Vision screening of illiterate populations

/ D0 { ITOBID IIEAIITIH ORGAIIIZATION Oachocerclasls Control Prograrme ln the Volta R:Lver Beeir Area IJJ t* \!) .,A ,r!: ./C .'iN otl Dpldeuiologloal Evaluatlon Unlt lechnlcal Docrment W\P/Ff[.fi 6 .e DR. B. IIffL,E:IORS VISION SCNE]NffiJIG OF IIJ.LITEI?ATE PO}UIJATIOTS a C) .-\ Ouagadougour May 1976 \) \ I IT o e / VISION SCREINrING OF II,IIIE}VITO POPI,I'A,TIO$S* by Dr. B. t\y1efors, OphthaS-nologlet Onchocerciasis ControL Programe i.:l the Volta Blver Baeln Area Ouagad,ougou, Upper Volta INTHODUCTIOTI Ihe present study deal-s with the introductton of the SJdgren Hard-Test as ar altencatlve to the E-Test for vislon screerriag of il.Lfterate populations. MATEIUI\IS AI{D IIEIHODS In moEt ophthalnological field surreya lt ls of great lnterest to test the visuaL acuity ln ord,er to estirate the amormt of reduced rislon ln a poBuLatlon. The most weLl-ktown and trled teet is certaln-\ir the illlterate E-test (Snelfen), ugett at 5 or 6 aeters distance ryith a metric or decinal sca1e, llhis teet however, qrthough very accurate, rngJr be cl.ifflcult for illiterate poptrlatlons to urderstand(nfoorc D65i Oanley 1971i Sarldes 1952) especial-\y if sultural and sooiolog:ical environmental factors are taken lnto consitleration (f.fppnann 1971i Ol-Iver, NawratslL t971i Sawlrlns 1964). [be experlence from a nrmber of fte]d sunreys in Ylest Af,rica durfng 1975 showetl that the llliterate E-Chart gave a relative\y high r.mtestabilityr 1.?. rllnabl1ity to leane the test arrt to glve re11abLe responsesrr (ltppnann 1971 ). Thus nuch lnfomation about the rrisual acuity ln some age groups was lost. Altenrative lLllterate teste have been used by several authors (gorg, Sunduark 1967; trEooks 1965i Koh1er, Stignnr 19771, Iippuann 1971i Nordlow, Joachl-msson 1962; hessl Austin 19@; SJUgren 1979). 1 o Untegt_a!_4l r An illlterate E-Chart (Aec:-uat scalee distance 5 neters) was used for visual acuity testing i:o flve rl11ages (fvory Coast a^nd Upper YoJ.ta). |Ihe test was lntro- cluced wlth e big netaLLlc E-test-[Ype in front of each case, w'ith explanations afil exercise a'rr also group tralnlng when possible to lmprove the cooperatlon. Each eye was tested separately without gLass comecti.on. The l-evels of, visua} aculty.examined.ln this etudy wer"r -tro (e/e)i or? (approx, A/g)i or3 (approx, 6/1a)onil 0r1' (6/60),[he teet was always Bezfo:med unler good drylfuht condltj.ons and the presentatlon of the test arranged to be as unlfozm as 1rcsslble, w"lth e naxi-urm of approrrmate\y 1O minutes spent for eaoh c&s€. fhe teet procdure was conlucted W one olhth41mis nurse or two tralned asslstants r:rrder cllrect strperv'lsion of the ophthaLtoologist, who participated, i-n testircg d,ifflcuLt ca,Ees. Parts of thie docrnent were pr€sentetl at the WHO late:reg"ional Meeting on the Preventlon of Bllrdness, Baghdad, 1976. t? -2- In five other viL1a6es (Ctrana and Upper Yolta) the Hand-fest lvas used.. The test-ca.nls were supplietl by Stil1e-Wernerf Sweden, in a stard.ard series comprising vlsual aculty levels ffia 5/5O io 5/7, metrj-c scale a'ral test ctistance 5 meterg. The test was presented in front of each patient with the 5/5O teet-type ard after explanatlons and tralnlng, the test was