- Email: [email protected]
Baltimore Vision Screening Project
Baltimore Vision Screening Project Mark W. Preslan, MD, I Audrey Novak, MS 2 Purpose: This study estimates the prevalence of common visual disorders (amblyopia, strabismus, refractive errors) in a group of inner-city school children. In addition, the study addresses the issue of access to care for vision-screening programs, specifically for children with recognized difficulties in obtaining routine medical care. Methods: School children from an inner-city elementary school were enrolled into a prospective vision-screening program combining the identification arm (screening) and diagnostic/treatment arm (ophthalmic examination). The screening consisted of Snellen E optotypes presented at a 1O-foot test distance. Each child failing the vision screening was examined by an ophthalmologist at the school using a standard protocol. This allowed the authors to examine all children identified through the vision-screening program. Results: Six-hundred eighty children were screened during the 1993 to 1994 school year. Eleven percent (76) failed the vision screening and were examined, 68 of whom failed the ophthalmic examination. The estimated prevalence of visual morbidity was as follows: amblyopia, 3.9%; strabismus, 3.1 %; and refractive errors, 8.2%. Conclusion: Amblyopia, strabismus, and refractive errors were found in relatively high frequencies for this population sample of inner-city children. These findings underscore the necessity of comprehensive vision-screening programs that integrate followup care. Children with limited access to specialized eye care must be provided with a mechanism for obtaining these services. Ophthalmology 1996;103:105-109
Amblyopia and strabismus are common ophthalmologic disorders in children. Numerous professional groups concerned with children's healthcare have advocated vision screening as a method of identifying and treating these disorders. I - 3 DeBecker and coworkers4 outlined some of the problems with previous vision-screening studies in their report on the negative predictive value of a preschool vision-screening program. In many cases, the true visual status of children either passing or failing the vision screening is not known reliably because of inadequate or incomplete reporting of follow-up examination findings. In their study, they found that their previously reported positive predictive value of 72.2% was actually 50% when Originally received: February 14, 1995. Revision accepted: September 25, 1995. I Department of Ophthalmology, University of Maryland School of Medicine, Baltimore. 2 Maryland Society to Prevent Blindness, Baltimore. Reprint requests to Mark W. Preslan, MD, Department of Ophthalmology, N6W46, University of Maryland Medical Center, 22 S. Greene St, Baltimore, MD 21201.
a standard examination protocol with gold standard criteria were applied to children failing the screening. They concluded that meticulous follow-up evaluation was necessary to interpret screening and disease frequency accurately. The reported prevalence of various ocular disorders in children is variable. For example, the prevalence of amblyopia has been reported to be as low as 2%5 to as high as 5% in 5- and 6-year-old children surveyed by the Rand Corporation. 6 In this same survey, strabismus was diagnosed in 4% (tropia) and 6.7% (phoria) of the children tested. Possible sources of variation in prevalence estimates for amblyopia, strabismus, and refractive errors in childhood include the demographic makeup of the population sampled and the methods used to sample the population. These estimates may not take into account various factors that may influence the observed prevalence, such as age at testing, race, and factors known to increase the risk of amblyopia and strabismus (prematurity, cerebral palsy, or birth injury). Population samples that may not represent the entire population can over- or underestimate the prevalence of
105
Ophthalmology
Volume 103, Number 1, January 1996
vision disorders. Little is reported or known about specific subgroups of the population, especially children in underserved areas. The development of programs like Head Start and the Early Periodic Screening Diagnosis and Treatment Program, which include a mandated visionscreening program, recognize the importance of surveillance for vision problems in those children. A recent study pertaining to access to care for poor children found that a higher percentage of poor children did not have a normal source of routine care when compared with children living above the poverty line.7 Medicaid coverage assisted in improving access to routine care for poor children but was less than adequate because of increasing numbers of needy children and dwindling resources. Those that fared the worst were not covered by Medicaid or private insurance. In an editorial commenting on the subject of vision screening, Lichte~ addressed the issue of access to followup care for children identified through vision-screening programs. He emphasized the importance of effective follow-up programs for all children, especially those with limited resources for healthcare, and suggested that a mechanism be developed to provide qualified eye care services through organized medicine. Access to care is an integral part of any vision-screening program. Equally important is education of parents and educators to the importance of follow-up care for students identified by a failed vision screening. The purpose of this study is to report the prevalence of childhood vision disorders in an inner-city school-aged population and to address the issue of access to care. A cooperative screening program was developed to screen an entire elementary school population and examine on site all children failing the vision screening.
