New ECG Morphologic Criteria for the Identification of Left Bundle Branch Capture

Authors, Journal, Affiliations, Type, DOI

• Authors: Óscar Cano, MD, PhD; Carmen Arveras, MD; Javier Navarrete, MD; Pablo Jover, MD; Valero Soriano, MD; Enrique Robles, MD; Josep Navarro, MD; Maite Izquierdo, MD, PhD; Joaquín Osca, MD, PhD; Luis Martínez-Dolz, MD, PhD; Pugazhendhi Vijayaraman, MD
• Journal: JACC: Clinical Electrophysiology 2025;11(9):2031–2044. ORIGINAL RESEARCH — CIED / Physiological Pacing. © 2025 by the American College of Cardiology Foundation, published by Elsevier. Manuscript received 7 January 2025; revised 9 April 2025; accepted 1 May 2025
• Affiliations: Hospital Universitari i Politècnic La Fe, Valencia, Spain; Instituto de Investigación Sanitaria La Fe; CIBERCV, Madrid; and the Geisinger Heart Institute, Wilkes Barre, Pennsylvania, USA (Vijayaraman)
• Type: Observational, single-centre, retrospective diagnostic-accuracy study (n=115), with a second phase of blinded qualitative validation by two independent observers and a third-observer tiebreak. No prospective arm, no control group — the comparator is within-patient LVSc during the same implant
• DOI: 10.1016/j.jacep.2025.05.006
• Disclosures: Dr Cano — speaker honoraria and consultancy fees from Abbott, Biotronik, Boston Scientific, Medtronic and MicroPort. Dr Vijayaraman — honoraria from Medtronic, Biotronik and Boston Scientific; consultant for Medtronic and Abbott; research funding from Medtronic; advisory board for Abbott; and holds a patent for a His bundle pacing delivery tool. All other authors report no relationships
• Why it matters here: the wiki's capture-confirmation criteria come chiefly from Jastrzębski's validation study (sources/lbb-capture-criteria-hrs-2021) and the EHRA consensus (sources/csp-ehra-2023). This paper re-tests those criteria in a new cohort and reports markedly lower sensitivities — see Contradictions

Overview

Identifying left bundle branch capture during LBBAP is the field's persistent bottleneck, and the accepted reference standard — QRS transition during decremental output pacing — is available in under 30% of cases. This study proposes two purely morphologic ECG criteria that need no calipers at all. The first is the paced V6 QRS upstroke/downstroke pattern: a fast/slow (or fast/fast) pattern indicates LBB capture, while slow/fast or slow/slow indicates LV septal capture. Quantitatively the V6 upstroke measured 34.6 ± 10.1 ms during LBBc versus 63.2 ± 12.1 ms during LVSc (P < 0.001), and an upstroke/downstroke ratio <1 achieved sensitivity 0.96 / specificity 0.95 (AUC 0.985). The second criterion is the appearance of new QRS downstroke slurring/notching in any lead — or an "M" pattern in the inferior leads — not present during lead penetration. In the validation phase, two independent observers using no measurements whatsoever classified 86.7% of LBBc and 88.8% of LVSc correctly (kappa 0.70). The criteria perform well in narrow QRS and non-cardiomyopathy patients and degrade substantially in dilated cardiomyopathy (70.4% vs 93%, P = 0.004). Two structural caveats bound the result: every patient was included because they had a QRS transition, so the criteria are derived in the minority who have the gold standard and then proposed for the majority who do not; and the study's own comparison shows the established criteria behaving very differently here than in the papers that proposed them.

Keywords

(No author keyword list is printed. Indexed under CIED — Physiological Pacing.)

