Epidemiologial significance of the local reaction to direct BCG vaccination

Bull. Org. mond. Sante 1969, 41, 45-62Bull. Wld Hlth Org. Epidemiological Significance of the Local Reaction to Direct BCG Vaccination As Assessed from a Study in Mongolia J. CHAVGANC,' R. HANAK 2 & H. G. TEN DAM3 Direct BCG vaccination has considerable operational advantages over the classical method of vaccinating only persons who do not react to tuberculin. In the present study it is shown that the direct method as applied in Mongolia does not cause any untoward reactions in persons who react to tuberculin and therefore can be considered a rational public health procedure. By reason of the repeated follow-up examinations, it was possible to obtain a clear picture of the development of the local BCG reactions and to test the hypothesis that these reactions may give information on pre-existing tuberculin sensitivity. It appeared impossible to deduce with any accuracy from either the local induration or the tissue destruction at the site of vaccination whether a person had tuberculin sensitivity before vaccination or not, and the results obtained with 2 different vaccines were inconsistent. It is therefore concluded that the local BCG reaction has no epidemiological significance in this respect. In many developing countries BCG vaccination is given by the direct method, i.e., without previous tuberculin testing. This method, which was recom- mended by the WHO Expert Committee on Tuber- culosis (1964), has considerable advantages over the classical 2-step procedure. In particular, it makes it possible to double the vaccination output and greatly increase the coverage. But the direct method may also have its inconveniences and, though they are certainly of a minor order of importance, these should be given due attention. In particular it should be ensured that the vaccination does not cause too much discomfort to those persons who happen to have a high degree of tuberculin sensitivity. Thus direct BCG vaccination should be restricted to the age-groups that include only a low proportion of these persons, and the effects of direct vaccination in these persons should be studied under local condi- tions. The present study deals especially with this latter aspect, but it was also designed to investigate the possibility of using the local BCG reaction for I Chief, National BCG Team, Republic Tuberculosis Dispensary, Ulan Bator, Mongolia. ' Senior WHO Medical Officer, Republic Tuberculosis Dispensary, Ulan Bator, Mongolia. 8 Scientist, Tuberculosis, Division of Communicable Diseases, World Health Organization, Geneva, Switzerland. determining the state of tuberculin sensitivity before vaccination. The latter information used to become available as a by-product of the tuberculin test when the population was screened before vaccination. The results of these tests were sometimes appreciated for the epidemiological information they could provide, and they were also used in assessment. Whereas accurate epidemiological information is nowadays obtained rather by applying special skin-testing techniques, the use of some test to select those persons who had little or no tuberculin sensitivity before vaccination remains of interest for certain assessment procedures. MATERIAL AND METHOD The study was carried out in adolescents and young adult students at the nursing school and the pedago- gical institute in Ulan Bator, Mongolia. Students who were known to have been BCG-vaccinated pre- viously or to have had tuberculosis were excluded. The students who were taken into the study were randomly divided into 3 groups. One group (about 20% of the study population) was given a placebo (saline). The other groups (approximately 40% of the study population each) were vaccinated with the International Reference Preparation for BCG (IRP) 2354 -45- 46 J. CHAVGANC, R. HANAK & H. G ten DAM TABLE I DISTRIBUTION OF THE STUDY POPULATION BY AGE, SEX AND VACCINE GROUP Age- ! Placebo IRP Glaxo vaccine group M F Total F Total M F Total(years)! 10-14 1 1 1 1 15-19 39 121 160 109 224 333 95 233 328 20-24 29 34 63 63 86 149 65 65 130 25-29 18 14 32 14 15 29 10 18 28 30-39 13 4 17 28 17 45 28 14 42 40-49 4 1 5 3 1 4 Total 99 173 272 28 34 562 201 32 53 and with the routinely used Glaxo vaccine (batch T 22 F) respectively. A dose of 0.1 ml was injected intradermally in the lower part of the deltoid region of the left arm. The distribution by age, sex, and vaccine group of the students given the first tuberculin test and actually vaccinated is shown in Table 1. At the time of vacci- nation all students were given a tuberculin test with 1 TU of RT 23 with Tween. Follow-up examina- tions were carried out 3, 4, 6, 9, 12, 18, 20, 23 and 26 days after vaccination, and also 12 weeks and 51/2 months after vaccination. At each