Ake Thörn, MD, MPH
Department of Public Health Sciences,
Division of Social Medicine, Stockholm, Sweden
Unit for Occupational Medicine
County Hospital, Boden, Sweden
961 64 Boden
Tel + 46 921 662 65, Fax + 46 921 502 62, E-mail: email@example.com
Paper presented at the Qualitative Evidence-based Practice Conference, Coventry University, May 15-17 2000.
Background: The research on the sick building syndrome is dominated by quantitative methodology, which focuses preventive and corrective strategies and actions on indoor air and the building as a purely physical structure. However, in a number of buildings, SBS is not resolved despite ambitious actions directed towards encountered presumed physical causes.
Methods: Based on historical documents and semi-structured interviews, this paper is a qualitative study of an office with long-standing building-related health problems.
Results: A state of reduced credibility developed between building owner, employees and employer, and was an important reason for why symptoms persisted after suspected causative defects in the building were corrected. This suggests that in sick building syndromes, symptoms are influenced by interactions between biomedical and psychosocial factors.
Conclusions: Failure to plan and act from an integrated bio-psycho-social perspective might lead to a building becoming chronically sick in spite of rational corrective measures. Case-studies might facilitate the identification and understanding of complex social dynamics and conflictive agendas within buildings, and therefore be a valuable complementary study approach to quantitative studies.
In medical practice much work concerns patients suffering from non-specific symptoms of the mucous membranes of the eyes and upper and lower respiratory tracts, skin reactions, fatigue, headache, nausea and concentration difficulties. In many cases, even after thorough medical investigations, no pathological alterations can be demonstrated, making a well defined diagnosis difficult. Historical research indicates that constellations of such symptoms have been common for centuries, and at various times attributed to different causes reflected in the designations given to them (1). These seem to evolve in an interplay between contemporary ideas, beliefs and knowledge both within and without the medical profession (1,2). Currently, denominations such as "multiple chemical sensitivity"(2), "electric hypersensitivity"(3) and "sick building syndrome"(4) are common. Their causes are attributed to factors in the environment, and this environmental attribution seems to be particularly prevalent today (1).
As in most medical research, studies on non-specific health problems are done with quantitative methodologies. However, increasingly it has been argued that health research needs to be complemented by qualitative methods (5). It has been pointed out that, in the study of environmental risks with disease outcomes consisting of non-specific symptoms, the need for the complementary use of perspectives and angles other than those traditionally used in epidemiological studies is particularly great (6). One possible approach could be the case study, which by tradition has been important in clinical research in the generation of fruitful hypothesis regarding diseases and their causes. For such purposes, case studies might also be valuable in the investigation of environmentally attributed syndromes such as the sick building syndrome (abbreviated SBS) (Table 1), which has become common during the past two decades (7,8).
Table 1. The most common Symptoms in Sick Building Syndrome (WHO 1983)
|Irritation of the eyes, nose and throat
Experience of dry skin and dry mucosa
Fatigue, headache, nausea, dizziness
Hoarseness, wheezing, pruritus and non-specific hypersensitivity
Research has demonstrated that many factors can be associated with the occurrence of SBS (4,8,9,10,11), which is why it is considered to be a multifactorial condition of unknown aetiology (11,12). However, the accumulated knowledge is broad and general, and hard to base preventive and corrective measures on (13). This might depend on the non-specific nature of the symptoms making SBS difficult to study with other than cross sectional designs. It might also depend on a confusion of exposure and outcome in the very SBS definition, which states that a sick building syndrome is a building where the population of inhabitants demonstrates an increased prevalence of subjective symptoms (14). Working within such a definition, the building itself including its inhabitants, not the individual person, ought to be the basic unit of investigation in epidemiological studies. Such investigations are, however, very difficult to realise. Case study methodology (15), on the other hand, is considered useful in situations similar to SBS, where the boundaries between a studied phenomenon (e.g. "outcome": subjective symptoms) and its context (e.g. "exposure": building including its population, activities, organisations and relations) are not clearly evident. The present paper presents a case study that examines and discusses processes in an office with long-standing health problems. The aim is to investigate the possibilities of case study design to enhance the understanding of SBS development, and to provide models for its practical handling.
