Copy of ECHoES Resource use and cost questionnaire for optometrists
(PDF download)
Components of a typical monitoring review
TABLE 29
Components of a typical monitoring review and skills required
View in own window
| Component | Description | Skills required |
|---|
| History | Discussion of patient-reported vision status in each eye and comparison with status at previous visit | Communication skills |
| Clinical examination: slit lamp biomicroscopy; anterior segment and macula | Clinical examination to ensure absence of VEGF-related adverse events and/or incidental other disease | Slit lamp and ophthalmoscopy skills |
| Visual acuity assessment | Visual acuity recorded as letters read on an ETDRS chart at 4 m (with/without mirrors) using previously recorded refraction. The results will then be recorded in the patient medical record | Test and interpret visual acuity |
| Administration of 1% tropicamide drops | Pupil dilatation. Drops will need to be administered 20 minutes before CF photography and spectral domain coherence tomography | Instillation of eye drops |
| CF photography (or equivalent CF image) | One good-quality photograph centred on the centre of the macula of each eye | Taking and interpreting retinal images |
| Spectral domain OCT | Cube scan of the posterior pole for each eye. Images will be acquired using a standardised protocol, which is pre-set on the OCT machine | Taking and interpreting OCT images |
| Final assessment | A retreatment decision will be made on the basis of the visual acuity data and interpretation of images obtained. The decision and rationale will need to be entered in the patient record | Ability to assess the need for retreatment and arrange necessary follow-up |
Resource use and costs associated with training
The cost of each of the three training activities was calculated by multiplying the time spent on each activity by the unit cost of optometrist time. The average cost per hour of optometrist time was £62.13 (SD £34.62) and its calculation was based on participants’ reports on salary and hours worked (through the health economics questionnaire). Given that our objective was to estimate the cost of optometrist training per monitoring review, each of the three components of costs in the table below was divided by the annual number of patients (after the changes in the practice would take place, as reported by each optometrist in reply to question Q7 in the health economics questionnaire), to obtain the cost of optometrist training per monitoring review, that is £0.89 (SD £1.080).
TABLE 30
Resource use and costs of training
View in own window
| Training type | Optometrist’s time (minutes), mean (SD) | Optometrist’s cost (£), mean (SD) |
|---|
| Attending webinars | 120 (0) | 124 (69.26) |
| Revisiting webinars | 90 (68.74) | 96 (111.58) |
| Consulting other resources | 64 (76.21) | 66 (112.35) |
| Observations | 48 | 48 |
Sensitivity analysis 1: three ranibizumab injections and consultations instead of one
TABLE 31
Care cost pathways for sensitivity analysis 1
View in own window
| Lesion status assessment | Observationsa (%) | Pathways cost (£),b mean (SD) |
|---|
| Experts | Optometrists | | |
|---|
| Reactivated | Reactivated | 795 (39.43) | 2548.83 (67.90) |
| Reactivated | Suspicious | 142 (7.04) | 103.61 (18.51) |
| Reactivated | Quiescent | 57 (2.83) | 51.29 (9.08) |
| Suspicious | Reactivated | 10 (0.50) | 118.12 (16.39) |
| Suspicious | Suspicious | 11 (0.55) | 57.04 (9.10) |
| Suspicious | Quiescent | 14 (0.69) | 52.96 (9.37) |
| Quiescent | Reactivated | 105 (5.21) | 117.14 (32.61) |
| Quiescent | Suspicious | 234 (11.61) | 78.31 (11.53) |
| Quiescent | Quiescent | 648 (32.14) | 51.98 (8.23) |
| Experts | Ophthalmologists | | |
|---|
| Reactivated | Reactivated | 736 (36.51) | 2495.81 (70.01) |
| Reactivated | Suspicious | 196 (9.72) | 153.18 (92.25) |
| Reactivated | Quiescent | 62 (3.08) | 77.01 (45.49) |
| Suspicious | Reactivated | 1 (0.05) | 2452.74 (n/a) |
| Suspicious | Suspicious | 17 (0.84) | 68.84 (31.004) |
| Suspicious | Quiescent | 17 (0.84) | 60.57 (17.16) |
| Quiescent | Reactivated | 35 (1.73) | 2493.45 (65.87) |
| Quiescent | Suspicious | 146 (7.24) | 150.34 (95.19) |
| Quiescent | Quiescent | 806 (39.98) | 75.28 (44.72) |
- a
The number of observations (i.e. vignettes) is 4032, namely 2016 referring to optometrists and 2016 referring to ophthalmologists.
- b
Pathway cost includes the cost of a monitoring consultation and downstream costs (e.g. injections and follow-up visits).