Patients and Methods Steuart Hill Elementary School was chosen for this study because of its location in a socioeconomically deprived region of Baltimore. The entire school population (preschool through second grade) was screened for this study, allowing us to assess visual morbidity in young children at an age when intervention for amblyopia and strabismus might be expected to be more successful. The school administration had a history of commitment to innovative programs aimed at reducing the impact of poverty on the children's education and lifestyle. The enrollment was 680 children (329 girls, 351 boys) with the following breakdown for each grade: (1) preschool, 125; (2) kindergarten, 213; (3) first grade, 165; and (4) second grade, 177. The racial distribution for the entire school was 75% AfricanAmerican, 22% white, and 3% other. A preexisting yearly vision screening protocol was conducted at the school by the health department. Trained health technicians visited the school and conducted hearing/vision screening for all preschool and kindergarten students. In addition, newly enrolled first graders were screened. A report of the screening results was entered into the student's record, and a recommendation was made to the parent or guardian to obtain an eye exami-
106
nation. The vision-screening method used by the technicians was the Titmus binocular screener (Titmus Optical, Petersburg, VA). Children failing the screening were offered a free examination through a health departmentsponsored clinic. Records from the 1991-to-1992 and 1992-to-1993 school years were reviewed and compared with figures obtained from the screening program conducted for this study. In each of the school years surveyed, 2% to 3% of preschool and 5% to 7% of kindergarten students were referred for further evaluation. Follow-up reporting was accomplished with the use of a return report card that was to be completed by the eye care professional examining the child and mailed to the health department or school. The vision screening for this study was sponsored independently and conducted by the Maryland Society to Prevent Blindness. In most cases, the acuity-screening technique used an individual Snellen E target at a distance of 10 feet. In children unable to respond to the Snellen E, vision was tested with HOTV or Lighthouse pictures. Children currently wearing spectacles were tested with and without their correction. The response for each child was recorded as normal (pass) if the acuity in each eye was equal to 10/10. If the acuity at the initial screening was 10/15 or less in one or both eyes, the child was rescreened at a second visit using the same acuity assessment method. If the acuity was found on retest to be less than 10/15 in one or both eyes, the child was referred for complete evaluation and was classified as abnormal (fail) on the screening record. In addition to the acuity screening, a screening cover test was performed on each child by the same technicians. Technicians were trained to refer any child with a manifest deviation on near target cover test. Failure of either portion of the screening test specified that the child be examined according to the following protocol. Each child was examined by the san'le ophthalmologist (M.W.P.), masked to the results of the screening. Informed consent was obtained from a parent or guardian according to the protocol approved by the University of Maryland Institutional Review Board. First, an assessment of fixation and ocular alignment was performed, using alternate cover testing with the child fixating on an accommodative target at 33 cm. Cyclopentolate 1 % and tropicamide I %ophthalmic drops were administered to each eye. An appropriate interval of time was allowed to reach optimum cycloplegia. The cycloplegic refraction (streak retinoscopy at 66 cm) and fundus examination (indirect and direct ophthalmoscopy) were performed. The examination score was designated "fail" if any of the following criteria were met: (1) manifest strabismic deviation with or without fixation preference; (2) refractive error (spherical equivalent) greater than +4.00 diopters or less than -0.50 diopters; (3) astigmatism (+ cylinder) greater than or equal to 2.00 diopters; (4) anisometropia greater than or equal to 1.00 diopter; (5) abnormality of ocular media; or (6) abnormal fundus examination. Arrangements were made for dispensing spectacles and/or continued treatment. Disposition of the children confirmed to have an ocular abnormality requiring treatment was addressed in one of three possible ways. In a
Preslan and Novak· Baltimore Vision Screening Project few cases, the child was already under care and vision was adequately corrected. Children whose expenses were covered by Medicaid were given spectacle prescriptions, and arrangements were made for follow-up care through the health department clinic. For those without any medical coverage, a volunteer optician and the Maryland Society to Prevent Blindness provided spectacles at minimal cost, and arrangements were made to obtain follow-up care through the health department clinic.
