Adusei A, et al. Spatiality in Health: the Distribution of Health Conditions Associated with Electronic Waste Processing Activities at Agbogbloshie, Accra. Annals of Global Health. 2020; 86(1): 31, 1–12. DOI: https://doi.org/10.5334/aogh.2630 ORIGINAL RESEARCH Spatiality in Health: The Distribution of Health Conditions Associated with Electronic Waste Processing Activities at Agbogbloshie, Accra Abenaa Adusei*, John Arko-Mensah*, Mawuli Dzodzomenyo*, Judith Stephens*, Afua Amoabeng*, Saskia Waldschmidt†, Katja Löhndorf†, Kwame Agbeko*, Sylvia Takyi*, Lawrencia Kwarteng*, Augustine Acquah*, Paul Botwe*, Prudence Tettey*, Andrea Kaifie†, Michael Felten†, Thomas Kraus†, Thomas Küpper† and Julius Fobil* Background: A walk through the Agbogbloshie e-waste recycling site shows a marked heterogeneity in the spatial distribution of the different e-waste processing activities, which are likely to drive clustering of health conditions associated with the different activity type in each space. Objective of study: To conduct a spatial assessment and analysis of health conditions associated with different e-waste activities at different activity spaces at Agbogbloshie. Methods: A choropleth showing the various activity spaces at the Agbogbloshie e-waste site was pro- duced by mapping boundaries of these spaces using Etrex GPS device and individuals working in each activity spaces were recruited and studied. Upon obtaining consent and agreeing to participate in the study, each subject was physically examined and assessed various health outcomes of interest via direct physical examination while characterizing and enumerating the scars, lacerations, abrasions, skin condition and cuts after which both systolic and diastolic blood pressure values were recorded alongside the admin- istration of open and close ended questionnaires. All individuals working within each activity space and consented to participate were recruited; giving a total of one hundred and twelve (112) subjects in all. Results: A study of the choropleth showed that health conditions associated e-waste processing activi- ties were clustered in a fashion similar to the corresponding distribution of each activity. While a total of 96.2% of all the study subjects had cuts, the dismantlers had higher mix of scars, lacerations and abra- sions. Abrasions were observed in 16.3% of the dismantlers. Scars were the most common skin condition and were observed on the skins of 93.6% of the subjects. Prevalence of burns among the study subjects was 23.1%. Developing hypertension was not associated with activity type and while a total of 90.2% of subjects had normal blood pressure and 9.8% of them were hypertensives. Finally, 98.2% of respondents felt the need to have a first aid clinic at the site with 96.4% and 97.3% willing to visit the clinic and pay for services respectively. Conclusion: We conclude that while the observed injuries were random and were due purely to accidents without any role of spatial determinants such as the configuration, slope, topography and other subter- ranean features of the activity spaces, a strong association between the injuries and activity type was observed. 1. Background monitors, 7.0 Mt of temperature exchange equipment The United Nations Environment Program, UNEP (2005) (cooling and freezing equipment), 11.8 Mt of large estimates that between 20 and 50 million tonnes of equipment, and 12.8 Mt of small equipment and the e-waste are generated annually worldwide, accounting global is projected to grow to 49.8 Mt in 2018, with an for about 5% of all municipal solid waste. In a recent annual growth rate of 4 to 5 per cent [1, 2]. Not only is global waste stream analysis, the composition of global this figure representing the fastest growing municipal quantity of e-waste generated in 2014 comprised of 1.0 waste stream, it also has the potential of increasing fur- Mt of lamps, 3.0 Mt of Small IT, 6.3 Mt of screens and ther. In spite of the unprecedented growth in the global quantities, there is only limited recycling technology for disposal and safe management especially in the develop- * University of Ghana School of Public Health, GH ing countries where most of the wastes end up and are † RWTH Aachen Technical University, DE recycled by informal means using rudimentary methods Corresponding author: Julius Fobil (jfobil@ug.edu.gh) [3, 4]. Art. 31, page 2 of 12 Adusei et al: Spatiality in Health In Ghana for example, before the arrival of electronic 2. Materials and methods waste at Agbogbloshie, the area was a wetland known as 2.1. Study Area Old Fadama and predominantly a place for selling agricul- The study was conducted at the Agbogbloshie e-waste tural food products. It was the place of escape for refugees site which is arguably one of the largest e-waste dumps running from the Kokomba and Nanumba (located in the in the world. Agbogbloshie, known also as Old Fadama, Northern part of Ghana) conflict [3]. The Agbogbloshie serves as a home to some 40,000 people who are among area which is now a major site for informal e-waste recy- the world poorest urban populations [3, 4]. It is one of the cling is less than one kilometre from the Central Business biggest slums ever created by urbanization in West Africa District of Accra and is about thirty-one hectares in size [3–5, 12]. It is bounded eastward by the Korle River and [3, 5, 6]. It is bounded south, west and Northwest by the westward by the highly polluted Odaw River which feeds Odaw River, which feeds into the Korle Lagoon. There is into the Korle Lagoon on the south-side. Agbogbloshie a popular yam and onion wholesale market close to the serves not just a home to thousands of informal sector e-waste recycling site. In totality, the scrap yard takes up workers and site for e-waste recycling, it is also noted for about one hectare [7]. The recycling site has emerged as an the popular Agbogbloshie Market where all major food interesting case study for many reasons including that it products and farm produce are sold; popular farm pro- is a hub of global sink for used electronic products where duce are onions, tomatoes, vegetables and yams [3]. This recovery of products from the waste stream has multiple draws thousands of residents of the city to the area on adverse impacts on the environment and human health. a daily basis – a principal concern to health authorities Ongoing research work at the site has shown that recy- because, the widespread environmental pollutants due cling and product recovery activities emit huge quantities to informal level e-waste recycling activities are likely to of toxic chemical mixtures into the ambient environment impact adversely on a wider population and pupils of the [4, 8, 9]. Known widely for a mix of both formal and infor- near-by basic schools [3, 5]. mal economic activities, which co-exist in an environment charged with tensions among the economic actors, ongo- 2.2. Participant Recruitment ing research