Colonoscopy, Colorectal, Polyp
Conditions
Keywords
Colon, Polyp, Colorectal, Colonoscopy, AI, Narrow Band Imaging, Polyp Prediction, Artificial Intelligence
Brief summary
The study objective is to establish the efficacy of the colorectal polyp Clinical Decision Support Device (CDSD) in clinical use.
Detailed description
Prospective, single arm, multicenter study with participating endoscopists comprising a range of clinical experience. Consecutive consented patients meeting the entrance criteria will undergo standard colonoscopy with the adjunctive use of CDSD. Sensitivity and specificity of CDSD-unaided and CDSD-aided predictions will be compared with the ground truth of histopathology.
Interventions
PROCEDUREColonoscopy
Standard of care colonoscopy, with added use of CDSD as an adjunctive tool
DEVICEClinical Decision Support Device
The investigational device consists of software residing on a dedicated computer enabling display of predicted diminutive polyp histology during colonoscopy.
Sponsors
Olympus Corporation of the Americas
Study design
Allocation
NA
Intervention model
SINGLE_GROUP
Primary purpose
OTHER
Masking
NONE
Masking description
The pathologist (outcomes assessor) is blinded to the endoscopist predictions and CDSD prediction.
Intervention model description
The study shall be a prospective, single arm, multicenter study.
Eligibility
Sex/Gender
ALL
Age
18 Years to No maximum
Healthy volunteers
No
Inclusion criteria
* ≥ 18 years of age * Willing and able to provide informed consent * Subjects undergoing colonoscopy
Exclusion criteria
* Polyposis syndromes including Familial Adenomatous Polyposis Syndrome * Inflammatory Bowel Disease * Hereditary Non Polyposis Colorectal Cancer * Severe coagulopathy * Subjects scoring less than 6 on the Boston Bowel Prep Score or scoring less than 2 in any colon segment. * No diminutive polyps detected during colonoscopy
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Specificity and Sensitivity CDSD-unaided and CDSD-aided Endoscopist Predictions of Diminutive Polyp Histology | Upon receipt of pathology results (within one week) | CDSD-unaided and CDSD-aided predictions will be compared with the ground truth pathology results. |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| 2. Sensitivity, Specificity, Negative Predictive Value (NPV), Positive Predictive Value (PPV) of Unaided Endoscopist Prediction Using White Light. | Upon receipt of pathology results (within one week) | pathology as ground truth |
| 3. Sensitivity, Specificity, NPV, PPV of Unaided Endoscopist Prediction Using NBI Compared With Pathology. | Upon receipt of pathology results (within one week) | pathology as ground truth |
| 1. Number of Unaided NBI Prediction and Pathology Result of Diminutive Polyps Detected by the Endoscopist and Confirmed by Pathology for Which CDSD Does Not Return a Prediction. | Upon receipt of pathology results (within one week) | pathology as ground truth |
| 5. Sensitivity, Specificity, NPV, PPV of CDSD-aided Endoscopist Prediction Compared With Pathology. | Upon receipt of pathology results (within one week) | pathology as ground truth |
| 6. Number of Pathology Results for Diminutive Polyps With Discordant Unaided Predictions Using NBI and Aided Predictions With CDSD | Upon receipt of pathology results (within one week) | pathology as ground truth |
| 4. Sensitivity, Specificity, NPV, PPV of CDSD Prediction Compared With Pathology. | Upon receipt of pathology results (within one week) | pathology as ground truth |
Countries
United States
Participant flow
Participants by arm
| Arm | Count |
|---|---|
| Single-arm CDSD un-aided vs. CDSD-aided predictions of diminutive polyp histology relative to ground-truth pathology. All patients will have both unaided and aided predictions.
