Executive overview

Across the GLP-1 receptor agonist clinical programs reviewed (spanning approximately 2014–2025), eCOA use is heterogeneous. Most trials prioritized clinical/biomarker endpoints (e.g., HbA1c, weight, ultrasound, DXA, troponin), while eCOA was used selectively to capture patient-centric domains such as appetite/hunger, disordered eating, treatment satisfaction, adherence, quality of life, usability of digital tools, and global impressions. The predominant eCOA modality is electronic Patient-Reported Outcomes (ePRO). Instances of eDiary-like digital dietary tools appear in feasibility/usability contexts. Electronic Clinician-Reported Outcomes (eClinRO) and eObsRO are rarely reported.

Based on the analyzed records, at least a dozen trials incorporated eCOA, encompassing more than fifteen distinct instruments or instrument families (e.g., VAS scales for appetite/symptoms; EAT‑26; TSQM‑9; TAPQ; IWQOL‑Lite‑CT; IW‑SP; TRIM‑D; DTSQs; CGI; GI-specific questionnaires; SUS; modified Rawl; digital dietary tools like MyFood24/Nutritics‑Libro/Intake24). Most were used as secondary or exploratory endpoints; notable exceptions used eCOA as primary endpoints in specific settings ^{1331321311282172142139910326026125625797922702830297271298244248163160158159161240276294296119170171172173174}.

What eCOA instruments were used and where

Appetite, hunger, satiety, and related symptoms (VAS scales; ePRO)

• Visual analogue scales (VAS; 0–10 cm) for appetite/satiety/hunger in a semaglutide study of individuals with obesity; used as secondary/exploratory endpoints ^260261256257.
• VAS for hunger, thirst, and anxiety in perioperative patients taking GLP‑1 RAs during a study of fasting and residual gastric content; used as secondary/exploratory endpoints ^{163160158159161}.
• VAS ratings of hunger, satiety, fullness, and prospective food consumption in a T2D population (exenatide plus dapagliflozin study); collected as self-reported scales (patient-reported), with electronic modality not explicitly specified in the source ¹¹⁹.

Significance: These VAS-based ePROs complement objective measures by quantifying perceived appetite/symptom burden, particularly in obesity and perioperative risk contexts ^{260261256257163160158159161119}.

Eating behavior/disordered eating (ePRO)

• Eating Attitudes Test (EAT‑26) questionnaire in semaglutide‑treated patients with type 2 diabetes and overweight; used for a primary endpoint (change in EAT‑26 score at 3 months) ^217216214213.
• Food diaries were kept for 14 days at the beginning and near end of the trial, though whether these were electronic was not specified ²¹⁴.

Significance: Demonstrates a case where a patient‑reported instrument (EAT‑26) was central to efficacy assessment, capturing behavioral change under GLP‑1 RA treatment in T2D ^217216214213.

Treatment satisfaction and adherence (ePRO)

• Abbreviated Treatment Satisfaction Questionnaire for Medication (TSQM‑9) and Treatment Adherence Perception Questionnaire (TAPQ) in adults who had lost >10% body weight on GLP‑1 RAs; used as secondary/exploratory outcomes (primary endpoint was objective weight maintenance via Bluetooth scales/EHR) ^9910398100.

Significance: These instruments quantify perceived treatment experience and adherence dynamics after GLP‑1 RA cessation, complementing device‑captured weight outcomes ^9910398100.

Weight‑related quality of life and self‑perception (ePRO)

• In tirzepatide (Mounjaro) Phase 3, Impact of Weight on Quality of Life‑Lite Clinical Trials version (IWQOL‑Lite‑CT) and Impact of Weight on Self‑Perception (IW‑SP) were used as eCOA instruments ²⁴⁰. These (and other PROs below) were used as secondary endpoints ^276294296.
• Additional patient‑reported measures collected electronically included:
  • Ability to Perform Physical Activities of Daily Living (APPADL)
  • Diabetes Treatment Satisfaction Questionnaire (DTSQ)
  • European Quality of Life 5 Dimensions 5 Levels (EQ‑5D‑5L) ^276294296

Significance: Highlights broader patient‑centric benefit assessment beyond glycemic endpoints in a T2D program ^240276294296.