has established that toxic chemical mixtures Various activity spaces were mapped using an Etrex geo- often generated during e-waste recovery activities result- graphic positioning system (GPS). The workers were found ing in high levels of exposures which are affecting a broad mainly in small groups at the various activity spaces. Since spectrum of the urban population, especially among the there were not many e-waste workers per activity space e-waste recovery workers [4, 6, 10, 11]. at the site, all workers who were working and willing to At the recycling location, there are various activities participate in the study were recruited and included in specific to each part of the site, though one would also the study. In total, 112 individuals located in the different observe that certain activities cut across the entire recy- activity spaces who were involved with the collection, sort- cling area. For example, there are specific places one ing, dismantling and burning of e-waste materials were would see only dismantling of fridges, air conditioners, therefore included in the study. car parts, televisions and computers taking place but sorting of wires seems to run across the entire site. The 2.3. Study Procedures e-waste recycling site therefore shows a striking hetero- In this study, we demarcated the entire recycling area geneity of spatial patterns of the different e-waste pro- into subunit spaces in which similar recycling activities cessing activities, which are likely to drive clustering of took place and defined by homogenous activity patterns physical injuries and other health conditions associated using a Global Positioning System (GPS) device Etrex 10 with the different activity type in space. Skin conditions GPS (Garmin, Kansas City); making use of longitudes were classified into scars, rashes, peeling and burns. and latitudes. We then recruited the recovery workers In this study, a Global Positioning System (GPS) device performing recycling activities in the demarcated areas (Etrex 10 GPS: Garmin, Kansas City) was used to demar- after they had consented to participate in the study and cate the area into activity spaces specific to recycling of then conducted direct examination of the skin for injuries specific recovered products from the waste stream. The and other visible skin conditions as markers of previous aim was to study the recyclers working in those activity injuries to the skin. In a simple survey, we asked questions spaces in order to estimate the burden of various physical about their age, level of education, marital status, num- injuries and health outcomes such as scars, rashes, skin ber of months/years in this recycling job, and number of peeling, and burns unique to the recycling activities in hours on the job per day. We also measured and recorded each demarcated area and across space. The study esti- systolic and diastolic blood pressure of all study subjects mated the prevalence of injuries using markers such as in order to estimate the prevalence of stress related car- cuts, abrasions, avulsions and lacerations sustained by diovascular conditions among the recyclers as it is widely the workers during e-waste processing/recycling as they reported that the e-waste workers conduct their activities performed these activities. Finally, the study also assessed under very stressful environment [13]. For this reason, a blood pressure levels among the e-waste workers in each questionnaire was employed to collect demographic data demarcated activity space in order to determine if blood on all participants and to elicit responses from the sub- pressure levels differed across each worker class and activ- jects as well as collect information on their knowledge and ity space because of pervasively stressful nature of the use of personal protective equipment (PPE). This helped in recycling environment. assessing injuries profile on the skin as well as whether or Adusei et al: Spatiality in Health Art. 31, page 3 of 12 not a first aid clinic was a need in the area. A translator 3.1. Mapping of different e-waste processing activities and three other trained research assistants assisted with Table 1 shows the geocodes and coordinates of the questionnaire administration. All study participants were regions within which specific e-waste processing activities assigned a unique identity (ID) comprising a prefix and a took place in space at recycling site. number. The prefixes were defined by the activities per- formed by the subjects. The following prefixes were used: • WB – West Bank, situated on the boundary line of Abossey Okai road where it intersects with the Odaw • COR – Collecting River – a region where the predominant activity is • SOR – Segregation sorting. • DIS – Dismantling • FA – Fridge area – Marks the beginning of the burning • BUR – Burning. area. • SE – South end – representing heavy burning area. Factors such as age, history of high blood pressure, old • SCP – South Central Point, represents a region of scars before starting work, injuries and burns not sus- minimal burning. tained at the e-waste site were assessed to control for pos- • SCWB – South Central West Point – represents region sible confounding. of predominant dismantling. Physical examination of the skin was conducted to • SWE – South West End – region close to Central numerate scars over skin as an indicator of injuries and Gospel Church where sorting and some dismantling these were classified into the different types of injuries took place. which were studied. Scars are broken patches in the skin • MWE – Middle West End – region close to the football indicating points of skin break due to cuts, puncture, sore, park where predominantly dismantling and minimal burns and other forms of injuries to the skin. During phys- sorting took place. ical examination, we counted the on hands, legs and over • BIF – a region near the Blacksmith Institute Facility the entire body-surface. Workers identified and differenti- where true recycling/refurbishing of recovered ated between scars which were sustained during e-waste products took place. work and those not associated with e-waste work. Systolic • NEW – North End West – a region close to the and diastolic blood pressure of all study participants were boundary line of the Abossey Okai road where sale of checked three consecutive times with a calibrated digital foodstuff is predominant activity. Omron Blood Pressure Monitor, by Omron Corporation, • NCP – North Central Point – a region where main Japan. The mean was found for both systolic and diastolic activity is dismantling. blood pressure for each study subjects in accordance with • WBE – West Bank End. World Health Organization (WHO) standard for diagnos- ing hypertension. Figure 1 shows choropleth or map of activity areas produced using the geo-coordinates presented in Table 2.4. Data Analysis 1 below. Points taken with the GPS device were entered into excel As seen on the choropleth, regions