examination the site of vaccination was inspected. If induration was present its transverse diameter (mm) was recorded. The type of the local tissue destruction, if any, was recorded (e.g., unperforated pustule, open or covered ulcer, scab, fresh scar) and its transverse diameter (mm) was also recorded. Retesting with tu- berculin (2 TU ofRT 23 with Tween) was performed at both the 12-week and the 51/2-month examina- tions. At these times the cervical and axillary regions of all the pupils were examined for lymph node enlargement. If lymph nodes were found that were larger than 1 cm in diameter, their number and size were recorded. Tuberculin testing and BCG vaccination were performed intradermally, according to the recom- mended techniques. 1 The tests and vaccinations were performed by 2 experienced nurses. These nurses also performed the reading of the tuberculin tests and 'Unpublished working documents WHO/TB/Techn. Guide/2 Rev. 4.65 and WHO/TB/Techn. Guide/3. A limited number of copies of these documents is available to persons officially or professionally interested on request to Distribu- tion and Sales, World Health Organization, 1211 Geneva, Switzerland. the examination of the local reactions. Each nurse performed all examinations of a particular student. For this purpose the study population had been divided into a number of groups that were randomly assigned to one of the nurses. However, owing to the temporary absence of one of the nurses, most of the local reactions were examined only by the other. The examinations for lymph node enlargement were performed by a physician. RESULTS Initial tuberculin sensitivity The results of the first tuberculin test are shown iA Fig. 1, for all vaccine groups combined. It is seen that before BCG vaccination almost half the study population reacted to tuberculin, and that this group of reactors can be clearly separated from the group of non-reactors, the proportion of persons with low- grade tuberculin sensitivity being small. Persons with a tuberculin reaction of 10 mm or more will therefore be considered as reactors (R in the tables), and those with less than 10 mm as non-reactors (NR in the tables). Post-vaccination tuberculin reactions The tuberculin reactions 12 weeks and 51/2 months after vaccination are shown, separately for the differ- ent levels of initial tuberculin sensitivity, in Table 2. It is seen that for the non-reactors (especially the 0-1-mm group, which comprises 42% of the study population) there is a clear difference between the reactions in the placebo group and those in the vac- cinated groups. The differences are statistically highly significant, which may be taken to indicate that both SIGNIFICANCE OF LOCAL REACTION TO DIRECT BCG VACCINATION vaccines produced tuberculin sensitivity in the non- reactors and therefore that vaccination was carried out effectively. The vaccinated reactors showed FIG. I PREVACCINATION TUBERCULIN REACTIONS TO 1 TU OF RT 23 WITH TWEEN IN THE ENTIRE STUDY POPULATION 0 2 n = 1367 4 6 8 LU IZ 14 lb 15* ZU ZZ Reaction size.(mm) I 28 30 WeO 91911 tuberculin reactions that were similar to those in the reactors given the placebo. Local vaccination lesions The means of the local indurations measured on the various examination days are shown in Fig. 2, separately for non-reactors and reactors, and for the IRP and Glaxo vaccine. No induration was noted in the placebo group. It is seen that the induration is largest during the first few days and subsides rapidly. In the reactors it is larger than in the non-reactors, and resembles the tuberculin reaction. No induration is found after 12 weeks or 5½/2 months. The correla- tion of local induration and prevaccination tubercul- in sensitivity has been summarized in Table 3. The distributions for the various days of examination (not shown) revealed that no excessive induration occurred. The development of the local tissue destruction, expressed in terms of the size of its transverse diameter, is shown in Fig. 3. It may be seen that in reactors the tissue destruction starts earlier than in the non-reactors and on the whole is somewhat larger. It should be noticed that there is a marked difference between the two vaccines. TABLE 2 MEAN 12-WEEK AND 5 :2-MONTH SUBREACTIONS FOR THE DIFFERENT LEVELS OF PREVACCINATION SUBSENSITIVITY Prevaccination Mean 12-week tuberculin reaction Mean 5 A-month tuberculin tuberculin reaction reaction(mm) Placebo IRP Glaxo Placebo IRP Glaxo 0-1 2.3 19.4 17.7 2.9 18.4 16.8 2-3 3.9 19.5 19.7 7.5 19.3 18.1 4-5 10.6 21.4 22.3 3.0 19.2 20.3 6-7 17.0 21.1 21.6 17.0 20.0 21.6 8-9 15.0 22.0 21.0 20.0 20.0 21.1 10-11 19.0 19.9 21.9 21.8 20.1 20.3 12-13 18.5 20.9 22.7 21.0 20.0 21.3 14-15 19.5 21.0 22.8 20.1 22.3 22.9 16-17 21.4 21.8 22.3 21.3 20.5 20.7 18-19 22.9 23.2 22.9 22.4 22.0 22.2 20-21 22.8 24.1 23.7 22.2 22.8 22.6 22-23 23.7 24.6 24.7 23.0 21.5 23.1 24-25 24.5 24.9 25.0 24.5 23.1 23.2 26-27 26.5 24.0 26.6 25.0 22.6 24.8 28-29 22.7 27.8 28.5 27.0 24.0 24.5 40 o 30 a ' 20 10 0. 