Quality Control Procedures and Baseline Values for Electroretinography, Perimetry, Color Vision, and Visual Acuity in an International Multicenter Study: Observations from a Safety Trial in Chronic Stable Angina Pectoris

Purpose To describe quality control procedures and baseline values of electroretinography (ERG), kinetic and static perimetry, color discrimination, and best-corrected visual acuity from a multicenter ocular safety study. Methods A multicenter prospective longitudinal randomized placebo-controlled study was conducted at 11 ophthalmic centers that had received certification following training, instruction, and monitoring. ERGs were obtained with the Espion E2 Ganzfeld console, perimetry with the Octopus 101 perimeter, color discrimination with the Lanthony desaturated D15 test, and best-corrected visual acuity with the Early Treatment Diabetic Retinopathy Study chart. Ophthalmic eligibility required satisfactory outcomes for ERG and perimetry by the second or third pre-inclusion attempts, respectively. Quality control for the ERG was undertaken by two central readers. Results The mean (SD) age of the 97 individuals was 63.5 (7.9) range, 44–83 years. The overall coefficients of variation (CVs) for the ERG peak times were less than those of the only comparable single-center study. The CV for the mean defect of standard automated perimetry was approximately one-third that of the Ocular Hypertension Treatment Study. With increasing age, ERG peak times and color discrimination Total Error Score increased while ERG amplitudes and isopter area all decreased. Conclusions The data illustrate the benefit of identical equipment, stringent on-site instruction and training, quality control, certification, and validation methods. The latter are recommended for planning and conducting multicenter trials using ERG and perimetry to monitor safety and/or efficacy of treatment intervention. Translational Relevance Stringent quality control procedures and reliable reference values are indispensable prerequisites for informative clinical trials.


Main objectives of the validation process:
The main objectives of the validation process are to validate each centre on the quality / reliability of the recordings and to confirm the feasibility of the electronic transmission from the investigator's centre to the central data base .

Performance of the quality review:
The quality assessment of the visual field examinations will require the following steps:

At the investigator's site:
The ophthalmological investigators will : -select / enrol volunteers ( no age range pre defined) after having checked that they have not participated in repeated visual field studies during the three years preceding this procedure ; therefore, the quality of the instruction given by the centre to the volunteer being investigated will be assessed , in addition any potential learning effect will be avoided. -perform the VF measurements (static and kinetic) in three selected healthy volunteers according to the procedure described in the study protocol ( see supplement 3) -transmit electronically to the central data base the visual fields ( graphs / plots/ calculated values) of 3 healthy volunteers ( the visual fields from the 3 should be sent together) .

Quality Control (QC) validation procedure of perimetric examination
-The visual electronic version of the visual field records will be reviewed and assessed by the independent quality control board of specialists in perimetry, members of the Scientific Ophthalmic Safety Committee. -The assessment for validation will be done on centers identification being coded ( blind assessment of the center) within a maximum of 15 days upon receipt ( at the central data base).

Annex 3 Visual field assessment with the Octopus 101 perimeter
Static and kinetic perimetry will be undertaken on each eye using the Octopus 101 perimeter, a standard commercially available instrument used in ophthalmology.
Static threshold perimetry will be undertaken within the central field (i.e. within a radius of 30° from the fixation point). The threshold response (i.e. the minimum brightness of a spot of light [the stimulus] necessary to evoke a response within the visual system) will be determined using Goldmann stimulus size III (subtending 0.431° at the eye) at each of 64 locations within the central field. The subject is required to indicate, by pressing a response button, whenever he/she perceives the spot of light.
Semi-automated kinetic perimetry (SKP) of the peripheral visual field (i.e. the visual field beyond an eccentricity of 30°) will be undertaken by separately presenting two moving spots of bright white light from the periphery towards the centre (the stimulus vector) at 15 angular intervals (i.e. meridian) at an angular velocity of 3°/s. The size of the two spots of light will be Goldmann III and Goldmann I (subtending an angle of 0.108at the eye) and the brightness 1000asb and 100asb, respectively. The subject is again required to indicate, by pressing a response button, whenever he/she perceives the spot of light. The line joining the location along each meridian at which the given stimulus is detected is termed the isopter.
The spots of light for both static and kinetic perimetry are presented against a white background of 10 cd/m 2 .

