Evaluation of fenitrothion for the control of malaria

Bulletin of the World Health Organization, 56 (3): 445-452 (1978) Evaluation of fenitrothion for the control of malaria R. E. FONTAINE,1 J. H. PULL,2 D. PAYNE,3 G. D. PRADHAN,4 G. P. Josm,5 J. A. PEARSON,6 M. K. THYMAKIs,7 & M. E. RAMos CAMACHO 7 Fenitrothion was evaluatedfor residual spraying in antimalaria programmes in a large- scale field trial near Kisumu, Kenya from 1972 to 1976. The insecticide was applied in a hyper/holoendemic malarious area of 200 km2 inhabited by about 50000 people. All houses and animal shelters were sprayed at a target dosage rate of2 g/m2 at 3-month intervalsfor a total of 8 consecutive spray rounds in 2 years. The malaria vectors Anopheles gambiae species A and B and A. funestus were reduced to negligible densities indoors and outdoors immediately after initiation of spraying and for 10 months after the last spray round. However, A. gambiae reappeared during the main wet season at densities high enough to re- establish low-level transmission for short periods. Spraying produced a marked and rapid decrease in both the incidence andprevalence ofmalaria. The daily probability ofacquiring malaria infection was reducedfrom 0.009 before spraying to 0.0003 under spray protection, a reduction of 96%. Data collected on a longitudinal basis indicated that sustained spray protection would reduce malaria prevalence to an asymptotic limit of 6.9% under the assumption that the inoculation and recovery rates remain stable. However, to attain malaria eradication in this type of epidemiological situation, complementary measures such as mass drug administration appear to be necessary. As part of the WHO Programme for the Evalua- tion and Testing of New Insecticides (1), a field research project was carried out jointly by the World Health Organization and the Government of Kenya from 1972 to 1976 to evaluate fenitrothion as a candidate insecticide for the control of malaria. The project was located 20 km west of Kisumu town on the shores of Lake Victoria, in a rural area with a relatively stable population where there were no 1 Extension Entomologist, University of California, Davis, CA 95616, USA. ' Epidemiologist, Division of Malaria and Other Parasitic Diseases, World Health Organization, 1211 Geneva 27, Switzerland. Reprint requests should be addressed to this author. 'Technical Officer, WHO/Nigerian Malaria Field Re- search Project, P.O. Box 671, Benin City, Nigeria. ' Entomologist, Semarang Subunit, Vector and Rodent Control Research Unit, c/o WHO Provincial Malaria Chief, Dinas Kesehatan Rakyat Propinsi Jawah Tengah, Jalan Pahlawan 1, Semarang, Indonesia. ' WHO Entomologist, P.O. Box 250, Dhanmandi, Dacca, Bangladesh. ' Sanitarian, Anopheles Control Research Unit No. 1, P.O. Box 503, Kaduna, Nigeria. 'Retired, formerly a WHO staff Member. antimalaria measures, such as drug administration or pesticide spraying for either public health or agricultural purposes. The climate is of the equatorial type, with an annual rainfall averaging 1335 mm during the 4-year period January 1972-December 1975. The mean temperature was 22.1°C with a maximum of 28.5°C and a minimum of 18.8°C. The climatological pat- tern favours perennial malaria transmission. The malaria in this region is of the hyper/holo- endemic type. Plasmodium falciparum is the domi- nant species and P. malariae is quite common; P. ovale and P. vivax are seldom encountered. The recognized malaria vectors, Anopheles gambiae (spe- cies A and B) and A. funestus, are both common throughout the area. The territory selected for the trial was divided into three zones (Fig. 1): the evaluation zone (treated area) with a population of 17 000; the comparison zone (untreated area) with a population of 3800; and the barrier zone with a population of 32 000. This last zone, 3-5 km wide, was to protect the popula- tion of the evaluation zone against reintroduction of insect vectors. 3707 445- R. E. FONTAINE ET AL. Comparison zone 122 km2 -- =: tKisumu Evaluation m: I t E= l <......,,vau ln one ,,.............._, XI.@>5*.. ::.@:¢ e86.km .............. *:**::e::* :* :X . ........