The use of jet-injectors in BCG vaccination*

Bull. Org. mond. Sante' 11970, 43, 707-720Bull. Wld Hlth Org. The Use of Jet-Injectors in BCG Vaccination* H. G. TEN DAM, C. FILLASTRE, G. CONGE, E. ORSSAUD, C. GATEFF, A. TANAKA, 0. ORTEGA RAMIREZ, R. COLLAS, J. WRIGHT, L. CHAMBON, M. BARME, U. B. TOMMASI, H. SARRAT, P. BRES, L. DIALLO, M. GAUTHIER, M. PIOT & J. GULD In mass vaccination programmes, the jet-injection of vaccine may have considerable operational advantages over the classical techniques. The technical performance of two models ofjet-injector, the Dermo-Jet and the Ped-O-Jet, in BCG vaccination was assessed in a number of studies which are reviewed by the authors. It is shown that the jet-injectors do not administer the full dosefor which they are calibrated and that the size of the vaccina- tion lesion varies more than after vaccination by syringe. By increasing the dosage considerably, the results of vaccination by jet-injection may be improved to a certain extent but the risk of unpleasant reactions is also increased. Jet-injectors have been used for many years as an alternative to the needle and syringe for administer- ing subcutaneous injections and, more recently, for the percutaneous application of vaccine in smallpox vaccination. The main advantages of jet-injectors are that, to an extent varying with the model, they are easy to handle and readily accepted by children and, in particular, their use allows, under certain circumstances, a large number of vaccinations to be made in a short time. With these advantages, the instruments are obviously highly attractive for use in mass vaccination programmes. Jet-injectors may be equipped with a nozzle that causes the vaccine to be deposited intradermally. Such a nozzle is used in smallpox vaccination and the results, in general, compare favourably with those obtained by the classical percutaneous tech- niques. It seemed possible that the same instruments might also be used to administer BCG vaccine although the careful intradermal injection of an * This article is a review of a number of investigations sponsored by, or made in co-operation with, the World Health Organization. The studies were carried out in France (C. Fillastre, G. Conge & E. Orssaud), in Gabon (C. Gateff), in Korea (A. Tanaka), in Nicaragua (O. Ortega Ramirez & H. G. ten Dam), in Niger (R. Collas & J. Wright) and in Senegal (L. Chambon, M. Barme, U. B. Tommasi, H. Sarrat, P. Bres, L. Diallo & M. Gauthier). The basic protocol for the studies and this review were prepared by H. G. ten Dam, Scientist, in consultation with M. Piot and J. Guld, Medical Officers, Tuberculosis, World Health Organization, Geneva. Separate reports of the studies in mimeographed form and reprints of this article are available to persons officially or professionally interested on request to Tuberculosis, World Health Organization, 1211 Geneva, Switzerland. exact dose is essential for this vaccination procedure. Investigations were prompted by operational con- siderations; many smallpox and BCG vaccination programmes can profitably be combined into a single programme. However, when smallpox vac- cination is given by jet-injector, it is generally held that BCG vaccine must be given in the same way, so that the combined programme is not retarded by a slower BCG vaccination technique. MATERIAL AND METHOD There is no simple way of estimating the immunity in man conferred by BCG vaccination; to assess the quality of the vaccination, use has to be made of various circumstantial effects. The most important of these is perhaps tuberculin sensitivity, which is invariably induced by vaccination, even if the BCG vaccine is of poor quality. This allergic response may not be directly related to immunity, but it is a quantitative phenomenon that varies with the dose of vaccine effectively given; the latter can thus be measured by means of a low-dose post-vaccination test-at least in persons who had no tuberculin sen- sitivity before vaccination (Edwards, Palmer & Magnus, 1953; WHO Tuberculosis Research Office, 1955a; Guld, 1957). It is important to realize that the level of tuberculin sensitivity after superficial and deep intradermal inoculations, and also after subcutaneous injections, is the same (Palmer & Edwards, 1953). Therefore, the studies reviewed in this paper have been based on the premise that any 2594 -707 H. G. TEN DAM AND OTIHERS difference between the vaccination techniques would be merely quantitative, at least as regards the post- vaccination tuberculin sensitivity. The allergic response to vaccination differs from individual to individual, even when a constant dose is administered in a uniform population. However, when vaccination and tuberculin testing have been performed properly, and when the vaccine used has not been too weak, the tuberculin reactions will show a normal, or at least a unimodal, statistical distribution. Thus, in general, the distribution is adequately characterized by the mean and the vari- ance. A disadvantage in the use of the allergic response for such assessment, and especially in comparisons, is that the level of tuberculin sensi- tivity, as revealed by a low-dose post-vaccination test, varies only slightly with the dose of vaccine administered. As an indication, for the usual strengths of vaccine and tuberculin, a dose