Title: Intra- and Inter- Examiner Repeatability of Cycloplegic Retinoscopy Among

Young Children

Running Head: Repeatability of Cycloplegic Retinoscopy

Authors: Sara J McCullough1, Lesley A Doyle1*, Kathryn J Saunders1

1 Biomedical Sciences Research Institute, School of Biomedical Sciences, University

of Ulster, Cromore Road, Coleraine, N. Ireland.

* Corresponding Author

Email: la.doyle@ulster.ac.uk

Keywords: vision science, optometry, retinoscopy, cycloplegia

Disclosure: The authors report no conflicts of interest and have no proprietary

interest in any of the materials mentioned in this article.

Word Count: 2102

1

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

Abstract

Purpose To evaluate the intra- and inter-examiner repeatability of cycloplegic retinoscopy in

young children aged 4-5 years old.

Methods Examiner 1 refracted all children in the first sample (n=108); firstly with masked loose

lenses, then using unmasked loose lenses (intra-examiner repeatability). Examiners

1 and 2 refracted all children in the second sample (n=97) using unmasked loose

lenses, blind to the child’s refractive error, presence/magnitude of habitual spectacle

correction and to each other’s findings (inter-examiner repeatability). Refractions

were performed on one eye chosen at random. Mean differences and 95% limits of

agreement (LOAs) and confidence intervals were calculated for intra- and inter-

examiner repeatability of sphere, cylinder and spherical equivalent refraction (SER).

Results Participants had a wide range of refractive errors (-1.50DS to +7.25DS; ≥4.50DC).

Mean differences (95% LOAs) were small for both intra- and inter-examiner

repeatability [Intra:Sphere 0.00D (-0.85, +0.85D), Cyl -0.03D (-0.68, +0.62D), SER -

0.06D (-0.90, +0.78D); Inter:Sphere -0.08D (-0.92, +0.76D), Cyl -0.08D (-0.75,

+0.59D), SER -0.13D (-0.95, +0.69D). A statistically significant proportional bias was

present for intra-examiner repeatability of cylinder (ρ=0.20, p=0.04) and SER

measurement (ρ=0.19, p=0.049). Proportional bias was not present for any other

measure (p>0.12). Examiners agreed on cylinder axis within ±20o in 71% of

refractions where astigmatism of -0.75D or higher was present. 80% of intra- and

inter-examiner measures fell within ±0.50D for spherical and cylindrical components.

Conclusions Differences of ±1.00D and ±0.75D or more for spherical and cylindrical measures

respectively can be considered significant when performing cycloplegic retinoscopy

on young children.

2

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

Introduction

Full cycloplegia has been recommended to obtain an accurate refraction in children

under 12 years of age1 particularly in the presence of suspected strabismus

(esotropia), latent hyperopia or where there is poor fixation or cooperation from the

child. Repeatability and validity of the measurement of refractive error using

cycloplegic autorefraction in children is well established2-6 and the agreement

between cycloplegic retinoscopy and non-cycloplegic refraction techniques has been

investigated.7-9 Zadnik et al.4 and Walline et al.10 measured intra-examiner

repeatability of cycloplegic retinoscopy in a small sample of healthy adults (n=40, 20

to 43 years old) but to date no published reports exist detailing intra- and inter-

examiner repeatability of cycloplegic retinoscopy in the population in which its use is

regarded as gold-standard. An appreciation of the repeatability of cycloplegic

retinoscopy is necessary in clinical practice and for epidemiological studies of

refractive error in order to determine a ‘real’ change in refractive error between

clinicians or over time. The purpose of the present study is to evaluate the intra- and

inter-examiner repeatability of refraction by cycloplegic retinoscopy in a group of

young children aged 4-5 years old.

Methods

Participants

A total of 198 children in their first year of formal education within mainstream

schools in Northern Ireland (Kindergarten equivalent) were recruited for the study.