Technical operation
The patient will be instructed in the required procedures and will undergo a short practice session prior to the start of the formal visual field examination. The distance refraction, corrected for a distance of 40 cm (i.e. the radius of the cupola of the Octopus 101 perimeter) will be worn in the form of thin-rimmed lenses positioned close to the eye in order to avoid shading parts of the visual field. The patient's lashes should not touch the lens. Each eye will be tested separately by covering the fellow eye with an opaque occluder. The subject´s pupils should not have been dilated prior to the visual field examination. The patient is positioned with the forehead touching the head rest and the chin in firm contact with the chin rest in such a way that the eye under examination is gazing at the fixation target (four green bars forming a fixation cross around the geometrical centre of the cupola). The noise level within, and surrounding, the examination room should be sufficiently quiet to enable the subject to concentrate on the task. The examination should ideally be undertaken in darkness although some minimal local ambient illumination is acceptable. Any change in the general lightening conditions within the room or the introduction of other distracters during the examination (e.g the telephone ringing, etc) should be avoided. The subject holds a response button in his/her hand which is pressed to indicate whenever he/she sees the (static or kinetic) stimulus. It is also necessary to ensure that the subject's sitting position is comfortable and relaxed, and that the chair is adjusted accordingly, so that the subject can easily tolerate the examination.
Additionally, a small camera within the perimeter permanently records the position and movement of the subject's pupil. The technician has to ensure that the head is adjusted so that the pupil is aligned in the centre of the monitor.
The instructions to the patient for static perimetry of the central field will be modified from those used for the Ocular Hypertension Treatment Study [OHTS]), namely: "Always look straight ahead at the centre between the four green bars (the fixation target). Other lights will flash one at a time at other positions around the centre of the bowl. Some lights will be bright, and others will be dim. Press the button whenever you see one of these flashes. You are not expected to see all of them. The best time to blink is just as you press the button." The instructions for static perimetry will be adapted for kinetic perimetry of the peripheral visual field. The patient will be reminded again to look steadily at the fixation target and press the response button whenever they become aware of a light moving into the edges (the periphery) of their vision.

Preparatory test procedure
Step 1 Kinetic perimetry will be performed with stimulus III4e, presented at an angular velocity of 3°/s along each of 16 meridians of the visual field (i.e. at 15° intervals). The order of the meridian examined is selected randomly by placing the origin of the vectors directed from periphery towards the centre, 10° outside of the age-corrected normal reference isopter. The direction of the vectors should be chosen in a way that the presumed visual field border/isopter is crossed by the vector almost perpendicularly. If there is a suspicion of an abnormal visual field (e.g. diabetic retinopathy or hemifield defect, e.g. after stroke) the patient may not see the vector movement in those regions affected by the disease. In this case, the technician will move the vector further in towards the centre of the field until the subject recognizes the stimulus. For quality assessment reasons, reaction time is assessed separately; for this purpose three "reaction time vectors" are placed within the central 30° radius of the intact visual field. The software will automatically correct the isopters for the individually measured reaction time.
Step 2 The smaller and less bright I3e stimulus will be presented at an angular velocity of 3°/s in the same manner as that described for the III4e stimulus, described above.

