Evidence Table 22Service Provision

Study detailsPatientsObserver GroupsOutcome MeasuresEffect SizeComments
Azuara-Blanco et al., 20076

Study design:
Prospective observational

Observer masked
Patient group: 671 referrals from community optometrists in Grampian, Scotland.

Inclusion criteria:
>18 years

All patients
N: 100 (165 randomised, 65 chose not to participate)
Age (mean): 67
M/F: 52/48
Mean IOP (mmHg): 26
Family history: 24
Black: 1
Glaucoma diagnosis (management decisions **) by consultant
  1. Normal & discharged: 35
  2. Suspect or OHT requiring review: 32
  3. Suspect or OHT requiring treatment: 8
  4. Glaucoma requiring urgent treatment: 2
Group 1:
3 community optometrists (CO) that had received in-house training by a consultant ophthalmologist and glaucoma specialist as part of glaucoma optometric service. Training included practical sessions, glaucoma clinics, teaching on diagnostic interventions

Group 2:
Junior (trainee) ophthalmologist

Group 3:
Consultant ophthalmologist

Examination methods:
Each CO examined all 671 referrals for:
  • Visual acuity (Snellen chart)
  • VF (threshold strategy 24-2 SITA)
  • Corneal thickness (ultrasound pachymetry)
  • Slit lamp biomicroscopy to assess anterior segment and optic disc
  • Goldmann tonometry
  • Refraction
  • Risk factors
The junior doctor and consultant ophthalmologist examined the 100 patients randomised into the study in the hospital out patient department with same tests except for IOP measurements
Inter-observer (consultant- optometrist) agreement for all management decisions (1–5)**
weighted kappa statistic κw
Mean (95%CI) κw = 0.53 (0.39 – 0.67) (moderate) 95% CI calculated by NCC-AC using SE 0.07 from studyFunding:
Scottish Executive Health Department

Limitations:
The method of weighting of the kappa statistic was not clearly defined and the kappa value agreement scale was not mentioned. It was assumed to be from (Landis and Koch 1977)

Additional Outcomes:

Notes:
The community optometrists were masked to randomised patient selection. Participants were required not to disclose details of previous consultations.
Inter-observer (junior doctor– consultant) agreement for all management decisions (1–5)**
weighted kappa statistic κw
Mean (95%CI) κw = 0.45 (0.31 – 0.59) (moderate) 95% CI calculated by NCC-AC using SE 0.07 from study
Inter-observer (junior doctor– optometrist) agreement for all management decisions (1–5)**
weighted kappa statistic κ w
Mean (95%CI) κ w = 0.45 (0.31 – 0.59) (moderate) 95% CI calculated by NCC-AC using SE 0.07 from study
Inter-observer (consultant- optometrist) agreement for diagnosis of glaucoma (4–5 v 1–3)**
weighted kappa statistic κw
Mean (95%CI) κw = 0.70 (0.54 – 0.87) (substantial) 95% CI calculated by NCC-AC using SE 0.083 from study
Inter-observer (junior doctor– consultant) agreement for diagnosis of glaucoma (4–5 v 1–3)**
weighted kappa statistic κw
Mean (95%CI) κ w = 0.54 (0.35 – 0.73) (moderate) 95% CI calculated by NCC-AC using SE 0.098 from study
Inter-observer (junior doctor– optometrist) agreement for diagnosis of glaucoma (4–5 v 1–3)**
weighted kappa statistic κw
Mean (95%CI) κ w = 0.22 (0.02 – 0.42) (fair) 95% CI calculated by NCC-AC using SE 0.101 from study
Inter-observer (consultant- optometrist) agreement for treatment required (3–5 v 1–2)**
weighted kappa statistic κw
Mean (95%CI) κw = 0.72 (0.57 – 0.86) (substantial) 95% CI calculated by NCC-AC using SE 0.076 from study
Inter-observer (junior doctor–consultant) agreement for treatment required (3–5 v 1–2)**
weighted kappa statistic κw
Mean (95%CI) κw = 0.55 (0.37 – 0.73) (moderate) 95% CI calculated by NCC-AC using SE 0.09 from study
Inter-observer (junior doctor–optometrist) agreement for treatment required (3–5 v 1–2)**
weighted kappa statistic κw
Mean (95%CI) κw = 0.62 (0.45 – 0.79) (substantial) 95% CI calculated by NCC-AC using SE 0.088 from study
Diagnosis of glaucoma (with reference standard defined by consultant)Group 1
Sensitivity: 0.76 (95% CI: 0.57–0.89)
Specificity: 0.93 (95% CI: 0.85–0.97)
Group 2
Sensitivity: 0.66 (95% CI: 0.48–0.81)
Specificity: 0.89 (95% CI: 0.80–0.95)
Treatment of glaucoma (with reference standard defined by consultant)Group 1
Sensitivity: 0.73 (95% CI: 0.57–0.85)
Specificity: 0.96 (95% CI: 0.88–0.99)
Group 2
Sensitivity: 0.64 (95% CI: 0.47–0.78)
Specificity: 0.90 (95% CI: 0.80–0.95)
Banes et al., 20008