WB, FA, SE and WBE to generate longitudes and latitudes and these were were mapped along the banks of Odaw River while regions exported into ArcGIS 10.1 for mapping. The completed NEW, WB, WBE and NCP bounded the Abossey Okai road. questionnaires were crosschecked by the study team. This was predominantly an area for sorting and the These were all entered into Epi InfoTM 7 software. They observed health conditions were mostly cuts, lacerations, were then entered into Microsoft Excel spread sheet which were subsequently exported to Stata Version 12. The data were cleaned and validated before analyses were Table 1: Coordinates of mapped out areas. conducted. Description X (Longitude) Y (Latitude) 2.5. Statistical Procedures BIF –0.22478 5.553944 The use of Pearson’s Chi square was employed to test for FA –0.22653 5.552306 differences in proportions across the different groups defined by the distinct activity spaces. Fischer’s exact MWE –0.22447 5.553111 p-value was determined where cell count was low; below NCP –0.225 5.553778 5. Analysis of variance (ANOVA) was conducted to deter- mine if differences existed across groups for continuous NEW –0.22642 5.554167 variables. SCP –0.22669 5.551056 SCWB –0.22522 5.551528 3. Results Findings of this study are presented to highlight the rela- SE –0.22731 5.550583 tionship between the different e-waste processing activi- SWE –0.224 5.552472 ties that take place at the Agbogbloshie dumpsite and WB –0.22397 5.553278 physical injuries as well as other skin conditions that the e-waste workers experience. WBE –0.21667 5.553278 Art. 31, page 4 of 12 Adusei et al: Spatiality in Health Figure 1: Spatial distribution of e-waste processing activities at Agbogbloshie. scars and rashes. The area marked as FA (fridge area) 3.2. Demographic Characteristics of the Study marked the beginning of e-waste burning, which Participants increased in intensity toward the area marked as SE One hundred and twelve (112) e-waste workers at dif- (South End) where heavy burning activities took place. ferent activity spaces comprising Collectors, Sorters, The observed injuries were cuts, lacerations, abrasions, Dismantlers and Burners were recruited into the study. burns, scars, rashes and skin peeling. Within the region The demographic profile of the participants is shown in described as the SCP (south central point), some mini- Table 2 below. mal burning was observed and the corresponding pre- Characteristic background of study participants con- dominant health conditions observed were burns, scars sidered were sex, ages, marital status, region, tribe, edu- and cuts. In the area marked as SCWP (south central cational background, length of work and daily working west point), dismantling was the predominant activity hours as captured in the questionnaire. All the study par- and the commonest the observed physical injuries were ticipants were males. The ages of the participants ranged mainly cuts, lacerations, scars rashes and burns. Sorting from 16 to 55 years. Sixty three (63) out of the 112 par- and some dismantling were major processing activities ticipants representing 56.3% were in their twenties and observed at the are labelled SWE (South West End) and 6.3% of them were 40 years and above. Those between the corresponding reported physical injuries were cuts, 20–29 years dominated other age groups across the lacerations, scars and rashes. In the region marked as activity spaces. There were more married workers than MWE (Middle West End), predominantly dismantling and the unmarried. Fifty participants (44.6%) and 62 (55.4%) some sorting took place. Again, the physical reported were unmarried and married respectively. Thirty-nine were mostly cuts, laceration, scars and rashes. The region (39) of them; representing 34.8% had no formal educa- marked as BIF (Blacksmith institute’s facility) where tion whereas 32.1% had primary education with 11.6% the association office is located, fabrication of e-waste and 19.6% having had Secondary and Junior education materials into cooking wares and pots (popularly called respectively. Majority (89.3%) hailed from the Northern “gyapa” in Ghana) was observed, but no subjects were Region of Ghana with 10.7% of them hailing from the recruited from this area. The area marked as NWE (North other regions of the country. Eighty-nine 89 (79.5%) of End West) which bordered the Abossey Okai road and the study participants were of Dagomba ethnicity. There were area marked as NCP (north central point) housed sales of two foreign nationals from Nigeria and Togo and while farm produce and refurbishing activities. The most domi- the majority of the workers had been working for 11–15 nant health conditions observed in this area were cuts, years (73.2%), 23.3% had been in the e-waste process- abrasions and scars, although less pervasive compared to ing business for 5–10 years. In addition, whereas 53.6% other activity spaces. had worked for more than 5 years, 33.9% and 11.6% of Adusei et al: Spatiality in Health Art. 31, page 5 of 12 Table 2: Demographic characteristics of study subjects. Characteristic Activity space P-value Collecting Sorting Dismantling Burning Total Age P = 0.010 <20 3 (8.6) 1 (6.3) 7 (18.0) 9 (40.9) 20 (17.9) 20–29 22 (62.9) 7 (43.8) 23 (59.0) 11 (50.0) 63 (56.3) 30 – 39 8 (22.9) 6 (37.5) 8 (20.5) 0 (0.0) 22 (19.6) 40+ 2 (5.7) 2 (12.5) 1 (2.6) 2 (9.1) 7 (6.3) Mean (SD) 26.1 (6.8) 28.1 (7.6) 24.5 (6.1) 22.1 (9.9) 25.3 (7.5) Marital status Unmarried 16 (45.7) 6 (37.5) 18 (46.2) 10 (45.5) 50 (44.6) Married 19 (54.3) 10 (62.5) 21 (53.9) 12 (54.6) 62 (55.4) Highest education P = 0.314 None 16 (45.7) 8 (50.0) 11 (28.2) 4 (18.2) 39 (34.8) Primary 9 (25.7) 4 (25.0) 13 (33.3) 10 (45.5) 36 (32.1) Junior High 3 (8.6) 0 (0.0) 6 (15.4) 4 (18.2) 13 (11.6) Secondary 7 (20.0) 4 (25.0) 8 (20.51) 3 (13.6) 22 (19.6) Region Northern 33 (94.3) 14 (87.5) 33 (84.6) 20 (90.9) 100 (89.3) Others 2 (7.1) 2 (12.5) 6 (15.4) 2 (9.1) 12 (10.7) Ethnic group Dagomba 33 (94.3) 14 (87.5) 27 (69.2) 15 (68.2) 89 (79.5) others 2 (5.7) 2 (12.5) 11 (28.2) 7 (31.8) 22 (19.6) Length of work P = 0.497 6–12 months 4 (11.4) 1 (6.25) 4 (10.3) 4 (18.2) 13 (11.6) 1–5 years 13 (37.1) 4 (25.0) 11 (28.2) 10 (45.5) 38 (33.9) >5 17 (48.6) 11 (68.8) 24 (61.5) 8 (36.4) 60 (53.6) Daily working hours 5–10 hours 9 (25.7) 3 (18.8) 9 (23.8) 5 (22.7) 26 (23.2) 11–15 hours 24 (68.6) 12 (75.0) 30 (76.9) 16 (72.7) 82 (73.2) Fischer’s exact p-value was used due to low cell count, below 5. the workers had worked for 1–5 years and 6–12 months of burners were observed to have had scars revealing a respectively. history of lacerations. Although abrasions were not as common among the e-waste workers compared to cuts 3.3. Injury experience among e-workers in the and lacerations, 38.5% of dismantlers and 18.9% of sort- different activity spaces ers were observed to have scars indicating past experi- Table 3 shows injury profile of the workers within the ences of this class