Safe Handling of Agricultural

Pesticides in Minnesota:
Results of a County-wide
Educational Intervention
J effrey H. Mandel, M.D, M.?H, M! Peter Catr, M.PH., Term Hillmer, M.I?H.,
Pamela R. Leonard, M.D, M.?H., Julia U Halberg, M.D., M.PH.,
Wayne 1: Sanderson, Ph.D, C.I.H., and Jack S. Mandel, Ph.D, M I H .
ABSTRACT: Safe pesticide handling in a group of Minnesota farmers, aged 40 years and
om, WIS inj7uenced through the use of a multifaceted, countyunde educational i n t m t w n .
Tw, i n t m t i o n counties and tuu control counties m e i ml ued in this effort. The i n t m -
twn consisted of mailed pesticide injinmation to farm households, educational programs on
pesticides f i r county physicians, elemoltary school training modules on pesticides and the use
of safe pesticide handling displays in key business areas by agricultural extension agents. Fiw
hundred eight farmers m e identipd as pesticide users (186 in the i n t m t w n counties and
322 in the control counties). The use of gloues and other protectiw clothing while handling
pesticides increased in the i n t mt i on group. I r n p r m t UMS greater in those who had used
protectiw equipment the least b e p e the intermtion. From these results, it appears that a
broad-based educational i nt enmt wn might haw a modest impact in hau farmers protect them-
selws when using pesticides.
he prevention of pesticide exposure among
farmers has the obvious benefit of avoiding
health problems from acute or chronic in-
toxication. Even though controversy exists
T as to the relationship of pesticide exposure
in the development of certain cancers such as lympho-
ma, leukemia, myeloma and others (Blair and Zahm,
1991), proper pesticide handling is obviously prudent.
Unfortunately, safe pesticide handling by farmers in
general has been shown to be less than optimum
(Avory and Coggon, 1994; Elkind and Coty-Salter,
1994; Hayes, 1980; Rucker, 1994; Winstead, 1993). This
appears to be true despite the fact that farmers under-
stand both the potential harmful effects of pesticides
and the benefits of protective equipment (Elkind and
Cody-Salter, 1994; Mandel, et al., 1996; Rucker, 1994).
Previous assessments have shown that with increas-
ing demands on their time, farmers are less likely to
practice safe handling techniques; that a farmeis atti-
tude toward safety in other situations is related to safe
pesticide use (Avory and Coggon, 1994; Ogdvie, 1990);
that labeling plays a role in proper pesticide use
(Hayes, 1980); that administrative control alone might
not effectively enhance safe pesticide use (Runyan,
The 1 0 u d of Rural H d t h 148 Vd. 16, No. 2
1994); that educational programs designed to affect
farmers' knowledge are difficult to implement and
measure (Rucker, 1994); that farmers self-report the
need for further education in pesticide protection
(Stone, et al., 1994); and that time urgency plays a role
in preventing farmers from using safe techniques (Elk-
ind and Cody-salter, 1994; Walker, 1988; Weigel, 1980).
Prior attempts to improve safe handling practices in
the United States have focused primarily on legisla-
tion, machinery engineering, protective equipment and
health provider education (Cordes and Rea, 1991). Im-
provements in farmers' other preventive health behav-
iors through education have been demonstrated (Mul-
lan, et al., 1996; Reding, et al., 1996; Goldsmith and
Sisneros, 1996).
This intervention study was initiated to gain further
insights into enhancing safe pesticide handling among
Minnesota farmers through an educational approad.
This effort was part of a countywide intervention pro-
gram designed not only to enhance safe pesticide han-
dling but also to increase breast cancer screening and
to increase awareness of skin cancer prevention. This
report provides the results of the safe pesticide han-
dling educational intervention.
Methods
The study was a community intervention trial con-
ducted in 1992. Four counties were selected on the ba-
sis of similar demographic and agricultural produc-
tion characteristics. Of these, two nonadjacent counties
were randomly selected to receive interventions. Uni-
versity of Minnesota Institutional Review Board ap
prod was obtained prior to the intervention effort.
The intervention included both a physician compo-
nent and a community component. The physician in-
tervention consisted of a seminar that was held in
both counties and provided information on exposure
to pesticides. The seminars were cosponsored by the
county medical societies and earned 2.0 category I
American Medical Association continuing education
credits. County medical society lists were used as the
basis for inviting physicians to attend the seminars.
More than 80 percent of the 72 primary care physi-
cians on these lists attended the seminars.
