Clinical evaluation of the Taiwan trachoma control programme

Bull. Org. mtond. SantM 1971, 45, 491-509Bull. Wid Hith Org. Clinical Evaluation of the Taiwan Trachoma Control Programme F. A. ASSAAD,' T. K. SUNDARESAN,2 C. Y. YANG3 & L. J. YEH4 Following a clintical trial in 1959-60 and a prevalence survey in 1960-61, a comnmunity- wide trachoma treatmnent programmne was instituted in Taiwan. A reexamination in 1968-69 of a subsample of the localities included in the 1960-61 survey provided the basis for a clinical evaluation of the programme. Except in areas of very high endemicity in 1960-61, trachoma appears to have ceasecd to be, or is in the process of ceasing to be, an important public health problem. A reduction in the nuinber of cases that were active in 1960-61 has since resulted in reducing both the incidence and the severity of the disease. Cases showing signs of past experience of the disease (healed cases), as well as cases with grave irreversible sequelae or complica- tions, belong to the pretreatment cohort and are moving alonig "time" without being replaced. In communities with an initially very high endemicity it is open to question whether the incidence of the disease has been reduced to the level at which it will fail to perpetuate itself. Thte results of a very limited survey conducted in 1970 are encouraging: a repeat sample slurvey is planned for 1972. INTRODUCTION Following a preliminary pilot phase that lasted for 18 months, the Taiwan Trachoma Control pro- ject started in 1954 as a Government, WHO, and UNICEF combined effort. Its objectives were (I) to reduce the disease load to a level at which it would cease to be a public health problem and (2) to pre- vent disabling sequelae and complications. Up to 1962 the project consisted essentially of (a) case-finding among new entrants in primary schools throughout the island and collective treat- ment of all those having active trachoma and/or conjunctivitis, follow-up examinations at the begin- ning of each school year, and repeat case-finding in the sixth (final) year, with treatment of all chil- dren showing active signs of either disease; and (b) " blanket-treatment ", without case-finding or 1 Medical Officer, Virus Diseases, Division of Commu- nicable Diseases, World Health Organization, Geneva, Switzerland. 2Statistician, Health Statistical Metlhodology, Division of Health Statistics, World Health Organization, Geneva, Switzerland. 3Director, Trachoma Control Centre, Taiwan. ' Head, Department of Ophthalmology, Kaohsiung, Pro- vincial Hospital, Taiwan. follow-up, of family contacts in areas where school surveys in 1955 indicated a prevalence of 50% or more. Up to 1961 the treatment given consisted mainly of the local application of an antibiotic ophthalmic ointment twice daily for 48 consecutive working days. In 1959 a critical appraisal of the project indi- cated that, notwithstanding good organization and performance, the effect on the disease had fallen short of expectations. To determine whether the poor results were due to an initially low rate of cure or a high rate of reinfection after cure, or both, a research project was set up; this comprised a therapeutic trial in 1959-60 (Assaad, Maxwell-Lyons & Sundaresan, 1968a) and an epidemiological sur- vey (Assaad, Maxwell-Lyons & Sundaresan, 1968b) in 1960-61. The object of the clinical trial was to determine the treatment schedules appropriate to the local forms of disease and to estimate the cure rates to be expected in collective treatment operations. As a result of the trial it was established that (1) the 48-day continuous treatment so far adopted in the control project was inadequate, (2) the inter- mittent schedule was superior to the 60-day con- 2745 -491 492 F. A. ASSAAD AND OTHERS tinuous schedule, and that (3) within the intermit- tent schedule, tetracycline hydrochloride was supe- rior to chlortetracycline. In accordance with these results the tetracycline hydrochloride intermittent schedule of treatment was adopted, late in 1961, throughout the project. The object of the epidemiological survey was to define the nature and magnitude of the disease problem, to indicate effective approaches for control, and to provide base-line data for later evaluation. This survey disclosed markedly different levels of endemicity in different communities. The highest prevalence of active trachoma was found consis- tently in the 15-19-year age group. There were no differences between the sexes in the rates of active or healed trachoma. The survey also showed that trachoma in Taiwan was not a family disease, many households having only one trachomatous member. It thus became clear that the risk of infection was spread throughout the community and was maintained over a wide age span; hence the failure of school treatment, as a single measure, to con- trol the disease. In 1962 the project was therefore redesigned to include the whole community in the control operations (Taiwan Health, 1963; Gonzalez, 1965). Treatment schemes Two schemes of treatment were adopted: (1) Intensive case-finding followed by treatment of all active trachoma cases I in communities in which the 1960-61 survey indicated an active tra- choma rate of less than 300%, and (2) mass treatment without prior case-finding in communities in which the 1960-61 survey gave an active trachoma rate of over 30%. Throughout the operation of the programme the pre-existing school campaign was maintained and case-finding and treatment were extended to children in all six grades. The standard schedule of treatment consisted of two courses of intermittent treatment, each course consisting of the local application of 1% tetra- cycline hydrochloride ophthalmic ointment twice daily on 5 consecutive days, this 5-day cycle being repeated every 4 weeks to a total of 6 cycles. The second course started during the following year in the same month as the first. 