Oropharynx: T3 or T4?
A man in his fifties has a p16-positive squamous cell carcinoma of the base of tongue. He needed a tracheostomy before any treatment, and there are nodes on both sides of the neck. The outside note says T3N2. Is it T3 or T4, and why does it matter?
Concepts first
HPV: human papillomavirus. p16: a protein the virus switches on, read on the biopsy by immunohistochemistry. ENE: extranodal extension, cancer breaking out through the lymph node capsule. AJCC: the American staging rules. Since Jan 1, 2026, p16-positive oropharynx uses AJCC Version 9; p16-negative still uses AJCC 8; trials such as KEYNOTE-689 enrolled by AJCC 8.
p16 stands in for HPV; it is not HPV.p16-positive means nuclear and cytoplasmic staining, moderate to strong, in at least 70% of tumor cells (College of American Pathologists, 2025). About 1 in 10 p16-positive tumors is HPV-negative (10.9% overall; about 3 in 10 at subsites other than the tonsil and tongue base). Version 9 calls p16-positive tumors HPV-independent, and stages them as p16-negative, when an HPV-specific test disagrees, or when p16 comes with a caveat: a subsite outside the tonsil and tongue base, a neck node with unknown primary, unusual morphology, or a population where HPV causes few of these cancers.
The numbers behind it
T: size first, then neighbors
Lingual epiglottis is T3
Mucosal spread from the base of tongue or vallecula onto the lingual surface of the epiglottis is not invasion of the larynx. Version 9 also moved the lingual surface of the epiglottis into the oropharynx chapter (still T3 when the tumor spreads onto it) and the adenoid and tubal tonsils into nasopharynx.
A tracheostomy is not a T criterion
Read the operative note: what blocked the airway?
p16-positive has one T4
No T4a or T4b: everything p16-negative calls T4b is plain T4 here. Version 9 calls T4 on extrinsic tongue muscle alone only with unequivocal imaging backed by clinical signs such as tongue tethering, fixation or deviation (Rai 2026). p16-negative keeps T4b: Lazy People Never Skip Coffee (lateral pterygoid, pterygoid plates, lateral nasopharynx, skull base, carotid encased).
T0 in the oropharynx
Only the HPV-associated chapter uses it: a neck node confirmed HPV-associated (p16 plus an HPV-specific test), with no primary found. A node not confirmed as HPV-associated and EBV-negative (EBV: Epstein-Barr virus) goes to the unknown-primary chapter; an EBV-positive node is staged as nasopharynx.
Key colors: orange, a trap; blue, a rule; grey, background.
N: three different rulers
pN0 needs at least 6 nodes examined; with fewer, use clinical N0.
Version 9 adds ENE to p16-positive N for the first time.AJCC 8 left it out: in ICON-S, the international collaboration that built that system (1907 HPV-positive patients, staged with the old 7th edition N, which had no ENE), survival was similar for the old N0, N1 to N2a and N2b, and the model grouped same-side nodes as N1, opposite or both sides (old N2c) as N2, and over 6 cm as N3 (O'Sullivan 2016). The node count after surgery came from 704 surgical patients (Haughey 2016). Version 9 now adds one N step for unequivocal ENE on exam or imaging, or proven on pathology (Rai 2026). In one Chinese single-center cohort of 191 patients treated with radiation, this moved 23% of N1 to N2 and 42% of N2 to N3 (Lv 2026).
Stage groups
For contrast, p16-negative (AJCC 8): I = T1N0 · II = T2N0 · III = T3N0, or T1 to T3 with N1 · IVA = T4a, or N2 · IVB = T4b, or N3 · IVC = M1.The same T2 tumor with two same-side nodes and no ENE is stage I if p16-positive and stage IVA if p16-negative.
Back to the base of tongue
Why was the airway secured?A bulky tongue base can obstruct at T3. It becomes T4 only when endoscopy or imaging shows a named neighbor invaded (in an airway story, usually the larynx); the tracheostomy itself is not staged.
Our patient, worked through.Say endoscopy shows a bulky tongue base filling the airway with the larynx free, and MRI shows no muscle invasion: T3. Nodes on both sides, no ENE: N2. Clinical stage II, and the outside note was right. If imaging shows unequivocal ENE, Version 9 makes it N3: stage III whatever the T. If MRI shows genioglossus invasion: T4, stage III. KEYNOTE-689 eligibility is still judged by AJCC 8 (next tab).
