I NATIONAL CENTER Series 11 For HEALTH STATISTICS Number 101 VITAL amd WEALTH STATISTICS DATA FROM THE NATIONAL HEALTH SURVEY Visual Acuity of Children United States Vision testing methods with uncorrected monocular and binocular visual acuity findings by age and sex among children 6-11 years of age. DHEW Publication No. (HSM) 73-1266 U.S. DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE Public Health Service Health Services and Mental Health Administration Washington, D.C. February1970
42
Embed
Visual Acuity of Children the Master Ortho-rater, for the test-ing of visual acuity of children who could read letters. Since no target was available for te&ing
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
INATIONAL CENTER Series 11For HEALTH STATISTICS Number 101
VITAL amd WEALTH STATISTICSDATA FROM THE NATIONAL HEALTH SURVEY
Visual Acuityof ChildrenUnited States
Vision testing methods with uncorrectedmonocular and binocular visual acuity
findings by age and sex among children
6-11 years of age.
DHEW Publication No. (HSM) 73-1266
U.S. DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE
Public Health Service
Health Services and Mental Health Administration
Washington,D.C. February1970
Series 11 reports present findings from the National Health ExaminationSurvey which obtains data through direct examination, tests, andmeasurements of samples of the U.S. population. The reports publishedto date (Nos. 1 through 34) have related to the adult program. Additionalreports concerning this group will be forthcoming and will be numberedconsecutively, 35, etc. The present report represents the first of a largenumber of reports of findings from the children and youth programs,Cycles II and III of the Health Examination Survey. These reports,emanating from :he same survey mechanism, will be p~blished inSeries 11 but will be numbered consecutively beginning with 101, thepresent report. It is hoped this will facilitate the efforts to provideusers with all of the data and only the data in which they are interested.
Public Health Service Publication No. 1000-Series 11-No. 101
For sale by the Superintendent of Documents, U.S. Government Printing OfticeTVwMrgton, D. C., 20402- Price 45 cents
NATIONAL CENTER FOR HEALTH STATISTICS
THEODORE D. WOOLSEY, Director
PHILIP S. LAWRENCE, SC. D., Associate Director
OSWALD K. SAGEN, pH.D.,, Assistant Director /or Health Statistics Development
WALT R. SIMMONS, M. A., Assistant Director /or Research and Scientific Development
ALICE M. WATERHOUSE, M. D., Medical Consultant
JAMES E. KELLY, D. D. S., Dental Advisor
EDWARD E. MINTY, Executive O//icer
MARGERY R. CUNNINGHAM, h/ormatian O//icer
DIVIS1ON OF HEALTH EXAMINATION STATISTICS
ARTHUR J. McDOWELL, Director
PAUL T. BRUYERE, M.D.,Assistant Director
HENRY W. MILLER, Chic/, Operations and Quality cor?tro~ Branch
JAMES T. BAIRD, Jr., (%e/, Analysis and Reports lhzrzc~
PETER v. v. HAMILL, M.D.,Medical Advisor, Children & Youth Program
ARNOLD EN GEL, M.D.,Medical Advisor, Adult Program
JAMES E. KELLY, D.D.s.,Dental Advisor
HAROLD J.DUPUY, Ph.D.,Psychological Advisor
COOPERATION OF THE BUREAU OF THE CENSUS
In accordance with specifications established by the
N;ational Health Survey, the Bureau of the Census, under acontractual agreement, participated in the design and selee-tion of the sample, and carried out the first stage of the fieldinterviewing and certain parts of the statistical processing.
Public Health Service Publication No. 1000-Series 1l-No. 101
Vision Examination ----------------------------------------------------The Testing Instruments --------------------------------------------Testing Methods ----------------------------------------------------Quality Control -----------------------------------------------------
Findings -------------------------------------------------------------Binocular Distance Acuity -------------------------------------------Binocular Near Acuity ----------------------------------------------Monocular Acuity --------------------------------------------------Agreement Bemeen Acuity Measures ---------------------------------Acuity as Measured by Two Methods ----------------------------------
Comparison With Other Studies -----------------------------------------
Appendix I. Statistical Notes -----------------------------------------The Survey Desi~--------------------------------------------------Reliability ---------------------------------------------------------Sampling and Measurement Error ------------------------------------Small Categories ---------------------------------------------------
Appendix 11, Target Specifications for Vision Testing --------------------Scoring Sheets Used for Master Ortho-Rater Test ----------------------Scoring Sheets Used for Landolt Rtig Charts --------------------------Specifications of Optotype Size, in Snellen Notation, and Number of
Optotypes Per Line for Targets Usedin theSurvey --------------------
1
2244
557
101012
12
14
15
16
2929293031
323234
35
...Ill
iv
THIS REPORT CONTAINS national estimates based on findings pornthe Health Examination Suvvey in 1963-65 onuncovrected monocular andbinoculav visual acuity levels of children 6-11 years of age.
For this part of the survey a nationwide probability sample of 7,417 chil-dren was selected to represent the roughly 24 million noninstitutional-ized childven 6-11 years of age in the United States. Of these, 7,119 or96 @rcent wwe examined.
Monocular and binocular central visual acuity at distance and near weremeasured without cycloplegics for each examinee as part of the stand-ardized examination. Testing with a commercial instrument at 12 levelswus done without glasses for all children who could vead lette?%. Thosewho could not vead letters were tested on approximately comparableLandolt ring chavts.
Nearly three-fourths of these children have at least “normal” (20/20 oiubetter uision in Snellen notation) binocular distance vision without con-vection. The proportion reaching similav levels at neav is only slightlyless.
The proportion with defective binoculav distance vision is small butshows a remarkably consistent increase with age, year by yeaY, forboth boys and G“rls.
Boys were found to have slightly but signiwantly better binocular visualacuity at both distunce and near than G“vls.
Binocular vision tends to exceed the better monocular vision which intwmzgenerally tends to be better than that for either eye alone.
Closer agveement was found between binoculav and better monocularacuity than between the acuities for the two eyes. The degree of asso-ciation between distance and near acuity was found to be highly signifi-cant, but of a 10WW ovder than for the other measures at distance only.
SYMBOLS
Data not available ------------------------ ---
Category not applicable ------------------- . . .
Quantity zero ---------------------------- -
Quantity more than O but less than 0.05---- 0.0
Figure does not meet standards ofreliability or precision ------------------ *
VISUAL ACUITY OF CHILDRENJean Roberts and Kenneth R. Duya,n, Division Of Hmlth Examination Statistics
INTRODUCTION.
This report presents uncorrected monocularand binocular visual acuity levels of children 6-11years of age in the noninstitutional population ofthe United States as estimated from the HealthExamination Survey findings of 1963-65.
The Health Examination Survey from whichthese data derive is one of the major programsof the National Center for Health Statistics, au-thorized under the National Health Survey Act of1956 by the 84th Congress as a continuing PublicHealth Service activity to determine the healthstatus of the population.
Three different programs are used in carry-ing out the National Health Survey.l One of these—the Health Interview Survey—is concerned princi-pally with the impact of illness and disability uponthe lives and actions of people. It collects informa-tion from samples of people by household inter-view. A second—the Health Records Survey—consists of followback studies based on vitalrecords, institutional surveys to establish sam-pling frames and provide data, and surveys basedon hospital reco~ds. The Health Examination Sur-vey is the third major program.
