- Email: [email protected]
A randomized controlled trial to determine the lowest effective dose for adequate mydriasis in premature infants
A randomized controlled trial to determine the lowest effective dose for adequate mydriasis in premature infants G. Vike Vicente, MD,a Monisha Bahri, MD,b Judith J. Palafoutas, BSN, RNC,c Hong Wang, MD,d,e and Nitin Mehta, MDc PURPOSE
To compare the mydriatic efficacy of different numbers of eye drops for retinal examination of premature infants.
METHODS
This pilot study enrolled preterm infants born at \32 weeks’ gestational age who were scheduled for retinopathy of prematurity screening examinations. In all study group assignments, the right (treatment reference group) eye received 3 eye drops per current neonatal intensive care unit protocol, whereas the left (test) eye was randomized at each examination to receive either: 0, 1, or 2 drops. Pupils were dilated with the use of a cyclopentolate 0.2% and phenylephrine 1% ophthalmic solution. Pupil size (in mm) was measured by a single examiner at 0 (baseline), 45, 90, and 120 minutes after instillation. Retinal examination occurred at the first opportunity to adequately visualize the peripheral retina. Comparison of the means was calculated using paired t test, t test, or analysis of variance, as appropriate. A total of 64 eye examinations were performed on 15 enrolled infants (mean gestational age, 28.7 2.6 weeks, 53% white; 40% male). No significant differences existed in mean pupil size between the 1-, 2- and 3-drop groups at baseline, 90 minutes, and 120 minutes. All 1-drop group eyes could be examined by 90 minutes. The 0-drop group maintained baseline size. Effective mydriasis was achieved in the test eye with 1 or 2 drops and was sustained to 120 minutes. Therefore, retinal examinations could be completed by 90 minutes in most infants with the use of 1 drop. A larger study is needed to determine the effect of iris color and severity of ROP on these findings. ( J AAPOS 2012;16:365-369)
RESULTS
CONCLUSIONS
C
ombination ophthalmic drops of phenylephrine 1% and cyclopentolate 0.2% have been shown to provide adequate pupillary dilation and are widely used for serial retinal examinations of premature infants undergoing screening for retinopathy of prematurity (ROP).1-4 This medication is not without untoward side effects, which may not be evident at the time of the examination5,6: feeding intolerance with delayed gastric emptying can occur 24 hours after the drops are instilled.
Author affiliations: aEye Doctors of Washington, Chevy Chase, Maryland; bDivision of Neonatology, Washington Hospital Center, Washington, DC; cDivision of Neonatology, Georgetown University Hospital, Washington, DC; dMedStar Health Research Institute, Hyattsville, Maryland; eGeorgetown and Howard Universities Center for Clinical and Translational Sciences, Washington, DC Supported by the Georgetown-Howard Universities Center for Clinical and Translational Science with an award (UL1RR031975) from the National Center for Research Resources. Submitted October 16, 2011. Revision accepted February 11, 2012. Correspondence: G. Vike Vicente, MD, Eye Doctors of Washington, 2 Wisconsin Circle, Suite 200, Chevy Chase, MD 20815 (email: [email protected]). Copyright Ó 2012 by the American Association for Pediatric Ophthalmology and Strabismus. 1091-8531/$36.00 http://dx.doi.org/10.1016/j.jaapos.2012.02.017
Journal of AAPOS
In premature infants, these may be soft signs of infection or necrotizing enterocolitis, which necessitates a sepsis workup and/or holding feedings. Despite the widespread use of cyclopentolate, great variation exists in practice and a paucity of data is available on the optimal dosing. A telephone survey of the neonatal intensive care units (NICUs) of the District of Columbia Metropolitan area revealed that although all the centers used the same eye drop preparation (cyclopentolate 0.2% and phenylephrine 1% ophthalmic solution), practices differed in the total number of drops instilled, frequency of instillation, and interval between instillation of drops and examination of eyes. The responses given were typically 2 or 3 drops per eye, but in an outpatient ophthalmology office setting, a single drop is usually deemed sufficient. With a goal of minimizing side effects by reducing the number of drops, we proposed a pilot study to help determine the dose effectiveness in preterm infants.