Participants were recruited from state primary schools that were non-selective in

academic ability and drew children from a range of socioeconomic backgrounds and

rural/urban environments. Intra-examiner repeatability was conducted in one sample

of children (n=108) and inter-examiner repeatability conducted in a second sample

(n=97). Seven children were included in both samples. Written informed consent

was obtained from the parents/guardians of the participants and verbal assent was

given by the participant on the day data collection took place. Data collection took

place on school premises during school time.

3

57

58

59

60

61

62

63

64

65

66

67

68

69

70

71

72

73

74

75

76

77

78

79

80

81

82

83

84

85

86

87

Ethical Approval

The study was approved by University of Ulster Research Ethics Committee and the

conduct of the study adhered to the tenets of the Declaration of Helsinki.

Procedures

The magnitude of spherical and cylindrical refractive error was assessed using

streak retinoscopy (Keeler Professional) by two experienced optometrists (Examiner

1-SJM, Examiner 2-KJS) at least 30 minutes after the instillation of one drop of 0.5%

proxymetacaine hydrochloride and one drop of 1.0% cyclopentolate hydrochloride in

each eye. Retinoscopy was performed on one eye chosen in a pseudorandom

fashion. Right eyes were measured using the right eye of the examiner holding the

retinoscope with their right hand and left eyes measured with the left eye of the

examiner holding the retinoscope in the left hand.11 Neutrality of the retinoscopy

reflex was achieved using a combination of spherical and minus cylindrical lenses

placed in a trial frame (Keeler Oculus Universal). Retinoscopy was carried out in a

darkened room while the child fixated on the retinoscope reflex. A ‘working’ distance

of 67cm from the retinoscope to the trial frame was maintained using a fixed string

attached to the retinoscope. The ‘working’ distance was checked using the string at

the start of the procedure, intermittently throughout and at the end of the procedure

when the examiner achieved neutralisation.

Intra-examiner repeatability

In the assessment of intra-examiner repeatability, Examiner 1 (SJM) refracted all

children within the first sample (n=108) firstly using masked loose lenses in

0.25DS/0.25DC steps held within a turntable and subsequently with unmasked loose

lenses in 0.25DS/0.25DC steps held within a trial lens case. The masked lenses

were marked with a number identifier and were decoded after data collection was

complete.

4

88

89

90

91

92

93

94

95

96

97

98

99

100

101

102

103

104

105

106

107

108

109

110

111

112

113

114

115

116

117

Inter-examiner repeatability

In the assessment of inter-examiner repeatability Examiners 1 (SJM) and 2 (KJS)

refracted all children within the second sample (n=97) using unmasked loose lenses

in 0.25DS/0.25DC steps from a trial lens case and were blind to each other’s

findings.

Statistical Analysis

Statistical analyses were performed using Stata 13.0 (StataCorp, College Station,

TX, USA). Spherical equivalent refraction (SER) was calculated using [Sphere+

(Cylinder/2)]. Mean differences, 95% limits of agreement (LOA’s) and confidence

intervals (CI’s) were calculated for the SER and the spherical and cylindrical

components. LOA’s were calculated as ‘mean difference ± (1.96 X standard

deviation)’ and have been used to allow for Bland Altman analysis and direct

comparison to previous studies.4,10 CI’s were calculated as ‘limit of agreement ± (1.66

X standard error)‘ and have also been included as recommended by McAlinden et

al.19 when reporting on the study of agreement and precision. While the spherical

and cylindrical components and cylindrical axis are most relevant to clinical practice,

SER and vector components (J0 and J45) have also been included in analysis. SER is

widely reported in epidemiological studies of refractive error12-14 and the use of vector

analysis allows for the consideration of both the magnitude and direction of

astigmatism. Bland and Altman15 plots were also used to inspect the intra- and inter-

examiner repeatability. Wilcoxon matched-pairs signed rank tests were used to

determine statistically significant intra- and inter-examiner differences. Spearman’s

correlations were used to identify proportional bias. A p value of less than 0.05 was

considered statistically significant.

