Colorectal and esophageal adenomas and early carcinomas
Conditions
Interventions
Group 1: Phase 1: Baseline (2 weeks, approx. 30 patients)
standard care (without AHEAD use)
retrospective data collection from medical records
content of the data (patient demographics: age, gender - comorbidities: ASA-classification, Charlson Comorbidity Index - medication: anticoagulation - procedure details: type, duration, complexity, te
the purpose is to establish the baseline hospitalization rate, to document the baseline processes, create a comparison group for the Hawthorne effect
anonymized data extraction after completion of phase 1.
Group 2: Phase 2: (3-4 weeks, approx. 45-50 patients)
Stand supply with AHEAD use, data integration is not visible to operating personnel
prospective data collection via AHEAD system
Data content (from patient portal, bevor procedure: self-assessment of health status, current symptoms, medication intake, social situation (care at home), preference for outpatient vs. inpatient
from endoscopy report during/after procedure: procedure type and details, ASGE grade (complexity), duration of procedure, technical success, complications during prucedure, protection classification
from the structured doctor's assessment, after procedure: vital parameters after procedure, clinical condition, pain (NRS), suspected complications, overall assessment of the risk)
The aim is to evaluate the feasibility of structured multi-source data collection and quantification of the Hawthorne effect
Data extraction in week 3-4.
Group 3: Phase 3: (12 weeks, approximately 180 patients)
AHEAD data collection with rule-based visualization
Sponsors
Universitätsmedizin Mannheim, II. medizinische Klinik
Eligibility
Sex/Gender
All
Age
18 Years to No maximum
Inclusion criteria
Inclusion criteria: Planned interventional endoscopic procedure: ERCP (Endoscopic Retrograde Cholangiopancreatography), EUS-FNB (Endosonographic Fine Needle Biopsy), Polypectomy (=1 cm), EMR (Endoscopic Mucosal Resection), ESD (Endoscopic Submucosal Dissection), Esophageal Dilatation/Bouging, other interventional procedures at the discretion of the PI Procedure in the outpatient/inpatient border area (i.e. the decision to hospitalize is clinically relevant) Capacity to consent: Patient is capable of consent and can give informed consent Language skills: sufficient knowledge of the german language for use of the patient portal (or availability of a translator)
Exclusion criteria
Exclusion criteria: Emergency procedures: Acute emergencies in which hospitalization is already certain before the procedure Planned hospitalization: Patients for whom hospitalization has already been determined for other reasons (e.g. other illness, planned further procedure) Lack of ability to consent: Patients who are unable to give consent (e.g. dementia, acute confusion) Refusal to participate in the study: Patient refuses to participate Technical impossibility: Patient cannot use patient portal (e.g. no smartphone/computer, no internet connection) and no support available Withdrawal of consent: Patient withdraws consent Protocol Violation: Major protocol violation Security Concerns: Unexpected security issues
Design outcomes
Primary
| Measure | Time frame |
|---|---|
| Goal 1: Technical feasibility of multi-source data integration Definition: Successful integration and completeness of data captured from all four data sources. Operationalization: - HIS data: =95% completeness (automatic transmission) - Patient portal: =80% completeness (patient-reported) - Endoscopy report: =98% completeness (doctor-documented) - Structured doctor assessment: =95% completeness (doctor-rated) Success criterion: All four data sources reach their target values in phase 3. Measurement time: Continuously during phases 2 and 3 Goal 2: User-friendliness (Usability) Definition: The AHEAD frontend is intuitive, efficient and acceptable to end users (doctors, nurses). Operationalization: - System Usability Scale (SUS): Average score =70 (corresponds to “good” to “excellent”) - Time required: data entry per patient <10 minutes - Acceptance rate: =80% of users would want to use the system regularly Success criterion: SUS =70 and time required <10 minutes in phase 3. Measurement time: End of phase 3 (SUS questionnaire, structured interviews) Goal 3: Data quality and reliability Definition: The data collected is consistent, concordant and reliable. Operationalization: - Concordance HIS ? physician assessment: Cohen’s Kappa (?) =0.7 for categorical variables - Concordance self-assessment ? physician assessment: ? =0.6 (lower threshold due to different perspectives) - Intra-rater reliability: ICC =0.8 (substudy with 10% re-assessment) Success criterion: concordance target values achieved in phase 3. Measurement time: Continuously during phases 2 and 3 Goal 4: quantification of the Hawthorne effect Definition: Measuring behavior change solely through structured data collection (without data aggregation/visualization). Operationalization: - Comparison of the hospitalization rate in phase 1 (baseline) vs. phase 2 (structured recording without visualization) - Effect size (Cohen’s h) with 95% confidence interval Success criterion: Hawthorne effect quantified (rega | — |
Secondary
| Measure | Time frame |
|---|---|
| Secondary goal 1: Technical system performance Metrics: - System uptime: =99% - Error rate: <1% - Response time: <2 sec. (normal load), <5 sec. (peak load) Secondary goal 2: Exploratory clinical outcomes Metrics (purely descriptive, no hypothesis testing): - Hospitalization rates in phases 1, 2, 3 (with 95% CI) - Length of stay (median, IQR) - Complication rates (safety monitoring) - Resource utilization (exploratory) Secondary goal 3: Subgroup analyzes (exploratory) Subgroups: - by procedure type (ERCP, EUS, polypectomy, esophageal dilatation/bougienage, etc.) - by ASA-classification (I/II vs. III/IV) - by procedure complexity (ASGE grade) - by time of day (morning vs. afternoon) Secondary goal 4: Learning curve analysis Question: Does user-friendliness and data quality improve over time (learning effect)? Analysis: Time series analysis of SUS scores, time expenditure, completion rates over phase 3 (12 weeks) Secondary goal 5: Classifier pre-training evaluation Question: How well does the classifier pre-trained with retrospective data perform on the pilot study data? Metrics: - Sensitivity, specificity for predicting hospitalization - Area Under the Curve (AUC) - Calibration (Calibration Plot) Note: This is a purely exploratory analysis in preparation for the definitive study. | — |
Countries
Germany
Contacts
Public ContactSebastian Belle
Universitätsmedizin Mannheim, II. medizinische Klinik
Outcome results
None listed