MATERIAL AND METHODS
Case building selection. Sources of information
In the most northern region of Sweden, the Labour Inspectorate identified two offices with non-specific health problems since at least 1990 and still ongoing in 1996. The study of one is presented here. Sources of information included interviews and historical documents (work environment investigations, health surveys, diaries, newspaper articles, etc.) provided by the employer, the local occupational health services (OSH), the Labour Inspectorate, individual employees and the local newspapers archives (Table 2).
Table 2. Sources of information used in the case study
|Health surveys performed by the local Occupational
Safety & Health Service (OSH) using the Örebro questionnaire (16).
Building construction drawings.
Work environment investigations by the local OSH, and the property owner and his consultants.
|Minutes from work-place meetings
Individual employees diaries
Injunctions from Labour Inspectorate
Interviews with selection (n=7) of
Material from local newspapers
The case building
The office, situated in a Swedish county capital, has two storeys and a cellar. It was constructed in 1982 with a concrete framework, outside brick walls and inside plaster discs covered with wallpaper. The floors were covered with linoleum flooring on a layer of putty. The roof is of wood covered with sheets of steel. Its ventilation is mechanical with both supply and exhaust air. Outdoor supply air is pre-heated by outgoing indoor air in a rotating heat-exchange device. The proprietor is a nation-wide property management company. The building was rented by the same office enterprise until its evacuation in 1996. At that time, seven women and 16 men worked in the office, 70% of them since 1982.
Selection of informants. Interviews
The purpose was to obtain a wide range of perspectives on the buildings health problems development and handling over the years, which is why the selection was intentional rather than random (17). The local OSH identified seven informants who, according to its opinion, would enable the exploration of different aspects of the buildings disease history: gender, management, employee and syndicate. They had all worked in the building since 1982/1983: five men (one the office manager and one the unions representative) and two women. Based on a loose pattern of open questions, the interviews focused on the buildings history during 1982-1996, with emphasis on the symptom evolution, perceived causes and how these were dealt with. They were tape-recorded, transcribed and analysed for patterns and regularities, which were grouped into different themes.
Symptoms and work-place investigations
After a year in the office, the employees started to experience mucosal irritation, fatigue, headache and concentration difficulties. Years later, these were reported to the employer and in 1988 and 1995 the OSH made health surveys (16), which demonstrated a high prevalence (>30%) of the type of non-specific symptoms mentioned.
In 1987 the buildings ventilation was investigated. No faults were found. Two years later, a leakage in the rotating heat-exchange device, leading to the recirculation of some exhaust air, was demonstrated and corrected. In 1994, continued symptoms led to further investigations by the owners consultants. Samples were taken from locations with conditions for mould growth, which on examination could not be demonstrated. Measurements of total volatile organic compounds (TVOC) were in the range 48-119 ng/l which was considered normal (18). Formaldehyde and dust were normal according to Swedish national threshold limits.
Some defects were found in the drainage and parts of the cellars concrete foundation had an elevated moisture (RH 95-99 %) content. The systems of joists on the first and second storeys, on the other hand, had normal moisture content (RH <70 %). Analysis of the flooring putty revealed that it contained casein. The glue beneath the linoleum floorings was wet and pipe passages between the storeys were not tight allowing the passage of air and possible contaminants between them. It was concluded that the health problems were probably due to emissions from flooring putty and glue in the cellar where chemical decomposition was facilitated by the moisture underneath.
In 1994-1995, the drainage system was reconstructed, the pipe passages between the storeys were sealed and all inner floorings were removed. Underlying glue and putty were removed, except in the cellar where only the glue could be removed. To prevent possible emissions from the remaining putty there, a ventilated inner floor was constructed to cover it. The removed linoleum was replaced on all floors by non-glued PVC-carpets.
Evacuations and later symptom development
The repair programme began in December 1994 and continued for several months with the personnel working as usual. During this period, the symptom prevalence increased as evidenced by a health survey, which is why the personnel were evacuated to another building in March 1995. According to the interviews, after the evacuation the symptoms disappeared.
When the repairs were complete, the personnel returned to the office in October 1995. Shortly afterwards, symptoms started to be reported again, leading the proprietor to examine the condition of the ventilation tubes. This operation started on 26 February 1996 with the removal of the inner ceilings in order to access the tubes. This was done by construction workers using ventilation masks while the office workers were working as usual. Visible dust was spread, leading the health and safety representative to close the building on 1 March. The personnel were again evacuated to other premises and never returned.
