Large diameter rigid lenses have enjoyed a renaissance over the past 5 to 10 years as manufacturing technologies, designs, and material science have progressed.1,2 Rigid lens fits remain a relatively small part of the overall contact lens market in the United States with market share estimates between 1 and 6%.1 The international community shows rigid lens market shares as high as 20% in some countries with scleral lenses accounting for 0.1% or less of all fits or refits.2 The relative infrequency with which these designs are fit coupled with the renewed interest in this modality driven by technological advancements suggests there may be a lack of familiarity with these designs among established practitioners and a need to ensure contemporary curriculum in educational programs are providing adequate instructional opportunities. Considerations for curricular content include design selection, fitting, and ability to identify adverse events.3–5
Manufacturer fitting guides are unique to each lens design. Though controversies exist over the optimal clinical endpoints for a physiologically acceptable fit, similarities exist for central and limbal clearances and desirable scleral landing relationships.6–9
Recent studies have evaluated observer performance estimating CCC using two-dimensional reference images10,11 and AS-OCT to assess lens settling over time.6,7,12–14 This study focused narrowly on neophyte skill at judging CCC in a corneoscleral design as defined by van der Worp,5 consistent with these study designs. Assessing limbal clearance and scleral zone apposition with the sclera are arguably more difficult than assessing CCC clinically. Including these endpoints would be necessary for a comprehensive study of fitter skills.
Multiple methods have been suggested for estimating scleral lens CCC (Table 1).6,10,11,15–20 No single method has emerged as a “gold standard” after factoring cost, convenience, and validity. The consistency of some estimation methods has been shown to vary with the experience of the observer and accuracy does not appear to be affected by number of lenses fit,10 years of practice, or residency training.11
Excessive CCC clearance is not without both optical and physiological concerns. Increases in corneal curvature induce greater amounts of higher order aberrations which are amenable to rigid lens correction, including scleral lens designs.21–23 There is a point unique to each patient where increases in vault begin to fail to adequately compensate for higher order aberrations, degrading the optical image and resulting in subjective complaints. This may necessitate a transition to reverse geometry designs, toric curves, or other modifications. Physiological considerations primarily center on concerns of potential hypoxia. Studies have assessed the respective contributions of material permeability, decreased transmissibility from lens center thickness plus reservoir thickness, and the adequacy of any tear mixing within the reservoir.7–9,24
A comprehensive study of fitting accuracy would necessarily include consideration of multiple variables such as CCC, limbal clearances, scleral zone apposition to the globe, visual performance, and physiological outcomes between cohorts of neophyte and experienced fitters, across a variety of different lens designs on regular and irregular corneas. Manufacturers offer a multitude of proprietary spherical and reverse geometry curves, transitional curves between the optic and limbal portions of each lens, scleral zone geometries, diameters, materials, and modifications. These interrelationships add confounding variables which would require a large-scale, longitudinal, prospective study with multivariate analysis to interpret their contributions to a physiologically and optically successful fit.
This study examines neophyte observer ability to accurately estimate corneoscleral lens CCC given a known lens center thickness as a biometric ruler obtained from AS-OCT values.
Protocols were approved by Research and Institutional Review Board of the Southern College of Optometry. Informed consent was administered in compliance with the declarations of Helsinki. Statistical analysis was performed using Excel 2010 (Microsoft, Santa Rosa, CA) and Analyse-it (Analyse-it ver. 2.26 Excel 12+, Leeds, LS#1HS, UK). Nonparametric statistics were used as indicated and significance levels were set at p ≤0.05 throughout. The distributions of observations were evaluated for normalcy by (1) histogram, (2) notched box-and-whisker plots with outliers comparing mean and median values, (3) Shapiro-Wilk test, and (4) normality plots. All observations and AS-OCT values were determined to be significantly different from a normal distribution. Nonparametric statistical analysis was performed to overcome the variability, the spread in the data, and non-normal distribution by comparing the median values. Data points were considered outliers if they fell within or exceeded the 1.5 interquartile ranges on box-and-whisker plots. The left eye observations included three such points (1) at 700 microns and (2) at 800 microns (Fig. 2). There were no outliers identified for the right eye data. Comparisons were made with and without outliers to assess impact on conclusions.