of injuries. The burners had the highest different activity spaces. Generally, cuts were the most prevalence (77.3%) of burns followed by dismantlers with frequent injuries in all the activity spaces as compared to 11.4% (Table 3). The burns experienced by the disman- the other injuries. A majority (96.2%) of all study subjects tlers could possibly be chemical burns. As they disassem- had one form of cuts or the other. Overall, cuts were most bled the WEEEs, the harmful chemicals could spill and get common (94.9%) among dismantlers compared to the into contact with their bodies to cause various degrees of other groups. In particular, the burners; in whom 90.9% injuries to the skin, including burns. cuts were observed, were observed to be highly exposed to risk of fire burns in addition to being the second group 3.4. Assessment of skin conditions of workers most prone to injury experience. Lacerations were the sec- Table 4 presents skin conditions of workers in each ond most common injury conditions observed on 46.6% worker-category. An assessment of recyclers’ skin revealed of workers across the activity spaces and whereas 74.4% that all dismantlers (100.0%), 90.5% of burners, 96.4% of dismantlers were observed to have lacerations, 54.5% of collectors and 87.5% sorters presented with scars of Art. 31, page 6 of 12 Adusei et al: Spatiality in Health Table 3: Injury experience among e-waste worker groups. Characteristic Activity space Statistic Assessed Collectors Sorters Dismantlers Burners Total P-value Injuriesb Cuts 30 (85.7) 13 (81.3) 37 (94.9) 20 (90.9) 100 (96.2) 0.799 Lacerations 10 (26.3) 5 (31.3) 29 (74.4) 12 (54.5) 56 (46.6) 0.208 Abrasions 1 (2.8) 3 (18.9) 15 (38.5) 3 (13.6) 22 (16.3) 0.038 Total no. of cases 35 (100.0) 16 (100.0) 39 (100.0) 22 (100.0) 112 (100.0) b Multiple responses allowed for various injuries and skin conditions. Fischer’s exact p-value was used due to low cell count, below 5. Table 4: Skin conditions among the different worker-category/activity space. Characteristic Assessed Activity space Statistic Collectors Sorters Dismantlers Burners Total P-value Skin conditionsb Rashes 27 (96.4) 14 (87.5) 39 (100.0) 19 (90.5) 99 (93.6) 0.275 Scars 3 (10.7) 3 (8.5) 4 (10.5) 6 (28.6) 16 (14.6) 0.201 Skin peeling 0 0 3 (7.9) 1 (4.8) 4 (3.9) 0.368 Cumulative no. of cases 35 (100.0) 16 (100.0) 39 (100.0) 22 (100.0) 112 (100.0) Burns <0.001 No 30 (93.8) 14 (93.3) 31 (88.6) 5 (22.7) 80 (76.9) Yes 2 (6.3) 1 (6.7) 4 (11.4) 17 (77.3) 24 (23.1) Cumulative burns 32 (100.0) 15 (100.0) 35 (100.0) 22 (100.0) 104 (100.0) Mean Scar #s/person 35 16 36 22 112 Number Mean 17.0 21.4 30.6 27.0 24.3 Stand. dev 9.8 10.5 12.1 10.9 12.3 P50 16 23 32 30 25.5 IQR 15 14.5 13 10 18 b Multiple responses allowed for various injuries and skin conditions. Fischer’s exact p-value was used due to low cell count, below 5. Stand dev – standard deviation, P50 – 50th Percentile, IQR-Intra Quartile Range. varying degree of density over the skin. Skin rash was 3.5. Association between injury levels and most common skin conditions (28.6%) among burners workers-characteristics compared to 10.7% among collectors and 10.5% among Among the different injury conditions, whereas cuts dismantlers. were more common among the workers as indicated in Generally, skin peeling was low among all worker- Table 5 age and marital status did not show any associa- groups and while the proportion of workers observed to tion with the injury conditions. However, the frequency have this skin condition was 7.9% among dismantlers and of scars increased with respect to the length of time spent 4.8% among burners, none was observed among collec- performing e-waste work and number of hours/day spent tors and sorters. Sustaining an injury and being afflicted working (Table 5). Overall, skin conditions did not show with a given skin condition were not associated with work- association with subject characteristics such as age, mari- ing within any activity space. As it was anticipated, the fre- tal status, length of time on the job and number of hours quency of burns was higher among burners (77.3%), than spent per day working. sorters (11.4%), dismantlers (6.7%) and collectors (6.3%) (Table 4). With a mean scar density of 30.6, scars on skin 3.6. Assessment of Systolic and Diastolic Blood Pressure surface were more widespread among the dismantlers An assessment of mean systolic and diastolic blood pres- than in burners (mean scar density = 27.0), than in sorters sure of the workers did not show evidence of differences (mean scar density = 21.4) and than in collectors (mean across the activity spaces and the workers generally had scar density = 17). normal blood pressure levels (Table 6). Adusei et al: Spatiality in Health Art. 31, page 7 of 12 Table 5: Association between injury levels and worker-characteristics. Characteristic Injuries Total no. of P-value Cuts Lacerations Abrasions cases Age 0.250 <20 14 (77.8) 4 (22.2) 2 (11.1) 18 20–29 46 (80.7) 12 (21.1) 5 (8.8) 57 30–39 14 (82.4) 4 (23.5) 2 (11.8) 17 40+ 2 (33.3) 2 (33.3) 2 (33.3) 6 Marital status 0.114 Unmarried 36 (81.8) 6 (13.6) 2 (4.6) 44 Married 40 (74,1) 16 (29.6) 9 (16.7) 54 Length of work 0.901 6–12 months 10 (83.3) 2 (16.7) 2 (16.7) 12 1–5 years 25 (75.8) 7 (21.2) 3 (9.1) 33 >5 40 (77.3) 13 (25.0) 6 (11.5) 52 Daily working/hours 0.416 5–10 hours 15 (71.4) 5 (23.8) 1 (4.8) 21 11–15 hours 59 (80.8) 15 (20.6) 9 (12.3) 73 Table 6: Mean systolic and diastolic blood pressure across activity spaces. Characteristic Activity space P-value Collecting Sorting Dismantling Burning Total Blood pressure Mean systolic Hg 123.1 ± 11.3 120.7 ± 11.7 121.7 ± 12.8 119.7 ± 3.9 121.6 ± 12.3 0.773 Mean diastolic Hg 72.8 ± 8.7 73.1 ± 10.5 74.1 ± 8.7 72.4 ± 9.5 73.2 ± 9.0 0.887 Hypertension 0.315 Normotensive 29 (82.9) 16 (100.0) 36 (92.3) 20 (90.9) 101 (90.2) Hypertensive 6 (17.1) 0 (0.0) 3 (7.7) 2 (9.1) 11 (9.8) Total 35 (100.0) 16 (100.0) 39 (100.0) 22 (100.0) 112 (100.0) ANOVA used to test for association. 