From UPMC Eye Center, Pittsburgh, Pennsylvania (Ho, Harvey, Scherer, Dhaliwal, Mah); and William Beaumont Hospital, Royal Oak, Michigan (Balasubramaniam).
The authors have no financial interest in the materials presented herein.
Study concept and design (J.S., D.K.D., F.S.M.); data collection (L.Y.H., T.M.H.); analysis and interpretation of data (L.Y.H., T.M.H., M.B.); drafting of the manuscript (L.Y.H., T.M.H., J.S.); critical revision of the manuscript (L.Y.H., T.M.H., M.B., D.K.D., F.S.M.); statistical expertise (M.B.); administrative, technical, or material support (J.S.)
Correspondence: Lawrence Y. Ho, MD, Associated Retinal Consultants, 3535 W Thirteen Mile Rd, Ste 344, Royal Oak, MI 48073. E-mail: firstname.lastname@example.org
Increased pupil size under dim conditions has been suggested to potentiate postoperative glare, halo, and star-burst visual phenomena.1,2 Although the importance of above-average pupil diameter is controversial among refractive surgeons, pupil size is still routinely measured in preoperative screenings.3 It has been suggested that correct assessment of pupil diameter under dim conditions is one of the essential steps of evaluating patients considering refractive surgery, not only for providing informed consent, determining a surgical plan, and predicting patient satisfaction/outcome, but also as an important inclusion or exclusion criterion.4
Several pupillometry devices exist, which incorporate light-amplification, infrared, digital photography, video, or aberrometry-based technology, and they have been evaluated previously.5–18 Our study compared scotopic pupil measurements using two high-technology devices (Iowa and Colvard pupillometers) with low-technology technique (Rosenbaum card with a cobalt blue light and red light source) to determine 1) whether a significant difference exists among the four techniques for each individual observer after adjusting for eye color, and 2) whether a significant difference exists between the two observers for the four techniques after adjusting for eye color.
Patients and Methods
The horizontal pupil diameter in 200 eyes of 100 healthy individuals was evaluated. Adult patients with ability to consent and excellent ocular health as determined by a screening examination were included. Individuals using eye or systemic medications that could affect pupil size were excluded from the study. The Rosenbaum card (J.G. Rosenbaum, Cleveland, Ohio) using both a red and cobalt blue light source, Colvard pupillometer (Oasis Medical, Glendora, California), and Iowa pupillometer (Henry Louis Inc, Iowa City, Iowa) were used to measure horizontal pupil diameter to the nearest 0.5 mm. The use of infrared pupillometry has been described previously.19–21
Measurements were performed in a windowless examination room with the lights off and the door cracked open 1 inch. The black-background monitor for the Iowa pupillometer was located to the side of the patients for all measurements. Individuals were given approximately 30 seconds to adjust to low-light conditions and were instructed to look into the distance. Pupillometry was performed separately by two observers using all four pupillometry techniques under ambient scotopic luminance conditions (<1 lux) as measured by a Lutron LX-101 digital lux meter (Lutron Electronics Inc, Taipei, Taiwan). The observers were masked to each other’s measurements. As hippus may influence pupil size,22 observers measured the largest maintained pupil diameter in the hippus cycle for all four techniques. The observer made one measurement with each modality per eye in contradistinction to averaging multiple measurements, and iris color for each individual was categorized as blue, brown, or hazel.