Key Takeaways

Abstract

• Background: "Identification of left bundle branch capture (LBBc) during left bundle branch area pacing remains challenging"
• Objectives: to validate the utility of new simple morphologic criteria for the identification of LBBc
• Methods: patients with proven LBBc based on the presence of QRS transition during decremental output pacing were included. The paced V6 QRS upstroke/downstroke was evaluated quantitatively and qualitatively, classified as fast or slow, with fast upstroke patterns associated with LBBc and slow upstroke patterns with LV septal capture (LVSc). Additionally, the appearance of paced QRS downstroke slurring/notching in any lead not previously present during lead penetration, and/or an "M" QRS pattern in inferior leads, was considered suggestive of LBBc. Accuracy was tested by independent evaluators using exclusively qualitative ECG morphologic features
• Results: 115 patients. Mean V6 upstroke duration during LBBc significantly shorter than during LVSc — 34.6 ± 10.1 ms vs 63.2 ± 12.1 ms; P < 0.001. Paced V6 upstroke/downstroke ratio <1: sensitivity 0.97 and specificity 0.95 (note: this contradicts the Results and Table 3 — see Limitations). A slurred/notched downstroke or "M" pattern in inferior leads was present in 91.1% during LBBc in baseline narrow QRS patients with inferior paced axis, with sensitivity 0.91 and specificity 0.87. Using exclusively qualitative criteria, independent evaluators identified the correct capture pattern in 87% of LBBc and 89% of LVSc cases, with diagnostic accuracy significantly lower among dilated cardiomyopathy patients: 70.4% vs 93%; P = 0.004
• Conclusions: simple ECG morphologic criteria can accurately identify LBBc during LBBAP "in patients with baseline narrow QRS and/or without cardiomyopathy"

Introduction — the gap being addressed

• The problem statement: identification of LBBc during LBBAP "remains challenging and is usually time consuming", with different diagnostic methods described — in contrast to His bundle pacing, "where the capture criteria are simple and clearly defined"
• The reference standard's own limitation, stated up front: QRS transition during decremental output pacing is "still considered the gold standard", but was present in only 26% of cases in the MELOS registry. Moreover it is usually a transient phenomenon visible shortly after lead deployment, and may disappear over time
• Programmed ventricular stimulation may identify QRS transitions but is time consuming and difficult to interpret
• The LBB potential route, and its known unreliability: an LBB potential with proof of capture — LBBp-to-V6-RWPT during intrinsic rhythm equal to stimulus-to-V6-RWPT during pacing ±10 ms — is "also considered a reliable criterion of LBBc, but it's only seen in around 20% to 77% of cases"
• Consequent dependence on measurement: "quantitative criteria based on precise ECG measurements using an electrophysiology recording system or digital calipers have been described and are frequently required"
• The hypothesis: paced QRS activation patterns during LBBc should resemble the morphology seen during normal AV conduction, and these patterns should be distinguishable from the cell-to-cell myocardial activation of LVSc using both quantitative and qualitative methods

Methods

Design and population

• Observational, single-centre, retrospective study
• Patients with conventional pacemaker or CRT indications who received an LBBAP lead and with proven LBB capture criteria based on QRS transition during decremental output pacing
• Institutional review board approval (Hospital Universitari i Politècnic La Fe); informed consent obtained
• Implantation: conventional technique — dedicated sheath positioned in the right ventricular mid septal area using anatomical references, unipolar pacing from the lead tip to select the insertion site, then a Medtronic SelectSecure 3830 lumenless lead deployed in the interventricular septum
• LBBc criteria evaluation: unipolar decremental output pacing from the lead tip, using asynchronous pacing at a heart rate at least 20 beats/min faster than the patient's baseline rhythm. Only patients with a documented QRS transition — NS-LBBP→S-LBBP or NS-LBBP→LVSP — were included

The two new morphologic criteria

Criterion 1 — paced V6 QRS upstroke/downstroke pattern (defined as a discrete variable, and also evaluated quantitatively to establish a cut-off):

• Fast/slow or fast/fast = typical of LBBc
• Slow/fast or slow/slow = typical of LVSc
• The authors began by studying spontaneous QRS patterns in patients with preserved AV conduction or AV block with a narrow escape rhythm

Criterion 2 — new QRS downstroke slurring/notching:

• The appearance of any QRS downstroke slurring/notching in any lead not previously present during lead penetration was considered suggestive of LBBc
• Additionally, in patients with an intermediate or inferior paced QRS axis (predominantly positive QRS in lead II and/or III), a slurred QRS downstroke or "M" QRS pattern in lead II or III was also considered indicative of LBBc
• The same slurring or notching occurring in the QRS upstroke was considered suggestive of LVSc

