Study design and setting
This cross sectional study was conducted in Gwanda District, one of Zimbabwe’s main sources of Imbrasia belina. Gwanda district lies in Matabeleland South province. In Zimbabwe, provinces are divided into districts, districts are further divided into smaller administrative units called wards and the wards are subdivided into villages.
Prior to the feasibility study, a series of community engagement (CE) strategies were implemented over a period of several months with the aim of securing support and to ensure that this research project is aligned to the sociocultural, political and economic contexts in Gwanda. The CE activities utilised in this study had been successfully carried out in a previous study in the area and had included advisory meetings with local leaders, workshops and research training for local youth (23). Through consultations with various stakeholders, Garanyemba (a ward with 8 villages) was identified as the ideal study site within Gwanda district because of its close proximity to Imbrasia belina hotspots (24). The inclusion criteria were males and females aged 10 years and over residing in Garanyemba, Gwanda district for at least a year. Previous surveys in Zimbabwe and Botswana reported that children as young as 10 years old participate in mopane worm harvesting in various ways (25). Children and adults not residing in Gwanda for at least a year were excluded. The study was approved by the Medical Research Council of Zimbabwe (Ref number MRCZ/A/2486) and the University of KwaZulu-Natal’s Biomedical Research Ethics Committee BREC (Ref number BE 327/19).
Sample size and feasibility criteria
Due to budgetary limitations, cost was also a significant consideration for feasibility assessment. The main study had to be accommodated within the available budget, therefore, approximate anticipated costs per day during fieldwork were calculated. To efficiently utilise the available financial resources, we determined that data would have to be collected from 10% of the total sample per day during the main study.
For the main study, a total sample size of 462 was calculated using Cochran’s formula (26) based on the assumption of an asthma prevalence estimate of 50%, a 5% margin of error at 95% confidence interval and adjusted for anticipated non-response (20%). We therefore intended to recruit 46 eligible participants in order to evaluate the feasibility of the study procedures over a two day period in the context of the available funds. The first day was allocated for the household questionnaire and the second day was for clinic data collection. It is on this basis and in line with other objectives for this pilot study, that feasibility criteria for success were developed as follows:
- The research team to conduct at least 90% of targeted household questionnaire interviews on day 1
- At least 80% of consenting participants who completed the household questionnaire to go to the clinic on day 2 for clinical data collection
- The selected clinic and skin prick test (SPT) procedures to comply with all the World Allergy Organisation’s (WAO) safety recommendations for SPTs (27)
- Prevalence of adverse/unintended reactions associated with SPTs to be less than 1%
- Prevalence of mopane worm sensitisation to be at least 10% with or without suspected cross reactivity with other allergens in the panel
- The mopane worm allergen would be deemed clinically relevant if at least one participant sensitised to it had lung function abnormalities and allergic airway inflammation suggestive of asthma.
Participant recruitment and data collection
For this feasibility study, participants were recruited from randomly selected households within a 1km radius around the clinic. Individuals of both genders aged 10 years and over up to a maximum of 4 participants per household were recruited. Written informed adult consent and in cases of minor participants, parental consent and child’s assent, were obtained.
The main questionnaire
For data collection at the household, a comprehensive questionnaire whose questions were extracted from previously validated and standardised questionnaires, was filled in by trained research assistants using Kobo Collect software (28). The first section of the questionnaire had questions pertaining to demographic and socio-economic characteristics such as age, gender, education, occupation and monthly household income. The second section of the questionnaire interrogated knowledge, attitudes and practices with respect to asthma in the community. A series of dichotomous questions and Likert-scale type questions on the knowledge of risk factors, signs, symptoms management as well as attitudes and perceptions for asthma were adapted from a previously validated instrument from the Chicago Community Asthma Survey (CCAS-32) (29). The third and fourth sections of the questionnaire were, respectively for the data collection of environmental exposure history (including residential history, occupational history and lifestyle factors such as smoking and alcohol consumption) and respiratory health questions including self-reported doctor diagnosed asthma and asthma symptoms. To cater for the wide age range of the study population, relevant questions were extracted from the previously validated International Study of Asthma and Allergies in Childhood (ISAAC) and (30) and The European Community Respiratory Health Survey questionnaire (ECRHS) (31) that has been used to collect information from adults. The feasibility outcome for the questionnaire was the total number that could be completed in one day against a set target of 46. Also important was to qualitatively assess the appropriateness of the questionnaire for the targeted audience. Therefore at the end of each interview, participants were requested to comment on the clarity of questions, the length of the questionnaire or any other observation. After completing the main questionnaire, participants were invited to go to the clinic for further tests carried out in 3 key steps.