Comparison pupillometry was performed using the Rosenbaum card with a technique often used in US Food and Drug Administration refractive surgery clinical trials. The card’s full-circle diameters increased in 1.0-mm increments and were verified to agree with the millimeter marks used for the other devices. The card was then held superior to the patient’s eye in the corneal plane. Each observer used the circles as a reference to measure the horizontal pupil first with a red light source and later with a blue light source. Red light was created using a halogen fiber optic transilluminator wrapped in a single layer of red electrical tape, and cobalt blue light was used from a standard direct ophthalmoscope to provide the lowest oblique illumination for measurement. The Colvard pupillometer was used as in previous studies.7,8,11,13–18
High-magnification infrared pupillometry was performed with the Iowa pupillometer, which consists of a charge coupled device camera tuned for infrared detection with three infrared side lamps for pupil illumination. Base-out prisms are placed over the camera lens so that both eyes are simultaneously displayed on the video monitor. Self-adhesive plastic rulers in 1.0-mm increments were placed horizontally over the individuals’ inferior orbital rim in the corneal plane (Fig 1). The system’s video overlay allowed movement of two vertical lines to a position tangential to the horizontal pupillary borders. The separation of the vertical lines was quantified from the adhesive rulers to provide the horizontal pupil diameter. The center of the monitor was used for measurement to avoid distortion and potential parallax-type error.
Figure 1. Technique Used for Iowa Pupillometer (Henry Louis Inc) Measurements.
Data were first analyzed using numerical and graphical techniques to assess whether they met the distribution assumptions of the statistical tests being used to analyze them. As the pupil diameter measurements (mm) of the left and right eyes for each patient for each observer using each technique are correlated with each other, the mean of these two measurements was taken to account for this association. A three-factor interaction generalized linear model was first fit to the data to determine if the variation in mean pupil diameter measurement was significant between the two observers, the four techniques, the three iris colors, each of these predictors taken two at a time (two-factor interaction terms), or by all three considered together (three-factor interaction term). Because the variation in mean pupil diameter measurements was not significantly affected by the combined (ie, interactive) effect of observers, techniques, and iris color, a simpler model was refit to these data without this term to next assess the significance of the effect of the three two-factor interaction terms on the variation in mean pupil diameter. This was iteratively performed until the simplest model was obtained, which best explained the predictors that significantly influenced the variation in mean pupil diameter measurements. Post-hoc pairwise comparisons were performed using the Bonferroni adjustment to correct for the number of pairwise comparisons performed. Unadjusted and adjusted P values <.05 were used for statistical significance. Statistical analysis was performed using PROC MIXED in the Statistical Analysis System (SAS) version 9.1.3 (SAS Institute, Cary, North Carolina).
Research was conducted in accordance with the tenets of the Declaration of Helsinki and was approved by the Institutional Review Board of the University of Pittsburgh. Informed consent was obtained from all individuals before participation in the study and information about the participants was kept confidential in accordance with HIPAA regulations.
Mean patient age was 32±9.95 years (range: 20 to 62 years). Forty-six men and 54 women were included. Table 1 lists the descriptive statistics for each observer’s pupil measurements using the four techniques for each eye color. Of the four methods, only the Colvard pupillometer averaged <6.0 mm for each observer regardless of iris color. In fact, neither observer measured a scotopic pupil >7.5 mm with the Colvard pupillometer. Figure 2 provides a graphical representation of the observers’ pooled pupil measurements for each modality separated by iris color. The graph illustrates the interactive effect of technique and observer adjusted for each eye color as well as the mean pupil diameter measurements obtained by each observer. From this graph, we can conclude that measurements of mean pupil diameters using both the Rosenbaum card with red light and Iowa pupillometer are fairly similar across each eye color between observers and the mean pupil diameters were the lowest using the Colvard pupillometer regardless of eye color for both observers. Also, observer 1 has lower mean pupil diameters than observer 2 when using the Rosenbaum card with blue light and higher mean pupil diameters than observer 2 when using the Colvard pupillometer.
