UBC Faculty Research and Publications

Vision screening for strabismus and amblyopia : a critical appraisal of the evidence Bassett, K.; Forbes, I. Apr 30, 1995

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52383-Bassett_K_et_al_Vision_screening.pdf [ 12.32MB ]
JSON: 52383-1.0045263.json
JSON-LD: 52383-1.0045263-ld.json
RDF/XML (Pretty): 52383-1.0045263-rdf.xml
RDF/JSON: 52383-1.0045263-rdf.json
Turtle: 52383-1.0045263-turtle.txt
N-Triples: 52383-1.0045263-rdf-ntriples.txt
Original Record: 52383-1.0045263-source.json
Full Text

Full Text

THE UNIVERSITY OF BRITISH COLUMBIAB.C. Office of Health Technology AssessmentDiscussion Paper Series\\\~- \ \\\\\\,APRIL 1995A CRITICAL APPRAISALOF THE EVIDENCEVISION SCREENING FORSTRABISMUS AND AMBLYOPIA:BCOHTA 95:10Centre for Health Servicesand Policy ResearchVision Screening for Strabismus and Amblyopia:A Critical Appraisal of the EvidenceBCOHTA 95:1DB.C. Office of Health Technology AssessmentCentre for Health Services & Policy ResearchS184 Koerner Pavilion2211 Wesbrook MallVancouver, B.C., CanadaV6T 1Z3K. Bassett, MD, PhDI. Forbes, MD, MHScApril 1995ACKNOWLEDGMENTSThe contributions of the following individuals are gratefully acknowledged:Internal review committee:Samuel B. Sheps, MD, MSc, FRCPCArminee Kazanjian, Dr. Soc.Kathryn D. Friesen, MScExternal Reviewers:Dr. Ross A. Kennedy, MD, CM, FRCS(C)Dr. I.B. Pless, MD, FRCP(C)Head, Department of Health Care andEpidemiologyFaculty of MedicineThe University of British ColumbiaChair, BCOHTASteering CommitteeAssociateDirector, Centrefor Health Servicesand PolicyResearchAssociate Professor, Department of Health Careand EpidemiologyUniversity of British ColumbiaProgramManager, BCOHfAUniversity of British ColumbiaClinical AssistantProfessorDept. of OphthalmologyFaculty of MedicineUniversity of British ColumbiaNationalHealth ScientistProfessorof Pediatrics & EpidemiologyMcGill University11TABLE OF CONTENTSACKNOWLEDGEMENTS ., iiFOREWORD iii•TABLE OF CONTENTS ivLIST OF TABLES : viEXEClJTIVE SUMMARy 1INTRODUCTION 41. BACKGROUND AND SIGNIFICANC~ 51.1 Amblyopia 61.1.1 Description and terminology 61.1.2 Natural history 61.1.3 Epidemiology 61.2 Strabismus 81.2.1 Description and terminology 81.2.2 Natural history 81.2.3 Epidemiology 81.3 Treatment of strabismus and amblyopia 91.3.1 Amblyopiatreatment 91.3.2 Strabismus treatment 101.4 Screening recommendations from various organizations 101.5 Current practice in British Columbia 122. SCREENING TESTSFOR AMBLYOPIA AND STRABISMUS 122.1 Tests for amblyopia 122.2 Tests for strabismus 132.3 Depth perception (stereopsis) tests 132.4 Screening for precursors 142.5 The gold standard ophthalmological examination 14IVLIST OF TABLESTable 1 Reconunended vision screening by pediatricians and family .physicians 11Table 2a Summary of research evaluating vision screening (as a clinical ..intervention): Israel 17Table 2b Summary of research evaluating vision screening (as a clinical .intervention): U.K 18Table 3a Summary of research evaluating vision screening as a diagnostic .procedure: program assessment 20Table 3b Summary of research evaluating vision screening as a diagnostic .procedure: tester assessment 20Table 4 Results of critical appraisal of studies evaluating vision ..screening ~ 21Table 5 Vision screening for the early detection of strabismus and .amblyopia in school-aged children evaluated against tenscreening parameters 30VIEXECUTIVE SUMMARYVision screening programs, conducted both prior to and at the time of school entry remain astime-honoured efforts for the early detection of potentially vision-damaging ocular disorders,particularly strabismus and amblyopia. At issue in this report, is whether the available scientificevidence supports these vision screening efforts.Definitions:1) Amblyopia is a clinical condition characterized by decreased vision in one or botheyes without detectable anatomical damage either in the eye or the visual pathway.2) Strabismus is a manifest deviation of the eyes from a single point of fixation.Strabismus is important because it is the most common cause of amblyopia.The report documents that vision screening programs, although relatively harmless to conduct andpotentially contributing to the discovery of very serious visual disorders, cannot be supported bydirect medium and long-term effectiveness evidence. That is, controlled trials have notestablished that children detected through any vision.screening program are ultimately better offbecause of the screening effort. Instead, evidence of testing validity and treatment efficacy forvision screening can only be supported by indirect evidence. Indirect evidence means, in thisinstance, that causality must be inferred through linking evidence for various program elements,rather than established through a completed screening program evaluation. The purpose of thisreport is to outline and critically examine this indirect evidence.The report concludes that, while the weight of scientific evidence favours vision screening ingeneral and at a much earlier age than school entry (preferably under age 3) the evidence does notestablish whether a particular population will be better off with either a school entry or apreschool program. Population benefit depends on the trade off between treatment effectivenessand population coverage for any target age. Thus, instituting new or altering old programsrequires empirical study to establish optimum population benefit.Screening programs for more common visual disorders, primarily refractive errors withoutamblyopia, are not considered in this report because earlier intervention in these conditions isunlikely to influence long-term visual outcome. 1,2The remainder of this summary outlines the literature reviewed and its relevance to the centralquestion:Central question: Is screening for vision-threatening ocular conditions, primarily amblyopiaand strabismus, supported by direct or indirect scientific evidence?Answer for direct evidence:No. Controlled trials have not established the overall efficacy or effectiveness ofvision screeningprograms.B.C. OFFICE OF HEALTH TECHNOLOGY ASSESSMENT 1Vision Screeningthe better eye. Treatment effectiveness evidence from long-term observational studies fromEngland and Israel cited above, for example, show that earlier vision screening, ideally under 3years of age, results in shorter and more effective treatment for amblyopia. As well, the samestudies describe the diminishing benefit from treatment for these screen detected childrenbeginning at age 4 to 5 years.7-10Indirect evidence also shows that the degree of sophistication of the tester and the ophthalmicequipment required for pre-school vision screening increases rapidly with decreasing age. Theprimary problem is that amblyopia detection without sophisticated optometric equipment andtraining involves unilateral acuity testing; testing dependent on children's cooperation andcomprehension. Several relatively weak cohort studies have shown that, other than newerautomated photographic techniques, the more knowledgeable and skilled the tester, the more.effective the program of amblyopia detection. 