Cover Image
close this book Impact of pesticide use on health in developing countries
close this folder Part II: Relating pesticide development, manufacturing, application techniques, and regulatory control to user safety
View the document Development of safe chemical pesticides
View the document Studying the effects of pesticides on humans
View the document Safer packaging and labeling of pesticides
View the document Improving pesticide regulation in the Third World: the role of an independent hazard auditor
View the document Safer pesticide-application equipment for small-scale farmers in developing Asian countries
View the document Field evaluation of protective equipment for pesticide appliers in a tropical climate

Part II: Relating pesticide development, manufacturing, application techniques, and regulatory control to user safety

Development of safe chemical pesticides

W.F. Tordoir

Occupational Health and Toxicology, Health, Safety and Environment Division, Shell international Petroleum, The Hague, Netherlands

Pesticides must meet many requirements, of which efficacy is fundamental. Product selection, therefore, always begins with efficacy testing; if these results are promising, additional work is carried out to check for human and environmental safety. However, commercial and technical requirements must also be met. Progress has been made in both increasing the selectivity (i.e., restricting the pesticidal activity to target species) and in improving efficacy. This has made dramatically lower application rates possible. In a number of cases, a reduction in intrinsic toxcity for mammals has also been achieved. Guided by a better understanding of intrinsic toxicity and good data on personal exposure during application, product development and product stewardship can contribute to safety through improved formulations, packaging, and advice on handling and personal protection. A comparison of three insecticides as representatives of organochlorines, organophosphates, and pyrethroids shows that, with the development of the pyrethroids, a major step forward has been made in the creation of safe pesticides. Similarly, the example of a highly toxic rodenticide demonstrates that a high efficacy and sophisticated product development can compensate for high toxicity.

It is only logical that in the history of the development of chemical pesticides, nearly all attention was focused initially on the ability of a substance to kill pests; concern for safety was secondary and followed somewhat later. The drive to develop safe pesticides is reflected in the introduction of pesticide registration and the increasingly stringent requirements to meet registration criteria. In 1950, only acute toxicity tests and a 30- to 90-day feeding study using rats were required for registration. By 1960, a 2-year study of effects on rats and a 1-year study using dogs were necessary. By 1970, most current requirements were in place (Table 1). The need for registration, combined with requirements for toxicity testing and restrictions for marketing certain pesticides, have, in turn, had a positive effect on the development of safe pesticides. ("Safe" is used here in relation to effects on human health.)

Table 1. Requirements and subrequirements met by the "ideal" pesticide.

Finding a molecule with specific pesticidal properties and low toxicity for nontarget species is still the first step in a long and complicated process. Today, however, modern pesticides must satisfy so many requirements that it seems easier to find a needle in a haystack than to discover a new molecule that meets the full specifications and still yields a profit. The essential requirements of a new pesticide are:

· Specific pesticidal activity;

· Safe for the crops on which it is applied and not seriously affecting nontarget species;

· Not inducing resistance in the target pests;

· Safe to health;

· Safe to the environment;

· Good market opportunities;

· Able to be produced in sufficient quantities in an acceptable manner (with respect to health, environment, and economics);

· Good formulation possibilities; and

· Stable when stored.

Failure to meet one of these requirements will jeopardize successful development of a project. Each of these core requirements can be split up into many individual subrequirements; even one unfulfilled subrequirement can be decisive in not pursuing product development.


Finding a pesticide

The usual process of selecting new chemicals for pesticide use is by elimination. The first step must be testing for pesticidal activity, thereby eliminating inactive candidates. Screening a range of molecules that are structurally similar to a known pesticide is often the approach adopted, although sometimes fundamental studies of the physiological properties or the biochemistry of the target pest provide clues for identifying the class of chemicals that would interfere with its essential biological processes.

Agrochemicals are also identified through a method of widespread screening - the testing of any new chemical against species representing target pests. Discovering a new pesticide often requires luck and scientific curiosity, combined with perseverance (Fig. 1).

Fig. 1. Stages and costs of agrochemical research and development.

Early testing of pesticidal activity is carried out in the laboratory; testing for plant toxicity and for the induction of resistance in pests follows shortly. Promising results in laboratory tests lead to small-scale field studies and limited toxicity data; initially, only ranges of oral and dermal acute toxicity are identified. If a new chemical survives to this stage, a market survey is carried out to obtain preliminary information about sales prospects. If a commercial future seems reasonable, further efficacy testing in the field and toxicity testing will be undertaken. The results of efficacy and toxicology testing are reviewed regularly and route scouting is used to determine the most economic, safe, environmentally sound, and technically feasible process of manufacture.

Assuming that all goes well (i.e., successful field-testing, no unpleasant surprises from toxicity testing, promising market prospects, and successful route-scouting), the first preliminary registration for experimental use in larger field studies is applied for (Fig. 2). These tests provide the opportunity for monitoring exposure of those applying the pesticides under realistic conditions. Exposure data are needed, in combination with animal toxicity information, for risk assessment. Early generation of exposure data facilitates the development of safe products and helps avoid unpleasant surprises (Tordoir 1980).

Fig. 2. Stages in developing and registering a new agrochemical. (*Field evaluation in various parts of the world.)

At this stage, process development begins. Laboratory and, later, pilot-plant production, using the most promising process found during route-scouting, are established. While toxicity and efficacy testing continue and the registration procedure expands, product development enters an engineering phase, in which a full-scale plant is designed. Experts are consulted to ensure an optimum design with respect to health, safety, and environment.

When the engineering phase is well under way, a crucial point is reached in the continuous process of reviewing the feasibility of the project. A construction site must be chosen, which may require obtaining permits, and construction begun. Once construction starts, the costs of the project rise exponentially; a thorough review of all aspects of the project is essential. All core requirements (see above) must be sufficiently met to justify the decision on plant construction.

The last phase, which is actually open ended, is the product development phase. Much research is needed to determine the best formulations to meet efficacy, application, health, safety, and environment requirements without becoming too costly. Another aspect to be considered is packaging. The amount of material per unit and the nature of the packaging can contribute to the safety of the product sold. Commonly, only a few of the many thousands of chemicals tested ever reach the manufacturing and marketing stage; the industry average is l in 25 000.

To confirm the safety of pesticides, field studies must be carried out once a product is to be marketed. Such studies assess and measure personal exposure and check for any biological or clinical effects. In many situations, biological monitoring is the method chosen for personal exposure measurement as it provides data on the total absorbed dose via all routes of entry (van Sittert 1986).

Ideally, regular health surveillance, including biological monitoring, should be carried out on workers engaged in pesticide manufacturing, formulation, and application. These data should also be used for epidemiologic evaluations of possible chronic adverse health effects.


Progress in improving safety

Pesticides in use at the turn of the century typically had a broad spectrum of biological activity and a low level of pesticidal activity. They were not particularly toxic to humans, but had to be applied in large quantities (compared with current application rates) to compensate for their low potency. The resulting human health hazard could, therefore, become significant.

Research has focused primarily on improving selectivity, i.e., enhancing the pesticidal activity on target species while leaving nontarget species, including man, more or less untouched. In the process, a higher efficacy is usually also obtained.

Qualitative selectivity is achieved when a pesticide interferes with a process or processes that occur only in the target species. For example, inhibition (by an acylurea pesticide) of chitin formation, which is essential for the formation of the legs, shield, etc., of insects, does not affect mammals. Interfering with photosynthesis in weeds does not harm birds and mammals as these organisms do not photosynthesize. Herbicides may, therefore, have a favourable toxicity profile (Table 2).

Table 2. Comparative toxicity of several herbicides, aspirin, and caffeine.

Quantitative selectivity, on the other hand, is based on differences between the target and nontarget species that determine susceptibility, e.g., skin-penetration rate, metabolism (toxification or detoxification), and excretion rate. A drawback of quantitative selectivity is the possibility that the target species becomes resistant. The properties that make nontarget species less susceptible to the pesticide may, in a number of generations, be acquired by the target species through genetic selection. Oppenoorth and Besemer (Inauguration speeches, Agricultural University, Wageningen, 1976, unpublished) point out that selectivity can also be improved by choosing the optimum time, place, and method of application; by improving the formulation; and by mixing a pesticide with other chemicals.

In certain cases, higher pesticidal activity can also be achieved by removing the less-active optical isomers (e.g., in pyrethroids) or less-active and nonactive components (e.g., removing 1,2-dichloropropane from a mixture of chlorinated C3 hydrocarbons that is sold as a soil fumigant).

Progress in improving the selectivity and efficacy of pesticides is demonstrated by the fact that application rates have decreased several orders of magnitude since the beginning of the century (Fig. 3). Lower application rates also mean less exposure for those applying the pesticides in the field. Insofar as the increase in pesticidal activity has been matched by an increase in target species selectivity, an overall reduction in the risk to nontarget organisms (including humans) has been achieved.

Fig. 3. Application rates (logarithmic scale) for crop protection chemicals plotted against the date of their introduction to illustrate the progressive increase in activity. Rates are representative values from the range employed in commercial practice for each compound (I.J. Graham-Bryce, Royal Society, London, UK, 1981).

Although the differences in toxicity between the target species and humans is of paramount importance, other factors can influence the health hazard. As skin contamination is the most important route of exposure during pesticide application, the degree of skin penetration of a substance has a significant bearing on the risk of intoxication. The amount of active ingredient used, the type of formulation, and warning properties (odour and taste) are other important factors.

Key toxicity parameters and other features of representatives of four groups of pesticides that have been developed since the 1940s are compared to illustrate the progress made in improving the safety of pesticides (Table 3). Dieldrin represents organochlorines, a highly effective group of insecticides that includes dichlorodiphenyltrichloroethane (DDT) and has been in use since the early 1940s. Dichlorvos belongs to the large group of organophosphates that became available in the 1950s (it is not possible to find a true representative of this group because of the large differences in toxicity among the related substances). Cypermethrin is a pyrethroid, a group that was developed in the 1960s and 1970s. The rodenticide, flocoumafen, represents the new generation of anticoagulants, the so-called "super-warfarins," developed in the past decade.

Table 3. Toxicity and other characteristics of three insecticides and one rodenticide.

Dieldrin has oral and dermal LD50 values (dose lethal to 50% of animals tested) that place it in the highly hazardous class IB in the World Health Organization's (WHO) classification system. Low NOAELs ("No adverse effect level") in the 90-day and chronic feeding studies also point to a high systemic toxicity. In spite of its toxicity and rapid skin penetration, lethal intoxication from occupational exposure has been rare over 40 years of use; only one case has been reported (IPCS 1989; van Raalte, unpublished paper presented at the Conference on occupational health, Caracas, Venezuela, 1965). Clinical recovery from intoxication, occurring during manufacture and formulation, was always complete (eager 1970).