0 as 10XL 47 - J. CHAVGANC, R. HANAK & H. G. ten DAM FIG. 2 DEVELOPMENT OF LOCAL BCG REACTIONS (INDURATION) 0 3 4 6 9 12 18 20 23 26 Day of follow-up examination A-C s 4_ FIG. 3 DEVELOPMENT OF LOCAL BCG REACTIONS (TISSUE DESTRUCTION) Non-reactors R P .---- Reactors J Non-reactors Glaxo vaccine ............. .Reactors ............- .* I/_ 0 34 ,6 .8 2 3 2 9 12 Day of follow-up examination 48 I 15 E ,4 10 a) a) _ n u 18. 20 .23 26 '91, - Ii 0 3 4 6 SIGNIFICANCE OF LOCAL REACTION TO DIRECT BCG VACCINATION 2 Al z 0; F'- 0)to 0 c0 to goF- 0 U) 0 04 04 U)go 0 04 04 v- 04i le C- 0) U) Cs 0 0 [I F- 0i 04 0 49 U) o E t .-*N O 0 )10 0 No sq 0 0 0 0 0 0 0 0 0 0 o 0 0 o °)ct 0) ( 0 0 0 00 00 00 00 00 E -C U rN- 0 f- D -'-'-' '- O1I 0Z_ N 0 0 0 0_ R 0 CD m 0 0 0 0 0 0 0 0 0 O E Z: o 00 OU) 0~ o^ 0 cn 0> 00 0)0_ 0 0-o 0 00O VZJN N0) 0) 0 0 I0) 0) 0) 0 ONO E t o~ r> 0__ 0 0 00o 0o C0 0o 00o E PO ac ot go (a0, 0 0 0) 00) 0goU) U) go '- 0)~~~~~~~~~~~~0m0 ; 5 Q 0)0I 0 F- 0 ~ 0 ~ ' ~ ~ ~ ~ ~ ' WI' 0) co 06 Q~~~~~~i 00 z r'- 0) ~~~~~r-CD 0) 0 002 E 00 - i Ut, r- L ) sa q oLO'q 0 cmt r- if co ul 4m 0) CD 0) 0)0 o a 0) > CE M ~ ~ ~ c) - '-l 0) 0) V-1 0) rl 0) V- '- 0) T O a ~~~ 0) V (0 '- 0~~~~~~ 'a P C c0i co P- co co i 0 0 0 0 V Z C4N N N 0 N _ - ( 0)O 0. 0 0 0 0 O _ _ - C) r'- oo gco r'- 0 ) 0 ) 0 ) 0 E q sq sq o q LsqORsq 'e 0L L' 0 0 i 0 0C E C-4 v Ln co m 0) C) 0) 0) 0) 0- 0) 0l 0- 0) 0 0E0(0 ~~~~~~~~~~~~~~~~ U) 0 ~~~~~~~~~~~~0 qao40) Lq0c0co ) 04 0 0 U) U0 U)co 0 0 o o -: - 0 0 V Z Ln cod U) F-- or E v-v- 'r. M n co to mj 0'- m U)) CDU 0m V- V- C go U 0) ) 4 0R F-. go F.. C F-' 01q 0 L-U q co go g V Z Q r- 0)vC 0L100) 06 0)0) to 4-0 v 04i '- 0 0-co0)c E v co OR r- 0g 0 go Q C>E C 0 0 W'- W co)co 0)0))mc '-) Om 0 C)D C4 00)m ) 0)CE 001~~~~~~~~~~~~Ve OR'~i U 04 U) C It C ( U) 10 U) C' co0 0)0Wv _ 04 04 '- 04 c ) U) co L6 r- g066 C C'- 0; 0) 0)C4~~~~~~~~~~0 qC nP or o4 c 2- E Al z z 04t 40 of cm r C-' .'r -41 0 0 U) to ui 04 ,o c-i co U) q-: In co U,) go go go co gD co c-i co E4 X s0 co a) 0o co o "'t 0 LO 0) 40 0) LOD go U) 0) co 0- 0) a) D0 0) cocim co0co go 0) o0)0)R ot co 0) 0) a v- 0 am 0TIiT 7iT' IT'-iHr T7E '- LocIq T1L Z z go X! 00 go r' O '- F- _I _ I I II jI E cm -W 04 C4 ~ ~ 0I~ C.1z r- co c)o 0) C5 o0 0- '- F' 0-0 Ez _ 0)o o ' I'I I II I I I I E LO 04 z tX _ o N > v 0 0 4 0 g co U _ _ EcoN llll4llll cl_ a - C Lo co U.) o c4L 04°104040 t- II I- I E co > c0 Xc4 004 '- ' I 777 E C4 0 4 0 0 ) 0 r' 0 I -4~'~ ~- liii II I I W U)~~ F-'- 0 0) 'a 0 0) ~ goo0toU) - E 08 co go 0 g 0 g 4 - F' LO C4 c > v00 m co L g o o 0m N Z S _ _ N _ N tO _ O _ _ __T__~~~~~~~~~~~~~~~~~~~U E E 4m z cq (c C1- "W ,4 04 to cow Q U) 04 04 go C>04 04 co0O 04 04 0D c-i 0co co0 N4 N co 04 C-i 04 ca go: N 04f- "W -i '- co- N "WN co-NN 4__N N -N N go <a. Z o- N- 0) 0 0 0 N) N U) ---N j4 _ 00a4 - C,, C0) CU)CD K c a c- 0 c- C 0 0 '- 0 c- 0 (a 0. (a 0L (0 0 ) r - 0 ) 0 a.0 O- go3 a. 0411 W 0W0a 0a (U.S ~ ~~ )( )()( )( )()( CC CC~~~~~~C0. (U 0.~ . U - U . U ~ U 0 U 0. (U 0. (U0rC 0 0 co 0 'e 0 0 0 0'D ax s( en-SC ss co< 0 C> >' -0> co 0 09 >' Cg>' 04' c Cli co ~ ) 0 2. 0 c 0 (o -oX. C 0 0) C -C 0 .)0(P cX C a P a)0L 0 r-0 *0 -aU 0 0 0 a) .0 z I- z cU Coz LU o I-z 0 F- z a -) W It I- LU - coO0 z z 0 CZ LU -j 0 -J Ii- 0 z 0 -j 0 ,v 1- E.Q). u .0 "- U*-i= co co 0> r- J. CHAVGANC, R. HANAK & H. G. ten DAM z 00 0d 40 00 P,: 0D CCi co CC04 Rd C', 000- 0" WI') La cin ,00 CO 04 in Cdi LO 0; 00 cC D0" 0 '9CC0' 0 In C4 0) ,-w C40i C' co, Ci co 0i Cdv CC( E c60 eq eq coru-C'm co r-0mCi if 'nO 0 ui gi e ~r- ,4 0)m 0cm )cm co CD CD 0) 0) 0c) 0) 0 ) 0) 0) 0 E (0 0 C,b 0 4m)r r- in 0) C') r- le r"2 ) co ig0w - I) co co cm, ln cm r. Csd eq V co C) r c Cd an 0' 04 0 0 L9 0)9- 0, 0) LO co 0r 0 04 coIt Lq In c'i eq Ci eq cq leq aIo Leu qu?0) 0 W0 Vc~ Ci an C4 r-. eq? "I ( ? CY! CR eq qR U) 4i 0)0) e00" 0) 0V Z C )V mn ,- 0) C') 0> 0- Ci0 00 0- in C's U) an C) CD C 0o 0) 0- 0o co 0- 10) r- in I- U) P- to co) of C4 Ci 2 ! Ci eq ci 0) c') 0q 0 0-: It Ci i U) 0) in E 0n 0 cm co- 110 toCim 0 0 a C co) co) 0) VZ co o) & C C 4 in v 0 0) 0) 0D C' 0)0a) C)C0) 0) 0) 0 ) co 0o 0- 0- 'C' in cm LO cmJ C') a' z z co 0- C) CC 04 CCl C', 0)0r) C's LO 0- a') C' 00 0) 40 C') C') 0) 004 0D CD 04 0 CD 0) 0): 00C's co CC4 40 Cdi in co co C' in C'R 00 co CY) CC4 C') co 0 04 co 0) in) 0 in to C') 0) C4 in C') 0) 04 0I ad CC co) 0 0 co, 0 Uin CC C', 40 'o14t C-4 co C4 co m Q 0 0) 0-- v co) 1- 0D 0 0) 0- ~~~ 00 o- '- r-.0 C 0 0o v T-. 