Try it: stage and trial check
Decide the stage in your head first, then set the boxes and check.
p16-positive with a caveat (subsite outside tonsil and tongue base, unknown primary, unusual morphology, HPV test negative) is HPV-independent: choose p16-negative.
Clinical stage, oropharynx only: p16-positive by AJCC Version 9, p16-negative by AJCC 8. The trial line uses KEYNOTE-689 entry rules, which were written in AJCC 8; performance status and the exclusions on the next tab still apply.
Memory hooks
Check yourself
Question 1p16-positive base of tongue, 3.5 cm, tracheostomy for stridor. MRI: tumor invades the genioglossus; larynx free. One 2 cm same-side node, no ENE. Stage?
T4 (extrinsic tongue muscle) N1, clinical stage III. The genioglossus made it T4, not the tracheostomy. If you said T2 by size, you forgot to check the neighbors first.
Question 2p16-positive tonsil cancer, 3.5 cm, two nodes on the same side (3 and 5 cm), no ENE on exam or imaging. Stage?
T2 N1, clinical stage I. With unequivocal ENE on imaging, Version 9 makes it N2, stage II. If p16-negative, the same tumor is T2 N2b, stage IVA.
Question 3The Question 2 patient has surgery: 6 positive nodes, no ENE on pathology. Pathological stage?
pT2 pN2, pathological stage II. After surgery, size and side stop counting: more than 4 nodes is pN2, and with ENE it would be pN3, still stage II with pT2. If you said pN1 because the nodes are small, that is the trap.
Question 4p16-positive base of tongue, 3 cm, mucosal spread onto the lingual epiglottis, one 2 cm node on the opposite side, no ENE. Stage?
T3 (lingual epiglottis) N2 (opposite side), clinical stage II. Not T4: the larynx was not invaded.
Question 5A p16-positive neck node, and no primary found after panendoscopy. How is it staged?
Version 9 treats p16 alone in an unknown primary as a caveat: confirm with an HPV-specific test. If HPV-associated, it is oropharynx T0; one same-side node of 6 cm or less without ENE is T0 N1, stage I. If not confirmed, it goes to the unknown-primary chapter.
Question 6The tumor is p16-positive but HPV in situ hybridization is negative. Does it matter?
Yes. Version 9 calls it HPV-independent and stages it with the p16-negative system. Prognosis sits in between: 55% alive at 5 years versus 81% when both are positive (Mehanna 2023). Do not use p16 alone to choose a de-escalation trial or a lighter plan.
Sources
Retrieved from PubMed and the American College of Surgeons, read on Oct 26, 2026.
- American College of Surgeons. AJCC Version 9 protocols for HPV-associated oropharynx and salivary glands, effective January 1, 2026; expert-panel summary of the N and stage group changes. facs.org
- Rai P, Holtzman AL, Bhatt AA. HPV-associated oropharyngeal squamous cell carcinoma: an imaging framework for AJCC Version 9 TNM staging. Radiol Imaging Cancer 2026. doi:10.1148/rycan.260013
- Lv W, et al. Validation of the 9th edition cTNM classification for HPV-positive oropharyngeal carcinoma. Clin Transl Radiat Oncol 2026. doi:10.1016/j.ctro.2026.101181
- Lydiatt WM, Patel SG, O'Sullivan B, et al. Head and neck cancers, major changes in the American Joint Committee on Cancer eighth edition cancer staging manual. CA Cancer J Clin 2017. doi:10.3322/caac.21389
- O'Sullivan B, Huang SH, Su J, et al. Development and validation of a staging system for HPV-related oropharyngeal cancer by the International Collaboration on Oropharyngeal cancer Network for Staging (ICON-S). Lancet Oncol 2016. doi:10.1016/S1470-2045(15)00560-4
- Haughey BH, Sinha P, Kallogjeri D, et al. Pathology-based staging for HPV-positive squamous carcinoma of the oropharynx. Oral Oncol 2016. doi:10.1016/j.oraloncology.2016.09.004
- Ang KK, Harris J, Wheeler R, et al. Human papillomavirus and survival of patients with oropharyngeal cancer. N Engl J Med 2010. doi:10.1056/NEJMoa0912217
- Mehanna H, Taberna M, von Buchwald C, et al. Prognostic implications of p16 and HPV discordance in oropharyngeal cancer (HNCIG-EPIC-OPC). Lancet Oncol 2023. doi:10.1016/S1470-2045(23)00013-X
- Lewis JS Jr, Beadle B, Bishop JA, et al. Human papillomavirus testing in head and neck carcinomas: guideline update. College of American Pathologists. Arch Pathol Lab Med 2025. doi:10.5858/arpa.2024-0388-CP
KEYNOTE-689: who got in, and what it showed
Pembrolizumab before surgery and again after it, for resectable stage III to IVA head and neck squamous cell carcinoma. FDA approved in June 2025 for tumors with PD-L1 CPS 1 or more.