-ison zone ;g-- Lake Vlctoria 0 2 WHIO 77111 Fig. 1. The trial area in Nyanza Province, Kenya. Fenitrothion (40% wettable powder) was applied inside dwellings at a fixed dosage of 2 g technical product/M2 at intervals of 3 months. The first cover- age of the treatment area was completed in August 1973 and the last in June 1975. During each of the 8 applications of the insecticide, house coverage was over 99%. The cost factors involved in the spraying opera- tions, based on an average of US$2100 per tonne of fenitrothion in 1975, were as follows: Insecticide and local shipping (78.3 tonnes) Salaries and wages Transport Spraying equipment Field station and maintenance Geographical reconnaissance Cost (US$) Percentage of total 170 140 65.11 59 740 22.85 18 260 06.99 5 770 02.22 5 460 02.08 1 950 00.75 261 320 100 The average cost of one spray round was $32 665 and each family compound cost nearly $6 to spiay once. The per caput cost per round in a population of approximately 40 000 people was $0.81, or $3.24 per annum. The activities carried out in both the evaluation and the comparison zones prior to and after the first application of fenitrothion in August 1973 were directed towards the collection of data related to: (a) parasitological variables (2), in particular malaria transmission indices and prevalence rates and their variations, under natural conditions or under the impact of the insecticide; (b) entomological variables, particularly the esti- mation of vector densities and infectivity and the impact of the insecticide on those variables following the application of the insecticide; (c) collection of vital statistics in areas with or without insecticidal protection; and (d) longitudinal observations regarding toxicity of the product and safety precautions regarding its use. PARASITOLOGICAL VARIABLES Malaria transmission indices Starting in August 1972, all susceptible infants aged 0-11 months were followed up once a month until found positive (a) prior to August 1973, in both the evaluation and comparison zones, to quantify the level of natural transmission in an undisturbed environment and (b) after August 1973 and up to April 1976 in the evaluation zone to assess the degree of residual transmission and in the compari- son zone to quantify the level of natural transmis- sion. 4 6 8km IL 4 JI I ..................... 446 FENITROTHION FOR MALARIA CONTROL 447 Table 1. Observed cumulated incidence of malaria in a cohort of 1000 newborn infants, in the absence of control measures, September 1972-August 1973 Observed monthly No. of New malaria cases Cumulated age- Age (months) parasite incidence susceptible in the cohort specific incidence rate (%) infants Per month Cumulated rate (%) 0 1000 2.4 24 24 3.6 b 1 976 18.5 181 205 20.5 2 795 22.8 181 386 38.6 3 614 28.8 177 563 56.3 4 437 25.2 110 673 67.3 5 327 22.9 108 781 78.1 6 219 28.2 62 843 84.3 7 157 23.8 37 880 88.0 8 120 18.8 23 903 90.3 9 97 (16.7) a 16 919 91.9 10 81 a Based on fewer than 10 observations. b Adjusted for 10 days' incubation. The positive cases found, after exclusion of the previously known positive cases, constitutes a state- ment of the number of new cases (or incidence) during the time stated. Blood specimens collected for microscopic examination were examined for 200 thick fields before being considered negative. The age-specific cumulated numbers and rates of new malaria cases detected by microscopic examina- tion in an initial cohort of 1000 newborn infants before and after fenitrothion application are shown in Tables 1 and 2, respectively. Establishment ofparasite prevalence rates Prevalence rates were established in the general population over 1 year of age. These surveys were carried out once every 6 months in groups selected at the beginning of the trial-8 groups of 250 people in the evaluation zone and 3 groups (total 1000 people) in the comparison zone. The results of these surveys are given in Table 3. ENTOMOLOGICAL VARIABLES a Vector densities Starting in August 1972, A. gambiae and A. funes- tus densities were measured using different types of collection in geographically representative locations a A series of papers on the entomological aspects of the project is in preparation. 448 R. E. FONTAINE ET AL. Table 2. Observed cumulated incidence of malaria in a cohort of 1000 newborn infants, October 1 973-April 1976 Area under spray coverage Comparison area Observed ~New malaria ObevdNew malariaAge mobthslerved p cases in Cumulated monthly No. of cases in Cumulated(months) morante N.onfati mthe cohot age-secific morate N.oefpt the cohort age-specificparasitce suanscpil incidence parasitce suanscpil incidenceincdene i fnt Per Cumu - rate (%) incdene i fnt Per Cumu- rate (%)rate(%) ~~month lated rt % month fated 0 1000 1000 2 2 0.3a 3.5 35 35 5.25 a 965 1 3 0.3 101 136 13.6 864 6 9 0.9 23.9 206 342 34.2 658 7 16 1.6 24.8 163 505 50.5 495 7 23 2.3 17.2 85 590 59.0 410 10 33 3.3 18.5 76 666 66.6 334 22 55 5.5 23.4 945 8 63 6.3 19.1 937 78 744 74.4 256 49 793 79.3 207 14 77 7.7 15.1 31 824 82.4 176 26 103 10.3 20.4 36 860 86.0 140 5 108 10.8 38.2 53 913 91.3 87 0 108 10.8 32.8 28 941 94.1 a Adjusted for an incubation period of 10 