1 The treatment was given on an individual basis except where two or more active cases were found in the same household, in which event all household members were treated. The control programme was carried out within the existing network of general health services, additional auxiliary personnel being temporarily employed as required to take part in the case- finding examinations and in supervising the treat- ment. At the Provincial Health Department the Trachoma Control Centre assumed the general direction and supervision of the programme. The case-finding examinations were carried out by local health workers trained specifically for the purpose and by college graduates temporarily re- cruited and trained. The treatment was adminis- tered in the community by adult members of the household, and in the schools by the class teachers. The treatment was directly supervised by " com- munity workers " temporarily recruited and trained for the purpose. Study population Communities on the island were divided, on the basis of administrative convenience, into four groups. Starting from 1962, one group was in- cluded in the programme every year. The opera- tion year coincided with the school year, i.e., from September to August. Approximately 10.6 million persons in about 1.9 million households were included in the community control programme (Taiwan Health, 1965; Taiwan Health, 1967; Taiwan Health, 1969). In communities receiving treatment No. 1, clinical examinations were made in almost 98% of the households and in over 95 % of the popula- tion. In all, 833 495 households (48.7%) were considered trachomatous. Nearly 600% of these households had one person in whom active tra- choma was diagnosed; the other 40% of the house- holds had two or more cases. In accordance with the criteria set out in the plan of operations, 3 091 011 persons (32.2% of the population) were found to require treatment. In communities receiving treatment No. 2, the plan of operations called for two courses of mass treatment after which the population was to be examined and active cases were to be treated as in treatment No. 1. Communities in the 1962-63 treatment group were examined in 1965. Of the 160 262 persons examined, 27 012 were considered to require treatment. They were distributed in 7 855 households of which 4 911 had one case of active trachoma and 2 944 had two or more. Because of the shortage of local funds the com- munities in the 1963-64 and 1964-65 treatment CLINICAL EVALUATION OF THE TAIWAN TRACHOMA CONTROL PROGRAMME 493 groups were not examined after the two courses of mass treatment. Instead they received one addi- tional course of mass treatment. Concerning the school campaign, the number of children enrolled at school increased from approxi- mately 1.9 million in 1962-63 to about 2.3 million in 1967-68. During this period the number requir- ing treatment decreased from about 500 000 to 260 000. Over 8000 of the persons in the community and over 95 % of the schoolchildren adhered to the prescribed treatment schedule. Untr-eated areas A number of localities selected at random,1 com- prising approximately 10% of the population on the island, were not included in the community treatment programme to serve as controls. How- ever, for ethical reasons the school campaign, established in 1954, was continued. These localities, therefore, served as controls for the assessment of the additional community programme as compared with the school campaign per se.2 Evaluation This paper deals with the clinical evaluation of the treatment programme. The evaluation was made late in 1968 and early 1969. The findings of the 1960-61 sample survey were used as the standard of comparison. The purpose was to determine whether the com- munity treatment programme had achieved its objective of significantly reducing the load of dis- ease in the population. The evaluation is based on the assessment of the changes over the period from 1960-61 to 1968-69 in the prevalence, incidence, and severity of the disease in communities of varying endemicity in 1960-61. METHODS The 1968-69 survey was limited to a subsample of the lins 3 included in the 1960-61 survey, the size of the subsample being determined by the 1 The island was stratified into 4 regions, northern, middle, southern, and eastern. One-tenth of the districts in cities, of the towns in county urban communities, and of the townships in county rural communities were drawn at random in each stratum separately. 2 In control communities 20-25 ° of children in all grades were found to require treatment annually. 3 The lin is the smallest administrative unit on the island, having on the average a population of 125 persons. amount of work the examining ophthalmologists could undertake. The following procedure was adopted for selecting the lins included in the evaluation. Lins in mountainous and outlying areas were excluded. The lins covered by the community treatment programme were then grouped by active trachoma prevalence rate in 1960-61 into 4 endemi- city groups, as shown below, and by year of commencement of treatment: Active trachoma prevalence rate (%) Group 1 <10 Group 2 10-<20 Group 3 20-<30 Group 4 >30 This gave 4 subgroups in each endemicity group, i.e., 16 groups in all. Approximately 60% of the lins in each of the 16 strata were drawn at random to form the subsample. Altogether 154 lins (popu- lation 20 550) in the treated communities were included in the evaluation (Table 1). For comparative purposes, lins in the communi- ties not included in the treatment programme, and which showed a change in population over the period from 1960-61 to 1968-69 not exceeding a 10% reduction or a 25 % increase, were included. Altogether 12 untreated lins (population 1 562) were included. The order in which the lins were visited was dictated by local convenience. Two ophthalmologists, one national and one a short-term WHO consultant, shared the clinical examinations. Half the households in each lin, selected by systematic subsampling, were allocated to each of the two examiners.4 Before embarking on the survey the two exam- iners, using the same criteria of diagnosis and classi- fication of trachoma (WHO Expert Committee on Trachoma, 1962; WHO Scientific Group on Trachoma, 1966-Annex 1) and adopting the same examination procedure (using focal illumi- nation and a monocular loupe), conducted a series of joint examinations to determine and resolve points of difference and ensure uniformity of diagnosis. ' The clinical examinations could not be completed dur- ing the stay of the short-term consultant; the national ophthalmologist continued the examinations after his depar- ture. Of the lins in the treated groups, 136 were examined by both ophthalmologists and 18 by the national ophthal- mologist alone. All of the 12 untreated lins were examined by both ophthalmologists. 