Data are collected in the Health ExaminationSurvey by direct physical examinations, tests, andmeasurements performed on the sample popula-tions under study. This method provides the bestway to obtain actual diagnostic data on the preva-lence of certain medically defined illnesses. It isthe only way to secure information on unrecog-nized and undiagnosed conditions and on a varietyof physical, physiological, and psychological
measures within the population. It also providesdemographic and socioeconomic data on the sam-ple population under study.
The Health Examination Survey is conductedas a series of separate programs referred to as“cycles.” Each cycle is limited to some specificsegment of the U.S. population and to certainspecified aspects of the health of that population.Data in the first cycle were obtained on the preva-lence of certain chronic diseases and on thedistribution of various physical and physiologicalmeasurements and other characteristics in a de-fined adult population, as previously described.2’3
For the second cycle, on which this reportis based, a probability sample of the Nation’snoninstitutionalized children 6-11 years” of agewere selected and examined. The examinationfocused primarily on health factors related togrowth and development. It included examinationsby a pediatrician and a dentist, tests administeredby a psychologist, and a variety of tests andmeasurements by a technician. ‘This survey plan,sample design, examination content, and operationhave been described previously. 4
Field collection operations for this cyclewere started in July 1963 and completed in De-cember 1965. Out of the 7,417 children selectedin the sample, 7,119 or 96 percent were examined.This national sample is representative of theroughly 24 million noninstitutionalized children6-11 years of age in the United States.
Each child, during a single visit, was given astandardized examination by the examining teamin the mobile units specially designed for use in
“ the survey. Prior to the examination, information
was obtained from the parent of the child consist-ing of demographic and socioeconomic data onhousehold members as well as medical history,behavioral and related data on the child to be ex-amined. Ancillary data for the child on gradeplacement, teacher’s ratings of his behavior andadjustment, and health problems of the childknown to the teacher were requested from theschool he attended. Birth certificates were ob-tained for verification of the child’s age and factsrelating to the child at birth.
Statistical notes on the survey design, reli-ability of the data, and sampling and measurementerror are shown in appendix I.
VISION EXAMINATION
The vision examination consisted of tests todetect and classify color vision deficiencies,both monocular and binocular tests to measurethe level of central visual acuity at distanceand near, tests for lateral phoria at distance andnear and for vertical phoria at distance, a testfor bilateral accommodation at distance, and dis-tance and near tests for binocularity. Except forcolor vision, tests were performed without glassesor other refractive lenses for those who normallywore them.
In addition, each child was given an eye ex-amination by the survey staff pediatrician. Thisincluded a careful inspection of the eyes for evi-dence of styes, conjunctivitis, blepharitis, nys -tagmus, and ptosis as well as tests to detect thepresence and type of strabismus.
This report is limited to the tests of monocu-lar and binocular central visual acuity at distanceand near. As indicated, all tests were done with-out correction.
The Testing Instruments
The need for consistent uniformity in testingin addition to space and time limitations were de-termining factors in the selection of a commercialinstrument, the Master Ortho-rater, for the test-ing of visual acuity of children who could readletters. Since no target was available for te&ingthe near vision of children who could not readand since difficulties were encountered during thepretest in using the distance Landolt ring slide
in the enclosed instrument, special charts ad-ministered separately from the instrument at bothdistance and near were developed for use withilliterates to be tested in the survey.
Shown in figure 1, the Master Ortho-raterdevice consists of a viewing Imx and two illumi-nated slide holders with two sets of test slidesmounted inside the metal case. Slides used to testdistance vision are mounted on an illuminateddrum located at the right side of the instrument;those used to test near vision, on an illuminateddrum at the left. A spring switch holds each drumaccurately at each possible position. Only theslide in focus is illuminated. Without changing theposition of the head of the examinee, the viewingbox is tipped up to a set position for distanceviewing and down slightly into a set position fornear. The instrument is also adjustable for dif -ferences in eye height (above the chinrest). Dis-tance targets or slides are viewed at a distanceof 26 feet simulated optically by means of convexlenses; near targets, at 13 inches.
This instrument permits rapid testing undercontrolled conditions of lighting and target dis-tance from the examinee. The effective illumina-tion on the target and the contrast between targetletters and background were maintain&d withinoptimum limits for such tests. 5
Selected targets developed by Dr. LouiseSloan of the Wilmer Eye Institute at Johns HopkinsUniversity for the Armed ForcesB were used inthe Master Ortho-rater during the survey. Thesetargets on the slides in the instrument consistedof lines of optotypes which were letters appro-priately graded in size from one line to the nextand arranged in decreasing size from the top tothe bottom of the slide to test at 12 levels fromones corresponding to 20/12 to 20/400 (Snellennotation). These levels consisted of the equivalentsof 20/12, 20/15, 20/17, 20/20, 20/25, 20/30, 20/40,20/50, 20/70, 20/100, 20/200, and 20/400 at dis-tance with the 12 corresponding equivalents atnear. The 10 unserifed letters used were of nearlyequal legibility and were arranged in randomorder—differing for each line, each eye, and fordistance and near. As previously described, theseletters met the recommendations of the Committeeon Optics and Visual Physiology of the AmericanMedical Association.7~ 8 The letters followed theSnellen principle with their height as well as. .
2
Figure1. The visiontesting,
their width being five times the width of the linesin the letters. The targets consisted of 10 lettersper line arranged in groups of 5 each for testingfrom 20/12 to 20/200 and 3 letters at 20/400(and their equivalents at near), as shown in ap-pendix II,
The comparability between these letter slidesand correspondingly sized slides of Landolt ringsused in the Master Ortho-rater has been investi-gated. ~
For testing nonreaders at distance, a specialwall chart was developed for use in the survey. Thetarget followed Landolt ring dimensions with ap-propriate sized optotypes, which were rings, andrandom symbol arrangement in five-symbolblocks from the corresponding Armed Forcesplate. Contrast between background and symbolsmet the same specifications used in the lettertargets. The chart was laminated with nonglare
Table A. Proportion of children testedon Landolt ring charts, by age: HealthExamination Survey, 1963-65
Age Total Boys Girls
I Percent
Total, 6-11 years-1=1==+=
6 years -------------7 years -------------8 years -------------9 years -------------10 years ------------11 years ------------
75.9 78.927.6 2;.;
;:: 3:01.9 1.21.1 1.1
72.625.37.72.72.61:1
I II I
plastic. Test symixds arranged five to a line weregraded in size from one line to the next for test-ing at a distance of 10 feet from the child atlevels of 10/5.35, 10/7.15, 10/8.95, 10/10.7,10/12.5, 10/14.3 , 10/19.65, 10/25, 10/35.7, 10/50,and 10/ 100—corresponding approximately to dis -tance levels of 20/10.7, 20/14.3, 20/17.9, 20/21.4,20/25, 20/28.6, 20/39.3, 20/50, 20/71.4, 20/100,and 20/200.
The near chart for use at 14 inches for chil-dren who could not read letters was similarlyadapted from one developed at the Wilmer EyeClinic of Johns Hopkins University. The five-symbol lines on this chart were graded in sizefrom one line to the next for testing at levels of14/14, 14/17.5, 14/21, 14/28, 14/35, 14/42, 14/56,14/70, 14/87.5, 14/112, and 14/140.
In all, 81 percent of the children were testedon letter targets and 19 percent on Landolt ringcharts or cards. The proportion tested with Lan-dolt rings ranged from 76 percent among those 6years of age to 1 percent for the n-year-oldgroup as shown in table A.
With completely comparable size optotypes,Sloan et al.g have found a correlation of 0.90 atdistance between letter targets of the type usedinthis survey and comparable Landolt ring slides,both used in the Master Ortho-rater. They furtherindicate that since this correlation is of approxi-mately the same magnitude as that obtained be-tween test and retest scores on the same target,l”the two forms of acuity test may be consideredessentially equivalent.