Study design:
Prospective observational

Observer masked
Patient group: patients from general glaucoma clinic. Moorfields Eye Hospital
Some patients had other ocular pathologies. Most patients had a diagnosis of POAG and were on medical treatment

Inclusion criteria:
NR

All patients
N: 54
Age (mean): NR
M/F: NR
No demographic data was reported
Group 1:
1 senior optometrist

Group 2:
1 general ophthalmologist (research fellow)

Examination methods:
Visual fields were carried out by a technician before assessment.
Both optometrist and research fellow carried out the following:
  • Clinical history of medication including adverse events
  • Slit lamp biomicroscopy to assess anterior segment and optic disc
    • Drawing of disc
    • Haemorrhages
    • Disc size
    • VF (24-2) plots were considered
    • Stable
    • Progressive
    • Unreliable
  • Goldmann tonometry
  • Management of patient according to clinical state was assessed
    • Continue with treatment
    • Change treatment
    • Stop treatment
    • Consider surgery
  • Length of time to next appointment
    • < 2 months
    • 3 months
    • 6 months
    • 1 year
    • Discharge
Inter-observer agreement for visual field assessment (right eyes)
kappa statistic κ* (% agreement)
= 0.81 (very good) (92%) (3 eyes had missing data and 4 eyes were disagreed upon)Funding:
NR

Limitations:
  • No confidence intervals for kappa
  • The kappa value agreement scale was not mentioned. It was assumed to be from (Landis and Koch 1977)
Additional Outcomes:

Notes:
* kappa was calculated excluding missing values Patients were randomly distributed to optometrist and research fellow by clerk but the optometrist did not see any postoperative or complicated cases.

The research fellow was masked to the observations of the optometrist
Inter-observer agreement for visual field assessment (left eyes)
kappa statistic κ* (% agreement)
= 0.80 (good) (91%)
Inter-observer agreement for management recommendations (right eyes)
kappa statistic κ* (% agreement)
= 1.00 (very good) (100%) (Group 2 had not recorded data for 3 eyes)
Inter-observer agreement for management recommendations (left eyes)
kappa statistic κ* (% agreement)
= 0.93 (very good) (98%) (6 eyes had missing data and 1 eye was disagreed upon)
Inter-observer agreement for follow up recommendations
kappa statistic κ* (% agreement)
= 0.97 (very good) (98%) (5 eyes had missing data and 1 eye was disagreed upon)
Banes et al., 20067

Study design:
Prospective + Retrospective observational study
Patient group:
350 patients attending glaucoma outpatient services at Moorfields, UK

Inclusion criteria: Exclusion criteria: All patients
N: 350
Age (median): NR
M/F: NR
Dropouts: 1 (one hospital record could not be retrieved)

No demographic data was reported
Group 1
4 certified optometrists with a College of Optometry diploma in glaucoma in hospital setting with patient assessment and management experienced gained from 3 – 10 years of 1–2 half day sessions/week. Training consisted of patient assessments in supportive environment with access to an ophthalmologist.