of vac- cine 10 times larger than the standard dose will result in a tuberculin reaction that is on average about 3 mm larger in diameter than that produced by the standard dose. Thus a large study population may be required to reveal the, presumably, small differences in allergic response resulting from similar vaccination techniques. Other effects that may be considered for assessing the quality of a vaccination are the local wheal that immediately follows an inoculation, the induration appearing in a few days at the inoculation site and the local lesion and scar which develop later. These effects depend, at least quantitatively, on the dose of vaccine but probably also on the injection technique and they are therefore unsuitable for comparing different vaccination techniques. A further effect, which apparently does not depend on the technique of injection and which is strongly related to the dose administered, is the occurrence of lymphadenitis, generally several months after vaccination. With strong vaccines, this unpleasant reaction is fairly common in new-born infants (Guld et al., 1955) but for obvious ethical reasons a study cannot be designed to make use of this effect. With the usual vaccine doses, and especially in schoolchildren (who represent a far more con- venient study population), the phenomenon is rare. Although local indurations, lesions and scars, as well as the incidence of lymphadenitis, are unsuitable as indications for assessing the positive qualities of the vaccination techniques, they are effects that may limit the application of BCG vaccine and, as such, they must be carefully considered. In view of the current interest in direct BCG vac- cination and re-vaccination, the populations included in most studies belonged to several epidemiological classes which, ideally, should be considered sepa- rately-namely, children who, according to the initial tuberculin test, are non-reactors and have not been vaccinated before (no BCG scar present), children who are reactors but who have not been vaccinated before, and children who have been vac- cinated before (BCG scar present). It should be noted that the latter group cannot be divided into non-reactors and reactors by means of the usual tuberculin test because the level of tuberculin sen- sitivity in vaccinated persons does not depend only on vaccination but also on tuberculin testing (Guld et al., 1968). The allergenic effect of vaccination is best mea- sured in non-reactors who have not previously been vaccinated, but in most studies this group was not singled out. Thus in Nicaragua, where more than half of the children had been vaccinated before, the results were analysed separately for those who had been vaccinated previously and for those who had not. Differences in post-vaccination tuberculin reac- tions between the various vaccination groups could therefore be expected to be relatively small since, after vaccination, the tuberculin sensitivity in reac- tors changes very little. Nevertheless, since the differences observed appeared to be statistically significant, no further subdivision into reactors and non-reactors was made. In the other studies, non-reactors and reactors were considered separately, but in several instances previously vaccinated children were not excluded from these groups. The inclusion of these children, who may have shown increased tuberculin sensi- tivity as a result of repeated tuberculin testing, could possibly have masked differences between the vac- cinated groups, especially if the vaccination tech- nique was less effective. The side-reactions, in particular the vaccination lesions, are generally more violent in reactors and in previously vaccinated children, and these groups are therefore of particular interest in this respect. In the studies reviewed here, the effects of vaccina- tion were measured in terms of the post-vaccination tuberculin reactions 9-15 weeks after vaccination, the induration at the site of injection 3-4 days after vaccination, and the local lesion (tissue destruction) 9-15 weeks after vaccination. On the occasion of the latter examination, the presence of axillary and cervical lymph node enlargement was also investi- 708 USE OF JET-INJECTORS IN BCG VACCINATION 709 TABLE I POST-VACCINATION TUBERCULIN REACTIONS, LOCAL INDURATIONS AND VACCINATION LESIONS IN SCHOOLCHILDREN VACCINATED WITH BCG a BY SYRINGE OR BY DERMO-JET; KOREA STUDY Tuberculin reaction Local induration Vaccination lesion after 15 weeks after 3 days after 15 weeks Population group and Means of concentration of vaccine b administration No. of Mean Van No. of Mean Van No. of Mean Van-(dose= 0.1 ml) sub- size arn- sub- size e sub- size ari- Jects (mm) ance jects (mm) ance jects (mm) ance Non-reactors without BCG scar Standard vaccine Syringe 285 6.2 38.4 320 3.3 4.4 295 5.3 5.8 Dermo-Jet 252 4.5 36.0 282 2.2 5.3 258 3.0 7.8 Concentrated vaccine Syringe 274 8.9 38.4 297 3.8 5.3 281 6.7 7.3 Dermo-Jet 266 6.7 42.3 298 3.1 9.6 271 4.2 10.2 Reactors without BCG scar Standard vaccine Syringe 93 18.0 23.0 98 7.5 9.0 94 7.6 5.3 Dermo-Jet 75 18.1 13.0 86 5.4 16.8 76 5.3 11.6 Concentrated vaccine Syringe 83 18.6 16.0 89 7.9 10.9 83 9.3 7.8 Dermo-Jet 103 19.0 6.3 113 9.1 31.4 103 6.3 7.8 Children with BCG scar Standard vaccine Syringe 45 13.2 42.3 52 5.0 7.3 47 6.0 5.3 