A prospective, observational design was used. A single participant (BS) with topographically regular corneas (OD: 43.7D/43.8D@116; OS: 43.2/44.0@178) was fit with a pair of Custom Stable (Valley Contax, Springfield, OR) corneoscleral lenses, 14.8 mm in diameter, in Boston XO (hexafocon A, Dk = 100) material with center thicknesses of 400 microns both eyes. Lens powers were −2.50 DS (OD) and OS −0.25 DS (OS). Sagittal depths were 4.05 and 3.84 mm right and left, respectively. Limbal clearance zones were ordered in a standard profile for both eyes whereas scleral landing zones were ordered as standard for the right eye and 1+ flat for the left eye to prevent seal-off. Fitting was conducted and refined in accordance with manufacturer fitting guide25 to achieve at least CCC = 150 to 250 microns, limbal clearance, and scleral landing without seal-off.
The approximate geometric centers of the lenses were assumed to be coincident with the optical centers as measured by a Nikon vertometer (Right Mfg. Co., Ltd., Maenocho, Itabashi-ku, Tokyo, Japan). A reference mark was placed on each lens over the optical centers using a fine-tipped permanent marker. Markings were reassessed and reapplied as needed before data collection to ensure wear and cleaning had not removed the reference points. The marks prevented clear views through them with a biomicroscope. To allow for this, observers were permitted to assess CCC from a point slightly nasal and adjacent to the reference marks. Lenses were allowed to equilibrate for at least 3 hours before data collection to allow for settling12,13 and rotation causing reference marks to vary in relationship to the pupillary axis.
Before data collection by each observer, a calibration procedure was performed to account for any lens settling. This consisted of lens center thickness and CCC measurements obtained by RTVue with cornea anterior segment module AS-OCT (2014–15, v. 220.127.116.11; Optovue, Fremont, CA) 26,000 A-scans per second using a scan line pattern of 2 by 2 mm for 0.04 seconds with a resolution of 1 by 1024 at 840 nm. CCC was then clinically assessed without fluorescein stain.
Observers were volunteers recruited by email from a cohort of fourth year, final semester interns (n = 34) during their on-campus clinical rotation. Each intern had two previous external clinical rotations with varying volumes and diversity of contact lens experiences. Twenty-nine (67.4%) of the available interns agreed to participate, resulting in a total of 58 (OD n = 29, OS n = 29) observations. Observers were provided instructions from a standardized script which included lens center thickness values for each eye obtained during calibration, a reference image labeled to identify different zonal layers they would encounter during the observation and to assist in orientation (see Supplemental Digital Content 1, available at http://links.lww.com/OPX/A231), and instructions to obtain estimates of CCC through or slightly nasal to the reference mark. All observations were collected on the same G2 Ultra Slit Lamp (Marco, Jacksonville, FL) using an optic section, under ×45 magnification, maximum illumination while in a dark room. Observers were instructed to begin with the light source 45 degrees off to the temporal side but allowed to perform a dynamic examination to simulate “real-world” practices. Elapsed time was not recorded. Estimates of CCC were verbalized by the observers to the investigator (NC) as microns of clearance to the nearest whole number.
The mean clinical estimates of CCC for each eye were OD 220.5 ± 121.0 microns (range 50 to 480 microns) and OS 398.0 ± 159.1 microns (range 140 to 800 microns). Outliers were found in the data for the left eye observations only. The mean clinical estimates of CCC after removal of outliers were OS 355.5 ± 99.47 microns (range 140 to 600 microns). The mean CCC by AS-OCT was OD 105.5 ± 11.7 microns (range 84 to 121 microns) and OS 340.8 ± 15.2 microns (range 315 to 362 microns).
A comparison of the median clinical observations to the calibrated measurements of CCC was significantly different on Mann-Whitney test, median OD = 200 to 108.5 microns (177.0; p = 0.001) and OS = 375.0 to 342.5 microns (260.0; p = 0.012) (Fig. 1). In the comparison of median values of clinical observations to AS-OCT after removal of outliers, OS = 370.0 to 342.5 microns (260.0; p = 0.047) (Fig. 2).
Results demonstrate a statistically significant difference between clinically observed estimates of CCC and those obtained by AS-OCT during a calibration procedure. The results were consistent even after outliers were removed for the left eye but of lower statistical significance. Additionally, neophyte clinical observer estimates demonstrated a greater statistically significant difference in the right eye with the shallower vault. The difference in means between observed and AS-OCT values were 115.0 microns (OD), 49.2 microns (OS), and 14.7 microns (OS) after removal of outliers, with an overall trend to overestimation. This is in contrast to previous findings of “an overall trend to underestimation by approximately 50 microns.”10 The ability of this participant to successfully wear the lenses beyond the days of data collection was not evaluated. Therefore, it is unclear whether or not this amount of CCC difference between the two eyes was clinically significant outside this period. Notwithstanding, it may be prudent for neophyte fitters to confirm CCC when equivocal and when using designs whose fitting guides advocate shallower clearances.15
The small sample size (n = 2 eyes) prevents validation of the use of lens center thickness as a biometric ruler in comparison to AS-OCT. Future studies may wish to obtain data from a larger cohort of fitters across a variety of CCC to perform a Bland-Altman analysis to assess agreement. Yeung and Sorbara attempted to assess observer estimates and variability in estimating CCC in a small study (n = 45) of participants with varying degrees of experience as defined by previous number of lenses fit, Novice = 0 to 5, Intermediate = 5 to 25, and Advanced = more than 25.10 They concluded there was a trend towards underestimation of approximately 50 microns regardless of previous experience with the two groups with the most experience showing less interobserver variability.10 Participants were asked to make estimates of CCC by viewing two-dimensional digital images rather than by actual dynamic, three-dimensional assessment as was used in this study, preventing direct comparisons.