4. Discussion • 50% of injuries were cuts and lacerations primarily A wide variety of hazards is associated with waste to hands and forearms during the disassembling recycling industry and while quite a reasonable num- process. ber is documented in the formal recycling sector, all • 30% were sprains and strains – associated with the the hazards have generally gone undocumented in the manual handling tasks and repetitive arm work in informal sector waste recycling industry. For instance, the disassembly process. a study of consumer waste recycling in Quebec found • 10% were bruising – mainly involved in the manual elevated exposures to airborne bacteria, noise, carbon handling of the TVs and computers from the storage monoxide (during winter months only) and ergonomic and disassembly processes. hazards [14]. Other specific types of waste recycling in the formal sector have noted elevated exposures to lead Previous studies on informal sector e-waste recycling in lead acid battery recycling and polybrominated diphe- at Agbogbloshie have also reported several risk fac- nyl ethers in electronic recycling [15, 16]. In respect of tors at play in the recycling process (Akormedi et al., specific health outcomes, a baseline data developed 2013, Asampong et al., 2015, Yu et al., 2016), over a 3-year period from 2007/08 to 2010/11 on the but to the best our knowledge, we report for the Occupational Health and Safety (OHS) risks associated first time, spatial clustering of health conditions with the e-waste recycling industry in Australia has associated with the informal e-waste recycling shown that: sector. Art. 31, page 8 of 12 Adusei et al: Spatiality in Health 4.1. Injury Types, Frequency and Distribution to those who worked for 1–5 years and 6–12 months Electronic waste (e-waste) recycling workers may respectively. A plausible explanation for the observed level encounter different types of hazards including the risk of injury experience is that the workers who had been on of injury, hearing loss, and exposure to toxic dusts and the job longer probably had become well adapted to the other noxious chemicals [17–20]. These hazards can cause hazards associated with the job and were less “risk-averse” permanent and serious health problems that could begin to those hazards. This assertion is supported by the obser- without workers being aware of them [14, 21–23]. In the vation that non-usage of personal protective equipment current study, injuries arising from e-waste processing (PPEs) was more common in those who had spent >5 activities which were observed among e-waste workers years (55.7%) as compared to 50.0% and 46.2% for those were classified as cuts, lacerations and abrasions. Conceiv- who had spent 1–5 years and 6–12 months on the job ably, cuts were most common among dismantlers because respectively. Our finding is quite consistent with that of a of the nature of the operations associated with the dis- study which evaluated health and safety hazards at a scrap mantling process. Dismantling activity involves the use metal recycling facility in Washington State and report- of hammer and physical force to disassemble electrical ing that the use of personal protective equipment and and electronic equipment component parts and for this exposure controls was generally low among the recycling reason; the activity is associated with frequent cuts due workers [7, 15, 20, 23–25, 32, 34, 35]. Use or non-use of to the physical force applied. The most prevalent injuries PPEs has a positive relationship with worker’s perception were cuts, as 96.2% were observed to have scars due to of risk associated with the activities or tasks they under- cuts and it made sense because sharps were the most took, which was suggestive that the use of PPEs was rela- abundant component materials in the e-waste stream. tively higher among workers who had been on the job for While the level of the different injury types was observed shorter time periods. This assertion is corroborated by the to be a function of the type of materials handled by the observation that use of PPEs was (53.9%) among workers e-waste workers, the frequency of the injuries could be who had worked for 6–12 months, (50.0%) among those explained by the level of safety measures in place and the who had worked for 1–5 years and (45.0%) among those level care observed by the workers. While some studies who had been on the job for more than 5 years [5, 12, have observed that workers who engaged in recycling of 36, 39]. Lastly, injuries were more common in and around cathode ray tubes (CRTs) sustained cuts than other inju- the palms as compared to the feet. This meant that as a ries from broken tubes, others have reported that e-waste first step toward improving health and safety, hand-glove workers were at a higher risk of work related accidents and usage would be much more important than for example more likely to suffer physical injuries and physical disa- the wearing of safety shoes. bilities than the general population [12, 24–30]. Lacera- tions and abrasions were observed to be a consequence 4.2. Skin Injuries, Type and Frequency of friction-based injuries, were second (46.6%) and third Burns did not only include burns due to fire-burns dur- (16.3%) most frequent injuries respectively and were ing e-waste burning process, but also chemical burns reported to result from falls and other accidents involving sustained as a result of coming into contact with harmful falling objects pulling/sliding over the skin. The 20-29 substances. The skin is the primary exposure surface to years group had the highest proportion (80.7%) of work- physical injuries and at the same time the natural expo- ers with cuts, probably suggesting that they were more sure barrier to chemical agents in environmental media prone to taking higher risks rather than being less careful and therefore the skin condition serves as a marker of pre- or while performing recycling work. This age group also vious and current exposure levels. Scars being markers of had the highest number of individuals with lacerations physical injuries to the skin were the most common skin and abrasions confirming our assertion that the observed conditions across all worker-categories, suggesting that high frequency of cuts was a consequence of increased all categories of workers were exposed to cuts and physi- intensity of physical activity and more tedious work rather cal injuries and dismantlers in whom scars were most than due to the type of materials handled. These results common were also observed to be engaged in the most are not substantially different from those of a study of vigorous and most tedious physical activity. In unpro- working conditions in a consumer waste recycling facility tected workers or workers without PPEs, it would be seen in Sweden in which workers reported regular exposures to that scar-density (i.e. the highest average scar-count per noise, ergonomic hazards, falls, lifting, and awkward pos- person) was greatest among dismantlers who used heavy tures as well as regular occurrence of accidents, injuries, tools to break apart, the components of electrical and and pain related to ergonomic hazards [25, 31–33]. electronic materials and therefore most prone to physical Into-the-bargain, the length of time on the job was injuries which were lowest among the collectors. In terms observed to be an important determinant of injury experi- of the frequency of skin condition, scars were common- ence among the e-waste workers and those who had spent est, followed by rashes and with skin peeling being the more than 5 years on the job sustained twice and 4 times least common across the worker-categories. We observed cut injury levels compared to those who had spent 1-5 years no cases of skin peeling among collectors and sorters and 6-12 months on the job respectively. Again, scar injury which meant that sorting and collection activities did