Table 1: Descriptive Statistics for Pupil Measurements of Each Observer Using Each Technique Separated by Iris Color
Figure 2. Mean Pupil Measurements Stratified by Observer, Pupil Measuring Technique, and Iris Color. B = Blue Iris, BR = Brown Iris, H = Hazel Iris
The results of how the four techniques compare to one another for each individual observer as well as how each technique between the two observers compares to each other after adjusting for eye color are shown in Table 2. The information shown in this table represents the post-hoc pairwise comparisons performed using the Bonferroni adjustment to correct for the number of pairwise comparisons in the generalized linear model described above. From the first six pairwise comparisons between techniques for observer 1, it can be concluded that the mean pupil diameter obtained using a Rosenbaum card with blue light is significantly higher than that for the Colvard pupillometer after adjusting for eye color (adjusted P<.0001) whereas it is significantly lower than the measurements obtained using the Iowa pupillometer and Rosenbaum card with red light after adjusting for eye color (adjusted P<.0001 and P=.0002, respectively). Also for observer 1, the mean pupil diameter obtained using the Colvard pupillometer is significantly lower than the measurements obtained using the Iowa pupillometer and Rosenbaum card with red light after adjusting for eye color (adjusted P<.0001 and P=.0001, respectively). No significant difference is noted in the mean pupil diameters obtained by observer 1 when using the Iowa pupillometer and Rosenbaum card with red light after adjusting for eye color (adjusted P=1.0000).
Table 2: Estimated Post-hoc Mean Differences in Pupil Diameters Between and Within Observers and Techniques After Adjusting for Eye Color
From the six pairwise comparisons among techniques for observer 2 shown in Table 2, it can be concluded that the mean pupil diameter obtained using the Rosenbaum card with blue light is significantly higher than the measurements obtained using the Colvard pupillometer after adjusting for eye color (adjusted P<.0001). It can also be concluded that the mean pupil diameters obtained using the Colvard pupillometer are significantly lower than the measurements obtained using the Iowa pupillometer and Rosenbaum card with red light after adjusting for eye color (adjusted P<.0001 and P<.0001, respectively). For observer 2, there is no significant difference in the mean pupil diameter measurements between the Rosenbaum card with blue light and Iowa pupillometer or Rosenbaum card with red light after adjusting for eye color (adjusted P=.4428 and P=.7714, respectively) or between the mean pupil diameter measurements obtained using the Iowa pupillometer and Rosenbaum card with red light after adjusting for eye color (adjusted P=1.0000).
The bottom four rows in Table 2 compare the mean pupil diameter measurements within each technique between the two observers after adjusting for eye color. The two observers have similar mean pupil diameters when using the Rosenbaum card with blue or red light and the Iowa pupillometer after adjusting for eye color (adjusted P=.2574, P=1.000, and P=1.000, respectively). The mean pupil diameter values are higher for observer 1 than observer 2 when using the Colvard pupillometer after adjusting for eye color (adjusted P=.037).
Table 3 represents the comparison of the three iris colors to each other irrespective of the observer and technique interaction. The mean pupil diameters of blue irides and brown irides are significantly higher than the measurements obtained for hazel irides after adjusting for the combined effect of technique and observer (adjusted P=.0271 and P=.0445, respectively). There is no significant difference in the mean pupil diameter measurement between blue irides and brown irides after adjusting for the combined effect of observer and technique (adjusted P=1.000).
Table 3: Estimated Mean Post-hoc Pairwise Differences in Pupil Diameters by Iris Color
Our study was designed to determine whether a significant difference exists among the four techniques for each individual observer after adjusting for eye color and if a significant difference exists between the two observers for each technique after adjusting for eye color. Because there is no gold standard for subjective pupillometry, we cannot assess the accuracy of each technique but we can make conclusions regarding which techniques seem to agree with one another and how the mean pupil diameter measurements vary with each technique in relation to one another.
For both observers, the mean pupil diameter measurements made with the Rosenbaum card with red light most closely approximated those of the Iowa pupillometer across the three iris colors. The measurements from the Colvard pupillometer always measured smaller in relation to the Rosenbaum card with red light, blue light, or the Iowa pupillometer regardless of the observer when adjusted for iris color. A difference is noted in mean pupil diameter measurements when observer 1 uses the Rosenbaum card with blue light compared to the Iowa pupillometer or the Rosenbaum card with red light, but this difference does not exist for observer 2 after adjusting for iris color. This suggests that for each observer, there are some differences among the four techniques for pupil diameter measurement after adjusting for iris color.
