Study design
We performed a randomized, double-blind, placebo-controlled, crossover trial collecting baseline and follow up measures of self-reported acute organophosphate poisoning symptoms and PchE activity in Nepali farmers randomized to either Group A (farmers spray with organophosphate in their first spray session and placebo in their crossover spray session) or Group B (farmers spray with placebo in first their spray session and organophosphate in their crossover spray session). The data presented was gathered by the principal investigator and a local partner over a 2-month period from September to November 2014.
Study area, population and recruitment
Very few studies examining occupational acute pesticide poisoning have been carried out in Nepal, however, the increasing pesticide consumption and health hazardous pesticide behavior among Nepali farmers, suggest a risk of significant exposure and poisoning [9, 20, 21]. This study was conducted in two purposely selected villages, Sukranagar and Jagatpur, in the Chitwan District of Nepal, where vegetable production is intensive and pesticide use extensive [22]. The study participants consisted of commercial vegetable farmers. Inclusion criteria were: male, minimum age 18 years, hand pressured backpack sprayer usage, and used to spray with moderately to extremely hazardous pesticides according to the World Health Organization’s (WHO) Pesticide Hazard Classification [23]. Exclusion criteria were: usual personal protective equipment use (respirator/mask with particulate filter, face shield, googles, gloves, boots, plastic poncho), unwilling to stay pesticide-free 7 days prior to each of the two spray sessions, or medical conditions interfering with PchE activity (liver disease, acute infection, chronic malnutrition, heart attack, cancer, obstructive jaundice, inflammation caused by various diseases, or use of pyridostigmine drugs) [24, 25].
The farmers were contacted through the management in their respective Vegetable Farmer Cooperative, who invited the farmers to a meeting regarding the study. The farmers were informed about the purpose of the study, its advantages and disadvantages, and were asked to volunteer for the study. All participating farmers signed an informed consent prior to participation. The study was approved by Nepal Health Research Council’s Ethical Review Board, Reg. no. 162/2014.
Intervention
Each farmer participated in two spray sessions, which each had an average duration of 2 h (minimum 1 h and 45 min; maximum 2 h and 15 min). Each spray session was preceded by minimum 7 days’ washout in which the farmers had to stay pesticide-free. In the organophosphate spray session the farmers sprayed with chlorpyrifos (organophosphate) 50% plus cypermethrin (pyrethroid) 5% EC, WHO Class II: Moderately hazardous, in the farmers’ usual dilution ratio. In the placebo spray session the farmers sprayed with the biopesticide Multineem, WHO Class U: Unlikely to present acute hazard in normal use. The dilution ratio for Multineem was 2 mL per L water. Multineem was purposely selected, as it is similar to the organophosphate pesticide in appearance and odor.
Data collection
A structured questionnaire interview and blood tests were conducted in each farmer’s home just before (baseline) and about 30 min after (follow up) each spray session by the local partner. Observations were made based on a checklist during each farmer’s spray session by the principal investigator and the local partner. The structured questionnaire interview and observation checklist were developed based on studies applying similar methods [13, 21, 26, 27]. Under the supervision of the principal investigator, the structured questionnaire interview was conducted face-to-face in Nepali by the local partner, and back translated into English later on the same day.
Self-reported acute organophosphate poisoning symptoms were assessed with the baseline question “Have you suffered from any of the following symptoms in the last 7 days?” and follow up question “Have you suffered from any of the following symptoms during or after spraying today?”. The definition of symptoms was based on WHO’s standardized list of clinical presentations of acute organophosphate poisoning [2]. Some of these clinical terms were considered difficult for the farmers to understand, therefore, these terms were translated into more understandable terms, the final list of self-reported acute organophosphate poisoning symptoms being: headache, dizziness, skin irritation, extreme tiredness, weakness, anxiety, excessive sweating, trembling hands, vomiting, diarrhea, abdominal pain, blurred vision, paralysis, salivation, tearing, lack of coordination, respiratory difficulties, confusion.