levels corresponded to laceration injury experience in that not unduly expose the e-waste workers to harmful sub- workers who spent more than 5 years on the job experi- stances that are part of technical formulation of electrical enced twice and 4 times laceration injury levels compared and electronic devices and which irritate the skin. On the Adusei et al: Spatiality in Health Art. 31, page 9 of 12 basis of the nature of recycling activity alone, dismantlers meteorological conditions [4, 6, 26, 29, 30, 38, 42, 43, 57]. were more likely to come into direct contact with both As such, complex spatial patterning can occur in urban air elemental and inorganic mercury, nickel, beryllium, lead quality making the variability of such phenomena difficult and several organic compounds such as flame retardants to characterize as different pollutants often exhibit differ- which are found in the e-waste materials and may spill ential spatial patterns (e.g., ozone vs. nitrogen dioxides) due to heavy cracking of the dismantling process. Some [58]. studies have shown that Hg may cause skin rashes, skin However, recent advancements made in the develop- discoloration, scarring as well as a reduction in the skin’s ment of spatial methods for studying spatial variation resistance to bacterial and fungal infections [26, 29, 31, in health outcomes have made it possible to study spa- 36, 40–42]. Mercury is ubiquitous in e-waste materials tial variability of more concrete health outcomes such as which is commonly found in thermostats, sensors, relays, injuries, skin conditions and scars with very high degree thermometers, switches commonly found on printed cir- of certainty; although admittedly, most public health cuit boards, mobile phones, batteries and in flat display and epidemiological studies have not fully embraced the panels and several studies have suggested that the use of application of advanced spatial methods probably due to mercury is likely to increase in flat panel displays in years limited understanding of the application of these new to come and this will further increase the risk of exposure spatial methods [47–49]. While there is scientific consen- to mercury [7, 22, 43–46]. sus that ecological studies are more reliably conducted over fairly large spatial areas over which multiple socio- 4.3. Spatiality, Hazard/Injury Distribution and Spatial economic and environmental factors acting over distinct Ordering spatial scales occur in more spatially explicit manner, a In spatial epidemiology, spatial frailty models are com- few other robust spatial statistical methods which can be monly used to estimate random effects – spatially explicit applied to study environmental health phenomena over ordering of events such as health outcomes or any other small spatial scales also exist and are widely applied with health events occurring in space [47–50]. At Agbogbloshie, high degree of success in spatial epidemiology [49, 51, 53, e-waste processing activities which are strongly associated 58–60]. Such methods include the one we applied in this with health outcomes; especially injuries that may under- study in which a hand-held GPS was used to map out small lie or determine the ordering of these health events across areas so that scars and other physical/concrete health space. The observed injuries were random and were due events were reliably counted on e-waste workers who purely to accidental events without spatial determinants work within these small spaces on daily basis. Despite the such as the configuration, slope, topography, and other fact that these methods are time consuming, they offer subterranean features. However, many of the observed large degree of control and flexibility and are therefore health conditions tended to cluster according to the conceivably reliable. type of activities and by extension, activity spaces, e.g. underlying variations in environmental exposures and dis- 5. Conclusions tribution of risk factors/hazards which in turn ultimately In conclusion, while the observed injuries were purely ran- determine the distribution of the health outcomes. Scars dom without any role of spatial determinants such as the for instance were observed to cluster within activity spaces configuration, slope, topography and other subterranean in which the main activity in the area was dismantling, features of the activity spaces, an association between the while burns tended to cluster around areas where the injuries and activity type was observed. For this reason, predominant activity was e-waste burning. On the con- a targeted occupational health and safety (OHS) program trary, skin condition did not show any spatial clustering will considerably minimize injury rate among the most at- suggesting that the determining factors (exposures) skin risk group (and in this case; the dismantlers) would help conditions were not location or space-dependent and are minimize the pervasive injuries among the workers. probably defused exposures in nature (Figure 1). This is consistent with the reports of other studies which showed 6. Limitation of Study that the geographical distribution of health outcomes is This study evaluated the relationship between job-task influenced by socio-economic and environmental factors and injury experience across worker-groups and because operating at different spatial scales [21, 37, 51–53]. Spatial some of the workers performed more than one task, it was variability in geographic events can be revealed with semi- not possible to objectively make a distinction between parametric Geographically Weighted Poisson Regression scars due to the different job-tasks. However, this limita- (sGWPR), models that can combine both spatially varying tion was offset by the objective enumeration of scar count and spatially non-varying parameters [21, 53–55]. Indeed, which was the main outcome of interest. measuring spatial relationships between socio-economic and environmental factors on the one hand and health Funding Information events on the other is a fairly new scientific endeavour Financial support for this study was provided under US and often very challenging; but forms a crucial part of spa- NIH/FIC GEOHealth planning grant #IR24TW009497-01 tial epidemiology [20, 48, 56]. To emphasize this point, air and GIZ of German Government. quality within urban environments involves a mixture of gaseous and particulate concentrations that are affected Competing Interests by a variety of emission sources, local topographies, and The authors have no competing interests to declare. Art. 31, page 10 of 12 Adusei et al: Spatiality in Health Author Contribution Environment International. 2011; 37(5): 921–928. All authors have contributed in drafting; review and DOI: https://doi.org/10.1016/j.envint.2011.03.011 approving the manuscript. 13. Yu EA, et al. Informal processing of electronic waste at Agbogbloshie, Ghana: workers’ knowledge about References associated health hazards and alternative liveli- 1. Baldé CP, et al. The Global E-waste Monitor 2017 hoods. Glob Health Promot. 