Methods This study was approved by the Institutional Review Board (IRB) of Georgetown University and was performed in compliance with
365
366
Vicente et al
the United States Health Insurance Portability and Accountability Act of 1996. Conditions of the IRB approval included surveillance of infant well-being and a limit of 4 observations per study examination. This prospective, randomized, controlled trial was conducted at Georgetown University Hospital’s Neonatal NICU from March 2007 to June 2008. Preterm infants scheduled to have their medically indicated routine ROP screening at postnatal age 4-6 weeks, according to the American Academy of Pediatrics/American Academy of Ophthalmology criteria,7 were enrolled after informed written consent was obtained from a parent or guardian. Infants with eye malformations were excluded from the study. The standard protocol at the Georgetown University Hospital NICU before the onset of this study was to instill 3 drops per eye of the mydriatic preparation cyclopentolate 0.2% and phenylephrine 1% ophthalmic solution, with each drop instilled at a 5-minute interval. The eye examination was accomplished when the pupil was sufficiently dilated to visualize the peripheral retina. For the purpose of the study, the infants were randomized into one of 3 groups; 0-, 1-, and 2-drop groups. In all groups, the right eye served as the treatment reference group, and the left eye served as the test eye. The treatment reference group eyes received 3 eye drops (3-drop group) per our standard protocol. The test eyes; received 0, 1, or 2 drops as per their randomization assignment to 0-, 1- or 2-drop group, respectively. The purpose of the 0-drop group was to evaluate any contralateral or crossover effect of a 3-drop instillation in the treatment reference group eye. The 2-drop group received the second drop 5 minutes after the first. The horizontal diameter of the pupils was measured with a Colvard pupillometer (OASIS Medical, Glendora, CA) at 0 minutes (baseline), and 45, 90, and 120 minutes after the instillation of the first cyclopentolate 0.2% and phenylephrine 1% ophthalmic solution. We speculated that 5 mm would be sufficiently dilated but did not specify an acceptable size before we analyzed the study findings. Effective mydriasis was defined as that pupil size where the observer was able to adequately visualize the peripheral retina and the ROP examination was able to be completed. At the end of the 120-minute study period, any eye unable to be examined because of suboptimal pupillary dilation was treated with additional drops of cyclopentolate 0.2% and phenylephrine 1% ophthalmic solution to improve mydriasis so that the ROP retinal evaluation could be completed. The number of additional drops instilled in the test eyes brought the total dose to 3 drops per eye, equal to the standard clinical protocol. A single unmasked ophthalmologist, (investigator GVV) randomized the infant, instilled eye drops per the randomization assignment, performed the pupillary measurements, determined iris color, and completed the ROP examinations. For possible further subgroup analysis, the iris color was described as blue for lightly pigmented irises and brown for darkly pigmented irises. The speculum insertions for pupillary measures were kept as brief as possible. Because most patients require serial ROP retinal examinations until the retina is mature, typically every 2 weeks, each study patient with immature retinas was randomly reassigned at the time of subsequent ROP evaluations to any of the 3 possible treatment options.
Volume 16 Number 4 / August 2012 Infants found to have threshold ROP disease while participating in the study were referred to a retina specialist for appropriate therapy and withdrawn from the study at that time. In addition, a clinically indicated ROP examination was deferred for infants who were medically unstable until they were more stable. All infants were monitored by the use of standard NICU practice, and their well-being was reviewed by the investigators for 24-hours pre- and post-study examination time, per conditions set forth by the Georgetown University IRB. The frequency of episodes of apnea and bradycardia and occurrences of feeding intolerance, as evidenced by gastric residual feedings of greater than 20% of the previous feeding, were documented.
Statistical Analysis Statistical analyses were performed with SAS version 9.1 (SAS institute, Cary, NC). Baseline characteristics of the study population are presented as means (SD), or frequency (percentages) for the whole study population and by the random groups. Analysis of variance and c2 tests were used to examine the difference in means and proportions among the random groups. In this preliminary analysis of this pilot study, we assumed the independence of multiple examinations for each participant as the result of the small sample size. A paired t test was used to compare the related two means between paired observations. T test or analysis of variance was used to compare the means between independent groups. In addition to this analysis, we performed a similar analysis including single examination for each participant. Two-tailed P values \0.05 were considered statistically significant.
Results A total of 15 infants were enrolled (mean gestational age, 28.7 2.6 weeks; white, 8 [53%]; male, 6 [40%]). The demographic characteristics of these infants by group assignment are shown in Table 1. The only statistically significant intergroup differences were race (the 1-drop group was 81.8% white) and age (the 2-drop group was about 1.5 weeks older than the 1-drop group at the time of the examination). Each infant underwent one or more ROP examinations and contributed to 36 paired observations. Four paired observations were excluded because of deviations from protocol, leaving 32 paired observations or 64 eyes available for analysis. The pupil size at baseline in all eyes across groups was similar (Figure 1). In the 0-drop group, the pupillary measurements remained at the baseline value out to 120 minutes, (P 5 0.55). The 1-drop, 2-drop, and 3-drop (treatment reference group) groups were significantly dilated over baseline, with no statistically significant difference among these groups at 45, 90, and 120 minutes, except between the 1-drop group and 3-drop groups at 45 minutes (Table 2A and 2B). The mean pupil size at time of ROP examination was 5.7 0.7 mm in the 3-drop treatment reference group. The test eyes in the 1-drop and 2-drop groups were examined at mean pupil
Journal of AAPOS
Volume 16 Number 4 / August 2012
Vicente et al
Table 1. Demographics by group
Eyes, no. GA at birth, weeks PMA, weeks Sex: male Race: white
367
Table 2A. Mean pupil size in millimeters over time
0 drops
1 drop
2 drops
3 drops (overall)
5 28.3 2.7
5 28.1 2.0
5 28.1 2.6
15 28.7 2.6
34.3 1.2 1 (20) 2 (40)
33.8 1.6 3 (60) 3 (60)
35.7 1.2 2 (40) 3 (60)
34.6 1.5 6 (40) 8 (53.3)
GA, gestational age; PMA, postmenstrual age. Participants included into the analysis of single examination for each child. Data are mean SD or n (%).