Results

Intra-Examiner Group

5

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

136

137

138

139

140

141

142

143

144

145

146

Participants were 50 males (46%) and 58 females (54%), with a mean age of

5.1±0.37 years (range 4.0 to 5.8 years). The majority of participants were white

(n=100, 93%) consistent with the demographics of the Northern Irish population.16 A

total of 56 right (52%) and 52 left eyes (48%) were examined for intra-examiner

repeatability.

The spherical and cylindrical components of refractive error, SER and vector

components determined by both Examiner 1 and Examiner 2 were not normally

distributed. Table 1. details the median, inter-quartile range and range for the

spherical, cylinder and vector components as determined by Examiner 1 for intra-

examiner repeatability.

Inter-Examiner Group

Participants were 44 males (45%) and 53 females (55%), with a mean age of

5.2±0.37 years (range 4.4 to 5.9 years). The majority of participants were white

(n=93, 96%). A total of 50 right (52%) and 47 left eyes (48%) were examined for

inter-examiner repeatability.

Table 2. details the median, inter-quartile range and range for the spherical, cylinder

and vector components as determined by Examiners 1 and 2 for inter-examiner

repeatability.

Intra-Examiner and Inter-Examiner Repeatability

Mean differences, standard deviations (SD), 95% limits of agreement (LOA’s) and

confidence intervals (CI’s) for intra-examiner and inter-examiner repeatability are

detailed in Table 3 and 4. The percentage of measures falling within ±0.25D and

±0.50D are also reported.

6

147

148

149

150

151

152

153

154

155

156

157

158

159

160

161

162

163

164

165

166

167

168

169

170

171

172

173

174

Figures 1 (A, B & C) & 2 (A, B & C) show the Bland and Altman plots for intra-

examiner and inter-examiner repeatability for the spherical and cylindrical

components and SER respectively. There was no statistically significant

proportional bias for the intra-examiner repeatability for the spherical component

(ρ=0.11, p=0.26). A statistically significant proportional bias was found for the intra-

examiner repeatability of the cylindrical component (ρ=0.20, p=0.04) and the SER

(ρ=0.19, p=0.049). No statistically significant proportional bias was found for inter-

examiner repeatability for the spherical and cylindrical components and the SER (all

Spearman correlations, p>0.12).

Inter-examiner Repeatability of Cylinder Axis

For those participants judged to have an astigmatic error of -0.75DC or higher by

either examiner [i.e. those cylinders likely to have a significant impact on visual

acuity17] (n=18), agreement of cylinder axis was within ±10o for 53% and within ±20o

for 71%.

Discussion

This is the first study to report the intra-examiner and inter-examiner repeatability of

cycloplegic retinoscopy in young children over a wide range of refractive errors

representative of the population at this age.18 Our results show that mean differences

between intra-examiner repeat measures for sphere, cylinder and spherical

equivalent refraction (SER) were small (<0.07D), representing less than one step in

the clinical procedure (0.25D) and showed no statistically significant differences.

There was no statistically significant difference between two examiners’ spherical

component measures and the mean inter-examiner difference was again small (-

0.08D) and represented less than one clinical step (0.25D). Statistically significant

differences were found between the two examiners when cylindrical component and

the SER were considered. On average, Examiner 1 over-estimated the magnitude of

the cylindrical component (mean difference= -0.08, 95% limits of agreement= -0.75

to +0.59) when compared to Examiner 2, which in turn resulted in a less positive

SER (mean difference= -0.13, 95% limits of agreement= -0.95 to +0.69). The limits