The community interventions consisted of an ele-
mentary school curriculum, educational materials
mailed to all farm households in the intervention
counties and panel displays utilized by County Exten-
sion Service staff. The school curriculum was designed
to educate children about the presence of pesticides in
and around the rural home, and it suggested ways to
prevent exposure to pesticides. Separate curricula
were developed for kindergarten through second
grade and for thitd grade through sixth grade Materi-
als were developed to serve as a free-standing school
unit intended to be presented as a special project.
Ea& public and private elementary school in the two
intervention counties was provided the curricula,
whi h were implemented with help from Minnesota
Extension Service representatives.
vention included an educational brochure with infor-
mation on safe pesticide handling practices, an emer-
gency action card for managing acute pesticide expo-
sures, poison center information and a washing ma-
chine magnet with instructions for laundering
pesticide-contaminated clothing. Materials were
mailed to each farm household in the intervention
counties; addresses were compiled from the list of
farm operators maintained by the Minnesota Agricul-
tural Statistics Service (MASS).
The panel display was developed by County Exten-
sion Service personnel to provide information on pes-
ticide exposure prevention and safe handling of pesti-
cides. The display was designed for use at sites such
as health fairs, training programs, government and
professional building lobbies and shopping malls in
these counties.
Intervention effectiveness was evaluated using pre-
and postintervention telephone surveys. Participants
were chosen randomly in a two-stage process. In the
first stage, a random sample of farm operations was
selected from the list of farms maintained by the
MASS. A farm was defined as any place from which
$1,OOO or more of agricultural products were pro-
duced and sold in a given year. A farm operation was
eligible for participation if there was at least one mem-
ber of the household age 40 years or older who had
lived on a farm for the last five years. IS age was
The materials mailed out on pesticide exposure pre-
This study uzzs supported by the National Institutefor Occupafional Safety
and Health, Demonstration Cancer Control Projects for Farmers, G o p a -
five Agreement UO3ICCU 506136. The Minnesota C a m Control in
Farmers Study Group comprises Eugene L. Anderson, Ph.D. Timothy R.
Church, Ph.D, Bryan Dmd , Ph.D., Sherry A. Gahring, M. S, Dean Herr-
feld, M. S, \ohn 'I: Kelley, M.D. George Maldonado, Ph.D, Anne McGeary,
B.A. Kiypen Macgregos B.S., Wanda W Olson, M.S., and J k i n
Vance, M.D. Forfurther information, contact: Jefiey H. Madel, M. D,
Medical Department, 3M Center, Building 220-3W-05, St. Pad, MN
551441000.
Ma&, ef al. 149 Spn'ng 2000
selected to maximize participants for the breast cancer
screening part of the study.) The second stage in-
volved selecting at random one eligible person from
the household, age 40 years or older, who had lived
on a farm for five years, to serve as the study respon-
dent. Toensure a balanced number of subjects by age
and gender, an approximately equal number of sub-
jects was sampled from each of six age-gender strata
using a weighted random sampling scheme. Pre- and
postintervention interviews were conducted with the
same respondents within the same households ap
proximately 18 months apart.
The s ur v ey contained items on demographic charac-
teristics, health insurance coverage, access to and use
of preventive health services, use of agricultural pesti-
cides, perceived adverse health effects related to pesti-
cide exposure, perceived effectiveness of protective
gloves, perceived effectiveness of other protective
clothing and percentage of time spent using protective
equipment (Mandel, et al., 1996). Pesticide users were
defined as those farmers who had personally mixed,
loaded or applied pesticides while farming.
A comparison of pre- and postintervention use of
chemically resistant gloves and other protective cloth-
ing was conducted among those who used pesticides.
A relative change statistic was used as an indication of
the interventions effectiveness. This was calculated by
comparing the pre- and postintervention difference in
the use of protective equipment 75 percent of the time
or more in the intervention and control counties. (A
ratio of the relative change was calculated by compar-
ing percentage of change in intervention counties vs.
control counties.) A relative-use statistic was also cal-
culated, which indicated the prevalence of postinter-
vention use of protective equipment 75 percent of the
time or more in the intervention counties, compared
to prevalence of use in the control counties. This post-
intervention prevalence of use was also calculated by
preintervention use strata (0 to 74 percent and 75 per-
cent to 100 percent). Seventy-five percent was the
point that was judged best to represent consistent pro-
tective equipment usage.
Results
Eighty-one percent (1,613) of the farms listed by
MASS were determined eligible on the initial meen-
ing. Of these, 1,049 individuals (65 percent) completed
both pre- and postintervention interviews. In this
group, there were 186 farmers in the intervention
counties who used pesticides and 322 in the control
counties who used pesticides.