It will be noted that the optotype sizes, foreither symbol or letter, differ slightly for 7 of the11 levels existing on both distance targets-theyare identical at 20/200, 20/100, 20/50, and20/25.Somewhat closer agreement was available on thenear tests where 7 of the levels were of identicalsize—the levels equivalent to 20/200, 20/100,20/50, 20/40, 20/30, 20/25, and 20/20.
When the distributions are combined as theyare in tables 1-4 for distance, this will mean thatthe proportions may possibly be slightly higherthan they would be if the targets had been com-pletely comparable at 20/70, 20/20, and 20/17 andslightly lower at 20/40, 20/30, 20/15, and 20/12.
Testing Methods
The testing order of right eye, left eye, andbinocular vision was maintained throughout thecycle. However, the sequence of near and distancetests was alternated for successive examinees.Testing order was randomized in this fashionto minimize any consistent bias for either testseries due to fatigue, practice, or learning thetarget letters. Every effort was used to secure amaximum level of participation on the part ofthe child so that the measurements would be asaccurate as possible. During the color test, whichpreceded the tests for visual acuity, it was pos-sible to determine without embarrassing the childby asking whether or not he could read letters.For those who could not read, visual acuity wastested on the Landolt ring charts as previouslyindicated. For the very slow reader, it was alsosometimes necessary to switch to the Landoltring charts to lessen the impact of fatique andloss of interest on the child’s performance, sincewith these charts the examiner had greater controlover the reading pace.
For children who could read well enough totake the regular examination, the following testprocedure was used. Each child was asked toread the line corresponding to an acuity level of20/30 (or the equivalent at near). If the childwas unable to do this with no more than the allow-able number of errors to’ ‘pass, ” he was presentedthe line corresponding to an acuity comparable to20/50. If the child again failed, he was started atthe 20/400 line. The child read the lines of pro-
gressively smaller letters until he failed or com-pleted the test.
In testing nonreaders and slow readers onthe Landolt ring chart or card, the examiner beganat the line with the largest rings (20/200 at dis-tance or the equivalent at near). The child wasasked to point in the direction of the “bite” inthe ring. The examiner continued this procedurefor the first ring on each line until one was missed.At that point the examiner went back to the pre-vious line and asked the child to read (point inthe direction of the “bite”) each symbol on theentire line. The same was done for the line orlines above and below it, to be sure the correctacuity level was determined.
To “pass” or to be able to read at a partic-ular level, no errors were allowed if the linecontained three letters, one in lines of five sym-bols, and three in lines of 10 letters. The visualacuity level or “score” for an examinee was thatwhich corresponded to the smallest letters orsymbols that the child was able to read with nomore than the allowable number of errors.
Quality Control
Vision tests were administered by the surveystaff examining dentist because that member ofthe survey team had the time available for them.The effect of this was to have these tests doneby a professional person who, once the necessaryspecial training had been given, was highly adeptat administering the tests. Each of the five dentistsemployed during the cycle was given training andpractice in vision testing techniques to insure theconsistency of test results. Further practice wasobtained during the “dry runs” preceding the startof the regular examinations at each of the 40 areasin which the mobile Health Examination Centerswere located.
A small-scale special study was conducted atDistrict of Columbia Village, a home for dependentchildren, prior to the start of the cycle to deter-mine the level of agreement between two of theexaminers in the testing of children. A high levelof agreement was found with more than three-fourths of the test results being identical or dif-fering by no more than one acuity level.
4
Midway during the cycle, at the Chicago lo-cation, replicate examinations were given to 95 ofthe 284 children examined there. The first testwas given by a different examiner than the second.Here a very high level of agreement was found onretest with nearly 90 percent of the tests differingby no more than one acuity level,
Test results appeared to remain consistentfor the various examiners throughout the cycle.The proportion of children rated as having normalor better vision showed essentially no differenceswhich might be attributable to the testers whenthe age and sex differences among examinees atthe various locations were removed.
Testing equipment and illumination werechecked periodically throughout the cycle to besure they were in good working order and met therequired standards.
FINDINGS
Binocular Distance Acuity
Nearly three-fourths (74.8 percent), or 19.5million, of the children 6-11 years of age in thenoninstitutional population of the United Stateshave at least “normal” or better than “normal”binocular vision at distance without correction,as estimated from Health Examination Surveyfindings in 1963-65. Nearly half are able to readat levels of 20/17 through 20/12 or better andabout 85 percent test at least 20/25 or better (table1 and fig. 2).
The median unaided binocular acuity was atdie 20/17.4 level. Hence, half of the children hadbetter than “normal” vision since they were able
Figure 3. Percent of children, 6-II years, reachingieve16 of 20/20 or better for binocular and monocu:Iar distance acuity, by age.
to read, at 20 feet, letters of a size that personswith what is generally considered to be “normal”vision could be expected to read at 17.4 feet, nearly3 feet closer to the target.
About 9 percent have moderately defectivebinocular vision, reading at best no further thanthe 20/30, 20/40, or 20/50 levels, while an ad-ditional 6 percent would test at the 20/70 level orless. Roughly 4 percent (3.6 percent) have acuityno better than 20/100 while less than 1 percent(0.8 percent) are unable to read at the 20/200level without correction.
The proportion of children unable to read atthe 20/200 level is too small to give a reliableestimate for this segment of the population. Yetit can be said with a fair degree of certainty thatthe actual proportion of children probably doesnot exceed 1 percent. This group will include thelegally blind, as well as those whose vision couldbe corrected to normal or near normal withlenses. However, neither the testing nor the ex-amination procedures in this cycle were sufficientto provide the basis for m&ing more preciseestimates of the prevalence of blindness sincethey did not include assessment of the limitationof visual fields or the degree of correctabilityof visual acuity.
No consistent pattern by age was foundamong those with at least normal vision, 20/20 orbetter fig. 3 and table 1). The proportions werethe lowest among the youngest-6 and 7 year olds—and the oldest age groups. It ranged from 72
percent for the 7-year age group to 78 for the 9year olds. However, the proportion with nearlynormal or better vision (20/25 or better) showeda slight, consistent, downward trend with age,which was more pronounced among the olderchildren, the proportions ranging from 88 percentat 6 years to 80 percent at 11 years.
At the lower end of the acuity scale a con-sistent increase in defective vision with age wasfound (fig. 4). For those testing no better than20/70, the proportion ranged from 1 percent at 6years to nearly 12 percent at 14 years of age.Or if the more seriously defective acuities areconsidered, the proportion ranges from less than1 percent at age 6 to nearly 8percent at 11 yearsat 20/100 or less and from less than 1 percent tonearly 6 percent at 20/200 or less.
Boys were found to have better binocularvisual acuity at distance than girls. About 77percent of the boys tested at least at the 20/20level compared with 72 percent for girls. The dif-ference in the proportion reaching at least the20/25 level—86 percent for boys for 84 percentfor girls— was less, but still differed more thanwould have been expected by chance in a sampleof the design and size of that used for the survey.At the lower extremes of the acuity scale, theproportion of boys was significantly less thanof girls (figs. 5 and 6; table 1).
16 -
14 -
12 -
10 -gu08
5‘e -
4 -
2 -
o_
.4’
Right Eye
.....
I I I I I I
6 7 a 9 10 II
AGE IN YEARS
Fi(JUt’6 L Percent of chiIdren,6-1I years, teStingno better than 20/70 for binocular and monoculardistance acuity, by age.
