Pathology
Conditions
Brief summary
The objective of this study is to evaluate precision of the Philips Dx system.
Detailed description
Slides were selected that contained clinically relevant histopathologic features that are generally encountered on surgical pathology slides. Twenty-one features, each selected from three different organs, were to be included to ensure that multiple tissue types were investigated. The study feature(s) as selected on each slide was defined as the selected feature. In total 420 selected features were acquired and the slides holding these features composed the slide set. The slideset consisted of 399 slides from 399 different participants. 1. Intra-system study. The full slide set was then divided over three subsets. Each subset was then scanned three times on one systems, with each subset being scanned on a different scanner. This means that each feature was scanned three times. Each of three pathologists read all three scans of the entire slide set. Intra-system precision was thereby determined. 2. Inter-system study. The full slide set was then scanned three times, each time on a different system, meaning that each feature was scanned three times. Each of three pathologists (different pathologists the ones used in the intra-system study) read all three scans of the entire slide set. Inter-system precision was thereby determined.
Interventions
None listed
Sponsors
Philips Digital & Computational Pathology
Study design
Observational model
CASE_ONLY
Time perspective
CROSS_SECTIONAL
Eligibility
Sex/Gender
ALL
Healthy volunteers
No
Inclusion criteria
* Left-over specimens from subjects who already received their diagnosis and have received their treatment in accordance with the standard of care * H&E glass coverslipped slides with human tissue obtained via surgical pathology * Selected slides fulfill the quality checks according to the Instructions for Use (lfU) * Selected slides must be between 1-5 years since accessioning * Selected slides and FOVs must contain a study feature that is: In it's natural environment (on slide and FOV); Readily observable (on slide and FOV); Not equivocal (on slide and FOV).
Exclusion criteria
* Selected slides contain indelible markings * Selected slides contain damaged tissue * More than one slide was selected for a patient (only one slide may be enrolled per patient).
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Agreement Rate | 2 months | The agreement rate between reads calculated over all selected features and pathologists. Readings were considered in agreement when the selected feature was indicated as 'present' or 'absent' in both readings. |
Countries
United States
Participant flow
Recruitment details
The study started enrollment in February 2016 and enrollment was completed in April 2016. The full set of 420 features were used for the intra-system study as well as for the inter-system sub-study.
Pre-assignment details
Features were evenly distributed over three magnifications (10x, 20x or 40x). 21 different feature types were included. Each feature type was divided over three organs.
Participants by arm
| Arm | Count |
|---|---|
| Precision Intra-system: At one study site the slides with selected features were evenly distributed over three different systems. Each slide was scanned three times on the same system. All scans were read by all three pathologists.
Inter-system: At one study site the full set of slides with selected features was scanned once on three different scanners. All three pathologists read all scans. | 399 |
| Precision Intra-system: At one study site the slides with selected features were evenly distributed over three different systems. Each slide was scanned three times on the same system. All scans were read by all three pathologists.
Inter-system: At one study site the full set of slides with selected features was scanned once on three different scanners. All three pathologists read all scans. | 420 |
| Total | 819 |
Baseline characteristics
| Characteristic | Precision |
|---|---|
| Region of Enrollment United States | 420 Study features |
Adverse events
| Event type | EG000 affected / at risk |
|---|---|
| deaths Total, all-cause mortality | 0 / 399 |
| other Total, other adverse events | 0 / 399 |
| serious Total, serious adverse events | 0 / 399 |
Outcome results
Primary
Agreement Rate
The agreement rate between reads calculated over all selected features and pathologists. Readings were considered in agreement when the selected feature was indicated as 'present' or 'absent' in both readings.
Time frame: 2 months
Population: For the intra-system sub-study 2 cases were not analyzed for possible reading bias due to an error in the EDC system.~For the inter-system sub-study 3 cases were not analyzed for different reasons such as possible reading bias due to an error in the EDC system (2) and a short washout period (1).
| Arm | Measure | Value (MEAN) |
|---|---|---|
| Intra-system Sub-study | Agreement Rate | 92.0 percentage of agreement |
| Inter-system Sub-study | Agreement Rate | 93.8 percentage of agreement |