0 C 0 co 0) 0o mP- leco P-0) 0)o 0 0"0o00o0--0)mc in C' in E P-~~~~~~~~0 0co0co) co '- 0- 0D '- 0m 0--C'00 2o WI,0 C) co V- 0) Ci 00 CO in- r- ra0) co '- c 40 E Cb m C~~~) V' v 4 o 0in 0I 0) '-V0 i E z CCr CC4 0> C4 00 (a C4 'C- CC4 C' to 0oC't)C4C 0407) CO,Uin C4 coUin co0D 04 CY) C4C' C4 00 0- C'co I' 0 0 0 40 C'C', In 04 Uin mt - 0 to 0o 0D 0 0) 0o le " in) in C') 0- C') Ci0 0) in toO)Ir in c.j co', wCi 01C> C4 '- 0D C') r- co '- 04 0) .EC4 04 CC 04CC CC 4i CC C 4CC C4 C4 04 04 04 C4 Cq cCi ~~to~~ 0 0~~0 0- 0 n0 r- 0- 0)1 r- 0C' 0 00i 0) 0 a. a. a. ta 0. to 0. to 0 co a. to a. to 0. to a. co a. to > -W t o5 0: 0 DL O0 co co co 'a I~~~~~~~- ,T-to (A (A CO>" > MC'>t Cmm C- to o. 04c co 04 '-to4 co -c -zcin 0 E 50 I Al 0 0 to4t 0 cm 0 co at)cm C a. 0 C_ 0 toto 0 -a C 0 -C ~0 z E 0 2 Al z w U) z w 11- z z w 0. 0 HU) w w z 0 L) w (D -J 0 -J Li- 0 z 0 -J w 0 C- E 0 04) L- r- r-E c-D ,ci 0.O.- a U -4--P co Cli IV" Lq v SIGNIFICANCE OF LOCAL REACTION TO DIRECT BCG VACCINATION After 12 weeks these differences have decreased and the following values are found: GlaxoIRP vaccine Non-reactors 6.1 mm 6.5 mm Reactors 7.0 mm 8.1 mm The correlation of local tissue destruction and pre- vaccination tuberculin sensitivity has been summa- rized in Table 4. Large reactions, which were most frequent on the sixth day after vaccination, never attained disquieting dimensions. The types of the lesions appeared to be closely related to their size. The large lesions seen during the first 2 weeks after vaccination were mainly unperforated pustules. Ulcers were rarely larger than 10 mm; only in 2 instances was an ulcer seen that was as large as 21 mm. In reactors the pustules appeared and per- forated somewhat earlier than in the non-reactors, but the further course of development was similar for both groups. The frequency of the different types of lesion at various times after vaccination is shown below for all vaccinated persons combined: Type of lesion 6 days No lesion 609 Unperforated pustule 199 Ulcer (with or without scab) 169 Dry scab or scar 2 Time after vaccination 26 days 75 12 18 days days 275 191 54 29 12 weeks 2 5 12 months 4 570 747 720 28 6 6 24 920 TABLE 5 DISTRIBUTION OF SCAR SIZES 5 Y, MONTHS AFTER VACCINATION Scar size IRP Glaxo vaccine (mm) NR R NR R 0 I 2 3 2 2 4 18 1 11 1 5 46 23 44 9 6 109 65 75 20 7 54 65 71 34 8 26 50 28 58 9 5 9 11 41 10 1 5 5 16 11 1 1 10 12 4 13 1 14 15 1 Total 261 219 248 195 Mean scar size (mm) 6.14 6.88 6.72 8.13 Variance 1.35 1.42 1.79 2.79 924 After 51/2 months practically all the vaccinated persons had a scar. The distribution of scar sizes is shown in Table 5. In reactors, especially those given Glaxo vaccine, the scars were larger than in non- reactors. For both vaccines the differences are statistically significant. It should be noticed, how- ever, that for the IRP and the Glaxo vaccine respec- tively, these differences are as small as 0.72 mm and 1.41 mm. The Glaxo vaccine produced slightly larger scars than the IRP. The mean scar size for persons vaccinated with Glaxo vaccine was 7.19 mm, that for persons vaccinated with the IRP 6.48 mm; the difference is statistically significant. Protruding scars were seen in a non-reactor vaccinated with the IRP and in 2 reactors vaccinated with Glaxo vaccine. The diameters of these scars were 6 mm, 9 mm and 9 mm respectively. Oedema A number of vaccinated persons showed local oedema, generally on the third day after vaccination. It never lasted longer than a few days, but sometimes was up to 8 cm in diameter. The distribution accord- ing to initial tuberculin sensitivity and vaccine group was as follows: Non-reactors Reactors IRP Glaxovaccine 1 2 22 11 Regional lymph node enlargement The sizes of the enlarged lymph nodes found 12 weeks and 5½/2 months after vaccination have been listed in Table 6, separately for each person who showed such enlargement. Whereas on the whole the incidence was low, it is seen that the great 51 TABLE 6 ENLARGED REGIONAL LYMPH NODES OBSERVED 12 WEEKS AND 5 A MONTHS AFTER VACCINATION a I nitial 12-week examination 5 :/-month examination Vaccine tuberculin group reaction Axillary Cervical Axillary Cervicalj(mm) Left IRight Left Right Left Right Left Right I Placebo IRP Glaxo vaccine 0 23 0 0 0 0 0 0 2 8 13 19 20 21 21 24 0 0 .0 0 0 5 9 13 15 16 16 17 19 20 20 21 21 21 23 26 15 15 10 25 15 15 15 15 10 20 10 20 10 10 15 15 10 10 10 15 15 10 15 10 20 10 10 10 10 10 20 20 15 15 15 30 10 10 10 15 15 10 10 20 20 15 15 10 10 a Diameters in mm. SIGNIFICANCE OF LOCAL REACTION TO DIRECT BCG VACCINATION majority had been caused by the vaccination. The incidence of lymph node enlargement seems hardly correlated with initial tuberculin sensitivity. Suppu- ration was not observed. DISCUSSION Acceptability of direct BCG vaccination The results confirm that when BCG vaccination is given without previous tuberculin testing no excessive reactions are seen in the persons who have tuberculin sensitivity before vaccination. Neverthe- less, it causes in those persons a certain discomfort, which is probably