Dermo-Jet 53 10.3 53.3 58 3.8 7.3 58 5.0 13.0 Concentrated vaccine Syringe 39 11.6 39.7 46 5.9 9.6 40 8.0 4.8 Dermo-Jet 56 12.6 39.7 58 5.4 15.2 57 6.9 16.0 a Vaccine; Glaxo batch FIO. b The concentrated vaccine was 2'1/2 times more concentrated than the standard vaccine. gated, although this was probably not the most suitable time for determining absolute frequencies. Except in Gabon, where the study was carried out among the general population, all the investiga- tions were carried out in schoolchildren. During the first visit, the children were given a low-dose tuber- culin test (1 TU or 2 TU of batch RT 23 with Tween 80 or 10 units of IP 48) and in several studies the children were examined for existing BCG scars. At the same time they were either vaccinated or given a placebo according to a random allocation. Post- vaccination testing was done with similar doses of tuberculin and all reactions (indurations, lesions, tuberculin reactions) were measured in millimetres. Enlarged lymph nodes were generally recorded in 1-cm groups. Further details on the composition of the groups in the different studies may be found in Tables 1-6. Two types of jet-injector were tested-namely, the manually operated Dermo-Jet1 and the foot- ' Manufactured by Soci6t6 AKRA, Pau, France. operated Ped-O-Jet.' For purposes of comparison, in all studies at least one group of children was vaccinated, according to the routine procedure, by syringe (Omega microstat) and needle (Eisele, platinum). In several studies, two or more different concen- trations or volumes of vaccine were used with a view to determining which dose administered by jet-injector would correspond most closely to the classical vaccination. In these instances, the instru- ments were generally used in rotation in the various eligible groups vaccinated by jet-injector in order to avoid differences arising from variations between one instrument and another. Only a few jet-injectors could be used in each study and these few instruments may not have been entirely representative of the particular model. In addition, the operators had had relatively little experience with jet-injectors which, if experience 2 Manufactured by the Scientific Equipment Manufactur- ing Corporation, Lodi, N.J., USA. H. G. TEN DAM AND OTHERS TABLE 2 POST-VACCINATION TUBERCULIN REACTIONS; LOCAL INDURATIONS AND VACCINATION LESIONS IN SCHOOLCHILDREN VACCINATED WITH BCG a BY SYRINGE OR BY DERMO-JET; NICARAGUA STUDY Tuberculin reaction Local induration Vaccination lesion after 10 weeks after 3 days after 10 weeks concentration of vaccine b Means of -No -| M - __| No _ofM Vi|(dose = 0.1 ml) administration No. of Mean Vani- No. of Mean Vani- No. of Mean 'Vani-sub- size ane sub- size ane sub- size ancejects (mm) nc jects (mm) nc jects (mm) Children without BCG scar Standard vaccine Syringe 185 10.8 37.6 210 5.7 12.3 195 5.8 4.8 Dermo-Jet 225 9.8 41.6 240 5.8 21.8 231 4.7 6.5 Concentrated vaccine Syringe 180 11.7 30.3 208 7.1 22.0 196 7.1 6.0 Dermo-Jet 199 10.0 42.6 221 6.4 21.8 206 5.4 5.7 Children with BCG scar Standard vaccine Syringe 307 12.0 26.4 333 7.1 9.3 322 6.4 2.9 Dermo-Jet 310 11.2 28.9 334 7.4 19.3 322 6.1 5.7 Concentrated vaccine Syringe 325 13.0 19.2 359 8.6 17.3 337 7.8 5.5 Dermo-Jet 298 11.5 24.8 329 8.4 24.8 309 6.6 3.4 a Vaccine; Glaxo batch FIO. b The concentrated vaccine was 21/2 more concentrated than the standard vaccine. counts, may have influenced unfavourably the results obtained with these instruments. RESULTS In some studies, the various vaccination groups, which had been formed by random allocation, showed slight differences, e.g., in the level of pre- existing tuberculin sensitivity, but in all cases these differences are satisfactorily explained as probable consequences of the allocation procedure and they will hardly have influenced the comparisons. In all studies, except those in France, there was at least one unvaccinated control group; thus it could be shown directly that the vaccine used had :an allergenic effect. In the studies in Korea, Nicaragua, Senegal and Gabon a placebo (diluent) was admin- istered by jet-injector and it could therefore be shown that the weakest vaccine dose administered TABLE 3 POST-VACCINATION TUBERCULIN REACTIONS, LOCAL INDURATIONS AND VACCINATION LESIONS IN SCHOOLCHILDREN VACCINATED WITH BCG a BY SYRINGE OR BY PED-O-JET; NIGER STUDY Tuberculin reaction Local Induration Vaccination lesion Means of after 13 weeks after 3-4 days after 13 weeks Population group and vaccine administration No. of Mean Varn_ No. of Mean Varn_ No. of Mean Van- sub- size ane sub- size ane sub- size acjects (mm) ance Jects (mm) ance jects (mm) ance Non-reactors Syringe 148 13.4 13.2 0.1 ml of 0.3 mg/ml vaccine Ped-O-Jet 198 11.1 18.5 Reactors and non-reactors Syringe 232 13.7 17.2 232 6.5 11.6 0.1 ml of 0.3 mg/ml vaccine Ped-O-Jet 296 9.9 16.0 296 5.0 12.9 aVaccineemploy-d:Gla I _ oI ,1ach17 a Vaccine employed: Glaxo batch T17D. 710 USE OF JET-INJECTORS IN BCG VACCINATION TABLE 4 POST-VACCINATION TUBERCULIN REACTIONS AND VACCINATION LESIONS IN SCHOOLCHILDREN VACCINATED WITH BCG BY SYRINGE OR BY PED-O-JET; SENEGAL STUDIES Tuberculin reaction after Vaccination lesion after _ _ansof 3 months 3 monthsPopulation group and vaccine administration l ~~~~No.of |Mean size Variance No. of (mean sie Vracsubjects (mm) e