Figure 1 The reference intervals for the normal III4e (pink) and I3e (orange) isopters recorded with the Octopus 101 perimeter
Step 3: Blind spot assessment At least 8 vectors (i.e. 30° apart and randomly selected) originating from the assumed centre of the blind spot and moving at an angular velocity of 2°/s are presented using the I4e stimulus.
The area of the blind spot is automatically calculated and the areal measure serves as a quality criterion of the patient's cooperation and attention. Magnification software is used to assess the blind spot more accurately.
Step 4, 5, and 6: The same procedure will be applied to the fellow eye after a rest period of 5 minutes.
Subsequently the area encircled by each isopter will be calculated.
Step 7: Automated static threshold perimetry After a 5-10 min. rest period, automated static perimetry will be undertaken. The subjects will be instructed in the new task, namely that the spot may appear in different locations with different brightness levels and that he/she has to press the button each time he/she perceives the spot, even if it is very dim, and that not all (only approximately 50% of the) presentations are expected to be perceived. A modified Program G1/G2 grid (incorporating one additional central test location, one test location within the centre of the blind spot and four additional test locations within the immediate vicinity of the blind spot) will be used, as show below: Altogether 64 locations are examined within the 30° central field, as illustrated in the table.  At the initial visit, static perimetry will be undertaken using a 4-2-1(1) dB staircase threshold assessment strategy: the initial stimulus is presented at a luminance level of 4 dB above or below the assumed threshold at the given location. The stimulus luminance is then increased / decreased in 4 dB steps until the stimulus is no longer perceived / perceived for the first time (→ reversal). Thereafter, the procedure is repeated in 2 dB steps until the next reversal is achieved and repeated again until a final reversal with a 1 dB step size is achieved. For each subsequent visit, the local threshold value obtained at the previous examination for the given patient is used as the local initial value for the starting point of the staircase in order to reduce the examination duration.
The local differential luminance sensitivity is estimated using a maximum likelihood (ML) method.
The thresholding procedure is carried out for each of the 64 locations in random order. It should be noted that the three staircase/bracketing procedures are not undertaken for each location in a single sequence. Instead, the computer stores the values of a given reversal at a given location. Specifically, it records the last value of the stimulus and the patient's response and selects the next location in random order. This is done in order to avoid local adaptation and loss of attention in regard to any given single test location. Thereby, each threshold sequence is put together by multiple events at each location with these events being presented in the correct order of brightness steps at each different location and at different time points. If the patient encounters a problem in response at any location, the thresholding sequence can be re-checked for the particular location(s).
Subsequently, the results are stored in an electronic database and also printed out on paper.
The results include the number of incorrect reponses to the false-negative and to the falsepositive catch trials, the magnitude of the mean deviation (MD) and other visual field indices as well as a display of the continuous video-pupillographic fixation control. Additionally, the printout contains visual acuity, refractive correction used, the level of the cupola luminance, a display of all vectors utilised for SKP, the mean reaction time for SKP, and the area of each of the two isopters and of the blind spot, corrected for the individual reaction time of the patient.

Calibration
The perimeter has an inbuilt self-calibration procedure that is activated automatically when the instrument is switched on. This procedure checks the luminance of the test and background stimuli. An error message is displayed if the calibration fails.
All examination data are stored together with the patient ID data on a computer file, printed out and, in parallel, electronically submitted to the central server.