Results The entire school population (total enrollment 680 children) was screened between September 1, 1993 and December 31, 1993. Of the 680 children, 173 failed the initial screening and underwent a rescreening. Seventy-six children failed the rescreening and were referred for ophthalmologic evaluation. These evaluations were performed at the school between October 1, 1993 and May 1, 1994. Table 1 contains a breakdown of the referrals by grade and whether the examination confirmed an ocular abnormality. The reason for referral was as follows: visual acuity less than 20/15 in both eyes, 48; visual acuity less than 10/15 in one eye only, 21; failed cover test, 21; observed external abnormality, 10 (some children were referred for multiple reasons). At the time of the screening, 22 children (3.2%) had been wearing spectacles previously prescribed by an eye care provider. Of those 22, 15 consistently were wearing the spectacles. Four of the 15 children wearing spectacles passed the screening with and without spectacles. The subsequent ophthalmic evaluation confirmed a normal examination. Five of the 15 spectacle wearers failed the vision screening with or without their spectacles. The ophthalmic examination confirmed the presence of persistent amblyopia. The remaining six children passed the vision screening only when wearing their spectacles. The seven patients who previously were prescribed spectacles but were not currently wearing them on a consistent basis all failed the screening. Of the seven children, three were accommodative esotropes, and four had refractive errors only. Table 2 shows a breakdown of abnormal findings by age group. Amblyopia was diagnosed in 27 children, for an estimated prevalence of 3.9%. The primary etiology for amblyopia was as follows: strabismus 44%, anisometropia (spherical) 26%, anisometropia (astigmatic) 18%, and bilateral ametropia 12%. Acuity in the amblyopic eye ranged from 20/40 to 20/200 at the time of examination. The deviation in all but one of the strabismic amblyopes was esotropia, with the remaining child having an exotropic deviation. The refractive error in the anisometropic amblyopic eye was almost equally divided between hyperopia, myopia, and astigmatism. The two children with bilateral ametropic amblyopia had spherical equivalent refractive errors of +8.00 and +8.50. Strabismus was identified in 21 children (3.1 %), all but two with esotropia. We believe that exotropia was identified infrequently because of the screening protocol con-
ducted on near targets. The addition of cover-test screening using distant targets may have identified children with intermittent exotropia. Of the esotropic children, 58% had refractive accommodative esotropia. Refractive errors requiring correction were observed frequently in this relatively young group of children. Myopia or compound myopic astigmatism (myopic spherical equivalent) was documented in 21 children (3.1 %), ranging from -0.75 to -9.00 diopters spherical equivalent (mean -2.26 diopters). Of the children with myopic refractive errors, more than half (12 of 21) were from the kindergarten or prekindergarten groups. Astigmatic refractive errors were considered visually significant if greater than 2.00 diopters in either eye. This level of astigmatism was found in 17 (2.5%) children. Anisometropia of greater than 1.0 diopter (absolute difference in spherical equivalent) was observed in 18 (2.6%) children. Hyperopic refractive errors greater than 4.00 diopters were observed in six children, two with bilateral amblyopia. The screening program correctly identified ocular abnormalities in 68 of 76 children (89.5%). The rate of false-positive results for this screening test was 10.5% (screening-test results for eight patients were identified incorrectly as abnormal). Children who passed the screening test were not examined for this study. Because we could not determine the false-negative rate (number of children with an abnormal examination that passed the screening test), we cannot comment on the validity of the screening test (sensitivity, specificity) in detecting amblyopia.
Discussion The prevalence of amblyopia and strabismus in this population sample was 3.9% and 3.1 %, respectively, which is within the range of previously published estimates for these two disorders. 6 ,9 The prevalence of either of these disorders could have been even higher if the screening protocol had been modified in the following manner. The acuity test used isolated optotypes. The crowding phenomenon is a well-recognized feature in amblyopia, in which acuity in the amblyopic eye is worse when an entire line of optotypes is presented. 6 The identification of more subtle amblyopia might have been facilitated if this was included as part of the acuity screening. The number of exotropic children identified was low. This probably would have been higher if the screening had included a distance cover test because control of intermittent exotropia usually is worse in the distance. The number of children requiring spectacles for treatment of refractive errors, amblyopia, or strabismus (57 or 8.3%) was markedly higher than the observed number of children actually wearing spectacles (15 or 2.2%) at the beginning of the school year. Some of the children lost their spectacles and could not afford to obtain a replacement. A considerable number of African-American children were found to be myopic, especially in the prekindergarten and kindergarten groups. A national survey conducted
107
Ophthalmology
Volume 103, Number 1, January 1996
Table 1. Student Enrollment by Grade* Grade
No.
Rescreen
Referral (%)
Abnormal (%)
Pre-kindergarten Kindergarten First grade Second grade
125 213 165 177
49 60 33 31
18 (14.4) 19 (10.2) 18 (9.6) 21 (11.1)
16 19 16 17
Total
680
173
76
68
• Children were rescreened if visual acuity was 10/ 15 or lower or if they failed the motility examination.
