Tina Baykaner
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Tina Baykaner has a total of 380 co-authors. Here are the top co-authors by number of shared publications:
- Sanjiv M. Narayan
Department of Medicine, Stanford University, Palo Alto, California
20 shared publications Muhammad Fazal
14 shared publications- Albert Joseph Rogers
Stanford University
13 shared publications - Paul J. Wang
Stanford University School of Medicine, CA (P.J.W., M.P.).
10 shared publications Paul J. Wang
Division of Cardiology, Department of Medicine, Stanford University School of Medicine, Stanford, USA
7 shared publications- Junaid A. B. Zaman
Keck School of Medicine of USC , Los Angeles , United States of America
6 shared publications
Latest Publications
- 2 July 2026
- Published by Wolters Kluwer Health in Circulation: Arrhythmia and Electrophysiology
- Vol. 19 (7) , e015039
Tau20-RETROmapping System Identifies Driver-Like Activation Patterns in Real-Time During Persistent Atrial Fibrillation
BACKGROUND: Conventional activation mapping of atrial fibrillation (AF) during clinical procedures is limited by low-amplitude, fractionated electrograms and cycle length variability. The Tau20-RETROmapping stimulator-recorder system (TauRhythm Therapies, United Kingdom) is an investigational device used to identify nonpulmonary vein AF drivers by real-time, high-density activation mapping of uniform wavefronts during AF. We validated the accuracy of the system and applied it to left atrial mapping for evidence of AF drivers. METHODS: Left atrial geometry was acquired using a 3-dimensional electroanatomic mapping system with high-density mapping catheters (CARTO 3 with Optrell, or EnSite X with HD Grid). Electrograms were recorded for 30 seconds at multiple left atrial sites. The Tau20-RETROmapping system generates activation maps of organized wavefronts. System performance was manually validated against randomly sampled local electrograms using a grid sweep of thresholds to identify Pareto-optimal parameters in sinus rhythm, atrial pacing, atrial tachycardia (AT), and AF. The system was then applied to identify putative driver activation patterns during persistent AF. RESULTS: We studied 24 patients undergoing pulmonary vein isolation for AF. Six patients presented in sinus rhythm or AT, in whom the system demonstrated a sensitivity of 100.0% (95% CI, 80.5%–100.0%) and a specificity of 100.0% (85.2%–100.0%). Eighteen patients were mapped in AF: the system had a sensitivity of 94.4% (86.2%–98.4%) and a specificity of 97.1% (89.9%–99.6%) when identifying organized AF wavefronts, and accurately identified the earliest activation in 89.6% (79.7%–95.7%) of waves. Stable propagation originating from the left atrial appendage (LAA) was observed in 3/18 patients, while 8/18 demonstrated competing propagation toward and away from the LAA. Conduction away from the LAA was detected along the anterior (8/18 patients), lateral (8/18 patients), and posterior (7/18 patients) walls. No propagation near the LAA was observed in 6/18 patients. CONCLUSIONS: The Tau20-RETROmapping system enables real-time activation mapping of AF, with the most convincing driver-like areas being identified around the LAA.
- 2 July 2026
- Published by Wolters Kluwer Health in Circulation: Arrhythmia and Electrophysiology
- Vol. 19 (7) , e015039
Tau20-RETROmapping System Identifies Driver-Like Activation Patterns in Real-Time During Persistent Atrial Fibrillation
BACKGROUND: Conventional activation mapping of atrial fibrillation (AF) during clinical procedures is limited by low-amplitude, fractionated electrograms and cycle length variability. The Tau20-RETROmapping stimulator-recorder system (TauRhythm Therapies, United Kingdom) is an investigational device used to identify nonpulmonary vein AF drivers by real-time, high-density activation mapping of uniform wavefronts during AF. We validated the accuracy of the system and applied it to left atrial mapping for evidence of AF drivers. METHODS: Left atrial geometry was acquired using a 3-dimensional electroanatomic mapping system with high-density mapping catheters (CARTO 3 with Optrell, or EnSite X with HD Grid). Electrograms were recorded for 30 seconds at multiple left atrial sites. The Tau20-RETROmapping system generates activation maps of organized wavefronts. System performance was manually validated against randomly sampled local electrograms using a grid sweep of thresholds to identify Pareto-optimal parameters in sinus rhythm, atrial pacing, atrial tachycardia (AT), and AF. The system was then applied to identify putative driver activation patterns during persistent AF. RESULTS: We studied 24 patients undergoing pulmonary vein isolation for AF. Six patients presented in sinus rhythm or AT, in whom the system demonstrated a sensitivity of 100.0% (95% CI, 80.5%–100.0%) and a specificity of 100.0% (85.2%–100.0%). Eighteen patients were mapped in AF: the system had a sensitivity of 94.4% (86.2%–98.4%) and a specificity of 97.1% (89.9%–99.6%) when identifying organized AF wavefronts, and accurately identified the earliest activation in 89.6% (79.7%–95.7%) of waves. Stable propagation originating from the left atrial appendage (LAA) was observed in 3/18 patients, while 8/18 demonstrated competing propagation toward and away from the LAA. Conduction away from the LAA was detected along the anterior (8/18 patients), lateral (8/18 patients), and posterior (7/18 patients) walls. No propagation near the LAA was observed in 6/18 patients. CONCLUSIONS: The Tau20-RETROmapping system enables real-time activation mapping of AF, with the most convincing driver-like areas being identified around the LAA.