6
80
70
i-
60
50
0
BINOCULAR BETTER MONOCULAR
r
—
60
~o6 7 8 9 10 II
[,, ,,,
6 7 8 9 10 II
AGE IN YEARS
Figure 5. Percent of children, 6-II years, with “normal”or better distancevisualacuity(20/200r better),byage and sex.
No really consistent trend byage was found then a leveling off anddroponly at 11 years (fig. 5).among boys or girls with at least nearly “normal” At the lower end of the acuity scale a con-(20/25 or better) or “normal” (20/20 or better) sistent increase with age in the proportion withacuity. Among those with at least’ ‘normal” vision, unaided vision no better than 20/70 or 20/100, wasthe proportion for boys increased from ages 7 to found among girls. The trend with age also ex-9 years then dropped consistently from 9 through isted among boys except age ~ years (fig. 6).11 years. Girls showed an increase to age 8 years
7
16
12
8
4
0
[
BINOCULAR
-6 II
AGE IN
16 r BETTER MONOCULAR
oL—J_u46 7 8 9 10 II
YEARS
Figure 6. Percent of children, 6-11 years, with distance visual acuity of 20/70 or less, by age and sex.
Binocular Near Acuity
The distribution of “normal” or almost “nor-mal” near binocular acuity among children of allages was found to be similar to that at distance.Slightly less than three-fourths (72.6 percent)tested at least 14/14 or better, while 89.2 per-cent tested 14/17.5 or better (talie 5 and fig. 7).These proportions do not differ significantly fromthose for similar levels at distance— 74.8 per-cent reaching the 20/20 level or better and 85.2percent, the 20/25 level or better.
The proportion having moderately defectivevision was about the same for near as for dis -
tance vision—9 percent. These are the percent-ages testing 14/21 through 14/35 at near and thesimilar levels of 20/30 through 20/50 at distance.
Relatively fewer children were found to havemore severely defective near than distance vision,with 2 percent testing no better than 14/49 at nearcompared with 6 percent reading no better than20/70” at distanceO
By age there was a consistent increase in theproportion testing at least 14/14 or better atnear from age 7 years on (fig. 8 and table 5), atrend not found at the similar levels of 20/20 orbetter for distance vision.better levels, with a slight
Figure 7. Percent of children, 6-II years, reaching specified acuity levels for binocular and monocular near
vision.
I,,,,,,,,,,,,,,,,111111111118s1,s1
!:‘*”;o~
6 7 e 9 10 II
AGE IN YEARS
Figure 8. Percent of children, 6-II years, reachinglevels of 14/14 or better for binocular and monocu-lar near acuity, by age.
year-old group, the proportions at each age werenearly all the same, with no slight, downwardtrend as existed for the similar levels at distance.
At the poorer end of the acuity scale, 14/49 orless, no trend by age was found-the proportionsranged from 1.5 percent at 6 years to 3.5 at 11(fig. 9). This is in sharp contrast to the consistentincrease of poorer distance acuity with age. Fromage 9 years on, the proportion was significantlylower than for the similar levels at distance—20/70 or less (fig. 4).
Boys were found to have slightly better binoc-ular near visual acuity than girls, as was alsofound true for distance vision. The proportion ofboys with “normal” or better vision reaching14/14 or higher levels was significantly greaterfor boys (75.3 percent) than girls (69.8 percent).
Figure 9. Percent of children, 6-II years, testingno better than 1~/49 for binocular and monocularnear acuity, by age.
At the nearly “normal” 14/17.5 level or better,the differences were less, but still significant.However, the proportions of boys and girls test-ing at the poorer end of the acuity scale weresimilar—2 percent. No consistent pattern with agewas found among boys or girls with better orpoorer acuity (figs. 10 and 11).
Monocular Acuity
More than three-fifths (63 percent) of thechildren, or an estimated 16.4 million, had at leastnormal distance acuity in their better eye. Roughly
80 percent tested near normal or better at dis-tance (20/25or better). About 8 percent had pooracuities of 20/70 or less” (table 2).
With age no really consistent trend in theproportion of children with better acuities at dis-tance is found although there was a marked in-crease from 7 through 9 years in the proportionof children with at least normal distance acuityin the better eye. At the other end of the acuityscale, there is a steady increase with age in theproportion having poorer acuity (20/70 or less)—the rates range from 2.3 per 100 children at age 6to 13.7 at age 11 years (figs. 3 and 4).
At near, as for distance, 63 percent have atleast “normal” acuity in the better eye (14/14 orbetter) while a larger proportion than for dis -tance—87.1 percent compared with 79. Y—reach
at least almost “normal” vision at near (14/17.5or better). The proportion with poor near acui -ties— 14/49 or less-is somewhat lower than fordistance vision, 2.9 percent compared with 7.6percent testing no higher than 20/70 in the bettereye (tables 2 and 6).
By age the proportion testing 14/14 or higherin the better eye increases from 7 years on, butonly from 7 through 9 years if those with near“normal” vision are also included. No trend withage for the proportion with poorer acuities isevident.
Binocular acuity tends to exceed the bettermonocular acuity which in turn generally exceedsthe acuity for either eye. At di$tance about 12percent more children test at least “normal”with their binocular acuity than with their bettermonocular acuity, while monocular acuity in eithereye runs about 10 percent below the percentagefor the better eye. If the near “normal” group isincluded, the differences drop to 5 percent betweenbinocular and better monocular, and 6 percentbetween better monocular and either eye. Thedifferences between these acuity measures isnegligible at the lower end of the acuity range(tables 1-4).
At near the proportion testing at least “nor-mal” is about 10 percent higher for binocularthan for better monocular, which in turn is 10percent above that for either eye. If the nearly“normal” group is also included, the differencein proportion reaching these levels is negligiMebetween binocular and better monocular and dropsto 6 percent between better monocular and thatfor either eye. Differences here are also negligibleat the lower end of the acuity scale (tables 5-8).
The proportion reaching better and pooreracuity levels tends to be similar for the rightand left eye (tables 3, 4, 7, and 8).
Agreement Between Acuity Measures
At both distance and near, a high order ofagreement was found between the binocular andbetter monocular acuity scores of children (tableB). The correlation of the scores at both distanceand near was +0.94. The extent of agreement forboys and girls was similar. Over 75 percent
10
BINOCULAR BETTER MONOCULAR
eo 80
1- b70 sz $+ 70
:++
a ●+ ●●
+++ .+ ●
k’ 60 ●*●*$+*Girls
●
●●
50 50
o~ o~6 7 a 9 10 II 6 7 8 9 10 II
AGE IN YEARS
Figure 10. Percent of children, 6-II years, with “normal” or better near visual acuity (1~/lqor better), by ageand sex.
1-Zuveun.
BINOCULAR6
[
4
Girls
2
6 7 8 9 10 II
BETTER MONOCULAR6
r
4
1. J
Girls
●**......9....*””*=- ‘*m. ,,mm* .mm mm,
.+*
2 ,.+Boys
t-
0 I I I I 1. J6 7 8 9 10 II
AGE IN YEARS
Figure il. Percent of children, 6-II years, wi{h near acuity of lW/~9 or less, by age and sex.
L1
Table B. Correlation between variousmeasures of visual acuity used in thisstudy, by sex: Health Examination Sur-vey,- 1963-65
Acuity measures
Right and left eye:I)istance ---------Near -------------
Bimocular andbetter monocular:
Distance ---------Near -------------
Distance and near:Binocular---- ----Better eye-------Right eye--------Left eye---------
showed no difference inacuitylevel betweenbettermonocular and binocular vision.