21,22However, while the weight of scientific evidence, albeit indirect, supports vision screening both atan earlier age than school entry and using sophisticated tests and testers, this does not necessarilymean that a program with these characteristics provides the optimum visual outcome for aparticular population. The optimum visual outcome benefit for any population will depend onpopulation coverage, the availability of testers and tests, as well as the likelihood of treatmentcompliance. Central among these issues for British Columbia is the trade off between visionscreening at school entry, when most children are efficiently gathered together, versus screeningat an earlier age, when complete population coverage is less likely but treatment is known to bemore effective. Thus, it is recommended that any move to eliminate current or establish newvision screening programs should depend on careful empirical evidence from visual outcomeprogram evaluations.B.C. OFFICE OF HEALTH TECHNOLOGY ASSESSMENT 3Vision Screeningknowledge claims in this area and insight about some of the problems with screening programevaluations.A review of the current efficacy and effectiveness evidence on vision screening is also reasonableat this time because this evidence is not likely to be clarified in the foreseeable future. On the onehand, the quality of direct evidence of screening program benefits will not likely improve becauseof the ethical problems associated with conducting a controlled trial. That is, delaying potentialvision-saving treatment in a cohort of children, a step necessary for a controlled trial, has rarelyoccurred in the past and seems increasingly unlikely in the future. As a result no definitive clinicaltrials designed to measure the effect of screening on long-term visual outcome are known to beunderway or are planned. On the other hand, the most valid, albeit mostly indirect observationalevidence comes from studies conducted more than a decade and often two decades ago, perhapswhen families were more likely to remain in one district long enough for adequate long-termassessments. The problems associated with mobility of populations is, ifanything, worsening.Nevertheless, while clinical trials will likely remain unethical and any study will likely provedifficult longitudinally, carefully-conducted cohort studies could be used to study vision screeningprograms. For example, this design could be used to compare existing with innovative programsin single or among matched populations in the same or among different districts over time.In summary, the overall conclusion that vision screening programs have not been provenbeneficial using scientifically-sound research remains solid and well-accepted in the literaturefound in various Western countries. Therefore, knowledge claims regarding andrecommendations for vision screening pr<:>grams from any professional group are based onopinion, not long-term outcome evidence.Our recommendation, therefore, is that, ifexisting programs are to continue or new programs areplanned, they should only be instituted as fully-evaluated pilot projects, using the best studydesign possible, in order to assess particular testers, tests and treatments.With any change in the provision ofvision screening in the province, a whole new series of publicpolicy problems will likely arise; not the least ofwhich is whether public health or primary carepractitioners should be responsible for providing and/or administering vision screening programs.1. BACKGROUND AND SIGNIFICANCEAmblyopia and strabismus are closely related pathogenetically and clinically. For example, on theone hand, poor visual acuity due to amblyopia in one eye can result in manifest deviation of thateye from normal alignment and result in the clinical diagnosis of strabismus. On the other hand,prolonged deviation of one eye from normal alignment could result in cortical suppression of thevisual input from that eye which could become amblyopic . In addition, the presence of eitheramblyopia or strabismus can result in a loss of binocular vision, also known as fine depthperception or stereopsis.The close inter-relationship between three eye parameters -- acuity, alignment, and stereoacuity -­have made natural history and epidemiological studies particularly difficult.l? That is, not onlyare the conditions often coincidental but they may also have codeterrnined etiology. ThusB.C. OFFICE OF HEALTH TECHNOLOGY ASSESSMENT 5Vision Screening(ii) anisometropia - inability to focus eyes simultaneously causing suppression of theimage in one eye; associated with hyperopia.(iii) large ametropia - both eyes may become amblyopic because of a failure to form afocused image in either eye.(iv) occlusion amblyopia - for example , complete blepharoptosis or capillary hemangioma.The epidemiology of amblyopia remains particularly problematic because this term remains toovague for rigorous assessment of even the most basic disease parameters . Hillis17 provides adetailed analysis of the problems associated with epidemiologic amblyopia research.Amblyopia is a common problem, with a generally-accepted prevalence estimate of 2%.17 Oliveret a/15 provide one of the few population-based estimates of amblyopia prevalence at differentage groups:Age Group (years)1.5-22 - 33 - 44 - 55 - 6Prevalence (%) results suggest a tendency for prevalence to increase with age. However, this tendencylikely reflects nothing more than the difficulty of diagnosing children under age 3 years.Prevalence estimates in an untreated population of4 to 6'year old children range from .8 to 5%,with the most widely cited estimate of2% (for discussion see Hillis17 and von Noordenl") . Theseauthors suggest that the wide discrepancy is likely the result ofvariation in screening programsensitivity rather than actual variation in population prevalence.Prevalence calculations for amblyopia are extremely problematic and must be viewed with cautionespecially for children younger than 3-4 years of age. This is in part due to variations in diseasedefinitions and testing methods. The major problem, however, is that for children in this agegroup, acuity testing is often very difficult (best described by Ingram;35 also Campbell-'v). Forexample, at the time of detection ofvisual impairment, it is difficult to distinguish sensory corticalsuppression from difficulties with testing younger children, either with or without correctivelenses. Moreover, even after treatment of suspected mild to moderate amblyopia, improvedvision in the most affected eye often remains suspect.Amblyopia is most commonly associated with strabismus. For example, in one unselected groupof 38,000 children screened between 1-1/2 and 3 years of age, 41% ofthe 1.3% of children withstrabismus had amblyopia. 13At all ages infants with unilateral strabismus had a higher prevalence of amblyopia than those withalternating strabismus .37B.C. OFFICE OF HEALTH TECHNOLOGY ASSESSMENT 7Vision ScreeningAcquired esotropia, the most common form of strabismus, usually develops between 1 1/2 and 3years of age, but may develop in younger and older children.!Incidence falls off rapidly after age 5. For example, only two new cases of strabismus were foundin a study of 1530 seven-year-old children who had previously been examined at age four. 421.3 Treatment of Strabismus and AmblyopiaTreatment of strabismus and amblyopia are considered separately in this section, although inpractice these conditions often are treated concurrently.1.3.1 Amblyopia TreatmentThe goals of amblyopia treatment are to acquire normal vision in both eyes, normal eye alignment,and ultimately normal depth perception. The main principle guiding treatment is to improve useof the affected eye by impeding use of the better eye. This is usually accomplished by occlusion,but other treatments are employed including penalization. 