Measurement protocol

• All measurements in an EP recording system using digital calipers at a sweep speed of 100 to 200 mm/s
• V6 upstroke measured from the initial onset of the QRS to the peak of the QRS in V6; during spontaneous rhythm this is the V6 RWPT in standard nomenclature
• V6 downstroke measured from the peak of the QRS in V6 to the lowest descending point in the same lead
• A predominantly positive QRS in V6 was necessary; where an S-wave >20% of the QRS amplitude in V6 was present, measurements were made in leads I or aVL instead
• Three measurements per patient, mean used
• Paced QRS patterns were evaluated during proven NS-LBBc and during previous LVSc obtained during lead penetration — but NOT during QRS transition, "to avoid potential confounding or mixed paced QRS patterns"

The four categories, by prespecified quantitative rule

• The 50 ms threshold was derived from the patients' own spontaneous rhythm — mean V6 upstroke 41.8 ± 18.6 ms and downstroke 60.2 ± 23.4 ms in those with preserved AV conduction

Blinded qualitative validation (the second phase)

• Two independent observers (C.A., P.V.) trained in the new criteria and blinded to any previous measurement or capture-pattern classification
• Only the V6 lead (predominantly positive QRS, <20% S-wave) was shown initially
• Disagreements resolved by a third independent observer
• A fast pattern was defined qualitatively by a "sharp" QRS onset/offset with a steep upstroke/downstroke slope; a slow pattern by a pseudo-delta wave appearance or a shallow slope
• No measurements allowed

Statistics

• Continuous variables as mean ± SD or median (IQR) if skewed; categorical as absolute frequencies, compared by chi-square or exact test
• Continuous compared by Student's t-test; tests 2-sided; P ≤0.05 significant
• ROC curves for the upstroke/downstroke ratio, overall and by subgroup
• Sensitivity, specificity, PPV, NPV for the binary patterns
• McNemar chi-square 2-sided test to compare sensitivities of established versus new criteria
• Kappa statistic for interobserver variability
• SPSS version 28

Results

The population (Table 1, N = 115)

• Age 68 ± 20 years; 51 female (44%); BMI 26 ± 5
• Hypertension 66 (57%); dyslipidaemia 54 (47%); diabetes 38 (33%); chronic kidney disease 23 (20%); dilated cardiomyopathy 29 (25%); ischaemic heart disease 11 (10%); valvular heart disease 28 (24%)
• Baseline LVEF 53 ± 16%; LVEF <50% in 44 (38%)
• Interventricular septum width 11.9 ± 2.6 mm; LA diameter 39 ± 8 mm
• Baseline QRS morphology: narrow QRS 16 (14%); isolated RBBB 14 (12%); bifascicular block 32 (28%); AV block with narrow escape 10 (9%); AV block with wide escape 23 (20%); paced with no escape 20 (17%)
• Pacing indication: SND 10 (9%); AVB 68 (64%); CRT 28 (27%)
• Transitions seen (Table 2): NS-LBBc→S-LBBc 85 (74%); NS-LBBc→LVSc 24 (21%); both 6 (5%) — all 115 therefore had a QRS transition by inclusion
• LB potential recorded in 75 (65%) — but 89% in baseline narrow QRS vs 33% in LBBB (P = 0.001) (see Contradictions)

Spontaneous rhythm as the reference (Table 2)

• Mean baseline V6 upstroke 42.9 ± 18.7 ms, downstroke 61.8 ± 23.8 ms
• Narrow QRS vs LBBB: 29.8 ± 5.3 ms vs 63.1 ± 16.5 ms (P < 0.001)
• Baseline upstroke/downstroke ratio 0.66 ± 0.20 (0.69 ± 0.19 narrow QRS vs 0.81 ± 0.33 LBBB; P = 0.078)
• 92.2% showed a fast/slow or fast/fast pattern during spontaneous rhythm; the 7.8% who did not were all except one patients with baseline LBBB morphology
• Downstroke slurring in inferior leads during spontaneous rhythm in 22 patients (19.1%)

The primary quantitative result

• Only 6 patients had a ratio >1 during proven LBBc, and 4 had a ratio <1 during LVSc — so the ratio <1 behaves almost as a clean binary
• Ratio <0.87 gave the best combination (SN 0.91, SP 0.99, PPV 0.99, NPV 0.92)
• Ratio <1: SN 0.96, SP 0.95, PPV 0.95, NPV 0.96 (Table 3's authoritative figures)
• AUC 0.985 (95% CI 0.967–1)