In step 1, a skin prick test was done to assess sensitisation to mopane worm and other locally relevant allergens included in the panel. The second step involved lung function assessment using spirometry and the third step was for FeNO tests to assess allergic airway inflammation. A qualified clinician was available on site in the event of any adverse reactions and arrangements were made at the nearby Provincial Hospital in the event that an emergency management became necessary.
Assessment of allergen sensitisation to mopane worm and other important allergens
Demographic and clinical data including anthropometric data, cigarette smoking, history of alcohol consumption, history of influenza or sinusitis, TB, allergic rhinitis and dermatitis were collected using a clinic data collection sheet designed specifically for this study. Since spatial differences are found in the distribution and sensitisation patterns of allergens (7, 32), it was necessary to also include other allergen extracts in order to identify the most relevant panel for the study area. The choice of allergens to include was based on our prior knowledge of the environmental characteristics of Gwanda district and a previous study in another part of Zimbabwe that used a similar panel (13). Their inclusion was necessary in order to identify participants that are uniquely sensitised to mopane worm. Allergen sensitization patterns were evaluated using 10 different inhalant allergen extracts. Participants were tested for allergen sensitisation to maize pollen, barley, 5 grass mix, cockroach, mosquito, the house dust mite (HDM) species Dermatophagoides pteronyssinus (D.pter.), Dermatophagoides farinae (D.far), and Tyrophagus putrescentiae, Alternaria and Imbrasia belina (mopane worm). Additionally, histamine (10mg/mL) and saline (0.9% NaCl) were included in the panel as positive and negative controls respectively. The skin prick test extracts used were sourced commercially (Stallergenes, France) with the exception of mopane worm which was prepared in-house.
Preparation of the mopane worm extract
Mopane worm allergy has only recently been recognised and there are currently no commercially available extracts for skin prick testing from any of the manufacturers. In order to perform SPT, a Mopane worm saline extract was prepared at the Biochemistry Department of the University of Zimbabwe. The use of in-house extracts and prick to prick testing is standard practice in allergy testing (33, 34). Established guidelines for extract preparation were followed (35, 36). Dried mopane worms were purchased from a Zimbabwean supermarket. Worm extract was prepared by subjecting, the worms to 3 cycles, a minute each, of alternate heating at 95°C and freezing at -195oC in liquid nitrogen. Thereafter 20 ml of Laemmli Sodium Dodecyl Sulphate (SDS) sample buffer (Bio-Rad Laboratories, Hercules, California, U.S) (37) was added to 2.5 g of mopane worms and sonicated for 10 minutes using an Ultra Turrax (IKA Labortechnik, Staufen, Germany). This was followed by centrifugation at 14 000 x g for 10 minutes. The supernatant was aspirated and stored at -20°C. Protein concentration was determined using spectrophotometry. The mopane worm in house Skin Prick Test was prepared by diluting the extract to a concentration of 1.437mg/dl using 0.9% sodium chloride.
Interpretation of Skin Prick Test Results
Trained technical staff conducted the skin prick tests at the clinic under the supervision of an allergy specialist. After locating the volar aspect of the forearm at least 2cm to 3 cm from the wrist, a drop of each test solution was placed on the skin in identical order for each participant tested. Each drop was immediately pricked with a sterile lancet and held against the skin for at least 1 second. Results were read 15–20 minutes following application (38). Skin-prick test wheal diameters exceeding 3 mm or greater than the saline control were considered as positive for sensitivity (39-41). This definition of sensitisation has been used in several other studies in Sub-Saharan Africa (2, 42, 43) including Zimbabwe (44, 45). An observation check list was used to rate the safety of the procedures against the WAO safety recommendations for SPTs.