The persistence of dieldrin in the environment was initially viewed as an advantage because it was associated with long-lasting pesticidal activity. However, in light of ecotoxicological consequences, it is now considered a major disadvantage. Although dieldrin and the other organochlorines have good records of safety in regard to human health, their use has been restricted because of ecotoxicological considerations and, in North America, because of a perceived cancer hazard.

Dichlorvos should also be classified as highly hazardous, because of its acute toxicity in laboratory animals and its skin penetrating properties. The 90-day and chronic feeding studies show NOAELs 10 and 20 times higher, respectively, than dieldrin, although observed carcinogenic effects in laboratory animals are not relevant to man. No reports of lethal work-related intoxications with dichlorvos are known, but other organophosphates, in particular parathion and malathion, have caused many deaths despite the availability of specific antidotes. Certain organophosphates, now withdrawn, cause irreversible neurological defects in patients recovering from intoxications (IPCS 1986). Not all organophosphates represented a step toward safer pesticides. However, their rapid degradation in the environment represented an improvement from an ecotoxicological point of view.

The pyrethroid, cypermethrin, is moderately hazardous (class II) orally and slightly hazardous (class III) when absorbed through the skin. NOAELs in the 90-day and chronic feeding studies are 10 times those of dichlorvos and 100 times higher than those of dieldrin. Metabolism and excretion are rapid and other toxicity parameters are also favourable. The lower toxicity of pyrethroids, combined with their slow skin penetration make them safer to use. Skin sensations (tingling) that occur with overexposure usually disappear within a few hours and leave no aftereffects. These sensations act as a warning signal that exposure should be better controlled.

Reported cases of occupational systemic poisoning with pyrethroids are rare. However, He et al. (1989) reported 229 cases in China, 2 of which were fatal. The explanation was a failure to observe basic principles of industrial hygiene (He et al. 1988, 1989). With respect to health and safety, however, the development and use of the pyrethroids was an improvement over the more toxic organophosphates.

Flocoumafen, a representative of the latest generation of anticoagulants, was included in this overview to illustrate that extreme toxicity according to the WHO classification, does not necessarily mean a high risk to human health. Although flocoumafen is classified as extremely hazardous (class IA), it is also an extremely efficient rodenticide. The oral and dermal LD50 values of 0.25 and 0.54 mg/kg, respectively, and a 90-day NOAEL apply to the pure substance, but the formulation used to poison bait is almost nontoxic because it contains only 50 ppm (w/w) of the active ingredient. In addition, during product development, a deliberate effort was made to make poisoned bait unattractive to nontarget species by incorporating it into wax blocks with a deterrent colour (blue) and a bitter taste. These measures have rendered a potentially hazardous product safe for practical use. In the unlikely event of an intoxication, a specific antidote (vitamin K) is available. No cases of human intoxication have been reported.



Pesticides must meet many requirements, the most important of which is efficacy. Therefore, selection always begins with efficacy testing. If results are promising, additional work will be carried out to test for human and environmental safety. Commercial and technical requirements must also be met. Progress in increasing specificity and improving efficacy has made possible dramatically lower application rates, resulting in a major reduction in exposure of people applying the substances. In a number of cases, a reduction in intrinsic toxicity for mammals has also been achieved.

Guided by better understanding of intrinsic toxicity and data on personal exposure during application, product development and product stewardship can contribute to safety through improved formulations and packaging and advice on handling and personal protection. The comparison of the major groups of pesticides shows that with the development of the pyrethroids, a major step forward has been made in the creation of safe pesticides. Similarly, rodenticide is an example of a substance whose high efficacy and sophisticated product development compensate for high toxicity, allowing a final product that is safe for humans.

Graham-Bryce, I.J. 1990. Shell Agriculture, 6, 20-22.

He, F.; Sun, J.; Han, K.; Wu, Y.; Yao, P.; Wang, S.; Liu, L. 1988. Effects of pyrethroid insecticides on subjects engaged in packaging pyrethroids. British Journal of Industrial Medicine, 45, 548-551.

He, F.; Wang, S.; Liu, L.; Chen, S.; Zhang, Z.; Sun, J. 1989. Clinical manifestations and diagnosis of acute pyrethroid poisoning. Archives of Toxicology, 63, 54-58.

IPCS (International Programme on Chemical Safety). 1986. Organophosphorus insecticides: a general introduction. World Health Organization, Geneva, Switzerland. Environmental Health Criteria 63.

____1989. Aldrin and dieldrin. World Health Organization, Geneva, Switzerland. Environmental Health Criteria 91.

Jager, K.W. 1970. Aldrin, dieldrin, endrin and telodrin: an epidemiological and toxicological study of long-term occupational exposure. Elsevier, Amsterdam, Netherlands.

van Sittert, N.J. 1986. Report of rapporteur. Toxicology Letters 33, 205-213.

Tordoir, W.F. 1980. Field studies monitoring exposure and effects in the development of pesticides. In Tordoir, W.F.; van Heemstra-Lequin, E.A.H., ea., Field worker exposure during pesticide application. Elsevier, New York, NY, USA. Studies in Environmental Sciences 7, pp. 21-26.



Studying the effects of pesticides on humans

M. Maroni

International Centre for Pesticide Safety, Busto Garolfo, Milan, Italy

Assessment of the health risks of pesticides to humans and regulation of their use are based on experiments on laboratory animals. Concern over the applicability of such studies prompted discussion at the 9th International Workshop of the International Commission on Occupational Health (ICOH). Highlights of the discussion and recommendations made at the workshop are presented in this paper, along with a review of difficulties associated with studies on people exposed to pesticides.

The use of human-exposure and health data for improving assessment of the toxicological risk of pesticides and setting regulations to govern their use was a topic of discussion at the 9th International Workshop of the International Commission on Occupational Health (ICOH 1990). There was concern that the regulation of pesticides is based mainly on data generated from animal-toxicology studies and only to a limited extent on information about effects on human health.

The aim of the workshop was to discuss studies on human exposure to pesticides and the resulting health effects, identify advantages and limitations on the use of such data by regulatory bodies compared with animal data, and to formulate recommendations for ICOH. Participants were 28 experts from 9 countries representing international organizations, governmental agencies, academia, and industry.

Presentations included: four on the generation of human-exposure data in field studies, using dermal-deposit measurements and biological monitoring; four on health effects in field workers applying pesticides, using biological and clinical tests; an epidemiologic study of a large farming population in Canada, investigating possible associations between the use of pesticides and health effects; three on the development of protocols for conducting epidemiological studies; the limitations of data on animal carcinogenicity for predicting carcinogenic risk in humans; and the activities of the International Centre for Pesticide Safety (ICPS).

Participants at the meeting endorsed five recommendations:

· Preliminary registration, registration, and reregistration of pesticides is frequently, if not exclusively, based on the evaluation of comprehensive animal data. To enhance and improve assessment of risk to human health, more use should be made of information based on human exposure and health effects.

· In addition to generating hypotheses, epidemiological studies should be designed and used to verify hypotheses. Particular attention should be paid to verification of specific exposures, selection of controls, and occurrence of mixed exposure. Whenever possible, prospective records on the use of pesticides should be kept, preferably linked to individuals or groups of individuals.

· Various techniques and methods for assessment of exposure, such as analyses of data on use, measurement of external exposure, and biological monitoring, each have a specific contribution to make. Combining the various techniques and methods will increase the accuracy of an assessment.

· Data on early biological effects may serve as indicators of possible health effects. Case-by-case evaluation will be helpful in assessing the clinical significance of a biological effect.

· Data on health or biological effects, obtained in situations with multiple exposures, should be used to stimulate measures to reduce pesticide exposure.

General discussion during the meeting highlighted several points. In assessment of toxicological risk of pesticides, the experimental model by which single compounds are tested is limiting. In practice, agricultural workers are commonly exposed to formulations and mixtures of pesticides. Although it is not feasible to test all possible mixtures or associations of compounds encountered by users, when there is reason to suspect that an interaction between two pesticides can occur (e.g., based on the mechanism of action), specific experiments with mixtures should be carried out. However, in most cases, mixed exposure in real life will remain an unprecedented experiment, emphasizing the importance of human observations and epidemiology in detecting additive or synergistic effects.

Because of lack of statistical power or other methodological flaws, experiments to investigate carcinogenicity in humans may produce false-negative results. On the other hand, animal experiments to investigate carcinogenicity of chemicals are carried out in extreme conditions, i.e., a very high dose is administered over a prolonged period. This method is appropriate to detect the carcinogenic potential of a molecule even through testing a limited number of animals, but it generates information that is more qualitative than quantitative. For pesticide users, who are exposed to doses four or five orders of magnitude smaller than those used on experimental animals, the assessment of the risk is a quantitative issue; lifetime risks of cancer of 10(-2) and 10(-8) are entirely different and would produce markedly diverging consequences in terms of risk management. Furthermore, high-low dose extrapolation with mathematical models has very weak scientific support and tends to be an administrative procedure rather than a real scientific evaluation of the risk.

Discussion highlighted the importance of data on personal exposure doses. In terms of information, an epidemiological study in a small cohort with accurate personal-exposure data may be equal or even better than one in a large cohort with poorly defined personal exposures.

Currently, health-risk assessment of human exposure to pesticides is mainly based on data from experiments on animals. Little information is available about effects on humans because of the lack of "strong" data from epidemiological studies. This problem may be overcome by applying a more scientific approach in health-risk assessment. Knowledge of the pharmacokinetics and pharmacodynamics of a compound in different species must be increased by carrying out studies on the mechanism of action.

To obtain "good" human data, investigators should critically review the protocol for future epidemiological studies and address the question of whether the effort required in such a study is proportional to the possible outcome. Specific questions should be addressed, such as the possibility of accurate exposure measurements, the selection of a suitable control group in case-control studies, and the feasibility of collecting accurate data on health status of exposed and unexposed people.

Cohort studies of factory workers who manufacture pesticides can minimize these problems. In these studies, controls can be selected among unexposed workers in the same factory or nearby factories where workers are exposed to different chemicals or processes. Employment records can be used to obtain dates of employment and job titles and, thereby, identify those who had the opportunity to be exposed and the duration of exposure. If subgroups of workers are exposed to only a few pesticides or chemicals used in production, then it is possible to relate excess incidence or mortality to a particular exposure. Such studies often suffer from having a very small sample size; evaluation of workers in more than one factory and in more than one country where the same pesticides are produced may be necessary.

In choosing a reference group, using two similar industrial populations for comparison eliminates the healthy-worker effect, which may appear when using the whole general population. One should, in such cases, select a second industry with workers of similar social and economic backgrounds so that diet, physical demands, and other life-style factors are likely to be similar.

One way to handle the comparison problem is the selection of the unexposed or lowest exposed group within the cohort as an internal reference. When this is done, care should be taken to assure similarities in duration of follow-up and in the size of the groups. In case-control studies, controls may come either from the general population, often referred to as community controls, or from hospitals where cases were identified. Three criteria have been proposed for the selection of controls:

· Controls should come from the same registry (e.g., hospital, mortality records) as the index cases.