The interviews provided a picture of an evolution of symptoms that began between 1982-83. Employees with symptoms initially looked for explanations in individual, personal factors, such as disease, increasing age, imagination, stress and concern owing to excessive workload and on-going changes of work organisation. Gradually, they became aware of colleagues suffering from similar symptoms. In 1985-86, the ensuing discussions led to the opinion that the symptoms were related to the office building itself. The health problems were then presented as a work environment issue to the employer, who responded by ordering the first investigation of the ventilation in 1987. According to the interviews, however, the investigations of the possible causes of the health problems were not carried out in an energetic way until the local OSH took a forceful position, which obliged the property owner to initiate the second ventilation investigation in 1989-90, as well as the other investigations made by his consultants in 1994-95.
The manner in which the remedial actions were carried out was severely criticised by all interviewees. By his actions, the property owner was considered to have demonstrated a lack of true interest in the health of his tenants. This was considered evidenced by the fact that potentially harmful and extensive repairs were undertaken in December 1994 to March 1995 while the building was normally occupied. The employees felt that the fact that the construction workers, who in February 1996 removed the inner ceilings for ventilation tube inspection, wore respiratory protection, while they themselves continued their normal activities unprotected once again provided evidence of a singular lack of concern for their health (see quotes in Vignette). Still further indications of this attitude were considered to be provided by a letter written by the environmental specialist of the property management company to its regional manager (Figure 1). All this seems to have resulted in the personnel having doubts as to whether the property owner really had found the causes of their health problems, or, if he had done so, he was suspected of having failed to deal with them effectively. The reasons for this presumed neglect were thought to be economic (see quotes in Vignette). The prerequisites for moving back into the building were felt to no longer exist: "I dont know if you should experiment with peoples health, we have been evacuated twice, now that could be enough."
Vignette: Quotations: credibility and economical issues.
|" Its like they dont take this seriously, they say that they have found the causes to the illnesses in this house, that it is putty and glue which have made us sick, and then they start taking away carpets and floors, and really let the poisons out when people are there, no, they dont take this seriously"|
not have SBS
|" I tell you, the psychological situation today for that house is zero. There has been so much going back and forward, I think that even if they succeed with their repairs, they have to bring in new people, because, I mean this is something that has been built up for years, so you are affected in a psychological way, too."|
not have SBS
|" Hell, we are grown up people, we are not deaf, all the time they come with double talk, and evidently they do not realise what they themselves have said or written. But pedagogical problems, they probably mean that we are a little, so to speak, paranoid, that it amuses us to complain."|
not have SBS
|" you hesitate to evacuate people during the repair, especially when the property owner and his consultants assure you that the personnel could keep working in the house without problems. You see, evacuations cost a lot, and these days, we have to save money, people are about to be dismissed because of lack of money."|
|"These are very sensitive things, I mean this is nothing unique for this office, there are lots of houses, but people dont dare touch it, it is so costly, and these days with unemployment."|
|"If this had been a farmhouse with 200 animals, looking tired, having red eyes, snoring noses, and vertigo, well, I tell you that the farmer would turn the building upside down in order to find out whats causing the problems. And then, if he finds it beneath an inner flooring he would remove every single cow before doing anything more, but we, we had to stay there."|
Figure 1. Quote from letter from the company´s environmental specialist to the regional manager
Several factors have been found to be associated with SBS: mechanical ventilation (10), particularly air conditioning (19), steam and evaporative humidification (20), volatile organic compounds (20), illumination (22), formaldehyde (23), dust, wall-to-wall carpets and textiles (24-6), noise (24), indoor temperature (27), work with photocopying (28), subordinate position in the work hierarchy (8), gender, tobacco smoking, atopic disease history and psychosocial discontent (21). The findings are often contradictory (29), and the scientific knowledge that would be able to support corrective measures is poorly developed (13).
