood glucose at or below 54 mg/dL, which is the current ADA gold standard threshold),
IV dextrose-containing solution must be provided within 15 minutes. So we now have two levels of hypoglycemia
instead of one: less than 70 managed within 30 minutes, and less than or equal to 54 managed within 15 minutes.
The negative outcomes associated with those very low blood sugar levels justify the faster response window.
The key is rechecking, so that you do not have persistent hypoglycemia go unrecognized — that's food for thought
as we continue rolling these out, and something I mentioned in the review. Generally, we provide IV dextrose, but
in the PACU, for a wide-awake patient, both the ADA and CDC support an oral carbohydrate solution as an
appropriate treatment option.
For inclusions: any patient with an intraoperative blood glucose measured between anesthesia start and end time,
and any postoperative blood glucose in the PACU. Exclusions: ASA Class 6, cases without a recorded blood glucose
value, labor epidurals and non-operative obstetric patients. Both operative and non-operative obstetric patients
will not be in this metric — a separate metric is needed for that population.
On time to treatment: the ADA recommends immediate treatment, and I think 15 minutes is reasonable as an
operationalization of 'immediate' in the OR. For context, there are some papers stating treatment needs to happen
within 5 minutes, which is part of why we debated and justified a time frame shorter than 30 minutes. We also
want to ensure that insulin solutions or AID devices are paused when treating hypoglycemia — that is coming in
future metrics. Nirav, Tony — should we open for questions, or shall I continue through the ranges?
Nirav J Shah [MPOG]: I think it's worth pausing for questions. I really like the differential between less than 70 and
less than or equal to 54, where we should move much faster. One thing we had discussed is whether MPOG data
resolution can accurately capture a 15-minute time period — I believe it can, whereas 5 minutes would be more
difficult. So I'd love feedback from the group on whether this tiered urgency reflects how they are already
managing hypoglycemia in practice.
The second point Beth raised is this notion of separating treatment from rechecking. Traditionally, we have
bundled those together in a hypoglycemia management measure. With Glucose 11, and now with the upcoming
ambulatory measures, we have started separating treatment from recheck. I'm curious whether the group would
support doing the same for hypoglycemia management. The benefit is that different interventions are needed for
cases flagged for failure to treat versus failure to recheck, and separating them could help sites design more
targeted interventions. I see a few things in the chat.
[Tariq Esmail [University of Toronto] via chat]: 54 mg/dL is 3.0 mmol/L — please include if possible.
[Nirav J Shah [MPOG] via chat]: Replying to Tariq Esmail: Yes — definitely will include SI units.
[Kim Finch (KFINCH1) [Henry Ford Health System – Detroit] via chat]: Is MPOG able to extract oral glucose
in the perioperative period?
[Meridith Wade [MPOG] via chat]: Replying to Kim Finch: Yes — if mapped correctly, we consider
oral glucose solutions as treatment. [Screenshot of treatment list shared]
[Kim Finch (KFINCH1) [Henry Ford Health System – Detroit] via chat]: We have trouble with juice
specifically.
[Xan Abess [Dartmouth] via chat]: Some providers may just use juice, ginger ale, etc.
Nirav J Shah [MPOG]: Is MPOG able to extract oral glucose in the perioperative time period? Meridith, you
responded to that — if the oral glucose is mapped correctly, we should be able to capture it. Dr. Brown has a
comment: if a blood gas is sent and was clearly diluted, a second one may be sent, and if it was greater than 70,
the case should pass. Some consideration for this scenario should be in the metric.
[Morgan Brown [Boston Children’s Hospital] via chat]: The one circumstance that comes up is when we
send a blood gas and it was clearly diluted. We may send a second one, and if it was > 70, you should
pass, I would think. Some consideration for this scenario should be in the metric.
[Blake Wilson [MyMichigan] via chat]: Is the timeframe measured from draw time or result time? A BMP
can take significant time to result. Also interested in PO extraction.