Concepts first
Perioperative
Neoadjuvant is before surgery, adjuvant is after. Perioperative is both, and KEYNOTE-689 is both.
Event-free survival (EFS)
Time until an event: progression that blocks surgery, local or distant recurrence, or death from any cause.
Hazard ratio (HR)
Event rate with the drug divided by the rate without it. 0.73 means 27% fewer events at any moment; 1 means no difference.
Major pathological response
10% or less viable invasive tumor left in the resected primary and all sampled nodes. Control specimens were scored too, but without treatment before surgery none qualified.
Combined positive score (CPS)
PD-L1 (programmed death ligand 1, the partner that the PD-1 receptor binds) stained tumor cells, lymphocytes and macrophages, divided by all tumor cells, times 100 (22C3 assay).
High-risk pathology
A margin under 1 mm or extranodal extension. Both arms then received cisplatin with the radiation.
The design
In the trial, pembrolizumab 200 mg every 3 weeks (17 doses); the label also allows 400 mg every 6 weeks. Adjuvant treatment was meant to start within 6 weeks of surgery; about 4 in 10 actually did, and about 8 in 10 by 8 weeks. Cisplatin 100 mg/m² every 3 weeks for 3 doses when a margin is under 1 mm or there is extranodal extension.
Who got in
Staging in the trial was AJCC 8. Since 2026 a p16-positive patient may be restaged by Version 9 (ENE can raise N); judge eligibility by the AJCC 8 rules the trial used.
Left out
- T4b, N3, or distant metastasis
- Unresectable, or the surgeon would not operate
- Nasopharynx, sinonasal, unknown primary
- Prior radiation or systemic therapy
- Active autoimmune disease, immunosuppression
- Hearing loss or neuropathy, grade 2 or more
- Performance status 2 or more (needed 0 or 1: fully active, or light work)
Who actually enrolled (CPS 1 or more, 682 patients)
What it showed
59.7 vs 29.6 months
Median event-free survival, CPS 1 or more.
About 1 in 10
Major pathological response: 9.4% overall, 9.8% with CPS 1 or more, 13.7% with CPS 10 or more; 0% in the control arm.
Not mature
Overall survival is immature at the first interim analysis, with no trend toward harm. Toxicity is real: fatal adverse reactions in 1.1% before surgery and 5% during adjuvant pembrolizumab; 17% stopped adjuvant pembrolizumab for side effects; hypothyroidism in about 1 in 4 (FDA label). Treatment-related grade 3 or higher events were similar in both arms (45% versus 43%) because radiation drives most of them (Uppaluri 2025).
Testing order was fixed: CPS 10 or more, then CPS 1 or more, then everyone (10 · 1 · all).
Rows 1 to 3: all 714 patients. Hypopharynx row: CPS 1 or more.
Distant control improved; local control did not.Distant spread was prevented more than local recurrence. The hypopharynx estimate is exploratory and tiny, but it points the wrong way (FDA label).
Four nuances
Six weeks of pembrolizumab rarely cost the operation
About 88% reached surgery in both arms. Progression that cancelled surgery was 4% with pembrolizumab and 1.7% in the control arm while waiting for surgery (15 and 6 patients); fewer declined surgery with pembrolizumab (3% versus 7%). The trial reimaged at week 6; so should you.
Hypopharynx: be careful
51 patients with CPS 1 or more, HR 2.28 (0.79 to 6.56), and fewer clean resections with pembrolizumab (74% versus 87%). Too small to conclude, too worrying to ignore.
p16-positive oropharynx: no real data
Only 4% of the CPS 1 or more population was HPV-positive, about 27 people. No subgroup estimate is worth quoting.