days. in both the evaluation and the comparison zones. In particular, man-vector contact was assessed by night-biting collections. (a) Prior to the application offenitrothion (August 1972-July 1973). In the evaluation zone, the biting rate indoors for A. gambiae was 8.4 per man per night as against 6.4 for A. funestus. The biting rate outdoors was not assessed until June 1973 before the start of the intervention phase. The combined rate for June and July averaged 1.2 bites per man per night for A. gambiae and 1.6 for A. funestus. In the comparison zone, the biting rate indoors for A. gam- biae was 10 per man per night as against 6.2 for A. funestus. (b) During the application offenitrothion (Septem- ber 1973-May 1975). After the first application of fenitrothion, A. gambiae and A. funestus densities indoors in all collections were virtually nil. After the second application, all collections were negative despite relatively high densities in the comparison 0.15 998 0.08 2 997 0.63 3 991 0.7 4 984 0.8 5 977 1.0 6 967 2.3 7 0.9 8 1.5 9 923 2.8 897 0.6 11 892 0.0 10 FENITROTHION FOR MALARIA CONTROL 449 Table 3. Parasite prevalence rates in the general population over 1 year of age before and afterfenitrothion application Crude prevalence rate Date of survey Evaluation area Comparison area Sept. 1972 64.5 (61.1) a 64.5 (59.96) a March 1973 58.1 (47.6) 58.3 (50.9) Sept. 1973 58.8 (54.2) 46.0 (41.1) Spraying started March 1974 32.1 (24.5) 56.6 (52.2) Sept. 1974 31.4 (27.2) 55.3 (51.1) March 1975 19.8 (16.0) 46.3 (41.7) Sept. 1975 24.5 (21.4) 55.3 (48.1) March 1976 17.2 (15.7) 49.0 (45.1) a Figures in parentheses are the P. falciparum rates. zone. The third application a coincided with the wet season of 1974 in March, April, and May when the monthly rainfall totalled 248, 268, and 133 mm, respectively, the highest recorded for any period during the trial. Under these conditions, A. gambiae reached high numbers in the comparison zone, averaging 92 per house in resting collections, where- as in the evaluation zone the density averaged only 0.14 per house. The biting rates were 0.23 per man per night in the evaluation zone and 53 in the comparison zone. Light traps produced the highest yield in the evaluation zone, averaging 1.2 per house against 166 in the comparison zone. Only two A. funestus females were collected. After the fourth application of the insecticide, densities were much lower in all collections owing to a combination of insecticidal effect and declining A. gambiae densities as confirmed by data from the comparison zone. A. funestus was absent in nearly all collections. In the second year of the spray phase, involving the last four of the eight spray applications, there were no exceptional conditions encountered. Cal- culated dosage rates corresponded with target rates and rainfall approximated normal trends and amounts. All collections of A. gambiae and A. funes- tus indoors were nil or virtually nil except during the wet season of 1975. Again, A. gambiae reappeared as it did in the corresponding period of 1974 after the third round despite higher dosage rates, insecticidal a The rate of application was less than 2 g/m'. build-up in houses from previous rounds, and a lower naturally occurring density in the comparison zone. Although the density indices in the collections were lower than the third round, the levels were sufficiently high to give rise to some transmission. The effect of fenitrothion spraying on the outdoor vector population involved three assessment methods: outdoor resting in pit shelters and granary huts, man-biting collections, and light traps. Man- baited net traps were employed only in the final three rounds of the trial. The results of outdoor collections showed trends similar to those indoors. The results of the spray phase clearly demon- strated a high level of effectiveness of fenitrothion residual spraying on the indoor and outdoor popula- tions of A. gambiae and A. funestus. All indices indicated that A. funestus was highly amenable to control in all seasons, but A. gambiae reappeared during the wet season at levels sufficient to maintain transmission for a limited period despite adequate spray coverage. (c) After the intervention phase. Results of house resting, man-biting, and light trap collections con- firmed a continuing insecticidal effect 399 days after the final application of fenitrothion. All collections were negative or nearly so for A. gambiae in the evaluation zone until the start of the wet season in April 1976, 312-343 days after the last application. With a steep rise in natural densities in April and May, A. gambiae again reappeared in the evaluation