Pattern distribution

• During LBBc (n = 115): fast/slow 105 (91.3%); fast/fast 5 (4%); slow/slow 4 (4%); slow/fast 1 (1%)
• Fast/slow or fast/fast during LBBc: SN 0.96, SP 0.97, PPV 0.96, NPV 0.96
• During LVSc (n = 115): slow/fast 96 (83.5%); slow/slow 15 (13%); fast/fast 3 (2.6%); fast/slow 1 (0.9%)
• Slow/fast or slow/slow during LVSc: SN, SP, PPV and NPV all 0.97

The slurring/notching criterion

• Descending-limb slurring/notches in 66.1% during LBBc vs 12.2% during LVSc
• Ascending-limb slurring was present in 16% during LVSc vs 1% during LBBc — the mirror image, and the reason upstroke slurring is read as LVSc
• When present, slurring/notching was predominantly in the inferior leads (82%) and less often precordial (18%)
• Among the 93 patients with an inferior/intermediate paced axis: slurred downstroke or "M" pattern in inferior leads in 73.1% of LBBc vs 9.7% of LVSc; SN 0.73, SP 0.90, PPV 0.88, NPV 0.77
• Restricting to patients with an inferior paced axis specifically: present in up to 91.1% during LBBc vs 13.3% during LVSc; SN 0.91, SP 0.87, PPV 0.87, NPV 0.91
• In the overall population the criterion is far weaker: SN 0.59, SP 0.91 (68/115 vs 10/115)

The blinded qualitative validation — the study's most practical result

• Without any measurement, observers classified during true LBBc: fast/slow 77.9%, fast/fast 8.9%, slow/slow 10.2%, slow/fast 3.1%
• During true LVSc: slow/slow 55.6%, slow/fast 33.2%, fast/slow 9%, fast/fast 1.8%
• Correct identification: 86.7% of LBBc and 88.8% of LVSc
• Interobserver agreement: kappa 0.70, percent agreement 0.85 — "considered optimal"

Subgroup findings, which are where the criteria break down

• Baseline narrow QRS vs LBBB (both with preserved AV conduction): paced V6 upstroke during LBBc 31.6 ± 4.9 ms vs 36 ± 7.9 ms (P < 0.017), with all narrow-QRS patients showing a ratio <1. Ratio 0.59 ± 0.10 vs 0.66 ± 0.13 (P = 0.063). A cut-off of 0.88 gave SN and SP of 1 in baseline narrow QRS patients
• Dilated cardiomyopathy (n = 29 vs 86): paced V6 upstroke during LBBc 41.5 ± 16.8 ms vs 32.3 ± 4.7 ms (P = 0.007); during LVSc 71.5 ± 14.6 vs 60.4 ± 9.8 ms (P < 0.001)
  • 5/29 DCM patients (17.2%) showed a slow/fast or slow/slow pattern during proven LBBc, whereas all non-DCM patients had fast/slow or fast/fast (P < 0.001)
  • Qualitative diagnostic accuracy: 93% (non-DCM) vs 70.4% (DCM); P = 0.004, mainly driven by more slow/slow patterns in DCM during LBBc (25.9% vs 7%; P = 0.013)
  • Interobserver agreement was 0.88/kappa 0.74 for DCM vs 0.84/kappa 0.68 for non-DCM

Head-to-head against the established criteria (Table 3) — the most consequential table here

• The established criteria are highly specific but insensitive in this cohort — "the specificity of both V6 RWPT and interpeak interval was high (ranging from 90% to 100%), but sensitivity was low, especially for the V6 to V1 interpeak interval (only 20%)"
• The new criteria beat them where it counts — the V6 upstroke <50 ms and the ratio <1 outperformed V6 RWPT <75 and <80 ms in narrow QRS or isolated RBBB (McNemar P < 0.001), but were not superior to V6 RWPT <85 ms or <100 ms in baseline wide QRS
• The new criteria remained accurate in patients who FAILED the V6 RWPT criterion — a fast/slow or fast/fast pattern was present during LBBc in 83.3% of narrow-QRS/isolated-RBBB patients with V6 RWPT ≥75 ms (vs 97.1% with <75 ms; P = 0.090), and in 83.8% of the wide-QRS group with V6 RWPT ≥80 ms (vs 86.6%; P = 0.658)