Before, after, or both?
The trial was not built to say which phase did the work. Do not skip the adjuvant year because the neoadjuvant doses went well.
More detail: pathology, subgroups, survival, CPS
The pathology came back better
Extranodal extension 25.5% versus 36.4%; positive margins, as defined in that analysis, 19.9% versus 29.9% (ASCO 2026). So fewer needed cisplatin: 107 of 275 (39%) versus 139 of 275 (51%) of those who started adjuvant treatment (FDA label).
Benefit held across surgical results
Event-free hazard ratios favored pembrolizumab in every surgical subgroup: clean resection (R0, no tumor at the margin) 0.67, positive margins 0.58, negative margins 0.78, extranodal extension 0.77. Several confidence intervals cross 1. Exploratory.
Overall survival: not yet
Immature at the first interim analysis (76% of planned deaths in CPS 1 or more). In CPS 10 or more, HR 0.72 was not statistically significant at this analysis.
CPS under 1 is outside the label
About 4% of the trial (27 of 714). The FDA indication requires CPS 1 or more by an FDA-authorized test.
The bigger picture
Before surgery, or with postoperative chemoradiation: positive. With definitive chemoradiation, or as maintenance alone after local treatment: negative.KEYNOTE-412 and JAVELIN added a PD-1 or PD-L1 inhibitor to definitive chemoradiation; IMvoke010 gave atezolizumab alone for a year after definitive treatment, which could include surgery. NIVOPOST-OP is not FDA-approved for this use; its high-risk definition was broad: extranodal extension, positive margins, 4 or more nodes, or multiple perineural invasion (Bourhis 2026).
Cisplatin stays the partner, even in good-prognosis HPV disease.Replacing cisplatin with cetuximab lowered 5-year survival from 84.6% to 77.9% in RTOG 1016, with no less toxicity (Gillison 2019); De-ESCALaTE HPV reached the same answer (Mehanna 2019).
Why high-risk pathology gets cisplatin in both arms.Two 2004 trials added cisplatin to postoperative radiation; a pooled analysis found the benefit in patients with extranodal extension or positive margins (Bernier 2004; Cooper 2004; Bernier 2005). KEYNOTE-689 kept that rule and added pembrolizumab on top.
Is KEYNOTE-689 for this patient?
Already operated, and the pathology is high risk?Adjuvant nivolumab with cisplatin-radiation improved disease-free survival in NIVOPOST-OP (HR 0.76; ages 19 to 74 only), but it is not FDA-approved for this use.
Back to our base of tongue
By AJCC 8, which the trial used: T4N2 would have qualified, T3N2 would not. Qualifying is not the same as best.Unequivocal ENE that makes him N3 by Version 9 does not change the AJCC 8 check. For a T4 HPV-positive tongue base, surgery is rarely the preferred local treatment; the trial holds about two dozen HPV-positive patients; and cisplatin-radiation, without immunotherapy, is the standard to beat. Bring one question to tumor board for the surgeon: would you operate for cure, and how would he swallow afterwards?
Memory hooks
Check yourself
Question 1p16-positive tonsil, T2N1, CPS 20. The patient asks for "the immunotherapy before surgery".
Not this trial's population: p16-positive needed T4. Standard local treatment (transoral surgery or cisplatin-radiation), no checkpoint inhibitor.
Question 2Oral cavity T4aN2b, resectable, CPS 0.
Inside the trial's stage rules, outside the FDA label (CPS 1 or more). Surgery and risk-adapted adjuvant treatment. If the pathology is high risk, NIVOPOST-OP supports adjuvant nivolumab with cisplatin-radiation, though it is not FDA-approved for this use.
Question 3After two doses the tumor is a bit larger but still resectable.
Operate. The decision rule is resectability, not shrinkage: early growth on a checkpoint inhibitor can be immune inflammation rather than true progression, and only 4% lost surgery to true progression. Continue the adjuvant plan.
Question 4Oral cavity T2 N2b (stage IVA), resectable, CPS 15, performance status 0, no autoimmune disease.
Offer it: pembrolizumab x2, reimage at week 6, surgery, then adjuvant treatment by pathology. This is the patient the trial was built for.