43,44 Occlusion and penalization !U"edescribed as follows:Occlusion is usually achieved using an adhesive patch. It may be full or part-time. Iffull­time occlusion is used, care is taken that treatment amblyopia does not occur in the bettereye. Occlusion is usually maintained until vision becomes normal, or until visual acuity isstable for at least 3 months . Further occlusion may be required due to the fact that, in50% of cases, deterioration in acuity occurs.U .Penalization, a relatively uncommon technique, involves blurring the vision in the bettereye using atropine or blurring lenses alone or in combination. Penalization is usually usedif occlusion fails or for maintenance therapy.!The greatest clinical difficulty with amblyopia therapy is deciding when to stop. This isparticularly true in younger children where testing visual acuity is problematic. Thus, clinicianshave difficulty determining ifvisual acuity has improved to a maximum, or alternately, if treatmentis stopped, whether visual acuity is degenerating.Acuity is usually treated before strabismus in part because it can be accomplished immediately inthe office with corrective lenses. Strabismus, in contrast, is usually delayed due to complicatingfactors such as, in the Canadian context, hospital operating room schedules. There are alsoinstances where improvements in acuity may lead to better alignment.R Subsequent treatmentmay require correction of residual strabismus.Many ophthalmology practice-based descriptive studies show dramatic amblyopia treatmentbenefit. Epelbaum et al, in a typical retrospective chart review of 407 patients with strabismicarnblyopia.J? noted that recovery of acuity of the amblyopic eye was maximal when occlusion wasinitiated before age 3 and decreased as a function of age until it was null by 12 years :B.C. OFFICE OF HEALTH TECHNOLOGY ASSESSMENT 9Vision Screening2) The 1988 U.S. Preventative Services Task Force recommends that:"vision screening is recommended for all children once before entering school,preferably at age 3 or 4. Routine vision testing is not recommended as acomponent of the periodic health examination of asymptomatic schoolchildren. "483) The Vision Advisory Committee to the BC. Ministry ofHealth, a committeeincluding representatives from both ophthalmology and optometry recommends that:"that all children have an eye examination by an eye doctor (ophthalmologist oroptometrist) by age 3 years." (personal communication)TABLE 1Recommended Vision Screening by Pediatricians and Family PhysiciansIAgeNewbornBy 6 months3.5 years5 yearsor olderExaminationPenlight examination of corneaNystagmusRed reflexesObjection to occlusionPenlight examinationStrabismusRed reflexesVisual acuityStrabismusFundus examinationVisual acuityStrabismusFundus examinationReferral CriteriaAny ocular pathologyRule out nystagmusAbnormal red reflexes or white reflexFixation to light or small toysNystagmus; any ocular pathologyCorneal light reflex test, cover textAbnormal red reflexes or white reflexAcuity of 20/40 or less in one or both eyesCorneal light reflex test, cover textAny ocular pathologyAcuity of 20/30 or less in one or both eyesCorneal light reflex test, cover textAny ocular pathology(After Crouch and Kennedy)49:1964) The BC Association of Optometrists, in conjunction with the Canadian Associationof Optometrists (COA):"encourages the implementation of effective vision screening programs for youngchildren. The earlier in life such programs can be applied the greater the potentialbenefit to the child. Irrespective of the performance ofvision screening programs,B.C. OFFICE OF HEALTH TECHNOLOGY ASSESSMENT 11Vision Screeningusually illiterate, children some form of "E" game or letter matching game (HOTV) is used . The. child is asked to point to appropriate letters on a near card, or in the direction that the E ispointing. The "E"s and letters are charted in diminishing sizes. These tests are generally acceptedas having high specificity, above 95%, however their sensitivity is low at 30%.25, 42 Testing foramblyopia, therefore, often involves additional tests for stereoscopic vision, which are discussedbelow.2.2 Tests for StrabismusThe most common screening tests for strabismus are the Hirschberg light reflex and theCover/Uncover tests. The Hirschberg test simply involves shining a light towards the eyes of achild and observing the symmetry of light reflected from the cornea. While simple to perform andrequiring little child cooperation, it is considered difficult to interpret and is seldom used alone.For the Cover/Uncover test, the child fixates on an object while the examiner alternately coverseach eye. A movement of the uncovered eye indicates strabismus. MacFarlane,50for example,reports typical Cover/Uncover efficacy of: sensitivity 43.5%, specificity 100%, and positivepredictive value 100%. Others have confirmed the low test sensitivity of the Cover/Uncovertest.51,52 The problem is that only about 50% of strabismus is grossly visible using the crudescreening tests such as Cover/Uncover tests. Detection of what are sometimes termed "micro­strabismic" children requires more sophisticated ophthalmic equipment and training.532.3. Depth perception (stereopsis) testsStereoscopic vision is one ofthe desired functional end points of normal bilateral visual acuity andalignment. The Titmus Stereofly and the Random Dot E tests are two of the more commonly­used stereoscopic vision tests . In both instances, a child wearing polarized glasses is asked toidentify objects or letters on cards visible only with symmetrical optical images.Testing for this "higher level" vision function is considered more efficacious than simple acuitytests for the identification of strabismus and amblyopia.54 The advantages cited include relativesimplicity, reliability, and the fact that no verbal response is required. Since the introduction ofthese tests in the 1970s, they have often been proposed as the method of choice for massscreening. However, evaluation has resulted in concerns regarding efficacy primarily because the"pass II criteria can be varied, particularly with the Titmus test, resulting in variable sensitivities andsome difficulty in comparing rates from different studies . Testing children under three years oldhas generally been unsuccessful due to the child's inability to cooperate.The Random Dot E test and the Titmus Stereofly are more effective than visual acuity tests indetecting strabismus, amblyopia or amblyogenic conditions in preschool children. Early efficacystudies showed the Random Dot E test more sensitive than the Titmus Stereofly although thefalse negatives are still considered unacceptably high.54Controversy remains, however. Hope,54 for example, compared two populations of children, oneat school (n=100, ages 5 to 15 years) and one in preschool (n=168, ages 3 to 4 years). In thepreschool group, the test was found to have an unacceptably low positive predictive value (17%) .B.C. OFFICE OF HEALTH TECHNOLOGY ASSESSMENT 13Vision Screening3. APPRAISAL OF RESEARCH REPORTING PRIMARY DATA3.1 METHODOLOGY3.1.1 Search StrategyElectronic searches were conducted using the NationalLibraryof Medicine MEDLINE (1985 topresent), HEALTH (1975 to present), HSTAR, CIJE, RIE, Nursing (1983 to present) andPSYCH INFO. Key search terms used were "schoolhealth services, vision screening, visiontests, program evaluation, amblyopia, strabismus, therapy and treatment." The searchwas limitedto English languagejournals and studieswith humans. Current content database searcheswerecontinued until the time ofpublication.A fugitive or non-electronic searchwas also conducted. Material that was examined includedreference lists in published articles, reports from other technology assessment offices, and other'agencies involved in systematic review of clinical