Although the mean pupil diameter measurement obtained by observer 1 was smaller compared with observer 2 while using the Rosenbaum card with blue light, no statistical difference was noted between the observers using this technique after adjusting for eye color. The mean pupil diameter measurement obtained by observer 1 was larger compared with observer 2 while using the Colvard pupillometer and this was statistically different after adjusting for eye color, suggesting that the use of the Colvard pupillometer is more observer dependent than the other three techniques after adjusting for iris color.
Interestingly, if one looks only at the mean differences of the mean pupil diameter between iris colors irrespective of the observer and technique, there was a statistical difference between the measurements of blue and brown irides compared to hazel irides, suggesting that iris color alone can also affect the mean pupil diameter measurements.
Based on the above observations, although one can make observations specific to each technique for individual observers and between observers for each technique after adjusting for iris color, our analysis concludes that the four techniques cannot be compared to each other regardless of the observer (ie, average the measurements for both observers for each technique), nor can the two observers be compared regardless of the technique (ie, average the measurements across all four techniques). This is because the mean pupil diameter measurements depend on a complex interaction between the observer and technique as well as iris color of the patient.
The Colvard pupillometer is one of the most commonly used modalities for refractive pupillometry.23 It is a rapid, portable, and technician-friendly method of subjective pupil measurement. Neither observer measured a scotopic pupil >7.5 mm with the Colvard pupillometer, despite the young study population, whereas large pupil measurements were obtained using the other techniques. This measurement discrepancy was evident early in the study, which led the observers to verify that the Colvard reticule matched the Rosenbaum card circles and adhesive ruler hash marks used for the Iowa pupillometer. In fact, all corresponded perfectly.
Several studies have found the Colvard pupillometer to have smaller mean pupil diameters than the modalities of comparison.7,11,15 Chaglasian et al15 and Pop et al8 found little, if any, statistically significant difference between mesopic card measurements and scotopic Colvard measurements, although Pop et al8 attributed this to bias. Our data suggest that under-estimation of larger pupil sizes may occur with the Colvard pupillometer but this is dependent on the interaction of the observer and the technique, and this study did not assess which technique is more accurate as there was no gold standard. Theoretically, it may be more beneficial to have larger pupil size measurements rather than smaller measurements when considering patient selection and surgical treatment plans in refractive surgery.
This study is not without limitations. All four modalities required subjective size estimation, thereby introducing the potential for human error. Observer bias is unavoidable with only two observers and four measurement methods. Another potential study limitation may be the lack of standardization of light reaching the retina. It was evident to both observers that the blue light source required more intensity than the red light source to visualize the pupil well enough for Rosenbaum measurement. The intensity, angle, and distance of the light source were not controlled, although every attempt was made to stimulate photoreceptors minimally. In addition, the measurements were always taken in the following order: Rosenbaum card with red light, then blue light, Iowa pupillometer, and Colvard pupillometer. One potential bias may exist with patient fatigue as the testing progressed, which may affect pupil size. Another potential observer bias would be that the observer already knew the measurement he or she recorded for the first method. This could influence the reading on the second, third, and fourth measurements. Randomization of the order for the pupillometry techniques would address these two issues. Finally, dynamic pupillometry may vary from one observer to the next as pupil size is affected by both internal and external factors not limited to psychological state, age, refraction, hippus, systemic medications, and illumination during measurement.19,24
Although it is difficult to derive conclusions from comparison of this study to others, our scotopic Colvard pupillometer values were smaller than most studies published before 2006. Several older studies found a scotopic pupil of 6.0 to 6.2 mm, despite varying testing conditions.7,8,14,16 Our mean Colvard measurements fell between the values found in the work of Chaglasian et al15 and Kohnen et al11 (means of 5.1 and 5.8 mm, respectively). We postulate that device-related differences, observer error, or bias may explain some of the discrepancy in the Colvard literature. Patient fatigue must also be considered, as the Colvard was always the last modality tested in our protocol.