Question 5He had pembrolizumab x2 and surgery. The specimen shows extranodal extension. Adjuvant plan?
Radiation with cisplatin 100 mg/m² every 3 weeks for 3 doses, plus pembrolizumab x3, then pembrolizumab alone x12.
Question 6Larynx T4a, the patient chooses organ preservation with chemoradiation. Add pembrolizumab?
No. Adding a checkpoint inhibitor to definitive chemoradiation failed (KEYNOTE-412, JAVELIN Head and Neck 100). KEYNOTE-689 applies only to the surgical path.
Sources
Retrieved from PubMed, the FDA label and ClinicalTrials.gov, read on Oct 26, 2026.
- Uppaluri R, Haddad RI, Tao Y, et al. Neoadjuvant and adjuvant pembrolizumab in locally advanced head and neck cancer. N Engl J Med 2025. doi:10.1056/NEJMoa2415434
- KEYTRUDA (pembrolizumab) prescribing information, sections 6.1 and 14.4. DailyMed, label updated July 2026. dailymed.nlm.nih.gov
- KEYNOTE-689 eligibility criteria. ClinicalTrials.gov NCT03765918. clinicaltrials.gov
- Merck. KEYNOTE-689 results presented at AACR 2025: median event-free survival, major pathological response, overall survival by population. Press release, April 27, 2025. businesswire.com
- The ASCO Post. Perioperative pembrolizumab plus surgery and risk-adapted standard adjuvant therapy (distant metastasis-free survival, locoregional control). Sept 25, 2025. ascopost.com
- Adkins D, Haddad R, Tao Y, et al. Efficacy by surgical outcomes in KEYNOTE-689. ASCO 2026, as reported by OncLive, July 8, 2026. onclive.com
- Machiels JP, Tao Y, Licitra L, et al. Pembrolizumab plus concurrent chemoradiotherapy versus placebo plus concurrent chemoradiotherapy (KEYNOTE-412). Lancet Oncol 2024. doi:10.1016/S1470-2045(24)00100-1
- Lee NY, Ferris RL, Psyrri A, et al. Avelumab plus standard-of-care chemoradiotherapy versus chemoradiotherapy alone (JAVELIN Head and Neck 100). Lancet Oncol 2021. doi:10.1016/S1470-2045(20)30737-3
- Haddad R, Fayette J, Teixeira M, et al. Atezolizumab in high-risk locally advanced squamous cell carcinoma of the head and neck (IMvoke010). JAMA 2025. doi:10.1001/jama.2025.1483
- Bourhis J, Auperin A, Borel C, et al. Nivolumab added to cisplatin and radiotherapy after surgery for squamous cell carcinoma of the head and neck at high risk of relapse (NIVOPOST-OP). Lancet 2026. doi:10.1016/S0140-6736(25)01850-1
- Bernier J, Domenge C, Ozsahin M, et al. Postoperative irradiation with or without concomitant chemotherapy for locally advanced head and neck cancer. N Engl J Med 2004. doi:10.1056/NEJMoa032641
- Cooper JS, Pajak TF, Forastiere AA, et al. Postoperative concurrent radiotherapy and chemotherapy for high-risk squamous-cell carcinoma of the head and neck. N Engl J Med 2004. doi:10.1056/NEJMoa032646
- Bernier J, Cooper JS, Pajak TF, et al. Defining risk levels in locally advanced head and neck cancers: a comparative analysis of concurrent postoperative radiation plus chemotherapy trials. Head Neck 2005. doi:10.1002/hed.20279
- Gillison ML, Trotti AM, Harris J, et al. Radiotherapy plus cetuximab or cisplatin in HPV-positive oropharyngeal cancer (NRG Oncology RTOG 1016). Lancet 2019. doi:10.1016/S0140-6736(18)32752-1
- Mehanna H, Robinson M, Hartley A, et al. Radiotherapy plus cisplatin or cetuximab in low-risk human papillomavirus-positive oropharyngeal cancer (De-ESCALaTE HPV). Lancet 2019. doi:10.1016/S0140-6736(18)32779-X
Transfusion in pregnancy: the number is not the trigger
Two pregnant women in clinic today with iron deficiency, hemoglobin around 10. Neither needs blood. Here is who does.
Concepts first
Anemia cutoffs by trimester: under 11, 10.5 and 11 g/dL (ACOG, the American College of Obstetricians and Gynecologists, 2021).