2017; 24(4): 90–98. DOI: Quantities, Flows, and Resources, in Sustainable https://doi.org/10.1177/1757975916631523 cycles program. Bonn/Geneva/Vienna: United 14. Lavoie J, Guertin S. Evaluation of health and safety Nations University (UNU), International Telecom- risks in municipal solid waste recycling plants. Jour- munication Union (ITU) & International Solid Waste nal of Air and Waste Management Association. 2011; Association (ISWA). Geneva. 2017; 116. 51(3): 352–360. DOI: https://doi.org/10.1080/104 2. Landrigan PJ, et al. The Lancet Commission on 73289.2001.10464278 pollution and health. Lancet. 2018; 391(10119): 15. Were FH, et al. Air and blood lead levels in lead 462–512. DOI: https://doi.org/10.1016/S0140-67 acid battery recycling and manufacturing plants in 36(17)32345-0 Kenya. Journal of Occupational and Environmental 3. Akormedi M, Asampong E, Fobil JN. Working Hygiene. 2012; 9(5): 340–344. DOI: https://doi.org/ conditions and environmental exposures among 10.1080/15459624.2012.673458 electronic waste workers in Ghana. International 16. Schecter A, et al. A newly recognized occupational Journal of Occupational and Environmental Health. hazard for US electronic recycling facility workers: 2013; 19(4): 278. DOI: https://doi.org/10.1179/204 polybrominated diphenyl ethers. Journal of Occupa- 9396713Y.0000000034 tional and Environmental Medicine/American College 4. Feldt T, et al. High levels of PAH-metabolites of Occupational and Environmental Medicine. 2009; in urine of e-waste recycling workers from 51(4): 435–440. DOI: https://doi.org/10.1097/ Agbogbloshie, Ghana. Sci Total Environ. 2014; JOM.0b013e31819b8c09 466–467: 369–76. DOI: https://doi.org/10.1016/j. 17. Burns KN, et al. Heart Rate, Stress, and Occupational scitotenv.2013.06.097 Noise Exposure among Electronic Waste Recycling 5. Asampong E, et al. Health seeking behaviours Workers. Int J Environ Res Public Health. 2016; 13(1). among electronic waste workers in Ghana. BMC DOI: https://doi.org/10.3390/ijerph13010140 Public Health. 2015; 15(1): 1065. DOI: https://doi. 18. Neitzel RL, Andersson M, Andersson E. Com- org/10.1186/s12889-015-2376-z parison of Multiple Measures of Noise Exposure in 6. Wittsiepe J, et al. Levels of polychlorinated Paper Mills. Ann Occup Hyg. 2016. DOI: https://doi. dibenzo-p-dioxins, dibenzofurans (PCDD/Fs) and org/10.1093/annhyg/mew001 biphenyls (PCBs) in blood of informal e-waste recy- 19. Neitzel RL, et al. Injury Risk and Noise Exposure cling workers from Agbogbloshie, Ghana, and con- in Firefighter Training Operations. Ann Occup Hyg. trols. Environ Int. 2015; 79: 65–73. DOI: https://doi. 2016; 60(4): 405–20. DOI: https://doi.org/10.1093/ org/10.1016/j.envint.2015.03.008 annhyg/mev088 7. Fuhriman DN. Dangerous Donations: Discarded 20. Tonne C, et al. Long-term traffic air and noise pollu- Electronics in Accra, Ghana. Pennsylvania State tion in relation to mortality andhospital readmission University; 2008. among myocardial infarction survivors. Interna- 8. Asante KA, et al. Multi-trace element levels and arse- tional Journal of Hygiene and Environmental Health. nic speciation in urine of e-waste recycling workers 2016; 219: 72–78. DOI: https://doi.org/10.1016/j. from Agbogbloshie, Accra in Ghana. Science of the ijheh.2015.09.003 Total Environment. 2012; 424(2012): 63–73. DOI: 21. Congdon P. Spatial variation in attributable https://doi.org/10.1016/j.scitotenv.2012.02.072 risks. Spatial and Spatio-temporal Epidemiology. 9. Amoyaw-Osei Y, et al. Ghana e-waste country 2015; 12: 39–52. DOI: https://doi.org/10.1016/j. assessment: SBC e-waste Africa project. sste.2015.02.002 GreenAd-Empa; 2011. 22. Cui J, Forsberg E. Mechanical recycling of waste, 10. Asampong E, et al. Health seeking behaviours electric and electronic equipment: a review. Journal among electronic waste workers in Ghana. BMC of Hazardous Materials. 2003; B99: 234–263. DOI: Public Health. 2015; 15: 1065. DOI: https://doi. https://doi.org/10.1016/S0304-3894(03)00061-X org/10.1186/s12889-015-2376-z 23. Julander A, et al. Formal recycling of e-waste leads 11. Srigboh RK, et al. Multiple elemental exposures to increased exposure to toxic metals: An occupa- amongst workers at the Agbogbloshie electronic tional exposure study from Sweden. Environment waste (e-waste) site in Ghana. Chemosphere. 2016; International. 2014; 73: 243–251. DOI: https://doi. 164: 68–74. DOI: https://doi.org/10.1016/j. org/10.1016/j.envint.2014.07.006 chemosphere.2016.08.089 24. Manhart A, et al. Informal e-waste management 12. Asante KA, et al. Human exposure to PCBs, PBDEs in Lagos, Nigeria-socio-economic impacts and and HBCDs in Ghana: Temporal variation, sources of feasibility of international recycling co-operations. exposure and estimation of daily intakes by infants. Öko-Institut e. 2011; 8–10. Adusei et al: Spatiality in Health Art. 31, page 11 of 12 25. Neitzel RL, et al. A mixed-methods evaluation of of Community Medicine. 2009; 35(3): 382–385. DOI: health and safety hazards at a scrap metal recycling https://doi.org/10.4103/0970-0218.69251 facility. Safety Science. 2013; 51: 432–440. DOI: 38. Leung A, Cai ZW, Wong MH. Environmental con- https://doi.org/10.1016/j.ssci.2012.08.012 tamination from electronic waste recycling at Guiyu, 26. Aucott M, McLinden M, Winka M. Release of mer- southeast China. Journal of Material Cycles and cury from broken fluorescent bulbs. Journal of the Waste Management. 2006; 8: 21–33. DOI: https:// Air and Waste Management Association. 2003; 53: doi.org/10.1007/s10163-005-0141-6 143–151. DOI: https://doi.org/10.1080/10473289. 39. Needhidasan S, Samuel M, Chidambaram R. 2003.10466132 Electronic waste – an emerging threat to the envi- 27. Basu N, et al. Multiple metals exposure in a small- ronment of urban India. Journal of Environmental scale artisanal gold mining community. Environmen- Health Science and Engineering. 2014; 12(1). DOI: tal Research. 2011; 111: 463–467. DOI: https://doi. https://doi.org/10.1186/2052-336X-12-36 org/10.1016/j.envres.2011.02.006 40. Fillion M, et al. A preliminary study of mercury 28. Chikarmane P, Narayan L, Chaturvedi B. Recy- exposure and blood pressure in the Brazilian cling Livelihoods: Integration of the informal recy- Amazon. Environmental Health. 2006; 5(29). DOI: cling sector in solid waste management in India. In https://doi.org/10.1186/1476-069X-5-29 Chintan Environmental Research and Action Group. 41. Moreira CM, et al. Effects of Mercury on myosin New Delhi: SNDT Womens’ University & Chintan ATPase in the ventricular myocardium of the rat. Environmental Research and Action Group; 2009 Comparative Biochemistry and Physiology. 2003; 29. Choi AL, et al. Methylmercury exposure and adverse 135: 269–275. DOI: https://doi.org/10.1016/ cardiovascular effects in Faroese whaling men. Envi- S1532-0456(03)00110-8 ronmental Health Perspectives. 2009; 117: 367–372. 