Diabetes‑specific PROs in fixed‑combination GLP‑1 RA therapy (PROs; electronic modality not specified)

• In Xultophy (insulin degludec/liraglutide) studies, TRIM‑D and DTSQs were used as patient‑reported outcome instruments; they were assessed as secondary endpoints, while the primary endpoint was change in HbA1c ^979227095271.

Significance: These PROs capture treatment impact on daily diabetes management and satisfaction in a GLP‑1 RA–containing regimen ^9792270.

Digital dietary tools and usability assessments (ePRO; eDiary‑like tools; usability)

• In adults with obesity initiating semaglutide (Wegovy), the feasibility/usability study evaluated:
  • Digital dietary assessment tools: MyFood24, Nutritics‑Libro, Intake24 (patient‑reported intake tools)
  • Usability instruments: System Usability Scale (SUS) and a modified Rawl Usability Questionnaire The study emphasized ePRO-based acceptability/usability rather than endpoint hierarchy ^{133132131128129130}.

Significance: Illustrates eDiary-like intake capture and formal usability evaluation as GLP‑1 RA programs integrate digital health tools in obesity management ^133132131128.

Clinician impressions (eClinRO; as characterized in the source)

• Clinical Global Impressions (CGI) scale was used as a secondary endpoint in an adolescent study of semaglutide (Wegovy) for antipsychotic‑associated obesity; the protocol synopsis did not explicitly designate eCOA modality, but the data summary identifies CGI as the eCOA used for secondary endpoints ²⁸³⁰²⁹.

Significance: CGI adds a clinician‑anchored appraisal of overall status in a complex adolescent obesity population ²⁸³⁰²⁹.

GI adverse event characterization (patient‑reported; electronic modality not specified)

• GI‑specific questionnaires to characterize gastrointestinal adverse events were administered in the NG101 study among adults on GLP‑1 agonists; instrument names and electronic modality were not specified, and endpoint role was not explicitly designated ^{170171172173174}.

Significance: Focused capture of GI tolerability can inform supportive care strategies around GLP‑1 RA–related side effects ^{170171172173174}.

Functional walking performance (designated as eCOA in the source)

• Initial Claudication Distance (ICD) and Absolute Claudication Distance (ACD) were described as an eCOA primary endpoint in a liraglutide study of T2D with lower‑extremity arterial disease ^7271298.

Significance: This dataset characterizes ICD/ACD as an eCOA primary endpoint in a vascular complications population treated with GLP‑1 RA ^7271298.

Device‑based endothelial function (designated as eCOA in the source)

• EndoPAT2000 was described as an eCOA used as a secondary endpoint in a liraglutide trial in non‑stroke volunteers ²⁴⁴²⁴⁸.

Significance: While commonly considered an instrumented physiological measurement, the source dataset classifies it as an eCOA in this context ²⁴⁴²⁴⁸.

Additional contextual ePRO use (not GLP‑1 RA–focused endpoints)

• In a cardiovascular genomic risk study where GLP‑1 RAs could be initiated when clinically indicated, health behaviors (physical activity, diet, tobacco use, sleep) were assessed by electronic self‑reported surveys (ePRO) to compute LE8; GLP‑1 RAs were not the primary focus of the trial’s eCOA outcomes ^{125126300123124}.

Where eCOA was not used (illustrative examples)

Many GLP‑1 RA trials relied exclusively on clinical, imaging, or laboratory endpoints, with no reported eCOA:

• Semaglutide vs. dulaglutide on epicardial adipose tissue (retrospective EMR and ultrasound endpoints) ^219220221222.
• Acute pulmonary embolism add‑on semaglutide (biomarkers, imaging; no eCOA) ^{178179180181182}.
• Preoperative residual gastric volumes with semaglutide (POCUS ultrasound; no eCOA) ^{140141142143144145}.
• Empagliflozin ± semaglutide in T2D with albuminuria (urine albumin endpoints; no eCOA) ^{207208209210211212}.
• Dulaglutide (Trulicity) approval program focused on HbA1c, fasting glucose, weight; no eCOA reported in the supporting regulatory documents ^{302301253278252}.
• Additional T2D/obesity and mechanistic studies without reported eCOA include fat redistribution, angiogenesis biomarkers, bone/tissue imaging, and several observational datasets ^{62606164632224252627232232242252262272281234}.