Discussion

Principal findings

• Two novel morphologic criteria for LBBc: (1) the paced V6 upstroke/downstroke pattern; (2) the appearance of QRS downstroke slurring/notching in any lead not previously present during lead penetration
• The motivating logic: multiple quantitative criteria exist "requiring specific and meticulous measurements during the implant procedure", while the gold standard — QRS transition during decremental output pacing — "is only obtained in <30% of cases". "Reproducible, identifiable by eyeball, fast and reliable diagnostic methods are warranted"
• The distinction from V6 RWPT: "The paced V6 upstroke derives from the well-known and established V6 RWPT but is measured from the real QRS onset and not from the pacing spike. In this manner, it does not include the isoelectric interval frequently present during LBBc and, occasionally, during LVSc"

The pathophysiological rationale

• Normal His-Purkinje activation is fast, in contradistinction to the slow cell-to-cell depolarisation of myocardial pacing — so patients with preserved intraventricular conduction or RBBB have a fast QRS upstroke, whereas true LBBB exhibits a slow QRS onset, as does RV pacing
• Therefore direct LBB capture should produce a fast QRS upstroke — but the authors are careful about the limit of the analogy: LBBc activation "is not identical to the spontaneous activation through the conduction system during normal AV conduction as LBBP is, by definition, nonselective and thus also associated with partial myocardial capture"
• The two-wavefront model explaining the downstroke: after the initial fast phase there is co-existence of two activation fronts — one through the conduction system and one from local myocardial cell-to-cell activation. "The amount of cell-to-cell depolarization and the collision of these 2 activation wave fronts would be responsible of the slow QRS downstroke and the potential appearance of 'notching' and 'slurring' in the descending limb" during LBBc
• The mirror image in LVSc: slow initial activation through LV septal myocardium (like RV pacing) with rapid ulterior conduction-system engagement — "initially slow QRS upstroke during which 'notching' and 'slurring' may occur once again as a consequence of these 2 different wave front's collision but this time with preponderance of slow myocardial activation"
• An explicit warning against an all-or-nothing reading: the balance between the two wavefronts "should not be considered as an 'all or nothing' behavior and dynamic changes in the QRS appearance during decremental output pacing may be identified, explaining the presence of subtle QRS morphologic changes seen during pacing threshold testing and cases with apparently more than one QRS transition identified." This is why the authors evaluated the pattern outside the transition
• The origins of the 50 ms cut-off: spontaneous-rhythm V6 upstroke 41.8 ± 18.6 ms and downstroke 60.2 ± 23.4 ms, which is what set "fast" at ≤50 ms

Previously reported criteria, and why they fall short here

• Only a few criteria are considered gold standard: QRS transition during decremental output pacing; a selective LBBc response during programmed ventricular stimulation; and visualisation of an LB potential with proof of capture (LB potential-to-V6 RWPT during spontaneous rhythm equal to spike-to-V6 RWPT ±10 ms)
• The rest are demoted: "The remaining criteria, including V6 RWPT or the V6 to V1 interpeak interval, may be highly suggestive of LBBc but cannot be considered definitive as they may be influenced by many different factors including the final lead position within the interventricular septum, the myocardial mass, the LV volumes, or the heart disposition within the chest. Moreover, all of them are quantitative criteria requiring precise measurements during the procedure"
• The practical stakes of getting it wrong, stated plainly: "The under-recognition of LBBc may have significant clinical implications and may lead to further unnecessary lead penetration attempts that may result in complications such as septal perforation."
• The forward-looking use case: "These quantitative measures may play a role for the automatic identification of the different LBBAP capture patterns by using specifically designed algorithms incorporated to the devices"

Study Limitations (the authors' own)