Xan Abess [Dartmouth]: Thanks, Beth — really nice review. Maybe a slight counter-argument to what you're going
after, Nirav: I get a little concerned with all this parsing of the measures, and I wonder about fidelity. The question
is: do we think we're really going to modify behavior with these measures? Especially if you have a 15-minute time
deadline, I can see providers being busy and documenting that they gave the dextrose a little beyond the 15
minutes. These things all seem clinically sound, but if you get too detailed with the measures, you start to risk
accuracy errors and lose buy-in from your team in the department.
Anthony L Edelman [MPOG]: We actually had some of that discussion internally. There was debate about whether
the right time was 5, 15, or 30 minutes. Treating a blood glucose at or below 54 as a critical value — everyone felt
it should be called out separately — but the timing and the fidelity of the data, and the reasonability of getting the
medication to the patient in an appropriate timeframe, certainly came up.
Beth Duggan, MD, MA [University of Alabama at Birmingham (UAB)]: I agree with all of the above. Whether it's 20
minutes or 15 minutes, something less than 30, because the goal is to say: this is a critical value, this is different
from 70. I think there are a lot of incidents where a blood sugar of 40 gets treated like a blood sugar of 69, when
we know the outcomes of those two numbers can be very different. How you set that metric — 20 minutes or 15
minutes — the key message is: this is something more severe and more dangerous than a blood sugar of 70. The
literature is pretty robust on that. I did not include the 3 mmol/L SI unit conversion in my slides — I should have,
and I'll do better at that.
Josh Goldblatt [Henry Ford Health System – Allegiance]: I'm very much in favor of splitting the recheck measure
from the treatment measure. That's how we operationalize the work: we have testing cadence and testing
expectations, and then what we're actually doing about it is a separate lane. At Henry Ford, we have a tiered
treatment quantity — half an amp of dextrose for Level 1 hypoglycemia, and a full amp for Level 2. We used 40
mg/dL as our Level 2 threshold and are changing it to 50, which is what the lab uses. We're still not at 54 per the
guidelines, but we're much closer. The separate recheck metric makes a lot of sense operationally, because we
don't know a patient's glucose reserve. We might treat with half an amp and think we're fine, but without knowing
they've exhausted their reserve and are on a nosedive, we could miss something. A separate monitoring metric is
really valuable.
[Josh Goldblatt [Henry Ford Health System – Allegiance] via chat]: I think he is highlighting the difference
between draw time and result time.
Tariq Esmail [University of Toronto]: I agree with the principle of what Beth was explaining, but I'm concerned
about the operational effect on behavior. Specifically, what timeline will be flagged as the start point? At our site,
we draw blood at a specific time — that label time is what shows up in Epic — but it may take 5 to 15 minutes to
get from the ROR through the RT lab, run the sample, and have the result available. The computer still shows the
draw time. So a provider may first see the result when it is already 15 minutes old, through no fault of their own.
That would generate flags perceived as unjust, and I want to make sure that 15-minute window accounts for when
the timeframe is actually measured from.
Anthony L Edelman [MPOG]: That's a valid concern — the difference between draw time and result time is
significant, especially when the sample is going to a primary lab rather than a bedside ABG machine. Patrick?
Patrick Henson [Vanderbilt University Medical Center]: Everybody has echoed what I'd say. I'm always in favor of
more measures and greater detail, and I would like to see the incidence of glucose values less than 54 across the
collaborative. We're limited by the values we check, but these are clearly meaningful data points we should want
to know. I endorse these measures, and I'd also like to see incidence data surfaced even if the treatment window
debate continues. I'd also frame it similarly to the blood pressure measures — we've separated out MAPs of 65
and 55 because drilling down on the actual serious cases has value. I would want to address individual cases where
patients had glucose below 54 as specific quality improvement opportunities.
Nirav J Shah [MPOG]: At the Coordinating Center, we need to figure out how to surface all these sub-components.
The measure is our primary unit for sharing data back to sites, but quality champions designing interventions need
to see data disaggregated — for example, are cases being flagged for failure to treat, or failure to recheck? We
need to find a way to share those elements with sites so they can design the right intervention.