· Control illness should be unrelated to the risk factors under study (i.e., a given risk factor under study should not predispose one to getting the control illness).

· Control illness should be similar to the illness under study, bearing in mind the factors influencing its appearance (or referral) in the registry.

One might want to add a second control if these criteria are not met.

If farmers are under study, then ideally, the control group should also be farmers. Farming is a unique way of life, not only in terms of chemical exposure, but also with respect to diet, physical activity, exposure to viruses, habits, and life-style. Opportunity for exposure to viruses, for example, may be affected by the presence of intermediate hosts, including livestock or even particular crops.

In carrying out epidemiological studies, accurate measurement of exposure is important because errors may reduce estimates of relative risk and dampen dose-response gradients. Misclassification of toxic substance increases the chance of attributing elevated cancer risk among exposed people to the wrong agent. Employment records are commonly used to assess exposure in cohort studies and interviews are used to determine exposure in case-control studies. However, these techniques are of use in both types of study and should be used, when possible, to test and improve the reliability of exposure evaluations.

Information can be obtained from different sources to increase confidence in exposure estimates. Sources include suppliers who sell pesticides to the subjects under study, records of applicator companies, interviews about the type of equipment used for application, and duration and frequency of application coupled with biochemical monitoring data. For populations, estimates of exposure may be developed by a panel of experts, including agricultural scientists, entomologists, and industrial hygienists familiar with patterns of pesticide use. Biochemical factors can provide better measures of delivered dose than estimates based on concentration of substances in patches and ambient air or those derived from job descriptions or interviews. However, biochemical measures are typically only available for a sample of study subjects, for only a few pesticides, and at only specific times.

Ideally, to allow epidemiological surveillance of a farming population using pesticides, it would be necessary to implement a system of registration that would identify the exposed subjects and specify and quantify exposure. This registry could be linked to the licence required in some countries to buy and use some classes of pesticides.

An experiment on the feasibility of this approach over a large geographical area is currently under way in Lombardy, Italy. The ICPS intends to establish an epidemiological observatory through which people and pesticide usage will be registered and specific epidemiological investigations carried out on selected subgroups of the farming population. This project will include about 500 000 farmers and cover an area equal to one-eighth of the country.

ICOH (International Commission on Occupational Health). 1990. Proceedings of the 9th International workshop, 2-4 May 1990, International Centre for Pesticide Safety, Busto Garolfo, Milan, Italy. ICOH, Geneva, Switzerland.



Safer packaging and labeling of pesticides

L.S. Dollimore

Shell International Chemical Company Ltd, London, UK

During the past 5 years, major developments in the area of pesticide packaging and labeling, particularly the use of pictograms, have improved safety for the user and the environment. Pictograms are symbols indicating essential advice and warning messages to the user. Introduced only a few years ago, they are now being used in at least 66 countries, mainly in the developing world. The most significant advances in packaging have been in the use of new plastics and container designs. New products, such as polyethylene terephthalate and fluorinated high-density polyethylene, are able to withstand aromatic solvents, allowing many new safety features, such as simple-to-use dispensers, to be built into containers.

In the past 5 years, major changes in pesticide labeling and packaging have contributed significantly to safeguarding not only end-users and those in the supply chain, but also the general public and the environment. A major stimulus was the International code of conduct on the distribution and use of pesticides (FAO 1986), which outlines rules of responsible practice for both governments and the crop-protection industry.



One of the recommendations of the Food and Agriculture Organization of the United Nations (FAO) was to "include appropriate symbols and pictograms whenever possible, in addition to written instructions, warnings and precautions." A pictogram is a symbol designed to convey a message, e.g., the no-smoking symbol. The message should be recognizable at a glance.

In the late 1970s, a specialist on information graphics, Shirley Parfitt, was working on an FAO project in Bangladesh when her illiterate maid was poisoned through mishandling of an agrochemical product. The incident convinced Parfitt of the need for pesticide labels to contain easily understandable symbols that convey safety warnings and advice. Using her artistic skills, she designed her own symbols and approached Shell International for sponsorship and agrochemical expertise. As a result of this collaboration and field-testing in six developing countries, a labeling system was developed (Fig. 1).

Fig. 1. Pictogram labeling system developed by Parfitt and Shell International.

In 1984, these pictograms were submitted to FAO for consideration and endorsement for worldwide use. FAO requested more exhaustive field-testing to ensure that the symbols were understandable to people of diverse cultures and with low levels of literacy. In response, the International Association of Agrochemical Manufacturers established a pictogram working group comprising labeling, safety, and environmental specialists and an FAO representative.

The working group reviewed previous work on agrochemical labeling systems using pictograms and examined general international and national warning and advice symbols, such as road and factory signs. Two decisions had to be made: which messages to portray and where to locate the symbols. Because of space limitations on labels, only the most important messages could be portrayed and these were to be selected on the basis of incident experience. For example, not wearing gloves when handling pesticide concentrates has led to far more incidents than smoking during product use. In fact no documented incident could be found relating to the latter. Space limitations also led the working group to recommend that the pictograms be placed in the hazard warning band recommended by the FAO.

Three graphic designers, including Parfitt, were given full background information, and asked to produce a set of pictograms of their own design (Fig. 2: bands A, B, and C). An important aspect of the designs was the linking of activity pictograms with advice pictograms, e.g., "when handling the concentrate - wear gloves."

Fig. 2. Sample label with three sets of pictograms presented by designers to the working group. (Original size of label = 18 x 28 cm.)

An international survey was carried out by field staff of the FAO, extension services, industry, and other organizations, using a questionnaire and interview procedure. The length of the interview ranged from 40 min to over 2 h in one case. Of the 3 000 questionnaires distributed, about I 000 were returned completed from 42 countries (Table 1).

Table 1. Regional distribution of pictogram survey.

The questionnaires were analyzed according to such factors as age, educational background, literacy, and occupation (Table 2). The most surprising finding was that only 80% of those interviewed understood the meaning of the skull-and-crossbones symbol, which has been in use for more than 50 years. It was reassuring, therefore, to find a high level of spontaneous understanding of the proposed pictograms; immediate understanding of the symbols ranged from 60% to 85% (Fig. 3). The most difficult concept to convey without explanation was the linkage of advice pictograms to activity pictograms. Of those interviewed, 90% considered pictograms a useful way to provide safety instructions on labels.

Table 2. Characteristics of the population surveyed (n = 940).

Fig. 3. Pictograms preferred by the majority of people surveyed.

Both the International Association of Agrochemical Manufacturers and the FAO began promoting the use of the recommended pictograms in 1988. A survey of 96 countries in 1990 revealed that 69% were using them (Table 3).

Table 3. Use of using pictograms, 1990.

It is reassuring to observe the high level of use in Africa and the Middle East and, to a lesser degree, in the Far East and Latin America, although an impediment to early use in some countries has been the need to change national legislation. Use of pictograms has been rejected in the USA, Canada, and many European countries, where they are thought to be for developing countries only. However, these countries have both illiterate and immigrant groups. They also make widespread use of pictograms in road signs and factory warnings. Moreover, many people do not read label instructions and warnings before using chemicals. Developed country arrogance is suspected in these cases, although once again the daunting task of changing legislation is no doubt a factor.



Pesticide containers provide a safety barrier between those handling the container and its chemical contents. Container design has a major effect on minimizing the risk of exposure in transferring the product to a spray tank. Liquid agrochemical products usually present the most difficult packaging problems and the greatest potential risk of exposure to the applicator.

Major changes in agrochemical packaging occurred during the 1980s and further progress is expected during the 1990s. These developments are a result of a number of factors:

· Introduction of package-performance tests by the United Nations;

· Introduction of high-activity and high-value pesticides leading to lower application rates and smaller package sizes;

· Advances in plastics and moulding technology; and

· Collaboration within the industry to improve user safety and convenience and product shelf-life.

Changes in small packages for the small-scale farmer have probably had the greatest impact on safety. In the past, most small containers were glass or tin plate: the former were too fragile and had narrow necks that caused spattering during pouring; the latter dented easily and had leaky seams and a metal insert under the cap, which was difficult and often hazardous to pry out. Both materials were severely limiting in terms of design flexibility. Furthermore, people frequently used the caps as measures. Although the common plastics (high-density polyethylene, polyvinyl chloride, and polyethylene) offered design flexibility and were suitable for aqueous products, they could not be used for aromatic solvents, which are the base for most liquid agrochemical formulations.

A major breakthrough was achieved with the discovery of polyethylene terephthalate (PET). This plastic is widely used for carbonated soft drinks and, more recently, for alcoholic spirits sold by airlines. When this polymer is biaxially stretched during blow-moulding, it produces a bottle that is compatible with aromatic solvents. The potential benefits of this plastic for the agrochemical industry were seen immediately by three companies, who collaborated to carry out the basic testing. PET bottles are now widely used for agrochemicals. They are transparent and very tough. New bottles have design features such as wide necks to avoid spattering.

However, farmers continued to use the cap of the PET bottles as a measuring container, thereby contaminating the outside of the bottle. Shell overcame this problem by designing a standard volume dispenser that is welded into the neck of the bottle during manufacture. Not only does this prevent the cap being used, it also allows for precise measurement and keeps the contents from spilling if the bottle is accidentally knocked over when the cap is off. A further safety benefit is that the bottles cannot be reused without cutting the neck off. PET bottles with dispensers have proven to be popular with small-scale farmers in Africa. The principal limitations of PET bottles are their unsuitability for ketones or products that are susceptible to water degradation because water vapour can permeate the bottle wall.

Two other recent developments in container technology have been the use of fluorinated high-density polyethylene (HDPE) and the design of multilayered containers. In the former, gaseous fluorine is included either during or after the blow-moulding process. This forms an extremely thin layer of fluorinated HDPE, which is highly resistant to the solvents commonly used in agrochemicals. A layer of only 150 A is usually considered sufficient.

Containers with walls comprising two or more polymer layers are produced by coextrusion. Bottle strength is attained from a support layer, usually HDPE, and resistance to the chemical product is provided by a layer of barrier material such as polyamide (nylon). The various layers are bonded together using adhesives.

Plastics permit considerable flexibility in packaging agrochemicals. For example, 5-L tins are not easy to handle when wearing protective gloves, whereas easy-to-grip handles can be incorporated into plastic containers of this size. A common problem with early plastic containers was the retention of product in the hollow handles. This problem has been overcome by pinching off the base of the handles during blow-moulding to prevent liquid ingress. Similarly, wider necks on plastic bottles prevent splashing of contents during pouring. Drainability of containers has also been improved to reduce risk if the container is reused for drinking water or foodstuffs. This is a particular advantage of fluorinated HDPE because fluorination makes the inner surface slippery and facilitates draining.