- 8 May 2026
- Published by Springer Nature in Journal of Interventional Cardiac Electrophysiology
- Vol. 69 (9) , 2675-2678
Focal pulsed field ablation for atrial tachycardia near the phrenic nerve: a two-case experience
Catheter ablation of atrial tachycardia (AT) adjacent to the phrenic nerve is limited by the risk of diaphragmatic injury with conventional energy sources. Pulsed field ablation (PFA) is a non-thermal modality with relative tissue selectivity that may reduce collateral damage. We describe two patients with symptomatic AT arising from the crista terminalis in close proximity to the phrenic nerve who underwent ablation with a novel focal PFA catheter (TactiFlex Duo, Abbott). Case 1: A 76-year-old man with recurrent AT after incomplete prior ablation received 12 focal PFA applications at the mid-crista terminalis. Phrenic nerve capture was preserved throughout, and sinus rhythm was restored without complication. Case 2: A 25-year-old woman with symptomatic supraventricular tachycardia underwent 12 low-energy PFA lesions at the posterolateral right atrium. Transient phrenic nerve palsy occurred but fully recovered by the end of the procedure. In both cases, durable sinus rhythm was achieved without long term phrenic injury. Focal PFA enabled acute termination of crista terminalis AT in close proximity to the right phrenic nerve, including one case with preserved phrenic capture throughout and one case with reproducible transient phrenic nerve palsy after each application with full recovery. These cases support the feasibility of focal PFA in phrenic-adjacent right atrial substrates, while underscoring that phrenic nerve stunning can occur.
- 8 May 2026
- Published by Springer Nature in Journal of Interventional Cardiac Electrophysiology
- Vol. 69 (9) , 2675-2678
Focal pulsed field ablation for atrial tachycardia near the phrenic nerve: a two-case experience
Catheter ablation of atrial tachycardia (AT) adjacent to the phrenic nerve is limited by the risk of diaphragmatic injury with conventional energy sources. Pulsed field ablation (PFA) is a non-thermal modality with relative tissue selectivity that may reduce collateral damage. We describe two patients with symptomatic AT arising from the crista terminalis in close proximity to the phrenic nerve who underwent ablation with a novel focal PFA catheter (TactiFlex Duo, Abbott). Case 1: A 76-year-old man with recurrent AT after incomplete prior ablation received 12 focal PFA applications at the mid-crista terminalis. Phrenic nerve capture was preserved throughout, and sinus rhythm was restored without complication. Case 2: A 25-year-old woman with symptomatic supraventricular tachycardia underwent 12 low-energy PFA lesions at the posterolateral right atrium. Transient phrenic nerve palsy occurred but fully recovered by the end of the procedure. In both cases, durable sinus rhythm was achieved without long term phrenic injury. Focal PFA enabled acute termination of crista terminalis AT in close proximity to the right phrenic nerve, including one case with preserved phrenic capture throughout and one case with reproducible transient phrenic nerve palsy after each application with full recovery. These cases support the feasibility of focal PFA in phrenic-adjacent right atrial substrates, while underscoring that phrenic nerve stunning can occur.