7

175

176

177

178

179

180

181

182

183

184

185

186

187

188

189

190

191

192

193

194

195

196

197

198

199

200

201

202

203

204

205

of agreements for both intra- and inter-examiner repeatability of cycloplegic

retinoscopy indicate that differences between repeat measures and between

examiners of ±1.00D or more for sphere and ±0.75D or more for cylinder can be

considered significant and denote a ‘real’ change in refractive error when using

cycloplegic retinoscopy in children. These findings are comparable to data available

for the repeatability of cycloplegic retinoscopy conducted in adults. Zadnik et al. 4

and Walline et al.10 reported intra-examiner mean differences (95% limits of

agreement) for the measurement of the spherical and cylindrical components as

0.075D (-0.87 to 1.02D) and 0.06D (-0.69 to 0.80D) respectively in a group of

healthy, young adults using cycloplegic retinoscopy. Our results show approximately

60% of both intra- and inter-examiner measures for sphere and spherical equivalent

refraction were within ±0.25D, increasing to over 80% within ±0.50D. Almost 80% of

intra-examiner measures and 70% of inter-examiner measures were within ±0.25D

for the cylindrical component increasing to almost 95% within ±0.50D for both intra-

and inter-examiner measures.

When investigating agreement, there are two potential sources of disagreement

between methods; fixed and proportional bias. Fixed bias occurs when one method

produces measures that are consistently higher or lower than the other by a constant

or fixed amount. Whereas proportional bias occurs when one method produces

values that are higher or lower than the other by an amount that is proportional to the

magnitude of the measured variable.18,19 A statistically significant proportional bias

was found for the cylindrical component and spherical equivalent refraction when

intra-examiner measures were compared. This bias was not evident in the spherical

component measures. Examiner 1 under-estimated the cylindrical component and

the level of hyperopic spherical equivalent refraction when using unmasked lenses

compared to when using masked lenses. The examiner may have had pre-

conceptions about the refractive error of the child when not masked to the

measurement which resulted in under-estimation when compared to the masked

results. There were few children within the study who were current spectacle

wearers (intra-examiner repeatability group n=9, inter-examiner repeatability group

n=8) and neither examiner was aware whether children habitually wore spectacles

when conducting retinoscopy. Few studies assessing repeatability of cycloplegic

8

206

207

208

209

210

211

212

213

214

215

216

217

218

219

220

221

222

223

224

225

226

227

228

229

230

231

232

233

234

235

236

237

238

refraction have reported on proportional bias; however inspection of Bland and

Altman plots from studies of the repeatability or reproducibility of cycloplegic

subjective refraction or autorefraction show increasing variability with increasing

ametropia.5,6 The typical procedure of cycloplegic retinoscopy in clinical practice is

better represented by the inter-examiner methodology of the present study, where

masked lenses were not used and where no proportional bias was found. These

data may be a more accurate reflection of the repeatability of cycloplegic retinoscopy

in a typical clinical setting.

Our findings show relatively good agreement between examiners for cylindrical axes

with 71% of recorded axes agreeing to within ±20o where astigmatism of -0.75DC or

more was present. The inter-examiner agreement in cylinder axis (53% within ±10o)

within the present study are similar to those reported by Walline et al.10 for repeat

measures by the same examiner using cycloplegic retinoscopy on adults; their

repeat measures agreed to within ±10o in 64% of astigmatic errors of -0.75DC or

higher.10

Strengths and Limitations

The data contain few myopic refractive errors which is reflective of the refractive

error distribution of young children of white ethnicity,14,20 however a wide range of

hyperopic and astigmatic errors were sampled.

Repeatability of cylindrical axis was assessed only between examiners and not

between repeat measures by the same examiner as it was not possible to

satisfactorily mask this component of the retinoscopy procedure in the latter

condition.

Both examiners were paediatric specialists, experienced in carrying out cycloplegic

retinoscopy in children. Whilst the findings of the present study may not be

generalizable to all clinicians, the limits of agreement reported for both intra- and

9

239

240

241

242

243

244

245

246

247

248

249

250

251

252

253

254

255

256

257

258

259

260

261

262

263

264

265

266

267

268

inter-examiner repeatability may be considered the ‘minimum’ difference required

before a true change in refractive error using cycloplegic retinoscopy in children can

be determined.

