USVI Public Records

Document text

e A Study of Environmental Health of the Industrial Area of St Croix University of the Virgin Islands Eastern Caribbean Center September, 2014 Acknowledgments The Eastern Caribbean Center of the University of the Virgin Islands would like to thank the following persons for their involvement in A study of Environmental Health of the Industrial Area of St Croix. We especially wish to recognize the support of Delegate Donna Christensen, who in August 2011 sought the assistance of the Eastern Caribbean Center in responding to residents’ request for a study of reported ailments in the vicinity of the industrial zone. We would like to thank Ms Myrtle Peters for her supervision that maintained quality assurance and efficiency during the data collection phase. We also recognize the field supervisor and interviewers’ professionalism, fortitude and enthusiasm that facilitated a timely completion of the data collection phase. In addition, we would like to express thanks to Ms Marissa Johnson Rogers and Ms Grisha Fleming, ECC administrative staff members, the former for her organizational proficiency in managing this project, and the latter for the design of the cover. Additionally, special thanks go to the Research Team of ECC, Ms Ayishih Bellew, Ms Corene Jn-Charles, Ms Sadio Thomas and Ms Deshona Williams whose dedicated teamwork from the beginning to the end of this project facilitated the successful production of this report. We would also like to thank our GIS Analyst, Mr Stevie Henry, for his skillful application of Geographic Information Systems technology in mapping the study areas for data collection, identifying the distinct geographic areas for field work and for preparing publication maps for this study. Most importantly, the Eastern Caribbean Center would like to express gratitude to all of the residents of St Croix who took time out of their busy schedules to participate in this important survey. Thank You! Frank L. Mills Interim Vice Provost for Research and Public Service September 2014 Asha DeGannes Acting Director, ECC September 2014 Eastern Caribbean Center University of the Virgin Islands Frank L. Mills, PhD Interim Vice Provost for Research and Public Service Asha DeGannes, PhD Acting Director Stevie Henry Ayishih Bellew Geographic Information Research Analyst III Systems Analyst Corene Jn-Charles Sadio Thomas Research Analyst I Research Analyst I iv Table of Contents Acknowledgements ....................................................................................... i Executive Summary ..................................................................................... vi I. INTRODUCTION ......................................................................................... 1 A. Background of the Environmental Health Study ................................. 1 B. Purpose of the Study .......................................................................... 3 II. METHODS .................................................................................................. 4 A. Selection of the Study Areas .............................................................. 4 B. Sampling Design ................................................................................. 7 C. IRT and the Survey Instrument ........................................................... 9 III. FIELD DATA COLLECTION ......................................................................... 14 A. The Environmental Health Study Questionnaire .............................. 14 B. Training and Data Collection ............................................................ 16 IV. ANALYSIS OF DATA ................................................................................ 17 A. Demographic Characteristics of Respondents ................................... 17 B. Environmental Health Impacts in the Target and Reference Areas .... 18 C. Environmental Health Impacts in the Target Zone ............................. 20 D. Environmental Health Impacts in the Reference Zone....................... 25 E. Diseases in the Target and Reference Zones ...................................... 29 F. Inferences of Field Data Collection .................................................... 31 (1) Differences in the level of impact of environmental factors ……31 (2) Relationship between health impacts and two independent variables..………………………………………………………………………………....34 v V. SUMMARY ............................................................................................ .. 37 VI. REFERENCES ............................................................................................. 39 Appendix I Survey Instrument Appendix II Frequency and Percentage Tables vi EXECUTIVE SUMMARY DEMOGRAPHIC CHARACTERISTICS  There were more females living near the South Shore Industrial Zone (SSIZ) (58%) and the reference area (53%) than males—42% and 47%, respectively.  In the target area, 16% of the children were under 18 years compared to 12% in the reference (or control) area.  Thirty-four percent (34%) of the residents who were sampled and who lived in the vicinity of the SSIZ were 55 years and over, and were proportionately smaller than the 40% of the same age in the control area.  The disparity by race and ethnicity is evident in that proportionately more Blacks (74%) live closer to the industrial area than in the reference area (66%), and similarly, a smaller percentage of Whites (2%) live near to the target area than in the reference area (9%).  The pattern of Latino residency is similar to that of Blacks: a larger percentage live within the shadow of the industrial zone (34%) than at a distance from the zone (31%).  In comparing the two areas under study, the less educated—41% with less than high school diploma—tend to live near the SSIZ compared to 28% in the reference area, and the proportion with college degrees and above was 15% in comparison with 21% in the reference section. ENVIRONMENTAL IMPACTS IN THE TARGET AND REFERENCE ZONES  Residents in the target area expressed the frequent the presence of, or reported higher frequencies of the occurrence of, (1) strong and irritating odors, (2) mold, (3) transportation emissions and exhaust and (4) factory smoke.  There is indication that the residents of the reference area do not consider themselves to be impacted to the same intensity as residents in the target zone.  The order of the impacting environmental conditions is not the same for both zones, indicating that the residents who live at a distance from the industrial zone do not experience the same level of morbidity as those who live in proximity to the industrial zone. vii  The order of impact of environmental conditions on residents near the south shore area are strong and irritating odors, followed by factory smoke and then by mold, in the control area the order is mold, followed by strong and irritating odors, followed by transportation emissions and exhaust.  Red mud dust is ranked fourth most by those who live near the industrial zone. However, it ranks very low in respect of its impact on the health of those who live at a distance from the industrial zone.  The target area is least impacted by industry drainage and runoff, incorrect disposal of chemical waste and improper sewage disposal; in the reference zone the lowest impact is imparted by incorrect disposal of chemical waste, industry drainage and runoff and red mud dust. INFERENCES FROM THE FIELD DATA COLLECTION  A difference-of-means test was conducted to reveal any statistically significant difference between the means of the groups concerning their perceptions of the impact of environmental conditions.  The mean measure for the target group was higher than the mean measure for the reference group.  The difference-of-means test concluded that there was a statistically significant difference between the target and reference mean measures.  One can conclude that persons within the target zone were much more likely, than those in the reference zone, to report that their households were impacted by negative environmental conditions. DISEASES IN THE TARGET AND REFERNCE ZONES  About 1 in every 5 persons (21.3%) in the target area reported they had experienced asthmatic conditions within the last five years, and this compares with 1 in 8 (12.5%) in the reference area.  There were more than twice as many residents near the industrial zone who indicate having chronic bronchitis—10.4%—compared to residents who live in the reference area—4.1%. I INTRODUCTION A. Background to the Environmental Health Study Around the world, scientists have searched for confirmation that pollution is harmful to life on earth. Indeed, many studies have provided evidence to support the above stated hypothesis, but opposing studies have stated otherwise. Environmental factors, especially pollutants in the air, water and soil, adversely affect the health of four to five billion persons worldwide annually. A number of specific research projects in various geographic locations around the world provide preliminary evidence of the relationship that is posited between a polluted environment and morbidity. In the United States itself, there is an implicit admission that the basic information about the health of Americans and the environment must be made available before the nation could take advantage of the new knowledge of the links between genetic predisposition and such factors as exposure to pollutants in the environment. Research has shown that long-term exposure to air pollution is linked to an increased risk of chronic respiratory illnesses (Nuvolone et al., 2011). In addition, studies of short-term exposures to high concentrations of air pollution have been associated with higher rates of asthma, bronchitis and other respiratory symptoms (Nuvolone et al., 2011). 2 Within the US Virgin Islands, there is widespread speculation that environmental pollutants on St Croix, mainly of industrial sources, have adversely affected the health of many of the island’s residents. For decades, residents have hypothesized that living and working in the South Shore Industrial Zone (SSIZ) have caused many to become afflicted with several respiratory diseases. In the South Shore area of St Croix, a noticeably unpleasant smell―similar to odors present in many urbanized, industrial areas of the mainland US―has intermittently plagued residents over many years. In February 2011, the Environmental Protection Agency (EPA) began a study of air pollution on St Croix. Results, released on August 18, 2011, indicated that concentrations of specific air pollutants measured at the three locations were below levels of concern, associated with health problems that result from short- or long-term exposure to toxins (Coward, 2011). Specifically, two of the 60 different compounds, benzene and 1,3 butadiene, measured by the monitoring devices were shown to appear at lower levels of health concern. The findings of the EPA were encouraging, but rest assured, that short- term study did not quell the complaints of health personnel concerning the potential impact of air pollution on health conditions in St Croix. Since May 2011, several incidences have occurred at Central High School, in which a foul odor sickened students and caused some to be sent some to the emergency room. In fact, Central High School was found to be uninhabitable and was closed on March 18, 2014 and did not reopen for the remainder of the school year. This was a strong indicator that more research on air pollution and its effects was essential on St Croix and the entire Territory. 