perforsned as w'ith the E-Test with corres- pording vlsual aculty levels exarnined. qpl test cond.itions. In order to obtai:a identical test conditlons for the E- ard lland-[est a set of separate E-{a,rtls, equal f,6 flqsimel ii-Chart (5 neters) were u6ed. lllre cantrs had a single test-type ln different sizes and the paper streets were ld.entical i.:a size and quality to the llanil4ards. Visua1 acuity teeting was conducted ln four vilLa6ee(enana antl Upper Volta) w'ith these E-Cards, using the sane mqrmer of presentation, tralning, test clistance arrl other conditions as for the Hand-test. Flnally, a sinple screerdr:g proced.ure was deslgned to cletect cases of visual impai-rment. llhe lland.-Test was used at 5 meters tllstance with presentatlon of thetest as mentioned above, but w:ith testing of only binocular rrision at the leveL of Or5 (e/rc and, 5/15), 0n1y cases that failed to pass this threstrold were then tested more in deta1l. Thls test procedure was conductetl in 20 villages in Northern Ghana. 2. Correlation between the E-[est and the Harrt-0est As the Hamd-Test has a less well-defined test-type than the geometrleal E, a comparison of the two tests was necessarSr to co:relate the Hand-fest for the actua]. levels of visr-nl aculty exqmined.. In flve villages (Cfrana and Upper Volta) both tests were performed with the saroe trntient with itlentical presentation of the tests, trainlngl test dlstance a,nc1 risuaL aculty IeveIs examined. Only the D-Test-Cards were used as E-[est, As the onler of the tests may be of i-mport&rrcel the E-[est was used as the first test, ar:d Yice verear irr an appro{i-mpte}y equal number of cases. Ihe requlrenents for a certaj.:o level of vlsual aculty were in accortlance rylth: 4 oonsecutlve co:rect posi- ti-ons, or 5 co:rect out of 5 shoninge, or 6 correct out of B strowlngs (Borg, Surrd.mark 1967i ltlordLow 1962). RESULT'S The populatlons examiraetl with the different methodologtes of rieuaL aouity testing are shovcn in table 1 . These populations resemble a stardant poprrlation, calculated fron this reglon (ltortnenc. Ghana a^nd Ivory Coastl l.XaJ.i and Upper Vo1ta). The age-groups correspond to those of epideruiological srurreyo of the Onchocercj-asis Control Progranme ln the VoLta River Basin Area. r+ Ad.drees r Bondegat.Zl 10251 S'fOCIfrOtM . Swed.en -5- The r'mtestablIlty rates obtainetl with the different methotls of visual acultytestr'n8 are ghown ln tabLe 2. rhere i,q a higtrty slgnlflcarrt dd;";;";;-fu.o;;i"between the totar results obtained with the ilriterate E-.cbart versus E-[est-cardsland also between both versions of E-Test against the Harid-fest. fhe d.lfference ismost prominent 11 the 5-9 yealps age-group but also evident ln feuales aged uorethan f0 /€&?so fhe two versions of ttaa-rest show no significarrt differenee concerFning untestability. lable 1. Poprrlation examined , by age-groups and sex 575 565 1 o9a Table 2. Untestability rates (It ) Age-Group E-[est Chart M rl E-Test Canls Hand-Test Hand-Test binocular soreenLng M F &[ F 22r5 o19 oro ' 0r7 1t1 72t5 1r1 1t1 3i 2r'l M F 5 -9 10-14 15-29 70-49 5o+ I 69,1 7rB or0 2t2 715 8512 16 15 711 17 tO 2019 50r, 215 oro 116 116 4411 4t3 4i 11r5 2016 21rD 1t5 or8 org 1r1 21 2 1 3 4 , , , , , 1 9 7 