not compensated by any protec- tion conferred by the vaccination. On the other hand, the whole population is spared the discomfort of a tuberculin test. It would therefore seem that, provided that the proportion of reactors in the population selected is low, direct BCG vaccination is an accept- able public health measure from the point of view of the population. Distinguishing reactors from non-reactors The present study was partly designed to investi- gate whether information on prevaccination tuber- culin sensitivity can be obtained from measurements of the local BCG reaction. The simplest information one could hope to extract from the measurements of the local BCG reaction is perhaps the proportions of reactors and non-reactors. Since, at least during the first few days after vaccination, there is a definite correlation between the induration at the vaccination site and the pre- vaccination tuberculin sensitivity, it should be pos- sible to establish a criterion by choosing the reac- tion size that divides the population in the same proportions as the tuberculin test did. If one con- siders, for instance, the reactions to the IRP 3 days after vaccination (Table 3; Appendix Table 1), it may be seen that the reactors (i.e., those who had a tuberculin reaction of 10 mm or more) have on the whole far greater induration than the non-reactors. In fact 203 out of 256 reactors have an induration of 14 mm or more, whereas only 49 of the 290 non- reactors do so. Thus, by considering all persons with a local induration of 14 mm or more as reactors, one could conclude that there were 252 reactors among the vaccinated persons (whereas in fact there were 256). This might seem a reasonably accurate estimate if it were not that the group of reactors so formed included only 79.3 % of the reactors and as many as 16.9% of the non-reactors to tuberculin, which means that the accuracy of the estimations will diminish the more the proportions of reactors and non-reactors differ from those in our study. Certainly, some correction could be applied to remedy this inconvenience, and the procedure might still be acceptable to the epidemiologist who had no other means of estimating the infection rate in the popula- tion. But even this desperate epidemiologist would lose interest if he knew that, by applying the same criterion to a population that was similar but vaccinated with Glaxo vaccine, he would find no more than 49 reactors whereas in fact there were 245! (see Appendix Table 2). The third day after vaccination Glaxo vaccine has not yet produced large indurations, and the correla- tion with prevaccination tuberculin sensitivity is less evident (note that 16.3% of the reactors show induration as against 2.6% of the non-reactors). Another criterion would have to be applied for the Glaxo vaccine: probably persons with an induration of 8 mm or more would have to be considered reactors. Unfortunately, many reactors and non- reactors have an induration of about this size and slight differences, for instance in reading techniques, might produce entirely different results. Furthermore the group of non-reactors would include only 66.2% of all the non-reactors to tuberculin and as many as 26.5% of the reactors. The accuracy would depend very much on the proportions existing in the vacci- nated population, unless appropriate adjustments were made. Be this as it may, the idea of using different criteria for different vaccines (or even vaccine batches?) is far from attractive. Certainly one part of the disad- vantage is that BCG, unlike tuberculin, is not a " standardized " product. It is produced in various laboratories in relatively small batches, which may show appreciable qualitative and quantitative differ- ences. From a further inspection of Table 3 it is clear that no advantages are to be gained by considering the measurements made on other days. Many an investigator has recalled the " Koch phenomenon " when observing a rapid development of local lesions in BCG-vaccinated tuberculin reac- tors. Without attempting to qualify the analogy, the reader may like to consult Table 4 and see whether any use can be made of the measurements of the tissue destruction. Obviously, during the first few days following vaccination there is hardly any local tissue destruc- tion, and not much information can be gained. 