zone. The resting density was 1.2 per house against 63.4 in the comparison zone and man-biting rates and light trap captures showed similar trends. In June, 379-399 days after the last spray round, natural densities in the comparison zone declined sharply but increased in the evaluation zone-a certain indication of diminishing insecticidal activity. The results of outdoor collections showed trends similar to those indoors. The data for the two types of collection (outdoor resting and man-biting) were largely negative or nearly so until the wet season when a resurgence of A. gambiae occurred, reflected in a June resting density of 1.6 per granary and a biting rate of 1.1 A. funestus remained virtually absent in the post-spray period. In summary, the entomology data in the post-spray period confirmed a pronounced insecticidal effect sufficient to provide a high degree of control of A. gambiae 371 days after spraying had ended. A. funestus was observed at relatively high densities in the comparison zone but was effectively sup- pressed in the evaluation zone. The fact that the R. E. FONTAINE ET AL. insecticide continued to demonstrate activity 371 days after spraying, with significant suppression of house resting densities of A. gambiae, shows that fenitrothion could be used at longer spray cycles and/or lower dosage rates than the those employed in the trial. Infectivity rates of the vectors Prior to insecticide application. Dissections for the detection of sporozoites were carried out with mos- quitos collected during pyrethrum spray and man- biting collections. Out of 31 488 A. gambiae exam- ined, 6.56% were sporozoite positive whereas 4.3% of a total of 30 157 A. funestus were positive. During the insecticide application. Out of 1010 A. gambiae dissected in the treated area, one was found positive for sporozoites in June 1974. In the comparison area, 4.22% out of 21 480 A. gambiae were found to be positive. All 73 A. funestus collected in the treated area were dissected and none was found to be infected whereas 3.02% (out of 11 192 dissected) were found to be sporozoite positive in the untreated zone. Vital statistics Payne et al. (3) have studied the modification observed in general and infant mortality rates with changes in transmission induced by malaria control measures. A spectacular indirect benefit was recorded in the general mortality. The annual crude death rate decreased from 23.9 to 13.5 per 1000 population in 2 years. This indirect benefit, already noted by Bruce-Chwatt et al. (4) and Gramiccia et al. (5), deserves attention in the assessment of the impact of control measures. The total infant death rate calculated from cohorts of 1000 newborn infants reflected, beyond doubt, the impact of vector control measures on the survival of infants since a reduction of 40.8% was observed in this rate (157 and 93 deaths per 1000 in the unpro- tected and protected areas, respectively). Toxicity offenitrothion and safety precautions Provision was made in the spray operations for the safe storage and handling of the insecticide and the protection of spraying personnel and inhabitants of the trial area (6). The following measures were enforced: (a) The use of protective clothing by spraymen consisting of water-repellant overalls, canvas ankle boots, broad-brimmed hats of flexible plastic, and a surgical-type face mask. (b) The issue of water and soap to spraymen for removal of spray residues from exposed skin after preparation of each pump charge. (c) Daily insecticide exposure limited to six hours. (d) Rapid transport of spray crews after work to the field station. (e) Daily showers for spraymen after work and issue of freshly laundered overalls each morning before spraying. In addition, the blood cholinesterase activity in spraymen and other workers having frequent contact with the insecticide was monitored weekly by the tintometric method. Although occasional operational difficulties were experienced owing to the necessity of removing spraymen with low cholinesterase activity from spraying duties, there were no clinical symptoms of toxicity observed in spraymen or in the inhabitants of the trial area or their domestic animals. DISCUSSION An assessment of house spraying with fenitrothion on malaria incidence, based on transmission indices, was derived from the monthly infant incidence rates shown in Tables 1 and 2. The data assumes an initial cohort of 1000 new- born infants in the prespraying phase from which is derived the age-specific cumulated numbers and incidence rates of new malaria cases. The observed cumulated incidence corresponds to a daily parasite inoculation rate of 0.00958 for the