5 3 Totalx 1417 21rB 17fl y'o 1or 1 17 r7 1112 % 4t5 611 5rl % 416 5r2 4r7 % iland-[estAge-Group Hand-testblnocuLar sereening I,I F 5-9 10 - 14 15-29 7o-49 5O+ 84 102 BO 11a 5B 88 85 11V 17a 47 111 115 110 144 B9 96 88 176 151 71 579 599 621 655 758 525 449 827 727 746 M F }I F M F 1 1 27 1 6 1 1 09 20 ?2 50 77 25 63 E-Test Chart E-Test Cards 91 Bota-l 929 462 467 569 594 1 167 2772 2e70 5 642 x) Rates adjusted for a6e and sex t I t t tl I tI I t I I I I I I 1 rI I I a I -4- Itre results of the comparison of the E-fest (single test-typeu) d the Hanit- Test are shown in table 1a and. 7b. A prelj-rnlnary test of the E- arid I'IaaC-test scales was done in a p1lot vlllage (taUte 5a). lach eye is correlated separately amd the orig:lnaI levels of the llard-Test are used, as defined by SJUgren (tSlS). Thor.rg[ the colrelation 1s relative\y good there is an obvlous and systematio tendency for the Ha.nd-test to be easier than the correspording E-[est 1evels. [he llanct-Test scale was then modified so that the nearest ma]-l-er test-type in the avail-ab1e star',la^rrl serles was used for the various levels of lri.sual aculty. This cotresporv{s to one metrj-c step, or a factor of 0r5 - OrB to the varj-ous test-types. rTae 5/5 test-type was thus slrnnged to 5/4i 5/715 to->/Si Z/t5 to D/10 arfi,fir:ally 5/rO to 5fiot and this new scale was then co:reIated. to the same E-scale asbefore. The result is shown ln table Jb, which also shows the total nrmber of eyes exami.ned, as no ctifference was for,ral d.ependlng on the ord.er of the tests. It was howevel not poesible to double-test every case a^nd then only rrisual aoulty other than l rO (S/+ SJUgren) was verlfied with the other test in two vi[ages. Ttrig erplalnsthe hlgh proportlon of rrisual impainnent in the table and was necessary 1n order to obtaln a sufficient nr.mber of cases in each category. Tbe comelation 1n table 5b is very good but with a slight tenlency for the 5/,4 test-type to be more easily seen than tne t rO E-Ievel. T[e Hand-Card designed,5/7 was also used., but gave a very obvious tendency of being uore tllfflcu-Lt to see than the 1 r0 E-Ievel, and was therefore rejected. DISCT'SSIO1,{ [estlng of the vtsua]. aoulty le a complex situation where not on\r test-types or angles of vision detezuine the result, but also prychological factors play an i-mportant role (Borg, sr.rntmark 1967i nro6rc $65i lip-pmann 79e2, 1971i fvordfow 1962).lhe E-'Iest has been much i-n use end the untestabl].lty is general\r consi-dered to be Iow from the age of 44 years, though important variations are seen (liptrmaffr 1971i Nordlow, Joachi-msson 1962; Oliver, Nawratskl 1971 i Sew:Ltz et aI. 1964). llre untestabillty rates four:d in this study with the E-Test are relativelyhi&r especially when usi-ng a visual chart. fhere are very few eomparabLe reportsfron field sunreysr but some results from the Cemereons indlcate about 5 11 totattmtestability with the illiterate E-Chart (Atderson, fhglsang 1974). This mA)r how- ever be a question of different test procedure or population examlned.. ltre pattenn of untestablllW 1n the age-groups and relatecl to sex 1s the sa$e ln thls str.lrly for alL verstons of tests arrl is elso confirmed. fnon previous fieLd surveys (anAer-son,I\iglsang 1974i canJ-ey 197r). The naJor part of r.