This could be due to methodological problems related to the subjective character of the SBS symptoms, which makes the syndrome difficult to delimit. The non-specific symptoms are common and highly prevalent in general populations with no relation to buildings (30). The diagnosis is based on non-specific symptoms, the presence of similar symptoms in co-workers, improvement of symptoms when away from the building, a lack of demonstrable biological markers, and the absence of any other likely diagnosis (31). With such diagnostic criteria, it is rather the system, i.e. the building inclusive of its population which is afflicted by the syndrome (32). Because of that, diagnoses cannot be made on an individual level, making longitudinal cohort or case-control studies difficult to realise. Much of the research on SBS is therefore based on cross-sectional designs with non-specific outcomes and crude exposure measures, which often consist of the building as such or its type of ventilation or the quantity of air turnover (11). When measurements of hypothesised relevant exposures are made, consistent associations between the measurements and the symptoms are rarely found (4). This might depend on the inability to measure the right substances or because it is not possible to record brief, intermittent concentration changes in the substances measured (33). Taken together, this leads to unusually large sources of systematic errors in the application of epidemiological methods which is why even those conclusions that can be drawn from the more consistent findings are vague.
The case building
In cases of sick buildings it seems prudent to identify and eliminate water damage, secure an effective ventilation and eliminate materials known to emit gases (33,34). Such actions are rational as there seem to be reasons for the acceptance of causal links between damp buildings and ill-health (35) and because ventilation appears as one of the most consistent SBS-related factors (10,19). Sometimes they are successful, but often they fail (36) as in the case presented. Could any reasonable explanation for this be found?
From 1987 to 1995, a number of suspected defects were successively found and dealt with. It was not unreasonable to believe that the faults encountered had contributed to the high prevalence of symptoms demonstrated in both 1988 and 1995. However, after a protracted period of repair, rational in relation to the suspected causes, symptoms returned 2 months after the building was reoccupied, and the perceived symptoms resulted in a new and permanent evacuation at the beginning of 1996. The interviews suggested that, by then, the personnel had lost faith in the property owners goodwill and repair actions. This mistrust seems to be founded on the interpretation made of his pronouncements and actions during the course of the buildings repair. The opinion that the owner really was not interested in their health and perhaps even refused to believe in the true nature of their complaints is not illogical, if seen from the background of the long-standing health problems, the initially fairly modest actions taken and the contradictory messages that were expressed in words and actions during the investigations and repairs 1994-1996 (Figure 1, Vignette). Finally, the interpretation of the proprietors actions and messages seems to have resulted in mistrust which prevented the personnel moving back to the building.
Biomedical and psychosocial models of symptom development
As a pathophysiological mechanism of SBS, it has been suggested that the symptoms are caused by a disturbance in sensory perception (37). Another mechanism, neurogenic switching (38), has recently been proposed. An impulse at one site (caused, for example, by chemical irritation in the respiratory mucosa) is thought to be redirected by means of liberated neuropeptides to some distant organ, for example the brain, where they are assumed to cause symptoms such as for example, headache. Such biomedical models are compatible (4,39,40) with the assumption that exposure to chemical emissions from building structures can cause SBS symptoms. Correlations between SBS symptoms and ventilation parameters and/or damp buildings (10,34,35) give some indirect support for such assumptions, which were also the point of departure for the repairs of the case office.
However, the symptoms returned despite the repairs. With a biomedical model, this might be supposed to be due to a sensitisation of the personnel after many years in the building, leading to reactions because of even minor exposure to hypothesised, unknown factors. Such explanations for why SBS often (36) remains in spite of ambitious and rational repair actions have neither been supported nor dismissed by empirical research.
However, in SBS research, there is empirical support (12,41,42) for psychosocial explanatory models. The opinion has been advanced that some cases can be psychogenic in origin or at least psychogenically mediated (43). In the case building, symptoms had existed for years and, from the personnels perspective, a comprehensible, well-founded mistrust and disappointment regarding the property owners attitude and repair actions had developed. In this context, psychosocial models provide equally reasonable explanations for continued symptom prevalence as do biomedical models. Within a psychosocial model, the reappearance of symptoms on returning to the repaired building could be seen as a stress reaction provoked by the interpretation that the health problems were not taken seriously by the proprietor. The links between the symptoms and the psychosocial environment could be various (12). First, the latter may directly act as a stressor and cause symptoms through psychophysiological mechanisms. Second, it may make the individual more sensitive to existent physical and chemical risk
An integrated biopsychosocial model of symptom development
Using the case study as a basis, an integration of the biomedical and the psychosocial perspectives on SBS development is proposed. A variable complex of physical factors (for example, moisture and unknown chemical emissions from putty) is assumed to lead to symptoms initially. However, it is also assumed that another variable complex of non-physical factors, such as, for instance, conflictive agendas between owners of buildings, employers and employees, can lead to intensification and maintenance of symptoms generated by the first factors. The complexes interact in terms of symptom development. Their respective significances for the occurrence of symptoms can in one and the same case vary from time to time. In the case presented, such an integrated perspective seems to explain the buildings disease history better than either a biomedical or a psychosocial model would do alone. Towards the end of the history, factors pertaining to the second complex seem to have become rapidly ever more important for the continuing symptoms. The development towards reduced credibility within the building is believed to have finally become the dominant reason why it became irremediably ill in spite of ambitious and, according to a biomedical model, rational remedial actions taken by the property owner.