494 F. A. ASSAAD AND OTHERS Table 1. No. of lins included in the 1968-69 evaluation survey Treatment Endemicity group group 1 2 3 Total 1962-63 lins 13 14 11 10 48 population 1 628 1 542 2 550 1 328 7 048 1963-64 lins 10 15 9 12 46 population 1 255 1 648 984 1 385 5 272 1964-65 lins 5 11 9 14 39 population 601 1 255 1 203 2 016 5 075 1965-66 lins 5 11 3 2 21 population 432 1 541 962 220 3 155 total lins 33 51 32 38 154 population 3 916 5 986 5 699 4 949 20 550 untreated lins 5 7 - - 12 population 503 1 059 - - 1 562 Neither of the two ophthalmologists who carried out the examinations in 1960-61 was available for the evaluation in 1968-69, but one of them was able to reexamine over 100 persons in one lin covered by both the 1960-61 and the 1968-69 sur- vey, and his findings were similar to those of the ophthalmologists who made the 1968-69 survey. From this and other evidence (Assaad & Maxwell- Lyons, 1967b) it is believed that the clinical findings discussed below were influenced very little by observer variation. RESULTS The survey started in September 1968 and was completed by the end of January 1969. Altogether 19 759 persons (96.20% of the population) in the sampled treated lins and 1489 persons (95.3% of the population) in the untreated lins were examined. The findings are presented in terms of: (1) Prevalence: total and active trachoma rates are given by age. (2) Incidence: the evaluation does not permit a direct measure of changes in incidence of the dis- ease. However, an indirect indication is provided by changes in prevalence in children under 10 years of age, i.e., by the difference between the prevalence rates among children born since the 1960-61 survey and those among children of corresponding ages born before the 1960-61 survey. (3) Severity: severity is measured by the degree of conjunctival and corneal involvement, the pro- portion of cases with disabling or potentially dis- abling lesions, and the mean score per case (see Annex 1). The results obtained in 1968-69 are compared with those in the same lins in 1960-61. Treated lins Prevalence. The changes 1 in the total and active rates are summarized below: Trachoina prevalence rates (%) Total trachoma Active trachonma 1960-61 50.0 21.8 1968-69 36.0 4.7 percentage reduction 27.9 78.4 The changes in prevalence in the 0-9-year age group are: Trachomna prevalence rates (%) Total trachomla A ctive trachoma 1960-61 16.5 61.1 1968-69 5.8 4.7 percentage reduction 65.1 70.6 1 Taking the number of trachoma cases in 1960-61 as x and that in 1968-69 as x' and the number of personis exam- ined in 1960-61 as y and that in 1968-69 as Y', the per- centage change is computed as follows: (Y 1X I= 1[ I(x x) CLINICAL EVALUATION OF THE TAIWAN TRACHOMA CONTROL PROGRAMME 495 The changes in prevalence for each age group (and by inference changes in incidence) are shown in Table 2 and Fig. 1 and 2. For total trachoma it is shown clearly that the younger cohorts born before 1960-61 were being replaced by new cohorts with a far lower prevalence rate. A marked reduction in active trachoma is seen in all the age groups up to 55 years. The reduction is less apparent among the older people. If the prevalence of active trachoma is subdivided into its two components, the non-cicatricial (Trl and TrIl) and cicatricial (Trlll) stages, it is seen that there is a marked difference in prevalence rates between 1960-61 and 1968-69; the load of Trl and Trll, however, is greatest in children between 2 and 16 years of age in both surveys. On the other hand, the distinct peak of Trlll that occurs around the age of 16 years in 1960-61 is not seen in the 1968-69 survey. The interpretation of this change in pattern is that a very high proportion of cured Trlll cases were not replaced by new cases. The prevalence rates for the four endemicity groups in the 1960-61 and 1968-69 surveys are shown in Table 3, and a reduction in the prevalence rates of total and active trachoma is seen in all groups, though it is relatively smaller in group 1. The reduction in prevalence rates in children below 10 years of age shows more or less the same pattern (Table 3). A study of the age-prevalence rates 1 shows that the shift along " time" of the cohorts born before 1960 was apparent in all groups but was again less marked in group 1. In 1968-69, in endemicity groups 1, 2, and 3, the prevalence of active trachoma was low in all age groups. In group 4, on the other hand, the active trachoma rate was 10-17% in age groups between 3 and 17 years; beyond that age a lower level prevalence was found. When active trachoma is broken down into its component stages, Trl plus Trll, and Trlll, it is found that the higher the endemicity the more marked the difference between 1960-61 and 1968-69 in both the Trl plus TrIl and the TrIII age-preval- ence rates. While the age of peak prevalence of Trl plus Trll was more or less similar for the two I The detailed tables of trachoma prevalence by stage and by age have been deposited in the WHO library. Single copies of the tables may be obtained on request to the Chief Librarian, World Health Organization, 1211 Geneva, Switzerland. survey periods, the Trlll prevalence rates in 1968-69 were very much lower than those for 1960-61 and do not show a distinct peak. It would therefore seem that irrespective of endemicity a high propor- tion of the cured Trlll cases were not replaced by new cases. Physical signs. The physical signs may be divided into conjunctival follicles and corneal infiltration (i.e., active lesions) on the one hand and conjunc- tival and corneal scars 2 (i.e., sequelae) on the other. Neovascularization of the cornea could be con- sidered as a sign of activity as well as one of the sequelae (in healed cases). Follicular involvement of the conjunctiva shows a definite shift towards mildness (Table 4). In par- ticular, the number of cases with an F3 diagnosis in 1968-69 was negligible (5) compared with the 129 cases in 1960-61. So far as corneal infiltration is concerned, 96.5 % of the cases of active trachoma in 1968-69 had no corneal infiltration, compared with 41.40% in 1960-61. On the other hand conjunctival scarring, an irre- versible sign, shows little change (Table 5). However, if the number of C3 cases is related to the total number examined, instead of to the num- ber of cases of cicatricial trachoma, a relative reduction of 46.2% over the period of study is seen. Furthermore, 64.70% of the C3 cases in 1968-69 were over 55 years of age, as compared with 45 % in 1960-61. Therefore, cases of excessive scarring are slowly being eliminated as they get older and are not being replaced by new ones. Corneal neovascularization shows the same trend (Table 6). Calculated as a percentage of the total population examined, the proportion of cases with pannus of 4 mm or more (V5) was 78.2 ,' smaller in 1968-69 than in 1960-61, whereas the propor- tion of cases with severe conjunctival scarring (C3) was 46.2% smaller. In 1960-61, 7.3% of the trachoma cases had dis- abling or potentially disablin{ lesions: in 1968-69, the corresponding proportion was 5.6%, a relative reduction of 22.1 %. The reduction in proportions is noted at all ages (Fig. 3). The frequency distribu- tion of cases by age shows a shift towards the older age groups in 1968-69 (Table 7), again demonstrat- ing a reduced rate of replacement by new cases. Further analysis by type of disabling or poten- 2Corneal scars are hereafter dealt with under " cases with disabling or potentially disabling lesions". F. A. ASSAAD AND OTHERS OD 00 N co 0 co 0) rN Nll Nt r-CN 0) CD N- 0) 6- 7 0 C*') C' 0) N- CD C)) CD~ r- 0 r- r- 0) 0) CD C) rN CY) CDco r-C') N- 0) a)r- co co C' CD 00 N CD CD co N CY) 0 0 C'i a) a) I'* CALO (O a) 0) 0) CD CD4 a) CD Lo M CO N (O 0CD N CD 0) 0)t o CD 0) cD CDN )Ns 0 r- Moo M o oo4 r LO oou:rsn aO o~ as cl r,- cl 9 N L C N OD CD q CD 0) Ns C t CD 0) CD CD N t N N t CD C' CD N CD N CD CD N 0) C__D N N N Ns CD C' 0 0) 0) CD 0) C) N N C' 0 C- C- CD .) 