&testabLe cases is found in chil-dren Less tban 10 years olil, and also in nomen, especlally i-n the oLder age-gloups. The funportance of sho!1ing single test-types to improve the cooperation has been -9mnha9fer1.!1 pone authors (ffoofs 1965i KohILr, Stigmar 1969; Llppna:en 1g71r O1iver,Nawratski 1971) anct thls is also the experience j.n thls rtudy. The trisua,l chart has-the disadv.anta6e of belng dlffictrlt to undersf,nnd, but also w:tth very few test-typesfor the 6/60 Level. It caJx not be considered as sufflclent to diagnose a rrlsual I-npafu:nent upol on-\y 2-5 tested. positlons orur fjnger-cor:ntlng in these cases meansless accuracy (Canfq 877). Patients without v:isual impairilent rnay uansge with thefirst line by chartce, and then not be able to eooperate iurther. fhis means a rlsk of nisintertrrretlng cases of r:ntestabllity as a rrisu,al lmpairment. - -5- 1l: d1arrtaqe of Br'oup trainlng has been snderllnea (eanrey 1977 alrd, 1g75j Iaruttgen,pers. colm,. ) axtd this undoubtedly nay tnp:rove the cooperation but on\y to a f.im{tedertentr as maII ohiLdren a.rd slqr women are difficuJ.t to involve in the trainlng. 9**p txalning "1se JmPlles a need for a very good ard acttve Local interpreter/leader ln the vi1_Iage. the llfLuence of crrlturaL deprlvation, lsolation anrt a 1.ow social Level hasbeen potnted out in some papers (lippmann 1971i Ollver, IlarryratsH. 19I1i Sarkies l95liSawitz 1954) ard. these factors,ay b; of great lmportnnce but are dlfflcrflt to eya-luate ln afrlcan populatlons. However, it stroufa be erophasizecl that the vlLtages e:camlned ln thls sttdy are situatetl in renote areas end with a very low soclo-"corro-nic Ieve}. Ihe Har:d-[est has given ]-ower rrntestabillty rates ln th1s etuily than the E-Test a^nd this conftums the ldeas of some authors (rrotrc l.965i oliver; Nawratskl 1971)thougfu it has neYer been testetl in the fieltt. [he Hand-iest seens to be easler to understarrcl and the d.ifference ln shown tnterest and cooperatlon of the populatlon isstrildngly in favour of thi-s test. However, both the E- and the Har:d-Test have thedisadvanta€e that a developed spati.al onieniatlon j.s necessarlr (aorg, s*rarr.t 1967 jFfooks 1965i lippann 1971). . .T1r" SjUgren Hand-fest was lxtroduced Ln 1g5g ln a modified version where theprlnclple of Snellen has been used. [he size of the fingers ard their lrrtenrentngangles are calorrlated so that the ar:gle of vislon is one olnute wtren the canl 1s seen at a certaj.:e di-stance, narked on the backslcle (Sl6gren 1g7g). However, ,roion\y the fingers but al-so the shape 1n general are lnportant for thls test-type asthe oval aBpearance naJr facllitate to locate the ctirection of the iitg;;; (#8,Srrndnark 1967). - . m93 u!r.!8 the orlginal metrlc scale of SjUgren, it does not dlrectJy corres-pond to the E-Test levels exa,ulned.. fhe reduction of-the test-type slze one metricstep gives a Yery good cornelation to the E-Test, but this has been tested for onlgrfour levers of rrisual aculfir. It should be strong\r emihasized, that the copelationfound in this stu{y is emplrlcal, as complicatlng factors are involved. e.g.fls6imq1 aJal metric scales. However, for screening purposes of il].lterate