Whether clinical significance can be derived from a larger pupil size in contemporary refractive procedures is still not known. Complaints such as glare, halos, and other visual disturbances are well-recognized with older laser platforms in patients with large pupils.25,26 Such visual disturbances may be a greater issue with intraocular procedures such as phakic intraocular lenses or refractive lens exchange, whereby the pupil may dilate beyond the optical zone or capsulor-rhexis edge.27 Multifocal implants may also have more pupil-dependent performance limitations than manufacturers admit.28,29 Therefore, pupillometry should still be performed preoperatively for refractive surgery patients even though the best pupillometry technique has yet to be determined and the pupil diameter measurements may depend on the observer, technique, and iris color among other factors.
- Hong X, Thibos LN. Longitudinal evaluation of optical aberrations following laser in situ keratomileusis surgery. J Refract Surg. 2000;14:S647–S650.
- Haw WW, Manche EE. Effect of preoperative pupil measurements on glare, halos, and visual function after photoastigmatic refractive keratectomy. J Cataract Refract Surg. 2001;27:907–916. doi:10.1016/S0886-3350(01)00871-9 [CrossRef]
- Leaming DV. Practice styles and preferences of ASCRS members—2003 survey. J Cataract Refract Surg. 2004;30:892–900. doi:10.1016/j.jcrs.2004.02.064 [CrossRef]
- Rosen ES. Quantify. J Cataract Refract Surg. 2002;28:203–204. doi:10.1016/S0886-3350(01)01326-8 [CrossRef]
- Wachler BS, Krueger RR. Agreement and repeatability of infrared pupillometry and the comparison method. Ophthalmology. 1999;106:319–323. doi:10.1016/S0161-6420(99)90070-2 [CrossRef]
- Boxer Wachler BS, Krueger RR. Agreement and repeatability of pupillometry using videokeratography and infrared devices. J Cataract Refract Surg. 2000;26:35–40. doi:10.1016/S0886-3350(99)00331-4 [CrossRef]
- Schnitzler EM, Baumeister M, Kohnen T. Scotopic measurement of normal pupils: Colvard versus Video Vision Analyzer infrared pupillometer. J Cataract Refract Surg. 2000;26:859–866. doi:10.1016/S0886-3350(00)00486-7 [CrossRef]
- Pop M, Payette Y, Santoriello E. Comparison of the pupil card and pupillometer in measuring pupil size. J Cataract Refract Surg. 2002;28:283–288. doi:10.1016/S0886-3350(01)01222-6 [CrossRef]
- Rosen ES, Gore CL, Taylor D, Chitkara D, Howes F, Kowalewski E. Use of a digital infrared pupillometer to assess patient suitability for refractive surgery. J Cataract Refract Surg. 2002;28:1433–1438. doi:10.1016/S0886-3350(01)01350-5 [CrossRef]
- Starck T, Liu Y, Prewett AL, Curup LG. Comparison of scotopic pupil measurement with slitlamp-based cobalt blue light and infrared video-based system. J Cataract Refract Surg. 2002;28:1952–1956. doi:10.1016/S0886-3350(02)01383-4 [CrossRef]
- Kohnen T, Terzi E, Buhren J, Kohnen EM. Comparison of a digital and a handheld infrared pupillometer for determining scotopic pupil diameter. J Cataract Refract Surg. 2003;29:112–117. doi:10.1016/S0886-3350(02)01898-9 [CrossRef]
- Twa MD, Bailey MD, Hayes J, Bullimore M. Estimation of pupil size by digital photography. J Cataract Refract Surg. 2004;30:381–389. doi:10.1016/S0886-3350(03)00619-9 [CrossRef]
- Kohnen T, Terzi E, Kasper T, Kohnen EM, Buhren J. Correlation of infrared pupillometers and CCD-camera imaging from aberrometry and videokeratography for determining scotopic pupil size. J Cataract Refract Surg. 2004;30:2116–2123. doi:10.1016/j.jcrs.2004.05.009 [CrossRef]
- Colvard M. Preoperative measurement of scotopic pupil dilation using an office pupillometer. J Cataract Refract Surg. 1998;24:1594–1597.