Josh Goldblatt [Henry Ford Health System – Allegiance]: One additional thought on the different threshold and
treatment timelines: the concern is operational — how would we write protocols to capture that nuance? Adding
that level of detail makes diabetes and glucose management even more complex, and it's already very complex. At
Henry Ford, we've taken the approach of: if you're under 70, treat it like it's serious every time. We're still working
on that. Our own analysis shows that most of our misses are in pre-op, with non-diabetic patients in the 60s who
are probably fine — but not all of them. We've created infographics and focused on the message that everybody in
this category needs to be taken seriously, and that giving someone in the 60s a half-amp of dextrose will not do
harm. We've simplified in order to improve reliability, so that's another consideration.
Anthony L Edelman [MPOG]: There is certainly a lot of nuance. My feeling on this particular measurement is that
the nuance is deserved, because the values really are driving providers to act in two different ways. A critical value
of 50 or 44 should be treated as a critical value — something genuinely different from a 69 in a marathon runner
who just runs on the lower side of things. While more nuance creates more complexity, I do think sometimes the
complexity is deserved.
Beth Duggan, MD, MA [University of Alabama at Birmingham (UAB)]: Josh, I totally appreciate what you're saying.
In an ideal world, everybody would act on a blood sugar less than 70 rapidly and we wouldn't need this nuance —
but people don't. The goal is to say: what is reasonable for most, and at what threshold do we get into real
danger? A blood sugar of 60 in a fasting marathoner is probably okay, but adding that nuance makes things
complicated. The goal is behavior change. I commend this group for trying to make a complicated topic usable in a
simple way. The literature on the difference in outcomes between 40 and 69 is robust, and I don't think you're
wrong to try.
Patrick Henson [Vanderbilt University Medical Center]: I equate this a little to the blood pressure measures — none
of us think a MAP of 56 is appropriate, and yet we've separated out thresholds of 65 and 55 because drilling down
on the actual serious cases has value. From a quality and practice improvement standpoint, individual cases where
patients had glucose below 54 represent meaningful review opportunities. I do want to add: these measures apply
across all patients, not just diabetics, and I think there is opportunity to parse this out further without necessarily
adding too many measures.
Anthony L Edelman [MPOG]: Thanks, Patrick. In the interest of time, we'll go to the recommendations and then a
vote. Beth, do you want to briefly walk through the future considerations?
Beth Duggan, MD, MA [University of Alabama at Birmingham (UAB)]: To add a few more things to consider:
obstetrics — we do need to come back to that. If a pregnant mother is hypoglycemic, there is clearly risk to the
fetus, so that requires a separate metric framework. The glucose reserve concept is exactly why a separate recheck
metric matters — if we are not retesting, we won't know if a patient is on a nosedive. CGMs and Automated Insulin
Delivery devices are coming — over 50% of Type 1 diabetics now use AID. CGM accuracy in the OR is something I'm
currently evaluating, but that will be a future metric. And the Type 1 diabetic is a very different patient from the
Type 2 diabetic, which may warrant separate metric pathways. All of these things are only getting more nuanced
as technology and understanding of these diseases progresses.
Nirav J Shah [MPOG]: Thank you. I'm going to launch the polls now. The first is simply: do we want to keep the
measures as is, or modify them based on our discussion? One vote per site, please.
[Meridith Wade [MPOG] via chat]: Replying to Morgan Brown: The success criteria for this measure
currently include both treatment and recheck of glucose within 30 minutes. If a second lab is sent, the
case would currently pass. We could consider only passing if that second lab value is > 70 mg/dL.
[Kate Buehler [MPOG] via chat]: Replying to Blake Wilson: For the standard MPOG extract out of Epic, the
observed time correlates to draw time. There may be some variation across sites, but MPOG typically
captures collection time for labs.
Nirav J Shah [MPOG]: Results: 73% and 60% voted to modify. Moving to the modification polls. The modification
Dr. Duggan suggested is limiting the response to 15 minutes for cases of severe hypoglycemia. One more question
after this: whether folks are interested in starting to separate the treatment and the rechecking.