- 26 April 2026
- Published by Wiley in Journal of Cardiovascular Electrophysiology
- Vol. 37 (6) , 1376-1380
Temporary Loss of Pacemaker Capture During Focal Pulsed Field Ablation: A Case of Reversible Lead–Energy Interactions
Background: Pulsed field ablation (PFA) is increasingly used for atrial arrhythmia ablation and is considered safe in patients with cardiac implantable electronic devices, though intraprocedural lead interactions remain undercharacterized. Case: We report a 75‐year‐old woman with a dual‐chamber pacemaker who underwent focal PFA for recurrent atrial tachycardia using a 9‐mm lattice‐tip catheter. Immediately following PFA near the superior vena cava‐right atrial junction, atrial lead capture was lost with transient right atrial lead threshold elevation and T‐wave oversensing. All parameters normalized within 1 h. Conclusions: This case highlights reversible lead–energy interactions and the importance of real‐time device monitoring when ablating near indwelling leads.
- 26 April 2026
- Published by Wiley in Journal of Cardiovascular Electrophysiology
- Vol. 37 (6) , 1376-1380
Temporary Loss of Pacemaker Capture During Focal Pulsed Field Ablation: A Case of Reversible Lead–Energy Interactions
Background: Pulsed field ablation (PFA) is increasingly used for atrial arrhythmia ablation and is considered safe in patients with cardiac implantable electronic devices, though intraprocedural lead interactions remain undercharacterized. Case: We report a 75‐year‐old woman with a dual‐chamber pacemaker who underwent focal PFA for recurrent atrial tachycardia using a 9‐mm lattice‐tip catheter. Immediately following PFA near the superior vena cava‐right atrial junction, atrial lead capture was lost with transient right atrial lead threshold elevation and T‐wave oversensing. All parameters normalized within 1 h. Conclusions: This case highlights reversible lead–energy interactions and the importance of real‐time device monitoring when ablating near indwelling leads.
- 13 February 2026
- Published by Elsevier in JACC: Clinical Electrophysiology
- Vol. 12 (5) , 1113-1127
Accuracy of Implantable Loop Recorders
- 13 February 2026
- Published by Elsevier in JACC: Clinical Electrophysiology
- Vol. 12 (5) , 1113-1127
Accuracy of Implantable Loop Recorders
- Open Access
- 27 January 2026
- Published by Elsevier in Indian Pacing and Electrophysiology Journal
- Vol. 26 (3) , 310-317
Artificial intelligence in atrial fibrillation – Timely diagnosis, risk assessment and personalized management
- 27 January 2026
- Published by Elsevier in Indian Pacing and Electrophysiology Journal
- Vol. 26 (3) , 310-317
Open AccessArtificial intelligence in atrial fibrillation – Timely diagnosis, risk assessment and personalized management
- 28 October 2025
- Published by Springer Nature in Journal of Interventional Cardiac Electrophysiology
- Vol. 68 (9) , 1797-1799
Four devices, one solution: gold-coated cardiac resynchronization therapy with a defibrillator in a patient with titanium hypersensitivity
Background Titanium hypersensitivity is a rare but important cause of recurrent complications in patients with cardiac implantable electronic devices (CIEDs). Despite titanium’s biocompatibility, hypersensitivity reactions can mimic device infection, leading to repeated device removals before diagnosis. Case presentation We report the management of a 67-year-old man with non-ischemic cardiomyopathy who presented to our institution after three failed cardiac resynchronization therapy with defibrillator (CRT-D) implantations over three years, each complicated by wound dehiscence and infection-like symptoms. The patient had been diagnosed with titanium hypersensitivity previously but prior attempts to mitigate exposure, including a Gore-Tex–wrapped device were unsuccessful. Therefore, a custom gold-plated CRT-D was obtained under U.S. Food and Drug Administration (FDA) compassionate use provisions. The gold-plated CRT-D was implanted via a right-sided approach. At one-year follow-up, the surgical site remained well-healed with no recurrent complications or device-related discomfort and left ventricular ejection fraction (LVEF) improved from 29% to 50%. Conclusions This case underscores the diagnostic challenges of titanium allergy, the limitations of conventional management, and the value of multidisciplinary collaboration to navigate regulatory pathways and deliver durable, patient-specific solutions. Graphical Abstract
- 28 October 2025
- Published by Springer Nature in Journal of Interventional Cardiac Electrophysiology
- Vol. 68 (9) , 1797-1799
Four devices, one solution: gold-coated cardiac resynchronization therapy with a defibrillator in a patient with titanium hypersensitivity