53 J. CHAVGANC, R. HANAK & H. G. ten DAM Similarly, after about 3 weeks both non-reactors and reactors have developed lesions, mostly in the narrow range of 4 mm-9 mm, and it will be difficult to determine which is which. It is after 2-3 weeks that confirmation is found of the common observation: practically all reactors have a marked lesion, whereas a great many of non- reactors have not (yet) (see Appendix Tables 3 and 4). Nevertheless, trying to find the proportion of reac- tors does not appear to be any easier than in the case of induration. On the contrary, the criterion happens to fall at a level where many reactions are found and therefore allows little leeway (the presentation of 2-mm groups seems already to be less appropriate). It is obvious that small variations, e.g., in the reading technique, would influence the results considerably, and again, different criteria would have to be estab- lished for the two vaccines. Selection ofa group of non-reactors Whether this selection is made for diagnostic purposes, e.g., to exclude the non-reactors from X-ray examination, or to select a population for assessment, the aim is to find a group that includes as many as possible of the non-reactors and few, if any, of the reactors. Considering, as before, firstly the data on local induration (Table 3), especially the percentages, it is seen that for the first few days there is no single satisfactory solution. However, as from the 12th day practically all the reactors appear no longer to have any induration, whereas many non-reactors still do so (see Appendix Tables 5 and 6). The observation applies, moreover, both to the IRP and to the Glaxo vaccine and is fairly consistent for several days. Turning now to the tissue destruction (Table 4), a somewhat similar picture is seen. As from the 12th day, practically all the reactors have local tissue destruction of at least 4 mm. On the other hand, many non-reactors are still found in the 0-1-mm group (Appendix Tables 3 and 4). It would therefore appear that, if these observa- tions could be confirmed, it might be possible, by examining the local induration or the local tissue destruction (or both) in the third week after vaccina- tion, to select a group comprising 25%°/-50% of the non-reactors and practically no reactors. Whether it is profitable, for diagnostic purposes, to select such a group, e.g., before proceeding to X-ray examination of the rest, is a question that can hardly be answered here, though it should be remarked that the procedure necessitates 2 visits at a few days' interval, as is necessary when a screening test is given. As regards assessment, however, the situation is quite clear: by means of the local indura- tion or the tissue destruction it is impossible to detect more than half of the non-reactors. Thus the assessment ofvaccination would have to be performed in a group of non-reactors that has been selected precisely on the basis of some reaction to this vaccination-a most unfortunate selection procedure indeed, which certainly cannot be presumed to produce a representative sample, as may be seen from Appendix Table 7, which shows the post- vaccination allergy after 12 weeks and after 51/2 months, separately for non-reactors who had some local induration 12 days after vaccination and for non-reactors who had not. The post-vaccination allergy in the latter group of non-reactors is signifi- cantly higher than in the former. From the present study it appears that not much epidemiological significance can be attributed to the local BCG reaction. Observations by Geser, Roy & Bl0cher (1966) in Burundi, Dahomey and Mauri- tania, though only at a few intervals after vaccination, tend to confirm the present results and make it unlikely that they are the exception to the rule. Studies (not yet published) carried out in Lesotho and Indonesia according to the same protocol as the present study also confirm these findings. On the otherhand, observations in Kenya (Egsmose, 1965), and Algeria (Grosset, Benhassine & Mokhtari, 1968; Oriol & Meriaux, 1968) seemed far more promising. Egsmose carried out 3 investigations in school- children. In the first investigation the children were vaccinated, in the shoulder region, with Japanese vaccine. The local induration measured 1, 2, and 3 weeks after vaccination did not provide a criterion for separating the reactors from the non-reactors. In his second and third studies, however, in which the children had been vaccinated in the forearm, the groups of children with a local induration of 10 mm or more included 80% of the reactors and not more than 5 % and 8% of the non-reactors respectively. In the studies in Algeria vaccination had also been carried out in the forearm. In the study reported by Grosset et al., 93% of the reactors and only 3% of the non-reactors showed local induration of 12 mm or more 4-5 days after vaccination. In the study by Oriol & Meriaux, more than 85% of the reactors and not more than 2.3% of the non-reactors showed local induration of 13 mm or more 3 days after vaccination. 