Annex 4 ERG study protocol Standard Ganzfeld electroretinogram (ERG)
Check all patient details and record appropriate data. Explain the procedures in full, and answer any questions the patient may have. Fundus photography should be avoided to be done within the hour preceding the start of the ERG procedure. Explain the effects and consequences of the eye-drops that are being used. The ERG will be recorded according to the ISCEV Standard for Full-field Clinical Electroretinography (see https://iscev.wildapricot.org/standards).
The following document serves to specify those parts where the standard recommendation (in the present paper 2004 update) allows several alternatives and to describe the exact procedure and stimulus parameters to be followed by any investigating centre.
The patient should have dilated pupils for all ERG recordings. Tropicamide is preferably recommended to be used . The electrodes are best fitted immediately after the installation of the dilating drops and prior to dark adaptation. DTL-Fibre electrodes are used for recording corneal responses. The use of topical anaesthesia is not necessary in case of the use of fibre electrodes. The reference electrodes should be placed on the zygomatic fossae. The ground electrode is placed on the central forehead. Impedances should be measured and be approximately the same for all three electrodes and ≤ 5kΩ.
The patient should then be positioned on the Ganzfeld chin rest, and the height of the chin rest, stool and table adjusted for maximum patient comfort and to align the subject's eyes with the central fixation spot and monitoring camera if fitted. DTL-Fibres electrodes will be used and provided to the centres by the Sponsor to ensure the homogeneity among the centres and the reliability among centres. The position of the transcorneal fibre in relation to the pupil should be recorded. The investigator must ensure that the fibre electrodes are always at or just above the rim of the lower lid. Pupils size should be recorded at the end of the session with a comparison ruler.
Dark adapted (DA) ERG: Dark adapt the patient for a full 20 minutes. This is best achieved in the test room and should be in total darkness. All light within the patient's section of the room should be extinguished, including any red 'safe' light. The patient should be attended at all times, preferably by the recording technician as well as any accompanying person. Curtains should be used to screen the subject from the instrument lights and screens. If it is unavoidable that the patient and operator cannot be screened off from each other, screen illumination should be reduced to the minimum required for operating comfortably once the operator has dark adapted. If it is necessary for the operator to leave the room during dark adaptation, then the patient should be asked to close their eyes whilst this happens, and also whilst the operator re-enters the room. Preferably a black curtain should shied the door. The outside illumination should also be reduced to allow minimum light contamination during re-entry.
Fit the fibre electrodes under the lowest possible level of red illumination, (use the minimum working level for the red light that is acceptable to the operator, best achieved where the operator has dark adapted with the patient), check and adjust if necessary the position of the fibres relative to the pupils under red light. Record whether change of electrode position was required. Position the patient on the chin rest in the previously adjusted Ganzfeld bowl. The patient should be instructed to look towards the fixation lamp, which is used on its lowest intensity setting. Patient fixation is monitored, if possible, using the infra red (IR) camera system within the bowl. Dark adapt the patient for a further 5 minutes.
The recordings detailed below should be recorded in all patients. Any variation in protocol, and the reasons for doing so, must be documented in the patients file. Dark adapted recordings should be obtained first with the lowest stimulus intensity, then successively with increasing intensity. Each intensity level should be replicated and recorded twice, and if good reproducibility is not present, further repeat(s) should be performed. Allow an inter stimulus interval (ISI) of 5 seconds between flashes at the low-intensity; of 10 seconds at the intermediate and higher intensities, and of 20 seconds at the upper levels of intensity. When the DA ERG responses have been satisfactorily recorded, the patient should be restored to photopic conditions using the background lighting of the Ganzfeld set to approx 34 cd/m2.

Light adapted (LA) ERG:
The patient should be reminded to keep his eyes fully open during light adaptation. The LA photopic recordings should be performed after 10 minutes light adaptation, again with the necessary number of replications. Each averaged response should be recorded twice. The background light should remain on at all times during photopic recording. When recording is complete, measure and record pupil diameter. Remove all electrodes and remind the patient of the effects of mydriasis routine safety precautions measures should be carefully followed.

Notes on recording the ERG
It should not be necessary to average the DA ERG data, but this may be needed if the waveform is poor or contaminated by artefact. Average enough to "clean up" the waveform form ERG, from artefacts but remember to maintain the allowed inter-stimulus intervals (ISI) at all times. Prior to every flash, instruct the patient to fixate, sit still and try not to blink or move their mouth until after the light has flashed, and open their eyes (comfortably) wide. Monitor patient compliance on the infrared camera display if possible. The patient should be told to keep their jaw and face relaxed, and to avoid clenching their teeth. Their eyes should always be open and looking at the fixation light in the Ganzfeld bowl.
Averaging is usually beneficial for good light adapted recordings. Start the 30Hz flicker stimulus and allow to run for 1-2 seconds before commencing averaging. 10-30 averages per trial should be adequate if the patient can maintain eye opening without excessive EMG (electromyographic, muscle artefact). For the "single flash" photopic ERG ("LA 3.0"), again start the flashing prior to commencing averaging. Flash at 1Hz; 5-10 averages per trial should be ample. Use a time-base of 200ms for DA ERGs and 100ms for LA ERGs. Filters should be 0.3 -1000. Any technical difficulties, such as an inability to maintain full pupil exposure during photopic flicker ERG recording should be recorded.
Create back-up of ERG data on CD at the end of each recording session.