Table1 lists demographic characteristics of pesticide
users who participated in the surveys in the interven-
tion and control Counties. There were small differences
between the two groups. Ten women in the interven-
tion counties indicated that heyhad personally
mixed, loaded or applied pesticides vs. 29 in the con-
trol group. A greater percentage of the intervention
group were high school graduates (50 percent vs. 42
percent), whereas 33 percent had postsecondary edu-
cation, compared with 47 percent in the control group.
Table2 contains information on the use of protec-
tive equipment (gloves and other protective clothing
su& as boots, bibs and hats) for pesticide users in the
intervention and control populations, before and after
the intervention. For use of chemically resistant gloves,
there was an increase in those using gloves 75 percent
or more of the time in the intervention counties (59
percent before vs. 71 percent after the intervention)
and in the control counties (56 percent vs. 65 percent).
After the intervention there was a modest shift to us-
ing other protective clothing 75 percent or more of the
time in the intervention group (21 percent vs. 34 per-
cent) and to a lesser degree in the control group (21
percent vs. 25 percent). Controlling for farm size (480
acres or less and greater than 480 acres) did not
change this general relationship. Belief in the harmful-
ness of long-term exposure and in the effectiveness of
protective equipment as well as pesticide applicator
certification status did not change following the
intervention.
The overall relative change ratio (intervention to
control counties) was 1.4 (confidence interval [CI]=O.9
to 1.8) for glove use and 2.9 (CI=1.8 to 3.3) for other
protective equipment use.
It was difficult to assess the effect of the interven-
tion in women because 8 of 10 in the intervention
group and 23 of 29in the control group said they
used protective equipment 75 percent of the time or
more Similarly, it was difficult to assess the interven-
tion effect by type of pesticide used. Most farmers in-
dicated that they used several pesticide types. The ef-
fect of the intervention by age was assessed within
two categories (40 to 49years and 50to 64 years). Rel-
ative change in intervention and control counties was
minimal (both CIs included 1.0). Relative change in
intervention and control counties was assessed by ed-
ucational level (up to and including high school and
postsecondary school). In those with a high school ed-
ucation or less, the change ratio was greater in the in-
tervention counties than in the control counties (rela-
The l oud of Rural Hd t h 150 Vd. 16, No. 2
Table 1. Characteristics of Pesticide Users by Intervention Status.
Intervention Control
Number (Total = Number (Total =
1%) Percentage 322) Percentage
Women (age in years)
45 0 49
50 to 64
65+
45 to 49
50 to 64
65 +
Never graduated from high Smool
High &I graduate
Postsecondary education
Marital status
Married or living as married
Not married
Employed for wages or salary
Yes
No
Income ($)
-=10,000
10,000 to <20,000
20,000 to <35,000
35,000 to <50,000
50, m+
Refused /unknown
Yes
No
Men (age in years)
Highest education level attained
Have source for regular / preventive health care
Amount of land in production (acres)
0 to 80
81 to 160
161 to 480
481 to 800
m+
Grow corn
Grow soybeans
Types of pesticide used
Livestock insecticides
Crop insecticides
Herbicides
Fungicides
2
8
0
20.0
80.0
0.0
19
10
0
65.9
34.5
0.0
59
98
19
33.5
55.7
10.8
130
144
19
44.5
49.1
6.5
32
93
61
17.2
50.0
32.8
34
136
151
10.6
42.4
47.0
171
15
91.9
8.1
294
28
91.3
8.7
36
150
19.4
80.6
59
263
18.3
81.7
10
26
57
43
36
14
5.4
14.0
30.6
23.1
19.4
7.5
17
56
108
76
48
17
5.3
17.4
33.4
23.6
14.9
5.3
174
12
93.5
6.7
298
24
92.5
7.5
4
20
94
45
23
153
147
2.2
10.8
50.5
24.2
12.4
82.3
79.0
12
29
139
92
49
275
264
3.7
9.0
43.3
28.7
15.3
85.4
82.0
80
121
178
28
43.0
65.1
95.7
15.1
127
188
313
38
39.4
58.4
97.2
11.8
tive change=l.4, CI=0.9 to 1.9 vs. relative change=
1.1, CI=0.8 to 1.7, respecbvely), although not statisti-
cally sigruficant.
Table 3 shows the prevalence ratio of intervention
vs. control county use of gloves and other protective
equipment 75 percent of the time or more after the in-
tervention, stratified by preintervention use Farmers
in the intervention county who had used gloves less
than 75 percent of the time prior to the intervention
had a higher postintervention use of gloves than f arm-
Mandel, et al. 151 spring 2m
Table 2. Re and Postintervention Pesticide Exposure Beliefs and Practices by Intervention Status Among
Pesticide Users.