The level of agreement between rightand lefteye acuities was also high, but of a markedlylower order of magnitude than that for binocularand better monocular vision. Better agreementwas found at distance (+0.78) than atnear (+0.57).No sexdifference wasfoundinthis measureeither.
Comparison of distance and near vision forthe various measures ofacuity shows asomewhatsimilar level of agreement for all four, rangingfrom +0.55 for better monocular acuity to+O.64for the left eye. These levels are substantiallylower than those for the right or left eyeatdis-tance but not near.
Acuity as Measured by Two Methods
In the foregoing sections of this report ithas been assumed thatthetestresults onthe lettertargets and the Landolt rings areroughlycom-parable for the levels discussed.
However, a further inspection of thedistri-butions for better and poorer acuities obtained
by both methods, asshown intables Canal 9-12,seems warranted.
It should be kept in mind that theoptotypesat the level equivalent to 20/21.4 are slightlylarger than those at 20/20 for a set distancebut offsetting this isthefact that only one errorwas allowed on lines of five symbols comparedwith three errors aHowed for passing lines of10 letters. Attheother levels theoptotype sizesare equivalent on the corresponding targets.
Relatively more children were found to haveat least normal acuity at distance on the Landoltrings (20/21 .4) than on the approximately cor-responding letter targets (20/20) at 6, 7, 8, and10 years of age while the reverse was found at 11years. At near, children throughout the age rangedid better on the Landolt rings. The reasons forthese differences are not readily apparent. Tosome extent they may reflect differences in thechildren tested or the differences in testing tech-nique in addition to the lack of complete com-parability in the targets. Only at 6 years and toa slight extent at 7 years will these differencesaffect the overall distributions of acuities.
COMPARISON WITH OTHER STUDIES
Strictly comparable findings for visual acuityamong children are not readily available frompublished reports. However, two large-scale stud-ies among selected groups do give the proportionof children in this age range who have “normal”or better vision.
Slataperll and Weymouth12 reported on thevisual acuity of patients in Houston, Texas,based on refractive examinations which includedmore than 1,700 among children 6-11 years of age.Morgan et al}3gave the results of vision testsbased on project-o-chart letters among some400 children 7-11 years, from two of the publicschools of Toronto, Canada. The latter group founda high correlation between tests and the project-o-chart letters and the Snellen letters (+0.96).Comparison of the findings from these threestudies are shown in table D.
12
Table C. Proportion of childrenas determined by Landolt ring1963-65
Level
Testing 20/20 or better on:Letter targets ---------------Landolt rings ----------------
Testing 20/70 or poorer on:Letter targets ---------------Landolt rings ----------------
Testing 14/14 or better on:Letter targets ---------------Landolt rings ----------------
Testing 14/49 or poorer on:Letter targets ---------------Landolt rings ----------------
examined and found to have betterand letter test, by age: Health
and poorer acuitiesExamination Survey,
All ages, 6 7 8 9 106-11 years
11years years years years years years
Percent
74.874.8
6.91.4
69.186.5
2.61.3
81.118.9
Table D. Proportion of children reachingvisual acuity levels of 20/20 or betterfrom selected studies? by age
1 UnitedAge States,
1963-651
6 years---7 years---8 years---9 years---
10 years---11 years---
1Private Schoolpatients children.
Hou;;on, To;&to,Texas Canada,19502 19523
Percentage
g ---:: 7251 8458 % 8358 69 8056 75 78
~Study by the HES.112study by Slataper. 13
3Study by Morgan et a10
67.774.1
2.21.2
42.586.3
3.0l.O
24.175.9
70.673.9
4.61.8
51.085,6
2.71.9
72.427.6
75.382.6
4.60.0
66.687.2
2.72.4
92.77.3
78.079.7
::;
73.992.7
$:
97.12.9
77*783.6
;::
77.993.3
2,20.0
98.11.9
73.260.5
11.411.8
78.885.8
::;
98.91.1
It might be expected that the findings of,.Morgan et al. 13 if their methods were’ similar,would be more nearly comparable to the presentfindings than those of Slataper. However ,thedatawould not suggest this. Sufficient informationisnot available to determine whether these sub-stantial differences result from the disparitiesin testing methods, the characteristics of thechildren, or both.
Only one available study investigated the re-lationshipof distance and near acuity measuresamong children. K~phart14in his study of nearly2,200 school children in grades three through12 found a slightly, but not significantly, lowerorder of agreement than was found in the presentstudy—t0.55 compared with +0.57. Agreaterdif-ference might have been expected sincehe usedthe checkerboard targets from the Ortho-raterrather than those used in the present study.
13
SUMMARY
National estimates based on Health Examina-tion Survey findings from a highly representativesample of children 6-11 years of age in the non-institutional population of the United States showthat:
1.
2.
3.
Nearly three-fourths, or 19.5 million, ofthese children have at least “normal” orbetter than “normal” (20/20 or better)binocular distance vision without correc-tion. The proportion reaching similarlevels at near is only slightly less.
No consistent pattern of better visualacuity at distance was found with age; atnear the proportion increased from 7years on.
The proportion with defective binoculardistance vision is small, but shows a re-markably consistent increase with age,year by year. At age 6, less than 1 per-cent tested no better than 20/100, whileby age 11, the percentage had increasedto nearly 8. This pattern prevails forboth boys and girls and is seen at thevarious poorer acuity levels—no betterthan 20/70, 20/100, or 20/200. A similartrend with age is not found for poorernear vision.
4.
5.
6.
7.
8,
Boys were found to have slightly but sig-nificantly better binocular visual acuityat both distance and near than girls.
Relatively more children were found tohave poorer acuity at distance than near—nearly 6 percent did not read above the20/70 level at distance compared with 2percent reaching just to the similar levelat near.
Binocular vision tended to exceed bettermonocular vision at both distance andnear. Slightly more than 60 percent had“normal” or better monocular vision com-pared with abut three-fourths for bin-ocular acuity.
The acuity in either eye tends in generalto be similar but somewhat less good thanthat for the better eye. The proportiontesting at least normal is about 10 per-cent below that for better monocularvision.
Closer agreement was found between bi-nocular and better monocular acuity thanbetween the acuities for the two eyes. Thedegree of association between distanceand near acuity was found to be highlysignificant but of a lower order than forthe other measures at distance only.
14
REFERENCES
lNa&ional Center for Health Statistics: origin, program,
wrd operation of the U.S. National Health Sumey. Vital rrndHmr?th Statistics. PHS Pub. No. 1000-Series l-No. 1. PublicHealth Sorvica. Washington. U.S. Government Printing Office,klK. 1963.
2National Center for Health Statistics: Plan and initialprogram of the IIealtb Examination Survey. Vital and HealthStatistics. PHS Pub. No. 1000&eries 1-No. 4. Public HealthService. Washington. U.S. Government Printing Offke, July196rl.
3National Center for Health Statistics: Cycle I of theHc?alth Examination Survey, sample and response, UnitedStates, 1960-1962. Vital and Health Statistics. PHS Pub. NIS.1000&!ries 11-No. 1. Public Health Service. Washington.U.S. Government Printing office, Apr. 1964.
4National Center for Health Statistics: plan, operation>
and response results of a program of children’s exsminations.Vital and Health Statistics. PHS Pub. No. 1000.Series l-No.L Public Health Service. W’asbington. U.S. Government Prinbing Office, Oct. 1967.
‘Norm, L. L.: hieasurement of visual acuity. A. M./LArcLOphth, 46:704-725, June 1951.