54 SIGNIFICANCE OF LOCAL REACTION TO DIRECT BCG VACCINATION These results seem to indicate that the site of injection may play an important role. However, this could certainly not explain entirely the general discordance between the various results obtained so far. From the present study it is seen that the local lesion largely depends on the kind of vaccine used, and several other factors (injection technique, quality of the vaccine) are also likely to have an influence on the local reaction. Thus it is most unlike- ly that the more promising results obtained are repro- ducible under varying conditions. Their epidemiolo- gical significance is therefore highly questionable. An important question that remains unanswered is whether the vaccination reaction itself may not be used directly for the evaluation of the quality of the vaccination. ACKNOWLEDGEMENTS The authors wish to express their thanks to Mrs Dashtseren, National BCG Team, and Miss V. MatkovA, WHO BCG Nurse, and the secretaries of the field team for their painstaking efforts during this study, and to Dr J. Guld for his advice and constructive criticism in the preparation of the report. RltSUMI IMPORTANCE EPIDEMIOLOGIQUE DE LA REACTION LOCALE APRES VACCINATION DIRECTE PAR LE BCG, TELLE QU'ELLE RESSORT D'UNE ETUDE EFFECTUEE EN MONGOLIE Le but de cette etude etait double: s'assurer que la vaccination directe (sans epreuve tuberculinique prea- lable) par le BCG est un procede d'immunisation qui peut etre employe sans inconvenient en Mongolie, et verifier l'hypothese selon laquelle il est possible, par l'examen de la reaction locale au BCG, d'evaluer le degre de la sensibilite tuberculinique preexistante. Les investigations ont port6 sur 1367 jeunes etudiants d'Oulan-Bator chez lesquels il n'existait aucun antecedent de tuberculose et qui n'avaient jamais et vaccines: 562 d'entre eux ont requ par voie intradermique la pre- paration internationale de reference (PIR) de BCG, 533 du vaccin Glaxo et 272 un placebo (solute salin). Tous ont subi simultan6ment une epreuve tuberculinique au moyen de I UT de RT 23 avec Tween. Des examens de controle ont eu lieu 3, 4, 6, 9, 12, 18, 20, 23 et 26 jours ainsi que 12 semaines et 51/2 mois apres la vaccination. Chaque fois, on a mesure en millimetres le diametre transversal de l'induration locale et/ou de la destruction tissulaire. A l'occasion des deux derniers contr6les, on a procede en outre a une nouvelle epreuve tuberculinique (2 UT de RT 23 avec Tween) et recherche une eventuelle hypertrophie ganglionnaire cervicale ou axillaire. Avant la vaccination, la moitie environ des sujets rea- gissaient 'a la tuberculine. L'epreuve tuberculinique post- vaccinale a permis de constater que la vaccination avait ete efficace. L'induration locale provoqu6e par le BCG a rapidement disparu et on n'a enregistre aucune reaction importante. La destruction tissulaire locale est apparue plus precocement et a ete legerement plus intense chez les sujets qui avaient reagi a la tuberculine que chez les 4 non-reagissants )). Les lesions ont W plus importantes apres emploi du vaccin Glaxo qu'apres vaccination par le PIR; cependant, apres 12 semaines, la difference moyenne n'atteignait que 2 mm au plus. Apres 51/2 mois, les sujets vaccines, dans leur quasi-totalite, presentaient une cicatrice dont les dimensions moyennes etaient de 6,9 mm (sujets reagissants) et de 6,1 mm (sujets non r6agissants) pour le vaccin PIR et de 8,1 mm (sujets reagissants) et de 6,7 mm (sujets non reagissants) pour le vaccin Glaxo. Un oed6me local, qui s'est rapidement resorbe, est apparu chez certains sujets qui avaient reagi au test tuberculinique. Les hypertrophies ganglionnaires regionales ont ete peu frequentes et sans relation appa- rente avec le degre de la sensibilite prevaccinale a la tuber- culine. II semble donc que s'il n'existe qu'une faible pro- portion de sujets reagissants dans les collectivites qu'on se propose de vacciner, la methode directe est susceptible d'etre acceptee par la population, a qui l'on epargne les inconvenients d'une epreuve tuberculinique generalisee. Trois jours apres avoir ete vaccines par le PIR, les sujets reagissants ont, dans l'ensemble, present6 une induration locale plus grande que les sujets non reagis- sants, mais il semble difficile de definir un critere permet- tant de determiner les proportions relatives de ces deux categories. De plus, un critere identique (par exemple une induration de 14 mm) ne peut en aucun cas etre applique aux sujets vaccines par le vaccin Glaxo. II est encore plus malaise de distinguer les sujets reagissants des non reagissants en se basant sur l'etendue de la destruction tissulaire locale. Par l'examen de l'induration ou de la destruction tissulaire (ou des deux) au cours de la 3e semaine suivant la vaccination, on pourrait parvenir a selectionner un groupe comprenant 25 i 50% de sujets non reagissants. On peut s'interroger sur l'int& ret diagnostique de cette recherche, qui necessite, comme l'epreuve tuberculinique, deux examens 'a quelques jours d'intervalle. Un groupe selectionn6 de cette maniere n'a par ailleurs, de toute evidence, aucune utilite lorsqu'il s'agit d'evaluer les r6sultats de la vaccination. 