Preintervention Postintervention
Intervention Control Intervention Control
Number
(Percentage) (Percentage) (Percentage) (Percentage)
Number Number Number
Percent time use hemically resistant gloves
0 to 74
75 +
0 to 74
75 +
Yes
No
Not sure
Percent time use other protective clothing
Believe that long-term exposure to pesticides is harmful
Effectiveness of protective clothing in preventing pesticide exposure
very
Somewhat
Not effective
Not sure
Currently certified
Certification lapsed
Never certified
Not sure
Pesticide applicator certification status
76 (40.9)
110 (59.1)
147 (79.0)
39 (21.1)
177 (95.2)
7 (3.8)
78 (41.9)
97 (52.2)
7 (3.8)
154 (82.8)
9 (4.8)
22 (11.8)
1 (0.5)
2 (1.1)
4 (2.2)
140 (43.5)
182 (56.5)
254 (78.9)
68 (21.1)
294 (91.3)
10 (3.1)
18 (5.6)
151 (46.9)
157 (48.8)
3 (0.9)
11 (3.4)
259 (80.4)
12 (3.7)
50 (15.5)
1 (0.3)
53 (28.5)
133 (71.5)
123 (66.1)
82 (33.9)
178 (95.7)
3 (1.6)
5 (2.6)
88 (47.3)
3 (1.6)
1 (0.5)
156 (83.9)
7 (3.8)
23 (12.4)
0 (0.0)
94 (50.5)
111 (34.5)
211 (65.5)
240 (74.5)
82 (25.5)
298 (92.5)
15 (4.7)
9 (2.8)
149 (46.3)
166 (51.6)
4 (1.2)
3 (0.9)
252 (78.0)
12 (3.7)
59 (18.3)
0 (0.0)
Table 3. Intervention Effectiveness Among Pesticides Users.
Preintervention Utilization Status
75 Percent Zero to 74 Percent
Chemically resistant glove use
Other protective clothing use
Relative use and 95% CI
Relative use and 95% CI
1.2
(0.9 to 1.7)
1.5
(1.0 to 2.2)
1 .o
(0.9 to 1.1)
1.1
(0.8 to 1.6)
1. Postintervention relative use of chemically resistant gloves and other protective clothing (75 percent of thetime postintenrention
[intervention vs. control counties] stratified by preintervention utilization status).
The Journal of Rural Health I52 Vol. 16, No. 2
ers in the control county (relative use=1.2). The inter-
vention county farmers who had used other protective
equipment less than 75percent of the time prior to
the intervention also had a higher postintervention
use of such equipment than did the control county
farmers (relative use=1.5).
The intervention county farmers showed an increase
in their safe handling of pesticides with these educa-
tional efforts. This was true for the use of both gloves
and other protective clothing. For intervention farmers
there was a larger effect among those who, prior to
the intervention, used protective equipment the least.
Although the magnitude of this effect is modest, these
farmers represent the desired group in terms of en-
hancing pestiade exposure protection.
The educational intervention used had the advan-
tage of reaching farmers from a variety of demograph-
ic backgrounds. In previous studies the specific ap
proach of labeling was thought to be the single most
important factor in how farmers handle pesticides
safely (Hayes, 1980). Education has been identified as
an important conceptual component of safe pestiade
use (Abrams, et al., 1991; Marwick, 1989; McQuiston,
et al., 1994; Rosival, 1985), but it has focused on the
speafic use of protective equipment instead of on cre-
ating an awareness, as this intervention attempted to
do. The benefits of the former approach have been dif-
ficult to demonstrate (Rucker, 1994).
Results from the study as a whole, where the inter-
vention was broader based, indicate that education
might be more helpful than previously thought. The
relative simplicity of conducting an educational inter-
vention like the one reported here makes it an attrac-
tive option compared with administrative and legisla-
tive efforts. The intervention omitted other approach-
es, such radio and television, which could enhance
and sust ai n desired effects. Even a modest result such
as the one demonstrated suggests that additional re-
search on educational interventions could benefit
farmers. It is not known whether the observed effects
will be long-lasting because only one follow-up evalu-
ation was conducted. Interestingly, beliefs in the harm-
fulness of long-term pesticide exposure did not
change before or after the intervention (95.2 percent
vs. 95.7 percent).