Group 2
3 medical clinicians (associate specialists) working part-time in glaucoma clinics for ≥ 10 years

Group 3
2 consultant ophthalmologists retrospectively reviewed the patient records and clinical decisions and made independent management decisions

Examination methods:
Optic disc assessment for glaucomatous damage or normal disc was performed independently of the main study using 134 stereo pairs of disc photographs. Results were compared to previously published data.

All patients had a visual field test performed by a technician before clinical assessment. The optometrists and medical clinicians then performed a structured clinical assessment on each of their 50 patients then used the clinical data to make management decisions on 5 aspects of patient care:
  1. Visual field status (stable, progression, unreliable, non-glaucoma, other)
  2. Clinical management 1 (no treatment, continue, start/increase treatment, reduce)
  3. Clinical management 2 (consider glaucoma surgery, consider cataract surgery, change treatment due to intolerance, reinforce compliance, discuss with consultant)
  4. Planned tests (disc photographs, HRT, VF, IOP phasing
  5. Time to next appointment in months (1–2, 3, 6 9 12, discharge)
Detection of glaucomatous disc
using 134 stereo pairs (with glaucomatous damage defined checking against previously published data)
Group 1
Sensitivity: range 77.8% – 88.2%
Specificity: range 76.0% – 79.0%
Group 2
Sensitivity: range 64.7% – 74.2%
Specificity: range 82.3% – 93.0%
Funding: NR

Limitations:
Mean kappa statistic not reported with confidence intervals

Additional outcomes:

Notes:
Patients allocated by clinic clerk on a sequential basis to specialist ophthalmologist or optometrist (50 patients each)

*Weighted kappa statistic κw

Weights assigned for time to next clinical appointment: 1.0 = agreement; 0.75 = 1 step away disagreement; 0.5 = 2 steps away disagreement ; 0.25 = 3 steps away disagreement, 0 = 4 steps away disagreement and disagreement for discharge and missing data

Kappa value agreement
0.00 to 0.2 = poor
0.21 to 0.40 = fair
0.41 to 0.60 = moderate
0.61 to 0.80 = good
0.81 to 1.00 = very good
Inter-observer agreement for visual field status (kappa statistic & % agreement)Group 3 (Consultant 1) v Group 1 κ = 0.33 fair (55%)
Group 3 (Consultant 2) v Group 1 κ = 0.27 fair (54%)
Mean κ = 0.30 fair
Group 3 (Consultant 1) v Group 2 κ = 0.22 fair (44%)
Group 3 (Consultant 2) v Group 2 κ = 0.21 fair (43%)
Mean κ = 0.22 fair
Inter-observer agreement for clinical management 1 (kappa statistic & % agreement)Consultant 1 v Group 1 (certified optometrists) κ = 0.67 good (79%)
N=199 (3% missing data)
Consultant 1 v Group 2 (general ophthalmologists) κ = 0.52 moderate (71%)
N=150 (5.3% missing data)
% agreement for clinical management 2Consider cataract surgery:
Group 3 (Consultant 1) v Group 1 94%
Group 3 (Consultant 1) v Group 2 91%
Consider glaucoma surgery:
Group 3 (Consultant 1) v Group 1 95%
Group 3 (Consultant 1) v Group 2 99%
Reinforce Compliance:
Group 3 (Consultant 1) v Group 1 97%
Group 3 (Consultant 1) v Group 2 99%
Discuss with consultant:
Group 3 (Consultant 1) v Group 1 72%
Group 3 (Consultant 1) v Group 2 81%
% agreement for planning of testsVisual Field:
Group 3 v Group 1 mean 62% (C1 & C2)
Group 3 v Group 2 mean 54% (C1 & C2)
Imaging:
Group 3 v Group 1 mean 73% (C1 & C2)
Group 3 v Group 2 mean 61% (C1 & C2)
Phasing:
Group 3 v Group 1 mean 98% (C1 & C2)
Group 3 v Group 2 mean 100% (C1 & C2)
Disc Photo:
Group 3 v Group 1 mean 91% (C1 & C2)
Group 3 v Group 2 mean 100% (C1 & C2)
Next clinic appointment weighted kappa statistic κw * and % agreementGroup 3 (Consultant 1) v Group 1 (certified optometrist) κw = 0.35 fair (79%)
Group 3 (Consultant 1) v Group 2 (general ophthalmologist) κw = 0.29 fair (73%)
Harper et al., 200056