- Chaglasian EL, Akbar S, Probst LE. Pupil measurement using the Colvard pupillometer and a standard pupil card with a cobalt blue filter penlight. J Cataract Refract Surg. 2006;32:255–260. doi:10.1016/j.jcrs.2005.08.061 [CrossRef]
- Spadea L, Giammaria D, Ferrante R, Balestrazzi E. Pre-excimer laser and post-excimer laser refractive surgery measurements of scotopic pupil diameter using 2 pupillometers. Ophthalmology. 2005;112:1003–1008. doi:10.1016/j.ophtha.2004.12.031 [CrossRef]
- Bradley JC, Anderson JE, Xu KT, Brown SM. Comparison of Colvard pupillometer and infrared digital photography for measurement of the dark-adapted pupil diameter. J Cataract Refract Surg. 2005;31:2129–2132. doi:10.1016/j.jcrs.2005.04.041 [CrossRef]
- Chaidaroon W, Juwattanasomran W. Colvard pupillometer measurement of scotopic pupil diameter in emmetropes and myopes. Jpn J Ophthalmol. 2002;46:640–644. doi:10.1016/S0021-5155(02)00556-7 [CrossRef]
- Loewenfeld IE. The Pupil; Anatomy, Physiology, and Clinical Applications. Ames, IA: Iowa State University Press; 1993.
- Lowenstein O, Loewenfeld IE. Electronic pupillography: a new instrument and some clinical applications. AMA Arch Ophthalmol. 1958;59:352–363.
- Loewenfeld IE, Rosskothen HD. Infrared pupil camera; a new method for mass screening and clinical use. Am J Ophthalmol. 1974;78:304–313.
- Kardon R. Anatomy and physiology of the autonomic nervous system. In: Miller NR, Newman NJ, eds. Walsh and Hoyt’s Clinical Neuro-Ophthalmology. 6th ed. Baltimore, MD: Williams & Wilkins; 2004:649–714.
- Duffey RJ, Leaming DV. US trends in refractive surgery: the 2006 ISRS/AAO survey. Presented at: Refractive Surgery Subspecialty Day, American Academy of Ophthalmology. ; November 11, 2006. ; Las Vegas, NV. .
- Winn B, Whitaker D, Elliott DB, Phillips NJ. Factors affecting light-adapted pupil size in normal human subjects. Invest Ophthalmol Vis Sci. 1994;35:1132–1137.
- Bailey M, Mitchell GL, Dhaliwal DK, Boxer Wachler BS, Zadnik K. Patient satisfaction and visual symptoms after laser in situ keratomileusis. Ophthalmology. 2003;110:1371–1378. doi:10.1016/S0161-6420(03)00455-X [CrossRef]
- Nixon WS. Pupil size in refractive surgery. J Cataract Refract Surg. 1997;23:1435–1436.