Conclusions

The 95% limits of agreement reported within the present study are useful for

clinicians and researchers wishing to identify ‘real’ refractive change between

measures made by the same examiner over time and/or between measures made

by different examiners. Differences of ±1.00D and ±0.75D or more for spherical and

cylindrical measures respectively can be considered significant when performing

cycloplegic retinoscopy on children aged 4-5 years. Over 80% of intra- and inter-

examiner measures of cycloplegic retinoscopy within the present study fell within

±0.50D for both spherical and cylindrical components.

Acknowledgements

This work was supported by a research grant from the College of Optometrists

(London, UK).

References

1. The Royal College of Ophthalmologists. Guidelines for the Management of

Strabismus in Childhood, 2012. Available from: https://www.rcophth.ac.uk/wp-

content/uploads/2014/12/2012-SCI-250-Guidelines-for-Management-of-Strabismus-

in-Childhood-2012.pdf

2. Harvey EM, Miller JM, Wagner LK, Dobson V. Reproducibility and accuracy of

measurements with a hand held autorefractor in children. Br J Ophthalmol.

1997;81:941-8.

10

269

270

271

272

273

274

275

276

277

278

279

280

281

282

283

284

285

286

287

288

289

290

291

292

293

294

295

296

297

3. Chat SW, Edwards MH. Clinical evaluation of the Shin-Nippon SRW-5000

autorefractor in children. Ophthalmic Physiol Opt. 2001;21:87-100.

4. Zadnik K, Mutti DO, Adams AJ. The repeatability of measurement of the ocular

components. Invest Ophthalmol Vis Sci. 1992;33:2325-33.

5. Steele G, Ireland D, Block S. Cycloplegic autorefraction results in pre-school

children using the Nikon Retinomax Plus and the Welch Allyn SureSight. Optom Vis

Sci. 2003;80:573-7.

6. Choong YF, Chen AH, Goh PP. A comparison of autorefraction and subjective

refraction with and without cycloplegia in primary school children. Am J Ophthalmol.

2006;142:68-74.

7. Saunders KJ, Westall CA. Comparison between near retinoscopy and cycloplegic

retinoscopy in the refraction of infants and children. Optom Vis Sci 1992;69(8):615-

22.

8. Chan OY, Edwards M. Comparison of cycloplegic and noncycloplegic retinoscopy

in Chinese pre-school children. Optom Vis Sci 1994;71(5):312-8.

9. Ozdemir O, Özen Tunay Z, Petriҫli IS, Ergintürk Acar D, Erol MK. Comparison of

non-cycloplegic photorefraction, cycloplegic photorefraction and cycloplegic

retinoscopy in children. Int J Ophthalmol. 2015;8:128-31.

10. Walline J, Kinney K, Zadnik K, Mutti DO. Repeatability and validity of

astigmatism measurements. J Refract Surg. 1999;15:23-31.

11. Safir A, Hyams L, Philpot J, Jagerman LS. Studies in refraction. I. The precision

of retinoscopy. Arch Ophthalmol. 1970;84: 49-61.

12. Negrel AD, Maul E, Pokharel GP, Zhao J, Ellwein LB. Refractive error study in

children, sampling and measurement methods for a multi-country survey, Am J

Ophthalmol. 2000;129:421-426.

11

298

299

300

301

302

303

304

305

306

307

308

309

310

311

312

313

314

315

316

317

318

319

320

321

322

323

324

325

326

327

328

329

330

331

13. Ojaimi E, Rose KA, Smith W, Morgan IG, Martin FJ, Mitchell P. Methods for a

population-based study of myopia and other eye conditions in school children: the

Sydney Myopia Study. Ophthalmic Epidemiol. 2005;12:59-69.