Study design:
Retrospective observational study
Patient group:
48 optic disc stereophotographs retrospectively selected from of glaucomatous and non glaucomatous patients attending glaucoma service in Greenwich Hospital, UK

Inclusion criteria:
Photographs that were representative of a wide range of disc appearances classified using a visual analogue scale (VAS) 0= definitely non-glaucomatous and 100= definitely glaucomatous by a glaucoma specialist. Matched visual field data was not available for the stereophotographs

All patients
N: 48
Age (median): NR
M/F: NR
Glaucomatous damage (defined by VAS):
  • Definitely non-glaucomatous ≤10): 11
  • Definitely glaucomatous ≥90): 15
  • Suspicious (11–89): 22
Patient demographics were not reported
Group 1
3 optometrists with 4 years accredited training ≥ 4 years post registration experience. None had specialist shared care expertise

Group 2
2 general ophthalmologists. One SPR and one associate specialist in medical ophthalmology. Neither had sub-speciality training although the associate specialist had responsibility for reporting on fundus/disc photographs

Examination methods:
Photographs had been taken with a standard fundus camera with stereopsis achieved through decentration of camera angle. They were examined through a Carl-Zeiss 2× stereoscopic viewer and standard light box

Each observer
  1. Estimated vertical cup disc ratio (VCD)
  2. Grading of narrowest rim width estimate
  3. Haemorrhage present or absent
Also graded using simple ranking/ordinal scales
4.

Focal pallor of neuroretinal rim

5.

Extent of peri-papillary atrophy

6.

Steepness of cup-edge

7.

Cribriform sign as present or absent

Inter-observer (ophthal-optom) agreement in estimating VCD
weighted kappa statistic κw*
Mean κw = 0.46 (moderate)
Range from 0.23 (fair) to 0.64 (substantial)
Funding:
College of optometrists

Limitations:
  • No confidence intervals available for Mean weighted kappa statistic or SD
  • No patient demographics
Notes:
Observers were presented photographs in a masked and random fashion with at least 5 days between the 2 assessments of each photograph

*Weighted kappa statistic κw
Weights assigned to each observation for VCD were equal to 1 minus (difference between estimates). 0.0 difference = 1, 0.1 difference = 0.9 weight etc until 1.0 difference = 0. Smaller disagreements were weighted more heavily Kappa value agreement (Landis and Koch 1977)
−1.00 to 0 = poor
0.01 to 0.2 = slight
0.21 to 0.40 = fair
0.41 to 0.60 = moderate
0.61 to 0.80 = substantial
0.81 to 0.99 = almost perfect
+1.00 = perfect
Inter-observer (ophthal-optom) agreement in estimating VCD
1 × standard deviation of difference scores
Mean SD = 0.19 (range 0.13 – 0.22) (4/6 mean differences were significantly different p<0.01)
Inter-observer (ophthal-optom) agreement in estimating rim:diameter ratio
weighted kappa statistic κw*
Mean κw = NR
Range from 0.29 (fair) to 0.65 (substantial)
Inter-observer (ophthal-optom) agreement in estimating rim:diameter ratio
1 × standard deviation of difference scores
Mean SD = NR (range 0.09 – 0.15) (3/6 mean differences were significantly different p<0.01)
Inter-observer (ophthal-optom) detection of disc haemorrhage as present or absent
(kappa statistic - unweighted)
Mean κ = 0.77 (substantial)
Range from 0.61 (substantial) to 0.91 (almost perfect)
% agreement ranges from 90–98%)
Inter-observer (ophthal-optom) agreement on neuroretinal rim pallor
weighted kappa statistic κw*
Mean κw = 0.23 (fair)
Inter-observer (ophthal-optom) agreement on peri-papillary atrophy
weighted kappa statistic κw*
Mean κw = 0.45 (moderate)
Inter-observer (ophthal-optom) agreement on steepness of cup edge
weighted kappa statistic κw*
Mean κw = 0.50 (moderate)
Inter-observer (ophthal-optom) agreement on cribriform sign
weighted kappa statistic κw*
Mean κw = 0.48 (moderate)
Harper et al., 200155