evidence. Finally, clinical practice input was .solicited from a practicing orthoptist providing visionscreening, optometrists, a pediatricophthalmologist, and representatives from the BC OptometryAssociation. Optometristswereparticularly helpful in providing reference material regardingepidemiology, treatmenteffectiveness, and screening program research.3.1.2 Inclusion Criteria -and Exclusion CriteriaStudies included in the critical appraisal process were those reporting primary data evaluatingscreeningfor early detection of visionthreatening ocular disorders in a geographically definedpediatric population. The screening function was defined in the broadest sense to mean testing toseparate children into disease or risk categories. Evaluationwas looselydefined as an effort tocompare a screened to an unscreened populationor to follow an entire screen positiveand/orscreen negative population over time.The requirement that a study evaluates screening resulted in elimination of descriptive studies,also known as case series or cross-sectional surveys. This meant excluding, from critical appraisal,most of the published literature including the recent report on schoolvision screening in BritishColumbia. Similar program descriptions have been noted for other areas of CanadaincludingOntario,59, 60 Manitoba,61 and New Brunswick,62 as well as other countries such as the UnitedKingdoms-' "and the United States.i! The bibliography lists approximately 50 observationalstudies that were systematically reviewed but excluded from the critical appraisal process.The requirement that the study evaluated screening of a geographic populationreflects theorientation of this report towards assessing primary or mass screening. Excluded are studies ofchildrenalreadywithin the health care systemsuch as ophthalmologic or optometric clinic-basedpopulations.Background material on pathology, epidemiology and treatment was also reviewed, as were anumber of reports providing expert clinical opinions.30,45,46,51,64,65,66,67,68,70,71,72B.C. OFFICE OF HEALTH TECHNOLOGY ASSESSMENT 15Vision ScreeningTable 2a. Summary of research evaluating vision screening as a clinical intervention: IsraelStudylSelling Population/Criteria Age Study I Screen Prelim. findings Gold standard . Diagnosis Treatment/ Outcome Commentrange deslnn enrolled test/testers % feiled Test/testers durati onNaumann Solicited Pilot Project I to 2-1/ Prospactlve 6400 Orthoptlsts: 342 feiled Ophth.: 166 treated 26 - CUE, 21 of 22 becama Sae Neumann 19871971 80000llered yr. old Descriptive 80% 133 (2.1%) S essumed A alternators for lonq-tsrm oulcomeChild Welfara 1968-70 Focus on Observation "Full" exam 104 newClinics Strabismus and LIght Reflex No details 29 known 6 losI 10 loll ow-upAmblyopia Covar rest 33 otherPPV (aill • 48%PPV lsi· 39%NeumaM Consecutive casas 7 mo. 10 1969-72 19,000 See above 97 failed Ophth : 97 CUS Glasses VA : 619 (42%) Convenience sample1987 or CUE 3 yrs. Prospective acuitv, slereo 78 esotrepla Occlusion 6/12(62%) 19/97 children lost toFoilow 'up Attended until e (lnitlel Analytic acuity , elternatlon end followed to ege 6 6/60 (6.4%1 follow-upminimum 01 ege 6 screenl No initial VA Minimum 3·6 yrs. of the diagnosedover age 2 yrs.• 50%had VIA 6/12 or bett erFriedman 98% pop. 1 102-1/ Prospect lve 38,000 Orthoptists (alii 2009 failed Ophth: S - (498) 1.3% Nol described See Friedman 1985 Progrem description1980 yr. old Analytic inspection full exam . CUS - (206) 0.6%Child Welfare 1968·76 IIghl reflex no details (Amblyopia assumed)Clinics cover text Includes refraction Other (781 .2%Rapid Retinoscopy 173 failed Dphlh:1E 84/173 Signif icant Glasses See Friedman 1985 Preliminary Report1974-76 (15, 0341 as ebove lefrectlve errors PenalizationAller Init ial screening 0.4% streight eyeUndilated pupils emblyopiaFriadman See Friedman 1980 1 to 2·1/ Prospective 39/15,084 See Friedman 1980 See Friedman Ophth, A - 6/12 or less, or Unt il ege 7 20 151%1 residual A Biased by required1966 Results of Screen Analy~ic Consecu- Repid retinoscopy 1980 dillerence of 2 lines on Glasses 6 (15%) bilateral A folloV'{·up lengthDetected 1974-77 tive cases screening eye chert Penalization 11 135%) A 6/12-6/16Non-Strabismic 7 117%) A 6/20-6/24Amblyople 2 (6%1 A 6/90Markad ametropia 8 full stereopslesfoilowed to 7-10 ylS.LEGEND: A = AmblyopiaS = StrabismusCUE = Constant Unilateral EsotropiaCUS = Conslant Unilaleral StrebismusGP = General PraclitionerVA = Visual Acully6/6 = normal acuily: 6/9 = 2/3 normal; 6/12 ~ 1/2 normal, etc .PPV = Positive Prediclive ValueBC OFFICE OF HEALTH TECHNOLOGY ASSESSMENTSchoo/ Vision Screening 17.'Table 3a. Summary of research evaluating vision screening as a diagnostic procedure: Program assessmentStudy/Setting Criteria/Population Age Study #% Screen Prelim. findings Gold standard Diagnostic PPV/NPV Commentsranee dasian enrolled test/testers % failed Test/testers outcomeFeldman 1980 Representative 5-1/2 yrs Prospective Experimentel Specialty Nurse 78 /10%) failed Ophth. exam No results N/C Best prospective cohortCanada samples of cohort group N=763 Illiterate E Test 19 (2.5%) glasses not described studyPreschools & schools geographic areas Matched: screened 24 (3%) VA 20/50age 6-12 monthssex earliersocia-economic•status Control group Same nurse 112 (15%1 failed Ophth . No results N/CN=746 Same test 12 11.6%1 glasses not describbdUnscreened 43 16%1 VA 20/50Kohler and Stlgmar G. All children starting 7 yrs. Prospective Experimental School nurses Failed: acuity < .9 Ophth. 111.7%) N/C1973/1978 school in one city Cohort group Linear' En chart full exam Vision-threateningSwedan ovar 3 yaar period Screened well-defined conditionCommunity-basad N=1530 tane 4) 1 AmblyopiaSchool Setting Control group See above See above See above 2914.5%) N/C Groups poorly matchedUnscreened 5 AmblyopiaN-648NPV 97.6%De Becher 1992 Random 4-112 to Prospective 11,814 P.H. Nurse 802 (6.8%1 total Orthop. end Ophth . 11 (7%) fail Vision-threaten- Result may be highCanadian representative 5-112 yrs Analytic HOTV 200 screen posi- 157 (78 .5%) seen 146 (93%1 pass ing conditions 25% sample not testedProvincial sample from Stereo-acuity . tive randomly- 43 not seen 1 amblvoplaINova Scotia) entire province test selected missed45 screen 36 180%) seen 18 Fail PPV 50% Result suspectnegative 18 Pass 20% not testedrandomlv selectedLEGEND: VA = Visual AcuityNC - Not CalculatedBC OFFICE OF HEALTH TECHNOLOGY ASSESSMENTSchoo/ Vision Screening 19Table 4. Results of critical appraisal of studies evaluating vision screeningCriterion , :;. )il:. <;.... .;. '-' ." .::';. :/:I . Representetive iihd . g~nii!alii~bfe "tci·. s~ree~ing ~~p~iatl~n L:,::. .~:}" :~si~ . . .:.:":.:::: .". :j:: . >" -=.:::- .: .~ :2. Communlty.settlng .:< ;,">·3. s::::~:~~:nlri;E~:o ~! i ~,~~·:n.~~~.r=·~,( · ::i~.i';':: ';;~i;~; - lJ~'~J~~~ ••~iiJ: ,ji~n~:~ i~·:·~.~h~~:::. ' . : : · i~:;i·: ·. : . i iii·i; i · · . ; ; i : ii: :i·i.. ::..} : ·.)t»~··j;t, .:) .:i;W:·.:;:1::..}fi~;:2):~.;]}.5. Adequatedesrtprlon 01 tests -':'-,~" ...,':;': :@A:.""·':1 6 . ~.l i lldad int~ip'i~(ijtioh'dfi~~(~dJ 'i\ }: ; :.\:- ~PI~ st~;;d ~Ui~~lJ i (i .U: ) i::; ii/::L ::Ti=:>::7; All subjecls'eligilile and enroliQiC:" t~'acc ciu ~ t ij d i'or;1:::: "... ..L,: . ;;.;: :;.,,.. ;:ri..8; Apjiiopiiat~ calc'ufalions..t i 6illls'iilvili f · Ii:;~'nd specl;i:c'It~ ~''' : . . .'f~ · :;::i.~t: \[: ·)t9. Appropriallic8'I~uiation~ 9IPtBit~f~ri~~{ '(} 'poSitlV8 an~·iiegatjit~ · pr~dlcii~~ :v~iiJe~ /:Jt:... . >.. . : ~' : '.· :·: .:· ~ :: .(f· t:·:10. Approprlaie usa 01 statlsticel.tests ',:' :::,.~..::~: ;::. ':!i:.,;... .:.. ;::: :.\ .;:;:;..