PchE activity was measured with a blood test using a Test-mate Che Cholinesterase System (Model 400) with a PchE Plasma Cholinesterase Assay Kit (Model 470). This system is considered to involve a reliable test method for measuring PchE inhibition of farmers due to organophosphate exposure [28, 29]. The blood test was performed in accordance with the Instruction Manual [30]. In short, after cleaning the farmer’s fingertip with water, soap and an alcohol swap, the fingertip was punctured with a sterile blood lancet, and a 10 microliter capillary tube was filled with blood. The capillary tube was inserted into an assay tube, and afterwards PchE Plasma Cholinesterase Reagent was added. A photometric analyzer calculated the test results. Given the large inter- and intraindividual variation of PchE, it is necessary to establish individual baseline PchE activities, adopted as a reference. The interindividual coefficient of variation is about 15–25%, whereas the intraindividual coefficient of variation is about 6–9% [31, 32]. PchE activities after exposure should be expressed as percentage change with respect to the individual baseline [31]. A 50% PchE inhibition from baseline is suggested as the level of acute overexposure and poisoning [5].
Descriptive variables partly included interviews about the background characteristics age, height, weight, marital status, educational level, farming experience, pesticide experience, medical condition, medication, tobacco status, alcohol consumption, names of pesticides the farmer usually uses, and partly observations of temperature, clothing during spraying, work practices during and after spraying, total area of spraying and if the backpack leaked. In addition to the PchE activity, we measured a hemoglobin (Hb) level with the Test-mate Che Cholinesterase System to evaluate whether the spray session resulted in dehydration of the farmers, which would be reflected as an increase in Hb level after spraying.
The informed consent, intervention, placebo and data collection procedure were pretested on one commercial vegetable farmer from Sukranagar. During pretesting of the intervention and placebo both the farmer and the local partner were unaware of the pesticide assignment (placebo) chosen by the principal investigator, however, both the farmer and the local partner believed that the placebo was organophosphate.
Sample size calculation
The sample size calculation was based on data from a previous study of 25 farmers in Chitwan whose acute organophosphate poisoning symptoms significantly increased after an organophosphate spray session (Neupane D. Pesticide exposure and its health effects among commercial vegetable farmers in Nepal. Unpublished masters’ thesis. University of Southern Denmark; 2012). The sample size was calculated using SAS 9.3 Proc Power for paired means (alpha 0.05, power 0.90). A sample size of 34 pairs was required. Forty-two farmers were enrolled taking into account possible dropouts and technical problems.
Randomization and blinding
The participating farmers were assigned a study ID number from 1 to 42. Forty-two opaque envelopes containing a study ID card with a study ID number were prepared by the principal investigator. A person not involved in the project (Khilendra R. Chaudhary) randomly allocated the envelopes to two buckets assigned Group A or Group B (ratio 1:1). The person wrote the assigned group on the study ID card, put the card back into the envelope, wrote the study ID number on the envelope, sealed the envelope and placed the envelopes in numerical order. The envelopes were placed in a secure location to which only the principal investigator had access.
On the day of each farmer’s first spray session, the principal investigator brought the numerically relevant sealed envelope, and opened it prior to group assignment. The principal investigator diluted the assigned pesticide (organophosphate or placebo) into the farmer’s backpack sprayer, while the farmer and the local partner stayed at a location where they could not see the mixing site. Afterwards, the backpack was handed to the farmer closed and ready to use. This procedure was repeated every time the farmer needed a backpack refill. Thus, the principal investigator assigned all farmers to their intervention, ensuring that the local partner and all farmers remained unaware of the pesticide assignment.
Data analysis
Double data entry, cleaning and statistical analyses were performed in SPSS version 22. P values lower than 0.05 were considered statistically significant. Mean (SD) was applied for parametric tests and median (IQR) for non-parametric tests. Body Mass Index was calculated as weight (in kilograms) over height squared (in meters).
To compare the farmers’ characteristics between Group A and Group B a two sample t-test was applied. If the normality assumption was violated the Mann Whitney U-test was applied. For categorical data the Pearson Chi-Square test was applied, or when more than 20% of the cells had expected values less than five, Fishers’ exact test (2×2 tables) and Likelihood Ratio (tables > 2×2) were applied.
To compare outcome measures between the organophosphate spray session and the placebo spray session, we used the model recommended by Altman and colleagues for analyzing crossover trials consisting of two sample t-tests or Mann Whitney U-tests [33]. First, we applied the tests to assess whether there was any period effect or treatment-period interaction, and given the lack of such effects, we applied the same tests to compare the two spray sessions.
To compare baseline levels within Group A and Group B respectively, and to compare baseline levels and follow up levels within the organophosphate and placebo spray session respectively, we applied the paired sample t-test when data were normally distributed and the Wilcoxon signed rank test when data were not normally distributed.