TABLE 32
Impact on cost-effectiveness of sensitivity analysis 1
View in own window
| Costs and effects | Optometrists (observations, n = 2016) | Ophthalmologists (observations, n = 2016) |
|---|
| Cost of a monitoring review (£), mean (SD) | 1047.03 (1213.05) | 1015.01 (1168.80) |
| Proportion of correct assessments (SD) | 0.844 (0.363) | 0.854 (0.353) |
| Incremental cost (£) (95% CI) | 32.02 (–60.87032 to 124.9) |
| Incremental benefit, proportion of correct assessments (95% CI) | –0.0099 (–0.045 to 0.025) |
| ICER, incremental cost per correct assessmenta | Dominated |
- a
The 95% CI around the ICER could not be defined.
Sensitivity analysis 2: one aflibercept injection instead of one ranibizumab
TABLE 33
Care cost pathways for sensitivity analysis 2
View in own window
| Lesion status assessment | Observationsa (%) | Pathways cost (£),b mean (SD) |
|---|
| Experts | Optometrists | | |
|---|
| Reactivated | Reactivated | 795 (39.43) | 1009.24 (45.50) |
| Reactivated | Suspicious | 142 (7.04) | 103.61 (18.51) |
| Reactivated | Quiescent | 57 (2.83) | 51.29 (9.08) |
| Suspicious | Reactivated | 10 (0.50) | 118.12 (16.39) |
| Suspicious | Suspicious | 11 (0.55) | 57.04 (9.10) |
| Suspicious | Quiescent | 14 (0.69) | 52.96 (9.37) |
| Quiescent | Reactivated | 105 (5.21) | 117.14 (32.61) |
| Quiescent | Suspicious | 234 (11.61) | 78.31 (11.53) |
| Quiescent | Quiescent | 648 (32.14) | 51.98 (8.23) |
| Experts | Ophthalmologists | | |
|---|
| Reactivated | Reactivated | 736 (36.51) | 956.50 (46.41) |
| Reactivated | Suspicious | 196 (9.72) | 153.18 (92.25) |
| Reactivated | Quiescent | 62 (3.08) | 77.01 (45.49) |
| Suspicious | Reactivated | 1 (0.05) | 2452.74 (n/a) |
| Suspicious | Suspicious | 17 (0.84) | 68.84 (31.004) |
| Suspicious | Quiescent | 17 (0.84) | 60.57 (17.16) |
| Quiescent | Reactivated | 35 (1.73) | 2493.45 (65.87) |
| Quiescent | Suspicious | 146 (7.24) | 150.34 (95.19) |
| Quiescent | Quiescent | 806 (39.98) | 75.28 (44.72) |
- a
The number of observations (i.e. vignettes) is 4032, namely 2016 referring to optometrists and 2016 referring to ophthalmologists.
- b
Pathway cost includes the cost of a monitoring consultation and downstream costs (e.g. injections and follow-up visits).
TABLE 34
Impact on cost-effectiveness of sensitivity analysis 2
View in own window
| Costs and effects | Optometrists (n = 2016) | Ophthalmologists (n = 2016) |
|---|
| Cost of a monitoring review (£), mean SD | 439.90 (460.90) | 425.61 (422.93) |
| Proportion of correct assessments (SD) | 0.844 (0.363) | 0.854 (0.353) |
| Incremental cost (£) (95% CI) | 14.29 (–19.91 to 48.49) |
| Incremental benefit, proportion of correct assessments (95% CI) | –0.0099 (–0.045 to 0.025) |
| ICER, incremental cost per correct assessmenta | Dominated |
- a
The 95% CI around the ICER could not be defined.
Sensitivity analysis 3: one bevacizumab injection instead of one ranibizumab
TABLE 35
Care cost pathways for sensitivity analysis 3
View in own window
| Lesion status assessment | Observationsa (%) | Pathways cost (£),b mean (SD) |
|---|
| Experts | Optometrists | | |
|---|
| Reactivated | Reactivated | 795 (39.43) | 242.23 (45.50) |
| Reactivated | Suspicious | 142 (7.04) | 103.61 (18.51) |
| Reactivated | Quiescent | 57 (2.83) | 51.29 (9.08) |
| Suspicious | Reactivated | 10 (0.50) | 118.12 (16.39) |
| Suspicious | Suspicious | 11 (0.55) | 57.04 (9.10) |
| Suspicious | Quiescent | 14 (0.69) | 52.96 (9.37) |
| Quiescent | Reactivated | 105 (5.21) | 117.14 (32.61) |
| Quiescent | Suspicious | 234 (11.61) | 78.31 (11.53) |
| Quiescent | Quiescent | 648 (32.14) | 51.98 (8.23) |
| Experts | Ophthalmologists | | |
|---|
| Reactivated | Reactivated | 736 (36.51) | 189.50 (46.41) |
| Reactivated | Suspicious | 196 (9.72) | 153.18 (92.25) |
| Reactivated | Quiescent | 62 (3.08) | 77.01 (45.49) |
| Suspicious | Reactivated | 1 (0.05) | 184.21 (n/a) |
| Suspicious | Suspicious | 17 (0.84) | 68.84 (31.004) |
| Suspicious | Quiescent | 17 (0.84) | 60.57 (17.16) |
| Quiescent | Reactivated | 35 (1.73) | 189.12 (38.002) |
| Quiescent | Suspicious | 146 (7.24) | 150.34 (95.19) |
| Quiescent | Quiescent | 806 (39.98) | 75.28 (44.72) |
- a
The number of observations (i.e. vignettes) is 4032, namely 2016 referring to optometrists and 2016 referring to ophthalmologists.