Group
Variable, min
n
Mean
SD
P valuea
0 drops
0 45 90 120 0 45 90 120 0 45 90 120 0 45 90 120
10 10 10 7 11 11 11 11 11 11 11 10 32 32 32 30
3.1 3.1 3.3 3.4 3.1 5.3 5.8 5.4 3.2 5.4 5.7 5.4 3.2 5.6 5.7 5.5
0.5 0.9 0.9 0.9 0.9 0.6 0.4 0.6 0.6 0.8 0.5 0.7 0.7 0.7 0.5 0.8
Referent 0.85 0.46 0.55 Referent \0.0001 \0.0001 \0.0001 Referent \0.0001 \0.0001 \0.0001 Referent \0.0001 \0.0001 \0.0001
1 drop
2 drops
3 drops
SD, standard deviation. Paired t test.
a
Table 2B. Effects of dosage on pupil size P value a
Baseline 45 minutes 90 minutes 120 minutes
1 vs 3 drops
2 vs 3 dropsa
1 vs 2 dropsb
0.34 0.01 0.08 0.07
0.17 0.28 0.34 0.59
0.68 0.34 0.46 0.83
Paired t test. t test.
a
FIG 1. Pupil size over time. *P \ 0.0001 vs baseline for all eye drop groups.
b
size of 5.6 0.7 mm and 5.7 0.6 mm, respectively. The results are graphically represented in Figure 1. None of the 10 test eyes in the 0-drop group dilated beyond baseline size and therefore all received 3 drops of cyclopentolate 0.2% and phenylephrine 1% ophthalmic solution at 120 minutes. In the other groups, only 1 patient required additional drops. In this infant, both the 2-drop test eye and the 3-drop treatment reference group eye required 1 additional drop at 120 minutes. The retinal examination was accomplished at 150 minutes. Of note, this patient had dark irises and stage 2, zone 2 ROP in both eyes. The retinal examination was possible when the pupil reached $5 mm. Of the 54 eyes with any number of drops instilled, 48 (88.9%) were dilated $5 mm at 45 minutes. Of the 6 eyes that were \5 mm at 45 minutes, all but 2 eyes achieved dilation .5 mm by 90 minutes. All of the 11 eyes in the 1-drop group achieved 5 mm by the 90-minute examination. Among the 3-drop treatment reference groups, there was no difference in pupillary dilation in millimeters at 45 minutes for sex (5.6 0.6 for males vs 5.6 0.8 for females, P 5 0.87); gestational age (5.6 0.6 for \28 weeks vs 5.6 0.8 for .28 weeks, P 5 0.91); or postmenstrual age (5.6 0.6 for 32-34 weeks vs 5.5 0.8 for .34-36 weeks vs 5.7 0.7 for .6 weeks, P 5 0.81).
In the planned subgroup analysis of iris color (Figure 2) of 32 3-drop treatment reference group eyes, the blue and brown irises reached their peak dilation by 45 minutes and 90 minutes, respectively. At 90 and 120 minutes there was a statistical difference between blue and brown irises (6.0 0.3 mm vs 5.5 0.6 mm, respectively, for 90 minutes [P 5 0.0044]; 5.9 0.5 mm vs 5.0 1.0 mm, respectively, for 120 minutes [P 5 0.009]), but no clinical difference was observed because both groups were able to be examined at 45 or 90 minutes. Because there were very few eyes with stage 2 or worse ROP, we are not able to detect any effect of the severity of ROP on mydriasis. Because some of these patients provided more than one examination, we performed a sensitivity analysis using a single examination for each child, and we observed a similar association pattern between pupil size and numbers of eye drops, iris color, and severity of ROP. However, most differences were not statistical significant. The lack of statistical significance is likely to be related to the small sample size (n 5 15). The review of the infants’ well-being for 24 hours preand poststudy examination time revealed that there were no clinically significant systemic complications attributable to the additional study procedures in any randomization group or subgroup of infants. No infant had to be
Journal of AAPOS
368
Vicente et al
FIG 2. Pupil size over time in 3-drop control eyes (n 5 32). A, By iris color (*P 5 0.0044, **P 5 0.0091). B, By ROP stage (*P 5 0.0108, **P 5 0.0103).
withdrawn from the study for intolerance of study procedures. Because our worst-case ROP finding was 1 infant with stage 2 ROP, who is discussed elsewhere, no infants had to be removed from the study for the purpose of treating threshold ROP.