Colonoscopy: Standard of care colonoscopy, with added use of CDSD as an adjunctive tool
Clinical Decision Support Device: The investigational device consists of software residing on a dedicated computer enabling display of predicted diminutive polyp histology during colonoscopy. | 1,252 |
| Total | 1,252 |
Withdrawals & dropouts
| Period | Reason | FG000 |
|---|---|---|
| Overall Study | Screen Fails/Withdrawal | 1,089 |
Baseline characteristics
| Characteristic | Single-arm |
|---|---|
| Age, Categorical <=18 years | 0 Participants |
| Age, Categorical >=65 years | 489 Participants |
| Age, Categorical Between 18 and 65 years | 763 Participants |
| Age, Continuous | 60.6 years STANDARD_DEVIATION 10.8 |
| Race/Ethnicity, Customized African American | 180 Participants |
| Race/Ethnicity, Customized Asian | 19 Participants |
| Race/Ethnicity, Customized Caucasian | 976 Participants |
| Race/Ethnicity, Customized Hispanic | 52 Participants |
| Race/Ethnicity, Customized Middle Eastern | 5 Participants |
| Race/Ethnicity, Customized Native American | 4 Participants |
| Race/Ethnicity, Customized Other | 16 Participants |
| Race/Ethnicity, Customized Pacific Islander | 0 Participants |
| Region of Enrollment United States | 1252 Participants |
| Sex: Female, Male Female | 601 Participants |
| Sex: Female, Male Male | 651 Participants |
Adverse events
| Event type | EG000 affected / at risk |
|---|---|
| deaths Total, all-cause mortality | 0 / 0 |
| other Total, other adverse events | 0 / 0 |
| serious Total, serious adverse events | 0 / 0 |
Outcome results
Primary
Specificity and Sensitivity CDSD-unaided and CDSD-aided Endoscopist Predictions of Diminutive Polyp Histology
CDSD-unaided and CDSD-aided predictions will be compared with the ground truth pathology results.
Time frame: Upon receipt of pathology results (within one week)
Population: Total number of polyps from eligible participants (N=1252) who completed the study and had a successful CDSD prediction.
| Arm | Measure | Group | Value (NUMBER) |
|---|---|---|---|
| CDSD Un-aided vs. CDSD-aided Predictions of Diminutive Polyp | Specificity and Sensitivity CDSD-unaided and CDSD-aided Endoscopist Predictions of Diminutive Polyp Histology | Sensitivity for CDSD aided | 90.8 percentage of polyps |
| CDSD Un-aided vs. CDSD-aided Predictions of Diminutive Polyp | Specificity and Sensitivity CDSD-unaided and CDSD-aided Endoscopist Predictions of Diminutive Polyp Histology | Sensitivity for CDSD unaided | 90.7 percentage of polyps |
| CDSD Un-aided vs. CDSD-aided Predictions of Diminutive Polyp | Specificity and Sensitivity CDSD-unaided and CDSD-aided Endoscopist Predictions of Diminutive Polyp Histology | specificity for CDSD aided | 64.7 percentage of polyps |
| CDSD Un-aided vs. CDSD-aided Predictions of Diminutive Polyp | Specificity and Sensitivity CDSD-unaided and CDSD-aided Endoscopist Predictions of Diminutive Polyp Histology | specificity for CDSD unaided | 59.5 percentage of polyps |
Secondary
1. Number of Unaided NBI Prediction and Pathology Result of Diminutive Polyps Detected by the Endoscopist and Confirmed by Pathology for Which CDSD Does Not Return a Prediction.
pathology as ground truth
Time frame: Upon receipt of pathology results (within one week)
Population: Patients that completed the study.
| Arm | Measure | Value (NUMBER) |
|---|---|---|
| CDSD Un-aided vs. CDSD-aided Predictions of Diminutive Polyp | 1. Number of Unaided NBI Prediction and Pathology Result of Diminutive Polyps Detected by the Endoscopist and Confirmed by Pathology for Which CDSD Does Not Return a Prediction. | 112 number of polyps without CDSD prediction |
Secondary
2. Sensitivity, Specificity, Negative Predictive Value (NPV), Positive Predictive Value (PPV) of Unaided Endoscopist Prediction Using White Light.
pathology as ground truth
Time frame: Upon receipt of pathology results (within one week)
Population: patients with polyps
| Arm | Measure | Group | Value (NUMBER) |
|---|---|---|---|
| CDSD Un-aided vs. CDSD-aided Predictions of Diminutive Polyp | 2. Sensitivity, Specificity, Negative Predictive Value (NPV), Positive Predictive Value (PPV) of Unaided Endoscopist Prediction Using White Light. | Sensitivity | 88.5 percentage of polyps |
| CDSD Un-aided vs. CDSD-aided Predictions of Diminutive Polyp | 2. Sensitivity, Specificity, Negative Predictive Value (NPV), Positive Predictive Value (PPV) of Unaided Endoscopist Prediction Using White Light. | Specificity | 61.1 percentage of polyps |
| CDSD Un-aided vs. CDSD-aided Predictions of Diminutive Polyp | 2. Sensitivity, Specificity, Negative Predictive Value (NPV), Positive Predictive Value (PPV) of Unaided Endoscopist Prediction Using White Light. | Negative predictive value (NPV) | 74.7 percentage of polyps |
| CDSD Un-aided vs. CDSD-aided Predictions of Diminutive Polyp | 2. Sensitivity, Specificity, Negative Predictive Value (NPV), Positive Predictive Value (PPV) of Unaided Endoscopist Prediction Using White Light. | Positive predictive value (PPV) | 80.4 percentage of polyps |