more than 20 years ago of teenagers 12 to 17 years of age reported the prevalence of myopia in African-Americans was almost half that observed in whites.1O In prevalence data from the Orinda Longitudinal Study of Myopia, myopia was observed in 7.5% of children (grades 1-5) surveyed in the predominantly white community of Orinda. I I The children in our study did not reach the peak incidence years for myopia and may reach prevalence levels for myopia similar to the Orinda Study. Perhaps the survey data from the 1970s should be revised by again looking at the prevalence of myopia in various ethnic groups. The data from our study only can suggest that this issue may need to be revisited. The finding of considerable visual morbidity in this population sample underscores the necessity for an effective vision-screening program, especially when considering the socioeconomic status of these children and their potential difficulties in accessing the healthcare system. In particular, this program integrated the identification arm (vision-screening program) with the diagnostic/treatment arm (ophthalmologic examination) in a manner that allowed excellent compliance with the screening referral recommendations. Without this type of integration, we might have found an artificially low prevalence of these disorders because of the lack of follow-up assessment. Evidence for problems in the existing screening protocol, which places the responsibility for follow-up care on the parent or guardian, can be found in the review of the 1991-to-1992 and 1992-to-1993 school year screenings. The percentage of children referred in previous years was somewhat lower than observed in this protocol. In addition, we noted that there were several children who were identified for evaluation the previous year without evidence of follow-up evaluation and/or treatment. This
should not be construed as a criticism of this school system's vision-screening protocol, but rather should alert us to the inherent problems in dealing with a large metropolitan area with limited school system resources. The purpose of exploring this issue is to emphasize that the problems of amblyopia, strabismus, and refractive errors exist in this group of children at levels equal to those previously reported in other studies. The observation that problems exist in the follow-up of children identified through vision screening is but one aspect of providing access to healthcare, where immunizations and routine health maintenance may be difficult to provide consistently for some children. 7 The data from the screening program (Table 1) illustrate an important point. The rescreening protocol was designed to reduce the number of over referrals (rescreening, n = 173, referral, n = 76) and was most beneficial in the preschool/kindergarten age groups. This is most likely due to the difficulties in using optotype vision screening in these children and emphasizes the need for retesting. Roughly one third of the original screening failures in the prekindergarten and kindergarten groups actually were referred for evaluation. A prime consideration in evaluating any screening program must be cost. In this particular investigation, all services and supplies were donated by the screeners and the examiner. Spectacles were provided for those children unable to obtain the glasses through healthcare coverage. In actual practice, if this procedure were recommended on a larger scale, some assessment of cost and time commitment would have to be undertaken. The vision screening team (Maryland Society to Prevent Blindness) regularly screens 'preschool programs on a volunteer basis. For this investigation, we estimated that more than 190
Table 2. Distribution of Abnormal Findings by Grade Grade
Amblyopia
Strabismus
Myopia
Anisometropia
Astigmatism
Pre-kindergarten Kindergarten First grade Second grade
8
4 4 7 6 21
7 5 2 7 21
5 4 4 5
2 5 6 4
18
17
Total
108
6 8
5 27
Preslan and Novak· Baltimore Vision Screening Project hours of screening time were provided. If parents could not afford spectacles, they were provided by the Maryland Society to Prevent Blindness for a nominal fee of$5. The diagnostic/treatment arm consisted of eight visits to the school at 3 to 4 hours for each visit. The total time spent on actual examination and recommendations was 30 hours. Obviously, it would be more efficient to examine these children through ophthalmologists in the community who would be comfortable with the evaluation and treatment of children. Those children unable to afford the examination and/or spectacles, because of financial status or lack of health coverage, would need an effective mechanism for obtaining follow-up care. This "pro bono" care should be distributed throughout the community so that it does not place an unfair burden on anyone institution or individual practice. Development of such a follow-up mechanism with monitoring by the school or health department might be an acceptable and more cost-effective alternative to the type of program described in this study. Parents would have to be educated about the importance of vision care and assistance provided in seeking appropriate care and treatment for their children. In summary, the prevalence of visual morbidity in a population sample of children 4 to 7 years of age is reported for amblyopia (3.9%), strabismus (3.1%), and refractive errors (8.5%). The population sample was relatively small compared with the total number of children in any given metropolitan area. Extrapolation of these findings to the entire population must be performed considering that many factors may raise or lower these statistics for any given subpopulation. Because children that passed the screening test were not examined for this study, the prevalence of any of these disorders could be even higher than what is reported (false negatives). However, these results add to the existing data on childhood visual disorders and help us to recognize the importance of effective programs addressing the identification and treatment of decreased vision in children. The ability of a child to participate in the educational experience is at least partially dependent on good vision. Amblyopia and strabis-
mus are treatable causes of decreased visual function and are unfortunate causes of permanent vision loss. Hopefully, continued emphasis, not only on identification of visual disorders, but also of diagnosis and treatment of these children so identified, will reduce the number of older children and adults with undetected and/or untreatable vision disorders.
References 1. American Academy of Ophthalmology. Infant and chil-
2. 3. 4. 5. 6. 7. 8. 9. 10. 11.
dren's vision screening. San Francisco: American Academy of Ophthalmology, 1991. American Academy of Pediatrics Committee on Practice and Ambulatory Medicine: Vision screening and eye examination in children. Pediatrics 1986;77:918-9. Eye care for the children of America: The American Association for Pediatric Ophthalmology and Strabismus. J Pediatr Ophthalmol Strabismus 1991;28:64-7. De Becker I, MacPherson HJ, LaRoche GR, et al. Negative predictive value of a population-based preschool vision screening program. Ophthalmology 1992;99:998-1003. Von Noorden GK. Binocular vision and ocular motility: theory and management of strabismus, 4th ed. St. Louis: Mosby, 1990; 280. Tychsen L. Binocular vision. In: Hart WM Jr, ed. Adler's physiology of the eye: clinical application, 9th ed. St. Louis: Mosby Year Book, 1992;837-8. St. Peter RF, Newacheck PW, Halfon N. Access to care for poor children: separate and unequal? JAMA 1992;267: 2760-4. Lichter, PRo Vision screening and children's access to eye care [editorial]. Ophthalmology 1992;99:843-4. Ehrlich MI, Reinecke RD, Simons K. Preschool vision screening for amblyopia and strabismus. Program, methods, guidelines, 1983. Surv Ophthalmol 1983;28:145-63. Sperduto RD, Seigel D, Roberts J, Rowland M. Prevalence of myopia in the United States. Arch Ophthalmology 1983;101:405-7. Zadnik K, Satariano WA, Mutti DO, et al. The effect of parental history of myopia on children's eye size. JAMA 1994;271:1323-7.