Three patients share one low hemoglobin
Anemic, not bleeding
Treat the cause: iron by mouth, or intravenous when oral fails or time is short (second and third trimesters only, usually 1000 mg in 1 or 2 doses). Expect about 1 g/dL in 2 weeks and the full effect at 4 to 6 weeks, so after about 36 weeks it may come too late. Transfuse for symptoms at rest, or consider it below 6 g/dL.
Bleeding now
The hemoglobin lags: early in a bleed it can still look normal. Follow measured loss, pulse and the shock index (heart rate divided by systolic pressure; 0.9 or more is an early warning even with normal pressure; over 1 is an abnormal sign in the WHO definition). Give products by protocol; check fibrinogen early and replace it if low.
Bled, now stable
First confirm the bleeding has stopped (think hematoma or retained tissue) and repeat the hemoglobin in 12 to 24 hours. Then intravenous iron, and transfuse for symptoms or instability. WOMB trial: see Question 4.
Hook: iron for the anemic, blood for the bleeding. Below six, consider blood even when all looks fine.Observational data link a maternal hemoglobin under 6 g/dL to abnormal fetal heart rate tracings, low amniotic fluid and fetal death, so ACOG advises considering transfusion for fetal reasons at that level (ACOG 2021). You do not need abnormal fetal testing first.
Does she need red cells?
Our two clinic patients, at about 10 g/dL: the last box. Iron, and a recheck in 2 to 4 weeks.
The familiar 7 g/dL comes from adults who were not pregnant.The 2023 AABB guideline (the main US blood banking society) makes a strong recommendation to consider transfusion below 7 g/dL in stable hospitalized adults; 7.5 g/dL for cardiac surgery and 8 g/dL for orthopedic surgery or cardiovascular disease (Carson 2023). Pregnant women were essentially not studied, so obstetric decisions rest on bleeding, symptoms, delivery timing and the fetus (ACOG 2021).
Bleeding after birth: measure, then act as one bundle
Act early.The 2025 guideline from the World Health Organization with the international obstetric (FIGO) and midwifery (ICM) federations diagnoses hemorrhage within 24 hours of birth at 300 mL of measured loss plus an abnormal vital sign (pulse over 100, shock index over 1, systolic under 100 or diastolic under 60), or at 500 mL measured; the evidence comes mostly from vaginal births. ACOG (2017) defines it as 1000 mL or more, or any loss with signs of low blood volume, and calls more than 500 mL at vaginal birth abnormal. Do not wait for 1000 mL.
E-MOTIVE: early detection, then everything at once
Early detection with a calibrated drape or weighing: eyeballing misses about half (sensitivity 48%, Cochrane 2025), and in E-MOTIVE the drape raised detection from 51% to 93%. Then together: Massage the uterus, Oxytocic drugs (no methylergonovine with hypertension or preeclampsia; no carboprost with asthma), Tranexamic acid, IV fluids (about 1.5 to 2 L of crystalloid, then blood products; less in preeclampsia, where fluid causes pulmonary edema), Examination of the genital tract, and Escalation. Severe outcomes fell 60% (vaginal births in African hospitals where the bundle was rarely used before; the gain in US units is likely smaller).
The hematology part: tranexamic acid
1 g intravenous over 10 minutes, as soon as hemorrhage is diagnosed and within 3 hours of birth. Repeat 1 g once if bleeding continues after 30 minutes or restarts within 24 hours. Not after 3 hours. Intravenous only: an intrathecal mix-up with spinal anesthetic can be fatal (FDA warning). Deaths from bleeding 1.2% versus 1.7% when given within 3 hours (WOMAN); every 15 minutes of delay costs benefit. Not for routine prevention at cesarean (Pacheco 2023).
Measure fibrinogen early
A level of 2 g/L (200 mg/dL) or less at the start strongly predicted severe bleeding (128 women; Charbit 2007). Replace it when low, to above 2 g/L: about 10 units of cryoprecipitate or about 4 g of concentrate, then recheck. A fixed 2 g dose given to women with normal fibrinogen did not help (FIB-PPH). OBS2 subgroups suggest replacement is not needed above 2 g/L.