Secondary
3. Sensitivity, Specificity, NPV, PPV of Unaided Endoscopist Prediction Using NBI Compared With Pathology.
pathology as ground truth
Time frame: Upon receipt of pathology results (within one week)
Population: Patients with polyps with NBI prediction
| Arm | Measure | Group | Value (NUMBER) |
|---|---|---|---|
| CDSD Un-aided vs. CDSD-aided Predictions of Diminutive Polyp | 3. Sensitivity, Specificity, NPV, PPV of Unaided Endoscopist Prediction Using NBI Compared With Pathology. | Sensitivity | 90.7 percentage of polyps |
| CDSD Un-aided vs. CDSD-aided Predictions of Diminutive Polyp | 3. Sensitivity, Specificity, NPV, PPV of Unaided Endoscopist Prediction Using NBI Compared With Pathology. | Specificity | 59.5 percentage of polyps |
| CDSD Un-aided vs. CDSD-aided Predictions of Diminutive Polyp | 3. Sensitivity, Specificity, NPV, PPV of Unaided Endoscopist Prediction Using NBI Compared With Pathology. | PPV | 74.6 percentage of polyps |
| CDSD Un-aided vs. CDSD-aided Predictions of Diminutive Polyp | 3. Sensitivity, Specificity, NPV, PPV of Unaided Endoscopist Prediction Using NBI Compared With Pathology. | NPV | 83.8 percentage of polyps |
Secondary
4. Sensitivity, Specificity, NPV, PPV of CDSD Prediction Compared With Pathology.
pathology as ground truth
Time frame: Upon receipt of pathology results (within one week)
Population: Patients with CDSD prediction
| Arm | Measure | Group | Value (NUMBER) |
|---|---|---|---|
| CDSD Un-aided vs. CDSD-aided Predictions of Diminutive Polyp | 4. Sensitivity, Specificity, NPV, PPV of CDSD Prediction Compared With Pathology. | Sensitivity | 86.5 percentage of polyps |
| CDSD Un-aided vs. CDSD-aided Predictions of Diminutive Polyp | 4. Sensitivity, Specificity, NPV, PPV of CDSD Prediction Compared With Pathology. | Specificity | 69.5 percentage of polyps |
| CDSD Un-aided vs. CDSD-aided Predictions of Diminutive Polyp | 4. Sensitivity, Specificity, NPV, PPV of CDSD Prediction Compared With Pathology. | PPV | 78.6 percentage of polyps |
| CDSD Un-aided vs. CDSD-aided Predictions of Diminutive Polyp | 4. Sensitivity, Specificity, NPV, PPV of CDSD Prediction Compared With Pathology. | NPV | 79.9 percentage of polyps |
Secondary
5. Sensitivity, Specificity, NPV, PPV of CDSD-aided Endoscopist Prediction Compared With Pathology.
pathology as ground truth
Time frame: Upon receipt of pathology results (within one week)
Population: Aided CDSD prediction
| Arm | Measure | Group | Value (NUMBER) |
|---|---|---|---|
| CDSD Un-aided vs. CDSD-aided Predictions of Diminutive Polyp | 5. Sensitivity, Specificity, NPV, PPV of CDSD-aided Endoscopist Prediction Compared With Pathology. | Sensitivity | 90.8 percentage of polyps |
| CDSD Un-aided vs. CDSD-aided Predictions of Diminutive Polyp | 5. Sensitivity, Specificity, NPV, PPV of CDSD-aided Endoscopist Prediction Compared With Pathology. | Specificity | 64.7 percentage of polyps |
| CDSD Un-aided vs. CDSD-aided Predictions of Diminutive Polyp | 5. Sensitivity, Specificity, NPV, PPV of CDSD-aided Endoscopist Prediction Compared With Pathology. | PPV | 76.8 percentage of polyps |
| CDSD Un-aided vs. CDSD-aided Predictions of Diminutive Polyp | 5. Sensitivity, Specificity, NPV, PPV of CDSD-aided Endoscopist Prediction Compared With Pathology. | NPV | 85.0 percentage of polyps |
Secondary
6. Number of Pathology Results for Diminutive Polyps With Discordant Unaided Predictions Using NBI and Aided Predictions With CDSD
pathology as ground truth
Time frame: Upon receipt of pathology results (within one week)
Population: NBI unaided prediction vs aided CDSD prediction
| Arm | Measure | Value (NUMBER) |
|---|---|---|
| CDSD Un-aided vs. CDSD-aided Predictions of Diminutive Polyp | 6. Number of Pathology Results for Diminutive Polyps With Discordant Unaided Predictions Using NBI and Aided Predictions With CDSD | 251 Polyps |