A larger study (n = 156) by Dinardo et al asked participants to estimate vault clearances of four of scleral lenses using static, two-dimensional images before and after training with a reference image.11 The authors conclude the use of such an image significantly improved participant accuracy and confidence but note that regardless of level of experience, years of practice, or residency training, overall accuracy was 20 ± 8%.11 Again, this study employed static image comparisons rather than a dynamic measure as in this study.
The presence of three outliers in the observational data of the left eye may originate from errors induced by disorientation among the various interfaces. Given their low frequency, this is likely due to observer inexperience rather than testing technique. Removal of these outliers did not change the direction or statistical significance of the results.
This study compared observer optical estimates of apparent CCC against AS-OCT measurements as the standard. Comparing non-orthogonal optical measurements (even through a reference point on the lens) to axial AS-OCT has been shown to introduce displacement artifacts related to refractive index of the contact lens when measuring the interaction of contact lens edge with the conjunctiva.26 Fourier domain devices such as the one utilized in this study have the advantage over time domain instruments by reducing distortions introduced by object movement and dispersion.27,28 Studies have proposed methods to compensate for these distortions for edge-conjunctiva interactions26 and corneal pachymetry29 but not for CCC.
Potential sources of bias in this study include observer selection bias and expectation bias.
The inclusion criteria required a single participant be selected with a normal corneal topography to avoid introducing confounding variables associated with the varying etiologies associated with irregular corneas. This may contribute to a selection bias which limits the generalizability to both regular and irregular cornea populations. Additional studies are needed to compare the accuracy of this method for each group and between groups.
Defining observers as “neophytes” is somewhat arbitrary and suggests an equivalency in clinical experiences which cannot be confirmed. Selecting observers from the final semester of their fourth year of training may mitigate differences in experience related to biomicroscopy skills but not clinical experience. The relative number of clinical encounters each observer had with scleral lenses was generally low but anecdotally varied. Two previous studies did perform a subgroup analysis which determined years of practice, residency experience, or number of scleral lenses fit did not appear to be associated with observer accuracy when comparisons were made to standard image only reductions in interobserver variability.10,11 The classification of scleral lenses used in each was not specified.
Observers were also asked not to discuss their individual experiences with the rest of the cohort participating in the study. An indeterminate number of observers did violate the protocol in this way. The previously cited studies could not separate out training effects from other variables10,11 and any such effects may be limited to the near term.11
This study provides some insights into the magnitude and directionality of error inexperienced clinicians may encounter when estimating CCC. Additional studies are required to validate these findings in larger cohorts of regular and irregular corneas to accurately gauge the impact on patient care.
Daniel G. Fuller
The Eye Center
Southern College of Optometry
1225 Madison Avenue
Memphis TN 38104
Received May 28, 2015; accepted October 6, 2015.
SUPPLEMENTAL DIGITAL CONTENT
Supplemental digital content 1, a reference image labeled to identify different zonal layers the observers would encounter during the observation and to assist in orientation, is available at http://links.lww.com/OPX/A231.
1. Nichols JJ. 2014 annual report: contact lenses 2014. Contact Lens Spectr 2015; 30(1): 22–7.
2. Morgan PB, Woods CA, Tranoudis GI, Helland M, Efron N, Jones L, Davila-Garcia E, Magnelli P, Teufl M, Grupcheva CN, et al. International contact lens prescribing in 2014. Contact Lens Spectr 2015; 30(1): 28–33.
3. Schornack MM. Scleral lenses: a literature review. Eye Contact Lens 2015; 41: 3–11.
4. Jacobs DS. Update on scleral lenses. Curr Opin Ophthalmol 2008; 19: 298–301.
5. van der Worp E, Bornman D, Ferreira DL, Faria-Ribeiro M, Garcia-Porta N, González-Meijome JM. Modern scleral contact lenses: a review. Cont Lens Anterior Eye 2014; 37: 240–50.