109
Amblyopia and strabismus are common ophthalmologic disorders in children. Numerous professional groups concerned with children's healthcare have advocated vision screening as a method of identifying and treating these disorders. I - 3 DeBecker and coworkers4 outlined some of the problems with previous vision-screening studies in their report on the negative predictive value of a preschool vision-screening program. In many cases, the true visual status of children either passing or failing the vision screening is not known reliably because of inadequate or incomplete reporting of follow-up examination findings. In their study, they found that their previously reported positive predictive value of 72.2% was actually 50% when Originally received: February 14, 1995. Revision accepted: September 25, 1995. I Department of Ophthalmology, University of Maryland School of Medicine, Baltimore. 2 Maryland Society to Prevent Blindness, Baltimore. Reprint requests to Mark W. Preslan, MD, Department of Ophthalmology, N6W46, University of Maryland Medical Center, 22 S. Greene St, Baltimore, MD 21201.
a standard examination protocol with gold standard criteria were applied to children failing the screening. They concluded that meticulous follow-up evaluation was necessary to interpret screening and disease frequency accurately. The reported prevalence of various ocular disorders in children is variable. For example, the prevalence of amblyopia has been reported to be as low as 2%5 to as high as 5% in 5- and 6-year-old children surveyed by the Rand Corporation. 6 In this same survey, strabismus was diagnosed in 4% (tropia) and 6.7% (phoria) of the children tested. Possible sources of variation in prevalence estimates for amblyopia, strabismus, and refractive errors in childhood include the demographic makeup of the population sampled and the methods used to sample the population. These estimates may not take into account various factors that may influence the observed prevalence, such as age at testing, race, and factors known to increase the risk of amblyopia and strabismus (prematurity, cerebral palsy, or birth injury). Population samples that may not represent the entire population can over- or underestimate the prevalence of
105
Ophthalmology
Volume 103, Number 1, January 1996
vision disorders. Little is reported or known about specific subgroups of the population, especially children in underserved areas. The development of programs like Head Start and the Early Periodic Screening Diagnosis and Treatment Program, which include a mandated visionscreening program, recognize the importance of surveillance for vision problems in those children. A recent study pertaining to access to care for poor children found that a higher percentage of poor children did not have a normal source of routine care when compared with children living above the poverty line.7 Medicaid coverage assisted in improving access to routine care for poor children but was less than adequate because of increasing numbers of needy children and dwindling resources. Those that fared the worst were not covered by Medicaid or private insurance. In an editorial commenting on the subject of vision screening, Lichte~ addressed the issue of access to followup care for children identified through vision-screening programs. He emphasized the importance of effective follow-up programs for all children, especially those with limited resources for healthcare, and suggested that a mechanism be developed to provide qualified eye care services through organized medicine. Access to care is an integral part of any vision-screening program. Equally important is education of parents and educators to the importance of follow-up care for students identified by a failed vision screening. The purpose of this study is to report the prevalence of childhood vision disorders in an inner-city school-aged population and to address the issue of access to care. A cooperative screening program was developed to screen an entire elementary school population and examine on site all children failing the vision screening.