<:~~;~i ;:.':.: :;.;: i/~::-:::;:~~/" ::':;;;i;;;;:- : ;: ~:i ::f~;De Becher I Feldman I Kohler I Neumann Friedman Friedman Neumann Oliver Ingram I Ingram I Bolger I Jarvis1971 1980 1985 1987 1977179 1986G I G P F G G G P F F F FG I G G G G . G G G P G F FG I G P F G G G F F F F FG I P G F F G G F G G P PG I F F F F G G P PIG G P PG I NC P F F NC P P P F P PP G G P G F P F P/F F P PNC NC NC NC NC NC NC NC NC NC NC PG I NC NC NC NC NC NC G NC NC NC FG I NC NC NC NC NC G NC NC NC NC PG = good (criterion wes adequately metl, F ~lair (criterion was inadequatelymet), P=poor (criterion was not met). NC=not calculatedBC OFFICE OF HEALTH TECHNOLOGY ASSESSMENTSchool Vision Screening 21applied to the correct individuals is a necessary component of an assessment of screening initiatedtreatment efficacy and effectiveness.Researchers used both intervention and diagnostic study designs to assess vision screening.Studies using an "intervention" design are listed in Table 2a and b; 2a for Israel, and 2b for theUnited Kingdom. Studies using a "diagnostic" design are listed in Table 3a and b; 3a for programassessments and 3b for tester assessments.The actual study quality using these designs, while variable, was quite poor. Moreover, theproblems with study quality were reciprocal. That is, on the one hand, while the interventionstudies followed children over time, these children were not assessed relative to an explicit or ahistorical control group. On the other hand, the diagnostic studies, three of which included someform of control group, did not follow children long enough to adequately compare treatmentoutcome for the two groups.. 3.2.2 Evidence of Diagnostic EfficacyThree research teams studying diagnostic efficacy3,4,5 removed most major sources of bias byincluding a control group in their study design. Incidentally, as mentioned above, all three ofthese studies examined children at the time of school entryIn two studies,3,5 a previously-preschool-screened population was re-examined at school entryand compared to a cohort of previously unscreened children.Feldman et aI3 provided stronger evidence because the two cohorts were matched for age, sex,and socio-economic status. In this study, a pediatric nurse tested children at school entry (meanage of 65 months) using the "illiterate E" visual acuity test. They showed a 50% lower overallprevalence ofuntreated eye disorders and 79% fewer moderate to severe vision problems in thepreviously screened group.Kohler et al,5 using an unmatched and poorly-described comparison group, also showed that "therisk of finding a significant eye disorder in a school entrant was more than 6 times greater for achild who was not examined in his preschool years, and the risk of finding an amblyotic child wasmore than ten times greater."A third study, de Becker et aI4 although using a different study design, also provided evidencethat their screening program of 4-1/2 to 5-112year old children accurately detected visualdisorders. These researchers randomly selected 200 screen negative children and subjected 78.5%ofthem to gold standard ophthalmologic examination. The study, which was designed tocalculate the negative predictive value oftheir screening program, showed incidentally that theirprogram had greatly reduced the prevalence ofvision threatening ocular disorders in the screennegative population. This was shown by the significant decrease in prevalence ofvisual disordersamong the 200 randomly selected screen negative children. .The explicit end-point of most diagnostic studies evaluating vision screening (Tables 3a and 3b)was the number of cases of visual disorders detected, both minor refractive errors, and the moreB.C. OFFICE OF HEALTH TECHNOLOGY ASSESSMENT 23Vision Screeningchildren and no one achieved normal vision. The relatively poor outcomes shown for theseusually severe forms of amblyopia is consistent with other studies of non-screen detected childrenwith these forms of amblyopia. Thus, some severe forms of amblyopia, even when discoveredthrough screening at one year of age, are largely resistant to treatment.Second, long-term visual outcome of screen detected constant unilateral strabismus cases,presumably with amblyopia, was explicitly studied .12,13 These studies are best consideredtogether as they are based on the same screen detected population. The authors report the visualoutcome for 78 of93 screen detected cases of unilateral esotropia detected when the childrenwere seven months to 3 years of age and followed until they were at least 6 years . While onlyfinal, not initial, acuity was reported, 50% achieved near normal and 67% acceptable (at least50% of normal) vision in the strabismic eye.In addition to the two series mentioned above, long-term visual outcome of screen detecteddisorders including both strabismic and straight eyed amblyopia was also studiedby Ingram et al9 and Oliver and Nawratzki.15,16 Ingram et al followed the children initiallyscreened at age 3 1/2 until at least age 5 and some childrenuntil age 11. They found that, whileacknowledging the difficulties of diagnosis and treating amblyopia, "most children achievednormal corrected visual acuity."Oliver and Nawratzki, in a methodologically much weaker study,15,16 found that 67% ofthe 1 1/2to 6 year old children with screen detected amblyopia achieved "good visual acuity."In summary, without cohort studies, the issue that remains is to what extent these screen detectedchildren have better ultimate visual outcome due to earlier detection and treatment of amblyopia.3.2.4 Evidence relating screening age to long-term visual outcomeSeveral studies explicitly examined the influence of screening age on diagnostic efficacy and long­term visual outcome. Two study types are identified: (1) studies that show lack of.benefit fromolder, school entry, screening programs; and (2) studies that show increasing benefit fromscreening younger children.Ingram et a[7,8 found vision screening at the time of school entry provided no improvement in"the incidence or severity ofamblyopia 3 or more years later. "8:236 Although they were onlydescriptive studies, they are included here because a screen detected population was followed fora significant length oftime. These studies found only 15 of the 120 screen detected children hadimprovement. As the authors note, without a control group the amount of deteriorationprevented cannot be estimated. Ingram et ai,10 in a methodologically-similar study, reports thatvisual acuity improved two lines on a standard eye chart in only 16 children out of the 108children identified through school entry vision screening.There is weak to moderate evidence to support the notion that significantly greater benefit isachieved if treatment is undertaken before three years of age, and perhaps as young as 18 monthsold. Neumann et ai12 provide the best evidence of treatment benefit of children detected underB.C. OFFICE OF HEALTH TECHNOLOGvASSESSMENT 25Vision Screening4. DISCUSSION OF FINDINGSVision screening programs, while sharing the common goal of facilitating normal binocular visualdevelopment, vary in almost every other sense, including tests, testers, and testing frequency, as'well as referral patterns and treatments applied. The almost complete lack of standardization,particularly the ~ack of treatment outcome definitions, makes programs very difficult to evaluateand compare. Nevertheless, some general observations can be made and are included in thesections that follow .4.1 The assumption of benefit from ophthalmologic interventionsA basic assumption underlying this research is that virtually