3 B. Purpose of the Study The deep-seated concerns of many Crucian residents within a one-mile radius of the industrial south-shore complex appear justified from a number of events prior to, and within, the current year. As a response to the expressed anxieties of residents of St Croix, the Delegate to Congress for the US Virgin Islands, Dr Donna Christensen, invited the Eastern Caribbean Center (ECC) of the University of the Virgin Islands (UVI) to initiate a study on the impact of environmental factors on health conditions in the SSIZ of St Croix. Upon examination of the necessary costs to conduct the study, researchers at ECC applied for, and received, funding for a Technical Assistance Grant from the Office of Insular Affairs of the US Department of the Interior. The proposed study aimed to seek out or to collect empirical evidence in relation to the belief that living in certain areas of St Croix contributed to the prevalence of health risk factors. In other words, researchers wanted to identify if a relationship exists between environmental conditions in and around the South Shore Industrial Zone and reports of adverse health conditions. The original plan included a comparison of environmental and health conditions of the SSIZ and an area of St Thomas. Upon examination of the contrasting socio- economic differences between St Thomas and St Croix, researchers decided to choose a control (or reference) area on St Croix. Indeed, this study addresses the apprehension of many residents of St Croix, and aspires to produce answers that may lead to solutions toward better environmental conditions and healthier surroundings. 4 II METHODS A. Selection of the Study Areas It was explained in the previous chapter that the original intent was to conduct a scientific sample survey in the marginal areas of the South Shore Industrial Zone (SSIZ) in St Croix, as well as in a geographic area of roughly similar size in St Thomas that would serve as a control. It was further explained that demographic characteristics appeared different enough to suggest the confinement of the study areas to St Croix only. The target and reference zones for this study were based on the location of estates in relation to the SSIZ. The SSIZ is bounded on the north along Route 68 from the intersection with Route 62 (Camacho’s Wholesale - Humbug Road) on the east. The SSIZ continues west until it meets the intersection with Route 681, then runs north until it reaches the Melvin H. Evans Highway (MHE Highway - Route 66) intersection at Sunny Isle. At this intersection, the zone turns west and runs along the MHE Highway’s west-bound lane until it intersects with the West Airport Road (Route 64). The western boundary continues south along the West Airport Road and intersects with the western boundary of Estate Betty’s Hope and continues in a straight line to the island’s southern shoreline. At the end of the SSIZ, the southern boundary runs east along the island’s southern shoreline and ends at the Estate Cane Garden parcel that intersects with the bend of Route 62, then continues north until it meets the intersection with Route 68, the northeastern extent of the SSIZ. 5 The Virgin Islands Zoning Code has designated two zoning districts for industrial land-use activities: I-1 Industry—Heavy, and I-2 Industry—Light. The total acreage on St. Croix with I-1 and I-2 designation equals approximately 3,861.47 acres. Ninety-seven percent (3,583.83 acres) of the acreage designated I-1 or I-2 is in the SSIZ. All of the estates that are within a one-mile buffer of the SSIZ comprise a part of the target zone of this study. Air quality monitoring stations identified used in previous EPA studies are also within the target zone boundaries. Figure 2.1 below shows a map of the central portion of St Croix that includes the industrial zone in the south central part of the island. The primary occupants of the SSIZ are Hovensa, LLC; the Diageo USVI, Inc. rum distillery; the St Croix Renaissance Group; the Anguilla Landfill; and the now-closed Harvey/ALCOA Alumina plant and its piles of residual red dust. An eight-year (August 2003 to June 2011) wind rose shows that the prevailing winds are from the east, varying from northeast to southeast. This strongly implies that residential estates in the zone defined by winds blowing from the northeast and southeast are very likely to impact the environment with airborne particulates from within industrial zone. The map also shows the demarcation of a geographic area that is defined by a perimeter of one mile around the industrial zone. The residential estates—among others—in the immediate vicinity of the industrial zone that were sampled include Barren Spot, Bethlehem Old Works, Clifton Hill, Diamond East, Mount Pleasant West, Profit, Ruby, Sion Farm, Strawberry Hill and Williams Delight (Figure 2.2). In the periphery of the 6 industrial zone, 833 housing units (HUs) were selected probabilistically from among all housing units in the 29 estates in this classification. Figure 2.1 The Geographic Target and Reference Areas of St Croix During the geographical conceptualization of the study design, it was decided that a control or reference area roughly similar in size in terms of the units to be selected was to be located in St Thomas. Upon further reflection, this plan gave way to the selection of a control area in St Croix that was distant enough from the industrial zone and in an upwind location. Care was observed to include residential areas that reflected essentially the same characteristics of 7 those estates on the periphery of the industrial zone. In this reference area, 690 HUs were randomly selected for interviewing. The geographic areas that contributed most units to the sampled estates include Altona, Beeston Hill, Glynn, Mary’s Fancy, Mon Bijou South, Mt Welcome, Rattan, Sion Hill, St Peters, and Work and Rest. B. Sampling Design The primary consideration in the choice of a sample size is that it should provide sufficient information to meet the objectives of the survey, i.e., to make valid inferences from the survey data. The specific survey sample design is systematic sampling, in which, like simple random sampling, every unit of the population has an equal probability of selection. This feature is particularly useful in situations in which population elements are available in a list form, as it is relatively easy to select every kth element after a random start. Before the sampled units were drawn in the two geographic areas of interest—each of which comprised several estates—all housing units in the designated areas were mapped with the use of Garmin GPS hand-held units. This procedure was the preferred mode of listing as it was uniquely precise in the field location of any HU, thus making it enormously easy to relocate any selected HU during the actual survey. This factor is of great value in an island where streets and roads without names are the norm. Field workers were trained in the use of these hand-held units and in the methods of geographic mapping, and were required to canvass all streets and roads within assigned 8 Figure 2.2 The Target Area in Proximity to the Industrial Zone areas. Every occupied and vacant livable structure in the two designated areas was recorded in the GPS unit with its latitude and longitude that is unique to that structure alone. When one structure housed multiple residential units, each unit would have the identical latitude and longitude designation, but would be differentiated in the master list on separate lines of the list. Once the field listings of all estates in the study areas were complete, their contents were dumped into another file at the main office that would allow the probabilistic selection of a specified number of HUs. Based on sampling 9 theory, it was decided that nt = 833 randomly selected units would provide adequate estimates with a margin of error of 4% in the target area adjacent to the SSIZ. Systematic sampling methods were also applied to the list of the total number of HUs in the target zone, and in this fashion the 833 units were marked for interviews. In a similar way, each HU in the reference area was listed with its latitude and longitude, from which nr = 690 HUs were selected in a systematic random way. The desired margin of error for estimates in the reference area was also 4%. The training period for the interviewers was followed by the data collection phase, in which interviewers were equipped with a list of the housing units in which they were expected to conduct interviews, except for selected HUs that were vacant at the time of the visit. These field workers used their Garmin GPS units to quickly negotiate their way to the randomly selected HUs on the list, and then attempted to conduct an interview in the occupied dwelling. C. IRT and the Survey Instrument The development of a framework is essential precondition prior to the construction of an effective instrument. It is fundamental because it is the step that is necessary to establish clearly in the researcher’s mind what it is that one is trying to measure. In this context, the term construct refers to the various attributes of the environment and their individual or collective impact on the health of the residents of the peripheral industrial zone, and which are of interest to be measured. It is therefore defined as a single characteristic, attribute, or dimension that is assumed to underlie a set of items. Latent trait is also used synonymously with construct. One can regard environmental health 10 as a latent trait because it is not directly visible, and is a concept that can take on ratings from low to high. One primary intent of the survey was to measure the perceived environmental effect on the health of residents in the vicinity of the industrial zone, and to assess the degree of self-identified morbidity among residents near the zone. This construct is herein referred to as Morbidity in the Vicinity of the Industrial Zone. Since one cannot directly measure a construct, items were devised to tap into the construct. As long as the items were directly related to what it is desirous of being measured, the construct is likely to be unidimensional. This is a desirable psychometric property for an instrument, for it ensures that there is only one single dimension in the construct of interest, and the level of this construct is the focus of the measurement. The method of measurement departs significantly from the Classical Test Theory method of treating ordinal categorical values as if they were linear interval measures. Instead, it makes use of Item Response Theory, from which a metric of desirable measures derived and on which parametrical statistics analysis can be conducted1. Considerable thought was given to the core items that constituted the survey instrument. Since the unidimensionality of the construct Morbidity in the Vicinity of the Industrial Zone demanded that the items be all directly related to the core dimension of impact on health conditions, a number of questions were adopted from previously tested questions in other national and international surveys, and adapted to the specific needs of this survey (see Bhopal et al., Toren et al., Najjar, and Nuvolone). 1Item Response Theory (IRT) has rapidly become mainstream as the theoretical basis for measurement …. due to the more theoretically justifiable measurement principles and the greater potential to solve practical measurement problems.” (Embretson, S.E., & Reise, S.P. (2000). 