42. Tamashiro H, et al. Methyl mercury toxicity in DOI: https://doi.org/10.1289/ehp.11608 spontaneously hypertensive rats (SHR). Bulletin 30. Huo X, et al. Elevated Blood Lead Levels of Chil- of Environmental Contamination and Toxicology. dren in Guiyu, an Electronic Waste Recycling Town 1986; 36: 668–673. DOI: https://doi.org/10.1007/ in China. Environmental Health Perspectives. 2011; BF01623567 115(7). DOI: https://doi.org/10.1289/ehp.9697 43. Patrick L. Mercury toxicity and antioxidants: part 31. Binion E, Gutberlet J. The effects of handling solid 1–role of glutathione and alpha-lipoic acid in the waste on the wellbeing of informal and organized treatment of mercury toxicity. Alternative Medical recyclers: a review of the literature. International Review. 2007; 7: 456–471. Journal of Occupational and Environmental Health. 44. Premalatha M, Tabassum-Abbasi TA, Abbasi 2012; 18(1): 43–52. DOI: https://doi.org/10.1179/1 SA. The Generation, Impact, and Management of 077352512Z.0000000001 E-Waste: State of the Art. Environmental Science and 32. Engkvist I-L. Working conditions at recycling cen- Technology. 2014; 44: 1577–1678. DOI: https://doi. tres in Sweden – physical and psychosocial work org/10.1080/10643389.2013.782171 environment. Application of Ergonomics. 2010; 45. Qiu B, et al. Medical investigation of e-waste deman- 41(3): 347–354. DOI: https://doi.org/10.1016/j. ufacturing industry in Guiyu town. In: Interna- apergo.2009.06.008 tional Conference on Electronic Waste and Extended 33. Perkins DN, et al. E-Waste: A Global Hazard. Lancet Producer Responsibility. Beijing: Greenpeace and Global Health; 2014. DOI: https://doi.org/10.1016/j. Chinese Society for Environmental Sciences; 2004. aogh.2014.10.001 46. Zeng X, et al. Solving e-waste problem using an 34. Eneh DC, Agunwamba JC. Managing hazardous integrated mobile recycling plant. Journal of Cleaner wastes in Africa: Recyclability of lead from e-waste Production. 2014; 90: 55–59. DOI: https://doi. materials. Release of mercury from broken fluores- org/10.1016/j.jclepro.2014.10.026 cent bulbs. Journal of the Air and Waste Manage- 47. Auchincloss AH, et al. A review of spatial meth- ment Association, Journal of Applied Sciences. 2011; ods in epidemiology, 2000–2010. Annu Rev Public 11: 3215–3220. DOI: https://doi.org/10.3923/ Health. 2012; 33: 107–22. DOI: https://doi. jas.2011.3215.3220 org/10.1146/annurev-publhealth-031811-124655 35. Sthiannopkao S, Wong MH. Handling e-waste 48. Beale L, et al. Methodologic Issues and Approaches in developed and developing countries: initiatives, to Spatial Epidemiology. Environmental Health Per- practices, and consequences. Science of the Total spectives. 2008; 116(8): 1105–1110. DOI: https:// Environment; 2012. DOI: https://doi.org/10.1016/j. doi.org/10.1289/ehp.10816 scitotenv.2012.06.088 49. Elliott P, Wartenberg D. Spatial epidemiology: 36. Azevedo BF, et al. Toxic Effects of Mercury on the Current approaches and future challenges. Envi- Cardiovascular and Central Nervous Systems. Jour- ron Health Perspect. 2004; 112(9): 998–1006. DOI: nal of Biomedicine and Biotechnology. 2012; 2012. https://doi.org/10.1289/ehp.6735 DOI: https://doi.org/10.1155/2012/949048 50. Fobil JN, et al. Mapping urban malaria and diarrhea 37. Kishore M, Kishore J. Waste Management: As a mortality in Accra, Ghana: evidence of vulner- Challenge to Public Health in India. Indian Journal abilities and implications for urban health policy. Art. 31, page 12 of 12 Adusei et al: Spatiality in Health J Urban Health. 2012; 89(6): 977–91. DOI: https:// Weighted Poisson Regression: Application to doi.org/10.1007/s11524-012-9702-x an ecological study. Spatial and Spatio-temporal 51. Bihrmanna K, Nielsena SS, Ersbøll AK. Spatial Epidemiology. 2016; 17: 1–13. DOI: https://doi. pattern in prevalence of paratuberculosis infection org/10.1016/j.sste.2016.02.001 diagnosed with misclassificationin Danish dairy 56. November V. Commentary. Environment and Plan- herds in 2009 and 2013. Spatial and Spatio-tempo- ning A. 2008; 40: 1523–1527. DOI: https://doi. ral Epidemiology. 2016; 16: 1–10. DOI: https://doi. org/10.1068/a4194 org/10.1016/j.sste.2015.10.001 57. Song Q, Li J. A review on human health conse- 52. Dedefoa M, Oljirab L, Assefa N. Small area cluster- quences of metals exposure to e-waste in China. ing of under-five children’s mortality and associated Environmental Pollution. 2014; 196: 450–461. DOI: factors using geo-additive Bayesian discrete-time https://doi.org/10.1016/j.envpol.2014.11.004 survival model in Kersa HDSS, Ethiopia. Spatial and 58. Pearce JL, et al. Characterizing the spatial distri- Spatio-temporal Epidemiology. 2016; 16: 43–49. bution of multiple pollutants and populations at DOI: https://doi.org/10.1016/j.sste.2015.11.003 risk in atlanta, georgia. Spatial and Spatio-tempo- 53. Kazembea LN, Kamndaya MS. Hierarchical spatial ral Epidemiology. 2016; 2016. DOI: https://doi. modelling of pneumonia prevalence when response org/10.1016/j.sste.2016.02.002 outcome has misclassification error: Applications 59. Chitunhu S, Musenge E. Spatial and socio-eco- to household data from Malawi. Spatial and Spa- nomic effects on malaria morbidity in children tio-temporal Epidemiology. 2016; 16: 35–42. DOI: under 5 years in Malawi in 2012. Spatial and https://doi.org/10.1016/j.sste.2015.11.002 Spatio-temporal Epidemiology. 2016; 16: 21–33. DOI: 54. Fu J, et al. Spatial distribution of polychlorinated https://doi.org/10.1016/j.sste.2015.11.001 biphenyls (PCBs) and polybrominated biphenyl ethers 60. Huang G, Lee D, Scott M. An integrated Bayesian (PBDEs) in an e-waste dismantling region in South- model for estimating the long-term health effects east China: Use of apple snail (Ampullariidae) as a of air pollution by fusing modelled and measured bioindicator. Chemosphere. 2011; 82: 648–655. DOI: pollution data: A case study of nitrogen dioxide con- https://doi.org/10.1016/j.chemosphere.2010.11.014 centrations in Scotland. Spatial and Spatio-temporal 55. Ribeiro MC, Sousa AJ, Pereira MJ. A coregionaliza- Epidemiology. 2015; 14–15: 63–74. DOI: https://doi. tion model can assist specification of Geographically org/10.1016/j.sste.2015.09.002 How to cite this article: Adusei A, Arko-Mensah J, Dzodzomenyo M, Stephens J, Amoabeng A, Waldschmidt S, Löhndorf K, Agbeko K, Takyi S, Kwarteng L, Acquah A, Botwe P, Tettey P, Kaifie A, Felten M, Kraus T, Küpper T, Fobil J. Spatiality in Health: The Distribution of Health Conditions Associated with Electronic Waste Processing Activities at Agbogbloshie, Accra. Annals of Global Health. 2020; 86(1): 31, 1–12. DOI: https://doi.org/10.5334/aogh.2630 Published: 18 March 2020 Copyright: © 2020 The Author(s). This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0 International License (CC-BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. See http://creativecommons.org/licenses/by/4.0/. Annals of Global Health is a peer-reviewed open access journal published by Ubiquity Press. OPEN ACCESS