Background Titanium hypersensitivity is a rare but important cause of recurrent complications in patients with cardiac implantable electronic devices (CIEDs). Despite titanium’s biocompatibility, hypersensitivity reactions can mimic device infection, leading to repeated device removals before diagnosis. Case presentation We report the management of a 67-year-old man with non-ischemic cardiomyopathy who presented to our institution after three failed cardiac resynchronization therapy with defibrillator (CRT-D) implantations over three years, each complicated by wound dehiscence and infection-like symptoms. The patient had been diagnosed with titanium hypersensitivity previously but prior attempts to mitigate exposure, including a Gore-Tex–wrapped device were unsuccessful. Therefore, a custom gold-plated CRT-D was obtained under U.S. Food and Drug Administration (FDA) compassionate use provisions. The gold-plated CRT-D was implanted via a right-sided approach. At one-year follow-up, the surgical site remained well-healed with no recurrent complications or device-related discomfort and left ventricular ejection fraction (LVEF) improved from 29% to 50%. Conclusions This case underscores the diagnostic challenges of titanium allergy, the limitations of conventional management, and the value of multidisciplinary collaboration to navigate regulatory pathways and deliver durable, patient-specific solutions. Graphical Abstract
- 1 October 2025
- Published by Wolters Kluwer Health in Circulation: Arrhythmia and Electrophysiology
- Vol. 18 (10) , e013692
Dual Epicardial and Endocardial Procedure (DEEP) for Persistent or Longstanding Persistent Atrial Fibrillation
BACKGROUND: Despite advances in endocardial catheter ablation (ECA) for persistent atrial fibrillation (PersAF), undertreatment persists, especially in ECA nonresponders and in longstanding PersAF (LSPersAF), with disappointing ablation results. These patients need effective clinical treatment options. METHODS: The DEEP (Dual Epicardial and Endocardial Procedure) was a prospective, multicenter, single-arm, investigational device exemption trial to establish the safety and effectiveness of a combined epicardial/endocardial ablation procedure with left atrial appendage exclusion for PersAF/LSPersAF. Eligibility included age 18 to 75 years; symptomatic PersAF/LSPersAF refractory to ≥1 Class I/III antiarrhythmic drug; and ≤2 previous failed ECAs. Two-stage hybrid ablation included ECA performed at 91 to 121 days after the epicardial first stage (including left atrial appendage exclusion), followed by a 90-day blanking and 90-day antiarrhythmic drug optimization period. Primary effectiveness was defined as freedom from documented atrial fibrillation/atrial flutter/atrial tachycardia episodes >30 seconds through the 12-month follow-up, absent Class I/III antiarrhythmic drugs, except previously failed antiarrhythmic drugs at doses not exceeding those previously failed. Primary safety was defined as a composite of device/procedure-related serious adverse events within 30 days of epicardial ablation and 7 days of ECA. RESULTS: Ninety patients enrolled from February 2015 to December 2020; 83.3% (75/90) were male and mean±SD age was 63.4±7.7 years. AF classification was 83.3% (75/90) PersAF/16.7% (15/90) LSPersAF, and 47.8% (43/90) had prior ECA. The composite serious adverse events rate was 6.7% (6/90 [95% CI, 2.5%–13.9%]; P P =0.0134 versus performance goal) and was 62.4% (53/85 [95% CI, 52.1%–72.7%]) through 2 years. CONCLUSIONS: A collaborative hybrid ablation approach to treating PersAF/LSPersAF is safe and effective. REGISTRATION: URL: https://www.clinicaltrials.gov ; Unique identifier: NCT02393885.
- 1 October 2025
- Published by Wolters Kluwer Health in Circulation: Arrhythmia and Electrophysiology
- Vol. 18 (10) , e013692
Dual Epicardial and Endocardial Procedure (DEEP) for Persistent or Longstanding Persistent Atrial Fibrillation
BACKGROUND: Despite advances in endocardial catheter ablation (ECA) for persistent atrial fibrillation (PersAF), undertreatment persists, especially in ECA nonresponders and in longstanding PersAF (LSPersAF), with disappointing ablation results. These patients need effective clinical treatment options. METHODS: The DEEP (Dual Epicardial and Endocardial Procedure) was a prospective, multicenter, single-arm, investigational device exemption trial to establish the safety and effectiveness of a combined epicardial/endocardial ablation procedure with left atrial appendage exclusion for PersAF/LSPersAF. Eligibility included age 18 to 75 years; symptomatic PersAF/LSPersAF refractory to ≥1 Class I/III antiarrhythmic drug; and ≤2 previous failed ECAs. Two-stage hybrid ablation included ECA performed at 91 to 121 days after the epicardial first stage (including left atrial appendage exclusion), followed by a 90-day blanking and 90-day antiarrhythmic drug optimization period. Primary effectiveness was defined as freedom from documented atrial fibrillation/atrial flutter/atrial tachycardia episodes >30 seconds through the 12-month follow-up, absent Class I/III antiarrhythmic drugs, except previously failed antiarrhythmic drugs at doses not exceeding those previously failed. Primary safety was defined as a composite of device/procedure-related serious adverse events within 30 days of epicardial ablation and 7 days of ECA. RESULTS: Ninety patients enrolled from February 2015 to December 2020; 83.3% (75/90) were male and mean±SD age was 63.4±7.7 years. AF classification was 83.3% (75/90) PersAF/16.7% (15/90) LSPersAF, and 47.8% (43/90) had prior ECA. The composite serious adverse events rate was 6.7% (6/90 [95% CI, 2.5%–13.9%]; P P =0.0134 versus performance goal) and was 62.4% (53/85 [95% CI, 52.1%–72.7%]) through 2 years. CONCLUSIONS: A collaborative hybrid ablation approach to treating PersAF/LSPersAF is safe and effective. REGISTRATION: URL: https://www.clinicaltrials.gov ; Unique identifier: NCT02393885.