subjects (mm) Vrac Non-reactors Syringe 43 9.4 17.6 43 4.8 3.1 0.1 ml of 0.5 mg/ml vaccine a Ped-O-Jet 43 10.1 24.6 43 3.5 6.6 Reactors Syringe 44 13.4 22.6 45 5.0 4.5 0.1 ml of 0.5 mg/ml vaccine a Ped-O-Jet 48 12.6 28.2 48 4.6 8.9 Non-reactors 0.1 ml of 0.35 mg/mlb vaccine c Syringe 121 14.8 15.5 123 3.4 1.3 Non-reactors 0.15 ml of 0.35 mg/ml b vaccine c Ped-O-Jet 123 14.2 21.0 123 3.3 2.6 Reactors 0.1 ml of 0.35 mg/ml b vaccine c Syringe 171 17.7 17.2 172 3.7 1.3 Reactors 0.1 ml of 0.35 mg/mI b vaccine c Ped-O-Jet 167 18.0 17.2 168 4.2 3.7 a Vaccine employed: IPD batch E43. b Half of each group was given the vaccine in half this strength. c Vaccine employed: IPD batch 66-59 B. in this way still produced measurable tuberculin sensitivity and that the lesions that developed were in fact due to BCG and were not, for example, merely effects due to the injection of a liquid. The mean values and variances for the post- vaccination tuberculin reactions, local indurations and vaccination lesions observed in the various groups are summarized in Tables 1-6. As regards the post-vaccination tuberculin reac- tions, the most interesting populations are those of non-reactors and of children not previously vac- cinated; the distributions for most of these are shown in Fig. 1-6. A comparison of the groups given equal (or supposedly equal) doses of the same vaccine shows that those vaccinated by jet-injector had a lower mean tuberculin reaction than those vaccinated by syringe. In most instances the dif- ferences were shown to be statistically significant but there is one exception. In the preliminary study in Senegal the group vaccinated by Ped-O-Jet actually showed a higher mean than the group vaccinated by syringe, but the 5% confidence limits of the difference of the means are - 1.3 and + 2.8, and this finding therefore does not contradict the general pattern. These results were furthermore confirmed in a Dermo-Jet study in Poland (Dr T. Olakowski, per- sonal communication). In this study, follow-up examinations were carried out 6 and 12 months after vaccination. In children who were non-reac- tors only one strength of vaccine was used (0.5 mg/ ml). The mean diameters of the post-vaccination reactions to 2 TU of batch RT 23 plus Tween 80 and of the scars, together with the variances, are given in Table 7. In all studies, total absence of a tuberculin reac- tion was observed more frequently in the groups vaccinated by jet-injector, as may be seen from Fig. 1-6. The variances were nevertheless mostly similar but as the means in the jet-injector groups were consistently smaller it is probable that the dose was not always administered as effectively by jet- injector as it was by syringe. Anticipating such a result, most workers included different vaccine dos- 71 712 H. G. TEN DAM AND OTHERS TABLE 5 POST-VACCINATION TUBERCULIN REACTIONS AND VACCINATION LESIONS IN A GENERAL POPULATION VACCINATED WITH BCGa BY SYRINGE OR BY PED-O-JET; GABON STUDY Tuberculin reaction after Vaccination lesion after Means of _ _ 9 weeks 9 weeksPopulation group and vaccine administration Population group andvcNo. of Mean size Variance No. of Mean sie Variance subjects I(mm) subjects (mm)z Non-reactors 0.10 ml of 1 mg/ml vaccine Syringe 58 12.4 14.8 58 4.8 1.8 Ped-O-Jet 60 11.5 17.6 60 4.4 5.7 0.15 ml of I mg/ml vaccine Ped-O-Jet 46 12.2 15.8 46 5.8 11.5 0.15 ml of 0.5 mg/ml vaccine Ped-O-Jet 82 11.5 12.2 82 5.2 7.7 0.10 ml of 1.5 mg/ml vaccine Ped-O-Jet 64 9.0 24.9 65 3.6 7.0 Reactors 0.10 ml of I mg/ml vaccine Syringe 75 14.9 75 5.9 2.7 Ped-O-Jet 74 14.2 74 5.1 7.1 0.15 ml of I mg/ml vaccine Ped-O-Jet 70 14.2 70 6.4 10.8 0.15 ml of 0.5 mg/ml vaccine Ped-O-Jet 83 14.3 83 5.8 9.3 0.10 ml of 1.5 mg/ml vaccine Ped-O-Jet 83 14.3 83 5.0 6.1 a Vaccine employed: IPD batch 6832. TABLE 6 POST-VACCINATION TUBERCULIN REACTIONS AND VACCINATION LESIONS IN SCHOOLCHILDREN VACCINATED WITH BCG a BY SYRINGE OR BY PED-O-JET; FRANCE STUDY Tuberculin reaction after Vaccination lesion after Means of 9 weeks 9 weeksPopulation group and vaccine administration . Population group andvaciNo. of Mean size Variance No of Mean size Variance subjects (mm) subjects (m Non-reactors Syringe 181 15.6 8.5 182 7.7 2.1 0.10 ml of 0.5 mg/ml vaccine Ped-O-Jet 180 13.3 14.6 181 6.2 3.3 Non-reactors Syringe 183 15.5 8.0 184 8.5 3.0 0.10 ml of 1 mg/ml vaccine Ped-O-Jet 182 13.8 16.3 183 7.0 5.3 a Vaccine employed: IPP batch 155. ages in their studies, in order to determine whether, by increasing the dose, the " loss " of vaccine could be compensated. That such a compensation is pos- sible is shown in the results obtained in Korea (Table 1 and Fig. 1). A vaccine 2½V2 times as con- centrated as the standard vaccine 1 gave, when admin- istered by a Dermo-Jet instrument, a slightly higher mean tuberculin reaction than vaccine of standard 1 Instead of the 5 ml prescribed by the manufacturer, 2 ml only of reconstitution fluid was used. USE OF JET-INJECTORS IN BCG VACCINATION TABLE 7 POST-VACCINATION TUBERCULIN REACTIONS AND VACCINATION SCARS, 6 AND 12 MONTHS AFTER BCG VACCINATION OF SCHOOLCHILDREN IN POLAND a Tuberculin reaction Vaccination scar Population Time after Means of __ -_ _-_ _ _ _-_ _-_ _vaccination administration No. of Mean size Variance No. of Mean size Variancel________ subjects (mm) _ subjects (mm) _ Syringe 50 11.2 19.4 50 