all ophthalmologic interventionsdesigned to diagnose and treat amblyopia and strabismus provide at least some benefit andrelatively little harm. This assumption was based on various factors:First, diagnostic investigations following screening tests require no invasive tests and can berepeated as often as necessary. In particular, false positive tests do not lead to vision threateningor patient damaging treatments. However, false positive tests do have significant financial andpsychological cost to children and their families.Second, it is assumed that some treatment benefit occurs even in the circumstance where visualacuity is only marginally improved, or degeneration does not occur, in the most amblyopic eyeand strabismus correction is only for cosmetic reasons. This assumption can be made foramblyopia because treatments involve relatively harmless, if carefully monitored, occlusiontherapy of the better eye. Surgical correction of alignment, however, is not without anestheticand technical risks .A high or even marginal benefit to harm treatment ratio, however, itself provides no justificationfor the screening effort. A positive benefit/harm ratio is merely a necessary, not a sufficientcondition to support these programs. Screening effort justification rests on whether earliertreatment in the natural history of these disorders improves the benefit to harm ratio.4.2 Data synthesis to reach o~erall conclusionsIn the absence of direct evidence of overall vision screening program benefit using a controlledtrial study design, it has been necessary to critically appraise indirect evidence ofvision screeningefficacy. This has meant examining studies of the various elements of a vision screening program:tests, testers, diagnostic accuracy, and treatment efficacy. One other possible method to evaluatescreening program efficacy is to combine studies examining the same population over time .Two series of studies, one from the United Kingdom7-10 and one from Israel,11-14 provide thisopportunity. In both instances, combining the studies allows a cohort of children, screened froman acceptably-representative geographic population, to be followed reasonably long-term,The strengths and limitations of the individual studies were detailed in the preceding section.Overall, the studies were poor and the evidence for vision screening was weak. The evidence wasB.C. OFFICE OF HEALTH TECHNOLOGY ASSESSMENT 27Vision Screening5. Summary and ConclusionTable 5 summarizes this vision screening assessment in relation to the 10 screening parametersoutlined by Wilson and Jungner.73 Vision screening, which shifts treatment to the latent phase ofchildhood visual disorders, remains largely unproved by direct experimental evidence. Directevidence requires an ethically unacceptable controlled trial in which children are randomlyassigned to and maintained in screened and non-screened groups. However, other researchdesigns did provide some indirect and partial evidence that vision screening results in at leastmoderate patient benefit and very little harm. The following points are also worth noting:a) An adequate positive predictive value can be achieved for amblyopia and strabismusacross a wide range of tests, testers and age groups.b) All studies reviewed here, while methodologically weak, consistently show, except inthe most severe forms of amblyopia, improvement in visual acuity from the time ofdiagnosis.c) No studies showed evidence of physical harm to visual functioning .d) School entry screening showed the weakest, while age 3 to 4 screening showed the bestevidence of treatment efficacy and effectiveness.Therefore, it seems reasonable to assume that moving the diagnosis back in the natural historythrough screening efforts is justified particularly in the preschool age groups.Available evidence points to younger ages as the optimal time to screen for potential vision- .threatening ocular conditions. School entry vision screening as a program directed at olderchildren is more difficult to justify in terms of the scientific evidence. However, it may be that, ina particular population and setting, comprehensive school entry screening may in fact providebetter overall population benefit than more limited screening at an earlier age. Similar to almost. any other aspect of this area of medicine, the costlbenefit of eliminating and/or substituting visionscreening programs requires empirical study.B.C. OFFICE OF HEALTH TECHNOLOGY ASSESSMENT 29Vision ScreeningAPPENDIX AData Synthesis of Israeli and United Kingdom StudiesIn the following analysis, studies are grouped according to country: six from Israel and four fromthe United Kingdom.IsraelFour of the six Israeli studies reported prospective observational results of a vision screeningprogram for children age 1 to 2 1/2 years, tested primarily by orthoptists and ophthalmologistsvisiting Child "Welfare" Clinics in Haifa, Israel over a 15 year periodbeginning in 1968.The initial report.U and more completely a subsequent report.U describe the screening programincluding the population screened, the examiners involved, and the techniques employed. Theinitial program reported in these studies was almost exclusively suited to detection of strabismusand used the following techniques :1) history taking from parents2) external inspection of the eyes3) corneal reflex test4) the cover testNeumann et alII reported on, among other things, the strabismus and strabismic amblyopiatreatment results from screening 6400 children during the first 18 months of the program. Ofthe342 screen positive children, 133 (2.1%) were found to have strabismus, 29 ofwhich werealready under care. Short term treatment efficacy is described for the 28 children that hadpreviously undiagnosed constant unilateral strabismus. Constant unilateral strabismus wasassumed associated with amblyopia since amblyopia testing is not possible in this age group.Twenty-one of the 22 children, followed for four to nine months became "alternators, II meaningtheir eyes did not shift during the cover test.The subsequent report--' presents more complete results of the first eight years of this screeningprogram involving 38,000 children which was 98% ofthe children born in this area during thestudy period. In the larger study, 498 children (1.3%) had constant unilateral strabismus andpresumably amblyopia. The long-term visual outcome and stability of children found to haveconstant unilateral strabismus in the earliest screen detected cohort (1969-72) were reported inNuemann et al.12 A total of78 of the 97 screen detected cases remained in attendance until atleast the age of six years.The researchers also assessed long term stability of treatment by examining 27 of the children 4 to10 years following the end of treatment in their eye clinic. Treatment included a combination ofglasses, occlusion of the fixating eye, the use of atropine drops, as well as surgery. They foundthat 82% of the children achieved normal or near normal visual acuity (6/12 or better) .Furthermore, they found 59% retained their vision when examined 4 to 10 years later. Nochildren developed full stereoscopic vision. While the study conclusions regarding the optimalage for screening are invalid because of the significant number of children lost to follow up, theB.C. OFFICE OF HEALTH TECHNOLOGY ASSESSMENT 31Vision ScreeningIngrarnlv further illustrated the limitation of school entry vision screening programs. This studywas a retrospective review of the 108 cases referred to ophthalmologists from school visionscreening in Kettering from 1976 to 1978. Referral criteria