14. O’Donoghue L, McClelland JF, Logan NS, Rudnicka AR, Owen CG, Saunders

KJ. Refractive error and visual impairment in school children in Northern Ireland. Br J

Ophthalmol. 2010;94:1155-9.

15. Bland JM, Altman DG. Statistical methods for assessing agreement between two

methods of clinical measurement. Lancet. 1986;1:307-310.

16. Northern Ireland Statistics and Research Agency [Internet], Census 2011, Key

Statistics for Northern Ireland [cited 20th August 15] Available from

http://www.nisra.gov.uk/Census/key_stats_bulletin_2011.pdf

17. Sandhu RK, Munoz BE, Swenor BK, West SK. Refractive error and visual

function difficulty in a Latino population. Ophthalmology. 2012;119:1731-1736.

18. Ludbrook J. Comparing methods of measurement. Clin Exp Pharmacol Physiol.

1997;24:193-203.

19. McAlinden C, Khadka J, Pseudovs K. Statistical methods for conducting

agreement (comparison of clinical tests) and precision (repeatability or

reproducibility) studies in optometry and ophthalmology. Ophthalmic Physiol Opt.

2011;31:330-8.

20. Giordano L, Friedman DS, Repka MX, et al. Prevalence of refractive error among

preschool children in an urban population: The Baltimore Pediatric Eye Disease

Study. Ophthalmology. 2009;116:739-746.

12

332

333

334

335

336

337

338

339

340

341

342

343

344

345

346

347

348

349

350

351

352

353

354

355

356

357

358

359

360

361

362

363

364

365

Tables Table 1. Median, inter-quartile range (IQR) and range for the spherical, cylindrical

and vector components of refractive error for intra-examiner repeatability masked

and unmasked measurements.

Refractive Error Intra-Examiner GroupMasked Measurement

Intra-Examiner GroupUnmasked Measurement

Median Sphere(IQR)

[Range]

+1.50D(+1.25 to +2.00D)[-1.00 to +7.00D]

+1.50D(+1.13 to +2.12D)[-1.50 to +6.00D]

Median Cylinder(IQR)

[Range]

-0.25DC(-0.75 to 0.00DC)[-4.50 to 0.00DC]

-0.50DC(-0.50 to 0.00DC)[-3.50 to 0.00DC]

Median SER(IQR)

[Range]

+1.38D(+0.88 to +1.81D)[-1.50 to +6.63D)

+1.25D(+1.00 to +1.81D)[-1.75 to +6.00D]

Median J0

(IQR)[Range]

0.10(-0.13 to 0.29)[-0.63 to 2.11]

0.13(-0.17 to 0.25)[-0.47 to 1.53]

Median J45

(IQR)[Range]

-0.01(-0.01 to 0.01)[-0.77 to 1.29]

-0.01(-0.06 to 0.01)[-0.56 to 1.64]

13

366

367

368

369

370

371

372

373

374

375

376

377

378

379

380

381

382

383

384

Table 2. Medians, inter-quartile ranges (IQR) and ranges for the spherical,

cylindrical and vector components of refractive error for inter-examiner repeatability

Examiner 1 and 2 measurements.

Refractive Error Inter-Examiner GroupExaminer 1 Measurement

Inter-Examiner Group Examiner 2 Measurement

Median Sphere(IQR)

[Range]

+1.50D(+1.00 to +2.50D)[-1.50 to +7.25D]

+1.50D(+1.00 to +2.50D)[-1.00 to +8.00D]

Median Cylinder(IQR)

[Range]

-0.25DC(-0.50 to 0.00DC)[-3.50 to 0.00DC]

-0.25DC(-0.50 to 0.00DC)[-3.50 to 0.00DC]

Median SER(IQR)

[Range]

+1.50D(+0.75 to +2.25D)[-1.75 to +6.75D)

+1.50D(+1.00 to +2.25D)[-1.38 to +7.25D]