Discussion The primary objective of this study was to compare the mydriatic efficacy of different numbers of eye drops to determine the lowest effective dose. We found that pupillary dilation can be adequately achieved with fewer than 3 drops, and for most infants, a single drop seems to be adequate. On the basis of the pupil size at the time of completed ROP examinations, a pupillary diameter of $5 mm was deemed adequate for visualization of the peripheral retina to complete a comprehensive ROP evaluation. Dosing protocols for use of these medications in newborns does not appear to be based on any evidence. Moreover, the effects of sex, race, iris color, postmenstrual age, and severity of ROP on mydriasis had not been evaluated systematically. Several studies in adults, in which the authors compared different doses of combination eye drops and single eye drops, have produced effective pupillary dilation but found that these results were confounded by age, sex, and iris color of the patients.8,9 Our findings suggest
Volume 16 Number 4 / August 2012 that in this population, pupil dilation is not affected by age and gender. As in adults, brown irises may have an effect on the pupillary dilation in newborn infants. It is hypothesized that the pigment in darker irises absorbs the dilating eye drops rather than the targeted muscle cells. Likewise eyes with more severe ROP, particularly with plus disease, may dilate less reliably because of a rigid iris. Our preliminary findings showed slower and less sustained dilation in darker eyes and perhaps those with more severe ROP at 90 and 120 minutes after instillation with no difference at baseline and 45 minutes. However, the data represent a number too small to definitively assess the influence of iris color and/or severity of ROP on dilation. Our data provide impetus for further study to confirm our findings. It can be difficult to instill eye drops in a premature infant’s eyes; eye drops can easily miss the mark in an active infant or wash out of a crying infant’s eye. Any eye drop(s) incorrectly instilled will not achieve mydriatic efficacy in that eye. Historically, the appeal of multiple drops is that it is presumed that at least 1 of the drops will be retained. For an ophthalmologist who is consulting in the NICU, inadequately dilated eyes means significant delays in performing the ROP screening examinations and thus the administration of multiple eye drops is the current norm in clinical practice. The effective mydriasis was achieved equally with 1, 2, or 3 drops. The best time for the ROP examination is during the time of sustained and optimum pupillary dilation. In our study, adequate dilation was achieved in 96% of test eyes by 90 minutes. All but 2 eyes were able to be examined by 120 minutes. Thus, adequate pupil size was maintained until at least 120 minutes, when our study observations ended. Knowledge of this window of pupillary dilation can improve efficiency of ROP examinations. It will serve to reduce or avoid the use of additional eye drops, minimizing the adverse effects of increased dosage. It will also reduce the number of reexaminations for lack of dilation. These subsequent examinations, with the repeated and prolonged insertion of eye speculum and handling of the infants, can be stressful and a cause of physiological instability. Targeting the time of the examination to the window of optimum mydriasis would result in a single examination being more likely to yield useful information regarding ophthalmic well-being without compromising the infant’s clinical stability. In a NICU contemplating further reducing the number of drops given, it may be necessary to reinforce nursing proficiency of eye drop administration to avoid ineffective administration. Differences in demographics between randomization groups were significant with regard to a large proportion of white infants in the 1-drop group over the other groups (P 5 0.026). Because iris color is highly correlated to race, this classification placed more blue eyes in the 1-drop group. Our findings that blue eyes respond more quickly and sustain dilation longer could perhaps exaggerate the
Journal of AAPOS
Volume 16 Number 4 / August 2012 effectiveness of one drop. Despite this concern, there was a statistical difference in slower response between 1-drop and 3-drop but not between 1-drop and 2-drop groups at 45 minutes (Figure 1). The limitations of this study include possible observer bias and a lack of external validation. A single pediatric ophthalmologist, who was unmasked to the group assignment, instilled the eye drops and conducted the study measurements. Sample variability was also limited by repeat examinations conducted on the same infants over time, yet a test of the sensitivity using a single examination for each infant showed similar results. Some of our statistics failed to reach significance. According to power calculations determined by our pilot data, a study to determine a statistical difference between 1, 2, and 3 drops would need to enroll at least 270 patients. We suggest that these preliminary findings may have clinical implications, particularly in nurseries in which 3 or more drops are currently used. Our goal is ultimately to reduce side effects from mydriatic drugs by using the lowest effective dose. We speculate that one carefully placed drop may be sufficient in most infants.
Journal of AAPOS
Vicente et al
369
References 1. Chew C, Rahman RA, Shafie SM, Mohamad Z. Comparison of mydriatic regimens used in screening for retinopathy of prematurity in preterm infants with dark irides. J Pediatr Ophthalmol Strabismus 2005;42:166-73. 2. Isenberg S, Everett S, Parelhoff E. A comparison of mydriatic eye drops in low-weight infants. Ophthalmology 1984;91:278-9. 3. Sindel BD, Baker MD, Maisels MJ, Weinstein J. A comparison of the pupillary and cardiovascular effects of various mydriatic agents in preterm infants. J Pediatr Ophthalmol Strabismus 1986;23:273-6. 4. Khoo BK, Koh A, Cheong P, Ho NK. Combination cyclopentolate and phenylephrine for mydriasis in premature infants with heavily pigmented irides. J Pediatr Ophthalmol Strabismus 2000;37:15-20. 5. Rush R, Rush S, Nicolau J, Chapman K, Nagvi M. Systemic manifestations in response to mydriasis and physical examination during screening for retinopathy of prematurity. Retina 2004;24:242-5. 6. Topical Mydriatics: The adverse effects of screening examinations for Retinopathy of Prematurity. Neoreviews 2003;4:163-5. 7. Section on Ophthalmology, American Academy of Pediatrics, American Academy of Ophthalmology American Association for Pediatric Ophthalmology and Strabismus. Screening examination of premature infants for retinopathy of prematurity. Pediatrics 2006;117:572-6. 8. Apt L, Henrick A. Pupillary dilatation with single eye drop mydriatic combinations. Am J Ophthalmol 1980;89:553-9. 9. Siderov J, Nurse S. The mydriatic effect of multiple doses of tropicamide. Optom Vis Sci 2005;82:955-8.