Patrick Henson [Vanderbilt University Medical Center]: Is the thought that this would be built into current
guidance, such that we have different treatment criteria for glucose less than 70 and glucose less than 54 for
success?
Nirav J Shah [MPOG]: Yes — we would need to work through that at the Coordinating Center to determine the
computational advantages and disadvantages for each approach, and then share that with the group.
Patrick Henson [Vanderbilt University Medical Center]: The current state gives us less than 70, right? We would
lose some of that data if we went to less than 54 only. I'm just curious whether we'd keep both.
Nirav J Shah [MPOG]: We would try to do this in a way that you wouldn't lose information. The poll results support
providing data for less than 54 within 15 minutes, and we'll build toward that. The final poll: are folks interested in
separating the recheck from the treatment as a mechanism for targeting interventions more effectively, or keeping
them together?
The poll results show support for separating them as well. We will start working on that. Dr. Duggan, thank you so
much — that was wonderful.
Measure Review — GLU-09, GLU-10, GLU-11
Anthony L Edelman [MPOG]: Dr. Duggan, thank you — great discussion and great review. We now move to Glucose
9, 10, and 11 with Dr. Henson from Vanderbilt. I have your slides here if you'd like me to share them.
Patrick Henson [Vanderbilt University Medical Center]: I did condense the MPOG review a bit, but happy to run
through these quickly. These measures were built three years ago, almost to the day, after reviewing the current
glucose measures, which used a blood glucose threshold of over 200 — a bit more relaxed than we would have
liked. We created GLU-09, 10, and 11, which cap the upper treatment threshold at 180, and include measures for
both intraoperative hyperglycemia and perioperative management. There is some intentional overlap between the
measures, but for centers who want to focus on one area or drill down, they are helpful. You'll notice there is no
measure for recheck alone — it's either treatment or recheck, or treatment only.
Exclusions, as Dr. Duggan already referenced, are fairly consistent: patients under 12 or under 18 depending on the
measure, high ASA class, obstetric and non-operative obstetric populations, short cases, and ambulatory cases less
than 4 hours — which is worth discussing.
Acute hyperglycemia is associated with a host of complications in surgical and perioperative patients, including
length of stay, infection, and mortality. Glycemic control across the preoperative, intraoperative, and
postoperative periods all influences outcomes, as we know from critical care and surgical inpatient literature.
Treatment with insulin to maintain blood glucose in the target range is associated with reduced risk of
complications — so not just accidental normoglycemia, but actively placing patients in the right range is associated
with improved outcomes.
We borrowed our treatment paradigms largely from the critical care literature — most notably, NICE-SUGAR in
2009, which established that we should not tightly control patients given the risk of hypoglycemia and worse
outcomes. That was a mixed medical and surgical population, and I think it accounts for much of what we see.
Capping the upper limit at 180 is consistent across the critical care and operative literature. Some groups may
benefit from a lower optimal target — cardiac and hepatobiliary patients come to mind — but from the broadest
consensus, 180 is appropriate.
There is also a large Vanderbilt pragmatic trial of an insulin dosing reminder versus a glucose check reminder that
did not result in improvement in postoperative hyperglycemia. Paradoxically, that actually may support the MPOG
approach of maintaining a management measure that includes a recheck as part of treatment, rather than
reminders alone. Additionally, in patients without diabetes, we estimate upward of 30% may have prediabetes or
stress-induced hyperglycemia, and their outcomes are disproportionately worse than normoglycemic non-diabetic
patients — so we should not exclude them from these measures.
One key point: you cannot enter these measures without a documented glucose check, so if you never check blood
glucose, you will ostensibly perform at 100%. Entering the pathway depends on your site's protocol for who gets
checked. At Vanderbilt, we check based on enhanced recovery pathways rather than universal testing, which is a
more targeted approach.