4.2 2.4 6 months Non-reactors Dermo-Jet 40 10.2 28.4 40 2.8 2.2 without an old BCG scar Syringe 26 11.5 22.4 65 b 4.8 2.8 12 months Dermo-Jet 40 8.6 23.9 77 b 2.9 3.2 Syringe 80 14.3 16.3 80 5.3 1.6 6 months Non-reactors Dermo-Jet 89 11.9 23.6 89 4.2 2.5 with an old BCG scar Syringe 60 13.8 22.8 127 b 5.9 2.0 12 months Dermo-Jet 66 12.5 24.3 151 b 4.8 2.7 a Polish vaccine, batch 357106; individual dose: 0.05 mgI0.1 ml. b Including the children also examined 6 months after vaccination. concentration delivered by syringe while the variances were similar. The lower 5% confidence limit of the difference of the means is -0.4 mm. This effect was however not seen in Nicaragua where the same vaccines and the same model of Dermo-Jet were used; a satisfactory explanation for the discre- pancy cannot be given. In the studies with the Ped-O-Jet instrument, dos- ages were included that had been increased by 50% or 100%. These increases were obviously not large enough. From Tables 3-6 and Fig. 2-6 it may be seen that the higher doses administered by Ped- O-Jet in no instance gave the same (or a higher) mean tuberculin reaction as the lower doses deliv- ered by syringe. In a few instances the values may seem quite close, for example, a dose of 0.15 ml administered by Ped-O-Jet (compared with 0.10 ml given by syringe) gave promising results in both Senegal and Gabon. Nevertheless, in these instances the lower 5% confidence limits of the differences in the mean between the Ped-O-Jet and syringe groups are -1.5 mm and -1.8 mm, respectively; this may represent a considerable difference in dose. To conclude that a still higher dose given by jet- injector would have resulted in the same reaction as a standard dose (0.10 ml) given by syringe would, therefore, be speculative. Although the mean tuberculin reactions associated with the higher doses administered by jet-injector were not altogether satisfactory, the variances were similar to those observed after vaccination by sy- ringe. Thus the variation in doses administered by jet-injector may not be greater than that in doses injected by syringe. Therefore it appears that the jet-injectors systematically injected a certain propor- tion of the dose. The size of the local induration 3-4 days after vac- cination appeared to vary independently of the technique of vaccination. A single observation of an evolving reaction provides little basis for com- parison and no conclusions with respect to the effectiveness of the jet-injection technique can be drawn. The observation does, however, show that vaccination by jet-injection gave rise to no serious effects at this time. To a certain extent, the comments on single observations apply also to lesion size, but 10 weeks or more after vaccination most lesions have passed the evolutive stage and comparisons are generally considered to be justified. From Tables 1-6 and 713 H. G. TEN DAM AND OTHERS FIG. 1 TUBERCULIN REACTIONS AND VACCINATION LESIONS 15 WEEKS AFTER BCG VACCINATION BY SYRINGE OR BY DERMO-JET IN NON-REACTORS WITHOUT AN OLD BCG SCAR; KOREA STUDY Non-reactors without a B CG scar 40 30 20 10 C B *;; 0 o 40 c0 ° 30 u 20 &10 0.(L O Tuberculin reaction Vaccination lesion Fig. 1-6 it can be seen that, for equal vaccine doses, the lesions produced by jet-injection are invariably smaller than those resulting from injection by sy- ringe. The means of injection may have influenced the lesion size, but the findings are also compatible with a smaller dose having been effectively injected, as the observations on post-vaccination tuberculin reactions indicated. Total absence of a lesion was more frequent in the groups vaccinated by jet-injec- tor, as may be seen from Fig. 1-6, and in some in- stances it was correlated with total absence of tuber- culin sensitivity, indicating that jet-injection some- times failed entirely. It is important to notice in Tables 1-6 that the variation of vaccination lesions in the groups vaccinated by jet-injector is generally greater than that in comparable groups vaccinated by syringe. As the variation of the tuberculin reac- tions between the groups is similar, it seems possible that there are individual differences in the way the vaccine particles are deposited. As might have been expected, higher doses of vac- cine, whether given by syringe or by jet-injector, produced larger lesions than smaller doses. Never- theless, in Korea (Table 1, Fig. 1) the stronger dose administered by Dermo-Jet gave rise to a mean lesion that was smaller than that produced by the weaker dilution administered by syringe, in both non-reactors and reactors, the differences being statis- tically significant at the 5% level. However, in previously vaccinated children in Korea, the mean lesion in those vaccinated by Dermo-Jet seemed to be larger. Since the mean tuberculin reaction was larger in non-reactors given the stronger vaccine by Dermo- Jet than in those given the weaker dose by syringe, the result of the vaccination by Dermo-Jet can, on the whole, be considered favourable, even though variations in size of the lesions in the Dermo-Jet groups were larger and some lesions more than 20 mm in diameter were seen. 714 USE OF JET-INJECTORS IN BCG VACCINATION FIG. 2 TUBERCULIN REACTIONS AND VACCINATION LESIONS 10 WEEKS AFTER BCG VACCINATION BY SYRINGE OR BY DERMO-JET IN CHILDREN