To compare the mydriatic efficacy of different numbers of eye drops for retinal examination of premature infants.
METHODS
This pilot study enrolled preterm infants born at \32 weeks’ gestational age who were scheduled for retinopathy of prematurity screening examinations. In all study group assignments, the right (treatment reference group) eye received 3 eye drops per current neonatal intensive care unit protocol, whereas the left (test) eye was randomized at each examination to receive either: 0, 1, or 2 drops. Pupils were dilated with the use of a cyclopentolate 0.2% and phenylephrine 1% ophthalmic solution. Pupil size (in mm) was measured by a single examiner at 0 (baseline), 45, 90, and 120 minutes after instillation. Retinal examination occurred at the first opportunity to adequately visualize the peripheral retina. Comparison of the means was calculated using paired t test, t test, or analysis of variance, as appropriate. A total of 64 eye examinations were performed on 15 enrolled infants (mean gestational age, 28.7 2.6 weeks, 53% white; 40% male). No significant differences existed in mean pupil size between the 1-, 2- and 3-drop groups at baseline, 90 minutes, and 120 minutes. All 1-drop group eyes could be examined by 90 minutes. The 0-drop group maintained baseline size. Effective mydriasis was achieved in the test eye with 1 or 2 drops and was sustained to 120 minutes. Therefore, retinal examinations could be completed by 90 minutes in most infants with the use of 1 drop. A larger study is needed to determine the effect of iris color and severity of ROP on these findings. ( J AAPOS 2012;16:365-369)
RESULTS
CONCLUSIONS
C
ombination ophthalmic drops of phenylephrine 1% and cyclopentolate 0.2% have been shown to provide adequate pupillary dilation and are widely used for serial retinal examinations of premature infants undergoing screening for retinopathy of prematurity (ROP).1-4 This medication is not without untoward side effects, which may not be evident at the time of the examination5,6: feeding intolerance with delayed gastric emptying can occur 24 hours after the drops are instilled.
Author affiliations: aEye Doctors of Washington, Chevy Chase, Maryland; bDivision of Neonatology, Washington Hospital Center, Washington, DC; cDivision of Neonatology, Georgetown University Hospital, Washington, DC; dMedStar Health Research Institute, Hyattsville, Maryland; eGeorgetown and Howard Universities Center for Clinical and Translational Sciences, Washington, DC Supported by the Georgetown-Howard Universities Center for Clinical and Translational Science with an award (UL1RR031975) from the National Center for Research Resources. Submitted October 16, 2011. Revision accepted February 11, 2012. Correspondence: G. Vike Vicente, MD, Eye Doctors of Washington, 2 Wisconsin Circle, Suite 200, Chevy Chase, MD 20815 (email: [email protected]). Copyright Ó 2012 by the American Association for Pediatric Ophthalmology and Strabismus. 1091-8531/$36.00 http://dx.doi.org/10.1016/j.jaapos.2012.02.017
Journal of AAPOS
In premature infants, these may be soft signs of infection or necrotizing enterocolitis, which necessitates a sepsis workup and/or holding feedings. Despite the widespread use of cyclopentolate, great variation exists in practice and a paucity of data is available on the optimal dosing. A telephone survey of the neonatal intensive care units (NICUs) of the District of Columbia Metropolitan area revealed that although all the centers used the same eye drop preparation (cyclopentolate 0.2% and phenylephrine 1% ophthalmic solution), practices differed in the total number of drops instilled, frequency of instillation, and interval between instillation of drops and examination of eyes. The responses given were typically 2 or 3 drops per eye, but in an outpatient ophthalmology office setting, a single drop is usually deemed sufficient. With a goal of minimizing side effects by reducing the number of drops, we proposed a pilot study to help determine the dose effectiveness in preterm infants.