prespraying phase (7). In the spraying phase, the same data as shown for the prespraying phase is given in Table 2 for the evaluation and comparison zones covering the period October 1973 to April 1976. In this period, the daily inoculation rate in the evaluation zone was 0.00037 compared with 0.009 in the unsprayed com- parison zone. Based on these rates, the effect of spraying reduced the probability of contracting malaria from once every 110 days in the comparison zone to once in 2700 days in the evaluation zone. From a residual constant inoculation rate h of 0.00037, it is possible to estimate the malaria pre- valence (Lx) in the youngest age group that could be reached if spraying of fenitrothion had continued for 450 FENITROTHION FOR MALARIA CONTROL 451 a longer period. According to the formula of Ross (8), Lx (limiting value of prevalence rate) is h/(h+r) and r is the daily recovery rate of single infection,a which is assumed to be 0.005 for P. falciparum. Using the above values of h and r, Lx will reach a value of 0.069 or 6.9%. Taking as baseline data the survey carried out in September 1973 at the time of spraying, the crude prevalence rates among the general population fell in the evaluation zone from 58.8 to 17.2 (a decrease of 70 %), whereas in the untreated zone, the prevalence rates were practically left unchanged (46.0% in September 1973 as against 49.0% in March 1976). However, residual transmission took place in the evaluation zone during the corresponding peak of transmission in the comparison zone. This is re- flected by the regression line of the P. falciparum prevalence rate which should have fallen more steep- ly if the reproduction rate Zo had been reduced to zero. In the case of Zo = 0, P. falciparum preva- lence should have fallen from 54.2% in September 1973 to 8.6% in September 1974-instead it was 27.2%. This was due to the remaining level of transmission. During the period October 1974-March 1976, the P. falciparum prevalence con- tinued to fall to a value of 15.7% with a small recrudescence in September 1975 when it reached 21.4%. It should be noted that prevalence and incidence were calculated for that fraction of the population which, according to fortnightly surveys, never left the treated area. The entomological evaluation has shown fenitro- thion to be a highly effective residual insecticide exert- ing a contact (and airborne) effect against A. gam- biae and A. funestus. The vector populations were radically reduced immediately following the finst round of indoor spraying and were maintained at a low level throughout eight rounds of the 2-year spray phase and for 1 year beyond the final spray round when observations were discontinued. The influence of indoor spraying pervaded all facets and stages of the vector population, as shown by the disappearance of vectors for long periods from unsprayed outdoor resting shelters, by the absence of outdoor biting, and by greatly reduced larval densi- ties in breeding habitats. The anthropophilic and a Taking superinfection into account would lead to a slightly lower value for the actual recovery rate R = 0.0048 as calculated by the formula R = h/[exp (h/r)-1] quoted by Dietz et al. (9). In this case the corresponding limiting value will be 7.2 %. endophilic preferences of the vectors enhanced the effect; this was best seen in results with the highly endophilic A. funestus, which was more vulnerable to control than the occasionally exophilic A. gambiae species A and the frequently exophilic species B. Although control was effective, numbers of A. gambiae rose in the treated area during the rainy season, and house densities and man-biting densities reached levels sufficient to resume malaria transmis- sion in the sprayed areas for about 10 weeks. CONCLUSIONS Indoor residual spraying with fenitrothion pro- vided good control of malaria incidence and preva- lence under conditions prevailing in the Kisumu area. The magnitude of the impact is clear from the decrease in the probability of infection from once every 110 days in the unsprayed area to once every 2700 days under spray protection, i.e., a 90% reduc- tion in the risk of contracting malaria. Under the holoendemic conditions existing in the trial area, continuation of residual spraying for a longer period would have reduced malaria preva- lence further. However, malaria eradication would require complementary measures such as mass drug distribution during the wet period. It is reasonable to expect that under less severe conditions varying from meso- to hyperendemicity, in East Africa or in other areas having similar transmission factors, fenitrothion residual spraying alone might