WITHOUT AN OLD BCG SCAR; NICARAGUA STUDY Children without a BCG ar Tuberculin reaction Vaccinati Size of reaction or lesion (mm) WHO 00741 In the study in Nicaragua and the studies with the Ped-O-Jet instrument the results were less favour- able. In several instances, and notably when the more satisfactory tuberculin reactions were observed, both the mean size and the variation in size of lesions resulting from vaccination byjet-injector were greater than those associated with vaccination by syringe; in no instance, however, was the same, or a higher, mean tuberculin reaction observed. The frequency of enlargement of axillary lymph nodes on the vaccination side differed from country to country, as may be seen from Table 8 in which the results for all vaccinated children have been com- bined for each country. A high frequency ofenlarged lymph nodes attributable to vaccination was seen only in Niger. In the group vaccinated by Ped- O-Jet the frequency of enlargement was significantly higher (P <0.05) than in the group vaccinated by syringe. Unfortunately, no placebo was administered by Ped-O-Jet in this study; thus, no further conclu- sions can be drawn. The relatively large numbers of enlarged lymph nodes observed in the Nicaragua study cannot be attributed to BCG vaccination, since the frequency was as high in the controls as in the vaccinated groups. Moreover, of the 53 enlarged nodes observed, 26 were found in the group of 247 children who received a placebo by syringe and the other 27 were in the group of 240 children who were given a placebo by Dermo- Jet infection. In all studies, the lymph nodes observed were rather small, their adherence to the skin (which may indicate suppuration at a later stage) was rare and suppuration did not occur. It should be kept in mind, however, that this com- plication may have been more frequent at a later stage. ADDITIONAL OBSERVATIONS Some studies were not limited to the original pro- tocol; thus, in Senegal, follow-up examinations were 715 H. G. TEN DAM AND OTHERS TABLE 8 FREQUENCY OF ENLARGEMENT OF AXILLARY LYMPH NODES ON THE SAME SIDE OF THE BODY AS THE VACCINATION IN STUDIES IN FIVE COUNTRIES Controls Syringe Jet-injector Country No. of No. of No. of No. of No. of No. of suJcs enlarged sujcs enlarged sujcs enlargedsubjects lymph nodes subjects lymph nodes ubjf lymph nodes Korea 372 0 835 5 820 3 Nicaragua 487 53 997 107 1 032 97 Niger 157 9 232 35 296 67 Senegal 123 5 226 14 231 8 France 365 6 362 4 more frequent. Observations 9 months after vac- cination confirm the results that have already been mentioned. The relative performances of the 3 Ped- O-Jet instruments employed were studied by com- paring the wheals produced by the injection. The variances of the mean wheal size appeared to differ significantly and the authors have emphasized that the instruments should be maintained in good FIG. 3 TUBERCULIN REACTIONS 13 WEEKS AFTER BCG VACCINATION BY SYRINGE OR BY PED-O-JET IN NON-REACTORS; NIGER STUDY Tuberculin reaction 40 30 20 0' 10 0 00E.. 401ob c 30 U 'In Number: 148 Mean: 13.4 0.1 ml of standard vaccine: Syringe' Number: 198 Mean: 11.1 0.1 ml of Ped.-O-Jet 8 12 16 20 24 Si'ze of reaction.(mm) W"o 00742 Lu _ 10 mechanical condition. The local wheal was also measured in studies made in France; as in Senegal, it was observed that the variance in the Ped-O-Jet group was much larger than in the syringe group. Moreover, some laboratory tests were carried out in the French studies. Vaccine, when delivered by the Ped-O-Jet instrument, showed a larger variation in the number of culturable particles than when seeded by pipette on the culture medium. Colori- metric determination of the volume delivered per dose showed similar dose-to-dose variations. DISCUSSION As a general conclusion it may be said that the jet-injectors appeared not to inject the entire dose for which they were calibrated. The importance of this characteristic and ways in which the results might be improved are discussed below. These considerations are closely connected with the problem of BCG vaccine dosages and are there- fore, in part, theoretical since it is, in fact, not known what dose of BCG vaccine will provide the maximum protection in man and it has therefore been the custom to administer the largest possible dose, i.e., a dose that produces not more than a tolerable- number of local, and especially regional, complica- tions such as suppurative lymphadenitis. It has been shown in controlled field trials that such a dose produces protection in man, and it is also known to produce post-vaccination tuberculin sensitivity al-- most as high as that attributable to infection with virulent mycobacteria. If the dose is reduced, the post-vaccination sensi- tivity is less and it must also be presumed that the. 716 t . - USE OF JET-INJECTORS IN BCG VACCINATION FIG. 4 TUBERCULIN REACTIONS AND VACCINATION LESIONS 3 MONTHS AFTER BCG VACCINATION BY SYRINGE OR BY PED-O-JET IN NON-REACTORS; SENEGAL STUDY Number: 123 Number: 123 Mean: 14.2 Mean:- 3.3 ~~~~~~~~~0.15 ml of 0.35 mng/mI voccine: Ped-O-Jet 16 20 24 0 4 8 12 16 20 Size of reaction or lesion (mm) WHO 00743 protection will be less, since until now there has been no evidence from prospective studies in man that this is not the case. Any reduction in dosage, other than that needed to reduce