• ⚠ The independent observers were trained on the same data set later analysed for SN, SP and predictive values. Their mitigation: ECGs were randomly displayed to the observers, and only the V6 lead was shown at an initial step, so they "believe that this methodological limitation did not significantly affect the final results"
• The LVSc comparison patterns were not contemporaneous with the LBBc ones — evaluated outside the transition but immediately before achieving LBBc. To show this was fair, the LVSc V6 RWPT and interpeak intervals inside and outside the transition were compared in the 30 patients who showed an NS-LBBc→LVSc transition, giving comparable means with all values within a 10-ms interval for both measures
• Subjectivity is inherent: "As with any qualitative measurement, the identification of paced QRS morphologic patterns is exposed to the presence of significant interobserver subjectivity" — addressed by using two independent observers
• Validation is internal only: "Prospective evaluation of the newly defined criteria should confirm their validity"
• DCM remains unsolved: "Patients with DCM still represent the most challenging population to establish a real diagnosis of LBBc and further research in this specific subgroup should be encouraged"

Conclusions and Perspectives

• Quantitative or qualitative assessment of the paced V6 QRS upstroke/downstroke pattern "is a reliable method for the identification of LBBc in the absence of QRS transition during decremental output pacing", as is the appearance of downstroke slurring/notching or an "M" pattern not previously present during lead penetration
• COMPETENCY IN MEDICAL KNOWLEDGE: identification of LBBc "remains challenging and time consuming"; these two criteria "can easily establish the diagnosis", and completely qualitative criteria "may avoid the need for precise measurements during LBBAP implant procedures"
• TRANSLATIONAL OUTLOOK: "The accuracy of these new ECG-based morphologic criteria in patients with baseline wide QRS and/or dilated cardiomyopathy is still limited and needs to be improved." Future incorporation into device automatic algorithms "may facilitate the identification of conduction system capture during follow-up"

Limitations of the document

Authors' own stated limitations

• The observers were trained on the same dataset later used to compute the diagnostic metrics — a training/testing overlap. The mitigations offered (random display order, V6-only at first) reduce but do not remove the concern, and the authors' phrasing ("we believe that this methodological limitation did not significantly affect the final results") is an assertion, not a demonstration
• The LVSc comparator was not truly contemporaneous — it comes from an earlier moment in the same implant ("during lead penetration"), not from the same beats. The 30-patient check that inside- and outside-transition values were comparable is a reasonable reassurance, but it is a subgroup check used to validate a comparison made in all 115
• Subjective assessment is intrinsic to the method, and prospective validation is explicitly still pending
• DCM patients remain unsolved by the authors' own admission, and further research is "encouraged"

Additional limitations evident in the study

• ⚠ The cohort is defined by the very gold standard the criteria are meant to replace — a circularity the paper never addresses. Inclusion required a documented QRS transition during decremental output pacing, and Table 2 confirms all 115 had one (85 NS→S, 24 NS→LVSc, 6 both). Yet the motivating premise is that this transition "is only obtained in <30% of cases" and the criteria are proposed precisely "in the absence of QRS transition." The criteria are therefore derived and validated exclusively in the minority who have the transition, and then recommended for the majority who do not — the population where they are needed is untested. None of the five stated limitations mentions this
• No external validation and no prospective arm, in a single centre, from a group with one author holding a patent on a His bundle pacing delivery tool and extensive industry relationships (Cano: five device manufacturers; Vijayaraman: Medtronic research funding). The technique described is also bound to one lead (Medtronic 3830 lumenless)
• Three mutually inconsistent statements of the same headline metric, with Table 3 authoritative at SN 0.96 / SP 0.95:

Table 3's raw fractions (110/115 LBBc, 6/115 LVSc) give SN 0.957 and SP 0.948, confirming 0.96/0.95. The abstract overstates sensitivity and the discussion swaps the two figures; a reader taking the abstract or the discussion at face value gets a different criterion performance than the data support