were visual acuity less than 6/18 withor without strabismus. These years were chosen because 1976 was the first year records werekept and after 1978 many of the children in this district were prescreened near age one withrefraction. The cases were drawn from seven general medical practices around a hospital referraleye clinic. The purpose was to compare initial with final acuity of the worst eye from the time ofreferral to last known acuity. Treatments were not specified in detail, rather they were referred toas the "usual" correction of refractive errors with glasses, as well as occlusion and surgery foramblyopia and strabismus when "advisable and practical". Duration offollow-up was notspecified. Improvement of more than two lines on a Snell or Sheridan-Gardiner chart was foundin 16 children. Thirteen of 18 children with severe amblyopia did not improve . Ingram estimates900 children were born each year in the practices from which the sample was derived. He notesthat only 16/900 improved (0.6%). The study does not indicate how many of these 900 childrenwere vision tested in school.Ingram et al9 provide a prospective analysis ofa screening program for 1507, 3 and 1/2 year oldchildren, followed until 5 to 11 years old. This represented between 70 to 75% ofthe populationin this district. Screening at this age detected 41% of visual defects. Twenty-six percent had beendetected earlier and 33% presented after 3 and 1/2 years old. Treatment included correctivelenses and occlusion therapy. Decision regarding treatment was very problematic forophthalmologists because ofdifficulties distinguishing blurred vision from either uncorrectederrors of refraction or amblyopia requiring occlusion. With ambiguous acuity in 50% of cases,improvement could not be assessed. In addition, failure of occlusion to result in improvementcould not be concluded because the child may not have needed the occlusion in the first place.Corrective lenses were very successful but only 3 of 16 children identified as having decreasedacuity at the initial age achieved normal vision. It was noted that 9 of 10 children with severeamblyopia presented prior to screening and never benefited from any therapy. The authorsconclude, based on this and other work, that therapy seems to depend on the presence or absenceof hypermetropia at age 1 rather than timing of diagnosis and therapy.DiscussionDespite the absence of a control group, these reports collectively provide some evidence ofpopulation based screening program benefit because they show improved visual outcome in asubstantial portion of the screen detected children. As mentioned in section 1.1.2, despite lack ofnatural history evidence, expert opinion considers improvement extremely unlikely in the absenceof treatment. Furthermore, because of the subtlety of the amblyopia and strabismus found,detection of these conditions is considered very unlikely outside of these vision screeningprograms. Indeed, as the authors note, their detection would likely otherwise occur much later inthe natural histories of the conditions.B.C. OFFICE OF HEALTH TECHNOLOGY ASSESSMENT 33Vision Screening1/2 years, in child welfare clinics. J Pediatr Ophthalmol Strabismus 1980;17(4):261-67.14 Friedman Z, Neumarm E, Abel-Peleg B. Outcome oftreatrnent ofmarked ametropia without strabismusfollowing screening and diagnosis before the age of three. J Pediatr Ophtha1mol Strabismus 1985;22(2):54-57.15 Oliver M, Nawratzki 1. Screening of pre-school children for ocular anomalies : II. Amblyopia:prevalence and therapeutic results at different ages. Br J OphthalmoI1971;55:467.16 Oliver M, Nawratzki 1. Screening of pre-school children for ocular anomalies: 1. Screening methodsand their practicability at different ages. Br J Oph~011971;55:464-66.17 Hillis A, Flynn IT, Hawkins BS. The evolving concept of amblyopia: a challenge to epidemiologists.Am J EpidemioI1993;118(2):192-205.18 Fielder AR. The management of squint. Arch Dis Child 1989;64:413-18.19 Epe1bawn M, Milleret C, Buisseret P, Dufier JL. The sensitive period for strabismic amblyopia inhwnans. Ophthalmology 1993 March 3;100(3):323-27.20 Ingram RM, Walker C. Refraction as a means ofpredicting squint or amblyopia in preschool siblingsof children known to have these effects. Br J OphthalmoI1979;63 :238-42.21 Jarvis SN, Tamhne RC, Thompson L, Francis PM, Anderson J, Colver AF. Preschool vision screening .Arch Dis Child 1990;65:288-94.22 Bolger PG, Stewart-Brown SL, Newcombe E, Starbuck A. Vision screening in preschool children:comparison of orthoptists and clinical medical officers as primary screeners. Br Moo J 1991 Nov23;303:1291-94.23 Canadian Association of Optometrists. The Canadian Association of Optometrists Briefto the Houseof Commons Standing Committee on Health and Welfare, Social Affairs, Seniors and the Status ofWomen. April 10, 1990.24 Buchan R. Vision services program: a basis for restructuring, Victoria (Be): Speech and HearingBranch, Preventive Services Division, Ministry ofHealth; 1987 Jan.B.C. OFFICE OF HEALTH TECHNOLOGY ASSESSMENT 35Vision ScreeningOpin OphthalmoI1992;3:735-40.38 Fielder AR. The management of squint. Arch Dis Child 1989;64 :413-18.39 Elston JS, Lee JP. Strabismus in childhood. Br J Hosp Med 1985;8-12.40 Cross AW . Medical progress: health screening in schools, part 1. J Pediatrics 1985 Oet;107(4):487-9441 Erlich MI, Reinecke RD, Simons K. Preschool vision screening for amblyopia and strabismus:programs, methods, guidelines, 1983. Surv Ophthalmol1983 Nov-Dec;26(3):145-163 .42 Kohler L, Stigmar G. Vision screening for four year old children. Acta Pediatr Scand 1973;62: 17-27.43 Cruz OA, Flynn IT. Amblyopia. In: Cibis GW, Cibis Tongue A, Stass-Isem ML, editors. Decisionmaking in pediatric ophthalmology. St-Louis, Missouri: Mosby, 1993:182.44 Ching FC, Parks MM, Friendly DS. Practical management of amblyopia. J Petriatr OphthalmolStrabismus 1986;23(1):12-16.45 Vaughan D, Asbury T, Riordan-Eva P. Objectives and principles oftherapy of strabismus. In: Generalophthalmology. 13th ed. Noralk CT: Appleton and Lange 1989:238-41,353,393.46 Holland WW, Stewart S. Screening and surveillance in childhood. In: Screening in health care: benefitor bane? London: The Nuffield Provincial Hospitals Trust 1990:73-97.47 Canadian Task Force on the Periodic Health Examination. Periodic health examination, 1989 update :3. Pre-school examination for developmental, visual and hearing problems. Can Med Assoc J 1989 Dec1;141:1136-40.48 U.S. Preventive Services Task Force ..Guide to clinical preventive services: an assessment of theeffectiveness of 169 interventions. In: Report of the US Preventitive Services Task Force. Guide to clinicalpreventive services. Baltimore : Willams and Wilkins, 1989:181-85.49 Crouch ER, Kennedy RA. Vision screening guidelines. In: Cibis GW, Cibis Tongue A, Stass-IsernML, editors. Decision making in pediatric ophthalmology . St-Louis, Missouri: Mosby, 1993:196-197.B.C. OFFICE OF HEALTH TECHNOLOGY ASSESSMENT 37Vision Screening62 Woodruff ME . Vision and refractive status among grade 1 children of the province ofNew Brunswick.Am J Optom Physiol Opt 1986;63:545-52.63 Hall SM, Pugh A, Hall DMB . Vision screening in the under-Ss. Br Moo J 1982 Oct 16;285: 1096-98.64 Cibis GW. Options in amblyopia therapy. In: Cibis GW, Cibis Tongue A, Stass-Isem ML, editors.Decision making in pediatric ophthalmology. St-Louis, Missouri : Mosby, 1993:184.65 Cunningham AI, Woodruff ME . The evaluation of vision screening procedures: a preliminary exercisein public health optometry. Can J Public Health 1978 Nov-Dec;69(l Suppl):24-27.66 Flynn JT, Cassady JC. Symposium, current concepts of amblyopia: current trends in amblyopiatherapy. Ophthalmology 1978 May;85(5):428-47.67 Helveston EM, Ellis FD, editors. Pediatric ophthalmology practice. In: Pediatric ophthalmologypractice. 