0 C- CD 0DCDD 0 0) N N N N 0) N 0 CD) D D 0) 0 __D~ N CD CD a o) 0 t 6 6 0 C>) Il CD O Nd- C') C') C' CD) CY) CY) CD No NNO C _ r- M CD 1- CDN- LO) coL Cl co tND CD CD LO LO 0) -I 0 1 N - m n 0 _ 0 CY) ,- N1 N co C') CD 0) 00 c'i CD N 0) N LO 00 co 0)CD N C') co 0 CD co CD 0 N 0 Lo co C' rN N o 0) C) N 00 CD0 CD L - r- m LO CD O CD- o _0 rN oo 0 N 0 0 NN ) N 0) N N C) C _L Coq 0 N ) 0 0) - C _ CD - ) C') N N N - ) 0 0 0) ' N Co N NC 0) _ D CD cD 0 0 0 0 CD 0 C C ) N 0 0 1 _ rN N rN q - C- N N N - 0)0 L N I N ' U) CD A^< o~~~ CI C')00 0 CDi CIO N CD CD4 CD CY) CY) N CD N CY) LO CD4 N CD) N CD N 0) Lo C') - LO LO r- 0) N N CD C) NI ID C)) N CD CD 6 N 0) N 0) N N 0 496 C.4 N- N~ CD CN CN N co - CD C) 6 co N CD CD0 CD0)c CD a) 6z__a)00 Z 0 CD a) a) co CD 0) -'- Z oCD 0) 14 i = o0 H _o CD a) C~CD 00 I l CD<) 0) co CD 0) a) 00 H.C CD 0)! Z00 10 0 z % a)> C.0a CD a) co C) a) CD Cn CD 0) co a) -0 CD CD .c CD -0 .0 0 C-) 0~ CD CD H-~ 0) 0 0 N1 2 .) CD C CD L0r) C n. ) CD 0 CD C D -C C CD _-C laO 'O C) .0 C - cn *1 0 CD 0 C) U- C0 I~Q CLINICAL EVALUATION OF THE TAIWAN TRACHOMA CONTROL PROGRAMME 0 5 10 15 20 25 30 35 40 45 50 55 60 65 Age (years) wHo io676 Fig. 1. Prevalence of total and active trachoma in relation to age: treated lins. tially disabling lesion reveals that the two features noted above-namely, the relative reduction (see below) and the shift to an older age group (see Table 7) were most marked in corneal lesions. Percentage of total trachomza cases Tcpe of lesion 1960-61 1968-69 F3C1, F2C2 or C3 6.0 4.0 trichiasis 4.8 5.2 pannus 4 mm 1.3 0.3 other central corneal opacities 3.7 2.3 total 7.3 5.6 70 60 In fact a higher proportion of trachoma cases had trichiasis in 1968-69-a relative increase of 8.1 %. However, when related to the total population examined a reduction of 22.0% in prevalence is noted. A summary picture of the severity of the disease shows that the mean score per case shows a reduc- tion from 4.4 in 1960-61 to 3.5 in 1968-69. The reduction is seen at all ages (Fig. 3). In the 1960-61 survey, the higher the endemicity the greater the degree of follicular involvement, i.e., Age ( years ) WHO 10677 Fig. 2. Prevalence of Tr I + Tr II and Tr Iil in relation to age: treated lins. 497 F. A. ASSAAD AND OTHERS Table 3. Prevalence of trachoma by stage, endemicity group, and year of survey: treated lins Endemicity group Characteirstic 1 2 3 4 1960-61 1968-69 1960-61 1968-69 1960-61 1968-69 1960-61 1968-69 All ages no. examined total trachoma no. % percentage reduction active trachoma no. percentag reductIon percentage reduction no. examined total trachoma no. percentage reduction active trachoma no. % percentage reduction 3 481 3 756 1 131 1 104 32.4 27.0 9.5 236 6.7 87 2.3 65.8 1 259 43 3.4 1 069 31 2.9 15.1 36 2.9 1 9 1.8 37.4 5 589 5 759 2 507 1 843 44.8 32.0 28.7 855 173 15.2 3.0 80.4 Age groups 0- 2 021 1 753 195 9.6 59 3.4 65.1 187 9.3 50 2.9 69.2 3 833 5 475 2 093 2 002 54.6 36.5 33.0 _ 956 236 24.9 4.3 82.7 9 years 1 482 1 816 277 86 18.7 4.7 74.7 271 1 8.3 69 3.8 4 229 2 830 66.9 1 689 4 769 2 169 45.4 32.0 435 39.9 9.1 77.2 1 568 529 33.7 522 33.3 79.2 1 533 179 11.7 65.4 153 10.0 70.0 14 -13 ___ 1960- 61 12 ____ 1968 - 69 1I 1 Mean score 0.00 Disabling lesions 7nQ .72 5 U 4Mt 3 2 0 Kt, Fig. 3. Proportion of trachoma cases with disabling or potentially disabling lesions, and mean score per case of trachoma, in relation to age: treated lins. 498 70 .- , 60 E0 = _ c5 31 - 4Z 0 <v = _ > 30 C Zq) = .2 a 20 cl, 10 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 Age (years) h I CLINICAL EVALUATION OF THE TAIWAN TRACHOMA CONTROL PROGRAMME Table 4. Follicular involvement of the conjunctiva in active trachoma cases by year of survey and endemicity: treated lins Total no. Degree of follicular involvement Endemicity S vy of activegroup ur ey trachoma FO Fl F2 F3 cases No. % No. NO. % No. % total 1960-61 3736 48 1.3 3050 81.6 509 13.6 129 3.5 1968-69 931 7 0.8 824 88.5 95 10.2 5 0.5 1 1960-61 236 2 0.8 212 89.8 22 9.3 - - 1968-69 87 1 1.1 84 96.6 2 2.3 - - 2 1960-61 855 13 1.5 755 88.3 78 9.1 9 1.1 1968-69 173 1 0.6 162 93.6 10 5.8 - - 3 1960-61 956 14 1.5 804 84.1 116 12.1 22 2.3 1968-69 236 2 0.8 202 85.6 30 12.7 2 0.8 4 1960-61 1 689 19 1.1 1 279 75.7 293 17.3 98 5.8 1 968-69 435 3 0.7 376 86.4 53 12.2 3 0.7 the proportion of cases classified as F2 and F3 was greater. In 1968-69, the same was true, but a very striking feature found in this survey was the reduc- tion in the number of active trachoma cases with F3 in group 4: only 3 cases of F3 among 435 active trachoma cases, compared with 98 among 1 689 in 1960-61 (Table 4). Concerning corneal infiltration, the percentage of cases diagnosed with I_>1 in 1960-61 shows a cor- relation with endemicity and ranges from 45.3% in group 1 to 64.7% in group 4. In 1968-69 the percentages varied from 1.2% in group 2 to 8.0 % in group 1-the substantial reduction thereby obscuring a relation with endemicity. Furthermore, in 1968-69 not a single case was reportedas having I > 2 as compared with 121 cases in 1960-61. Conjunctival scarring is correlated to endemicity in both surveys. The maximum percentage reduc- Table 5. Cicatricial involvement of the conjunctiva in cicatricial trachoma (Tr Ill + Tr IV) by year of survey and endemicity: treated lins Total no. Degree of cicatricial involvement Endemicity of I_ __I_ _I group Survey cicatricial c0 c1 C2 C3 j cases No. No. J No. % No. % total 1960-61 7 187 195 2.7 5 570 77.5 993 13.8 429 6.0 1968-69 6 612 358 5.4 954 74.9 1 034 15.6 266 4.0 1 1960-61 1 054 44 4.2 879 83.4 99 9.4 32 3.0 1968-69 1 058 90 8.5 849 80.2 94 8.9 25 2.4 2 1960-61 2 239 67 3.0 1 821 81.3 254 11.3 97 4.3 1968-69 1 734 96 5.5 1 343 77.5 247 14.2 48 2.8 3 1960-61 1 743 38 2.2 1 352 77.6 257 14.7 96 5.5 1968-69 1 873 79 4.2 1 414 75.5 295 15.8 85 4.5 4 1960-61 2 151 46 2.1 1 518 70.6 383 17.8 204 9.5 1968-69 1 947 93 4.8 1 348 69.2 398 20.4 108 5.5 499 F. A. ASSAAD AND OTHERS Table 6. Neovascularization of the cornea (all trachoma cases) by year of survey and endemicity: treated lins Total no. Degree of neovascularization Endemicity ofEndemicity Survey trachoma VO Vl, V2 V3, V4 V5 cases No. % No. % No. % No. % total 1960-61 8561 1 433 16.7 6854 80.1 163 1.9 111 1.3 1968-69 7 118 657 9.2 6 245 87.7 187 2.6 28 0.4 1 1960-61 1 131 235 20.8 874 77.3 12 1.1 10 