The intervention involved four primary areas: ef-
forts through the school systems; direct mailing of
pesticide information to all defined farms in the coun-
ties; use of County Extension Service agents to in-
crease awareness of safe pestiade use; and attempts to
inform community physicians about pesticides. The
role of each of these efforts was not ascertained sepa-
rately in the analyses. However, the latter two efforts
involved the smallest amounts of time and might,
with greater emphasis, be areas that could further en-
hance safe pesticide use
There were several potential problems with t hi s
study. The results of the analyses are based on phone
interviews. Information on pesticide handling was not
validated in either the intervention county or the con-
trol county. Because participating farmers were over
age 40 years, results are not generalizable to younger
groups. Also, these results might not represent other
parts of the country, where different pesticides and
different pesticide handling practices might be used.
In summary, more broad-based educational inter-
ventions could be important in creating an awareness
of potential problems related to pesticide exposure
and in increasing farmers' use of protective equip
ment. More attention needs to be paid to enhancing
and sustaining the results of such education. There
appears to be a potential for improving the safety of
pesticide handling with a relatively simple educational
input. This could have important public health impli-
cations for t h i s occupational group.
References
Abrams, K, Hogan, D, & Maibach, H. (1991). Pesticide-related der-
matoses in agricultural workers. Occupational Medicine: State of
the Art Reviews, 6,4&&492.
Avory, G, & Coggon, D. (1994). Determinants of safe behavior in
farmers when working with pesticides. Occupatwnnl Medicine,
44,236-238.
Blair, A, Q Zahm, SH. (1991a). Cancer among farmers. Occupational
Medicine: State of the Art Reviews, 6, 415-428.
Cordes, D, & Rea, D. (1991). Preventive measures in agricultural set-
tings. Occupatwnal Medicine: State of the Art Reviews, 6, 541-550.
Elkind, P, & Cody-salter, H. (1994). Farm stressors: Thehazards of
agrarian l i fe Annals of Agricultural and Environmental Medicine, 1,
2S27.
among California fannworkers: Preliminary findings. Journal of
Rural h l f h , 12, 343-348.
Environmental Science Hialth, 15, 100S1021.
W, & Mandel, JS. (19%). Factors associated with safe use of ag-
ricultural pesticides in Minnesota. Journal of Rural Health, 12,
301-309.
Goldsmith, D, & Sineros, G. (1996). Cancer prevention strategies
Hayes, W. (1980). Factors limiting injury from pesticides. Journal of
Mandel, JH, Cam, P, Hillmer, T, Leonard, P, Halberg, I Sanderson,
Marwick, C. (1989). Educating farmers, physicians who treat them,
uandel, et a!. 153 Spring Zoo0
about rural lifes potential health hazards. Journal of the A m -
can Medical Association, 261, 343.
McQuiston, T, Coleman, N, Wallerstein, A, Marcus, I Morawetz, I. &
Ortlieb, D. (1994). Hazardous waste worker education: h g -
termeffects. Journal of Occupufionnl Medicine, 36,1310-1323.
Mullen, P, Gardiner, J Rosenman, K. Zhw Z, & Swanson. M. (1996).
Skin cancer prevention and detection practices in a Michigan
fann population following an educational intervention. h r n a l
of Rural Health, 12, 311-320.
grain fmmers. Unpublished doctoral dissertation. University of
Iowa, A m , IA.
Reding, D, Fisdter, V, Gunderson, P, Lappe, K, Anderson, H, & Cal-
vert, G (1996). Teens team skin cancer prevention JournaI of
Ogilvie, L. (1990). Agricultural safety equzpment usage: A surupy of Iowa
Rurnl Hearth, iz, 265-272.
Rosival, L. (1985). Pestiades [Review]. Scandinavian Journal of Mbrk,
Enoironment and Health, 11, 189-197.
Rucker, M. (1994). Attitudes and clothing practices of pesticide ap
plicators. Protecfk CIothing Systems and h4aterials, 55,79-94.
Runyan, J. (1994). Farm safety: Nay mles on pesticides. Agricultural
Outlook, 210, 24-26.
Stone,
Greenhouse applicators perceptions and use of personal pro-
tective equipment. J o u d of Emhmnmt al Health, 57, 1622.
Walker, J. (1988). Self reported stress symptoms in farmers. Journal of
Clinical Psychology, 1, 10-16.
Weigel, R. (1980, May/-). Helping farmers handle stress. Journal
Winstead, C. (1993). How pesticides are handled in a rural North
Carolina county: A survey of farmers. AmeticPn AssoCintirm of
Occupational Health Nursing, 41, 24-32.
Padgett, S, Witersteen, W, Shelley, M, & Chisholm. S. (1994).
of Extension, 3 7 4 .
Vd. 16, No. 2