11 The principal advantages of the Likert format in the standard questionnaire are that it is familiar, that it is relatively easy for respondents to answer well- formulated questions, and that it often does not consume a lot of time (refer to Chapter 3 for a detailed description of the questionnaire). However, because the responses produce raw-score data at an ordinal level of measurement, the traditional practice of transforming the scores into percentages for ordering purposes may be misleading due to the fact that the distances between the percent scores do not have direct meaning. (For example, the difference between 90 and 95 percent satisfaction on two items does not represent the same intensity of satisfaction as that between two other items at 75 and 80 percent.) It is therefore highly inappropriate to compute averages and standard deviations on these ordinal data, as the results can be rather deceptive. The Rasch Rating Scale Model (RSM) was applied to the survey data to transform them from an ordinal into an interval scale with the measurement property of equal units upon which standard arithmetic operations can be performed. The values of the items in Sections A and B in the survey instrument were rescaled from logits to range from 0 to 100—although this latter scale is not a percent scale—in order to produce a meaningful scale for score interpretation and parametric data analysis2. The smaller the number on the transformed scale—i.e., the closer to 0 of the lowest item measure or item calibration—the easier it was for the respondent to endorse the frequency of the impact of environmental effects on the respondent’s health, and the greater morbidity among residents. On the other hand, the bigger the number—i.e., the closer to 100 is the highest item measure—the more easy it was for the respondent to endorse the item as never having, or only scarcely having, an 2The logit is a unit of measurement that results when the Rasch Rating Scale Model is used to transform raw scores obtained from ordinal data to log-odds ratios on a common interval scale. 12 impact on their health conditions. Consequently, the lower the level of the environmental factor on the scale, the bigger the impact on residents in the zone, and the higher the item on the continuum, the smaller was the extent of the effect on the residents. While the percent distribution of each item provides basic information of the raw scores, the counts themselves are not able to differentiate between the levels of morbidity. Moreover, percentages cannot be treated as if they are on an interval scale because the distances between them have no direct meaning. All that one can infer is that there is an ordering of the percent values for the items, but one cannot say with any certainty what the size of the gap is. By transforming the raw scores into an interval scale the reader is able to readily determine which items are reflective of a high degree of impact, and which ones are reflective of low degrees of influence. Advantages in the use of the method applied here are that with interval-scale measures, immediate responses are available for the questions: What is the prevalence rate of morbidity among residents in the proximity of the industrial zone compared to residents who live in a selected area further afield? Is there a significant difference between the mean measure of the environmental impact in the industrial zone compared to the reference area? What does a hierarchical ranking of the intensity of the environmental impacts say about residency and health in the two geographic areas? How can the logit3 calibrations of the items along the morbidity continuum be used for direct comparison with measures among those who experience inhalants at work, or those who work with the industrial zone? 3A logit is a pure, abstract, interval-level unit of measurement that results when the Rasch model transforms ordinal raw scores to a common metric. 13 Additional results that are forthcoming from the application of the model include: data on an hierarchical ranking of the level of impact, a measure that helps to locate each environmental item on the unidimensional continuum of the morbidity construct, a measure of the reliability of each statement that is used in the instrument, and a measure of the validity of each question or statement. (The terms statement, item and question are used interchangeably in this report.) Perhaps the model’s most substantive feature is that the measures that are produced are all at an interval level, thus enabling the construction of an objective hierarchy of all the statements. They also permit one to say by how much one environmental condition is ranked above, or below, another. Another valuable feature of the RSM analysis is that unlike the traditional method of comparing raw-score means and standard deviations from different surveys in different locations—which is inappropriate—measures derived from the application of the RSM can be compared statistically for the significance of differences. For example, it is possible to compare the mean of the measures from the target area with that from the reference area. 14 III FIELD DATA COLLECTION A. The Environmental Health Study Questionnaire The survey instrument for this research study was developed by researchers at the Eastern Caribbean Center of the University of the Virgin Islands. The final instrument consisted of four sections: (1) environmental incidences, (2) adverse health conditions, (3) exposure to smoke and harmful inhalants and (4) demographic questions (see Appendix II). The first section required participants to reveal their perceptions of the impact of eleven (11) environmental incidences within their households in the last five years by responding Never, Scarcely, Sometimes or Always. Those environmental incidences included red mud dust, transportation emissions and exhaust, factory smoke, mold, incorrect disposal of waste, industry drainage and runoff, improper sewage disposal, contaminated drinking water, strong and irritating odors, landfill odor and other materials in the cistern. The second section asked participants to indicate if someone in their household suffered from ten (10) health conditions within the last five years by responding Yes or No. If the response was Yes, participants were asked to indicate the number of household members who suffered from the particular health condition. The health conditions included asthma, chronic bronchitis, lung cancer, emphysema, chronic obstructive pulmonary disease (COPD), 15 allergies, black lung disease, acute respiratory distress syndrome (ARDS), airborne substances, and mesothelioma. The third section of the questionnaire asked about exposure to smoking and harmful inhalants at work. The three (3) questions in this section inquired if the respondent smoked within the last five years and, if yes, the frequency of the smoking (e.g. Daily, Once a month, 1-2 times a year, Never). The same question was asked about anyone in the household, in order to reveal exposure to second-hand smoke. The third question in this section asked if anyone in the household was consistently exposed to harmful inhalants at work. The final section asked a series of demographic questions, including the respondent’s age, ethnic background, income, etc. Due to the fact that only one person in the household was responding to the survey, a method of selecting the person for whom the demographic questions would refer to was utilized. Specifically, if only one person lived in the household, that person’s information was recorded in the demographic section. However, if more than one person lived in the household, the following process was used to select the person for whom the demographic questions would refer. If one (1) person in the housing unit suffered from a condition mentioned in Section B, that person’s demographic information was recorded. If more than one (1) person in the housing unit suffered from a condition, the person (with a condition) with the most recent birthday was selected for demographic responses. If no one in the HU has suffered from a condition, the person with the most recent birthday was selected for responses in the demographic section. 16 B. Training and Data Collection Upon selection of the sample, a training of twenty-four (24) field interviewers and two (2) field supervisors took place on Saturday, March 29, 2014. The majority of the field interviewers and both of the supervisors were experienced survey interviewers who previously worked with the Eastern Caribbean Center on the annual Virgin Islands Community Survey and decennial Census. During the training, interviewers carefully reviewed the survey instrument and interviewing techniques. Additionally, the GIS analyst facilitated a three-hour segment of the training on finding selected housing units. During this segment of the training, the GIS analyst reviewed proper map use, functions of the GPS unit and held an exercise in the field. Data collection began on April 4, 2014 and concluded on May 2, 2014. Field interviewers reported that residents in selected households were very cooperative and eager to complete the survey interview. The willingness of participation among residents demonstrated the sincere concern regarding environmental conditions and their impact on health conditions on St Croix. 17 IV ANALYSIS OF DATA A. Demographic Characteristics of Respondents The survey instrument that was used in the field appears in Appendix II. Section D houses the demographic questions that were posed to the respondents. In each randomly selected housing unit (HU) in both of the two areas under study, only one person in the unit was selected for the interview. If one occupant of the HU had experienced any of the morbidities listed, that person’s demographic information was recorded. In situations in which more than one occupant suffered one of the listed conditions, the one with the most recent birthday was selected for interview. This approach introduced a degree of randomness that minimized bias in the source of the information. If no one in the household had suffered a disease or illness, the occupant with the most recent birthday was selected. Of those for which demographic information was collected, a total of 306 persons in the target district and 175 in the reference district were classified as having suffered from one of the listed conditions. Tables 27-34 in Appendix I show the demographic characteristics of the persons that suffered from any of the illnesses. Proportionately, there were more females in the target area (58%) and the reference area (53%) than males—42% and 47%, respectively. In the target area, 16% of the children were under 18 years compared to 12% in the reference area. However, 34% of the residents who were sampled and who lived in the vicinity of the SSIZ were 18 55 years and over, and were proportionately smaller than the 40% of the same age in the control area. The disparity by race and ethnicity is evident in that proportionately more blacks (74%) live closer to the industrial area than in the reference area (66%), and similarly, a smaller percentage of whites (2%) live near to the target area than in the control area (9%). The pattern of Latino residency is similar to that of blacks: a larger percentage live within the shadow of the industrial zone (34%) than at a distance from the zone (31%). In comparing the two areas under study, the less educated—41% with less than high school diploma—tend to live near the SSIZ compared to 28% in the reference area, and the proportion with college degrees and above was 15% in comparison with 21% in the reference section. While interviewees asserted that they spent virtually the same number of hours, i.e., 12 hours of more, in their neighborhoods—76.5% and 76.0% in the target and reference sections, respectively—13% of respondents who lived in the reference district worked in the industrialized belt, compared to 7% who lived nearby and worked in the SSIZ. B. Environmental Health Impacts in the Target and Reference Areas Health is often regarded as the most highly valued human asset. Hence any persistent threat to that desirable state, be it environmental or otherwise, tends to arouse strong reactions in potential victims. Morbidity, disease and poor health derived from environmental conditions are of no less concern to residents who believe they are negatively impacted, for these afflictions not only affect 19 the quality of life, but also life expectancy. This section presents the reaction of residents to perceived threats to their health, and this is captured in Tables 1 to 11 in Appendix I. The responses were provided to one of four response categories in each of 11 statements—Never, Scarcely, Sometimes and Always. The assumption is made that when a household occupant responded Never to a statement in Section A of the questionnaire, the intent was to indicate that there is no impact that has ever been felt or perceived from the particular morbidity. Similarly, a response of Always demonstrated the highest level of reaction to the health stimulus, and that Scarcely reflected some degree of impact, but less than Sometimes. There is thus a gradient of effect in health condition from Never to Always. The first effort here then is to combine, for each item, all the responses that could be labeled as showing some level of impact by the environmental condition, hence providing evidence of the level of illness. The impactive conditions of health to which occupants responded included the following: red mud dust; transportation emissions and exhaust; factory smoke (with fine, suspended particulate matter); mold; incorrect disposal of chemical waste; industry drainage and runoff; improper sewage disposal; contaminated drinking water; strong and irritating odors; landfill odor; and other materials in the cistern (i.e., things other than water). 