• The abstract and body describe the "M"-pattern subgroup differently. The abstract says the pattern was present in 91.1% "in baseline narrow QRS patients with inferior paced QRS axis"; the body attributes the 91.1% to "only patients with an inferior paced QRS axis" (and separately reports a 76.1% figure for narrow-QRS patients with inferior or intermediate axis). The 91.1% figure's denominator is therefore ambiguous across the paper
• The cohort is unusual in ways that bound generalisability. 74% had an NS-LBBc→S-LBBc transition and only 21% NS→LVSc, so the "superior" separation may partly reflect a population unusually rich in selective capture. Pacing indication was 64% AV block and only 9% sinus node disease, and 25% had DCM — a different mix from the bradycardia-dominated SDL registry
• Septal thickness is reported (11.9 ± 2.6 mm) but never used, despite depth being the mechanism by which lead position determines the ECG. No analysis relates lead depth to the upstroke pattern, even though the discussion names "final lead position within the interventricular septum" as a factor influencing the established criteria
• The "no measurement needed" claim is narrower than the framing suggests. The qualitative arm worked (86.7%/88.8%), but the study excluded the transition beats themselves as confounded, required a predominantly positive V6 (falling back to I or aVL otherwise), and the criteria degrade to 70.4% accuracy in DCM and are not superior to existing criteria in baseline wide QRS. So "eyeball the V6 lead" is reliable chiefly in the narrow-QRS, non-cardiomyopathy patient
• No clinical outcome is attached to any of it. Improved capture adjudication is assumed to reduce unnecessary penetration and perforation — the paper raises that risk explicitly — but no perforation, attempt-count, procedure-time or follow-up data are reported, so the benefit remains inferred
• Rating rationale: rated medium. Strengths: n=115 — comparable to the largest capture-criteria cohort in this wiki — with a blinded two-observer qualitative validation, a measured kappa (0.70), a prespecified four-category rule, and an unusually candid head-to-head table that reports the established criteria's weakness rather than only its own strength. Weaknesses: single-centre, retrospective, no external or prospective validation, a cohort selected by the reference standard it seeks to replace, observers trained on the evaluation set, and a headline metric stated three different ways. It is a genuinely useful clinical contribution whose evidence base is one internal validation — hence medium, not high

Key Concepts Mentioned

• concepts/Left-Bundle-Branch-Area-Pacing — the technique; this paper addresses how to identify its capture pattern
• concepts/Conduction-System-Pacing — the umbrella category
• concepts/Cardiac-Refractoriness — not addressed here, but the QRS-transition gold standard this paper works around is the output-dependent capture transition the LBBAP page describes
• Paced V6 upstroke/downstroke pattern — the new criterion; fast/slow or fast/fast = LBBc, slow/fast or slow/slow = LVSc
• The V6 upstroke, and how it differs from V6 RWPT — measured from the real QRS onset, not the pacing spike, so it excludes the isoelectric interval
• QRS downstroke slurring/notching, and the "M" pattern in inferior leads — the second criterion; upstroke slurring conversely suggests LVSc
• The two-wavefront collision model — conduction-system activation plus local myocardial activation, whose balance determines the upstroke/downstroke appearance and the site of notching
• LBBc vs LVSc — the distinction the criteria exist to draw; the paper uses LVSc (left ventricular septal capture) throughout
• QRS transition during decremental output pacing — the reference standard, and this cohort's inclusion criterion
• V6 RWPT and V6–V1 interpeak interval — the established criteria, re-tested here and found specific but insensitive
• Interobserver kappa 0.70 — the reported reproducibility of the purely qualitative assessment

Key Entities Mentioned

• concepts/Left-Bundle-Branch-Area-Pacing — primary concept page
• Medtronic SelectSecure 3830 — the lumenless lead used for every implant; the standard lead of the lumenless literature in this wiki
• Hospital Universitari i Politècnic La Fe, Valencia — the single investigating centre
• Geisinger Heart Institute — Vijayaraman's institution; the same author appears across sources/lbbap-jacc-2019 and the wider capture-criteria literature
• MELOS registry — cited for the 26% QRS-transition figure that motivates the whole paper
• Óscar Cano — corresponding author; disclosed relationships with Abbott, Biotronik, Boston Scientific, Medtronic and MicroPort
• Pugazhendhi Vijayaraman — co-author; Medtronic research funding and a patent on a His bundle pacing delivery tool
• American College of Cardiology Foundation / Elsevier — publisher (JACC: Clinical Electrophysiology)

Wiki Pages Updated

• Created: wiki/sources/ecg-lbbap-criteria-jaccep-2025.md
• Updated: wiki/concepts/Left-Bundle-Branch-Area-Pacing.md
• Updated: wiki/sourceindex.md
• Updated: wiki/wikiindex.md