Quantitative studies of sick buildings seem to fail to appreciate the social dynamics from which neat tidy data are gathered (44). By offering deeper perspectives on the evolution of SBS, case studies might contribute to its understanding and facilitate the creation of models for practical use. The present study suggests that neglect of a psychosocial perspective could lead to a building becoming chronically sick in spite of, from a biomedical perspective, rational remedial actions. Therefore, it suggests the use of an integrated biopsychosocial perspective, which implies that suspected causative factors should be removed rapidly and with all efforts made to ensure effective communication in such ways that trust is maintained between proprietor, employer and employees.
1. Shorter E. From paralysis to fatigue: a history of psychosomatic illness in the modern era. New York: Free Press, 1992.
2. Shorter E. Multiple chemical sensitivity: pseudodisease in a historical perspective. Scand J Work Environ Health 1997; 23 suppl 3: 35-42.
3. Bergquist U, Vogel E (eds). Possible health implications of subjective symptoms and electromagnetic fields. A report by a European group of experts for the European Commission, DG V, Stockholm; Arbetslivsinstitutet, Arbete och Hälsa 1997:19, 1997.
4. Apter A, Bracker A, Hodgson M, Sidman J, Leung W-Y. Overview. Epidemiology of the sick building syndrome. J Allergy Clinical Immunol 1994; 94 Suppl 2 pt 2:277-288.
5. Scheff TJ, Starrin B. Qualitative methods in the health sicences. Eur J Publ Hlth 1997;7:355-356.
6. Spurgeon A, Gompertz D, Harrington JM. Modifiers of non-specific symptoms in occupational and environmental syndromes. Occup Environ Med 1996; 53: 361 -366.
7. Norlén U, Andersson K (red) Bostadsbeståndets inneklimat (Indoor climate in dwellings) Elib-rapport nr 7. Statens Institut för Byggnadsforskning, Gävle 1993 (in Swedish).
8. Burge PS, Hedge A, Wilson S. Sick building syndrome: a study of 4373 office workers. Appl Ind Hyg 1987; 31: 493-504.
9. Kreiss K. The epidemiology of building-related complaints and illness. Occup Med 1989; 4: 575 - 592.
10. Mendell MJ. Non-specific symptoms in office workers: A review and summary of the epidemiologic literature. Indoor Air 1993; 3: 227-236.
11. Stenberg B. Office illness. The worker, the work and the workplace. Umeå University Medical Dissertations, New Series No 399, Umeå 1994.
12. Eriksson N. The psychosocial work environment and illness among office workers. Doctoral thesis at the Department of Sociology, Umeå University, Nr 2, 1996.
13. Kreiss K. The sick building syndrome in office buildings- a breath of fresh air. New Engl J Med 1993; 328: 877-878.
14. World Health Organization (WHO). Indoor Air Pollutants: Exposure and Health Effects. Copenhagen: WHO 1983. EURO Reports and Studies,no 78.
15. Yin RK. Case Study Research. Design and Methods. Applied social research series. Volume 5. Beverly Hills: sage Publications, 1987.
16. Andersson K, Fagerlund I, Stridh G, Larsson B. The MM-questionnaires: a tool when solving indoor climate problems. Örebro (Sweden): Department of Occupational and Environmental Medicine, Örebro Medical Center Hospital, 1993.
17. Elder NC, Miller WL. Reading and evaluating qualitative research studies. J Family Pract 1995;41:279-285.
18. Berglund B, Johansson I. Health effects of volatile organic compounds in the indoor air. Final report to the National Board of Health and Welfare. Archives of the Center for Sensory Research 1996; 3, Issue 1. Stockholm University and Karolinska Institute, 1996.
19. Mendell MJ, Smith AH. Consistent pattern of elevated symptoms in air-conditioned office buildings: A reanalysis of epidemiologic studies. Am J Pub Health 1990; 80: 1193-1199.