- b
Pathway cost includes the cost of a monitoring consultation and downstream costs (e.g. injections and follow-up visits).
TABLE 36
Impact on cost-effectiveness of sensitivity analysis 3
View in own window
| Costs and effects | Optometrists | Ophthalmologists |
|---|
| Cost of a monitoring review (£), mean (SD) | 137.43 (91.78) | 131.89 (77.12) |
| Proportion of correct assessments (SD) | 0.844 (0.363) | 0.854 (0.353) |
| Incremental cost (£) (95% CI) | 5.54 (–0.834 to 11.916) |
| Incremental benefit, proportion of correct assessments (95% CI) | –0.0099 (–0.045 to 0.025) |
| ICER, incremental cost per correct assessmenta | Dominated |
- a
The 95% CI around the ICER could not be defined.
Sensitivity analysis 4: only monitoring review cost, no pathway cost
TABLE 37
Care cost pathways for sensitivity analysis 4
View in own window
| Lesion status assessment | Observationsa (%) | Pathways cost (£),b mean (SD) |
|---|
| Experts | Optometrists | | |
|---|
| Reactivated | Reactivated | 795 (39.43) | 51.79 (8.49) |
| Reactivated | Suspicious | 142 (7.04) | 51.81 (9.26) |
| Reactivated | Quiescent | 57 (2.83) | 51.29 (9.08) |
| Suspicious | Reactivated | 10 (0.50) | 50.64 (7.12) |
| Suspicious | Suspicious | 11 (0.55) | 57.04 (9.10) |
| Suspicious | Quiescent | 14 (0.69) | 52.96 (9.37) |
| Quiescent | Reactivated | 105 (5.21) | 51.49 (8.01) |
| Quiescent | Suspicious | 234 (11.61) | 52.21 (7.69) |
| Quiescent | Quiescent | 648 (32.14) | 51.98 (8.23) |
| Experts | Ophthalmologists | | |
|---|
| Reactivated | Reactivated | 736 (36.51) | 76.09 (43.66) |
| Reactivated | Suspicious | 196 (9.72) | 76.59 (46.13) |
| Reactivated | Quiescent | 62 (3.08) | 77.01 (45.49) |
| Suspicious | Reactivated | 1 (0.05) | 89.70 (n/a) |
| Suspicious | Suspicious | 17 (0.84) | 68.84 (31.004) |
| Suspicious | Quiescent | 17 (0.84) | 60.57 (17.16) |
| Quiescent | Reactivated | 35 (1.73) | 76.71 (38.22) |
| Quiescent | Suspicious | 146 (7.24) | 75.17 (47.59) |
| Quiescent | Quiescent | 806 (39.98) | 75.28 (44.72) |
- a
The number of observations (i.e. vignettes) is 4032, namely 2016 referring to optometrists and 2016 referring to ophthalmologists.
- b
Pathway cost includes the cost of a monitoring consultation and downstream costs (e.g. injections and follow-up visits).
TABLE 38
Impact on cost-effectiveness of sensitivity analysis 4
View in own window
| Costs and effects | Optometrists (observations, n = 2016) | Ophthalmologists (observations, n = 2016) |
|---|
| Cost of a monitoring review (£), mean (SD) | 51.90 (8.36) | 75.60 (44.31) |
| Proportion of correct assessments (SD) | 0.844 (0.363) | 0.854 (0.353) |
| Incremental cost (£) (95% CI) | –23.70 (–26.09 to –21.31) |
| Incremental benefit, proportion of correct assessments (95% CI) | –0.0099 (–0.045 to 0.025) |
| ICER, incremental cost per correct assessment (95% CI) (£) Fieller’s methoda | 2389.07; lower limit: 535, upper limit: –943b |
- a
We can be 95% confident that the way of performing a monitoring review with the larger point estimate for effect (i.e. monitoring review performed by ophthalmologists) represents bad value compared with the alternative.
- b
The lower confidence limit for the ICER is positive, and the upper limit is negative, because the ICER point estimate (see ) is in the south-west quadrant of the cost-effectiveness plane, hence positive, but the CI spans both the south-west (positive ICER) and south-east quadrants (negative ICER). The upper confidence limit of –943 lies in the south-east quadrant, which is represents a win-win situation (monitoring by optometrists is less costly and more effective). In contrast to the previous analyses, this sensitivity analysis shows a clear difference in costs and uncertainty in the effects.