the number of complications to a tolerable level, is therefore unjus- tifiable. Consequently, even the slightest reduction in post-vaccination allergy, which may indicate a considerable reduction in dosage, should be avoided. When using a jet-injector it therefore seems advisable to increase the dosage in order to obtain a level of post-vaccination tuberculin sensitivity that is at least as great as that which occurs after vaccination by syringe. Unfortunately, this procedure will affect the size of the lesion. The results so far obtained indicate that the mean lesion size will probably not be excessively large, but that the variation may be considerably greater than after vaccination by sy- ringe. Therefore, when the dose is increased, the number of excessively large lesions must be expected to be relatively high. In practice, it is the frequency of large lesions, rather than the mean lesion size, that determines the acceptability of vaccination. This greater variation in lesion size therefore limits the possibility of increasing the dose given by jet- injector. However, in most countries, the present practice is to allow a, perhaps unjustifiably, large safety margin; thus an adequate increase should often be possible. As regards the means of increasing the vaccine dose, either the concentration of vaccine or the volume delivered may be increased, as indicated in Tables 1-6. The former method was more or less successful in Korea but seemed to fail in the studies in Nicaragua and in France. In Niger, only one dose was given by jet-injector; in Senegal, the vaccine was administered in standard and 50% strengths but, as significant differences were not obtained, the separate results have not been reported. In Gabon, both methods of increasing the dosage were used; increasing the volume seemed to produce a better result but it should be remarked that a single instrument was used in each group. The differences observed may therefore have been the result of tech- nical differences between instruments. In new-born infants, suppurative lymphadenitis, rather than the local lesion, is the most serious side-reaction. Its incidence has been shown to vary strongly with the dose (Guld et al., 1955; WHO Tuberculosis Research Office, 1955b). In none of the studies discussed in this article was special attention given to new-born infants and a general recom- mendation to use an increased dose of BCG when 717 H. G. TEN DAM AND OTHERS FIG. 5 TUBERCULIN REACTIONS AND VACCINATION LESIONS 9 WEEKS AFTER BCG VACCINATION BY SYRINGE OR BY PED-O-JET IN NON-REACTORS; GABON STUDY Tuberculin reaction Ni N N Ni lumber-.. 58 Mean: 12.4 lumber: 60 Mean: 11.5 1 l lomber: 46 Mecn: 12.2 lumber: 64 Mean: 9.0 I . 12 16 20 24 0 4 8 Size of reaction or lesion (mm, Vaccination lesion Number: 58 Mean. 4.8 0.1 ml of 1tmg/mI vaccine: Syringe . I 1 ,I 1 Number: 60 Mean: 4.4 0.1 ml of 1 mg/ml vaccine: Ped-O-Jet Number: 46 Mean: 5.8 0.1 ml of 0.15 mg/mI vacci ne: Syringe LR | t |w_J Number: 65 Mean: 3.6 0.1 ml of 1.5 mg/ml vaccine: Ped-O-Jet 1 1 12 16 20 WHO 00744 vaccinating by means of jet-injector does not apply to this group. More studies are, therefore, required. The present studies were not designed to assess the operational aspects of BCG vaccination by jet-injector but some incidental observations are worth mentioning. The vaccination procedure seemed to be relatively easy. Handling of the instru- ments, however, required a greater physical capacity than vaccination by syringe and the maintenance and repair of the instruments appeared to require some skill or even special facilities. In this con- nexion, it was remarked that the vaccinator who does not check the local wheal may not notice a sudden failure of the instrument. The vaccination output can be considerably higher than in a programme of conventional vaccination 40 30 20 10 0 40 2 30 0 -E 20 0 i 10 C 0 o 40 0 CPe12 30 e a 20a- 10- a 40 301- 20 10 01 0 4 8 - v 718 10s, USE OF JET-INJECTORS IN BCG VACCINATION FIG. 6 TUERCULIN REACTIONS AND VACCINATION LESIONS 9 WEEKS AFTER BCG VACCINATION BY SYRINGE OR BY PED-O-JET IN NON-REACTORS; FRANCE STUDY 40 30p- 20p- 10k e *0 0 o 40 44 ° 30 v 02 20 & 10 O0 Sif 40 - 30 20 10 0 Number: 180 Mean: 13.3 Number: 182 Mean: 13.8 4 8 12 j- lzlAzl _ . a 16 20 24 0 4 Size of reaction or lesion by syringe provided that large numbers of subjects are waiting in line for vaccination and that there are no formalities (such as registration) to be applied. In practice, such a situation may be rare. * * C. FILLASTRE, M6decin chef de la station pilote du Centre international de 1'Enfance, Chateau de Longchamp, Carrefour de Longchamp, Bois de Boulogne, Paris 16e France G. CONGE, Charg6e de recherche a l'INSERM, 3, rue Lton-Bonnat, Paris 16e, France E. ORSSAUD, Administrateur charg6e des 6tudes statis- tiques 'a la station pilote du Centre international de l'Enfance, Chateau de Longchamp, Carrefour de Longchamp, Bois de Boulogne, Paris 16e, France C. GATEFF, Adjoint technique au Secr6taire Gen6ral de I'OCEAC, P. B. 288, Yaound6, Cameroun 0.1 ml of 0.5mR/mI vaccine: Syringe Number: 181 Mean: 6.2 0. 