Patients and Methods Steuart Hill Elementary School was chosen for this study because of its location in a socioeconomically deprived region of Baltimore. The entire school population (preschool through second grade) was screened for this study, allowing us to assess visual morbidity in young children at an age when intervention for amblyopia and strabismus might be expected to be more successful. The school administration had a history of commitment to innovative programs aimed at reducing the impact of poverty on the children's education and lifestyle. The enrollment was 680 children (329 girls, 351 boys) with the following breakdown for each grade: (1) preschool, 125; (2) kindergarten, 213; (3) first grade, 165; and (4) second grade, 177. The racial distribution for the entire school was 75% AfricanAmerican, 22% white, and 3% other. A preexisting yearly vision screening protocol was conducted at the school by the health department. Trained health technicians visited the school and conducted hearing/vision screening for all preschool and kindergarten students. In addition, newly enrolled first graders were screened. A report of the screening results was entered into the student's record, and a recommendation was made to the parent or guardian to obtain an eye exami-
106
nation. The vision-screening method used by the technicians was the Titmus binocular screener (Titmus Optical, Petersburg, VA). Children failing the screening were offered a free examination through a health departmentsponsored clinic. Records from the 1991-to-1992 and 1992-to-1993 school years were reviewed and compared with figures obtained from the screening program conducted for this study. In each of the school years surveyed, 2% to 3% of preschool and 5% to 7% of kindergarten students were referred for further evaluation. Follow-up reporting was accomplished with the use of a return report card that was to be completed by the eye care professional examining the child and mailed to the health department or school. The vision screening for this study was sponsored independently and conducted by the Maryland Society to Prevent Blindness. In most cases, the acuity-screening technique used an individual Snellen E target at a distance of 10 feet. In children unable to respond to the Snellen E, vision was tested with HOTV or Lighthouse pictures. Children currently wearing spectacles were tested with and without their correction. The response for each child was recorded as normal (pass) if the acuity in each eye was equal to 10/10. If the acuity at the initial screening was 10/15 or less in one or both eyes, the child was rescreened at a second visit using the same acuity assessment method. If the acuity was found on retest to be less than 10/15 in one or both eyes, the child was referred for complete evaluation and was classified as abnormal (fail) on the screening record. In addition to the acuity screening, a screening cover test was performed on each child by the same technicians. Technicians were trained to refer any child with a manifest deviation on near target cover test. Failure of either portion of the screening test specified that the child be examined according to the following protocol. Each child was examined by the san'le ophthalmologist (M.W.P.), masked to the results of the screening. Informed consent was obtained from a parent or guardian according to the protocol approved by the University of Maryland Institutional Review Board. First, an assessment of fixation and ocular alignment was performed, using alternate cover testing with the child fixating on an accommodative target at 33 cm. Cyclopentolate 1 % and tropicamide I %ophthalmic drops were administered to each eye. An appropriate interval of time was allowed to reach optimum cycloplegia. The cycloplegic refraction (streak retinoscopy at 66 cm) and fundus examination (indirect and direct ophthalmoscopy) were performed. The examination score was designated "fail" if any of the following criteria were met: (1) manifest strabismic deviation with or without fixation preference; (2) refractive error (spherical equivalent) greater than +4.00 diopters or less than -0.50 diopters; (3) astigmatism (+ cylinder) greater than or equal to 2.00 diopters; (4) anisometropia greater than or equal to 1.00 diopter; (5) abnormality of ocular media; or (6) abnormal fundus examination. Arrangements were made for dispensing spectacles and/or continued treatment. Disposition of the children confirmed to have an ocular abnormality requiring treatment was addressed in one of three possible ways. In a
Preslan and Novak· Baltimore Vision Screening Project few cases, the child was already under care and vision was adequately corrected. Children whose expenses were covered by Medicaid were given spectacle prescriptions, and arrangements were made for follow-up care through the health department clinic. For those without any medical coverage, a volunteer optician and the Maryland Society to Prevent Blindness provided spectacles at minimal cost, and arrangements were made to obtain follow-up care through the health department clinic.
Results The entire school population (total enrollment 680 children) was screened between September 1, 1993 and December 31, 1993. Of the 680 children, 173 failed the initial screening and underwent a rescreening. Seventy-six children failed the rescreening and were referred for ophthalmologic evaluation. These evaluations were performed at the school between October 1, 1993 and May 1, 1994. Table 1 contains a breakdown of the referrals by grade and whether the examination confirmed an ocular abnormality. The reason for referral was as follows: visual acuity less than 20/15 in both eyes, 48; visual acuity less than 10/15 in one eye only, 21; failed cover test, 21; observed external abnormality, 10 (some children were referred for multiple reasons). At the time of the screening, 22 children (3.2%) had been wearing spectacles previously prescribed by an eye care provider. Of those 22, 15 consistently were wearing the spectacles. Four of the 15 children wearing spectacles passed the screening with and without spectacles. The subsequent ophthalmic evaluation confirmed a normal examination. Five of the 15 spectacle wearers failed the vision screening with or without their spectacles. The ophthalmic examination confirmed the presence of persistent amblyopia. The