- 1 February 2025
- Published by Wolters Kluwer Health in Circulation: Arrhythmia and Electrophysiology
- Vol. 18 (2) , e012860
Comparing Phenotypes for Acute and Long-Term Response to Atrial Fibrillation Ablation Using Machine Learning
BACKGROUND: It is difficult to identify patients with atrial fibrillation (AF) most likely to respond to ablation. While any arrhythmia patient may recur after acutely successful ablation, AF is unusual in that patients may have long-term arrhythmia freedom despite a lack of acute success. We hypothesized that acute and chronic AF ablation outcomes may reflect distinct physiology and used machine learning of multimodal data to identify their phenotypes. METHODS: We studied 561 consecutive patients in the Stanford AF ablation registry (66±10 years, 28% women, 67% nonparoxysmal), from whom we extracted 72 data features of electrograms, electrocardiogram, cardiac structure, lifestyle, and clinical variables. We compared 6 machine learning models to predict acute and long-term end points after ablation and used Shapley explainability analysis to contrast phenotypes. We validated our results in an independent external population of n=77 patients with AF. RESULTS: The 1-year success rate was 69.5%, and the acute termination rate was 49.6%, which correlated poorly on a patient-by-patient basis (φ coefficient=0.08). The best model for acute termination (area under the curve=0.86, Random Forest) was more predictive than for long-term outcomes (area under the curve=0.67, logistic regression; P <0.001). Phenotypes for long-term success reflected clinical and lifestyle features, while phenotypes for AF termination reflected electrical features. The need for AF induction predicted both phenotypes. The external validation cohort showed similar results (area under the curve=0.81 and 0.64, respectively) with similar phenotypes. CONCLUSIONS: Long-term and acute responses to AF ablation reflect distinct clinical and electrical physiology, respectively. This de-linking of phenotypes raises the question of whether long-term success operates through factors such as attenuated AF progression. There remains an urgent need to develop procedural predictors of long-term AF ablation success.
- 1 February 2025
- Published by Wolters Kluwer Health in Circulation: Arrhythmia and Electrophysiology
- Vol. 18 (2) , e012860
Comparing Phenotypes for Acute and Long-Term Response to Atrial Fibrillation Ablation Using Machine Learning
BACKGROUND: It is difficult to identify patients with atrial fibrillation (AF) most likely to respond to ablation. While any arrhythmia patient may recur after acutely successful ablation, AF is unusual in that patients may have long-term arrhythmia freedom despite a lack of acute success. We hypothesized that acute and chronic AF ablation outcomes may reflect distinct physiology and used machine learning of multimodal data to identify their phenotypes. METHODS: We studied 561 consecutive patients in the Stanford AF ablation registry (66±10 years, 28% women, 67% nonparoxysmal), from whom we extracted 72 data features of electrograms, electrocardiogram, cardiac structure, lifestyle, and clinical variables. We compared 6 machine learning models to predict acute and long-term end points after ablation and used Shapley explainability analysis to contrast phenotypes. We validated our results in an independent external population of n=77 patients with AF. RESULTS: The 1-year success rate was 69.5%, and the acute termination rate was 49.6%, which correlated poorly on a patient-by-patient basis (φ coefficient=0.08). The best model for acute termination (area under the curve=0.86, Random Forest) was more predictive than for long-term outcomes (area under the curve=0.67, logistic regression; P <0.001). Phenotypes for long-term success reflected clinical and lifestyle features, while phenotypes for AF termination reflected electrical features. The need for AF induction predicted both phenotypes. The external validation cohort showed similar results (area under the curve=0.81 and 0.64, respectively) with similar phenotypes. CONCLUSIONS: Long-term and acute responses to AF ablation reflect distinct clinical and electrical physiology, respectively. This de-linking of phenotypes raises the question of whether long-term success operates through factors such as attenuated AF progression. There remains an urgent need to develop procedural predictors of long-term AF ablation success.