Methods This study was approved by the Institutional Review Board (IRB) of Georgetown University and was performed in compliance with
365
366
Vicente et al
the United States Health Insurance Portability and Accountability Act of 1996. Conditions of the IRB approval included surveillance of infant well-being and a limit of 4 observations per study examination. This prospective, randomized, controlled trial was conducted at Georgetown University Hospital’s Neonatal NICU from March 2007 to June 2008. Preterm infants scheduled to have their medically indicated routine ROP screening at postnatal age 4-6 weeks, according to the American Academy of Pediatrics/American Academy of Ophthalmology criteria,7 were enrolled after informed written consent was obtained from a parent or guardian. Infants with eye malformations were excluded from the study. The standard protocol at the Georgetown University Hospital NICU before the onset of this study was to instill 3 drops per eye of the mydriatic preparation cyclopentolate 0.2% and phenylephrine 1% ophthalmic solution, with each drop instilled at a 5-minute interval. The eye examination was accomplished when the pupil was sufficiently dilated to visualize the peripheral retina. For the purpose of the study, the infants were randomized into one of 3 groups; 0-, 1-, and 2-drop groups. In all groups, the right eye served as the treatment reference group, and the left eye served as the test eye. The treatment reference group eyes received 3 eye drops (3-drop group) per our standard protocol. The test eyes; received 0, 1, or 2 drops as per their randomization assignment to 0-, 1- or 2-drop group, respectively. The purpose of the 0-drop group was to evaluate any contralateral or crossover effect of a 3-drop instillation in the treatment reference group eye. The 2-drop group received the second drop 5 minutes after the first. The horizontal diameter of the pupils was measured with a Colvard pupillometer (OASIS Medical, Glendora, CA) at 0 minutes (baseline), and 45, 90, and 120 minutes after the instillation of the first cyclopentolate 0.2% and phenylephrine 1% ophthalmic solution. We speculated that 5 mm would be sufficiently dilated but did not specify an acceptable size before we analyzed the study findings. Effective mydriasis was defined as that pupil size where the observer was able to adequately visualize the peripheral retina and the ROP examination was able to be completed. At the end of the 120-minute study period, any eye unable to be examined because of suboptimal pupillary dilation was treated with additional drops of cyclopentolate 0.2% and phenylephrine 1% ophthalmic solution to improve mydriasis so that the ROP retinal evaluation could be completed. The number of additional drops instilled in the test eyes brought the total dose to 3 drops per eye, equal to the standard clinical protocol. A single unmasked ophthalmologist, (investigator GVV) randomized the infant, instilled eye drops per the randomization assignment, performed the pupillary measurements, determined iris color, and completed the ROP examinations. For possible further subgroup analysis, the iris color was described as blue for lightly pigmented irises and brown for darkly pigmented irises. The speculum insertions for pupillary measures were kept as brief as possible. Because most patients require serial ROP retinal examinations until the retina is mature, typically every 2 weeks, each study patient with immature retinas was randomly reassigned at the time of subsequent ROP evaluations to any of the 3 possible treatment options.
Volume 16 Number 4 / August 2012 Infants found to have threshold ROP disease while participating in the study were referred to a retina specialist for appropriate therapy and withdrawn from the study at that time. In addition, a clinically indicated ROP examination was deferred for infants who were medically unstable until they were more stable. All infants were monitored by the use of standard NICU practice, and their well-being was reviewed by the investigators for 24-hours pre- and post-study examination time, per conditions set forth by the Georgetown University IRB. The frequency of episodes of apnea and bradycardia and occurrences of feeding intolerance, as evidenced by gastric residual feedings of greater than 20% of the previous feeding, were documented.
Statistical Analysis Statistical analyses were performed with SAS version 9.1 (SAS institute, Cary, NC). Baseline characteristics of the study population are presented as means (SD), or frequency (percentages) for the whole study population and by the random groups. Analysis of variance and c2 tests were used to examine the difference in means and proportions among the random groups. In this preliminary analysis of this pilot study, we assumed the independence of multiple examinations for each participant as the result of the small sample size. A paired t test was used to compare the related two means between paired observations. T test or analysis of variance was used to compare the means between independent groups. In addition to this analysis, we performed a similar analysis including single examination for each participant. Two-tailed P values \0.05 were considered statistically significant.
Results A total of 15 infants were enrolled (mean gestational age, 28.7 2.6 weeks; white, 8 [53%]; male, 6 [40%]). The demographic characteristics of these infants by group assignment are shown in Table 1. The only statistically significant intergroup differences were race (the 1-drop group was 81.8% white) and age (the 2-drop group was about 1.5 weeks older than the 1-drop group at the time of the examination). Each infant underwent one or more ROP examinations and contributed to 36 paired observations. Four paired observations were excluded because of deviations from protocol, leaving 32 paired observations or 64 eyes available for analysis. The pupil size at baseline in all eyes across groups was similar (Figure 1). In the 0-drop group, the pupillary measurements remained at the baseline value out to 120 minutes, (P 5 0.55). The 1-drop, 2-drop, and 3-drop (treatment reference group) groups were significantly dilated over baseline, with no statistically significant difference among these groups at 45, 90, and 120 minutes, except between the 1-drop group and 3-drop groups at 45 minutes (Table 2A and 2B). The mean pupil size at time of ROP examination was 5.7 0.7 mm in the 3-drop treatment reference group. The test eyes in the 1-drop and 2-drop groups were examined at mean pupil
Journal of AAPOS
Volume 16 Number 4 / August 2012
Vicente et al
Table 1. Demographics by group
Eyes, no. GA at birth, weeks PMA, weeks Sex: male Race: white
367
Table 2A. Mean pupil size in millimeters over time
0 drops
1 drop
2 drops
3 drops (overall)
5 28.3 2.7
5 28.1 2.0
5 28.1 2.6
15 28.7 2.6
34.3 1.2 1 (20) 2 (40)
33.8 1.6 3 (60) 3 (60)
35.7 1.2 2 (40) 3 (60)
34.6 1.5 6 (40) 8 (53.3)
GA, gestational age; PMA, postmenstrual age. Participants included into the analysis of single examination for each child. Data are mean SD or n (%).