A common constraint to the use of new plastic containers is that many developing countries require agrochemical containers to be manufactured locally to minimize foreign-currency expenditures. Although PET blow-moulding is becoming common because of the high-volume needs of the soft drinks industry, the technology needed to produce HDPE and coextrusion bottles restricts their manufacture to developed countries.

Speculating on future developments in agrochemical containers and packaging for developing countries, it is useful to consider recent advances in the USA and Europe. Returnable containers, probably the most important innovation in terms of safety, are designed to have a long life. Made of stainless steel or plastics, they can be returned to the supplier for refilling. This eliminates the problems of container misuse and disposal. Returnable containers are also linked to a closed system designed to minimize the risk of exposure during transfer from the container to the tractor or aircraft. However, this approach requires a good supply and transport infrastructure, often lacking in developing countries.

A resurgence of interest in water-soluble sachets has also occurred as this technology has improved. Although these sachets were originally designed to contain powders, they are now being used for liquid formulations. Finally, there are signs that rigid plastic containers may be replaced in many cases by "bag-in-the-box" packaging, which has proved successful for the supply of wine. This will certainly facilitate disposal and prevent unauthorized reuse.

FAO (Food and Agricultural Organization of the United Nations). 1986. International code of conduct on the distribution and use of pesticides. FAO, Rome, Italy. 31 pp.



Improving pesticide regulation in the Third World: the role of an independent hazard auditor

M.E. Loevinsohn

Consultant in Applied Ecology, Butare, Rwanda

Central to the environmental and health hazards created by the expanding use of pesticides in developing countries is the weakness of national regulatory agencies. International efforts to support these institutions include the establishment of a Hazard Audit Organization to assess the pesticide industry's adherence to accepted standards of health and environmental protection. An independent evaluation by a hazard auditor may be attractive to all parties in the long-standing confrontation over the control of pesticide technology: the industry, public interest groups, developing and developed countries, and international agencies. One approach to implementing the concept is proposed and initial responses to the proposal are reported.

The papers presented at this symposium join a growing body of evidence of the effect on human health and the environment caused by the rapid increase in pesticide use in developing countries. Chemicals of sometimes extreme human or environmental toxicity are transported, stored, used, and discarded in ways that expose people and other nontarget organisms to significant hazard.

The urgent need for effective regulation of these hazards, however, contrasts starkly with national capacity, which is limited in much of the Third World. More than 50% of developing countries have no legislation enabling government to regulate the marketing of pesticides or limit their availability to particular areas or users; in Africa, the proportion is 76% (FAO 1989). Even where adequate legislation exists, regulatory agencies are often unable to assess pesticide hazards in light of local conditions or to enforce the decisions they reach, because of a lack of qualified personnel, inadequate resources, or interference.

Current initiatives, at the international level, aimed at improving this state of affairs are discussed in this paper. A novel approach, the concept of a "pesticide hazards auditor," who would build upon and supplement these initiatives has been promoted over the last year with support from the International Development Research Centre (IDRC). Initial reactions to this proposal from developing and developed countries, international agencies, the pesticide industry, and consumer, environmental, and labour groups are described.


International initiatives

Technical assistance and information exchange

A number of bilateral and multilateral aid organizations have launched programs aimed at increasing the skills and resources available to pesticide regulatory agencies in developing countries. The Food and Agriculture Organization (FAO), United Nations Environmental Program (UNEP), the World Health Organization (WHO), the US Agency for International Development (USAID), and Germany's Agency for Technical Cooperation (GTZ), among others, are providing training, analytical equipment, information systems, continuing support in the evaluation of risks and benefits, and advice on legislative reform. In several cases, as in FAO's programs in the Far East and Africa, this assistance is organized on a regional basis. The task, however, is immense; fewer than one-quarter of developing countries claim to have received any technical assistance (FAO 1989).

A major cause of concern is the international trade in highly toxic pesticides, particularly the export to developing countries of products banned or severely restricted in the country of manufacture. The USA, UK, and the European Community have instituted schemes to notify importing countries of shipments of unregistered or severely restricted pesticides. In practice, notifications are often received well after the pesticides have arrived and do little to enable importing countries to control hazardous imports (Pallemaerts 1988).

Nongovernmental organizations (NGOs), with support from many developing countries, have mounted a determined lobbying effort within the governing councils of FAO and UNEP in favour of more restrictive schemes based on the principle of "prior informed consent" (PIC), whereby a designated authority in the importing country must explicitly agree to the import before it can take place (Anon. 1990a). Late in 1989, both organizations adopted complementary PIC procedures, after first refusing to do so. The Commission of the European Community is currently considering a draft directive that would incorporate PIC into European law (T. Casey, Consultant to the Directorate General for the Environment, Commission of the European Community, June 1990, personal communication) and the proposed Pesticide Export Reform Act would do the same in the USA (Anon.1990b).

Although PIC, as operated by FAO and UNEP, will extend to many of the pesticides that have been implicated most often in human poisoning, the degree to which it will actually improve the regulation of such hazards is open to question. Pesticides banned or severely restricted by 10 or more countries will be the first to be covered by the scheme, followed, probably in late 1990, by those so treated by five or more countries. Thereafter, substances labeled "banned or severely restricted" in a decision by any additional country will be included. A working group will determine whether formulations based on WHO class IA (extremely hazardous) compounds should be covered as well (Anon. 1990a). A recent report by the British-based Pesticides Trust (1989) contends, however, that several class 1B (highly hazardous) pesticides that have frequently been involved in poisoning incidents may escape the informed consent provisions.

The scheme hinges on a government's ability to evaluate and act on the notices it receives, and it is precisely this capacity that is deficient in many instances. As well, PIC begins with the decisions industrialized countries have taken to protect health and the environment within their own jurisdictions. Industry has often claimed that a different balance of risks and benefits may lead developing countries to judge acceptable a number of pesticides strictly controlled in industrialized countries (Willis 1986).

The argument works as well, however, in the other direction, e.g., the application methods and worker protection typical in much of the Third World may result in operators being dangerously exposed when using products not subject to any significant restriction in industrialized countries. These determinations can only be made in the light of local conditions, emphasizing once again the need for effective national regulation.

The drafting of the International code of conduct on the distribution and use of pesticides (FAO 1986) is another major initiative that addresses the weakness of pesticide regulation in developing countries. The code calls on the pesticide industry at all levels, as well as exporting nations, international agencies, and public-sector organizations to assume a share of responsibility for ensuring safety in the use of pesticides. The code's provisions are entirely voluntary and there has been considerable controversy over the extent to which they are respected in practice.

Two reports (ELC 1987; Pesticides Trust 1989) prepared for the Pesticides Action Network (an international group of NGOs) allege widespread infringements of the Code, for the most part by industry, in all developing countries investigated. Evidence is presented regarding misleading advertising, inappropriate packaging, poor quality control, and marketing of banned and dangerous products. Governments in developing countries also cite widespread failure by industry, as well as other parties, to abide by the Code's provisions (FAO 1989). In response to these findings, the FAO Conference has asked the Director-General to report by next year on the feasibility of transforming the Code into a convention that governments could make legally

binding within their jurisdictions. Effective enforcement of such legislation, however, would come up against both the vague wording of many of the Code's provisions and, once again, the limited resources available to Third World governments.

The industry perspective

At Ciba-Geigy Ltd, a principal pesticide manufacturer, the FAO Code is accepted apparently without reservation and has been incorporated into the Agriculture Division's quality policy (Anon. 1988a). It is seen as being consistent with the principle of "product stewardship," which entails continual monitoring and periodic internal audits (Anon. 1988b).

The true measure of corporate commitment to these policies and principles is in their application, particularly in cases where there may be conflict with short-term profitability. Ciba-Geigy claims, in several instances, to have voluntarily refrained from marketing products where evidence suggested that they could not be safely used (as required under section 5.2.3 of the Code). For example, chlordimeform (Galecron) was removed from the Latin American market following reports of poisoning; dichlorvos (Nuvan) and phosphamidon (Dimecron) were considered too toxic for agricultural application in the Philippines and Burkina Faso, respectively.

Officers of the company point to a range of initiatives aimed at reducing risks to health and the environment, including improvements in formulations and packaging and an increased emphasis on safety training. Progress is slow but continual, they say, yet little credit is given to these efforts by the company's critics.


Pesticide hazard auditor

Crisis and opportunity

The pesticides industry finds itself under increasing pressure from national authorities, international bodies, and environmental and consumer groups (GIFAP n.d.). Its public image has suffered from a series of widely publicized disasters (Seveso, Bhopal, and the Rhine), as well as from more localized crises, such as the contamination of ground water in Italy's Po valley.

An alternative to confrontation may be found in an historical analogy. By the late 18th century, there had emerged in Britain a large number of common-law corporations engaged in commerce and manufacturing. A highly speculative and unregulated market in corporate stocks developed, leading to several spectacular financial failures. Investors and creditors led the resulting public demand for investigation, which required the services of independent accountants. By the early 19th century, it had become common practice to call upon such skilled outsiders to assist in settling disputes and bankruptcies and, increasingly, to attest to the soundness of enterprises seeking investment or credit. It is to these developments, given legal support in 1844, that the Anglo-American tradition of independent financial auditing can be traced (Anderson 1984).

The recent trend in the United States toward "environmental auditing" appears to have a similar history. A growing number of firms whose activities may give rise to pollution and occupational-health hazards have retained independent environmental auditors to help ensure compliance with regulatory standards and to oversee internal auditing procedures. Once again, the need of companies to maintain investor and creditor confidence and to safeguard their public images appears to have been as crucial in this decision as court-sanctioned or regulatory requirements (Palmisano 1989)

Companies producing and marketing pesticides in developing countries should have their practices, with regard to impact on health and the environment, examined by an independent pesticide-hazards auditor. To the extent that a company's good name or image has value in a competitive environment, a hazard auditor might help create a market-based mechanism for ensuring compliance with accepted standards that would reinforce official regulation. For the system to gain acceptance in the industry, it must embody certain characteristics:

· Independence - the auditor must be seen to have no link, direct or indirect, with the company being examined.

· Authoritativeness - the audit must be based on explicit and recognized standards, as have been codified for financial auditing in the form of generally accepted accounting principles. The FAO Code (FAO 1986) is subscribed to by all parties and might provide one of the bases for defining acceptable corporate practice with respect to pesticide hazards.

· Expertise - the individuals performing the audit must inspire confidence by their demonstrated technical knowledge and mastery of the standards underlying the hazard audit.

· Openness - while respecting proprietary and commercial information whose disclosure might prejudice a company's interests, the detailed and material conclusions of the auditor must be made public if its function is to be fulfilled. Similarly, the company must be prepared to make available to the auditor all relevant documents and records.

· Service - beyond assessing a company's compliance with accepted standards, a financial auditor often provides advice on internal auditing procedures. Similarly, the hazard auditors would make a more useful contribution (and not only to the company) if they suggested changes in, say, a company's environmental- and health-monitoring programs that would enable problems to be identified earlier.