- 1 January 2025
- Published by Wolters Kluwer Health in Circulation: Arrhythmia and Electrophysiology
- Vol. 18 (1) , e013023
Engineering of Generative Artificial Intelligence and Natural Language Processing Models to Accurately Identify Arrhythmia Recurrence
BACKGROUND: Large language models (LLMs) such as Chat Generative Pre-trained Transformer (ChatGPT) excel at interpreting unstructured data from public sources, yet are limited when responding to queries on private repositories, such as electronic health records (EHRs). We hypothesized that prompt engineering could enhance the accuracy of LLMs for interpreting EHR data without requiring domain knowledge, thus expanding their utility for patients and personalized diagnostics. METHODS: We designed and systematically tested prompt engineering techniques to improve the ability of LLMs to interpret EHRs for nuanced diagnostic questions, referenced to a panel of medical experts. In 490 full-text EHR notes from 125 patients with prior life-threatening heart rhythm disorders, we asked GPT-4-turbo to identify recurrent arrhythmias distinct from prior events and tested 220 563 queries. To provide context, results were compared with rule-based natural language processing and Bidirectional Encoder Representations from Transformer-based language models. Experiments were repeated for 2 additional LLMs. RESULTS: In an independent hold-out set of 389 notes, GPT-4-turbo had a balanced accuracy of 64.3%±4.7% out-of-the-box at baseline. This increased when asking GPT-4-turbo to provide a rationale for its answers, a structured data output, and in-context exemplars, to a balanced accuracy of 91.4%±3.8% ( P P <0.05). Results were consistent for GPT-3.5-turbo and Jurassic-2 LLMs. CONCLUSIONS: The use of prompt engineering strategies enables LLMs to identify clinical end points from EHRs with an accuracy that surpassed natural language processing and approximated experts, yet without the need for expert knowledge. These approaches could be applied to LLM queries for other domains, to facilitate automated analysis of nuanced data sets with high accuracy by nonexperts.
- 1 January 2025
- Published by Wolters Kluwer Health in Circulation: Arrhythmia and Electrophysiology
- Vol. 18 (1) , e013023
Engineering of Generative Artificial Intelligence and Natural Language Processing Models to Accurately Identify Arrhythmia Recurrence
BACKGROUND: Large language models (LLMs) such as Chat Generative Pre-trained Transformer (ChatGPT) excel at interpreting unstructured data from public sources, yet are limited when responding to queries on private repositories, such as electronic health records (EHRs). We hypothesized that prompt engineering could enhance the accuracy of LLMs for interpreting EHR data without requiring domain knowledge, thus expanding their utility for patients and personalized diagnostics. METHODS: We designed and systematically tested prompt engineering techniques to improve the ability of LLMs to interpret EHRs for nuanced diagnostic questions, referenced to a panel of medical experts. In 490 full-text EHR notes from 125 patients with prior life-threatening heart rhythm disorders, we asked GPT-4-turbo to identify recurrent arrhythmias distinct from prior events and tested 220 563 queries. To provide context, results were compared with rule-based natural language processing and Bidirectional Encoder Representations from Transformer-based language models. Experiments were repeated for 2 additional LLMs. RESULTS: In an independent hold-out set of 389 notes, GPT-4-turbo had a balanced accuracy of 64.3%±4.7% out-of-the-box at baseline. This increased when asking GPT-4-turbo to provide a rationale for its answers, a structured data output, and in-context exemplars, to a balanced accuracy of 91.4%±3.8% ( P P <0.05). Results were consistent for GPT-3.5-turbo and Jurassic-2 LLMs. CONCLUSIONS: The use of prompt engineering strategies enables LLMs to identify clinical end points from EHRs with an accuracy that surpassed natural language processing and approximated experts, yet without the need for expert knowledge. These approaches could be applied to LLM queries for other domains, to facilitate automated analysis of nuanced data sets with high accuracy by nonexperts.