remaining six children passed the vision screening only when wearing their spectacles. The seven patients who previously were prescribed spectacles but were not currently wearing them on a consistent basis all failed the screening. Of the seven children, three were accommodative esotropes, and four had refractive errors only. Table 2 shows a breakdown of abnormal findings by age group. Amblyopia was diagnosed in 27 children, for an estimated prevalence of 3.9%. The primary etiology for amblyopia was as follows: strabismus 44%, anisometropia (spherical) 26%, anisometropia (astigmatic) 18%, and bilateral ametropia 12%. Acuity in the amblyopic eye ranged from 20/40 to 20/200 at the time of examination. The deviation in all but one of the strabismic amblyopes was esotropia, with the remaining child having an exotropic deviation. The refractive error in the anisometropic amblyopic eye was almost equally divided between hyperopia, myopia, and astigmatism. The two children with bilateral ametropic amblyopia had spherical equivalent refractive errors of +8.00 and +8.50. Strabismus was identified in 21 children (3.1 %), all but two with esotropia. We believe that exotropia was identified infrequently because of the screening protocol con-
ducted on near targets. The addition of cover-test screening using distant targets may have identified children with intermittent exotropia. Of the esotropic children, 58% had refractive accommodative esotropia. Refractive errors requiring correction were observed frequently in this relatively young group of children. Myopia or compound myopic astigmatism (myopic spherical equivalent) was documented in 21 children (3.1 %), ranging from -0.75 to -9.00 diopters spherical equivalent (mean -2.26 diopters). Of the children with myopic refractive errors, more than half (12 of 21) were from the kindergarten or prekindergarten groups. Astigmatic refractive errors were considered visually significant if greater than 2.00 diopters in either eye. This level of astigmatism was found in 17 (2.5%) children. Anisometropia of greater than 1.0 diopter (absolute difference in spherical equivalent) was observed in 18 (2.6%) children. Hyperopic refractive errors greater than 4.00 diopters were observed in six children, two with bilateral amblyopia. The screening program correctly identified ocular abnormalities in 68 of 76 children (89.5%). The rate of false-positive results for this screening test was 10.5% (screening-test results for eight patients were identified incorrectly as abnormal). Children who passed the screening test were not examined for this study. Because we could not determine the false-negative rate (number of children with an abnormal examination that passed the screening test), we cannot comment on the validity of the screening test (sensitivity, specificity) in detecting amblyopia.
Discussion The prevalence of amblyopia and strabismus in this population sample was 3.9% and 3.1 %, respectively, which is within the range of previously published estimates for these two disorders. 6 ,9 The prevalence of either of these disorders could have been even higher if the screening protocol had been modified in the following manner. The acuity test used isolated optotypes. The crowding phenomenon is a well-recognized feature in amblyopia, in which acuity in the amblyopic eye is worse when an entire line of optotypes is presented. 6 The identification of more subtle amblyopia might have been facilitated if this was included as part of the acuity screening. The number of exotropic children identified was low. This probably would have been higher if the screening had included a distance cover test because control of intermittent exotropia usually is worse in the distance. The number of children requiring spectacles for treatment of refractive errors, amblyopia, or strabismus (57 or 8.3%) was markedly higher than the observed number of children actually wearing spectacles (15 or 2.2%) at the beginning of the school year. Some of the children lost their spectacles and could not afford to obtain a replacement. A considerable number of African-American children were found to be myopic, especially in the prekindergarten and kindergarten groups. A national survey conducted
107
Ophthalmology
Volume 103, Number 1, January 1996
Table 1. Student Enrollment by Grade* Grade
No.
Rescreen
Referral (%)
Abnormal (%)
Pre-kindergarten Kindergarten First grade Second grade
125 213 165 177
49 60 33 31
18 (14.4) 19 (10.2) 18 (9.6) 21 (11.1)
16 19 16 17
Total
680
173
76
68
• Children were rescreened if visual acuity was 10/ 15 or lower or if they failed the motility examination.
more than 20 years ago of teenagers 12 to 17 years of age reported the prevalence of myopia in African-Americans was almost half that observed in whites.1O In prevalence data from the Orinda Longitudinal Study of Myopia, myopia was observed in 7.5% of children (grades 1-5) surveyed in the predominantly white community of Orinda. I I The children in our study did not reach the peak incidence years for myopia and may reach prevalence levels for myopia similar to the Orinda Study. Perhaps the survey data from the 1970s should be revised by again looking at the prevalence of myopia in various ethnic groups. The data from our study only can suggest that this issue may need to be revisited. The finding of considerable visual morbidity in this population sample underscores the necessity for an effective vision-screening program, especially when considering the socioeconomic status of these children and their potential difficulties in accessing the healthcare system. In particular, this program integrated the identification arm (vision-screening program) with the diagnostic/treatment arm (ophthalmologic examination) in a manner that allowed excellent compliance with the screening referral recommendations. Without this type of integration, we might have found an artificially low prevalence of these disorders because of the lack of follow-up assessment. Evidence for problems in the existing screening protocol, which places the responsibility for follow-up care on the parent or guardian, can be found in the review of the 1991-to-1992 and 1992-to-1993 school year screenings. The percentage of children referred in previous years was somewhat lower than observed in this protocol. In addition, we noted that there were several children who were identified for evaluation the previous year without evidence of follow-up evaluation and/or treatment. This