20. Jaakkola JJ, Miettinen P. Type of ventilation system in office buildings and sick building syndrome. Am J Epidemiol 1995; 141:755 - 765.
21. Norbäck D, Torgén M, Edling C. Volatile organic compounds, respirable dust, and personal factors related to prevalence and incidence of sick building syndrome in schools. Br J Ind Med 1990; 47: 733-741.
22. Robertson AS, McInnes M, Glass D, Dalton G, Sherwood Burge P. Building sickness, are symptoms related to the office lighting? Ann Occup Hyg 1989; 33:47-59.
23. Olsen JH, Dössing M. Formaldehyde induced symptoms in day care centers. Am Ind Hyg J 1982; 43:366-370.
24. Skov P, Valbjörn O, Pedersen BV. Influence of indoor climate on the sick building syndrome in an office environment. Scand J Work Environ Health 1990; 16:363-371.
25. Nexö E, Skov P, Gravesen S. Extreme fatigue and malaise - a syndrome caused by malcleaned wall-to-wall carpets? Ecology of Disease 1985; 2: 415-418.
26. Jaakola JJK, Tuomaala P, Seppänen O. Textile wall materials and sick building syndrome. Arch Environ Health 1994; 49: 175-181.
27. Jaakola JJK, Heinonen OP, Seppänen O. Sick building syndrome, sensation of dryness and thermal comfort in relation to room temperature in an office building: need for individual control of temperature. Environmental International 1989; 15: 85-94.
28. Skov P, Valbjörn O, Pedersen BV. Influence of personal characteristics, job-related factors and psychosocial factors on the sick building syndrome. Scand J Work Environ Health 1989; 15:286-295.
29. Thörn Å. Case report on a sick building: Analysis and interpretation in the context of its disease history. Scand J Soc Med 1994; 22: 228-34.
30. Lipscombe JA, Satin KP, Neutra RR. Reported symptom prevalence rates from comparison populations in community-based environmental studies. Arch Environ Health 1992; 47: 263 - 269.
31. Redlich CA, Sparer J, Cullen MR. Sick-building syndrome. Lancet 1997;349:1013-16.
32. Thörn Å. The sick building syndrome: A diagnostic dilemma. Soc Sci Med 1998;47:1307-12.
33. Menzies D, Bourbeau J. Building-related illnesses.New Engl Med J1997; 337:1524-1531.
34. Hodgson M. The sick-building syndrome. State of Art Rev Occup Med 10:167-175, 1995.
35. Howden-Chapman P, Isaacs N, Crane J, Chapman R. Housing and health: the relationship between research and policy. International Journal of Environmental Health Research 1996; 6:173-185.
36. Rothman AL, Weintraub MI. The sick building syndrome and mass hysteria. Neurologic Clinics 1995; 13: 405-412.
37. Berglund B, Lindvall T. Sensory criteria for health buildings. Indoor Air 90. Plenary lectures. V. pp 65-79. Canada Mortgage and Housing Corporation. Ottawa, 1990.
38. Meggs WJ. Neurogenic switching: A hypothesis for a mechanism for shifting the site of inflammation in allergy and chemical sensitivity. Environmental health Perspectives 1995; 103: 54-56.
39. Stolwijk JAJ. Sick-building syndrome. Environ Health Perspect 1991; 95:99-100.
40. Sundell J, Kjellman N-IM. Inomhusmiljöns betydelse för allergi och annan överkänslighet. En kunskapssammanställning (The importance of the indoor environment for allergy and other hypersensitivity. A sate-of-the-art review), National Institute of Public Health, 1994 (in Swedish).
41. Salvaggio JE. Psychological aspects of "environmental illness", "multiple chemical sensitivity" and building-related illness. J Allergy Clin Immunol 1994;94(2 Pt2):366-77.
42. Ryan CM, Morrow LA. Dysfunctional buildings or dysfunctional people: An examination of the sick building syndrome and allied disorders. J Consult Clin Psychol 1992;60:220-24.
43. Bachmann MO, Myers JE. Influences on sick building syndrome symptoms in three buildings. Soc Sci Med 1995;40:245-51.
45. Crawford JO, Bolas SM. Sick building syndrome, work factors and occupational stress. Scand J Work Environ Health 1996;22:243-50.