Study design:
Retrospective observational study
Patient group:
48 optic disc stereophotographs retrospectively selected from of glaucomatous and non glaucomatous patients attending glaucoma service in Greenwich Hospital, UK

Inclusion criteria:
Photographs that were representative of a wide range of disc appearances classified using a visual analogue scale (VAS) 0= definitely non-glaucomatous and 100= definitely glaucomatous by a glaucoma specialist. Matched visual field data was not available for the stereophotographs

All patients
N: 48
Age (median): NR
M/F: NR
Glaucomatous damage (defined by VAS):
  • Definitely non-glaucomatous≤10): 11
  • Definitely glaucomatous ≥90): 15
  • Suspicious (11–89): 22
Patient demographics were not reported
Group 1
6 optometrists with 4 years accredited training. 2 had 1 year of post-registration experience, 2 had 4 years of post-registration experience and 2 had ≥ 10 years of post-registration experience. None had been involved in shared care schemes or had specialist training. All employed full or part-time in primary care optic role.

Group 2
6 general ophthalmologists: 2 SPR and 2 SHOs and 2 consultants with subspecialty expertise in glaucoma.

Examination methods:
Photographs had been taken with a standard fundus camera with stereopsis achieved through decentration of camera angle. They were examined through a Carl-Zeiss 2× stereoscopic viewer and standard light box

Each observer
  1. Estimated vertical cup disc ratio (VCD) uncorrected for disc size
  2. Grading of narrowest rim width estimate
  3. Haemorrhage present or absent
The features were discussed between each observer and the researcher prior to grading. All 12 observers had opportunity to read instructions for grading criteria
Inter-observer (ophthal-optom) agreement in estimating VCD
weighted kappa statistic κw*
Mean (95%CI) κw = 0.36 (0.31 – 0.41) (fair)
Range for κw from 0.06 (slight) to 0.63 (substantial)
Funding:
NR

Limitations:
  • No patient demographics
Notes:
Observers were presented photographs in a masked and random fashion with at least 5 days between the 2 assessments of each photograph

*Weighted kappa statistic
Weights assigned to each observation for VCD were equal to 1 minus (difference between estimates). 0.0 difference = 1, 0.1 difference = 0.9 weight etc until 1.0
difference = 0.
Smaller disagreements were weighted more heavily

Kappa value agreement (Landis and Koch 1977)
−1.00 to 0 = poor
0.01 to 0.2 = slight
0.21 to 0.40 = fair
0.41 to 0.60 = moderate
0.61 to 0.80 = substantial
0.81 to 0.99 = almost perfect
+1.00 = perfect
Inter-observer (ophthal-optom) agreement in estimating VCD
1 × standard deviation of difference scores
Mean (95%CI) SD = 0.18 (0.17 – 0.20)
Range 0.10 – 0.28 (25/36 mean differences were significantly different p<0.01 or <0.001 or <0.0001)
Inter-observer (ophthal-optom) agreement in estimating rim:diameter ratio
weighted kappa statistic κw*
Mean (95%CI) κw = 0.35 (0.29 – 0.41) (fair)
Range for κw from −0.01 (poor) to 0.77 (substantial)
Inter-observer (ophthal-optom) agreement in estimating rim:diameter ratio
1 × standard deviation of difference scores
Mean (95%CI) SD = 0.11 (0.11 – 0.12)
Range 0.08 – 0.15 (23/36 mean differences were significantly different p<0.01 or <0.001 or <0.0001)
Inter-observer (ophthal-optom) detection of disc haemorrhage as present or absent
(unweighted kappa statistic)
Mean (95%CI) κ = 0.42 (0.37 – 0.47) (moderate)
Range 0.12 (slight) to 0.72 (substantial)
Spry, 1999142 & Gray, 200052
[Bristol Shared Care Glaucoma Study]