2nd ed. Toronto: C.V. Mosby, 1984:33-35,77-90.68 Reinecke RD . Current concepts in ophthalmology. N Engl J Moo 1979 May;300(20):1139-41.69 Schecter M, LeBlanc FE. Critical appraisal ofpublished literature. In: Hanstroidle W, Spitzer 0,McPeek B, Mulder DS, and McKneally MF, Editors. Principles and practice of research : strategies forsurgical investigations. New York:Springer-Verlag, 1986:112-117.70 Kohler L, Stigmar G. Testing for hypermetropia in the school vision screening programme. ActaOphthalmoI1981;59:369,374-77.71 Ingram RM. Refraction as a basis for screening children for squint and amblyopia. Br J Ophthal1977;61 :8-15.72 Ingram RM, Walker C, Billingham B, Lucas J, Dally S. Factors relating to visual acuity in childrenwho have been treated for convergent squint. Br J Ophthalmol1990 Feb;74(2):82-3.73 Wilson JMG, Jungner G. Principles and practice of screening for disease. Geneva: WHO (PublicHealth Paper) 1968:34.B.C. OFFICE OF HEALTH TECHNOLOGY ASSESSMENT 39Vision ScreeningChalmers TC, Smith H, Blackburn B. A method of assessing the quality of randomized controltrial. Controlled Clin Trials 1991 ;2:31-49.Ching FC, Parks :MM, FriendlyDS. Practical management of amblyopia. J Pedriatr OphthalmolStrabismus 1986;23(1):12-16.Cibis GW. Options in amblyopia therapy. In: Cibis GW, Cibis Tongue A, Stass-Isern Ml., editors.Decision making in pediatric ophthalmology. St-Louis, Missouri: Mosby, 1993:184.Cogen MS, Ottemiller DE. Photorefractor for detection of treatable eye disorders in preverbalchildren. AlaMed 1992;62(3):16-2Q.Coyte PC, Dewees DN, Trebilcock MI. Medical malpractice - the Canadian experience. N Eng JMed 1991;324:89-93.Cross AW. Medical progress : health screening in schools, Part I. J Pediatrics 1985Oct;107(4):487-94.Crouch ER, Kennedy RA. Vision screening guidelines. In: Cibis GW, CibisTongue A, Stass-IsernML, editors. Decision making in pediatric ophthalmology. St-Louis, Missouri: Mosby,c1993:196-197.Cruz OA, Flynn IT. Amblyopia. In: Cibis GW, Cibis Tongue A, Stass-Isern ML, editors. Decisionmaking in pediatric ophthalmology. St-Louis, Missouri: Mosby, c1993:182.Cunningham AI, Woodruff ME. The evaluation ofvision screeningprocedures: a preliminaryexercise in public health optometry. Can J Public Health 1978 Nov-Dec;69(1 Suppl):24-27.De Becker ill, MacPherson HJ, Laroche GR, Braunstein J, Cottle R, McIntyre L et al. Negativepredictive value of a population-based preschool vision screening program. Ophthalmology 1992June;99(6):998-1003.Duckman R. The incidence ofvisual anomalies in a population of cerebral palsied children. J AmOptom Assoc 1979;50(9):1013-6.Ehrlich NIT, Reinecke RD, SimonsK. Preschool vision screeningfor amblyopiaand strabismus :programs, methods, guidelines, 1983. Surv Ophthalmo11983 Nov-Dec;26(3):145-163.Elston JS, Lee JP. Strabismusin childhood. Br J Hosp Med 1985;8-12.Epelbaum M, Milleret C, Buisseret P, Dufier JL. The sensitive period for strabismic amblyopia inhumans. Ophthalmology 1993 March 3;100(3):323-27.Fabian G, Sjostrand 1. A longitudinal study of a population based sample of astigmatic children.II. The changeability of anisometropia. Acta Ophthalmol1990 Aug;68(4):435-40.B.C. OFFICE OF HEALTH TECHNOLOGY ASSESSMENT 41Vision ScreeningHillis A, Flynn IT, Hawkins BS. The evolving concept of amblyopia: a challenge toepidemiologists. Am J Epidemiol 1993; 118(2):192-205.Holland WW, Stewart S. Screening and surveillance in childhood. In: Screening in health care:benefit or bane? London: The Nuffield Provincial Hospitals Trust, 1990:73-97.Hope C, Maslin K. Random dot stereogram E in vision screening of children. Aust N Z JOphthalmoI1990;18(3):319-24.Ingram RM. Amblyopia: the need for a new approach? Br J OphthalmoI1979;63:236-7.Ingram RM. Amblyopia: neither screening nor treatment is satisfactory. BMJ 1989:298:204.Ingram RM. Refraction as a basis for screening children for squint and amblyopia. Br J Ophthal1977;61 :8-15.Ingram RM. Review of children referred from the school vision screening program in Ketteringduring 1976-8. Br Med J 1989 April 9;298:935-6.Ingram RM. The problem of screening children for visual defects. Br J Ophthalmol1977;61 :4-7.Ingram RM, Holland WW, Walker C, Wilson 1M, Arnold PE, Dally S. Screening for visualdefects in preschool children. Br J Ophthalmol 1986;70:16-21.Ingram RM, Rogers S, Walker C. Occlusion and amblyopia. B. Orthopt. J 1977;34(11)) 1-22.Ingram RM, Walker C. Refraction as a means of predicting squint or amblyopia in preschoolsiblings of children known to have these effects. Br J Ophthalmol 1979;63 :238-42.Ingram RM, Walker C, Billingham B, Lucas J, Dally S. Factors relating to visual acuity in childrenwho have been treated for convergent squint. Br J Ophthalmol1990 Feb;74(2):82-3.Ismail H, Lall P. Visual acuity of school entrants. Child Care Health Dev 1981;7:127-34.Jarvis SN, Tamhne RC, Thompson L, Francis PM, Anderson J, Colver AF. Preschool visionscreening. Arch Dis Child 1990;65 :288-94.Jervis B. A case for pre-school vision screening. Can J Public Health 1978 Nov-Dec;69(1 Suppl):45-48.Kennedy R, Sheps SB, Bagaric D. Field trial of the "otago" photoscreener. Can J Ophthalmol. Inpress.Kennedy RA, Sheps SB. A comparison of photoscreening techniques for amblyogenic factors inchildren. Can J OphthalmoI1989;24(6):259-64.B.C. OFFICE OF HEALTH TECHNOLOGY ASSESSMENT 43Vision ScreeningReinecke RD, Simons K. A new stereoscopic test for amblyopia screening. Am J Ophthal1974;78(4):714-21.Romano PE. Advances in vision and eye screening: screening at six months of age . Pediatrician1990;17:134-41.Rona RJ, Allsop M, Morris R, Morgan M. Referral patterns after school medical examinations.Arch Dis Child 1989;64:829-33.Rosner 1. Parents as screeners for strabismus in their children. J Visual Impairment and Blind1988 May;193-94. .Sanke RF . Amblyopia. AFP 1988 Feb;37(2):275-78.Schecter M, LeBlanc FE. Critical appraisal ofpublished literature. In : Hanstroidle W, Spitzer 0,McPeek B, Mulder DS, and McKneally MF, eds. Principles and practice of research: strategies forsurgical investigations. New York.Springer-Verlag, 1986 :112-117.Shaw DE, Minshull C, Fielder AR, Rosenthal AR. Amblyopia: factors influencing age ofpresentation. Lancet 1988 July 23;207-209.Simons K, Reinecke RD. A reconsideration of amblyopia screening and stereopsis. Am JOphthalmoI1974;78:707-13.Stewart-Brown SL, Haslum M. Screening of vision in school: could we do better by doing less?BrMed J 1988 Oct 29;297:1111-13.Stewart-Brown SL, Haslum Mbl, Howlett B. Preschool vision screening: a service in need ofrationalisation. Arch Dis Child 1988;63 :356-59.Sullivan M. Results in the treatment of anisometropic amblyopia. Am Orth J 1976;26:37-42.Thomas J, Mohindra I, Held R. Strabismic amblyopia in infants. Am J Optom & Physiol Optics1979;56(3): 197-201.TriefE, Morse AR. Strabismus and amblyopia. J of Visual Impairment and Blind 1988 Oct;327-30.TriefE, Morse AR. An overview of preschool vision screening. J of Visual Impairment and Blind1987 May; 197-99.U.S . Preventive Services Task Force. Guide to clinical preventive services: an assessment of theeffectiveness of 169 interventions. In: Report of the U.S . Preventitive Services Task Force. Guideto clinical preventitive services. Baltimore: Williams and Wilkins, 1989:181-185.B.C. OFFICE OF HEALTH TECHNOLOGY ASSESSMENT 45Vision Screening


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items