Group
Variable, min
n
Mean
SD
P valuea
0 drops
0 45 90 120 0 45 90 120 0 45 90 120 0 45 90 120
10 10 10 7 11 11 11 11 11 11 11 10 32 32 32 30
3.1 3.1 3.3 3.4 3.1 5.3 5.8 5.4 3.2 5.4 5.7 5.4 3.2 5.6 5.7 5.5
0.5 0.9 0.9 0.9 0.9 0.6 0.4 0.6 0.6 0.8 0.5 0.7 0.7 0.7 0.5 0.8
Referent 0.85 0.46 0.55 Referent \0.0001 \0.0001 \0.0001 Referent \0.0001 \0.0001 \0.0001 Referent \0.0001 \0.0001 \0.0001
1 drop
2 drops
3 drops
SD, standard deviation. Paired t test.
a
Table 2B. Effects of dosage on pupil size P value a
Baseline 45 minutes 90 minutes 120 minutes
1 vs 3 drops
2 vs 3 dropsa
1 vs 2 dropsb
0.34 0.01 0.08 0.07
0.17 0.28 0.34 0.59
0.68 0.34 0.46 0.83
Paired t test. t test.
a
FIG 1. Pupil size over time. *P \ 0.0001 vs baseline for all eye drop groups.
b
size of 5.6 0.7 mm and 5.7 0.6 mm, respectively. The results are graphically represented in Figure 1. None of the 10 test eyes in the 0-drop group dilated beyond baseline size and therefore all received 3 drops of cyclopentolate 0.2% and phenylephrine 1% ophthalmic solution at 120 minutes. In the other groups, only 1 patient required additional drops. In this infant, both the 2-drop test eye and the 3-drop treatment reference group eye required 1 additional drop at 120 minutes. The retinal examination was accomplished at 150 minutes. Of note, this patient had dark irises and stage 2, zone 2 ROP in both eyes. The retinal examination was possible when the pupil reached $5 mm. Of the 54 eyes with any number of drops instilled, 48 (88.9%) were dilated $5 mm at 45 minutes. Of the 6 eyes that were \5 mm at 45 minutes, all but 2 eyes achieved dilation .5 mm by 90 minutes. All of the 11 eyes in the 1-drop group achieved 5 mm by the 90-minute examination. Among the 3-drop treatment reference groups, there was no difference in pupillary dilation in millimeters at 45 minutes for sex (5.6 0.6 for males vs 5.6 0.8 for females, P 5 0.87); gestational age (5.6 0.6 for \28 weeks vs 5.6 0.8 for .28 weeks, P 5 0.91); or postmenstrual age (5.6 0.6 for 32-34 weeks vs 5.5 0.8 for .34-36 weeks vs 5.7 0.7 for .6 weeks, P 5 0.81).
In the planned subgroup analysis of iris color (Figure 2) of 32 3-drop treatment reference group eyes, the blue and brown irises reached their peak dilation by 45 minutes and 90 minutes, respectively. At 90 and 120 minutes there was a statistical difference between blue and brown irises (6.0 0.3 mm vs 5.5 0.6 mm, respectively, for 90 minutes [P 5 0.0044]; 5.9 0.5 mm vs 5.0 1.0 mm, respectively, for 120 minutes [P 5 0.009]), but no clinical difference was observed because both groups were able to be examined at 45 or 90 minutes. Because there were very few eyes with stage 2 or worse ROP, we are not able to detect any effect of the severity of ROP on mydriasis. Because some of these patients provided more than one examination, we performed a sensitivity analysis using a single examination for each child, and we observed a similar association pattern between pupil size and numbers of eye drops, iris color, and severity of ROP. However, most differences were not statistical significant. The lack of statistical significance is likely to be related to the small sample size (n 5 15). The review of the infants’ well-being for 24 hours preand poststudy examination time revealed that there were no clinically significant systemic complications attributable to the additional study procedures in any randomization group or subgroup of infants. No infant had to be
Journal of AAPOS
368
Vicente et al
FIG 2. Pupil size over time in 3-drop control eyes (n 5 32). A, By iris color (*P 5 0.0044, **P 5 0.0091). B, By ROP stage (*P 5 0.0108, **P 5 0.0103).
withdrawn from the study for intolerance of study procedures. Because our worst-case ROP finding was 1 infant with stage 2 ROP, who is discussed elsewhere, no infants had to be removed from the study for the purpose of treating threshold ROP.