0.9 1968-69 1 104 124 11.2 960 87.0 15 1.4 5 0.5 2 1960-61 2 507 404 16.1 2 054 81.9 31 1.2 18 0.7 1968-69 1 843 170 9.2 1 638 88.9 31 1.7 4 0.2 3 1960-61 2 093 336 16.1 1 699 81.2 31 1.5 27 1.3 1968-69 2 002 181 9.0 1 759 87.9 57 2.8 5 0.2 4 1960-61 2 830 458 16.2 2 227 78.7 89 3.1 56 2.0 1968-69 2 169 182 8.4 1 889 87.1 84 3.9 14 0.6 tion in C3 cases (41.5 % when related to trachoma cases, and 53.1 % when related to total population examined) is seen in group 4 (Table 5). The degree of corneal neovascularization is also, in general, related to endemicity in both 1960-61 and 1968-69; however, the reduction in cases of V5 is not related to endemicity (Table 6). In summary, it would seem that the reduction in the degree of conjunctival involvement, whether measured by the presence of follicles or of scars, was most marked in the group of highest endemi- city, group 4. The reduction in the degree of corneal infiltra- tion (see above) was very large compared with the reduction in conjunctival follicular involvement and is noted in all 4 endemicity groups. The less marked reduction in VS cases is similarly seen in the 4 endemicity groups. In both the 1960-61 and the 1968-69 surveys, the higher the endemicity the higher the proportion of cases with disabling or potentially disabling lesions. The reduction in the proportion of cases with disabling or potentially disabling lesions in 1968-69 is apparent in nearly all age groups. A study of the changes in frequency of cases with different dis- abling lesions (Table 8) shows that, except in group 1, trichiasis was seen in a higher propor- tion of trachoma cases in 1968-69 than 1960-61. However, if the number of trichiasis cases is related to the total population examined (i.e., prevalence) then the prevalence of trichiasis reduction in all 4 groups: shows a relative Endemicity Percentage change in the proportioni of group trichiasis cases in relationt to: Total no. of Total population trachoma cases examiiined 1 - 18.6 -26.3 2 +17.0 - 16.5 3 +15.0 -23.1 4 + 8.0 -26.4 A summary of the complex picture of changes. in the degree of conjunctival and corneal involve- ment and in the frequency of sequelae and com- plications is given by the mean score per case: Endemicity Mean score per trachonia case 1960-61 1968-69 1 2 3 4 3.3 2.8 3.8 3.3 4.4 3.6 5.4 4.1 Except for the youngest age groups, the reduction in the mean score per case is seen in all ages irrespec- tive of endemicity group. Untreated lins Altogether, 1 489 persons, or 95.3 % of the popu- lation, in the untreated lins were examined. In 1960-61, the untreated lins showed a lower active trachoma rate than that for the treated. For com- parative purposes a group of treated lins was formed by matching treated lins by endemicity in 1960-61 and the changes in the population since. This 500 CLINICAL EVALUATION OF THE TAIWAN TRACHOMA CONTROL PROGRAMME o - i 0 o NW a, (0 (a CC N N N N 7 l(a UL) 00 N oo O m m _ N N r _ N C) (0 LD (a a, ,.- -N N N N N (o 0 N N (a NN ) CD 0 (a N N N N CN a) a, (a0 N r- L6 00CO 0) CC) 0 0 0 0 0 N N(0 0 0 0o 0 CC 0 0 0 0 0 (D O N 0a _ _ C1N _ N t N N N C) T (D CD (aT (aN CC)s (a CN (a NI CC) (a CY) N1 a) N (a N' (a ' " a 0 a, a, m0 0 00 a, N N m0 0 0 m0 a 0 ) C. 6 N (D N 0 N N m 6 m0CC) N- CC Nl co (a , 't (a 0 aT N0 CCm CD r- N CN C 0 0 0 N ICY) 0 0 0 0 0 1.V k 11 - .J V ) (a I£) D£ 1D Lo u N N t (a aD - V Cu o6 a) (a)CD (D 0) C u) 0 -avtoU O0m, a) 0 O 0) 0) E C o mCD .C O a) C O0 11 mu- 2 group comprised 72 lins. In the untreated linls and the matched treated lins (m-treated lins) the total and active trachoma prevalence rates and the observed reduction therein are shown in Table 9. The differences between the two groups of lins are more marked among children of 10 years of age and below, and these figures are also shown in Table 9. The difference in the age pattern of the disease between the two groups of lins (Fig. 4) can be summarized as follows: (1) Total trachoma: the shift along " time " is less marked in the cohorts born before 1960 in the untreated lins. (2) Active trachoma: in 1960-61 a similar peak prevalence was observed in the age group 15 -19 years in both groups of lins. In 1968-69 the peak had moved into the 25-29-year age groups in both the untreated and m-treated lins. However, there was a distinctly greater difference in the levels in the two groups of lins. In contrast to the untreated lins where there is little difference between 1960-61 and 1968-69 Trn plus TrIl age-prevalence line graphs either in shape or magnitude, a time lag and a flattening of the 1968-69 curve are marked features in the in-treated lins (Fig. 5). Trlll age-prevalence line graphs give the same pattern in both groups but run at appreciably lower levels in the m-treated lins. The difference in the behaviour of Trl plus Trll and Trlll in the two groups of lins is reflected in the magnitude of the reduction: 1960-61 1968-69 reduction (%) TrachomaCa prer aleiCce r-ates ( °,) LUtitreated lins Trea(ted lins Ttr + Tr!I TriII Tri + TrI TrlII 3.7 7.7 3.8 8.4 2.6 2.6 28.8 66.0 1.6 1.1 58.8 86.4 Thus, a high rate of spontaneous cure of Trlll cases in the untreated lins is not matched (as in the treated tins) by an equally marked reduction in the incidence of new cases. Concerning the degree of involvement of the con- junctiva or cornea the main differences between the two groups can be summarized as follows. In pre- cicatricial cases (Tr +TrIl), a follicular involve- ment of the conjunctiva of F>2 was reported in untreated lins in 6 cases in 1960-61 and 5 cases in 1968-69 and by contrast in m-treated linis in co a) 0) '1 0) II 501 6 z 0 0 0 -6 Q) O: _ 0 -E E E c c i- Un ._1 co CC) 0 a N N CY, L) C11 CZ ~0 a) ao C,) 0 C/a a) .0 0 a) -0 co Cn V 3:.Z_ C .s a) 0 co 0._ E 0 0 c CU 0) C C . a) Ca E 0a CD a, 0 z 0 (O 0) 00 (a0 0)OC CD 0 C, - 502 F. A. ASSAAD AND OTHERS Table 8. Distribution of trachoma cases with disabling or potentially disabling lesions by endemicity group and year of survey: treated lins *~~~~~F Cl CetaIonaTotal Trichiasis Pannus > 1 mmToaEndemicity Survey F2 C2 or C3 cenral corneal Total gru Suvy trachoma F21 r~ pcte X ;cases No. % No. % No. % No. % No. % 1 1960-61 1 131 33 2.9 39 3.4 10 0.8 28 2.4 51 4.5 1968-69 1 104 26 2.4 31 2.8 5 0.4 19 1.7 33 2.9 2 1960-61 2 507 107 4.3 100 3.9 18 0.7 67 2.6 138 5.5 1968-69 1 843 49 2.7 86 4.6 4 0.2 34 1.8 88 4.7 3 1960-61 2 093 107 5.1 101 4.8 27 1.2 83 3.9 150 7.1 1968-69 2 002 88 4.4 111 5.5 5 0.2 47 2.3 120 5.9 4 1960-61 2 830 263 9.3 174 6.1 56 1.9 139 4.9 283 10.0 1968-69 2 169 121 5.6 144 6.6 14 0.6 67 3.1 162 7.4 ___ __._ 64 and 10 respectively. In cicatricial trachoma (TrIII4TrIV) a conjunctival involvement with scars of C .2 was diagnosed in the untreated lins in 79 cases (including 17 cases of C3) in 1960-61 