1 mlof 0.5 mg/ml vaccine:Ped-O-Jet Number: 183 Mean: 7.0 L i t 0 I mlof 1.0 mg/mI vaccine: P.d-O-Jet 8 12 16 20 (mm) WHO 00745' A. TANAKA, WHO Statistician, Regional Tuberculosis Advisory Team, c/o WHO Regional Office for the Western Pacific, P.O. Box 2932, Manila, Philippines 0. ORTEGA RAMIREZ, Chief, Tuberculosis Programme, Ministry of Health, Nicaragua R. COLLAS, Medecin du projet de lutte contre la tubercu- lose (Niger-5), OMS, B.P. 739, Niamey, Niger J. WRIGHT, Medecin-chef des services antituberculeux de la R6publique du Niger L. CHAMBON, Directeur de l'Institut Pasteur de Dakar, Senegal M. BARME, Chef de laboratoire de l'Institut Pasteur de Dakar, Senegal U. B. TOMMASI, Mission OMS au Sen6gal H. SARRAT, Chef de laboratoire de l'Institut Pasteur -de Dakar, Senegal P. BRES, Sous-Directeur de l'Institut Pasteur de Dakar, Senegal L. DIALLO, Medecin-Inspecteur des 6coles, Dakar, S6n6gal M. GAUTHIER, Mission OMS au Sen6gal 6 719 720 H. G. TEN DAM AND OTHERS RtSUMI EMPLOI D'INJECTEURS SOUS PRESSION POUR LA VACCINATION AU BCG Les r6sultats i attendre, sur le plan technique, de l'emploi pour la vaccination au BCG de deux types d'injecteurs sous pression, le Dermo-Jet et le Ped-O-Jet, ont ete evalu6s au cours d'une serie d'etudes menees sous les auspices ou avec la participation de l'OMS au Senegal, au Niger, en Coree, au Nicaragua, au Gabon et en France. Dans chacune de ces etudes, on a compare les rende- ments respectifs de l'injection sous pression et de la vaccination classsique pratiquee au moyen d'une se- ringue et d'une aiguille. Les resultats ont e estim6s en termes de sensibilite tuberculinique postvaccinale, des dimensions de l'induration et de la destruction tissulaire locales et de l'incidence des hypertrophies ganglionnaires regionales. La vaccination pratiquee a l'aide d'un injecteur calibre de maniere a fournir la meme dose (0,1 ml) que la seringue suscite une sensibilite tuberculinique de valeur moindre que celle que confere la vaccination par la technique classique. On peut obtenir de meilleurs resultats en augmentant la dose de vaccin injectee sous pression. Ainsi, en Cor6e, l'injection par Dermo-Jet d'une dose 2,5 fois sup6rieure a entraine un niveau de sensibilite tuberculinique comparable 'a celui obtenu par injection ai la seringue d'une dose simple. Par contre, lors d'essais pratiques dans les memes conditions au Nicaragua, aucun effet analogue n'a e't observe. Dans les etudes menees avec le Ped-O-Jet, les doses n'ont pas e augmentees de plus de 50 a 100%. Bien qu'ap- paremment plus eleve qu'apres administration d'une dose simple, le niveau de la sensibilite tuberculinique ainsi r6alisd n'a en aucun cas atteint celui resultant de l'injection a la seringue d'une dose simple. Si les dimensions moyennes des r6actions tubercu- liniques, apres vaccination sous pression, etaient plus petites, les variances, en revanche, etaient du meme ordre qu'apres vaccination ia la seringue. II semble donc qu'avec les injecteurs, et de fa9on systdmatique, seule une certaine proportion de la dose soit effective- ment introduite dans la peau. D'autre part, pour des doses egales de vaccin, I'injection sous pression provo- quait des l6sions vaccinales dont les dimensions moyennes etaient plus petites qu'apres la vaccination 'a la seringue. Cette observation est compatible avec la conclusion, 6nonc6e ci-dessus, selon laquelle la dose reellement inject6e est plus faible, mais il se peut aussi qu'elle soit le resultat direct de l'emploi d'une technique d'injection diff6rente. La variance des lesions provoquees par la vaccination sous pression etait plus elevee que celle des lesions succedant a la vaccination classique. La variance des reactions tuberculiniques etant similaire avec les deux techniques, il est possible qu'il existe des differences individuelles dans les modalites d'introduction du vaccin dans la peau. Dans toutes les etudes autres que celle menee en Coree, les dimensions moyennes des lesions, et en particulier leurs variances, ont 6te plus elevees apres administration d'une dose plus forte par injection sous pression qu'apres injection 'a la seringue d'une dose simple. La difference etait particulierement nette pour les dosages qui conferaient un niveau plus satisfaisant de sensibilite tuberculinique postvaccinale. L'importance pratique de cette observation ne peut etre negligee, car cela limite les possibilites d'augmenter les doses administrees par injection sous pression. Au Niger, on a releve une plus forte incidence des hypertrophies ganglionnaires apres emploi du Ped-O-Jet qu'apres vaccination ai la seringue. Ailleurs, la frequence de ce genre de reaction a ete du meme ordre dans les deux groupes de sujets vaccines. Selon les auteurs, il convient, si l'on utilise l'injecteur sous pression, d'augmenter les doses individuelles de facon a obtenir un niveau de sensibilite tuberculinique postvaccinale 6quivalant a celui qui resulte de la vacci- nation par la technique classique, mais alors le risque de reactions indesirables s'accroit egalement. REFERENCES Edward, L. B., Palmer, C. E. & Magnus, K. (1953) BCG vaccination, Geneva, (World Health Organization: Monograph Series, No. 12), pp. 51-64 Guld, J. (1957) Bull. Wld Hlth Org., 17, 225-248 Guld, J. et al. (1955) Brit. med. J., 2, 1048-1054 Guld, J. et al. (1968) Bull. Wld Hlth Org., 39, 829-836 Palmer, C. E. & Edwards, P. Q. (1953) Brit. med. J., 1, 363-368 WHO Tuberculosis Research Office (1955a) Bull. Wld Hith Org., 12, 123-141 WHO Tuberculosis Research Office (1955b) Bull. Wld Hlth Org., 12, 143-167