20 C. Environmental Health Impacts in the Target Zone One source asserts that “raw scores have special conditions, restrictions, and assumptions which make them very treacherous to analyze and too ambiguous for the solution of contemporary measurement problems.”4 In order for the arithmetic of statistical analysis to be useful, it must be done with equal interval, constant-unit, linear measures. To this end, the Rasch Rating Scale Model was utilized to construct linear measures from the concrete raw response data for which counts are the media. These measures are also required for the derivation of sound, reliable and valid inferences about the group from which the data are collected. Table 4.1 presents measures for the 11 items of environmental impact that were implicitly rated by the randomly selected residents in the target zone. Column (3) shows the scaled measures corresponding to the serial number of the items on the instrument (in column (1)), with a label of each item given in column (2). The scaled measures in column (3) illustrate that the items in their transformed state from their raw scores are in a range from 0 to 100, with a mean of 49.8. The measures are ordered in size from the largest at the top to the smallest at the bottom, and these collectively create the construct or latent trait—Morbidity in the Vicinity of the Industrial Zone—described in II.B above. The scaled values are interpreted to mean that the closer they are to the bottom of the column, the greater the endorsability of the residents of those environmental factors that impact them. Similarly, the closer the item is to the top of the column, the lower is the endorsability of the residents on those 4Bezruczko, N. & Linacre, J.M. (2005). Measurement theory foundations. In N. Bezruczko (Ed.). Rasch measurement in health sciences (8- 34). Maple Grove, MN: JAM Press. 21 environmental conditions that hardly, or do not, impact them. Column (5), with the item calibrations in logits, presents the same information as in column (3) on another scale. A hierarchical ranking of the items is provided in column (6), from the environmental condition with the least effect at the top to the most numerously cited impacting condition at the bottom. Column (4) provides a measurement index of reliability for each item. All of the items have acceptable infit mean square values—not shown in the table—between -0.5 and +0.5. These fit Table 4.1. Measures and Rank of Environmental Impacts in the Target Zone Item No. Item Measure S.E. Measure (in logits) Rank (Least impact to most impact) (1) (2) (3) (4) (5) (6) 5 Incorrect disposal of chemical waste 66.0 1.44 1.48 1 6 Industry drainage and runoff 65.1 1.39 1.39 2 7 Improper sewage disposal 60.8 1.19 1.00 3 10 Landfill odor 54.2 0.96 0.40 4 11 Other materials in the cistern 52.9 0.93 0.29 5 8 Contaminated drinking water 47.6 0.80 -0.20 6 1 Red mud dust 45.3 0.76 -0.41 7 3 Factory smoke 42.8 0.72 -0.63 8 2 Transportation emissions and exhaust 41.3 0.70 -0.77 9 4 Mold 38.1 0.67 -1.07 10 9 Strong and irritating odors 33.5 0.63 -1.48 11 Mean 49.8 0.93 0.00 S.D. 10.4 0.28 0.95 Reliability 0.99 Note: The numbers in col. (1) refer to the serial numbers of the items in the survey instrument (see Appendix II). The measures in col. (3) are rescaled to range from 0 to 100, and are equivalent to the logit measures in col. (5). 22 values impart whether an item is compatible, or not compatible, with the model. Acceptable fit values lie between 0.5 and 1.5 logits. From column (3), one observes that item 9—Strong and irritating odors— with a transformed scale value of 33.5, has the smallest measure of all. This signifies that the item in relation to occasions of pungent and foul odors was the easiest for residents to endorse as having an effect on their health. Similarly, item 4 with a measure of 38.1—Mold—was the second most frequently endorsed item affecting their health. The third easiest item that residents endorsed in relation to its effect on their wellness was Transportation emissions and exhaust. At the other end of column (3), starting at the top, the largest item calibration—item #5 on Incorrect disposal of chemical waste—has a value of 66.0, and was concomitantly identified by the residents in the vicinity of the industrial zone as the least of the set of conditions that impact their health. This item was ranked last. Table 5 in Appendix I confirms that less than 1 in 10 (or 9.4%) respondents identified this item as having an impact on their health. This would seem logical, given that any disposal of waste would be limited in geographical scope. The condition with the second least impact is that of item 6 with a measure of 65.1—Industry drainage and runoff. This is not surprising relative to the similarity in type to the previous item. About 92% of those responding have never been affected by this environmental condition. The item that speaks to Improper sewage disposal, with a measure of 60.8, ranks the third lowest in terms of impact on residents adjacent to the SSIZ. The ranking of all of the other items in between these mentioned may be read from the table with the help of the last column. 23 Figure 4.1 presents a graphic illustration of the information in columns (1), (2), (3) and (6) in Table 4.1. It maintains the same rank order of the items as in the table, with the items that elicited the most frequent impact on health at the bottom of the graphic, and the least frequent form of ill effects at the top. It allows the reader not only a quick view of the location of all of the items on the health impact continuum, but it also permits a comparison of the relative gap between any pair of items. The map also shows the distribution of respondents on the left of the vertical line and of the items on a common logit metric. The values that are listed on the left margin derive from the logit scale that has been transformed to range between 0 and 100. The mean of the items is generally set to 0.0 logits, which on the transformed scale is 49.8, and for the residents the mean is 28.2. The means of the distributions of the respondents and the items are shown with the letter M on each side of the vertical line, and S represents ±1 standard deviation while the T represents two standard deviations. A valuable characteristic of the map is the illustrated vertical distance between the items. It is not only useful to decision makers to know that one item has a higher impact on residents than another, but by how much may be even more important. This is a valuable property of the logit scale of the measures that is not inherent in a comparison of percentages that is derived by adding and dividing raw scores. Since the transformed scale is also an interval scale, the equal distances anywhere up and down the map scale are of equal size. 24 Figure 4.1 Section A Item-Respondent Map of Residents Near the Industrial Zone Measure Respondents| Items <least impact> <less frequent impact on health> 80 + | . | | | | | . | |T 70 + | | | . | 5 Disposed chem waste 6 Industry drainage & runoff | | | 7 Improper sewage disposal | 60 . +S | | . | | . | 11 Materials in cistern T| 10 Landfill odor . | | 50 . +M | # | 8 Contaminated drinking water | .# | | .# | 1 Red mud dust 2 Transportation emissions & exhaust S| 4 Mold 40 . + 3 Factory smoke .## |S | ### | | .###### | | 9 Strong irritating odors | .#### | 30 M+ |T . | .##### | | | . | .###### | | 20 + S| | | . | | .###### | | | 10 .############ + <most impacted> <more frequent impact on health> Each "#" is 12: Each"." is 1 to 11 25 D. Environmental Health Impacts in the Reference Zone In experimental research, the control or reference group, for the sake of comparison, does not receive the treatment the experimenter is interested in studying. The treatment in this case is exposure to the immediate environment of the industrial zone on the south shore. The location of reference zone was therefore deliberate in the selection of an area that was demographically similar in most respects to the target zone, and more particularly, outside of the immediate area of impact by its upwind location. The format of Table 4.2 is similar to that of Table 4.1. It addresses the same 11 environmental impact items that were posed to residents in the target area. Table 4.2. Measures and Rank of Environmental Impacts in the Reference Area Item No. Item Measure S.E. Measure (in logits) Rank (Least impact to most impact) (1) (2) (3) (4) (5) (6) 5 Incorrect disposal of chemical waste 68.7 3.52 1.75 1 6 Industry drainage and runoff 62.3 2.69 1.14 2 1 Red mud dust 58.7 2.32 0.78 3 10 Landfill odor 58.7 2.32 0.78 4 7 Improper sewage disposal 58.2 2.28 0.73 5 11 Other materials in the cistern 50.6 1.74 0.00 6 3 Factory smoke 50.0 1.71 -0.06 7 8 Contaminated drinking water 48.7 1.64 -0.20 8 2 Transportation emissions and exhaust 37.8 1.23 -1.25 9 9 Strong irritating odors 33.6 1.14 -1.66 10 4 Mold 30.0 1.09 -2.02 11 Mean 50.7 1.97 0.00 S.D. 11.8 0.71 1.15 Reliability 0.96 Note: The numbers in col. (1) refer to the serial numbers of the items in the survey instrument (see Appendix II). The measures in col. (3) are rescaled to range from 0 to 100, and are equivalent to the logit measures in col. (5). 26 One of the first points of note is that the mean of the item measures in the control area is almost about the same as in the target area in Table 4.1. This is a positive indication that the respondents in the target and reference areas were interpreting the statements in essentially the same way. There is indication, however, that the residents of the reference area do not consider themselves to be impacted to the same intensity as residents in the target zone. A second aspect that is noticeable in Table 4.2 is that the rank order of the items is not the same, indicating that the residents who live at a distance from the industrial zone do not experience the same level of morbidity as those who live in proximity to the industrial zone. Thirdly, whereas the order of impact on those near the south shore area are strong and irritating odors, followed by factory smoke and then by mold; in the control area the order is mold, followed by strong and irritating odors, followed by transportation emissions and exhaust. Fourth, while red mud dust is ranked fourth by those who live near the industrial zone, it ranks very low in respect of its impact on the health of those who live at a distance from the industrial zone. Fifth, while the target area is least impacted by industry drainage and runoff, incorrect disposal of chemical waste and improper sewage disposal; in the reference zone the lowest impact is imparted by incorrect disposal of chemical waste, industry drainage and runoff and red mud dust. The mean of all respondents in the target zone—as distinct from the mean of the items—was 30.4 (SD = 11.2). This may be observed from the map in Figure 4.1 above. The map in Figure 4.2 displays similar information for the respondents in the reference area. The mean of the respondents on this map is 23.1 (SD = 9.7). The higher mean value of the respondents in the target zone is a reflection of the more frequent responses of Always, Sometimes or Scarcely 27 compared to Never. The lower mean value that residents in the control area gave is attributed to the frequency with which they responded to the lower categories like Never to indicate that it was only on occasion that they were subjected to environmental distress. The map provides another opportunity to visually observe the gaps between the different levels of impact of environmental conditions as expressed by the respondents who are in close proximity to the industrial and those who reside further afield. One observation that is evident is the relatively high level of occurrence of red mud dust that is noted on the south shore periphery, whereas it ranks third from the bottom in the reference area in terms of a health issue. 