and in 94 cases (including 10 cases of C3) in 1968-69. Table 9. Total trachoma and active trachoma prevalence rates in the untreated and m-treated lins Trachoma prevalence rates (%) Survey Total trachoma Active trachoma All ages-untreated lins 1960-61 40.9 11.5 1968-69 36.3 5.3 reduction (%) 11.3 53.8 All ages-m-treated lins 1960-61 41.0 12.4 1968-69 31.9 2.8 reduction (%) I 22.2 77.6 < 10 years of age-untreated lin 1960-61 5.3 4.9 1968-69 4.8 3.5 reduction (%) 10.9 29.5 < 10 years of age-m-treated lins 1960-61 7.6 7.2 1968-69 2.9 2.3 reduction (%) 61.8 67.8 In contrast, in the m-treated lins, the corresponding figures are 430 cases (including 118 cases of C3) in 1960-61 and 358 (including only 17 cases of C3) in 1968-69. The following numbers of cases were diagnosed in the absence of pannus. TrI-TriII TrIV 1960-61 1968-69 1960-61 1968-69 untreated lins 29 22 36 28 m-treated lins 256 78 289 156 The clearest difference between the two groups lies, however, in the frequency distribution by age of cases with disabling or potentially disabling lesions in 1968-69 (Table 10). In 1960-61 more than half the cases were in the age groups below 55 years in both m-treated and untreated lins. In 1968-69 the distribution was the same in the untreated lins while in the m-treated group almost 63 % of cases were in the age groups over 55 years of age. The contrast between the untreated and m-treated groups is even more marked in the age distribution of tri- chiasis cases in 1968-69 (Table 10). DISCUSSION The feature that stands out vividly from the 1968-69 evaluation is the shift of the load of disease to older age groups. This can best be illustrated by comparing total trachoma prevalence in the age groups 0-64 years in 1960-61 with that in the age groups 10-74 years' (i.e., the same cohort) 1 The shift of 10 years corresponds approximately to the interval between the two surveys. CLINICAL EVALUATION OF THE TAIWAN TRACHOMA CONTROL PROGRAMME 503 90019°o r TotalI Tr untreated lins 80 70- 70 f / _ __~~~~~~~~~~~~~~~~~~1960 -61 60- tQ60! __/ - ____~~~~~~~~~~~~~~~~1968-69 50- 5 40 0L 20 --10 X- Activ Tr_ 1 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 Age (years) 100 90 nm-treated lins TotalT80 _ 70- 60- 50- > 40 30- 30 10 - -Active Tr 0 - --- -- - 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 Age (years )WHo I679 Fig. 4. Prevalence of total and active trachoma in untreated and in m-treated lins in relation to age. in 1968-69. The results show, irrespective of a change in examiners, a rate of 48.8% in 1960-61 and almost the same rate, 49.3 %, in 1968-69.1 The prevalence of total trachoma indicates the load of infection the population has been sustaining. Since active cases constitute the source of infection, the prevalence of active trachoma reflects not only I As compared with a rate of 34.2 % in the age groups 0-64 years (in 1968-69). the present as well as the recent past experience of the disease but also gives an indication of the risk of infection to which the population is actually exposed. The 1968-69 evaluation was based on a repeat survey of a number of administrative units, the lins. No attempt was made to trace and follow up active trachoma cases over the period of 8 years and cure rates cannot, therefore, be computed. Never- F. A. ASSAAD AND OTHERS S e 4 w Iw 3 a, I I Ia , 4a) IO 10 0 ~0 ~0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 7 Age (years) ,r+ Tr N1.. 00ft 4-- -9 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 Age (years) wHo 10680 Fig. 5. Prevalence of Tr I + Tr II and Tr IlIl in untreated and in m-treated lins in relation to age. Table 10. Age distribution of trachoma cases showing disabling or potentially disabling lesions by year of survey: untreated and m-treated lins Cases with disabling or potentially disabling lesions Age group Total Trichiasis (years) 1960-61 1968-69 1960-61 1968-69 No. % No. % No. % No. % Untreated lins <55 20 58.8 13 54.2 1 5 51.7 1 3 56.5 >55 14 41.2 1 1 45.8 14 48.3 10 43.5 Total 34 100.0 24 100.0 29 100.0 23 100.0 m-treated lins <55 93 55.4 39 37.1 63 51.2 36 35.6 >55 75 44.6 66 62.9 60 48.8 65 64.4 Total 168 100.0 105 100.0 123 100.0 101 100.0 untreated lins 10 ____ 1968-69 103 ~TrlI4+ Tr II 0 0 504 10 5 CLINICAL EVALUATION OF THE TAIWAN TRACHOMA CONTROL PROGRAMME theless, a comparison of the findings in the age groups 0-64 years in 1960-61, and in the same cohort 10 years later, i.e., 10-74 years in 1968-69, gives an indication of the degree of reduction in the prevalence of active trachoma compared with 1960-61 (though probably an underestimate, as the number of new cases and reinfections among pre- viously healed cases cannot be assessed). A decrease in the prevalence of active trachoma from 22.3 % in 1960-61 to 4.7% in 1968-69 represents a reduc- tion of 79.7%. How much this reduction affects incidence can only be assessed indirectly by the difference in total trachoma among those born after the 1960-61 survey and those of corresponding ages born before. A reduction of 65.1 % is noted. The cohorts born before 1960 have therefore been replaced by a new cohort exposed to a far lower risk of infection. The lower the risk of infection, the shorter the duration of the active stages and the speedier the healing (Bobb & Nichols, 1969; Detels, Alexander & Dhir, 1966; Dhir, Detels & Alexander, 1967). It is of interest to note the difference in the number of healed cases in the children 0-7 years of age in both surveys. 1960-61 Total Tr Healed Tr 17 0 29 0 59 0 78 0 102 1 118 1 143 2 145 0 1968-69 Total Tr Healed Tr 0 0 2 0 20 4 30 2 32 2 49 12 45 11 61 10 It has been postulated that the cohort formed of those who were in the age group 15-19 years during the 1960-61 survey had been exposed at an earlier period to a relatively higher risk of infection and had been carrying with them their load of the dis- ease in the form of Trlll, hence the peak prevalence at the age of 16 years in the 1960-61 survey (Assaad & Maxwell-Lyons, 1966). It is of interest to note that, irrespective of the great reduction in TrIII, a peak is still discernible in the results of the 1968-69 survey, presumably from the same cohort, at the age of 24 years, i.e., 8 years later (Fig. 2). There is no change, however, in the prevalence of active trachoma in the age groups over 64 years- 3% in each survey. Thus it would seem that the oldest age group does not benefit from a reduction in the risk of infection and that the active cases in this age group mostly represent the old complicated cases carried over from earlier periods. The marked reduction in the degree of involve- ment of the conjunctiva by follicles and of the cornea by infiltration supports the obsei-vation made in an earlier study that a reduction in the risk of infection leads to a