Indications and contexts where eCOA appears

• Obesity/weight management: Appetite/symptom VAS, QoL/weight‑perception (IWQOL‑Lite‑CT, IW‑SP), treatment satisfaction (TSQM‑9), adherence (TAPQ), digital dietary tools and usability (MyFood24, Nutritics‑Libro, Intake24; SUS; modified Rawl), perioperative symptom VAS ^{26026125625724027629429699103133132131128163160158159}.
• Type 2 diabetes: Disordered eating (EAT‑26, primary), diabetes‑specific PROs (TRIM‑D, DTSQs), daily living function and generic QoL in tirzepatide T2D program (APPADL, EQ‑5D‑5L), appetite VAS in obese T2D (with exenatide), vascular walking performance (ICD/ACD) ^{21721621421397922702762942961197271298}.
• Special populations/contexts: Adolescent antipsychotic‑associated obesity (CGI), non‑stroke volunteers (EndoPAT2000 as designated), GI‑AE characterization in GLP‑1 RA users (GI‑specific questionnaires) ^{283029244248170171172173174}.

Endpoint roles and examples

• Primary eCOA endpoints:

  • EAT‑26 for disordered eating in semaglutide‑treated T2D patients ^217216214213.
  • ICD/ACD (as designated in the source) in liraglutide‑treated T2D with lower‑extremity arterial disease ^7271298.

• Secondary or exploratory eCOA endpoints:

  • Appetite/symptom VAS in semaglutide obesity and perioperative fasting contexts ^{260261256257163160158159161}.
  • Treatment experience and adherence (TSQM‑9; TAPQ) post‑GLP‑1 RA cessation ^9910398100.
  • QoL/weight perception (IWQOL‑Lite‑CT; IW‑SP), APPADL, DTSQ, EQ‑5D‑5L in tirzepatide Phase 3 ^240276294296.
  • TRIM‑D and DTSQs in Xultophy combination therapy ^9792270.
  • CGI in adolescent antipsychotic‑associated obesity ²⁸³⁰²⁹.
  • EndoPAT2000 in non‑stroke volunteers (as designated) ²⁴⁴²⁴⁸.

• Not designated (feasibility/usability or unspecified):

  • Digital dietary tools and usability (MyFood24, Nutritics‑Libro, Intake24; SUS; modified Rawl) focused on feasibility rather than endpoint hierarchy ^133132131128.
  • GI‑specific AE questionnaires (instrument names/modality not specified) ^{170171172173174}.
  • Appetite VAS in T2D with exenatide plus dapagliflozin (electronic modality and endpoint hierarchy not explicitly specified) ¹¹⁹.

Key takeaways and implications

• ePRO dominates: Most eCOA usage centers on patient‑reported symptoms (appetite, hunger, perioperative discomfort), behavior (disordered eating), treatment experience (satisfaction, adherence), and QoL/functional status (IWQOL‑Lite‑CT, IW‑SP, APPADL, EQ‑5D‑5L) ^{260261256257217214213991032402762942969792270119163160158159}.
• Regulatory programs employ eCOA primarily as secondary measures: Pivotal programs like tirzepatide used multiple PROs as secondary endpoints alongside clinical primary endpoints (HbA1c, weight), reflecting a supportive role for patient‑centric evidence ^240276294296.
• Notable primary eCOA use exists in focused settings: EAT‑26 as a primary endpoint in disordered eating (T2D) and (as designated in the dataset) ICD/ACD in PAD‑complicated T2D demonstrate contexts where eCOA can be central to the research objective ^{2172162142137271298}.
• Digital tools are emerging: Diet logging platforms and usability instruments are being evaluated alongside GLP‑1 RA therapy, indicating integration of eDiary‑like capture and user‑experience assessment in obesity care pathways ^{133132131128129130}.
• eClinRO/eObsRO remain rare in this dataset: Apart from CGI being noted as a secondary eCOA in one adolescent study, clinician‑ or observer‑reported eCOAs are not prominent; the bulk of eCOA is patient‑reported ²⁸³⁰²⁹.

Methodology and scope

This summary synthesizes the analyzed trial records spanning approximately 2014–2025 across GLP‑1 RA indications (type 2 diabetes, obesity/weight management, and special contexts). It identifies eCOA instruments explicitly reported in the processed data, preserves their original attributions (including primary/secondary roles where stated), and highlights cases where modality or endpoint hierarchy was unspecified. Examples are cited directly from the dataset’s references to maintain traceability ^{1331321311282172142139910326026125625797922702830297271298244248163160158159161240276294296119170171172173174}.