Lung function assessment
Lung function was assessed using a portable office spirometer (KoKo® Legend) in line with the American Thoracic Society (ATS) and European Respiratory Society (ERS) Standardization guidelines (46). The spirometric parameters that were measured included Forced Vital Capacity (FVC); forced expiratory volume in 1 s (FEV1); forced expiratory flow between 25% to 75% of FVC (FEF25–75%); the Peak Expiratory Flow Rate and FEV1/FVC ratio. A skilled health worker conducted all the spirometry tests after demonstrating the appropriate breathing manoeuvre and explaining it in vernacular language to the participants. Tests were performed while the participants were seated. At least 3 satisfactory measurements were done for each participant after which the best, according to the ATS/ERS guidelines, was selected for analysis(46, 47). Lung function testing was discontinued if the participant was unable to produce acceptable results after 8 attempts. Lung function parameters were expressed as a percentage of the predicted normal values according to the European Community for Coal and Steel (ECCS) reference equation with a 10% adjustment for ethnicity (48).
Fractional exhaled nitric oxide (FeNO) test to assess allergic airway inflammation
Fractional exhaled nitric oxide measurement is a recognised non-invasive method for assessing allergic airway inflammation (49, 50). The FeNO test discriminates between different types of asthma and guides therapy. It is a novel way of confirming an asthma diagnosis and defining asthma theratypes. For participants with an abnormal measurement from at least one of the spirometric parameters, a Fractional exhaled nitric oxide (FeNO) test to assess allergic airway inflammation was recommended. A hand-held portable nitric oxide sampling device (NIOX VERO® Airway Inflammation Monitor (NIOX VERO); Circassia, Oxford, UK) was used according to the manufacturer’s instructions and in line with the current American Thoracic Society /European Respiratory Society (ATS/ERS) recommendations (51). A trained nurse carried out the tests to determine fractional exhaled NO (FeNO) from each eligible participant. The technique involved inspiration of NO-free air via a mouthpiece to total lung capacity, followed immediately by full exhalation at an even rate through the mouthpiece into the apparatus. The nurse carefully explained the technique, in vernacular, to the participants before carrying out the assessment. The ATS/ERS guidelines make evidence based recommendations in the interpretation of FeNO levels and they have been used in African studies (52, 53). Elevated FeNO levels greater than 50ppb in adults and greater than 35ppb in children indicate eosinophilic airway inflammation that is suggestive of probable allergic asthma. FeNO levels 25-50ppm in adults and 20-35ppb in children indicate possible airway inflammation but should be interpreted with caution and within the context of other clinical data collected such as the self-reported asthma symptoms and the skin prick test results (54).
Data analysis
The software selected for data analysis was Stata Release 13 (StataCorp, Texas, USA) (55). Recruitment rate was calculated as a percentage of the target (46 participants) in one day. Participants’ comments after completing the questionnaire were reviewed qualitatively. Response rate for clinic data collection was calculated as a percentage of those who completed the questionnaire the day before. Descriptive statistics were computed to summarise the variables. Median and interquartile range were used to summarize the continuous variables that were not normally distributed. Categorical variables were summarised as frequencies and percentages. To compare median values of spirometry and FeNO measurements between the participants sensitised and not sensitised to mopane worm extract, the Wilcoxon sum rank test (or the Mann-Whitney test) was used. To detect differences in proportions of self-reported asthma and respiratory symptoms between participants sensitised to mopane worm and those who were not, the Fisher’s exact test was used. This test was selected because of the small sample size and the expected frequency is less than 5 each of the cells. A new discrete variable named ‘Polysensitisation’ was generated which indicated how many allergens each individual was sensitised to. Significance was considered for p values less than 0.05 for all the statistical tests performed.