Benefits of a hazard auditor

From the company's perspective, a positive and unqualified attestation from the hazard auditor would provide authoritative confirmation that the company was acting on the high standards to which it laid claim. This would help reassure the increasingly restive society within which agrochemical companies operate and at the same time serve to differentiate the firm from less responsible competitors.

Among developing countries, those whose national regulation is the weakest would stand to benefit most from a hazard audit. The audit would provide an immediate form of control of pesticide hazards, based on the application of broadly accepted principles to the local context in which the products are marketed and used. In no sense, however, should the hazard auditor be seen as substituting for national regulation over the longer term.

A financial audit, in most industrialized countries, is sanctioned by law and backed by administrative and legal measures that ensure compliance with accepted norms. Either form of audit, financial or hazard, relies on market forces and corporate self-interest to raise and maintain an industry's standards. Internal and external audits may lessen the requirements for government enforcement, benefiting developing countries with operational, if constrained, regulatory systems. However, public supervision is still essential to ensure that these mechanisms function efficiently.

For NGOs and their allies on one hand and the pesticides industry and its supporters on the other, the hazard audit may represent one element of a solution to a long-running conflict that, for both, has absorbed considerable energy and resources.


Implementing the concept

Initial steps

A description of the hazard audit (Loevinsohn 1989) was sent to some 150 organizations on all sides of the debate. The concept was further discussed at two scientific conferences and in meetings with some of the major organizations. The response has been generally positive. A meeting of representatives of the main sectors has been suggested to explore in greater detail whether a consensus is attainable and to chart further action.

Participants at such a meeting could discuss its outcome in their respective constituencies and, if general agreement is obtained, working groups could be formed to define "accepted standards," develop procedures for the audit teams, and prepare a draft charter for a Hazard Audit Organization. The output of the working groups would be reviewed at a further meeting involv

ing all major actors. At the same time, the concept would be given wider circulation through print and other media.

Several case studies could be conducted to build confidence and gain experience. These would take place in developing countries whose governments support the aims of the audit.

If the studies were judged successful, the Hazard Audit Organization might be established by a substantial portion of companies in the industry, the major NGOs engaged in campaigning, and other groups representing the public. The support of influential governments in the North and South, key professional associations, and leading international bodies would also be essential.

Structure and function

An autonomous, nonprofit Hazard Audit Organization would have, as its primary task, external hazard audits of companies involved in the manufacture and sale of pesticides in developing countries. Financing would be provided by participating companies, the members, and the industry association, Groupement international des associations rationales de fabricants de produits agrochimiques (GIFAP), as well as firms outside this body. Companies would be charged on a cost basis for each audit, but would also make annual contributions toward the organization's administrative expenses.

General supervision, policy formulation, and the further development of "accepted standards" would be the responsibility of a governing council whose members would be drawn from four broad sectors: the pesticides industry; national regulatory agencies and international bodies (e.g., FAO, WHO, and UNEP); research institutions and professional associations; and consumer, producer, and environmental organizations. Relative proportions remain to be negotiated, but no sector should be allowed to dominate. A technical subcommittee would be responsible for planning and setting terms of reference for individual audits, selecting team members, and reviewing their reports. A small secretariat would also be required. Well-qualified auditors would be drawn from professional associations, international agencies, and national regulatory bodies in the North and South. Retained initially as consultants or on secondment, auditors might eventually be hired by the Hazard Audit Organization.


The FAO Code of Conduct (FAO 1986) may provide a framework of generally accepted principles on which to base the hazard audit, but in many respects the Code's provisions lack specificity. What, for example, constitutes "safe use" or an "unacceptable hazard"? An operational definition of these terms might be based on the practice of well-established regulatory agencies. A residue concentration or exposure level that falls within the range of what

different agencies take to be permissible can be said to be "generally acceptable." The variation in national tolerances to health hazards appears to be greatest with respect to chronic effects which, in statistical terms, are often weak and uncertain. The consensus is generally clearer for acute effects in spite of the preeminent threat to populations in the Third World (Jeyaratnam 1985).


Rather than focus on one company's operations worldwide, the hazard audit might be conducted in one developing country at a time and involve all participating companies that do business there. In this way, it should be possible to cover several countries each year. Given the number of firms and the range of their activities, the auditors would have to rely on sampling techniques, as do financial auditors. The hazards entailed in different aspects of companies' operations might be stratified by severity and risk on the basis of published information, reports from government agencies and NGO groups, and the experience of the technical committee and auditors. Giving greatest weight to the most severe and probable hazards, a sample of practices would then be drawn and assessed in relation to the standards that had been defined.

The audit team would examine company documents and facilities, interview employees, and investigate the distribution of products and the manner in which they were employed. Auditors would also consider information from regulatory bodies, research institutions, and producer, consumer, or environmental groups. Where it is deemed necessary, the team might undertake or commission research that would enable it to reach an informed opinion.

Every effort would be made to ensure the active support of governments of the countries in which the audits are performed. The Hazard Audit Organization and host governments might work out different relations, according to the latter's needs and desires. Following their investigations, auditors would be well placed to report to the government on the effectiveness of national regulation and provide some advice on remedies, possibly focusing on aspects that the government had identified beforehand as problematic. An audit that covered perhaps several months would not, however, provide an opportunity for extensive technical assistance, although the team might make a useful contribution by identifying critical needs for other agencies to follow up.


The auditor's report would express a considered opinion regarding the extent of a company's adherence to accepted standards of conduct. Where deficiencies were noted, the report would detail how practices should be improved to meet standards. This might entail, for example, changes in labeling, packaging, promotional material, educational programs, or restrictions on the availability of the product in that market.

As suggested above, the auditor's report would be made public, except for commercially sensitive or proprietary information. At the company's request, dissemination might be delayed a few months to permit it to bring its practices into line with the recommendations.

It is conceivable that, at some point, the conclusions of the auditor may conflict with the judgement of the national regulatory agency. For example, the former may find that a company should not be marketing a certain product, given pesticide practices in that country, even though the agency might have recently renewed the product's registration. The audit is of a company; it is not intended to limit a government's prerogative to evaluate risks as it sees fit. In the face of an auditor's public report, however, a decision to permit continued use would call for an alternative interpretation of the evidence or a demonstration of overriding benefits. In this way, the hazard auditor might serve to raise the standards of risk assessment and to open it to public scrutiny.


Initial reactions

Several dominant themes emerged among 58 written replies to the hazard auditor proposal (Loevinsohn 1990). Of the opinions expressed, 5 (9%) were negative and 53 (91%) were positive in varying degrees.

Increase support to national regulatory authorities

The most widely voiced view (39% of the 49 detailed responses) was that greater emphasis should be placed on evaluating and assisting regulatory agencies. Some industry respondents felt that the focus on industry alone was unfair and others, from several sectors, thought the auditor's recommendations would more likely be acted on if government was more closely involved in the process.

An audit that puts developing country governments on the same footing as industry makes no sense; the weakness of national regulation is universally acknowledged and is the underlying rationale for initiatives such as the Code of Conduct and the pesticide hazard auditor. Although some assistance to national authorities might take place within or along with the audit, supported from nonindustry sources, there is a danger of overlap with existing or planned programs of agencies, such as FAO, were this function to take on a much greater significance. Closer integration with FAO is indeed a possibility and, in that context, the audit might extend to other sectors addressed by the Code of Conduct.

Ensure the auditor's independence from industry

The second most frequent comment (22% of the 49 detailed responses) concerned the danger of the hazard audit being dominated by the pesticides industry and of its serving to legitimize pesticide use. Several respondents believed that these risks could only be avoided by complete financial autonomy and by excluding industry representatives from the governing and technical bodies of the audit organization.

Reasonable safeguards against domination should be built into the scheme. There would be justification, for example, in excluding company representatives from the technical committee where audits would be planned. Sanctions should be available against companies that, in their advertising or labeling, misconstrue audit results to imply an endorsement of their products. Funding from other sources could be sought to dilute the dependence on industry, but the self-financing character of the proposal is one of its chief attractions. Any attempt to exert undue influence would lead other sectors to withdraw their support from the scheme and lose companies the commercial benefit they derive from an independent audit.

Furthermore, financing by industry is economically rational. An independent audit can legitimately be seen as part of the regulation required to minimize the external costs to which pesticides give rise when they damage human health or the environment. Outside financing of the audit would amount to a subsidy, leading to greater use than if real costs were reflected in market price (Repetto 1985; Brader 1990) and distorting the choice between chemical-based pest control and alternative techniques.

Emphasize the incentives for industry compliance

A number of respondents suggested measures to increase the benefits to a company that agreed to be audited and implement the audit's recommendations. These include proposing that multilateral and bilateral aid organizations make a satisfactory audit report a requirement in their procurement programs. Developing country governments could similarly agree to purchase only from manufacturers who have received such an evaluation from the auditor. Governments might also make the external hazard audit a legal requirement, as is the case for financial audit in many countries.

Measures such as these might indeed usefully increase incentives for compliance and penalize companies who remain outside the scheme to gain, for example, a price advantage. Additional incentives may be particularly important to smaller manufacturers based in Third World countries where public opinion is often poorly informed. Many of these firms produce hazardous pesticides and are often not affiliated with national or international industry associations.

Increase collaboration with FAO and other United Nations agencies

Several respondents wrote that, as the hazard auditor aims at improving compliance with accepted standards, particularly those embodied in the FAO Code of Conduct, a closer relation with FAO should be sought. Some questioned the need for an independent audit organization.

I investigated the possibility of an association with FAO and other United Nations (UN) agencies. Senior FAO officials recognized the value of the auditor, particularly as a possible means to improve compliance with the Code should it become a convention made binding under national law. However, two difficulties were mentioned. FAO, the officials declared, would not accept the financial link with industry that the proposal envisages. The other problem involves openness in reporting, which is crucial to the functioning of the audit mechanism. Because FAO is responsible to its member governments, problems might arise if it were to publish reports that were critical of national administrations.

Similar concerns were voiced by officials of UNEP and WHO. These constraints might be loosened if there were widespread support for the innovation at the highest levels. However, the independence required of an external auditor would be more readily assured within an organization that is itself independent of the major actors.



The hazard auditor concept holds promise for improving the regulation of pesticide hazards in developing countries by creating a new market-based mechanism complementary to, and supportive of, national structures and international programs. It has already attracted widespread, if still provisional, support, but further progress toward implementing the concept will require collaboration among parties grown accustomed to confrontation. The proposal does not assume an congruency of interests among these parties, only that each side believe its interests are served by independent evaluation. Individual actors may conclude that the risks attendant on creating and operating this novel mechanism outweigh its benefits. It is not possible to predict the outcome of what will be a long process of negotiation. The prudent option, for all concerned, is to judge at each step whether the hazard auditor as it is emerging represents an improvement on what currently exists.

Anderson, R.J. 1984. The external audit (2nd ed.). Copp Clark Pitman, Toronto, ON, Canada.