Required stimulus parameters
The white standard flash (SF) should be set at 3 cd.s/m2 with 4 ms stimulus.

a) Dark adapted ERG: For each response the stimulus intensity will be as follows:
-

II-ERG REVIEW PRIOR TO THE START OF THE STUDY (i.e. validation procedure)
Before the start of the study, three ERG tests will be performed by each centre in order to validate each centre on the quality of the recordings and on the feasibility of ERG procedure including the electronic transmission to the central data base. The three ERGs should be sent together.
The data will be fully anonymised and the evaluation will be performed blind treatment, patient identification and to the site.
The quality validation will be done by the central reviewer within a maximum of 15 days upon receipt. A copy of the validation certificate will be transmitted to each investigator, to the main investigator of the corresponding centre, and to the sponsor.
The centre cannot start to include patients until the centre has not been approved by the central reader. Therefore, it is only after validation by the central reader that a centre will be formally admitted to the study and become able to take data from patients.

III-ERG PERFORMANCE BY THE INVESTIGATORS DURING THE STUDY
ERG will be performed by the investigator according to the "ERG procedure" described in Supplement 1.
ERG will be digitally recorded using the ESPION E2 equipment and DTL -fibre electrodes.
Before the start of an ERG recording, a "patient's file" will be created by the qualified technician(s) pre-selected in the centre. This file will contain the patient ID number, his/her date of birth (d.o.b.), and gender (these information are mandatory as the clinical assessment is done in comparison with the expected values of an age matched population).
The ERG will be performed by a qualified technician. The amplitude (v) and peak time (ms) of each component (a-and b-waves, flicker response) will be measured by cursor positioning. The cursor positioning will be marked on the waveforms. Prior to the start of the recording, the technician will ask the patient his/her date of his/her birthday in order to check that the reported personal data are correct. At the end of the ERG recording procedure, the technician will hand over to the investigator (OPH/electrophysiologist) the waveforms for review/validation prior to the sending to the central data base (the technician does not have the "right" to send the waveforms to the central data base). The investigator will review the ERG performed by the technician, eventually will reposition the cursor(s) (i.e. re-measurement) and make the clinical assessment of the ERG. The so called "reviewed ERG" will be printed-out (source document to be included into the patient's medical file) and the completed electronic document of the ERG will be saved on the Espion E2, then transmitted to the central data base. In addition a back-up (CD) of all electronic waveforms will be performed.
The ERG tests should be sent using an electronic transmission to the Central Data Base immediately after their performance (i.e. within the day of recording). Within 5 days upon receipt at the central reader's office the qualitative and quantitative assessments of the central reader will be received by the ophthalmologist (investigator).

IV-1: ERG sequences description
The central reviewer will have a direct access to the central data base up loaded with the ERGs performed/validated by the investigators.
For each ERG recording (i.e. dark/light adapted) a qualitative and quantitative analysis will be performed by the central reader (measurement of amplitude and implicit times of each component). This review has to be done within 5 days upon receipt.
The baseline ERG must be of high quality as all other waveforms will be compared to this reference one (so called baseline). Therefore, if the central reviewer is: satisfied with the quality of the ERG performed at the selection visit (ASS1), this one will become "the baseline ERG". This "validation" will be transmitted to the computerised centre for randomisation and treatment allocation (CCRTA).
not satisfied with the quality of the ERG performed at the selection visit, an additional ERG will be requested by the central reviewer to be done by the investigator (OPH) at the preinclusion visit within a maximum of 7 days. Then the same quality process will take place: ➢ if the quality is still insufficient the patient will not be eligible, ➢ if the quality is good, this ERG performed at the pre-inclusion visitwill be the baseline ERG; its validation will be transmitted to the CCRTA and the patient will be potentially eligible as regards to this criteria.

IV-2: ERG central review procedure
The central reader will assess the ERG quality and will be responsible for its analysis and measurements. This will be done for any ERG recorded (i.e. all visits performed).
The quality and quantitative evaluation of the central reader will be sent to the investigator within 5 days working upon receipt.
An overall medical review will then be done for each completed patient.

IV-2-1: Quality control
The quality control will be done on:

IV-2-2: ERG measurements
The central reader will assess the investigator's reported measured components of the different responses and may or may not re-position the cursor(s) according to his evaluation. All data after having been assessed will be either validated (measured value agreed) or modified (measured value not agreed upon i.e. a re-positioning of the cursor being done). These values will be entered directly into the "Central Reader Database". The main analysis of the ERGs will be performed with the data from the central data base.