Median J0

(IQR)[Range]

0.13(-0.13 to 0.25)[-0.75 to 1.41]

0.13(-0.13 to 0.25)[-0.70 to 1.64]

Median J45

(IQR)[Range]

-0.01(-0.01 to 0.01)[-0.49 to 1.64]

-0.01(-0.01 to 0.01)[-0.60 to 1.15]

14

385

386

387

388

389

390

391

392

393

394

395

396

397

398

399

400

401

Table 3. Mean differences, standard deviations (SD), 95% limits of agreement

(LOA’s) and 95% confidence intervals (CI’s) for the spherical, cylindrical and vector

components of refractive error for intra-examiner repeatability and the percentage of

repeat measures falling within ±0.25D and ±0.50D.

INTRA-EXAMINER REPEATABILITYRefractive

ErrorMean

Difference (SD)(D)

95% LOA’s

(D)

95% CI’s(D)

Statistically significant difference?

Within ±0.25D

Within ±0.50D

Sphere 0.00(0.43)

-0.85 to +0.85

-0.99 to 0.99

Noz=-0.194, p=0.846

66% 86%

Cylinder -0.03(0.33)

-0.68 to +0.62

-0.78 to +0.73

Noz=-1.240, p=0.215

79% 94%

SER -0.06(0.43)

-0.90 to +0.78

-1.04 to +0.92

Noz=-0.672, p=0.501

59% 81%

J0 0.04(0.21)

-0.37 to +0.45

-0.41 to 0.49

Noz=-0.462,p=0.644

J45 -0.02(0.11)

-0.24 to +0.20

-0.26 to 0.22

Noz=-0.077,p=0.939

15

402

403

404

405

406

407

408

409

410

411

412

413

414

415

416

417

Table 4. Mean differences, standard deviations (SD), 95% limits of agreement

(LOA’s) and 95% confidence intervals (CI’s) for the spherical, cylindrical and vector

components of refractive error for inter-examiner repeatability and the percentage of

between examiner measures falling within ±0.25D and ±0.50D.

INTER-EXAMINER REPEATABILITYRefractive

ErrorMean

Difference(SD)(D)

95% LOA’s

(D)

95% CI’s(D)

Statistically significant difference?

Within ±0.25D

Within ±0.50D

Sphere -0.08(0.43)

-0.92 to +0.76

-1.07 to +0.91

Noz=-1.724, p=0.085

64% 89%

Cylinder -0.08(0.34)

-0.75 to +0.59

-0.87 to +0.71

Yesz=-2.011, p=0.044

67% 94%

SER -0.13(0.42)

-0.95 to +0.69

-1.10 to +0.84

Yesz=-3.257, p=0.001

59% 89%

J0 -0.02(0.25)

-0.51 to +0.47

-0.58 to 0.54

Noz=0.282,p=0.778

J45 0.10(0.33)

-0.55 to +0.75

-0.64 to 0.84

Noz=0.858p=0.391

16

418

419

420

421

422

423

424

425

426

427

428

429

430

431

432

433

434

435

Figures

17

436

437

438

439

440

Figure 1. Bland and Altman plots for intra-examiner repeatability of (A) spherical

component, (B) cylindrical component and (C) spherical equivalent refraction. The

18

441

442443

444

solid black line represents the mean difference and the dashed black lines represent

the 95% limits of agreement. The shaded grey areas encapsulate intra-examiner

differences of less than or equal to ±0.50D.

19

445

446

447

448

449

450

451

452

453

454

455

456

457

458

459

460

Figure 2. Bland and Altman plots for inter-examiner repeatability of (A) spherical

component, (B) cylindrical component and (C) spherical equivalent refraction. The

20

461

462463

464

solid black line represents the mean difference and the dashed black lines represent

the 95% limits of agreement. The shaded grey areas encapsulate inter-examiner

differences of less than or equal to ±0.50D.

21

465

466

467

468

469