28 Figure 4.2 Section A Item-Respondent Map of Residents in the Control Area Measure Respondents - Items <less impact> | <less frequent impact on health> 80 . + | | | | |T | | | 70 + | 5 Incorrect disposed of chemical waste | | | | | |S 6 Industry drainage and runoff | 60 + | 1 Red mud dust 7 Improper sewage disposal 10 Landfill odor | | | . | | | . |M 11 Materials in cistern 50 + 3 Factory smoke . | 8 Contaminated drinking water | . | | . | T| | 40 . + |S . | 2 Trans emit & exhaust | | . | | 9 Strong irritating odor S| . | 30 + 4 Mold | | .# |T | | M| | .### | 20 + | | | | | .#### S| | | 10 .############ + <less>|<frequent> Each "#" is 20: Each "." is 1 to 19 29 E. Diseases in the Target and Reference Zones In a study on air toxins in St Croix, the EPA regional administrator was quoted as saying, “Improving air quality for the people of St Croix is a priority for EPA. The communities near these industrial facilities face health and environmental challenges from air pollution.”5 The same source indicated that both Bethlehem Village and Central High School were impacted by two key pollutant volatile organic compounds, benzene and 1,3-butadiene, which are associated with pollution from refineries. It was also reported that the chemical carbon disulfide, which is associated with strong odors, was highest at Bethlehem Village and was likely the effluent of a nearby rum distillery.6 Another study indicated that “air pollution has been the cause of numerous health problems including asthma and cancer” (Najjar, 2011). Section B of the questionnaire addresses a number of specific illnesses that were put to respondents. These included: asthma, chronic bronchitis, lung cancer, emphysema, COPD (chronic obstructive pulmonary disease), allergies, black lung disease, ARDS (acute respiratory distress syndrome), airborne substances, mesothelioma and any other environmentally induced ailment. The data results are reported in Tables 12 to 22. Each respondent was asked whether she or he had suffered from asthma in the last five years. The two simple responses were either Yes or No. Table 12 and Figure 12 in Appendix I show that about 1 in every 5 persons (21.3%) in the target area reported they had experienced asthmatic conditions within the last five years, and this compares with 1 in 8 (12.5%) in the reference area. This does indicate a substantive difference, and it appears to conform to 5 Mhtml:file://C:\OIA\STXZONE\PROPOSAL\EPA Issues Final. Downloaded 9/20/14 6 Ibid. 30 the current knowledge that residents in the industrial zone’s periphery are exposed to air pollutants that contribute to this disease. The prevalence of chronic bronchitis, shown in Table 13 and Figure 13, is exhibited in a similar pattern: proportionately, there were more than twice as many residents near the industrial zone who indicate having chronic bronchitis—10.4%—compared to residents who live further afield—4.1%. The most frequently reported illness in both geographic areas under study was the occurrence of allergies (Table 17 and Figure 17). Near the south shore industrial area, 37.1% of the respondents in the survey indicated that they suffered from allergies, and 27.4% of those responding in the reference area admitted to having allergies. For most of the other illnesses, the percentage of occurrence among the respondents was less than five percent. Another aspect of Section B was that if the respondent within the selected household answered in the affirmative that someone living in the household had suffered from a specific illness, the respondent was then asked how many within the household suffer from that condition. These results were captured in Table 23 and illustrated in Figure 23 (in Appendix I). It is not surprising that the illnesses that stand out were allergies, 21.5% and 15.3%, in the industrial zone and reference area, respectively, asthma (11.0% and 5.6%), and chronic bronchitis (5.2% and 1.8%). The effect of smoking is evident in the data in Table 24. The top panel of the table refers to the person in the household who was responding. Among those who reported being ill, the percentage was higher for those who smoked daily (10.2%) compared to those who did not report any illness (8.3%). Among the other members in the household, the disparity is even more evident: among 31 daily smokers, 7.1% reported illness compared to 3.9% who were not ill. The pattern was similar in the graph in Figure 24. The data in Table 25 indicate an incipient relationship between illness and the presence of harmful inhalants at work. In each classification of exposure — 1 to 2 times a year, Once a month, and Daily, ― the percent of those reporting illness was always higher than those without illness. One notes that those who report an illness present (10.6%) were more than twice that of those without an illness (4.7%). This relationship between the report on one’s health and regular exposure to inhalants at work will be analyzed further below. F. Inferences of Field Data Collection (1) Differences in the Level of Impact of Environmental Factors Before conducting any advanced statistical tests on the data, scores produced from the environmental incidences and adverse health conditions sections were transformed into measures. Specifically, the ordinal scores derived from the participants’ responses (e.g., Always, Scarcely, Sometimes, Never) were analyzed by a statistically rigorous methodology within the field of Item Response Theory (IRT). This statistical method transformed ordinal level responses into interval level measures that can be subjected to arithmetic operations. A difference of means test, or t-test, was conducted to reveal any statistically significant difference between the means of the groups concerning their perceptions of the impact of those listed environmental conditions in their households. The mean measure for the target group was 30.43 and the mean 32 measure for the reference group was 11.36. It is noticeable that the mean measure for the target was higher than the mean measure for the reference group, showing a difference of 19.07. The t-test for independent samples produced a t value of 23.60 and a p value of .000. Therefore, the t-test concluded that there was a statistically significant difference between the target and reference mean measures. One can conclude that persons within the target zone were much more likely, than those in the reference zone, to report that their households were impacted by negative environmental conditions. Figure 4.3 utilized the box plot graph to illustrate the difference between the measures of the two groups. A box plot is a graphical representation of a five-number summary of a distribution that includes the minimum, the maximum, and the three quartiles – the 25th, 50th and 75th – and captures other important features of the distribution. The horizontal line in the box represents the median of the distribution, or the middle score. Half of the measures, or 50%, lie above the median and half of the measures lie below the median. Whiskers extend from both ends of the box. On the lower end of the distribution, the whisker extends to the lowest measure that is not an outlier. An outlier is an observation point (or a measure, in this case) that is distant from other observations. On the upper end, the whisker also extends to the highest value that is not an outlier. The circles beyond the whiskers on the lower and upper end of the distribution of the measures for the target and reference groups represent the outliers. The asterisk on the upper end of the reference zone represents an extreme outlier. The box plots of the target and control zones in Figure 4.2 demonstrate the vast difference between the groups’ perceptions of the presence of negative 33 environmental factors. The median of the reference group is aligned with the first quartile. Therefore, it is apparent that an overwhelming number of respondents in the reference area did not perceive that their households were impacted by the listed environmental conditions (e.g., red mud dust, strong and irritating odors, etc.). The two distributions represented in the box plot illustrate the distinction between the reported conditions of those living in the separate zones. Moreover, one can conclude that target area residents found it easier to endorse the fact that they have been exposed to air pollution and other forms of harmful environmental substances. Figure 4.3. Box plot of target and reference measures 34 2) Relationship between Health Impact and Two Independent Variables In Section C above, measures were provided for respondents in the target zone. These measures of items were given in Table 4.1 and shown graphically in Figure 4.1. The focus at that time was on the construct Morbidity in the Vicinity of the Industrial Zone, to determine if there was a unidimensional construct along which the items could be hierarchically ordered. This effort was successful as there was only one item that showed very slight misfit. Equally important in the graphic was the distribution of the respondents on the left side of the vertical line. Each of the 496 respondents in the target area had a measure that was derived from the raw scores each provided in the ordinal categories of the 11 rated items. Each of these person measures reflects the location of each respondent relative to the items on the latent variable. They also reflect the situation in which a respondent who answered Sometimes or Always frequently to environmental impacts would have a higher raw score than one who frequently answered Never or Scarcely to the same items. High person measures would place a respondent high on the logit or transformed scale, and low measures would locate a respondent low on the scale. The purpose of this section is to estimate the statistical relationship between the outcome variable—the measure of environmental health impact—and two independent variables, and to determine if there is interaction between the two predictor variables. The person measures comprise the outcome variable for the factorial or two- factor ANOVA as the interval scale of this variable meets the criterion for the dependent variable in this type of linear model. The two main factors derive from date in two questions on the questionnaire. Question 25 asked if anyone in the responding household was consistently exposed to harmful inhalants 35 while at work. The other predictor variable was Question 36 which asked if the respondent worked in the South Shore Industrial Area. The two factors were dichotomized. Responses to Question 25 on harmful inhalants were reclassified into two categories, Never and Sometimes, the latter combining daily, once a month, and 1 to 2 times a year. Question 36 that relates to work in the South Shore