6Sloan, L. L.: New test charts for the measurementofviqua[ acuity at fnr and near distances. Arn.J.Ophth. 48(6Yt307-813,Dec. 1959.
7Ja~~on, E.; Black, M. hf.; Ewing, A. E.; Lancaster,
W. B.i and Fagin, R.: Committee on Standardizing TestCards for Visual Acuity. Transam%one of the Section ofOphthalmology of the American Medical Association, 1916.pp. 383-888.
8Jackson, E.; Black, M. M.; Ewing, A. E.; Lancaster,
W. B.; and Fagin, R.: Report of the American Committee onOptics and VisuaI Physiology: Report on Tests and Recordsof Visual Acuity. Transactions of ths Section of Ophthal-mology of the American Medical Association, 1930. pp. 355-363.
9sloan, L. L.; Rowland, W. M.; and Altman, A.: Compari-
son of three types of test target for the measurement of visualacuity. Quart. Rev. Oph$h. 8(1):4-16, Mar. 1952.
10Altman, A., and Rowland, W. M.: Measures of acuitywith optical simulation of distance. Qrzart. Rev. Ophth. 8(l):1-3, Mar. 1952.
llSlataper, F. J.: Age norms of refraction and vision.A.M.A. Arch. Ophth. 43(3):466-481, Mar. 1950.
12Weymouth, F. W.: Visual acuity of children, in M. J.i-lirsch and R. E. Wicks, eds., Vision of children. Philadel-phia. Chilton Co., 1963. pp. 119-145.
13Morgan, A. L.; Crawford, if. S.; Pashby, T. J.; andGaby, J. R.: A eurvey of methods used to reveal eye defectsin school children. Canad. M.A.J. 67:29-34, July 1952.
14Kephart, N. C.: Visual skills and their relation toschool achievement. Am.J. Ophth. 36:794-799, June 1953.
—000
15
DETAILED TABLES
Table 1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
Percentageof children,6-11 years, reaching or reaching and exceedingspecifiedacuity levels for uncorrectedbinocular distanceacuity, by sex and age: UnitedStates, l963-65-----------------------------------------------------------------
Percentageof children,6-11 years, reaching or reaching and exceedingspecifiedacuity levels for uncorrectedmonoculardistanceacuity in the,bettereye,by sexand age: United States, l963-65-------------------------------------------------
Percentageof children,6-11 years, reaching or reaching and exceedingspecifiedacuity levels for uncorrectedmonoculardistanceacuity in the right eye, by sexand age: United States, 1963-65-------------------------------------------------
Percentageof children,6-11 years, reaching or reaching and exceedingspecifiedacuity levels for uncorrectedmonocular distanceacuity in the left eye, by sexand age: United States, l963-65-------------------------------------------------
Percentageof children,6-11 years, reaching or reaching and exceeding specifiedacuity levels for uncorrectedbinocularnear acuity, by aex andStates, l963-65--------------------------------------------G--------?!?~-.!?~~~!
Percentageof children,”6-11years, reaching or reaching and exceedingspecifiedacuity levels for uncorrectedmonocular near acuity in the better eye, by aexand age: United Statea, l963-65-------------------------------------------------
Percentageof children,6-11 yeara, reaching or reaching and exceedingspecifiedacuity levels for uncorrectedmonocularnear acuity in the right eye, by sex andage: United States, l963-65-----------------------------------------------------
Percentageof children,6-11 years, reaching or reaching and exceedingspecifiedacuity levels for uncorrectedmonocular near acuity in the left eye, by sex andage: United States, l963-65-----------------------------------------------------
Percentageof children,6-11 years, reaching or reaching and exceedingspecifiedacuity levels for uncorrected binoculardistance acuity on the Master Ortho-rater letter slides, by sex and age: United States, 1963-65---------------------
Percentageof children,6-11 years, reaching or reaching and exceedingspecifiedacuity levels for uncorrected binoculardistance acuity on the Landoltringschart, by sex and age: United States, 1963-65-------.----------........---------
PercentageJofchildren,6-11 years, reaching specifiedacuity levels for uncor-rected binocularnear acuity on the Master Ortho-raterletter slides,by aex andage: United States, 1963-65-----------------------------------------------------
Percentageof children,6-11 years, reaching specifiedacuity levels for uncor-rected binocularnear acuity on the Landoltrings charts, by sex and age: UnitedStates, l963-65-----------------------------------------------------------------
Page
17
18
19
20
21
22
23
24
25
26
27
28
16
●
Table 1. Percentageof children,6-11years,reachingor reaching-andexceedingspecifiedscuityleve1s for un-correctedbinoculardistanceacuity,by sex and sge:UnitedStstes1963-65
6 years ------------------7 yeara------------------8 years ------------------9 yeara------------------10 yeara-----------------U years-----------------
Wk4e 2. Percentage of children, 6-11 years, reaching or reaching and exceeding specified acuit levels for un-??corrected monocular distance acuity in the better eye, by sex and age: United States, 1 63-65
Table 3. Percentage of children, 6-11 years, reaching or reathing and exceeding specified acuity levela for un-corrected monocular distance acuity in the right eye, by aex and age: United Statea, 1963-65
; aa------------------9 yeara------..-..-------10 years -----------------11 years-----------------
19
Table4. Percentageof children,6-lL years,reachingor reachingand exceedingspecifiedacuitylevelsfor ti-correctedmonoculardistanceacuityin the lefteye, by sex and age:UnitedStates,1963-65
6 years ------------------7 years------------------8 years ------------------9 years------------------10 years-----------------11 years-----------------
43.943,651.257,354.654.8
72.074,074.274.672.167.4
82.885.383.682.377.874.3
93.491.189.087.581.979.5
98.397.494.793.487.787.2
I
20
Table 5. Percentage of children, 6-11 years, reaching or reaching and exceeding specified acuity levels for un-corrected binocular near acuity, by sex and age: United Statea, 1963-65
14/14 1:;:9 Less14/17.5 14/21 14/2S 14/35 14/140 than
be%er 14/70 14/140
Percentage
72.6_
75.960.868.174.478.378.9
75.3
79.963.669.377.881.380.7
69.8
71.957.966.971.075.276.9
72.6
75.960.868.174.47s.378.9
75.3
79.963.669.377.881.380.7
69.8
71.957.966.971.075.276.9
16;6—
14.225.320.516.012.010.9
14.5
11.322.819.013.810.98.9
18.7
17.127.822.118.313.113.0
89.2—
90.186.188.690.490.389.8
S9.8
91,286,488.391.692.289.6
88.5
89.085.789.089.388.389.9k
4.9_
4.2
:::4.04.33.5
5.1
3.48.5
M3.53.”5
4.7
5.05.95.13.3
:::
2.t_
3.43.2
;:!2.E2.:
2.1
x1.61.81.91.9
3.5
0.8_
0.81.0
:::0.50.7
0.7
M0.90.5
::;
0.9
0.91.50.7
::;0.4
1,6_
1.21.71.21.11.52.9
1.6
R1.50.6
;::
1.6
1.51.60.9
:::2.5
Cumulative percentage
94.1
94.393.394.594.494.693.3
94.9
94.694.994.996.395.793.1
93.2
94.091.694.192.693,493.5
96.9_
97.796.596.597.397.495.8
97.0
97.997.196.598.197.695.0
1
96.7
97.695.996.496.797.196.7
97.7
98.597.597.398.197.996.5
97.7
98.697.797.498.698.195.9
97.6
98.597.497.197.897.697.1
99.3_
99.799.298.599.299.499.4
99.3
99.599.498.999.299.799.3
99.2
.00.099.098.099.498.999.6
0.5_
0.2
:::0.70.50.4
0.5
$$0.90.7
:::
0.6
0.;0.7
u0.3
99.8—
99.999.899.399,999.999.8
99.8
99.999,799.899.9100.099.7
99.8
100.099.898.7100.0;;.:
.