Anonymous. 1988a. FAO code of conduct 17 important messages. In Product quality manual (Appendix 3). Agriculture Division, Ciba-Geigy Ltd, Basel, Switzerland.

1988b. Product quality concept. Agricultural Division, Ciba-Geigy Ltd, Basel, Switzerland.

1990b. White paper proposals. Pesticides News, 8, 7-9.

1990a. London guidelines "prior informed consent." International Register for Potentially Toxic Chemicals Bulletin, 10(1), 4-6.

Brader, L. 1990. Integrated pest management: successes and constraints. Paper presented at the FAO/UNEP workshop on integrated pest management, 12-15 June 1990, Kishniev, Moldavia, USSR.8 pp.

ELC (Environment Liaison Centre). 1987. Monitoring and reporting the implementation of the international code of conduct on the use and distribution of pesticides (the FAO Code). ELC, Nairobi, Kenya.

FAO (Food and Agriculture Organization of the United Nations). 1986. International code of conduct on the distribution and use of pesticides. FAO, Rome, Italy. 31 pp.

1989. International code of conduct on the distribution and use of pesticides: analysis of responses to the questionnaire by governments. FAO, Rome, Italy. AGP: GC/89/BP.1.

GIFAP (Groupement international des associations rationales de fabricants de produits agrochimiques). n.d. The role of GIFAP. GIFAP, Brussels, Belgium.

Jeyaratnam, J. 1985. Acute pesticide poisoning: a Third World problem. World Health Forum, 6, 39.

Loevinsohn, M. 1989. The hazard auditor and improved pesticide regulation in the Third World. International Development Research Centre, Ottawa, Canada. Consultant's report, 10 pp.

Loevinsohn, M. 1990. Evaluation of responses to the pesticide hazard auditor proposal. International Development Research Centre, Ottawa, Canada. Consultant's report, 9 pp.

Pallemaerts, M. 1988. Developments in international pesticide regulation. Environmental Policy and Law, 18, 62-68.

Palmisano, J. 1989. Environmental auditing: past, present and future. Environmental Auditor, 1, 7-20.

Pesticides Trust. 1989. The FAO code: missing ingredients. Pesticides Trust, London, UK.

Repetto, R. 1985. Paying the price: pesticide subsidies in developing countries. World Resources Institute, Washington, DC, USA.18 pp.

Willis, G.A.1986. Pesticide residue problems in developing countries in Asia: some contributions of industry. Paper presented at the 2nd session of the working group on pesticide residue problems in Asia, 2-5 April 1986, Chiang Mai, Thailand.



Safer pesticide-application equipment for small-scale farmers in developing Asian countries

Md. Jusoh Mamat, A.N. Anas, and S.H. Sarif Hashim

Many tropical Asian countries produce lever-operated knapsack sprayers locally. The quality of the sprayers ranges from poor to moderate, yet large numbers of these sprayers are bought and used in the region. Some of their unsafe features are: sharp edges on the tank body, poor positioning of the pump body and linkage system, leak-prone connections, and poor quality materials. Factors leading to the continued production of inferior sprayers include the lack of incentives for local manufacturers to produce good quality sprayers, farmers' poor appreciation of quality and safety features, and the lack of national minimum product standards. Overcoming these problems requires the establishment of research programs in national agricultural research institutes to design safer and more efficient knapsack sprayers that can be produced locally, upgrading of farmers' training programs to improve their appreciation of good quality sprayers and safer spraying practices, and establishing minimum standards for knapsack sprayers.

Unlike North America or Europe, the greatest concern relating to the escalating use of pesticides in developing Asian countries is the personal contamination of those applying the chemicals. Ironically, in cases of occupational poisoning or failed pest control, farmers put the blame on the pesticide. Seldom, if ever, is the method or apparatus used for application blamed. Yet, the design and condition of the equipment owned and used by farmers can contribute to the safe and efficient application of pesticides (Anas et al. 1987; Jusoh Mamat and Anas 1988).

In many developing Asian countries, the lever-operated knapsack sprayer is still the most commonly used pesticide applicator among small-scale farmers (Adam 1976; Fraser and Burrill 1979; Heinrichs et al.1979; Litsinger et al.1980; Prasadja and Ruhendi 1980; Lim et al. 1983; Jusoh Mamat et al. 1985). Major agricultural countries in South Asia produce lever-operated knapsack sprayers locally. However, the quality of these sprayers, in terms of efficiency and safety, ranges from very poor to moderate. In Malaysia, for example, the lever-operated knapsack sprayers (96% locally manufactured) used by rice farmers were not well designed and did not satisfy the minimum product standards set by the World Health Organization (Anas et al.1987).


Unsafe features of locally produced sprayers

In using a lever-operated knapsack sprayer, the operator is exposed to two kinds of hazard: direct physical harm such as wounds and bruises caused by the equipment and contamination by the pesticides being applied. Hazardous features of knapsack sprayers have been discussed in detail (Fisher and Deutsch 1985; Anas et al.1987; Thornhill 1987).

Sharp edges on the tank body - Sharp edges are found at the bottom of the tank body, the body skirting, and the threaded knob where the pump lever is attached. These sharp edges can rub against and injure the operator's back or buttock area (if contaminated by pesticides, wounds can be even more serious). Sharp protrusions also cause injuries if the operator falls.

Narrow straps made of unsuitable material - Knapsack sprayer straps are often too narrow (less than 5 cm) and made of hard, coarse materials. When a full load of 15-20 L of pesticide is carried, a hard narrow strap will dig into the shoulder muscles causing skin bruises. Because of the constant movement of the tank on the operator's back, bruises can appear after only two or three rounds of spraying with a full load. Straps may snap under a heavy load and cause the tank to fall, wounding the operator. If the straps give way early in the work period, farmers may replace them with even less appropriate materials, such as plastic rope, which is very uncomfortable, or soft cloth rope, which can absorb spilled or splashed pesticide and become a persistent source of contamination.

Tank weight and balance - Most locally produced sprayers are made of brass and are heavy even when empty (over 6.5 kg). Carrying such a heavy sprayer often taxes the energy of slightly built Asian workers. This is especially true for women workers in the rubber and oil-palm plantations of Indonesia, Malaysia, and Thailand. The tank's centre of gravity, often located in the upper half, will cause the operator to be unbalanced and fall easily when walking on difficult terrain, such as wet and muddy paddy fields.

Small tank port and shallow basket strainer - A tank port smaller than 12 cm and basket strainers less than 5 cm deep can create an air lock. As a result, diluted pesticide can overflow and splash onto the operator's hands or legs as it is poured into the tank, particularly if this is done in haste.

Leak-prone spraying components - Leaky sprayer components are a major cause of pesticide contamination among small-scale farmers. Pump cylinders, air chambers, and cut-off valves may leak because of worn-out parts, such as the piston head, ball bearings, and washers, or because they are not greased properly. Leaks at hose connection points are mainly due to the use of crimp ferrule; continual use and flexing of the hose causes cracking.

Positioning of components - If the pump body and the air chamber are not firmly secured to the tank or are awkwardly positioned at the side of the tank body, misalignment of the piston head in the cylinder and leakage at weld points in the air chamber can easily result from knocks and falls during spraying operations. Misalignment was found in 64% of the knapsack sprayers owned and used by farmers in the rice-bowl area of Malaysia (Anas et al. 1987).

Large mesh and absence of strainers and filters - A common cause of contamination during spraying operations is blockage of the nozzle opening by solid foreign matter when strainers and filter mesh are too large, i.e., larger than 1 mm for strainers and 0.5 mm for filters, or completely absent. In Southeast Asian countries, water used to dilute the pesticides comes from irrigation canals, small streams, or rain-water storage tanks, which contain much foreign debris. As a result, nozzle blockage is common. Farmers often attempt to free blockages with sharp hard objects, such as pins or steel wire (possibly damaging the nozzle opening) or by blowing into the nozzle, contaminating both their hands and mouths. In a survey in Malaysia's rice-bowl area, 89% of 193 knapsack sprayers had strainers with a mesh size larger than 1-mm and none had filters at the cut-off valve and the nozzle (Anas et al.1987).

Absence of an agitator - Not all lever-operated knapsack sprayers produced locally are fitted with an agitator. When wettable powder (WP) formulations are used, agitators keep the powder suspended in the solution, thus preventing it from settling or becoming aggregated and blocking the nozzle opening during spraying.

Length and design of the spray lance - All locally produced knapsack sprayers currently available are equipped with a 50-cm spray lance. This is too short to prevent drifting spray droplets from contaminating the operator. A 1-m lance, curved slightly at the front end, could make directing the spray much easier and decrease contamination due to dirt (Jusoh Mamat and Anas 1988).

Poor quality brass used for tank body - According to Anas et al. (1987), 50-58% of sampled sprayers had body indentations and evidence of corrosion. These were caused by knocks and falls, testifying to poor user habits as well as the poor quality brass sheets used by manufacturers. In Malaysia, local manufacturers produce sprayers with a price range from 60 to 120 MYR (2.6875 Malaysian ringitts (MYR) = 1 US dollar (USD) in 1991) according to brass content or thickness of the brass sheet used for the tank body. Sprayers in the lower price categories tend to be unsafe because they are easily dented and corroded and are thus too prone to leaking.


Improving local production of sprayers

The continued production of poor-quality sprayers in developing Asian countries cannot be blamed solely on the profit motive of the manufacturers. The lack of incentives for local manufacturers to produce good-quality knapsack sprayers, the attitudes of farmers, and the lack of government standards are equally responsible.

There are few economic incentives for local manufacturers to improve quality. In Malaysia, for example, local manufacturers, who produce about 300 000 sprayers annually, have virtually no foreign competition. Even though their sprayers are inferior to imported models in quality and safety, the overwhelming majority of farmers choose to buy cheaper, locally produced sprayers. Any improvement would merely add to the cost of production without any market advantage. Moreover, improvements of certain components of the sprayer system can involve producing an entirely new model. This may require changes in manufacturing equipment and possibly retraining workers, which may result in an initial decrease in the efficiency of production.

On the demand side, farmers' poor appreciation of good quality and safety features also discourages improvement. In Malaysia, farmers and plantation workers pay little attention to the proper use of pesticides especially with regards to safety (Zain 1977; Zam 1980; Basri 1981; Heong 1982; Normiya 1982; Ooi et al. 1983; Heong et al. 1985, 1987; Hussein et al. 1985; Anas et al. 1987; Jusoh Mamat et al. 1987; Anon. 1990). Consequently, they fail to recognize or avoid sprayers with poor safety features. As long as the sprayers function cheaply, they will use them without concern for their own health and safety. Such attitudes cause complacency among local manufacturers about the quality of their sprayers.

The lack of national product standards for knapsack sprayers also contributes to the continued production of unsafe sprayers. In most Asian developing countries, monitoring quality or standardized testing of locally produced machines and equipment is left almost entirely to manufacturers. Generally, governments have made little effort to establish standards or enforce them.