6. Sonsino J, Mathe DS. Central vault in dry eye patients successfully wearing scleral lens. Optom Vis Sci 2013; 90: e248–51.
7. Michaud L, van der Worp E, Brazeau D, Warde R, Giasson CJ. Predicting estimates of oxygen transmissibility for scleral lenses. Cont Lens Anterior Eye 2012; 35: 266–71.
8. Compañ V, Oliveira C, Aguilella-Arzo M, Mollá S, Peixoto-de-Matos SC, González-Méijome JM. Oxygen diffusion and edema with modern scleral rigid gas permeable contact lenses. Invest Ophthalmol Vis Sci 2014; 55: 6421–9.
9. Jaynes JM, Edrington TB, Weissman BA. Predicting scleral GP lens entrapped tear layer oxygen tensions. Cont Lens Anterior Eye 2015; 38: 44–7.
11. Dinardo A, Lotoczky J, Rosen C, Norman CW, Buckingham R, Spalding A, Middleton A. Introducing a scleral lens fit assessment guide based on tear layer thickness. Global Specialty Lens Symposium, Las Vegas, Nevada, January 23–26, 2014. Available at: http://www.pentavisionevents.com/ckfinder/userfiles/files/Dinardo%20-%20%20Introducing%20a%20Scleral%20Lens%20Fit%20Assessment%20Guide.pdf
. Accessed May 15, 2015.
12. Caroline P, Andre M. Scleral lens settling. Contact Lens Spectr 2012; 27(5): 56.
13. Kauffman MJ, Gilmartin CA, Bennett ES, Bassi CJ. A comparison of the short-term settling of three scleral lens designs. Optom Vis Sci 2014; 91: 1462–6.
15. Sorbara L, Maram J, Fonn D, Woods C, Simpson T. Metrics of the normal cornea: anterior segment imaging with the Visante OCT. Clin Exp Optom 2010; 93: 150–6.
16. Viestenz A, Vogt S, Langenbucher A, Walter S, Behrens-Baumann W. [Biometry of the anterior segment using optical coherence tomography: evaluation of different devices and analysis programs]. Ophthalmologe 2009; 106: 723–8.
17. Sandler SF, Zelefsky JR, Dorairaj S, Arthur SN, Ritch R, Liebmann JM. Intra-observer and inter-observer reliability and reproducibility of slit-lamp-adapted optical coherence tomography for evaluation of anterior chamber depth and central corneal thickness. Ophthalmic Surg Lasers Imaging 2008; 39: 299–303.
18. Gemoules G. A novel method of fitting scleral lenses using high resolution optical coherence tomography. Eye Contact Lens 2008; 34: 80–3.
19. DeNayer GW. The scleral lens vault estimating central corneal clearance. Contact Lens Spectr 2014; 29(9): 46.
21. Maeda N, Fujikado T, Kuroda T, Mihashi T, Hirohara Y, Nishida K, Watanabe H, Tano Y. Wavefront aberrations measured with Hartmann-Shack sensor in patients with keratoconus. Ophthalmology 2002; 109: 1996–2003.
22. Sabesan R, Johns L, Tomashevskaya O, Jacobs DS, Rosenthal P, Yoon G. Wavefront-guided scleral lens prosthetic device for keratoconus. Optom Vis Sci 2013; 90: 314–23.
23. Marsack JD, Ravikumar A, Nguyen C, Ticak A, Koenig DE, Elswick JD, Applegate RA. Wavefront-guided scleral lens correction in keratoconus. Optom Vis Sci 2014; 91: 1221–30.
24. Bergmanson JP, Ezekiel DF, van der Worp E. Scleral contact lenses and hypoxia: theory versus practice. Cont Lens Anterior Eye 2015; 38: 145–7.
26. Sorbara L, Simpson TL, Maram J, Song ES, Bizheva K, Hutchings N. Optical edge effects create conjunctival indentation thickness artefacts. Ophthalmic Physiol Opt 2015; 35: 283–92.
27. Wojtkowski M, Bajraszewski T, Targowski P, Kowalczyk A. Real-time in vivo imaging by high-speed spectral optical coherence tomography. Opt Lett 2003; 28: 1745–7.
28. Wojtkowski M, Srinivasan V, Ko T, Fujimoto J, Kowalczyk A, Duker J. Ultrahigh-resolution, high-speed, Fourier domain optical coherence tomography and methods for dispersion compensation. Opt Express 2004; 12: 2404–22.
29. Moezzi AM, Sin S, Simpson TL. Novel pachometry calibration. Optom Vis Sci 2006; 83: 366–71.