Study design:
RCT

Evidence level:
+

Duration of follow-up:
2 years

Computer generated random numbers and allocation concealment
Patient group: glaucoma patients and glaucoma suspects attending glaucoma clinic

Setting: Bristol Eye Hospital, UK

Inclusion criteria: Exclusion criteria:
  • <50 years
  • Unstable glaucoma
  • Normal tension glaucoma
  • Narrow angle glaucoma
  • Other coexisting ocular pathology
  • Extensive field loss (>66/12 missed points on Henson 132 point threshold related suprathreshold examination
  • Best corrected VA in either eye worse than 6/18
All patients
N: 403

Group 1 (HES)
N: 200
Age (mean ± SD): 69.4 ± 8.8
M/F: 115/85
Mean glaucoma suspects
Male: 48
Female: 30 Family history: 35
Previous cataract extraction: 14
LogMAR both eyes (mean ± SD): 0.06 ± 0.18
Drop outs: 38 (died = 7, moved = 2, general health = 6, lost to follow up = 23)

Group 2 (CO)
N: 203
Age (mean ± SD): 68.0 ± 8.3
M/F: 103/100
Mean glaucoma suspects
Male: 51
Female: 44
Family history: 48
Previous cataract extraction: 8
LogMAR both eyes (mean ± SD): 0.06 ± 0.17
Drop outs: 19 (died = 5, moved = 4, general health = 3, other = 7)
Group 1
Routine follow up** in Hospital Eye Service (HES) comprising by a general ophthalmologist:
  • VF analysis with Henson CFS2000/CFA3000
  • Single IOP measurement using Goldmann Applanation Tonometry (GAT)
  • Vertical cup-disc ratio (VCD) using direct ophthalmoscopy or indirect binocular ophthalmoscopy
Group 2
Structured 6 monthly follow-up at specially trained (instruction through lectures and demonstrations from study researchers) Community Optometrist (CO) comprising:
  • VF analysis using
    Henson CFA 3000 132 point threshold related suprathreshold examination
  • Repeat VF examination on 50% patients
  • Single IOP measurement using GAT
  • VCD using direct ophthalmoscopy or indirect binocular ophthalmoscopy (dilated pupil)
Examination methods:
A research clinic reference standard (RCRS) examination was performed on each patient at baseline pre-randomisation and 2 year follow up comprising:
  • VF analysis using Henson CFA 3000 132 point threshold related suprathreshold examination
  • Repeat VF examination
  • Triple IOP measurement using GAT
  • VCD using direct ophthalmoscopy or indirect binocular ophthalmoscopy (dilated pupil)
  • Stereophotographic analysis of VCD by observer 1
  • Stereophotographic analysis of VCD by observer 2
Mean number of points missed on visual field testing ± SD
Better Eye
Group 1: 7.9 ± 12.0
Group 2: 6.8 ± 10.8
Difference between means: 0.07 (95% CI: − 1.86, 2.04)
p value: 0.94 (ANCOVA)* not signif.
Funding:
MRC, International Glaucoma Association, R&D Directorate NHS Executive South and West and Avon Health Authority

Limitations:

Notes:
*ANCOVA: analysis of covariance was performed for each outcome variable comparing the 2 follow up groups adjusting for baseline measurements.
Control was also considered for age, sex, time from recruitment to follow up, treatment at baseline, treatment at any time (any/none) and diagnosis (glaucoma suspect/established POAG)

$Adjusted Intraclass Correlation Coefficient (ICC):
The ICC is an equivalent to a quadratic weighted kappa statistic as a chance corrected measure of agreement which corrects for systematic bias, weighting discrepancies according to square of the differences between the paired measurements.