In Malaysia, a positive step has been taken with the establishment of the Standard and Industrial Research Institute (SIRIM). Its role is to produce standards and approve products that conform. However, SIRIM has no authority to enforce these regulations, and even the product approval scheme has not yet been implemented for knapsack sprayers. Without such a scheme, local manufacturers have no incentive to alter their products to meet minimum safety standards.


Improving safety in design and use

Because manufacturers are not likely to produce better models, it is suggested that local government research institutions should accept this responsibility. Research institutions should develop a new sprayer design and provide blueprints, free of charge, to local manufacturers. Besides optimizing ergonomic and safety features, the new design must also take into account the production capability of the manufacturers and the availability of materials, which together determine the unit price of the product. Good, imported knapsack sprayers are available, but at a price double or triple that of locally manufactured sprayers. The challenge is to produce a new knapsack sprayer as good as the imported ones, if not better, at the same price as current local models.

Good equipment will not improve pesticide-application technology among small-scale farmers without training in its proper use and maintenance. A critical step will be to encourage the establishment of sprayers' clinics within existing farmers' cooperative centres. These clinics, besides selling approved knapsack sprayers, storing spare parts, and providing repair services, must also provide advice and training for farmers in the safe use of the sprayers.

Regulatory measures, such as setting local minimum standards for knapsack sprayers and mandatory use of protective clothing during spraying operations must also be considered. These measures are useless without effective enforcement.



Regardless of future improvements in spraying techniques and technology, there is much room for improvement in current conventional practices, especially in tropical developing countries. Although the inefficiencies of conventional foliar spraying are well recognized (Matthews 1983; Hislop 1988; Zeren and Moser 1988), its versatility makes it attractive to small-scale farmers of the Asian region. Any significant improvement in the current practice, be it in the equipment itself, the technique of spraying, safety attire, or the attitude and perception of the farmers, will, therefore, produce long-lasting benefits. The advent of a revolutionary new application technique is less likely to replace conventional spraying than to augment it in this part of the world.

Adam, A.V. 1976. The importance of pesticides in developing countries. In Gunn, P.L.; Steven, J.G.R., ea., Pesticides and human welfare. Oxford University Press, Oxford, UK. Pp. 115-130.

Anas, N.; Jusoh Mamat, Md .; Heong, K.L.; Ho, N. K. 1 987. A field observation of lever operated knapsack sprayers owned by the rice farmers in the Muda Irrigation Scheme. In Proceedings of the international conference on pesticides in tropical agriculture, 23-25 September 1987, Kuala Lumpur, Malaysia. Malaysian Plant Protection Society, Kuala Lumpur, Malaysia.

Anonymous.1990. Laporan ringkas bancian kegunaan racun makhluk perosak di kalangan peserta den pekerja FELCRA Seberang Perak. Jabatan Pertanian, Kuala Lumpur, Malaysia.

Basri, M.W.1981. Study on the use of and hazards posed by certain insecticides on tobacco and vegetables in Peninsular Malaysia. Crop Protection Branch, Department of Agriculture, Kuala Lumpur, Malaysia.

Fisher, H.H.; Deutsch, A.E. 1985. Lever-operated knapsack sprayers: a practical scrutiny and assessment of features, components and operation - implication for purchasers, users and manufacturers. International Plant Protection Center, Oregon State University, Corvallis, OR, USA.

Fraser, F.; Burrill, L.C. 1979. Knapsack sprayers: use, maintenance, accessories. International Plant Protection Center, Oregon State University, Corvallis, OR, USA.

Heinrichs, E.A.; Saxena, R.C.; Chellia, S.1979. Development and implementation of insect pest management systems for rice in tropical Asia. In Sensible use of pesticides. Asian and Pacific Council, Taiwan. FFTC Book Series, 14, 208-247.

Heong, K.L.1982. Pest control practices of rice farmers in Tanjung Karang, Malaysia. In Proceedings of the 3rd international meeting on the perception and management of pests and pesticides, Nairobi, Kenya, 21-25 June 1982. Insect Sci. Appl., 5(3), 221-226.

Heong, K.L.; Ho, N.K.; Jegatheesan, S. 1985. The perception and management of pests among rice farmers in the Muda Irrigation Scheme, Malaysia. Malaysian Agricultural Research Institute, Kuala Lumpur, Malaysia. Report 105.

Heong, K.L.; Jusoh Mamat, Md.; Ho, N.K.; Anas, A.N.1987. Sprayer usage among rice farmers in the Muda area, Malaysia. In Proceedings of the international conference on pesticides in tropical agriculture, 23-25 September 1987, Kuala Lumpur, Malaysia. Malaysian Plant Protection Society, Kuala Lumpur, Malaysia.

Hislop, E.C. 1988. Electrostatic ground-rig spraying: an overview. Weed Technology,2,94-105.

Hussein, M.Y.; Jusoh Mamat, Md.; Azmi, M.A.; Monyvellu, S. 1985. Transfer of pesticide application technology to small farmers: practical aspects on approaches and constraints. Paper presented at the workshop and course on pesticide application technology, 21 -26 October, Universiti Pertanian Malaysia, Serdang, Selangor, Malaysia. Malaysian Plant Protection Society, Kuala Lumpur, Malaysia.

Lim, G.S.; Hussein, M. Y.; Ooi , A.C.P.; Zain, M.B.A.R. 1983 . Pesticide application technology in annual crops in Malaysia. In Lim, G.S.; Ramasamy, S., ea., Pesticide application technology. Malaysian Plant Protection Society, Kuala Lumpur, Malaysia. Pp.13-41.

Litsinger, J.A.; Price, E.C.; Herrara, R.T.1980. Small farmer pest control practices for rainfed rice, corn and grain legumes in three Philippines provinces. Philippine Entomologist, 4, 65-86.

Matthews, G.A.1983. Problems and trends in pesticide application technology. In Lim, G.S.; Ramasamy, S., ea., Pesticide application technology. Malaysian Plant Protection Society, Kuala Lumpur, Malaysia. Pp.163-170.

Jusoh Mamat, Md.; Anas, A.N. 1988. Herbicide application technology for irrigated rice in Malaysia. In Lam, Y.M.; Cheong, A.W.; Azmi, M., ea., Proceedings of the national seminar and workshop on rice field weed management, Penang. Malaysian Agricultural Research and Development Institute, Kuala Lumpur, Malaysia. Pp. 221-229.

Jusoh Mamat, Md.; Heong, K.L.; Rahim, M.1985. Principles and methodology of pesticide application techniques. In Proceedings of the workshop and course on pesticide application technology, 21-26 October 1985, Universiti Pertanian Malaysia, Serdang, Selangor, Malaysia. Malaysian Plant Protection Society, Kuala Lumpur, Malaysia.

Jusoh Mamat, Md.; Anas, A.N.; Heong, K.L.; Chan, C.W.; Nik Mohd Nor, N.S.; Ho, N.K.; Zaiton, S.; Fauzi, A. 1987. Features of lever operated knapsack sprayer considered important by Muda rice farmers in deciding which sprayer to buy. In Proceedings of the international conference on pesticides in tropical agriculture,23-25 September 1987, Kuala Lumpur, Malaysia. Malaysian Plant Protection Society, Kuala Lumpur, Malaysia.

Normiya, R.1982. Problems in transfer, delivery and acceptance of rice technology. Malaysian Agricultural Research and Development Institute, Kuala Lumpur, Malaysia. Rural Sociology Bulletin 12.

Ooi, A.C.P.; Heong, K.L.; Lim, B.K.; Mazlan, S. 1983. Adoption of pesticide application technology by small-scale farmers in peninsular Malaysia. In Lim, G.S.; Ramasamy, S., ea., Pesticide application technology. Malaysian Plant Protection Society, Kuala Lumpur, Malaysia. Pp.148-158.

Prasadja, I.; Ruhendi. 1980. Farmers' existing technology end pest control practices for food crops at three locations in Jogjakarta Province. Agency for Agricultural Research and Development, Central Research Institute for Agriculture, Bogor, Indonesia. 59 pp.

Thornhill, E.W. 1987. Lever-operated sprayer selection. In Proceedings of the international conference on pesticides in tropical agriculture, 23-25 September 1987, Kuala Lumpur, Malaysia. Malaysian Plant Protection Society, Kuala Lumpur, Malaysia.

Zain, M.B.A.R. 1977. Survey on the use of pesticides on tobacco in peninsular Malaysia. Crop Protection Branch, Department of Agriculture, Kuala Lumpur, Malaysia. Report 1.

Zam, A.K. 1980. Bancian pengurusan den kawalan serangga perosak padi di rancangan perairan tanjung karang den krian. Laporan Cawangan Pemeliharaan Tanaman, Jabatan Pertanian, Kuala Lumpur, Malaysia.

Zeren, Y.; Moser, E. 1988. Effects of electrostatic charging and vertical air current on deposition of pesticide on cotton plant canopy. Agricultural Mechanization in Asia and Latin America, 19(1), 55-60.



Field evaluation of protective equipment for pesticide appliers in a tropical climate

G. Chester, A.V. Adam, A. Inkmann-Koch, M.H. Litchfield, R. Sabapathy, and C.P. Tuiman

To provide practical advice and information on personal protection used during pesticide application in tropical climates, a field study was conducted in Thailand. Items assessed included protective garments worn by workers mixing and loading the organophosphorous insecticide methamidophos and by sprayers applying the diluted formulation for several hours per day to a cotton crop using knapsack sprayers. Garments made of various materials were assessed for their acceptability to the workers, their comfort and durability, and the degree of protection they offered. Tyuek garments proved to be uncomfortable under tropical conditions. Kleenguard and cotton garments were acceptable to the workers; cotton proved more durable and comfortable over 6 days of use. Nitrile gloves and face shields worn by the mixer-loaders were also found to be suitable under field conditions. The effective use of protective equipment must go hand in hand with safe handling practices and good personal hygiene.

The heat and humidity in tropical countries often make it difficult for farmers to wear recommended protective clothing when handling and applying pesticides. They either suffer excessive heat discomfort when wearing such equipment or remove it to work more comfortably, thus risking increased exposure. To improve this situation, the Groupement international des associations rationales de fabricants de produits agrochimiques (GIFAP) and the Food and Agriculture Organization (FAO) undertook a field evaluation of personal protective equipment in Thailand (Working Group 1989). This paper describes the study and its findings.


Materials and methods

Assessment of protective equipment

The Working Group examined materials of potential use for protective garments, using criteria such as protection against pesticides, durability, heat-exchange properties, comfort, availability, and cost. Three materials were identified for field evaluation: polypropylene (Kleenguard EP), polyethylene (Tyvek S1422), and cotton. Two-piece protective garments were made from these materials. The upper garment was a double-apron design, the lower garment a pair of trousers.