Industrial Area is already in binary form, Yes or No. The (SAS) output presented in Table 4.3 immediately informs us whether any of the two independent variables had an effect on the dependent variable. The first panel of the table tells us about the overall significance of the model— the F value is 2.74, with 3, 492 df, and this is statistically significant, p < 0.05. The important aspects of the second panel in the table are the significance values of the independent variables. The first noticeable feature is that there is a significant main effect of consistent exposure to harmful inhalants at work. The F-ratio is highly significant, indicating that the attitude of a worker about environmental impacts is affected by her or his persistent exposure to harmful inhalants. This means that even when the place of work is ignored, the consistent exposure to harmful inhalants influences one’s perception of environmental conditions. The table further informs that the main effect IndusZ is not significant, with p = 0.402. This imparts that if the place of work of a respondent is in the SSIZ, which alone does not affect her or his attitude to the overall impact of the environment. It is further noticeable from the second panel that the IndusZ*Exposed interaction of the two predictor variables is not significant, with p = 0.645. This further means that one predictor variable does not 36 influence the other. (In graphic form, not shown, the two predictor lines are parallel, further affirming the lack of interaction between the two variables.) Table 4.3 Two-Way ANOVA Source df Sum of Squares Mean Square F Value Pr > F Model 3 1028.01 342.67 2.74 0.0427 Error 492 61487.90 124.98 Corrected Total 495 62515.91 Source df Type III SS Mean Square F Value Pr > F IndusZ 1 87.91 87.91 0.70 0.4021 Exposed 1 606.62 606.62 4.85 0.0280 IndusZ*Exposed 1 26.63 26.63 0.21 0.6446 Note: IndusZ is Work in industrial area, Exposed is Exposed to harmful inhalants at work. 37 IV SUMMARY Research on air pollution provides evidence of a relationship between long-term exposures to airborne toxins and a higher prevalence of asthma, bronchitis and other adverse respiratory conditions. The vast majority of studies on air pollution were conducted in industrial areas of the mainland United States and other urban environments. In highly populated, urban environments, human contact with environmental conditions that negatively influence respiratory health is not startling news. Outside of the Caribbean, many people would be surprised to hear that some residents of the beautiful island of St Croix have been consistently exposed to harmful pollutants in the air. For years, residents of St Croix have discussed the damaging effects, on human health, of working and/or living in the South Shore Industrial Area. This study of the environmental health of the industrial area of St Croix came forth to provide evidence toward, or against, the discussions that positively correlate environmental conditions and respiratory illnesses on the island. This study sought out to reveal differences, if any, between the perceptions of residents of the target and reference areas concerning the presence of environmental occurrences and the prevalence of adverse health conditions. Indeed, differences between the target and reference groups were apparent regarding the reporting of the impact of negative environmental conditions in the households of the target area residents. In fact, target area residents were more likely to report that their households were impacted by 38 strong and irritating odors, mold, transportation emissions and exhaust and factory smoke. The lower mean of the measures in the reference group reflected the higher frequency of responses of Never concerning the occurrences of negative environmental conditions. In fact, results of the IRT analysis of the data indicated that that the residents of the reference area do not consider themselves to be impacted to the same intensity as residents in the target zone. Furthermore, inferential statistics displayed a statistically significant difference between the intensity of environmental conditions in the target versus reference zones. In other words, more residents in the target zone reported the existence of those environmental conditions within their households and neighborhoods. In conclusion, the results of this study have offered empirical evidence that adverse environmental conditions that could lead to declining respiratory health exist at higher levels for residents in and around the South Shore Industrial Area of St Croix. Indeed, the conversations that many Crucians have met a premature death as a result of environmental conditions in the industrial area may be more than that of suspicion. This study represents a new beginning of scientific study of the relationship of industrial emanation and respiratory health of residents in industrial areas of St Croix. We encourage the results of this study to be shared with health care providers, educators, policy makers and all residents of the Territory. This research is not only important for the health and future of St Croix residents, but it is necessary for the overall prosperity of the entire US Virgin Islands. 39 References Bhopal, R. S., Moffatt, S., Pless-Mulloli, T., Phillimore, P. R., Foy, C., Dunn, C.E. & Tate, J.A. (1998). Does living near a constellation of petrochemical, steel, and other industries impair health? Occupational Environmental Medicine, 55, 812-822. Coward, S. (2011, February 11). EPA begins study of air pollution near HOVENSA oil refinery on St Croix, US Virgin Islands. Caribbean Press Releases. Retrieved from http://www.caribbeanpressreleases.com. Embretson, S.E., & Reise, S.P. (2000). Item Response Theory for Psychologists. Mahway, NJ: Lawrence Erlbaum Associates, Publishers. Najjar, Y. S. H. (2011). Gaseous pollutants formation and their harmful effects on health and environment. Innovative Energy Policies, 1, 1-9. Nuvolone, D., Maggiore, R., Maio, S., Fresco, R., Baldacci, S., Carrozz, L., Pistelli, F. & Viegi, G. (2011). Geographical information system and environmental epidemiology: a cross-sectional spatial analysis of the effects of traffic-related air pollution on population respiratory health. Environmental Health, 10, 1-12. Toren, K., Brisman, J. & Jarvholm, B. (1993). Asthma and asthma-like symptoms in adults assessed by questionnaire. Chest, 104, 600-608. Appendix I Frequency and Percentage Tables _____________________________________________________________________________________________ Table Page _____________________________________________________________________________________________ Table 1. Frequency of Impact of Red Mud Dust on Household by Zone ................................................... I- 5 Table 2. Frequency of Impact of Transportation Emissions and Exhaust on Household by Zone ............. I- 6 Table 3. Frequency of Impact of Factory Smoke on Household by Zone ................................................... I- 7 Table 4. Frequency of Impact of Mold on Household by Zone .................................................................. I- 8 Table 5. Frequency of Impact of Incorrect Disposal of Waste on Household by Zone .............................. I- 9 Table 6. Frequency of Impact of Industry Drainage and Runoff on Household by Zone ......................... I- 10 Table 7. Frequency of Impact of Improper Sewage Disposal on Household by Zone .............................. I- 11 Table 8. Frequency of Impact of Contaminated Drinking Water on Household by Zone ......................... I- 12 Table 9. Frequency of Impact of Strong and Irritating Odors on Household by Zone .............................. I- 13 Table 10. Frequency of Impact of Landfill Odor on Household by Zone ................................................... I- 14 Table 11. Frequency of Impact of Other Materials in the Cistern on Household by Zone ......................... I- 15 Table 12. Someone in Household has Suffered from Asthma by Zone ...................................................... I- 16 Table 13. Someone in Household has Suffered from Chronic Bronchitis by Zone .................................... I- 17 Table 14. Someone in Household has Suffered from Lung Cancer by Zone .............................................. I- 18 Table 15. Someone in Household has Suffered from Emphysema by Zone ............................................... I- 19 Table 16. Someone in Household has Suffered from Chronic Obstructive Pulmonary Disease by Zone ... I- 20 Table 17. Someone in Household has Suffered from Allergies by Zone .................................................... I- 21 Table 18. Someone in Household has Suffered from Black Lung Disease by Zone ................................... I- 22 Table 19. Someone in Household has Suffered from Acute Respiratory Distress Syndrome by Zone ....... I- 23 Table 20. Someone in Household has Suffered from Airborne Substances by Zone .................................. I- 24 Table 21. Someone in Household has Suffered from Mesothelioma by Zone ............................................ I- 25 Table 22. Someone in Household has Suffered from Some Other Illness by Zone .................................... I- 26 Table 23. Persons who Suffered from Illness by Zone ............................................................................... I- 27 Table 24. Smokers in Household by Presence of Any Illness ..................................................................... I- 28 I- 1 Table 25. Someone in Household Exposed to Harmful Inhalants at Work by Presence of Any Illness ..... I- 29 Table 26. Income of the Respondent Households by Zone ......................................................................... I- 30 Table 27. Sex of Persons who Suffered from Illness by Zone .................................................................... I- 31 Table 28. Age of Persons who Suffered from Illness by Zone ................................................................... I- 32 Table 29. Race of Persons who Suffered from Illness by Zone .................................................................. I- 33 Table 30. Ethnicity of Persons who Suffered from Illness by Zone ............................................................ I- 34 Table 31. Years in Residence of Persons who Suffered from Illness by Zone ........................................... I- 35 Table 32. Hours Spent in Neighborhood of Persons who Suffered from Illness ........................................ I- 36 Table 33. Area of Work of Persons who Suffered from Illness by Zone .................................................... I- 37 Table 34. Educational Attainment of Persons who Suffered from Illness by Zone .................................... I- 38 I- 2 Appendix I Percentage Charts ______________________________________________ Figure Page _____________________________________________________________________________________________ Figure 1. Frequency of Impact of Red Mud Dust on Household ................................................................. I- 5 Figure 2. Frequency of Impact of Transportation Emissions and Exhaust on Household .......................... I- 6 Figure 3. Frequency of Impact of Factory Smoke on Household ................................................................ I- 7 Figure 4. Frequency of Impact of Mold on Household ............................................................................... I- 8 Figure 5. Frequency of Impact of Incorrect Disposal of Waste on Household ........................................... I- 9 Figure 6. Frequency of Impact of Industry Drainage and Runoff on Household ...................................... I- 10 Figure 7. Frequency of Impact of Improper Sewage Disposal on Household ........................................... I- 11 Figure 8. Frequency of Impact of Contaminated Drinking Water on Household ...................................... I- 12 Figure 9. Frequency of Impact of Strong and Irritating Odors on Household ........................................... I- 13 Figure 10. Frequency of Impact of Landfill Odor on Household ................................................................ I- 14 Figure 11. Frequency of Impact of Other Materials in the Cistern on Household ....................................... I- 15 Figure 12. Someone in Household has Suffered from Asthma ................................................................... I- 16 Figure 13. Someone in Household has Suffered from Chronic Bronchitis ................................................. I- 17 Figure 14. Someone in Household has Suffered from Lung Cancer ........................................................... I- 18 Figure 15. Someone in Household has Suffered from Emphysema ............................................................ I- 19 Figure 16. Someone in Household has Suffered from Chronic Obstructive Pulmonary Disease ................ I- 20 Figure 17. Someone in Household has Suffered from Allergies ................................................................. I- 21 Figure 18. Someone in Household has Suffered from Black Lung Disease ................................................ I- 22 Figure 19. Someone in Household has Suffered from Acute Respiratory Distress Syndrome .................... I- 23 Figure 20. Someone in Household has Suffered from Airborne Substances ............................................... I- 24 Figure 21. Someone in Household has Suffered from Mesothelioma ......................................................... I- 25 Figure 22. Someone in Household has Suffered from Some Other Illness ................................................. I- 26 Figure 23. Persons who Suffered from Illness ............................................................................................ I- 27 I- 3 Figure 24. Smokers in Household ................................................................................................................ I- 28 Figure 25. Exposure to Harmful Inhalants at Work ..................................................................................... I- 29 Figure 26. Income of the Respondent Households ...................................................................................... I- 30 Figure 27. Sex of Persons who Suffered from Illness ................................................................................. I- 31 Figure 28. Age of Persons who Suffered from Illness ................................................................................. I- 32 Figure 29. Race of Persons who Suffered from Illness ............................................................................... I- 33 Figure 30. Ethnicity of Persons who Suffered from Illness ......................................................................... I- 34 Figure 31. Years in Residence of Persons who Suffered from Illness ......................................................... I- 35 Figure 32. Hours Spent in Neighborhood of Persons who Suffered from Illness ........................................ I- 36 Figure 33. Area of Work of Persons who Suffered from Illness ................................................................. I- 37 Figure 34. Educational Attainment of Persons who Suffered from Illness ................................................. I- 38 I- 4 Table 1. Frequency of Impact of Red Mud Dust on Household by Zone Responses Number Percent Number Percent Total 504 100.0 486 100.0 Never 307 60.9 465 95.7 Scarcely 31 6.2 6 1.2 Sometimes 84 16.7 14 2.9 Always 82 16.3 1 0.2 18 Target Reference 60.9 6.2 16.7 16.3 95.7 1.2 2.9 0.2 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 110.0 Never Scarcely Sometimes Always Figure 1. Frequency of Impact of Red Mud Dust on Household Target Reference Percent I- 5 Table 2. Frequency of Impact of Transportation Emissions and Exhaust on Household by Zone Responses Number Percent Number Percent Total 504 100.0 486 100.0 Never 283 56.2 398 81.9 Scarcely 46 9.1 21 4.3 Sometimes 119 23.6 48 9.9 Always 56 11.1 19 3.9 18 Target Reference 56.2 9.1 23.6 11.1 81.9 4.3 9.9 3.9 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 Never Scarcely Sometimes Always Percent Figure 2. Frequency of Impact of Transportation Emissions and Exhaust on Household Target Reference I- 6 Table 3. Frequency of Impact of Factory Smoke on Household by Zone Responses Number Percent Number Percent Total 502 100.0 486 100.0 Never 241 48.0 448 92.2 Scarcely 60 12.0 12 2.5 Sometimes 156 31.1 19 3.9 Always 45 9.0 7 1.4 18 Target Reference 48.0 12.0 31.1 9.0 92.2 2.5 3.9 1.4 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 Never Scarcely Sometimes Always Percent Figure 3. Frequency of Impact of Factory Smoke on Household Target Reference I- 7 Table 4. Frequency of Impact of Mold on Household by Zone Responses Number Percent Number Percent Total 504 100.0 486 100.0 Never 285 56.5 355 73.0 Scarcely 36 7.1 24 4.9 Sometimes 109 21.6 70 14.4 Always 74 14.7 37 7.6 18 Target Reference 56.5 7.1 21.6 14.7 73.0 4.9 14.4 7.6 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 Never Scarcely Sometimes Always Percent Figure 4. Frequency of Impact of Mold on Household Target Reference I- 8 Table 5. Frequency of Impact of Incorrect Disposal of Waste on Household by Zone Responses Number Percent Number Percent Total 504 100.0 486 100.0 Never 457 90.7 478 98.4 Scarcely 13 2.6 2 0.4 Sometimes 24 4.8 5 1.0 Always 10 2.0 1 0.2 18 Target Reference 90.7 2.6 4.8 2.0 98.4 0.4 1.0 0.2 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 Never Scarcely Sometimes Always Percent Figure 5. Frequency of Impact of Incorrect Disposal of Waste on Household Target Reference I- 9 Table 6. Frequency of Impact of Industry Drainage and Runoff on Household by Zone Responses Number Percent Number Percent Total 504 100.0 486 100.0 Never 466 92.5 471 96.9 Scarcely 3 0.6 3 0.6 Sometimes 18 3.6 11 2.3 Always 17 3.4 1 0.2 18 Target Reference 92.5 0.6 3.6 3.4 96.9 0.6 2.3 0.2 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 Never Scarcely Sometimes Always Percent Figure 6. Frequency of Impact of Industry Drainage and Runoff on Household Target Reference I- 10 Table 7. Frequency of Impact of Improper Sewage Disposal on Household by Zone Responses Number Percent Number Percent Total 504 100.0 487 100.0 Never 445 88.3 468 96.1 Scarcely 13 2.6 3 0.6 Sometimes 28 5.6 12 2.5 Always 18 3.6 4 0.8 18 Target Reference 88.3 2.6 5.6 3.6 96.1 0.6 2.5 0.8 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 Never Scarcely Sometimes Always Percent Figure 7. Frequency of Impact of Improper Sewage Disposal on Household Target Reference I- 11 Table 8. Frequency of Impact of Contaminated Drinking Water on Household by Zone Responses Number Percent Number Percent Total 502 100.0 486 100.0 Never 348 69.3 447 92.0 Scarcely 24 4.8 6 1.2 Sometimes 75 14.9 21 4.3 Always 55 11.0 12 2.5 18 Target Reference 69.3 4.8 14.9 11.0 92.0 1.2 4.3 2.5 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 Never Scarcely Sometimes Always Percent Figure 8. Frequency of Impact of Contaminated Drinking Water on Household Target Reference I- 12 Table 9. Frequency of Impact of Strong and Irritating Odors on Household by Zone Responses Number Percent Number Percent Total 503 100.0 486 100.0 Never 133 26.4 363 74.7 Scarcely 62 12.3 32 6.6 Sometimes 237 47.1 76 15.6 Always 71 14.1 15 3.1 18 Target Reference 26.4 12.3 47.1 14.1 74.7 6.6 15.6 3.1 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 Never Scarcely Sometimes Always Percent Table 9. Frequency of Impact of Strong and Irritating Odors on Household Target Reference I- 13 Table 10. Frequency of Impact of Landfill Odor on Household by Zone Responses Number Percent Number Percent Total 504 100.0 486 100.0 Never 378 75.0 466 95.9 Scarcely 41 8.1 5 1.0 Sometimes 69 13.7 13 2.7 Always 16 3.2 2 0.4 18 Target Reference 75.0 8.1 13.7 3.2 95.9 1.0 2.7 0.4 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 Never Scarcely Sometimes Always Percent Figure 10. Frequency of Impact of Landfill Odor on Household Target Reference I- 14 Table 11. Frequency of Impact of Other Materials in the Cistern on Household by Zone Responses Number Percent Number Percent Total 500 100.0 485 100.0 Never 399 79.8 448 92.4 Scarcely 28 5.6 9 1.9 Sometimes 38 7.6 22 4.5 Always 35 7.0 6 1.2 18 Target Reference 79.8 5.6 7.6 7.0 92.4 1.9 4.5 1.2 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 Never Scarcely Sometimes Always Percent Figure 11. Frequency of Impact of Other Materials in the Cistern on Household Target Reference I- 15 Table 12. Someone in Household has Suffered from Asthma by Zone Responses Number Percent Number Percent Total 653 100.0 487 100.0 Yes 139 21.3 61 12.5 No 514 78.7 426 87.5 Target Reference 21.3 78.7 12.5 87.5 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 Yes No Percent Figure 12. Someone in Household has Suffered from Asthma Target Reference I- 16 Table 13. Someone in Household has Suffered from Chronic Bronchitis by Zone Responses Number Percent Number Percent Total 653 100.0 487 100.0 Yes 68 10.4 20 4.1 No 585 89.6 467 95.9 Target Reference 10.4 89.6 4.1 95.9 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 Yes No Percent Figure 13. Someone in Household has Suffered from Chronic Bronchitis Target Reference I- 17 Table 14. Someone in Household has Suffered from Lung Cancer by Zone Responses Number Percent Number Percent Total 653 100.0 486 100.0 Yes 6 0.9 2 0.4 No 647 99.1 484 99.6 Target Reference 0.9 99.1 0.4 99.6 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 Yes No Percent Figure 14. Someone in Household has Suffered from Lung Cancer Target Reference I- 18 Table 15. Someone in Household has Suffered from Emphysema by Zone Responses Number Percent Number Percent Total 653 100.0 485 100.0 Yes 7 1.1 1 0.2 No 646 98.9 484 99.8 Target Reference 1.1 98.9 0.2 99.8 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 Yes No Percent Figure 15. Someone in Household has Suffered from Emphysema Target Reference I- 19 Table 16. Someone in Household has Suffered from Chronic Obstructive Pulmonary Disease (COPD) by Z Responses Number Percent Number Percent Total 653 100.0 483 100.0 Yes 11 1.7 7 1.4 No 642 98.3 476 98.6 Target Reference 1.7 98.3 1.4 98.6 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 Yes No Percent Figure 16. Someone in Household has Suffered from COPD Target Reference I- 20 Table 17. Someone in Household has Suffered from Allergies by Zone Responses Number Percent Number Percent Total 653 100.0 486 100.0 Yes 242 37.1 133 27.4 No 411 62.9 353 72.6 Target Reference 37.1 62.9 27.4 72.6 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 Yes No Percent Figure 17. Someone in Household has Suffered from Allergies Target Reference I- 21 Table 18. Someone in Household has Suffered from Black Lung Disease by Zone Responses Number Percent Number Percent Total 652 100.0 485 100.0 Yes 1 0.2 1 0.2 No 651 99.8 484 99.8 Target Reference 0.2 99.8 0.2 99.8 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 Yes No Percent Figure 18. Someone in Household has Suffered from Black Lung Disease Target Reference I- 22 Table 19. Someone in Household has Suffered from Acute Respiratory Distress Syndrome (ARDS) by Zone Responses Number Percent Number Percent Total 653 100.0 486 100.0 Yes 20 3.1 12 2.5 No 633 96.9 474 97.5 Target Reference 3.1 96.9 2.5 97.5 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 Yes No Percent Figure 19. Someone in Household has Suffered from ARDS Target Reference I- 23 Table 20. Someone in Household has Suffered from Airborne Substances by Zone Responses Number Percent Number Percent Total 652 100.0 486 100.0 Yes 27 4.1 14 2.9 No 625 95.9 472 97.1 Target Reference 4.1 95.9 2.9 97.1 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 Yes No Percent Figure 20. Someone in Household has Suffered from Airborne Substances Target Reference I- 24 Table 21. Someone in Household has Suffered from Mesothelioma by Zone Responses Number Percent Number Percent Total 653 100.0 483 100.0 Yes 6 0.9 3 0.6 No 647 99.1 480 99.4 Target Reference 0.9 99.1 0.6 99.4 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 Yes No Percent Figure 21. Someone in Household has Suffered from Mesothelioma Target Reference I- 25 Table 22. Someone in Household has Suffered from Some Other Illness by Zone Responses Number Percent Number Percent Total 829 100.0 684 100.0 Yes 9 1.1 7 1.0 No 820 98.9 677 99.0 Target Reference 1.1 98.9 1.0 99.0 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 Yes No Percent Figure 22. Someone in Household has Suffered from Some Other Illness Target Reference I- 26 Illness Number Percent Number Percent Total Number of Persons in Responding Households 1,741 100.00 1,285 100.0 Asthma 191 11.0 72 5.6 Chronic Bronchitis 91 5.2 23 1.8 Lung Cancer 6 0.3 2 0.2 Emphysema 9 0.5 1 0.1 COPD 13 0.7 7 0.5 Allergies 375 21.5 196 15.3 Black Lung Disease 1 0.1 1 0.1 ARDS 25 1.4 15 1.2 Mesothelioma 6 0.3 3 0.2 Other 13 0.7 9 0.7 Reference Target Table 23. Persons who Suffered from Illness by Zone 11.0 5.2 0.3 0.5 0.7 21.5 0.1 1.4 0.3 0.7 5.6 1.8 0.2 0.1 0.5 15.3 0.1 1.2 0.2 0.7 0.0 5.0 10.0 15.0 20.0 25.0 Figure 23.  