ICC = <0.2 “slight agreement”;
ICC = 0.21–0.40 “fair agreement”;
ICC = 0.41–0.60 “moderate agreement;
ICC = 0.61–0.80 “substantial agreement;
ICC =0.80 “almost perfect agreement.

**For HES group mean time to first follow up 10.7 ± 5.4 months (range 3 – 24 months) Median number of visits within 2 year period was 2.8 (range 0–8)

Additional outcomes:
RCRS v HES (all outcomes and RCRS v CO (all outcomes
Mean number of points missed on visual field testing ± SD
Worse Eye
Group 1: 20.2 ± 21.6
Group 2: 18.3 ± 19.9
Difference between means: 0.04 (95% CI: − 3.49, 3.40)
p value: 0.98 (ANCOVA)* not signif.
Mean IOP (mmHg) ± SD
Better Eye
Group 1: 19.3 ± 5.1
Group 2: 19.3 ± 4.7
Difference between means: 0.26 ± (95% CI: −1.21, 0.68)
p value: 0.59 (ANCOVA)* not signif.
Mean IOP (mmHg) ± SD
Worse Eye
Group 1: 19.1 ± 5.5
Group 2: 19.0 ± 5.3
Difference between means: 0.53 ± (95% CI: −1.58, 0.51)
p value: 0.32 (ANCOVA)* not signif.
Cup disc ratio ± SD
Better Eye
Group 1: 0.72 ± 0.12
Group 2: 0.72 ± 0.13
Difference between means: 0.00 (95% CI: −0.02, 0.03)
p value: 0.70 (ANCOVA)* not signif.
Cup disc ratio ± SD
Worse Eye
Group 1: 0.74 ± 0.13
Group 2: 0.74 ± 0.14
Difference between means: 0.00 (95% CI: −0.03, 0.03)
p value: 0.70 (ANCOVA)* not signif.
VCD (inter centre agreement)
Right Eye
Mean Difference: −0.05 (95% CI: −0.03, − 0.07)
$Adjusted ICC: 0.50 (moderate agreement)
N=360
VCD (inter centre agreement) Left EyeMean Difference: 0.05 (95% CI: 0.03, 0.07)
$Adjusted ICC: 0.54 (moderate)
N=358
IOP mmHg (inter centre agreement) Right EyeMean Difference: 0.4 (95% CI: −0.05, 0.85)
$Adjusted ICC: 0.45 (moderate)
N=388
IOP mmHg (inter centre agreement) Left EyeMean Difference: 0.6 (95% CI: 0.13, 1.07)
$Adjusted ICC: 0.40 (fair)
N=388
VF points missed (inter centre agreement) Right EyeMean Difference: 1.1 (95% CI: −0.38, 2.58)
$Adjusted ICC: 0.55 (moderate)
N=287
VF points missed (inter centre agreement) Left EyeMean Difference: 0.7 (95% CI: −0.80, 2.20)
$Adjusted ICC: 0.61 (substantial)
N=287
Theodossiades & Murdoch, 2001148

Study design:
Prospective observational
Patient group:
Volunteers from Moorfields Eye Hospital glaucoma clinics, UK

Inclusion criteria:
Wide range of normal and glaucomatous disc features

All patients
N: 50
Age (median): NR
M/F: NR
Glaucomatous damage (defined by consultant): Patient demographics were not reported
Group 1
8 community optometrists based in high street optometric practices. 6 also worked part-time in the hospital eye service but not for glaucoma.
Optometrists received 2 hours of lectures on assessment of optic nerve head