decrease in the severity of the clinical manifestations ofthe active disease (Assaad, Maxwell- Lyons, Sundaresan, 1968b). However, as is to be expected, a reduction in the risk of infection does not influence the prevalence of severe conjunctival or corneal scarring in those already affected. As occurs with other communicable diseases, the reduction in the number of cases was smallest in the group that already had the lowest endemicity. Moreover, it is in this group that the non-sampling errors (e.g., the missing of a few cases during the case-finding examinations or the non-adherance of some others to the prescribed treatment) and the sampling errors would have the most marked effect on the estimates. Furthermore, as they were the communities with the lowest trachoma rate, popula- tion movement would tend to have a larger effect on the disease picture. On the other hand, because of the higher socio-economic development in these communities, there is a greater chance that the effects of the treatment programme would be sustained, if not further accelerated. The programme has resulted in narrowing the gap in endemicity between the different communi- ties, in particular between the groups 1, 2, and 3. Endemicity Active trachoma prevalence rate (00) group 1960-61 1968-69 2 3 4 6.7 15.2 24.9 39.9 2.3 3.0 4.3 9.1 The reduction in disease endemicity in groups 1, 2, and 3 could lead to a sustained lowering of the risk of infection. On the other hand it is open to question whether the prevalence of recent and pro- gressive lesions (Trl and Trll) in group 4 in 1968-69 has been reduced to the level at which it will fail to perpetuate itself. The results of a very limited survey conducted in 1970 by one of the two ophthal- mologists who took part in the evaluation survey in 2 lins in group 4 are, however, shown in Table 11 and are encouraging. The overwhelming size of the treated communi- ties and their proximity to the untreated areas would affect statistical comparisons. Nevertheless, Age (years) < I<1 2 3 4 5 6 7 505 506 F. A. ASSAAD AND OTHERS Table 11. Survey of prevalence of total trachoma and active trachoma in endemicity group 4 Year of survey 1961 1968 1970 No. % No. % No. % total examined 221 100.0 250 100.0 196 100.0 total trachoma 190 86.0 159 63.6 99 50.5 active trachoma 150 67.9 60 24.0 27 13.8 valuable information is obtained from the follow-up of untreated lins. In 1960-61 the untreated lins gave a picture con- sistent with a low or moderate trachoma endemicity. The age-prevalence line graphs suggest that a ten- dency to a reduction in the risk of infection had already set in prior to the 1960-61 survey and that this tendency has been maintained. The prevalence of Trlll follows a course that had been anticipated in a previous study (Assaad & Maxwell-Lyons, 1966); that is, the shift of peak prevalence to older age groups, in fact to the 25-29-year age group. The reduction in the prevalence of Trlll, representing the great bulk of spontaneous healing, is not matched by a decrease in incidence of the same magnitude, as evidenced by a reduction of only 10.9% in total trachoma in children up to 10 years of age,' com- pared with a reduction of 61.8%in the m-treated lins. In this connexion, the fact that the main bulk of pre-cicatricial trachoma is in children of school age and that little difference is seen between the 1960-61 and 1968-69 surveys in prevalence rates in the untreated lins would support the assumption made that the school programmes per se in Taiwan were inadequate as a control measure (Assaad, Maxwell- Lyons & Sundaresan, 1968a). The striking feature in the untreated lins, however, is the continued replacement of cases with grave 1 A tendency for an accelerated healing is, however, noted. lesions, as evidenced by the absence of a change in the age distribution of these cases in 1968-69. In conclusion it can be said that the progressive change in the trachoma picture in young people, together with a concomitant sustained socio- economic development, would indicate that the reduction in infection will be maintained, if not accelerated, on the greater part of the island. In areas of excessively high endemicity in 1960-61 a very marked reduction in active trachoma has nevertheless left behind a moderate incidence of trachoma of moderate intensity. Considering that these communities are the ones that have had the smallest share of socio-economic development in Taiwan, the likelihood of a sustained lower trend is less certain. As part of the Trachoma Control Programme, the mass campaign in schools will be maintained up to the 1971-72 school year, the routine case-finding examinations acting as a surveillance measure. Furthermore, the 16 townships (hsiangs) in which school examinations in 1969-70 showed an active trachoma rate of 10% or more are considered as areas in which the disease still presents a public health problem. In 1970 the population in these areas will be examined. Two courses of treatment will be given to active (and suspect) trachoma cases. In 1972 a sample survey will be conducted. In addi- tion, case-finding and selective treatment have been instituted since late 1969 in the communities excluded from the original community treatment programme. ACKNOWLEDGEMENTS The authors acknowledge with sincere thanks the sup- port of the Commissioners of the Provincial Health Department, Taiwan, at the time of these surveys, Dr C. H. Yen and Dr T. C. Hsu, in carrying out these studies. They also acknowledge with gratitude the immense amount of work done by Dr S. P. Dhir in conducting some of the examinations, and the help of Mr P. Narishkin in processing the data. CLINICAL EVALUATION OF THE TAIWAN TRACHOMA CONTROL PROGRAMME 507 RESUME EVALUATION CLINIQUE DU PROGRAMME DE LUTTE CONTRE LE TRACHOME A TAIWAN En 1954, apres une etude pilote prealable, une cam- pagne de masse contre le trachome a ete lancee dans toutes les ecoles primaires de Taiwan. En 1959, on a soumis le projet a une evaluation critique qui a revele qu'en depit d'une organisation et d'un deroulement satis- faisants cette action n'avait pas donne les resultats attendus. En 1959/60, un essai clinique a montre la supe- riorite du traitement intermittent sur le traitement continu de 48 jours applique jusqu'alors. A la suite d'une enquete epidemiologique menee en 1960/61, il est apparu que l'ensemble de la population et des groupes d'Age autres que le groupe d'Age scolaire etaient exposes au risque d'infection. A la fin de 1962, on a decide de reviser le programme de lutte et de l'etendre a 1'ensemble des habitants de l'ile. Dans les collectivites ouj, d'apres l'enquete de 1960/61, la prevalence du trachome actif etait inferieure a 30%, le depistage a conduit a la mise en traitement de tous les cas