should not be construed as a criticism of this school system's vision-screening protocol, but rather should alert us to the inherent problems in dealing with a large metropolitan area with limited school system resources. The purpose of exploring this issue is to emphasize that the problems of amblyopia, strabismus, and refractive errors exist in this group of children at levels equal to those previously reported in other studies. The observation that problems exist in the follow-up of children identified through vision screening is but one aspect of providing access to healthcare, where immunizations and routine health maintenance may be difficult to provide consistently for some children. 7 The data from the screening program (Table 1) illustrate an important point. The rescreening protocol was designed to reduce the number of over referrals (rescreening, n = 173, referral, n = 76) and was most beneficial in the preschool/kindergarten age groups. This is most likely due to the difficulties in using optotype vision screening in these children and emphasizes the need for retesting. Roughly one third of the original screening failures in the prekindergarten and kindergarten groups actually were referred for evaluation. A prime consideration in evaluating any screening program must be cost. In this particular investigation, all services and supplies were donated by the screeners and the examiner. Spectacles were provided for those children unable to obtain the glasses through healthcare coverage. In actual practice, if this procedure were recommended on a larger scale, some assessment of cost and time commitment would have to be undertaken. The vision screening team (Maryland Society to Prevent Blindness) regularly screens 'preschool programs on a volunteer basis. For this investigation, we estimated that more than 190
Table 2. Distribution of Abnormal Findings by Grade Grade
Amblyopia
Strabismus
Myopia
Anisometropia
Astigmatism
Pre-kindergarten Kindergarten First grade Second grade
8
4 4 7 6 21
7 5 2 7 21
5 4 4 5
2 5 6 4
18
17
Total
108
6 8
5 27
Preslan and Novak· Baltimore Vision Screening Project hours of screening time were provided. If parents could not afford spectacles, they were provided by the Maryland Society to Prevent Blindness for a nominal fee of$5. The diagnostic/treatment arm consisted of eight visits to the school at 3 to 4 hours for each visit. The total time spent on actual examination and recommendations was 30 hours. Obviously, it would be more efficient to examine these children through ophthalmologists in the community who would be comfortable with the evaluation and treatment of children. Those children unable to afford the examination and/or spectacles, because of financial status or lack of health coverage, would need an effective mechanism for obtaining follow-up care. This "pro bono" care should be distributed throughout the community so that it does not place an unfair burden on anyone institution or individual practice. Development of such a follow-up mechanism with monitoring by the school or health department might be an acceptable and more cost-effective alternative to the type of program described in this study. Parents would have to be educated about the importance of vision care and assistance provided in seeking appropriate care and treatment for their children. In summary, the prevalence of visual morbidity in a population sample of children 4 to 7 years of age is reported for amblyopia (3.9%), strabismus (3.1%), and refractive errors (8.5%). The population sample was relatively small compared with the total number of children in any given metropolitan area. Extrapolation of these findings to the entire population must be performed considering that many factors may raise or lower these statistics for any given subpopulation. Because children that passed the screening test were not examined for this study, the prevalence of any of these disorders could be even higher than what is reported (false negatives). However, these results add to the existing data on childhood visual disorders and help us to recognize the importance of effective programs addressing the identification and treatment of decreased vision in children. The ability of a child to participate in the educational experience is at least partially dependent on good vision. Amblyopia and strabis-
mus are treatable causes of decreased visual function and are unfortunate causes of permanent vision loss. Hopefully, continued emphasis, not only on identification of visual disorders, but also of diagnosis and treatment of these children so identified, will reduce the number of older children and adults with undetected and/or untreatable vision disorders.
References 1. American Academy of Ophthalmology. Infant and chil-
2. 3. 4. 5. 6. 7. 8. 9. 10. 11.
dren's vision screening. San Francisco: American Academy of Ophthalmology, 1991. American Academy of Pediatrics Committee on Practice and Ambulatory Medicine: Vision screening and eye examination in children. Pediatrics 1986;77:918-9. Eye care for the children of America: The American Association for Pediatric Ophthalmology and Strabismus. J Pediatr Ophthalmol Strabismus 1991;28:64-7. De Becker I, MacPherson HJ, LaRoche GR, et al. Negative predictive value of a population-based preschool vision screening program. Ophthalmology 1992;99:998-1003. Von Noorden GK. Binocular vision and ocular motility: theory and management of strabismus, 4th ed. St. Louis: Mosby, 1990; 280. Tychsen L. Binocular vision. In: Hart WM Jr, ed. Adler's physiology of the eye: clinical application, 9th ed. St. Louis: Mosby Year Book, 1992;837-8. St. Peter RF, Newacheck PW, Halfon N. Access to care for poor children: separate and unequal? JAMA 1992;267: 2760-4. Lichter, PRo Vision screening and children's access to eye care [editorial]. Ophthalmology 1992;99:843-4. Ehrlich MI, Reinecke RD, Simons K. Preschool vision screening for amblyopia and strabismus. Program, methods, guidelines, 1983. Surv Ophthalmol 1983;28:145-63. Sperduto RD, Seigel D, Roberts J, Rowland M. Prevalence of myopia in the United States. Arch Ophthalmology 1983;101:405-7. Zadnik K, Satariano WA, Mutti DO, et al. The effect of parental history of myopia on children's eye size. JAMA 1994;271:1323-7.
109