55 J. CHAVGANC, R. HANAK & H. G. ten DAM REFERENCES Egsmose, T. (1965) Acta tuberc. scand., 46, 220 Geser, A. Roy, L. A. & Bl0cher, C. (1966) Bull. Wld Hith Org., 35, 609 Grosset, J., Benhassine, M. & Mokhtari, L. (1968) Rev. Tuberc. Pneumol., 34, 551 Oriol, 0. & Meriaux, B. (1968) Bull. Wld Hlth Org., 39, 837 WHO Expert Committee on Tuberculosis (1964) Wld Hlth Org. techn. Rep. Ser., 290, 11 APPENDIX TABLE 1 CORRELATION OF PREVACCINATION TUBERCULIN INDURATION WITH LOCAL VACCINATION REACTION (INDURATION) 3 DAYS AFTER VACCINATION WITH IRP Local vaccination induration (mm) 0 2 4 6 8101214161820222426 28>30Total 0 30 33 75 25 15 13 9. 13 11 2 3 229 2 3 10 6 2 3 1 2 2 1 30 4 2 4 2 1 1 4 2 16 6 1 4 2 2 1 1 1 1 8 1 2 1 4 10 1 1 3 4 1 10 12 4 1 2 1 2 1 1 1 14 6 44 - 8 16 4 2 1 4 4 6 3 2 26 18 7 1 1 5 6 13 10 4 3 1 51 20 3 1 1 2 3 10 26 10 8 2 2 68 22 5 1 1 8 7 7 6 7 1 43 24 3 1 6 1 3 5 5 1 1 1 27 26 1 2 2 2 1 8 28 1 2 1 4 .30 TotalI60 33 9134 25 5 26 61 80~ 48 35 20 5 2 1 546 56 .- E C 0 to .° M C ._ c U 0 .0 C 0 4- 0(J 0~ SIGNIFICANCE OF LOCAL REACTION TO DIRECT BCG VACCINATION APPENDIX TABLE 2 CORRELATION OF PREVACCINATION TUBERCULIN INDURATION WITH LOCAL VACCINATION REACTION (INDURATION) 3 DAYS AFTER VACCINATION WITH GLAXO VACCINE Local vaccination induration (mm) 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 >30 Total 0 6 50 102 54 5 3 220 2 1 11 8 20 4 1 4 5 3 1 14 6 4 1 5 8 1 2 5 1 1 10 10 1 3 1 1 6 12 1 3 5 2 1 12 14 2 1 3 6 1 1 1 15 16 2 3 7 6 3 2 23 18 7 4 7 10 7 2 4 41 20 16 3 8 12 8 8 1 3 2 61 22 6 1 2 8 11 13 6 2 1 50 24 4 1 2 5 5 4 2 23 26 1 1 1 3 2 3 11 28 1 1 1 3 .30 Total 47 55 14111 6 45 2614 7 21 514 57 E E 0a C C U .0 3 .0 a c 0. a p ._ 5 J. CHAVGANC, R. HANAK & H. G. ten DAM APPENDIX TABLE 3 CORRELATION OF PREVACCINATION TUBERCULIN INDURATION WITH LOCAL VACCINATION REACTION (TISSUE DESTRUCTION) 12 DAYS AFTER VACCINATION WITH IRP Local tissue destruction (mm) 0 12 3 4 567 8 910 1112 1314 1516 17 1819.20Total 0 109 3 6 11 33 21 18 9_2 1 1 1 215 2 12 2 7 1 1 1 1 25 4 6 1 132 1 1 15 6 1 2 1 5 1 10 8 11 1 1 1 3 10 2 43 9 12 1 5 2 2 10 14 2 1 4 1 8 16 3 8 8 3 3 2 27 18 2 11 149 5 1 1 43 20 1 9 16 17 14 1 58 22 1 77 1055 35 24 1 3 5 7 4 4 24 26 1 32 6 28 { 1 2 1 4 >30 Total127 4 1143 104 9665 3024141 492 58 E E 4-0 tu as -D C 0 U) Q co E _ IL SIGNIFICANCE OF LOCAL REACTION TO DIRECT BCG VACCINATION APPENDIX TABLE 4 CORRELATION OF PREVACCINATION TUBERCULIN INDURATION WITH LOCAL VACCINATION REACTION (TISSUE DESTRUCTION) 12 DAYS AFTER VACCINATION WITH GLAXO VACCINE Local tissue destruction (mm) 0 1 2 3 4 5 61 7 8 9 1011|12|13|14|1S|16|17|18|19T>20|Total 0 100 5 6 6 19 12 20 13 5 4 2 3 1 196 2 5 1 2 3 1 121 1 17 4 1 12 2 13 1 11 6 1 2 2 5 8 1 3 4 1 1 10 10 1 1 2 1 1 6 12 2 1 5 3 1 12 14 2 2 4 4 3 15 16 1 5 10 4 1 1 22 18 119 9 3 4 36 20 1 9 313 23 47 22 4 4 12 1036 1 40 24 1 4 4 5 3 1 1 1 20 26 1 2 1 1 1 1 7 28 2 1 3 >30 TotalI109 67 635 61 866736 19 85 1 1447 59 E E 0- c .o C s 0 ._ a ._ a ._o 0~ J. CHAVGANC, R. HANAK & H. G. ten DAM APPENDIX TABLE 5 CORRELATION OF PREVACCINATION TUBERCULIN INDURATION AND LOCAL VACCINATION REACTION (INDURATION) 12 DAYS AFTER VACCINATION WITH IRP Local vaccination induration (mm) 0 2 4 6 8101214161820222426 28.30 Total 0 53 8 54 61 18 2 196 2 7 2 8 6 2 25 4 4 2 6 2 14 6 7 1 8 8 4 4 10 9 1 10 12 1q 10 14 7 18 16 19 2 21 18 44 44 20 55 1 1 1 58 22 32 32 24 25 25 26 7 7 28 3 3 >30 Total 286 10 64 76 25 1 2 1 465 60 E E .2 co 0 C c 0 C 0.0 -a_b SIGNIFICANCE OF LOCAL REACTION TO DIRECT BCG VACCINATION APPENDIX TABLE 6 CORRELATION OF PREVACCINATION TUBERCULIN INDURATION AND LOCAL VACCINATION REACTION (INDURATION) 12 DAYS AFTER VACCINATION WITH GLAXO VACCINE Local vaccination induration (mm) 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 .>30 Total 0 6 50 102 54 5 3 220 2 1 1 8 20 0 16 4 5 3 1 14 6 -4 1 5 8 1 2_5 1 1 10 10 1 3 1 1 6 12 1 3 5 2 1 12 14 2 1 3 6 1 1 1 15 16 2 3 7 6 3 2 23 18 7 4 7 10 7 2 4 41 20 16 3 8 12 8 8 1 3 2 61 22 6 1 2 8 1 1 13 6 2 1 50 24 4 1 2 5 5 4 2 23 26 1 1 1 3 2 3 11 28 1 1 1 3 ->30 61 E E c 0 .Q-aEtl5 - E U a .0 .a c 0 a a 0. Total 47 L 55 141 110 7 45 26 14 7 514 J. CHAVGANC, R. HANAK & H. G. ten DAM APPENDIX TABLE 7 POST-VACCINATION TUBERCULIN REACTIONS, 12 WEEKS AND 5 Y2 MONTHS AFTER VACCINATION, IN NON-REACTORS WHO HAD SOME LOCAL INDURATION ON THE TWELFTH DAY AFTER VACCINATION AND IN NON-REACTORS WHO HAD NOT Local vaccination reaction Vaccine Time after after 12 days used vaccination |Present |Absent No. 152 63 12 weeks Mean 19.3 20.4 Variance 21.8 9.5 IRP No. 138 58 5 Y. months Mean 18.2 20.1 Variance 28.0 14.3 No. 104 99 12 weeks Mean 16.6 20.1 Variance 33.4 17.5 Glaxo vaccine- No. 94 95 5 2 months Mean 15.3 18.9 Variance 38.5 28.1 62