A review of materials for protective gloves, using criteria such as protection against a range of pesticide formulations, cost, and availability, revealed that nitrile rubber was one of the most appropriate. Gauntlet-style gloves were made from this material and used by the mixer-loaders in the study.

A face shield of simple design was constructed with a transparent visor mounted with elastic so that the transparent vertical section rested about 2.5 cm from the face. These were worn by the mixer-loaders in the study.

Study site and crop

The study was conducted on five smallholder farms in the vicinity of Patananikom, Lopburi District, Thailand, in September 1988. The cotton crop under cultivation at the time was up to 1.5 m tall. Weather conditions, including temperature, humidity, and wind speed, were monitored during the study.

Pesticide and application equipment

The organophosphorous insecticide, methamidophos, was used in the study. The formulation, Tamaron 600 SL, contained methamidophos at 600 g/L; 30 mL was diluted to 17 L in each knapsack tank to give a methamidophos concentration in the spray solution of about 1.06 g/L. The application rate was I L/ha. The spray equipment consisted of semiautomatic knapsack sprayers of stainless-steel construction with a capacity of 17 L.


Twenty pesticide workers were recruited locally and divided into two groups of 10. Each group consisted of two mixer-loaders and eight spray operators. All workers wore standard cotton T-shirts and shorts under the protective garments.

Study design

The study was carried out in two stages. A prestudy was conducted to evaluate the comfort, acceptability, and effectiveness of protective garments made of the nonwoven materials, Kleenguard and Tyvek, to select one for use in the main study. In the main study, we evaluated garments made of the selected material and cotton.

On day 1 of the prestudy, one group of workers wore Kleenguard protective garments, the other group wore Tyvek garments. All workers wore cotton sampling garments beneath their protective clothing to assess its permeability to the insecticide during work. Upon completion of the day's operations, the protective and sampling garments were removed carefully from each operator. Each garment was turned inside out and wrapped in aluminium foil. Each foil package was placed in a polyethylene bag labeled with the worker's number, day, type of garment, and date. All garments were kept in refrigerated storage before analysis.

On day 2, workers in both groups wore new Kleenguard protective garments without sampling garments underneath to assess comfort. On day 3, this procedure was repeated using Tyvek protective garments.

On days 1 and 6 of the main study, one group of workers wore Kleenguard protective garments and the other group wore cotton. All workers wore cotton sampling garments under their protective clothing to assess their protection against insecticide. Upon completion of the day's work, the protective and sampling garments were removed and stored as described for the prestudy.

On day 2 (main study), new protective garments were issued to the two groups of workers (Kleenguard for one group, cotton for the other) for daily assessment of comfort and durability. Each worker wore the same garment until either the study team agreed that it could no longer be worn because of deterioration or until the end of the study.

Regular assessments were also made of the performance and condition of protective gloves and face shields. The opinions of the mixer-loaders were obtained through the use of a questionnaire.

Daily operations

To standardize the daily work as much as possible, the two mixer-loaders in, each group prepared Tamaron spray solution for the spray operators in their respective group. Each mixer-loader handled the same quantity of the insecticide formulation.

The spray operators applied the same volume of insecticide (seven knapsack tanks) each day. Operators worked sufficiently far apart to minimize the possibility of cross-contamination due to drift from other sprayers.

At the end of each day's operations, the protective garments were washed with detergent and water using a standard procedure and dried overnight.

Assessment of protection

Permeation was derived by measuring the amount of methamidophos found on the protective and sampling garments. Methamidophos was extracted from the garments using methanol for cotton and ethyl acetate for Kleenguard and Tyvek. The extracts were subjected to analysis by high-performance liquid chromatography (HPLC) using an ultraviolet detector for the determination of methamidophos. The results were corrected for losses of methamidophos through exposure to light, storage, and transit.

The permeation value, expressed as a percentage, was the amount of methamidophos found on a sampling garment, divided by the total amount found on the protective and sampling garments. Permeation data for each group were compared using Student's t-test after exclusion of outlying values (Tukey 1977).

Evaluation of comfort, thermal comfort, and durability was achieved through the use of a detailed questionnaire and from observations made by the study team.

Health surveillance

The workers were given a medical examination I day before the start of the prestudy and were kept under medical surveillance during both phases of the study. Blood samples were collected at regular intervals to test for cholinesterase activity. Measurement was carried out in situ using the tintometer method.



Total working time on the I 1st day of the prestudy was about 4 h. As the study proceeded, work rate increased. On the last day of the main study, working time had decreased to about 2.5 h. Because spray equipment broke down frequently, one person was employed full-time to look after the equipment and ensure continuity of spraying.

Over the 9-day study period, shade temperatures during work ranged from 26 to 33°C. Relative humidity, measured on 3 days, was 52-81%. Wind speed was low, with gusts of up to I m/s recorded occasionally.

During the prestudy, all 20 workers said that they felt comfortable when working in the Kleenguard protective garments (Table 1). By contrast, 11 of the 20 workers said that the Tyvek garments were uncomfortable and this was reinforced by responses to questions about problems during spraying; 10 of the 16 sprayers complained that the material stuck to the skin and caused the garments to wrinkle. Most of the workers with Tyvek garments felt hotter as a result of wearing them. On the basis of questionnaire results and observations made by the study team,it was decided that Kleenguard garments would be used in the main study.

Table 1. Worker assessment of two types of protective garment. Kleenguard and Tyvek.

Workers wearing both Kleenguard and cotton stated that the garments were comfortable to wear every day of the main study (Table 2). Most workers in both groups experienced no problems. When problems did occur, they were due to the individual fit of the garment rather than general physical discomfort. Although most workers felt somewhat hotter when wearing the garments, this was not sufficient to cause undue stress or loss of working efficiency. By day 5, fewer workers wearing cotton garments felt hotter.

Table 2. Comparison of Kleenguard (K) and cotton (C) protective garments regarding comfort and acceptability (numbers represent positive answers).

Most of the cotton protective garments were unaffected by wear over the first 3 days (Table 3). From day 4 on, an increasing number of cotton garments showed signs of slight chafing, mainly around the shoulders, lower back, around the buttock area, and knees. There was no evidence of tearing or splitting.

Table 3. Number of garments chafing or showing wear over 6 days of use.

Kleenguard garments showed evidence of slight chafing on day 1. By day 3, the chafing was pronounced and tearing and splitting of the garments was also noted. This was more evident in the garments worn by workers who carried the knapsack sprayers, which rubbed against their shoulders and lower back. The damage was progressive; by day 6, all Kleenguard garments were affected, some producing severe chafing. The chafing was mostly on the shoulders and lower back.

The mean permeation value for the Kleenguard garments on day I of the main study (41.8%) was similar to that found during the prestudy (37.7%) (Table 4). The mean permeation value for the cotton garments (18%) was significantly lower than that of the Kleenguard garments (p < 0.01). After 6 days of wear, including daily washing, the permeation values of the two sets of protective garments had not increased and, in fact, were somewhat lower than on day 1. Permeation of the cotton garments was still significantly lower than that for the Kleenguard garments on day 6 (p < 0.01).

Table 4. Permeation (%) of protective clothing by insecticide.

The permeation rate varied widely within each group of workers. This is consistent with findings from other worker-exposure studies, which show that the amount of contamination varies with individual workers (Wolfe et al. 1972).

Three of the four mixer-loaders found the protective gloves uncomfortable to wear. However, all four believed that their hands were being protected. Some experienced difficulty in gripping equipment when wearing the gloves, and two pairs of gloves had to be replaced because of abrasion or splitting. Workers found the face shields comfortable to wear; no misting of the visors occurred. The mixer-loaders had no difficulty in keeping the face shields on during the work periods. There were some complaints of reflection from the white garments onto the face shields.

Health surveillance

No clinical signs or symptoms of intoxication by methamidophos were reported or observed in anyone involved in the study. Although there were slight variations in cholinesterase activity in blood samples from workers during the study, all values were within normal ranges.



Cotton protective garments of the design used in this study were comfortable to wear in hot and humid climates. They showed little deterioration over the period of use and provided as much or more protection from the organophosphorous insecticide than the synthetic materials.

When the benefits of availability and general low cost are taken into account, cotton has major advantages as a material for protective garments required by pesticide workers in tropical climates. Further investigation into the protective properties of cotton, i.e., variations in weight, thickness, and weave, would be worthwhile.

The design of the protective garments was intended to provide some degree of flexibility under field conditions. For example, workers are able to wear the top and trousers separately or together over their normal work clothing depending on the kind of work they are engaged in. Although this design is recommended by the Working Group, it may not be acceptable esthetically in all cultures and other designs may be more appropriate for some types of pesticide application. However, the general principle of the design proved to be successful in this study.

Nitrile rubber gloves, although somewhat uncomfortable, can be worn by mixer-loaders in hot and humid conditions and are, therefore, suitable for handling pesticide concentrates under these conditions. The simple face shield used in this study is also recommended.

The opportunity was taken during the study to seek the workers' opinions about acquiring and wearing gloves and face shields in normal practice. Commitment to wearing these protective items varied, even if they were made readily available. This type of response is common (Jeyaratnam 1982) and underlines the need to change attitudes through training and education programs.

Many pesticides can be handled and applied safely without the need for protective clothing; workers simply wear lightweight work clothing covering most of the body. However, where protective clothing is necessary for use with more hazardous pesticides, cotton garments of the type described in this paper are recommended for use in tropical conditions.

Finally, protective equipment cannot be regarded as the only answer for all aspects of personal protection. This equipment does not necessarily reduce pesticide contamination unless it is used and maintained properly. Its use must go hand in hand with other important protective measures: principally the avoidance of contamination, good personal hygiene, and the correct use and maintenance of safe application equipment.

Acknowledgments - The dedicated technical support provided by Mr Sant Kongsomboon (Bayer Thai Ltd), Dr Apichai Daorai (ICI Asiatic Ltd), and Dr Sujin Chantarasa-ard (Shell Thailand Ltd) is gratefully acknowledged. We also shank the Thai Pesticide Association for their encouragement and support of this project.

Jeyaratnam, J. 1982. Health hazards awareness of pesticide applicators about pesticides. In van Heemstra, E.A.H.; Tordoir, W.F., ea., Education and safe handling in pesticide application. Elsevier Scientific Publishing, Amsterdam, Netherlands. Pp. 23-30.

Tukey, J.W. 1977. Exploratory directory data analysis. Addison-Wesley, Reading, MA, USA.

Wolfe, H.R.; Armstrong, J.E.; Staiff, D.C.; Comer, S.W.1972. Exposure of spraymen to pesticides. Archives of Environmental Health, 25,29-33.

Working Group(Working Group on Protective Clothing for Hot Climates). 1989. Field evaluation of protective clothing materials in a tropical climate. Groupement international des associations rationales de fabricants de produits agrochimiques and Food and Agriculture Organization, Bangkok, Thailand. 4 pp.