0.2—
0.10.2
::!0.10.2
0.2
R0.20.1
0.;
0.2
0.21.3
0.20.1
100.0——
100.0100.0100.0100.0100.0100.0
100.0
100.c1100.0100.0100.0100.0100.0
100.0
100.0100.0100. O100.0100.0100.0
21
Table6. Percentage of children, 6-11years, reachingor reachingand exceedingspecifiedacuitylevelsfor un-correctedmonocularnearacuityin the bettereye, by sex and age:UnitedStatea,1963-65
Table 7. percentage of children, 6-11 yeara, reaching or reaching and exceeding apecified acuity levels for un-corrected monocular near acuity in the right eye, by aex and age: United States, 1963-65
All ages, 6-11 yeara-------------------------- 100.0
66.544.545.6:3.;
61:9
52.4
62.041.445.751.755.259.0
84.077.377.982.384.681.4
79.6
81.776.376.479.481.882.3
91.087.”790.691.990.487.3
88.9
89.988.087.9Ss.189.789.8
95.394.294.095.694.891.2
94.2
95.493.793.893.094.894.3
96.395.795.596.896.393.2
95.8
97.796.195.194.695.795.4
6 years---------------------------------------------7 years---------------------------------------------8 years---------------------------------------------9 years ---------------------------------------------10 years--------------------------------------------11 years--------------------------------------------
U
All ages, 6-11 years-------------------------- 100.0
99.498.997.298.497.798.4
100.0100.0100.0100.0100.0100.0
6 yeara---------------------------------------------7 years---------------------------------------------8 years---------------------------------------------9 years ---------------------------------------------10 years --------------------------------------------11 years --------------------------------------------
23
Table 8. Percentage of children, 6-11 years, reaching or reaching and exceeding speci.fied acuity levels fcm un-corrected monocular near acuity in the left eye, by aex and age: United Statea, 1963-65
Sex and age
Acuity laval
14/14 14/49 Less14/17.5 14[21 14/28 14/35 and 14/140
Table 10. Percentage of children, 6-11 years, reachin or reaching and exceeding specified acuity levels for uncorrected bin-focular distance acuity on the L-redoc rings chart, by sex and age: United States, 1963.65
6 years-----------------7 years -----------------8 years-----------------9 years-----------------10 years----------------11 years----------------
4.83.6:.;
0:0
23.628.428.320.952.938.7
45.254.259.445.983.338.7
68.975.288.386.083.372.2
87.98S.595.193.087.9LOO.O
93.292.297.093.095.7100.0
98.196.7100.0100.0100.0100.0
99.498.5100.0100.0100.0100.0
99.899.3100.0100.0100.0100.0
99.899.3100.0100.0100.0100.0
100.0100.0100.0100.0100.0100.0
100.0100.0100.0100.0100.0100.0
26
Table 11. Percentageof children,6-11 years, reaching specifiedacuity levels for uncorrected:~ci&ar near acuity on the Master Ortho-raterletter slides, by sex and age: United States,
The sample design for the second cycle of theHealth Examination Survey, similar to the one used forthe first cycle, was that of a multistage, stratifiedprobability sample of loose clusters of persons inland-based segments. Successive elements dealt within the process of sampling are primary sampling unit(PSU), census enumeration district (ED), segment,household, eligible child (EC), and finally, the samplechild (SC).
At the first stage, the nearly 2,000 PSU’S intowhich the United States (including Hawaii and Alaska)has been divided and then grouped into 357 strata foruse in the Current Population Survey and the HealthInterview Survey were further grouped into 40 super-strata for use in Cycle II of the Health ExaminationSurvey. The average size of each Cycle II stratum was4.5 million persons,’ and all fell between the limits of3,5 and 5.5 million. Grouping into 40 strata was donein a way that maximized homogeneity of the PSU’S in-cluded in each stratum, particularly with regard to de-gree of urbanization, geographic proximity, and degreeof industrialization. The 40 strata were classified intofour broad geographic regions (each with 10 strata) ofapproximately equal population and cross-classified intofour broad population density groups (each having 10strata). Each of the 16 cells contained either two or threestrata. A single stratum might include only one PSU(or only part of a PSU as for example New York Citywhich represented two strata) or several score PSU’S.
To take account of the possible effect that the rateof population change between the 1950 census and 1960census might have had on health, the 10 strata withineach region were further classified into four classesranging from those with no increase to those with thegreatest relative increase. Each such class containedeither two or three strata.
One PSU was then selected from each of the 40strata. A controlled selection technique was used inwhich the probability of selection of a particular PSUwas proportional to its 1960 population. In the controlledselection an attempt was also made to maximize thespread of the PSU’S among the States. While not everyone of the 64 cells in the 4 x 4 x 4 grid contributes a
PSU to the sample of 40 PSU’S, the controlled selectiontechnique ensured the sample’s matching the marginaldistributions in all three dimensions and being closelyrepresentative of all cross-classifications.
Generally, within a particular PSU, 20 ED’s wereselected with the probability of selection of a particu-lar ED proportional to its population in the ,age group5-9 years in the 1960 census, which by 1963 roughlyapproximated the ~pulation in the target age group forCycle II. A simil~ method was used for selecting onese~”ent (cluster of households) in each ED. Each ofthe resultant 20 segments was eirher a lmmded areaor a cluster of households (or addresses). All of thechildren in the age range properly resident at the ad-dress visited were EC. Operational considerations madeit necessary to reduce the number of prospective ex-aminees at any one location to a maximum of 200. TheEC to be excluded for this reason from tie SC groupwas determined by systematic subsampling.
The total sample included 7,417 children in the 6-11 -year-age group with approximately 1,000 in each of thesingle years of age and from 25 different States.
Reliability
Measurement processes employed in the surveywere highly standardized and closely controlled. Ofcourse this does not mean that the correslxmdence be-tween the real world and the survey results is exact.Data from the survey are imperfect for three majorreasons: (1) results are subject to sampling error, (2)the actual conduct of a survey never agrees perfectlywith the desi~, and (3) the measurement processesthemselves are inexact even though standardized andcontrolled.
The first report on Cycle 114 describes in detailthe faitifdlness with which the sampling design wascarried out. It notes that out of the 7,417 samplechildreq, the 7,119 who were examined-a responserate of 96 percent—gave evidence that they were ahighly representative sample of children of this agein the noninstitutional population of the United States.The response levels for the various demographic sub-groups, including those for age, sex, race, region, popu-lation density, parent’s educational level, and family in-
come, show no marked differentials. Hence, it appearsunlikely that nonresponse could bias the findinga muchin these respects.
Measures used to control the quality of data fromthis survey in general have been cited previously; 4 thoserelating specifically to the vision tests are outlined inan earlier section of this report.
Data recorded for each sample child are inflatedin the estimation process to characterize the largeruniverse of which the sample child is representative.The weighta used in this inflation process are a prod-uct of the reciprocal of the probability of selectingthe child, an adjustment for nonresponse caaes, and apost-stratified ratio adjustment which increasea pre-cision by bringing survey results into closer alignmentwith known U.S. population figures by color and sexfor single years of age, 6-11.