Discussion The primary objective of this study was to compare the mydriatic efficacy of different numbers of eye drops to determine the lowest effective dose. We found that pupillary dilation can be adequately achieved with fewer than 3 drops, and for most infants, a single drop seems to be adequate. On the basis of the pupil size at the time of completed ROP examinations, a pupillary diameter of $5 mm was deemed adequate for visualization of the peripheral retina to complete a comprehensive ROP evaluation. Dosing protocols for use of these medications in newborns does not appear to be based on any evidence. Moreover, the effects of sex, race, iris color, postmenstrual age, and severity of ROP on mydriasis had not been evaluated systematically. Several studies in adults, in which the authors compared different doses of combination eye drops and single eye drops, have produced effective pupillary dilation but found that these results were confounded by age, sex, and iris color of the patients.8,9 Our findings suggest
Volume 16 Number 4 / August 2012 that in this population, pupil dilation is not affected by age and gender. As in adults, brown irises may have an effect on the pupillary dilation in newborn infants. It is hypothesized that the pigment in darker irises absorbs the dilating eye drops rather than the targeted muscle cells. Likewise eyes with more severe ROP, particularly with plus disease, may dilate less reliably because of a rigid iris. Our preliminary findings showed slower and less sustained dilation in darker eyes and perhaps those with more severe ROP at 90 and 120 minutes after instillation with no difference at baseline and 45 minutes. However, the data represent a number too small to definitively assess the influence of iris color and/or severity of ROP on dilation. Our data provide impetus for further study to confirm our findings. It can be difficult to instill eye drops in a premature infant’s eyes; eye drops can easily miss the mark in an active infant or wash out of a crying infant’s eye. Any eye drop(s) incorrectly instilled will not achieve mydriatic efficacy in that eye. Historically, the appeal of multiple drops is that it is presumed that at least 1 of the drops will be retained. For an ophthalmologist who is consulting in the NICU, inadequately dilated eyes means significant delays in performing the ROP screening examinations and thus the administration of multiple eye drops is the current norm in clinical practice. The effective mydriasis was achieved equally with 1, 2, or 3 drops. The best time for the ROP examination is during the time of sustained and optimum pupillary dilation. In our study, adequate dilation was achieved in 96% of test eyes by 90 minutes. All but 2 eyes were able to be examined by 120 minutes. Thus, adequate pupil size was maintained until at least 120 minutes, when our study observations ended. Knowledge of this window of pupillary dilation can improve efficiency of ROP examinations. It will serve to reduce or avoid the use of additional eye drops, minimizing the adverse effects of increased dosage. It will also reduce the number of reexaminations for lack of dilation. These subsequent examinations, with the repeated and prolonged insertion of eye speculum and handling of the infants, can be stressful and a cause of physiological instability. Targeting the time of the examination to the window of optimum mydriasis would result in a single examination being more likely to yield useful information regarding ophthalmic well-being without compromising the infant’s clinical stability. In a NICU contemplating further reducing the number of drops given, it may be necessary to reinforce nursing proficiency of eye drop administration to avoid ineffective administration. Differences in demographics between randomization groups were significant with regard to a large proportion of white infants in the 1-drop group over the other groups (P 5 0.026). Because iris color is highly correlated to race, this classification placed more blue eyes in the 1-drop group. Our findings that blue eyes respond more quickly and sustain dilation longer could perhaps exaggerate the
Journal of AAPOS
Volume 16 Number 4 / August 2012 effectiveness of one drop. Despite this concern, there was a statistical difference in slower response between 1-drop and 3-drop but not between 1-drop and 2-drop groups at 45 minutes (Figure 1). The limitations of this study include possible observer bias and a lack of external validation. A single pediatric ophthalmologist, who was unmasked to the group assignment, instilled the eye drops and conducted the study measurements. Sample variability was also limited by repeat examinations conducted on the same infants over time, yet a test of the sensitivity using a single examination for each infant showed similar results. Some of our statistics failed to reach significance. According to power calculations determined by our pilot data, a study to determine a statistical difference between 1, 2, and 3 drops would need to enroll at least 270 patients. We suggest that these preliminary findings may have clinical implications, particularly in nurseries in which 3 or more drops are currently used. Our goal is ultimately to reduce side effects from mydriatic drugs by using the lowest effective dose. We speculate that one carefully placed drop may be sufficient in most infants.
Journal of AAPOS
Vicente et al
369
References 1. Chew C, Rahman RA, Shafie SM, Mohamad Z. Comparison of mydriatic regimens used in screening for retinopathy of prematurity in preterm infants with dark irides. J Pediatr Ophthalmol Strabismus 2005;42:166-73. 2. Isenberg S, Everett S, Parelhoff E. A comparison of mydriatic eye drops in low-weight infants. Ophthalmology 1984;91:278-9. 3. Sindel BD, Baker MD, Maisels MJ, Weinstein J. A comparison of the pupillary and cardiovascular effects of various mydriatic agents in preterm infants. J Pediatr Ophthalmol Strabismus 1986;23:273-6. 4. Khoo BK, Koh A, Cheong P, Ho NK. Combination cyclopentolate and phenylephrine for mydriasis in premature infants with heavily pigmented irides. J Pediatr Ophthalmol Strabismus 2000;37:15-20. 5. Rush R, Rush S, Nicolau J, Chapman K, Nagvi M. Systemic manifestations in response to mydriasis and physical examination during screening for retinopathy of prematurity. Retina 2004;24:242-5. 6. Topical Mydriatics: The adverse effects of screening examinations for Retinopathy of Prematurity. Neoreviews 2003;4:163-5. 7. Section on Ophthalmology, American Academy of Pediatrics, American Academy of Ophthalmology American Association for Pediatric Ophthalmology and Strabismus. Screening examination of premature infants for retinopathy of prematurity. Pediatrics 2006;117:572-6. 8. Apt L, Henrick A. Pupillary dilatation with single eye drop mydriatic combinations. Am J Ophthalmol 1980;89:553-9. 9. Siderov J, Nurse S. The mydriatic effect of multiple doses of tropicamide. Optom Vis Sci 2005;82:955-8.