Group 2
Consultant ophthalmologist with specialist interest in glaucoma

Examination methods:
Both undilated eyes of each patient were first examined by the consultant ophthalmologist using slit lamp biomicroscopy and one eye selected for examination by optometrist.
Optometrists assessed one undilated eye through a direct ophthalmoscope of each patient for the following parameters:
  1. Vertical disc diameter
  2. Vertical cup disc ratio (VCD)
  3. Neuroretinal configuration
  4. Cup shape
  5. Neuroretinal rim colour
  6. Vessel configuration
  7. Haemorrhage
  8. Extent of peri-papillary atrophy
  9. Health status of optic nerve head
These were then used to give a final opinion on presence or absence of glaucomatous damage
Inter-observer agreement in Vertical disc diameter
weighted kappa statistic κw*
Mean (95%CI) κw = 0.34 (0.26 – 0.42) (fair)Funding:
International Glaucoma Association

Limitations:
  • No patient demographics
  • Weighting method for VCD and vertical disc diameter was not reported
  • Observer masking was not reported
  • Patients were not recruited in a randomised or consecutive fashion.
Notes:
Kappa value agreement based on (Landis and Koch 1977)
0.00 to 0.2 = poor
0.21 to 0.40 = fair
0.41 to 0.60 = moderate
0.61 to 0.80 = good
0.81 to 1.00 = very
Inter-observer agreement in VCD
weighted kappa statistic κw*
Mean (95%CI) κw = 0.84 (0.81 – 0.87) (very good)
Inter-observer agreement in Neuroretinal configuration
kappa statistic κw
Mean (95%CI) κw = 0.67 (0.58 – 0.76) (good)
Inter-observer agreement in Cup shape
kappa statistic κw
Mean (95%CI) κw = 0.66 (0.58 – 0.74) (good)
Inter-observer agreement in Neuroretinal rim colour
kappa statistic κw
Mean (95%CI) κw = 0.32 (0.25 – 0.38) (fair)
Inter-observer agreement in Vessel configuration
kappa statistic κw
Mean (95%CI) κw = 0.53 (0.40 – 0.65) (moderate)
Inter-observer agreement in Haemorrhage
kappa statistic κw
Mean (95%CI) κw = 0.67 (0.45 – 0.89) (good)
Inter-observer agreement in Peri-papillary atrophy
kappa statistic κw
Mean (95%CI) κw = 0.22 (0.14 – 0.29) (fair)
Inter-observer agreement in Health status of optic nerve head
kappa statistic κw
Mean (95%CI) κw = 0.62 (0.53 – 0.70) (good)
Health status of optic nerve head (reference standard defined consultant)Sensitivity: 0.90 (95% CI: 0.86 – 0.94)
Specificity: 0.73 (95% CI: 0.66 – 0.80)

Abbreviations: NR=not reported, NA=not applicable, M/F=male/female, N=total number of patients randomised, SD=Standard Deviation, SE(M)=Standard Error (of the mean), ITT – Intention to Treat etc

From: Appendix D, Evidence tables

Cover of Glaucoma
Glaucoma: Diagnosis and Management of Chronic Open Angle Glaucoma and Ocular Hypertension.
NICE Clinical Guidelines, No. 85.
National Collaborating Centre for Acute Care (UK).
Copyright © 2009, National Collaborating Centre for Acute Care.

Apart from any fair dealing for the purposes of research or private study, criticism or review, as permitted under the Copyright, Designs and Patents Act, 1988, no part of this publication may be reproduced, stored or transmitted in any form or by any means, without the prior written permission of the publisher or, in the case of reprographic reproduction, in accordance with the terms of licences issued by the Copyright Licensing Agency in the UK. Enquiries concerning reproduction outside the terms stated here should be sent to the publisher at the UK address printed on this page.

The use of registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant laws and regulations and therefore for general use.

The rights of National Collaborating Centre for Acute Care to be identified as Author of this work have been asserted by them in accordance with the Copyright, Designs and Patents Act, 1988.

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.