In the second cycle of the Health Examination Sur-vey the sample was the result of three stages of selec-tion-the single PSU from each stratum, the 20 segmentsfrom each sample PSU, and the sample children fromthe eligible children. The probability of selecting anindividual child is the product of the probability ofselection at each stage.
Since the strata are roughly equal in populationsize and a nearly equal number of sample children wereexamined in each of the sample PSU’S, the sample de-sign ia essentially self-weighting with respect to thetarget population, that is, each child 6-11 years of agehad about the same probability of being drawn into thesample.
The adjustment upward for nonresponse is intendedto minimize the impact of nonresponse on final esti-mates by imputing to nonrespondents the characteristicsof “similar” respondents. Here “similar” respondentswere judged to be examined children in a sample PSUhaving the same age (in years) and sex as children notexamined in that sample PSU.
The post-stratified ratio adjustment used in thesecond cycle achieved most of the gains in precisionwhich would have been attained if the sample had beendrawn from a population stratified by age, color, andsex, and makes the final sample estimates of populationagree exactly with independent controls prepared by theBureau of the Census for the U.S. noninstitutional popu-lation as of August 1, 1964 (approximate midsurveypoint) by color and sex for each single year of age 6-11.The weights of every responding sample child in eachof the 24 age, color, and sex classes is adjusted upwardsor downwards so that the weighted total within the classequals the independent population control.
In addition to children not examined at all, therewere some whose examination was incomplete in oneprocedure or another. The extent of missing data forthe visual acuity test is shown in table I.
Sampling and Measurement Error
In the present report, reference has been made toefforts to minimize bias and variability of measure-ment techniques.
The probability design of the survey makes possiblethe calculation of sampling errors. The sampling erroris used here to determine how imprecise the surveytest results may be because they come from a samplerather than from the measurements of all elements inthe universe.
The estimation of sampling errors for a study ofthe type of the Health Examination Survey ia difficultfor at least three reasons: (1) measurement error and“pure” sampling error are confounded in the data—it isnot easy to find a procedure which will either completelyinclude both or treat one or the other separately, (2)the survey design and estimation procedure are com-plex and accordingly require computationally involvedtechniques for the calculation of variances, and (3) from
Table I. Number of examinees and extent of missing visual acuity tests, by age: HealthExamination Survey, 1963-65
Item missing
All examinees --------------------------- -
Not given vision battery-----------------------
Parta only missing:Distance
Right eye----------------------------------Left eye-----------------------------------Binocular----------------------------------
Total II 6 7 8 9 10 11years years years yeara years years
7,119
36
87
%
707260
1.111
12
434632
323527
Number of examinees
1,241—
10
201816
:;15
1,231—
4
1097
977
1,184—
4
z4
433
1,160—
5
676
565
1,192—
1
:4
233
30
the survey are coming thousanda of ,statistics, many forsubclasses of the population for which there area smallnumber of cases. Estimates of sampling error areobtained from the sample data and are themselvessubject to sampling error when the number of casesin a cell is small or even occasionally when the numberof cases is substantial.
Estimates of approximate sampling variability forselected statistics used in this report are presented intable 11, These estimates have been prepared by a repli-cation technique which yields overall variability throughobservation of variability among random subsamplesof the total sample, The method reflects both “pure”sampling variance and a part of the measurement vari-ance.
In accordance with usual practice, the intervalestimate for any statistic may be considered the rangewithin one standard error of the tabulated statistic, with68-percent confidence; or the range within two standarderrors of the tabulated statistic, with 95-percent confi-dence. The latter is used as the level of significance inthis report.
An overestimate of the standard error of a differ-ence d- ~- ~ of two statistics : and ~ is given bythe formula ~= ( S< + SYZ)1/2 where SXand SY are thesampling errors, respectively of x and Y, shown intable 11.
Small Categories
In some tables magnitudes are shown for cells forwhich the sample size is so small that ‘the samplingerror may be several times as great as the statisticitself. Obviously, in such instances the statistic has nomeaning in itself except to indicate that the true quantityis small. Such numbers, if shown, have been includedin the belief that they may help to convey an impressionof the overall story of the table.
Table 11. Standard error expressed in per-centage for percent of persons with a speci-fied acuity level: United States, 1963-55
Test and age
Both tes~
All ages,6-11 years---
6 years--- .---------,7 years -------------8 years -------------9 years -------------10 yeara ------------11 years ------------
Letter test
All ages,6-11 years---
6 years -------------7 years --------------8 years -------------9 years -------------10 years ------------11 years ------------
* U. S. GGVESN,SENT PRINTJNG OFFICE :1973 515-213/65 35
-—--..—
OUTLINE OF REPORT SERIES FOR VITAL AND HEALTH STATISTICS
Public Health Service Publication No. 1000
Series 1.
Series 2.
Sevies 3.
SeVies 4.
Sevies 10.
Series 11.
Series 12.
Series 13.
Series 14.
SeYies 20.
Series 21.
SeTies 22.
P)’o,gvams and collection procedures.— Reports which describe the general programs of the National
Center for Health Statistics and its offices and divisions, data collection methods used, definitions,and other material necessary for understanding the data.
Data evaluation and methods research. —Studies of new statistical methodology including: experi -
mental tests of new survey methods, studies of vital statistics collection methods, new analytical
techniques, objective evaluations of reliability of collected data, contributions to statistical theory.
Analytical studies. — Reports presenting analytical or interpretive studies based on vital and healthstatistics, carrying the analysis further than the expository types of reports in the other series.
Documents and committee reports.= Final reports of major committees concerned with vital and
health statistics, and documents such as recommended model vital registration laws and revised birthand death certificates.
Data from the Health Interview Survey. —Statistics on illness, accidental injuries, disability, use of
hospital, medical, dental, and other services, and other health-related topics, based on data collectedin a continuing national household interview survey.
Data from the Health Examination Survey. —Qata from direct examination, testing, and measure-ment of national samples of the population provide the basis for two types of reports: (1) estimatesof the medically defined prevalence of specific diseases in the United States and the distributions ofthe population with reqpect to physical, physiological, and psychological characteristics; and (2)analysis of. relationships among the various measurements without reference to an explicit finiteuniverse of persons.
Data from the Institutional Population .%rveys. — Statistics relating to the health characteristics ofpersons in institutions, and on medical, nursing, and personal care received, based on nationalsamples of establishments providing these services and samples of the residents or patients.
Data from the Hospital Discharge Sumey.— Statistics relating to discharged patients in short-stayhospitals, based on a sample of patient records in a national sample of hospitals.
tits on health resources: manpower and facilities. — Statistics on the numbers, geographic distri-bution, and characteristics of health resources including physicians, dentists, nurses, other healthmanpower occupations, hospitals, nursing homes, and outpatient and other inpatient facilities.
.Data on mortality. —Various statistics on mortality other than as included in annual or monthlyreports —special analyses by cause of death, age, and other demographic variables, also geographicand time series analyses.
Data on natality, marriage, and divorce. —Various statistics on natality, marriage, and divorce otherthan as included in annual or monthly reports —special analyses by demographic variables, alsogeographic and time series analyses, studies of fertility.
Data from the National Natality and Mortality Surveys. —Statistics on characteristics of births anddeaths not available from the vital records, based on sample surveys stemming from these records,including such topics as mortality by socioeconomic class, medical experience in the last year oflife, characteristics of pregnancy, etc.
For a list of titles of reports published in these series, write to: C)ffice of Information
National Center for Health StatisticsU.S. Public Health ServiceWashington, D.C. 20201
kJ.S. DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE
F
UBLIC HEALTH SERVICEealth Services and Mental Health Administration