Persons who Suffered from Illness  Target Reference I- 27 Table 24. Smokers in Household by Presence of Any Illness Target Responses Number Percent Number Percent Respondent is a Smoker Total 659 100.0 481 100.0 Never 583 88.5 420 87.3 1 to 2 times a year 6 0.9 1 0.2 Once a month 15 2.3 11 2.3 Daily 55 8.3 49 10.2 Other Household Member is a Smoker Total 659 100.0 481 100.0 Never 626 95.0 431 89.6 1 to 2 times a year 4 0.6 4 0.8 Once a month 3 0.5 12 2.5 Daily 26 3.9 34 7.1 No Illness Illness Present 13.7 86.3 18.1 81.9 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 At Least One Person in Household Smokes No One in Household Smokes Percent Figure 24. Smokers in Household No Illness Illness Present I- 28 Table 25. Someone in Household Exposed to Harmful Inhalants at Work by Presence of Any Illness Target Responses Number Percent Number Percent Total 659 100.0 481 100.0 Never 611 92.7 414 86.1 1 to 2 times a year 9 1.4 7 1.5 Once a month 8 1.2 9 1.9 Daily 31 4.7 51 10.6 No Illness Illness Present 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 Never 1 to 2 times a year Once a month Daily Percent Frequency of Exposure Figure 25.  Exposure to Harmful Inhalants at Work No Illness Illness Present I- 29 Table 26. Income of the Respondent Households by Zone Responses Number Percent Number Percent Total 571 100.0 431 100.0 Less than $5,000 85 14.9 52 12.1 $5,000-$9,999 53 9.3 14 3.2 $10,000-$14,999 51 8.9 41 9.5 $15,000-$24,999 94 16.5 53 12.3 $25,000-$34,999 114 20.0 79 18.3 $35,000-$49,999 96 16.8 83 19.3 $50,000-$74,999 50 8.8 62 14.4 $75,000-$99,999 19 3.3 25 5.8 $100,000 and over 9 1.6 22 5.1 Target Reference 14.9 9.3 8.9 16.5 20.0 16.8 8.8 3.3 1.6 12.1 3.2 9.5 12.3 18.3 19.3 14.4 5.8 5.1 0.0 10.0 20.0 Less than $5,000 $5,000-$9,999 $10,000-$14,999 $15,000-$24,999 $25,000-$34,999 $35,000-$49,999 $50,000-$74,999 $75,000-$99,999 $100,000 and over Percent Household Income Figure 26. Income of the Respondent Households Reference Target I- 30 Table 27. Sex of Persons who Suffered from Illness by Zone Responses Number Percent Number Percent Total 307 100.0 176 100.0 Male 130 42.3 83 47.2 Female 177 57.7 93 52.8 Target Reference 42.3 57.7 47.2 52.8 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 Male Female Percent Figure 27. Sex of Persons who Suffered from Illness Target Reference I- 31 Table 28. Age of Persons who Suffered from Illness by Zone Responses Number Percent Number Percent Total 307 100.0 176 100.0 Under 18 49 16.0 21 11.9 18-24 26 8.5 17 9.7 25-39 47 15.3 21 11.9 40-54 80 26.1 47 26.7 55-69 77 25.1 51 29 70+ 28 9.1 19 10.8 Target Reference 16.0 8.5 15.3 26.1 25.1 9.1 11.9 9.7 11.9 26.7 29.0 10.8 0.0 10.0 20.0 30.0 Under 18 18-24 25-39 40-54 55-69 70+ Percent Age Figure 28. Age of Persons who Suffered from Illness Reference Target I- 32 Table 29. Race of Persons who Suffered from Illness by Zone Responses Number Percent Number Percent Total 307 100.0 175 100.0 Black 228 74.3 116 66.3 White 6 2.0 16 9.1 Asian 0 0.0 1 0.6 Native American 6 2.0 3 1.7 Pacific Islander 2 0.7 0 0 Other 65 21.2 39 22.3 Target Reference 74.3 2.0 0.0 2.0 0.7 21.2 66.3 9.1 0.6 1.7 0.0 22.3 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 Black White Asian Native American Pacific Islander Other Race Figure 29. Race of Persons who Suffered from Illness Target Reference I- 33 Table 30. Ethnicity of Persons who Suffered from Illness by Zone Responses Number Percent Number Percent Total 307 100.0 176 100.0 Not Hispanic 201 65.5 121 68.8 Hispanic 106 34.5 55 31.3 Target Reference 65.5 34.5 68.8 31.3 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 Not Hispanic Hispanic Percent Figure 30. Ethnicity of Persons who Suffered from Illness Target Reference I- 34 Table 31. Years in Residence of Persons who Suffered from Illness by Zone Responses Number Percent Number Percent Total 306 100.0 175 100.0 Less than 1 year 17 5.6 14 8.0 1 to 5 years 74 24.2 28 16.0 6 to 10 years 59 19.3 31 17.7 11 years or more 156 51.0 102 58.3 18 Target Reference 5.6 24.2 19.3 51.0 8.0 16.0 17.7 58.3 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 Less than 1 year 1 to 5 years 6 to 10 years 11 years or more Percent Figure 31. Years in Residence of Persons who Suffered from Illness Target Reference I- 35 Table 32. Hours Spent in Neighborhood of Persons who Suffered from Illness by Zone Responses Number Percent Number Percent Total 306 100.0 175 100.0 Less than 12 hours 72 23.5 42 24.0 12 hours or more 234 76.5 133 76.0 Target Reference 23.5 76.5 24.0 76.0 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 Less than 12 hours 12 hours or more Percent Figure 32. Hours Spent in Neighborhood of Persons who Suffered from Illness Target Reference I- 36 Table 33. Area of Work of Persons who Suffered from Illness by Zone Responses Number Percent Number Percent Total 306 100.0 175 100.0 South Shore Industrial Area 22 7.2 22 12.6 Other 284 92.8 153 87.4 Target Reference 7.2 92.8 12.6 87.4 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 South Shore Industrial Area Other Percent Area of Work Figure 33. Area of Work of Persons who Suffered from Illness Target Reference I- 37 Table 34. Educational Attainment of Persons who Suffered from Illness by Zone Responses Number Percent Number Percent Total 306 100.0 173 100.0 Less than High School 125 40.8 48 27.7 High School/GED 89 29.1 53 30.6 Some College 46 15.0 36 20.8 College Grad 31 10.1 22 12.7 Graduate/Prof. Degree 15 4.9 14 8.1 Target Reference 40.8 29.1 15.0 10.1 4.9 27.7 30.6 20.8 12.7 8.1 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 Less than High School High School/GED Some College College Grad Graduate/Prof. Degree Educational Attainment Figure 34. Educational Attainment of Persons who Suffered from Illness Target Reference I- 38 1 Good day, my name is (your name). The University of the Virgin Islands is conducting a survey about health conditions in this area. Your housing unit was randomly selected to participate in this survey. The purpose is to gather information from residents on St Croix regarding conditions that affect one’s health. All information obtained is strictly confidential. May I continue with the interview? Please respond to each statement or question as they refer to you (or your housing unit). Section A: Please indicate the frequency of each event by stating Never, Scarcely, Sometimes or Always for each statement. Within the last 5 years, this household has been impacted by…… Never Scarcely Sometimes Always 1. Red mud dust     2. Transportation emissions and exhaust     3. Factory smoke     4. Mold     5. Incorrect disposal of chemical waste     6. Industry drainage and runoff     7. Improper sewage disposal     8. Contaminated drinking water     9. Strong and irritating odors     10. Landfill odor     11. Other materials in the cistern (i.e., things other than water)     Section B: Please respond to each statement by answering Yes or No. Within the last 5 years, someone living in your household has suffered from… Yes No If yes, how many persons in this household suffer from (or suffered from) this condition? 12. Asthma   __ 13. Chronic bronchitis   __ Environmental Health Study Interviewer’s Name_________________________ Date ___/___/___ HU ID#__________________________________________________ Vacant  UHE  Final Disposition - ECC use only  Place barcode ID sticker here APPENDIX II 2 Yes No If yes, how many persons in this household suffer from (or suffered from) this condition? 14. Lung cancer   __ 15. Emphysema   __ 16. Chronic obstructive pulmonary disease (COPD)   __ 17. Allergies   __ 18. Black Lung Disease   __ 19. Acute respiratory distress syndrome (ARDS)   __ 20. Airborne substances (explain) _______________________________   __ 21. Mesothelioma   __ 22. Other (explain) ___________________________________________ __ Section C: Please indicate the level of occurrence by stating Daily, Once a month, 1 to 2 times a year or Never for each statement. In the past 5 years… Daily Once a month 1 to 2 times a year Never 23. Did you smoke cigarettes or some other substance?     24. Did some other person in your household smoke cigarettes or some other substance?     25. Was anyone in this household consistently exposed to harmful inhalants at work?     Section D: The following section asks demographic questions. Answer the questions about the one (1) selected person. - If one (1) person in the HU has suffered from a listed condition, use that person for Section D. - If more than one (1) person in the HU has suffered from a condition, choose a person (with a condition) with the most recent birthday for Section D. - If no one in the HU has suffered from a condition, choose the person with the most recent birthday for Section D. 26. Is (the selected person’s name)…? Male  Female  27. Which age category best describes him/her? Under 18  18-24  25-39  40-54  55-69  70 +  28. What is his/her ethnic background? Black  White  Asian  Native American  Pacific Islander  Other  3 29. Is he/she Latino/Hispanic? Yes  No  30. Which category describes his/her current level of education? Less than high school  High school graduate/GED  Some college  College graduate  Graduate or Professional degree  31. Which category best describes the annual income of this household?  Less than $5,000  $5,000 to $9,999  $10,000 to $14,999  $15,000 to $24,999  $25,000 to $34,999  $35,000 to $49,999  $50,000 to $74,999  $75,000 to $99,999  $100,000 and over 32. How long has he/she lived in this home?  Less than 1 year  1 to 5 years  6 to 10 years (Skip to #34)  11 years or more (Skip to #34) 33. Where did he/she live before moving here? _______________________________ (If in St Croix, write the name of the estate. If other than St Croix, write the name of the island, state or country.) 34. On April 1, 2014, how many people lived in this household?     10 or more  35. The number of hours per day that he/she spends at home or in this neighborhood is…? Less than 12 hours  12 hours or more  36. Does he/she work in the South shore industrial area? Yes  No  Thank You! (turn over) 4 Notes: _________________________________________________________________________________________________ _________________________________________________________________________________________________ _________________________________________________________________________________________________ _________________________________________________________________________________________________ _________________________________________________________________________________________________ _________________________________________________________________________________________________ Interviewer’s Signature:_____________________________________ e