stop transmission. The findings suggest a point of diminishing returns in the capability of the insecticide applied as a residual spray to contain explosive increases in the vector population without the addition of supple- mentary measures, as more frequent residual spray- ing or higher dosage rates may not result in a higher level of control. The high cost of the insecticide would be prohibi- tive for most programmes at the dosage rates and frequency of spray rounds employed in this trial. Its use in antimalaria programmes should be restricted to selective focal situations if the regimen of spraying employed in the trial were followed. Longer cycles and lower dosages might be adequately effective in certain epidemiological situations and need to be considered to reduce costs. Toxicological data showed that the insecticide is safe to use when recommended precautions are observed. 452 R. E. FONTAINE ET AL. ACKNOWLEDGEMENTS We wish to express our gratitude to Dr Z. Onyango and Dr John Roberts of the Kenya Ministry of Health and to Mr S. Achapa, Division of Vector-Borne Diseases, Nyanza Province for their continuous support of the project. Our thanks are due to Dr J. Hamon and Dr C. Pant, Division of Vector Biology and Control, and to Dr B. Grab, Division of Health Statistical Methodology, World Health Organization, Geneva for their constant guidance during the implementation of the field research project. We are also grateful to Dr T. Lepes and Dr A. Noguer, Division of Malaria and Other Parasitic Diseases, WHO, Geneva for their constructive suggestions. Special acknowledgement is made to Mr H. Rafatjah, Division of Vector Biology and Control, WHO, Geneva for his active participation in the planning and execution of the survey, as well as for his expert and constant assistance in the implementation of the trial. The studies were supported in part by a research grant from the United States Agency for International Development, Washington, DC, USA. REFERENCES 1. WRIGHT, J. W. Bulletin of the World Health Organiza- tion, 44: 11-12 (1971). 2. MACDONALD, G. The epidemiology and control of malaria. London, Oxford University Press, 1957. 3. PAYNE, D. ET AL. Bulletin of the World Health Organization, 54: 369-377 (1976). 4. BRUCE-CHWATT, L. J. ET AL. An experimental malaria control scheme in Ilaro, a semi-rural holo-endemic area of southern Nigeria. Report on five years' results 1949-1953. Information Bulletin No. 3, Lagos Depart- ment of Medical Services, 1955. 5. GRAMICCIA, G. & HEMPEL, J. Journal of tropicat medicine and hygiene, 75: 187-192 (1972). 6. WHO Technical Report Series, No. 513, 1973. 7. PULL, J. H. & GRAB, A. Bulletin of the World Healtk Organization, 51: 507-516 (1974). 8. Ross, R. Proceedings of the Royal Society, A, 92: 206-230 (1916). 9. DIETZ, K. ET AL. Bulletin of the World Health Organi- zation, 50: 347-357 (1974). RESUME EVALUATION DU FENITROTHION POUR LA LUTTE CONTRE LE PALUDISME Un essai pratique a grande echelle, qui s'est deroule entre 1972 et 1976 au Kenya, 'a proximite de Kisumu, a permis d'evaluer 1'efficacite de pulverisations a effet remanent de f6nitrothion dans des programmes anti- paludiques. L'insecticide a 0t6 applique dans une zone impaludee hyper- ou holoendemique de 200 km2 oii vivent quelque 50 000 ruraux. Les pulverisations ont e effectu&es sur 2 ans i raison de 2 g/m2 dans toutes les habitations et abris pour animaux lors de 8 toume'es s6par6es par un intervalle de trois mois. On a enregistre immediatement apres le debut des operations la reduction jusqu'a des densites negligeables des populations de vec- teurs du paludisme Anopheles gambiae, especes A et B, et Anopheles funestus, 'a l'interieur aussi bien qu'a F'ext&- rieur des habitations; ces effets ont persiste pendant dix mois apres la derniere application d'insecticide. Les den- sites de A. gambiae constat6es lors de sa reapparition at% cours de la principale saison de pluies ont toutefois ett& assez elev6es pour declencher la reprise d'une transmis- sion de faible intensite pendant de courtes periodes. Les. pulverisations ont eu pour effet une diminution pronon- cee et rapide de l'incidence et de la prevalence du palu- disme. Le taux journalier d'inoculation a ete reduit pour- la population prot6g6e a 0,00037 contre 0,009 avant les pulverisations, ce qui represente une diminution de 96%. de la probabilite journaliere de contracter l'infection paludique. On estime, sur la base d'une enquete longitu-- dinale, qu'une protection permanente de ce type reduirait la prevalence du paludisme a la valeur limite de 6,9 %, en supposant que les taux residuels d'inoculation et de guerison journalieres demeurent constants.