Massive transfusion
Activate the protocol for loss over 1500 mL still going, more than 2 units given with bleeding ongoing, a shock index over 1, or suspected clotting failure. Do not wait for unit counts. Start with the protocol pack; in obstetrics fibrinogen falls first, so give it early and switch to lab or viscoelastic guidance (ROTEM or TEG) once results are back. Keep ionized calcium above 1.1 mmol/L, temperature above 36 °C with a fluid warmer, pH above 7.2, platelets above 75 x 109/L.
Which blood
Every pregnant woman
Type and antibody screen early. If antibodies are present, compatible blood can take hours: tell the blood bank days before a high-risk delivery.
RhD-negative mother
RhD-negative red cells; group O RhD-negative is the emergency unit. In a massive bleed the blood bank may switch to RhD-positive cells: life first. Give anti-D after an RhD-positive baby, dosed by Kleihauer-Betke or flow cytometry after a large bleed or cell salvage, and within 72 hours if she received RhD-positive platelets.
Any woman who may become pregnant
UK guidance selects K-negative red cells (Kell antigen negative), because anti-K causes fetal anemia (Milkins 2013). Most US blood banks do not do this routinely, and it must never delay emergency blood. Ask your blood bank.
Sickle cell disease
Extended Rh (C, c, E, e) and K matched; check her antibody history. Her usual hemoglobin is often 6 to 8, so the pregnancy thresholds do not apply: treat by symptoms and complications, and do not raise it above about 10 g/dL. Planned transfusion is not routine: observational data suggest benefit, but there is no adequate randomized trial; consider it after severe prior complications (BSH, Oteng-Ntim 2021).
Declines blood
Before labor, write down item by item what she accepts (fibrinogen concentrate, cryoprecipitate, cell salvage, tranexamic acid). Intravenous iron early, erythropoietin if needed. Deliver where cell salvage and interventional radiology exist, and agree in advance on an earlier threshold for hysterectomy.
Memory hooks
Check yourself
Question 128 weeks, hemoglobin 6.8, ferritin 6, no symptoms, normal fetal testing. The night resident says "under 7, transfuse".
No transfusion today. The 7 g/dL rule comes from inpatients who were not pregnant (AABB 2023). She is stable, has no symptoms, fetal testing is normal, and she has about 10 weeks to term. Give intravenous iron, look for the cause of such severe deficiency (hidden bleeding, hemoglobinopathy, folate), send a type and antibody screen, and recheck in 1 to 2 weeks. Rethink if she bleeds, develops symptoms, goes into labor, or does not respond as delivery approaches.
Question 230 weeks, hemoglobin 5.6, abnormal fetal testing.
Transfuse: below 6 g/dL is a fetal indication, and here the fetus already shows it. Send the antibody screen first, give one unit at a time and reassess (circulatory overload risk), and involve maternal-fetal medicine on fetal causes and delivery timing. Then find and treat the cause.
Question 3Placenta accreta, planned cesarean at 35 weeks, hemoglobin 8.5.
The number is not the trigger, the operation is. Intravenous iron now, delivery at a center with a multidisciplinary accreta team, antibody screen days ahead, large-bore access, at least 4 to 6 crossmatched units with plasma ready, massive transfusion protocol and cell salvage on standby. Once bleeding starts, transfuse early and ahead of losses: accreta bleeding outruns the vital signs.
Question 4Day 1 after a 1500 mL hemorrhage. Hemoglobin 6.8, walking, mild fatigue, stable vital signs.
Usually no transfusion, once you confirm the bleeding has stopped and the repeat hemoglobin is stable. Intravenous iron is preferred over oral after a large bleed. WOMB trial (hemoglobin 4.8 to 7.9 g/dL 12 to 24 hours after a bleed of 1000 mL or more, no severe symptoms): women not transfused were slightly more tired at day 3 and 1 week, had no more complications, and 1 in 8 later needed blood for symptoms; non-inferiority was not formally met (Prick 2014). WOMB included women down to 4.8 g/dL. Below about 6 g/dL many clinicians still transfuse, but that is practice, not trial evidence.
Question 5Ongoing bleeding at 1200 mL, fibrinogen 1.6 g/L (160 mg/dL), 2 hours after birth.