- Tahzib NG, Bootsma SJ, Eggink FA, Nuijts RM. Functional outcome and patient satisfaction after Artisan phakic intraocular lens implantation for the correction of myopia. Am J Ophthalmol. 2006;142:31–39. doi:10.1016/j.ajo.2006.01.088 [CrossRef]
- Kawamorita T, Uozato H. Modulation transfer function and pupil size in multifocal and monofocal intraocular lenses in vitro. J Cataract Refract Surg. 2005;31:2379–2385. doi:10.1016/j.jcrs.2005.10.024 [CrossRef]
- Hutz WW, Eckhardt HB, Rohrig B, Grolmus R. Reading ability with 3 multifocal intraocular lens models. J Cataract Refract Surg. 2006;32:2015–2021. doi:10.1016/j.jcrs.2006.08.029 [CrossRef]
Descriptive Statistics for Pupil Measurements of Each Observer Using Each Technique Separated by Iris Color
|Iris Color||Pupillometry Technique||Mean±Standard Deviation (Median [Range]) (mm)|
|Observer 1||Observer 2|
|Blue (n=34)||Rosenbaum card with red light||7.24±0.92 (7.25 [5.5–9.0])||7.21±0.84 (7.0 [5.75–9.0])|
|Rosenbaum card with blue light||6.59±0.75 (6.75 [5.0–8.0])||7.00±0.95 (6.5 [6.5–9.0])|
|Colvard handheld pupillometer||5.69±0.81 (5.75 [4.5–7.25])||5.35±0.76 (5.5 [4.0–6.75])|
|Iowa pupillometer||7.40±0.88 (7.5 [6.0–9.0])||7.28±0.93 (7.5 [5.25–8.75])|
|Brown (n=106)||Rosenbaum card with red light||7.17±0.94 (7.25 [5.0–9.5])||7.12±1.14 (7.0 [4.5–9.5])|
|Rosenbaum card with blue light||6.45±0.93 (6.5 [4.0–9.0])||6.73±1.19 (6.75 [4.5–9.5])|
|Colvard handheld pupillometer||5.80±0.8 (6.0 [3.5–7.5])||5.31±0.88 (5.5 [3.25–7.0])|
|Iowa pupillometer||7.28±0.98 (7.25 [4.5–9.5])||7.21±1.13 (7.25 [4.5–9.5])|
|Hazel (n=60)||Rosenbaum card with red light||6.84±1.17 (7.0 [4.75–8.75])||7.05±1.19 (6.88 [4.75–9.5])|
|Rosenbaum card with blue light||6.31±1.05 (6.25 [4.0–8.0])||6.83±1.05 (6.88 [4.75–9.0])|
|Colvard handheld pupillometer||5.55±1.03 (5.5 [3.5–7.25])||5.08±0.71 (5.0 [3.75–6.5])|
|Iowa pupillometer||6.86±1.12 (6.88 [4.75–8.5])||6.95±1.17 (6.88 [4.75–9.0])|
Estimated Post-hoc Mean Differences in Pupil Diameters Between and Within Observers and Techniques After Adjusting for Eye Color
|Device||Observer||Estimated Mean Difference (mm)||t Value||PValue|
|BL – Colvard||O1||0.7275||5.16||<.0001||<.0001|
|BL – Iowa||O1||−0.7425||−5.26||<.0001||<.0001|
|BL – RL||O1||−0.6475||−4.59||<.0001||.0002|
|Colvard – Iowa||O1||−1.47||−10.42||<.0001||<.0001|
|Colvard – RL||O1||−1.375||−9.74||<.0001||<.0001|
|Iowa – RL||O1||0.095||0.67||.5013||1.0000|
|BL – Colvard||O2||1.555||11.02||<.0001||<.0001|
|BL – Iowa||O2||−0.3425||−2.43||.0158||.4428|
|BL – RL||O2||−0.3125||−2.21||.0276||.7714|
|Colvard – Iowa||O2||−1.8975||−13.45||<.0001||<.0001|
|Colvard – RL||O2||−1.8675||−13.23||<.0001||<.0001|
|Iowa – RL||O2||0.03||0.21||.8318||1.0000|
|BL||O1 – O2||−0.37||−2.62||.0092||.2574|
|Colvard||O1 – O2||0.4575||3.24||.0013||.0370|
|Iowa||O1 – O2||0.03||0.21||.8318||1.0000|
|RL||O1 – O2||−0.035||−0.25||.8043||1.0000|
Estimated Mean Post-hoc Pairwise Differences in Pupil Diameters by Iris Color
|Iris Color||Estimated Mean Difference (mm)||Standard Error||Degree of Freedom||t Value||PValue|
|Blue – Brown||0.0855||0.0983||97||0.87||.3867||1.0000|
|Blue – Hazel||0.2854||0.1071||97||2.67||.009||.0271|
|Brown – Hazel||0.1999||0.0806||97||2.48||.0148||.0445|