For the measures being reviewed today: all three include patients with a documented blood glucose over 180
mg/dL (10 mmol/L), and all exclude high ASA class, short-duration cases, short outpatient cases, and non-labor
obstetric procedures. One variable between the measures is the age range — GLU-09 and GLU-10 include an age
split I think we could clarify, while GLU-11 includes only patients 18 or older.
[Tariq Esmail [University of Toronto] via chat]: 180 mg/dL is 10 mmol/L.
Patrick Henson [Vanderbilt University Medical Center]: For my review: I think the 90-minute time frame is a bit
relaxed, but likely appropriate given current practice. I think 60 to 120 minutes is a reasonable range — I would
prefer shorter, but I understand we may need some flexibility. Inclusion of all patients with hyperglycemia is
appropriate; we should not limit to diabetic patients. 180 is an appropriate threshold. Excluding obstetric deliveries
is appropriate — treating hyperglycemic parturients is not necessarily recommended, the treatment threshold is
unclear, and depending on timing, we run the risk of impacting the fetus.
Some open questions for the group: first, the preoperative phase — MPOG's algorithm declares pre-op as within 4
hours of case start. Inpatients who have blood glucose checked on the floor may fall outside this window and, to
my understanding, cannot succeed on this measure. Those patients might actually benefit significantly from
appropriate treatment and management. How do we account for them? Second, almost half of our institution's
cases are excluded due to the outpatient less-than-4-hours criterion — ambulatory measures are in development,
but are we comfortable maintaining this while we work on that? Third, the age range is conflicting: the measures
are described as adult measures, but GLU-09 and GLU-10 include adolescents, which I think should either be
clarified or the measures standardized to 18 and older. Fourth, ASA Class 5 is excluded — a small cohort, but
potentially one that benefits from monitoring. I'd like to put that out to the group for discussion.
Anthony L Edelman [MPOG]: Great, thank you — fantastic, in-depth review. A couple of comments before opening
to the floor. On the pediatric question for GLU-09 and GLU-10: the Pediatric Committee reviewed the glucose
measures about a year to a year-and-a-half ago and recommended keeping patients 12 and older in that group.
The reason for the split between GLU-09/10 and GLU-11 is that GLU-11 is the treatment-alone measure. In the
pediatric world, providers don't always necessarily treat — they may recheck. So leaving the adolescent age group
in the recheck category is still appropriate, and excluding them from the treatment-specific measure is also
appropriate. That's how we ended up with the age splits.
On the pre-op present 4-hour window: the measure generally starts when pre-op present is documented in the
chart. The 4-hour window only applies when there is no discrete pre-op present document for that patient. So it
doesn't affect every patient — just those without a pre-op present documented.
Nirav J Shah [MPOG]: To go back to the pre-op time period: MPOG does have blood glucose data from up to 4
hours before pre-op, but because the measure starts when the patient rolls into pre-op holding, that earlier
information — for inpatients who had glucose checked on the floor or ICU — is not being used in the measure. It's
an opportunity for systemic improvement in hyperglycemia management across care teams that we're not
currently capturing.
Kunal Karamchandani [UT Southwestern]: Great presentation. I have a concern about applying this metric to non-
diabetics. At our institution, it's going to be a hard sell to ask everyone coming in for a procedure to have their
blood sugar checked. Is the data strong enough to support expanding this to all-comers? What percentage of non-
diabetics are actually above 180? Is it cost-effective?
Patrick Henson [Vanderbilt University Medical Center]: These measures don't prescribe checking — what they do
not do is exclude non-diabetics, which I think is appropriate. A significant number of non-diabetic patients are
hyperglycemic perioperatively, and their outcomes are exponentially worse compared to normoglycemic non-
diabetic patients. At Vanderbilt, we parse out our enhanced recovery pathways to consolidate glucose checking in
a targeted way rather than testing everyone.
[Kim Finch (KFINCH1) [Henry Ford Health System – Detroit] via chat]: At Henry Ford, we transitioned from
testing only diabetics to testing all patients. We found that about 13–15% of our patients presenting as
non-diabetic had elevated blood sugars and had never been diagnosed with diabetes.