Activate the massive transfusion protocol now: the loss is under 1500 mL, but bleeding continues with fibrinogen under 2 g/L, which is suspected clotting failure, one of the triggers, and the strongest early sign this bleed will become severe. Tranexamic acid 1 g over 10 minutes (still within 3 hours), treat the cause (tone first, then trauma, tissue, clotting), and raise fibrinogen above 2 g/L with about 10 units of cryoprecipitate or about 4 g of concentrate, rechecking after each dose.
Sources
Retrieved from PubMed, the World Health Organization, Cochrane and the FDA, read on Oct 26 and 27, 2026.
- American College of Obstetricians and Gynecologists. Anemia in pregnancy. ACOG Practice Bulletin No. 233. Obstet Gynecol 2021. doi:10.1097/AOG.0000000000004477
- Committee on Practice Bulletins, Obstetrics. Postpartum hemorrhage. ACOG Practice Bulletin No. 183. Obstet Gynecol 2017. doi:10.1097/AOG.0000000000002351
- World Health Organization, FIGO, ICM. Consolidated guidelines for the prevention, diagnosis and treatment of postpartum haemorrhage. October 2025. who.int
- Coomarasamy A, Devall AJ, Bell S, et al. Diagnosis and treatment of postpartum haemorrhage: a race against time. Lancet 2026. doi:10.1016/S0140-6736(26)01031-7
- Gallos I, Devall A, Martin J, et al. Randomized trial of early detection and treatment of postpartum hemorrhage (E-MOTIVE). N Engl J Med 2023. doi:10.1056/NEJMoa2303966
- Cochrane review CD016134. How accurate are tests to diagnose postpartum haemorrhage? 2025. cochrane.org
- WOMAN Trial Collaborators. Effect of early tranexamic acid administration on mortality, hysterectomy, and other morbidities in women with post-partum haemorrhage (WOMAN). Lancet 2017. doi:10.1016/S0140-6736(17)30638-4
- Gayet-Ageron A, Prieto-Merino D, Ker K, et al. Effect of treatment delay on the effectiveness and safety of antifibrinolytics in acute severe haemorrhage. Lancet 2018. doi:10.1016/S0140-6736(17)32455-8
- US Food and Drug Administration. Risk of inadvertent intrathecal injection of tranexamic acid. fda.gov
- Pacheco LD, Clifton RG, Saade GR, et al. Tranexamic acid to prevent obstetrical hemorrhage after cesarean delivery. N Engl J Med 2023. doi:10.1056/NEJMoa2207419
- Charbit B, Mandelbrot L, Samain E, et al. The decrease of fibrinogen is an early predictor of the severity of postpartum hemorrhage. J Thromb Haemost 2007. doi:10.1111/j.1538-7836.2007.02297.x
- Wikkelsø AJ, Edwards HM, Afshari A, et al. Pre-emptive treatment with fibrinogen concentrate for postpartum haemorrhage: randomized controlled trial (FIB-PPH). Br J Anaesth 2015. doi:10.1093/bja/aeu444
- Collins PW, Cannings-John R, Bruynseels D, et al. Viscoelastometric-guided early fibrinogen concentrate replacement during postpartum haemorrhage: OBS2. Br J Anaesth 2017. doi:10.1093/bja/aex181
- Prick BW, Jansen AJ, Steegers EA, et al. Transfusion policy after severe postpartum haemorrhage: a randomised non-inferiority trial (WOMB). BJOG 2014. doi:10.1111/1471-0528.12531
- Carson JL, Stanworth SJ, Guyatt G, et al. Red blood cell transfusion: 2023 AABB international guidelines. JAMA 2023. doi:10.1001/jama.2023.12914
- Pavord S, Daru J, Prasannan N, et al. UK guidelines on the management of iron deficiency in pregnancy. Br J Haematol 2020. PMID 31578718
- Milkins C, Berryman J, Cantwell C, et al. Guidelines for pre-transfusion compatibility procedures in blood transfusion laboratories. Transfus Med 2013. doi:10.1111/j.1365-3148.2012.01199.x
- Oteng-Ntim E, Pavord S, Howard R, et al. Management of sickle cell disease in pregnancy. A British Society for Haematology guideline. Br J Haematol 2021. doi:10.1111/bjh.17671
- Malinowski AK, Shehata N, D'Souza R, et al. Prophylactic transfusion for pregnant women with sickle cell disease: a systematic review and meta-analysis. Blood 2015. PMID 26302758