individuels; dans les familles comptant 2 cas ou davantage, tous les membres ont ete traites. Dans les collectivites a prevalence superieure A 30%, on a instaure la therapeutique de masse sans depistage prealable. Le traitement a consiste en deux series de traitement inter- mittent A I'aide d'une pommade ophtalmique A 1% de tetracycline. Aux fins de comparaison, on a exclu de la campagne de masse 10% de la population totale de l'ile mais, pour des raisons d'ethique, le traitement des eco- liers, commence en 1954, a ete poursuivi partout. Afin de voir si le programme de traitement de masse applique A 1'ensemble de la population avait atteint son objectif, c'est-a-dire une reduction substantielle de la morbidite trachomateuse, on a reexamine en 1968/69 un sous-echantillon des localites incluses dans l'enquete de 1960/61. L'evaluation clinique a pris comme base hes changements intervenus durant la periode 1960/61- 1968/69 dans la prevalence, l'incidence et la gravite de la maladie. Dans les collectivites traitees, la prevalence globale du trachome actif est passee de 22,3 % en 1960/61 a 4,70 en 1968/69, soit une diminution de 79,7 %. Neanmoins le trait le plus caracteristique mis en evidence lors de l'enquete de 1968/69 est le deplacement de la morbidite vers les groupes d 'age superieur. La reduction de l 'inci- dence peut etre mesurde indirectement par la difference entre le nombre de cas de trachome total parmi les sujets nes apres l'enquete de 1960/61 et parmi les sujets d'age correspondant nes avant cette enquete. La reduction est de 65,1 %. 11 s'ensuit que les cohortes nees avant 1960 ont fait place a des cohortes exposees ia un risque d'infection de loin inferieur. Ce declin notable du risque d'infection a eu pour conse- quence d'attenuer dans une mesure importante la gravite des manifestations cliniques du trachome actif (follicules conjonctivaux et infiltrats corneens). Cependant, comme il fallait s'y attendre, la rdduction du risque n'a eu aucune influence sur la prevalence des sequelles graves (cica- trices conjonctivales et corneennes) chez les anciens malades. Le traitement de masse a eu pour rdsultat de reduire l'ecart existant entre les differentes collectivites sous le rapport de l'enddmicite trachomateuse, surtout parmi celles oi l'enquete de 1960/61 avait montre une prevalence faible ou moderee de l'infection. De la baisse de 1'endemicitd dans ces groupes de population pourrait rdsulter un declin permanent du risque d'infection. Il reste a voir si la prevalence des lesions recentes et evolu- tives dans les collectivites at endemicite initialement tres dlev6e a ete ramende a un niveau tel que la transmission serait interrompue. Les donndes recueillies 'a ce sujet en 1970 lors d'une enquete limitee sont encourageantes, mais ces populations continuent a beneficier du traitement. Dans les collectivites non traitees, on note une reduc- tion de 53,8 %Y de la prevalence du trachome actif. Nean- moins, le deplacement de la morbidite vers les groupes d'age superieur est moins net que dans les collectivites traitees, et la diminution, d'une enquete fa l'autre, du nombre des cas de trachome total parmi les enfants de moins de 10 ans n'est que de 10,9 %.. Le fait le plus saillant est le renouvellement incessant des cas porteurs de lesions graves. A la fin de 1969, on a mis en route un programme de depistage et de traitement selectif dans ces collectivites. REFERENCES Assaad, F. A. & Maxwell-Lyons, F. (1966) Bull. Wld Hlth Org., 34, 341-355 Assaad, F. A. & Maxwell-Lyons, F. (1967a) Amer. J. Ophthal., 63, 1327-1353 Assaad, F. A. & Maxwell-Lyons, F. (1967b) Bull. Wid Hlth Org., 36, 885-900 Assaad, F. A., Maxwell-Lyons, F. & Sundaresan, T. (1968a) Bull. Wld Hlth Org., 38, 565-576 Assaad, F. A., Maxwell-Lyons, F. & Sundaresan, T. (1968b) Bull. Wld Hlth Org., 39, 567-586 Assaad, F. A., Maxwell-Lyons, F. & Sundaresan, T. (1969) Bull. Wld Hith Org., 41, 181-194 Bobb, A. A. & Nichols, R. L. (1969) Amer. J. Ophthal., 67, 2 Detels, R., Alexander, E. R. & Dhir, S. P. (1966) Amer. J. Epidem., 84, 81 508 F. A. ASSAAD AND OTHERS Dhir, S. P., Detels, R. & Alexander, E. R. (1967) Amer. J. Ophthal., 64, 1 Gonzalez, C. L. (1965) Wld Hlth Org. Publ. Hlth Pap., No. 29 Kupka, K., Nizetic, B. & Reinhards, J. (1968) Bull. Wid Hlth Org., 39, 547-566 Reinhards, J., Weber, A., Nizetic, B., Kupka, K. & Maxwell-Lyons, F. (1968) Bull. Wld Hlth Org., 39, 497-545 Sowa, S., Sowa, J., Collier, L. H. & Blyth, W. (1965) Spec. Rep. Ser. med. Res. Coun. (Lond.), No. 308 Taiwan's Health, 1963 (1964) Department of Health, Taiwan Provincial Government, Republic of China Taiwan's Health, 1965 (1966) Department of Health, Taiwan Provincial Government, Republic of China Taiwan's Health, 1967 (1968) Department of Health, Taiwan Provincial Government, Republic of China Taiwan's Health, 1968 and 1969 (1970) Department of Health, Taiwan Provincial Government, Republic of China WHO Expert Committee on Trachoma (1962) Wld Hith Org. techn. Rep. Ser., No. 234 WHO Scientific Group on Trachoma Research (1966) Wld Hlth Org. techn. Rep. Ser., No. 330 CLINICAL EVALUATION OF THE TAIWAN TRACHOMA CONTROL PROGRAMME Annex I RECORDING OF PHYSICAL SIGNS Physical sign Conjunctival follicles (upper tarsal conjunctiva) Conjunctival scars Trichiasis and/or entropion Symbol F C Limbus and cornea: pannus vessels (measured from the upper limbus) infiltration jLlst beyond cor- neal vessels limbal follicles Herbert's pits corneal scars V I LF a Hpa Kop Degree of insolvement (1) involving less than 1/3 of the surface area (2) involving 1/3-2/3 of surface area (3) involving the entire surface area (1) deviation of upper tarsal conjunctival vessels, and/or fine scattered superficial scars in upper tarsal conjunctiva, or scars of any severity or extent in other conjunctival sites (2) moderate, readily recognizable scarring with no shortening or distortion of the upper tarsus (3) dense scarring of the upper tarsal conjunctival tissue (1) <1.0 mm extension (2) 1.0-<2.0 mm extension (3) 2.0-<3.0 mm extension (4) 3.0-<4.0 mm extension (5) >4 mm extension (1) minimal (seen only with monocular loupe) (2) barely recognizable by naked eye (3) dense opacification (1) one to three typical follicles (2) more than three, but not involving entire upper lunula (3) entire upper lunula involved (4) cornea encircled or two rows of follicles above (1) minimal, resulting in slight or no visual loss b (2) pupillary area involved, moderate visual loss (3) resulting in gross visual loss in one eye (4) resulting in gross visual loss in both eyes (eco- nomic blindness) Score Immature Mature follicles follicles 2 2 4 3 6 1 4 l 9 16 1 2 3 4 5 11 2 3 1 4 9 16 a Not included in the scoring system (Assaad & Maxwell-Lyons, 1967a). tb Assessed by objectise examination. 509