[Jonathan Paul [Columbia University] via chat]: Agree with Josh.
[Blake Wilson [MyMichigan] via chat]: 60 minutes seems reasonable if we are speaking about result time.
Same argument applies as for hypoglycemia.
[Graciela Beatriz Mentz [MPOG] via chat]: Very interesting discussion. Need to move to my next meeting.
Nirav J Shah [MPOG]: These measures essentially assume that providers have made the appropriate decision about
whether to check. The question of which population to actually check — maybe diabetic patients plus high-risk
cases — is something we've discussed. We've thrown around the idea of a measure targeting that specific
population, but it's still in the planning phases. I did want to get input on the timing component. Right now the
window is 90 minutes. Should it be something less — 60 minutes? What is the appropriate time frame?
Josh Goldblatt [Henry Ford Health System – Allegiance]: At Henry Ford, across the whole hospital, insulin is due
within 30 minutes of when you know it's needed — whether PRN or because a high glucose came back. Part of the
rationale is that glucose is so rapidly variable that if you wait longer, you're not sure what you're actually treating.
That said, when a chem panel comes back in 45 minutes, the right approach may be to recheck with a finger stick
to confirm it's still current. So 30 minutes for the metric may not always be operationally achievable, but I would
support something shorter than 90 minutes for treatment alone.
Patrick Henson [Vanderbilt University Medical Center]: I think it's hard to align exactly on 30, 60, or 90 minutes
given the overlap between measures, but I agree. If someone has a blood glucose result and waits 89 minutes, that
could be two hours from the actual high value. At the very least, by that point they should probably recheck to
confirm the value hasn't risen or dropped. I think less than 90 minutes is appropriate. Whether that's an
institutional recommendation or an MPOG measure change is a separate question.
Anthony L Edelman [MPOG]: Any other comments? We're at time. Nirav, would you like to move to a vote?
Nirav J Shah [MPOG]: I'm going to launch the first poll. The modifications would be: should we add ASA Class 5
cases? My thought is that if we add ASA Class 5, we should have a broader discussion about adding it across all
measures where appropriate — not just this one.
Patrick Henson [Vanderbilt University Medical Center]: I agree — just wanted to raise it for discussion. I appreciate
it.
Nirav J Shah [MPOG]: Poll results: keep GLU-09 and GLU-10. So we'll update the age language to accurately reflect
adolescent inclusion per the Pediatric Committee recommendation, and we'll continue to exclude ASA Class 5 and
6. There was a segment that wanted to modify GLU-11, so I'm launching one more poll. I'm curious what folks
think about the timetable. Then on the MPOG side, we'll make sure any change doesn't conflict with or be
inconsistent with the other glycemic management measures.
Patrick Henson [Vanderbilt University Medical Center]: I'd ask a follow-up: would it be possible to add, in the
measures that include both treatment and recheck, a differentiation between the treatment time and the recheck
time, so those times could be different? If you made a change to GLU-11, the treatment window there should be
aligned with treatment windows in GLU-09 and GLU-10 — but that's a bigger question.
Nirav J Shah [MPOG]: Totally agree. And it goes back to what we were doing with the ambulatory measures —
separating them out. Poll results: yes, there is support for reducing that time period, which makes a lot of sense.
Tony, do you want to bring us home?
Closing Remarks
Anthony L Edelman [MPOG]: We're over time, so I'll be quick. I want to thank Dr. Henson and Dr. Duggan for really
great, in-depth reviews, and thank everyone for the great discussion. Unfortunately, we don't have time for Dr.
Chopra or for the updates today, but I think we can get the updates in at least at the next meeting. Thanks,
everyone, for your time — sorry we went over.
[Beth Duggan, MD, MA [University of Alabama at Birmingham (UAB)] via chat]: Such an honor to be asked
to participate. Thank you!
Josh Goldblatt [Henry Ford Health System – Allegiance]: Thanks.
Kunal Karamchandani [UT Southwestern]: Nice job.