populationsthe Harrd-Test has lrnportarrt advarrtages ard shows a comelati.on that 1s suffi-cient tope:mit categorJ.zing of vislon leveIs with a good, accuracy. SI]T,IMANY lwo verslons of the llliterate E-Iest have been used. together w:lth tJre SjdgrenHanl-Test for vlsr.pL acuity teetlng in op&rthaLuol-ogical field sunreys of illlteratepopulations In lfest Af:rica. lhe illlterate E-chart noqlr glve a high proportion of lmtestabllity because ofd'ifficultj-es to r.urlersta$d the test. rhis problen can be signlflcant]y reduced by Ssing a single E-lest-$rpr. on separate cartds. Ihe Hand-Tesf has glven slgnlfloarrtlyless untestabllity than both versi-ons of the E-test in thls studJr. It is posstble to co:relate the ilard-Test direct\y to the E-Test results forcerta:in levels of visual aculty ard using a slightly ,&itrea metrlc sca1e. the E-Test is still better defined and. preferable to use for a detailed. risualacuity testlng of a limited nr.mber of cases, iut for vislon screening of i-Lllteratepopurations i.n field surreys the Hand-Test is a better arternati.ne. -6- Iab1e 5. Comelation E-Test versus Hand-I est for for.u leve1s of visual aerri Table 5a: Oulginal metrie scale of Hand_Test (SSUgren) (each eye exami^ned comelated separately) E-Test Heud-Test HanI-[est or7 or7 0rl 726 )t 0 o 357 1 4 11 o 16 o 1 2 1 6 o o 4 4 127 36 17 7 783 Tabl-e 3b: Modified metric scale of Hanrt_Test (each eye exqmj-ned, correlated, separate\y) E-[est (r = or84) 1rO or7 B orl 0r1 ,/4 716 o 0 724 5/5 4 o B1 o o 117 o B5 5/10 o 117 5/zo o o o 36 76 720 117 962B9 ,6 (3 = 0199) 5/so I I I I o I I I I I -7 - NEFERENCES Anderoon, J et 41. (lSl+) str-rOies on onchocerciasis ln the Untted Ca:ueroonRepublic. Comparison of popul-atlons with ar:d without 0nchocerca volrmlus[raas. Roy. Soc. frop. Ivied. Hvs ,at @, 7, 19o-2o8 2. 3. 4. 12. 13. Borg, G. & sundrnalk, E. (tS0Z) A conparative str:dy of visuaL acn:ltJr tests forchlld.ren, Acta E:trthal., E, 1OD-117 trfooks, 0. (ts0l) vislon test for child,ren, Brit. J. ophthral ,1 491 112 Garrley, J.?. (Docr-ment llSl1) I{6thode de mesure de I' t6 visuelle Ie te:rai:r , 7pp. 5, Ganley, J.p. & Bi1es, J.E. (lSlZ) prevalence onchoc rrisua-l and IN the U of Ghana 3 ,p. fig. 6, Kohler, l. & Stigmar, G. a paediat .(lgeZ) vision screening of four-year-o1d chtLd,ren,Act 7. B. 9. 10. 11. al 14. 15. llplmaon, o. Ogez) ny" sereenlng, Archives of ophthaLmology, 68 November trlp1mann, o. (tglt) vtsion soreening of young chlrd.ren, 4.J.?.H., g]-, g Nordlow, yf. & Joaehimsson, S. UgAz) Screerring test for vlsr.raL acrrtty ln for.rr_year-o1d chiLdren, Acta onhuraL., 1[Q Oliver, M. & Nawratskl, tr. (lglt) screenlng of pre-sehoor chil-dren for ocurararromalies, BrLt. J. Ophthal.t 55t 462- ?ress, E. & Austtsr _9. (tSOe) screenlng of pre-school ctrildren for anb\ropla,JAI4A, 2041 9e 27 Mqy sarkles, J.II'n. (lgEz) ocurar onchocereiasis, Brit. J. ophthal., b g1 Saw-Itz, R.A., Reed, R,B. & Valadlan, I. (SA+) Testabllity of pre_school chlrclren for vlsion screenlng, J. pediat: ophthar., i, l|,l34, oct. sj6gren, II. (l7ll) A new series of test-canls for detennlrring visual aculty Inehlldren, Acta optrthal , t 1T Fr6vention de Ia cr6cit6. Rapport drrrn groupe drdtr:de d.e LtoMs. s6rie Rapp. [ech.,1977t 518, Genbve 1'