1494
that allergy is responsible for most of the features. Ventilation engineering and architecture do not have a strong tradition of human-based research, and building services management research is in its infancy. In the UK, a parliamentary select committee is looking into the health effects of the indoor environment.21 Meanwhile, the Government, despite being the largest employer of office workers in Britain, is resistant to change. Thus Government buildings have more problems man those in the private sector,6 yet the Department of the Environment has decided that the recommendations for better building service design and management should be left to "those setting professional standards", and "consideration" should be given as to whether further research is required. Recommendations for improved design and maintenance standards were not supported.22 The solutions include treating the workers’ environment as seriously as that of computers, returning control of the environment to individual workers (ventilation engineering standards require only 80% of workers to be satisfied), and providing better training, careers, and supervision for building services workers, who should relate to the occupants of the building as well as to the owners. There are many lessons for the design of new buildings. Occupational health services have given low priority to office workers in the past; a recent NATO expert group recommended occupational surveillance of office workers with regular questionnaires.23 Employers would then know which of their buildings were sick or healthy, could make informed management decisions when new buildings were required, and know where remedial action should be taken. 1. Franck C.
Eye symptoms and signs in buildings with indoor climate problems ("office eye syndrome"). Acta Ophthalmol 1986; 64: 306-11.
2. Franck C, Skov P. Foam at inner eye canthus in office workers, compared with an average Danish population as control group. Acta Ophthalmol 1989; 67: 61-68. 3. Franck C, Skov P. Evaluation of two different questionnaires used for diagnosing ocular manifestations in the sick building syndrome on the basis of an objective test. Indoor Air 1991; 1: 5-11. 4. Burge PS, Finnegan M, Horsfield N, et al. Occupational asthma in a factory with a contaminated humidifier. Thorax 1985; 40: 248-54. 5. Burge PS, Robertson AS, Hedge A. Validation of self-administered questionnaire in the disgnosis of sick building syndrome. Indoor Air 1990; 1: 575-80. 6. Burge PS, Hedge A, Wilson S, Harris Bass J, Robertson AS. Sick building syndrome; a study of 4373 office workers. Ann Occup Hyg 1987; 31: 493-504. 7. Anderson I, Lundquist GR, Proctor DF. Human perception of humidity under four controlled conditions. Arch Environ Health 1973; 26: 22-27. 8. Burge PS, Jones P, Robertson AS. Sick building syndrome; environmental comparisons of sick and health buildings. Indoor Air 1990; 1: 479-83. 9. Skov P, Valbjørn O. The sick building syndrome in the office environment; the Danish town hall study. Environ Int 1987; 13: 339-44. 10. Skov P, Valbjørn O, Pedersen BV Influence of personal characteristics, job-related factors and psychological factors on the sick building syndrome. Scand J Work Environ Health 1989; 15: 286-95 11. Skov P, Valbjørn O, Pedersen BV. Influence of indoor climate on the sick building syndrome in an office environment. Scand J Work Environ Health 1990; 16: 363-71. 12. Finnegan MJ, Pickering CAC, Burge PS. Sick building syndrome; prevalence studies. Br Med J 1984; 289: 1573-75.
13. Harrison
J, Pickering CAC, Faragher EB, Austwick PKC. An investigation of the relationship between microbial and particulate indoor air pollution and the sick building syndrome. Indoor Air 1990; 1:
149-54 14. Preller L, Zweers T, Boleij JSM, Brunekreef B. Gezondheidsklachten en klachten over het binnenklimaat in kantoorgebouwen. DirectoraatGeneraal van de Arbeid 1990; S 83. 15. Mendell MJ, Smith AH. Consistent pattern of elevated symptoms in air-conditioned office buildings: a reanalysis of epidemiological studies. Am J Publ Health 1990; 80: 1193-99. 16. Jaakkola JJK, Heinonen OP, Seppanen 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. Environ Int 1989; 15: 163-68. 17. Jaakkola JJK, Heinonen OP, Seppanen O. Mechanical ventilation in office buildings and the sick building syndrome: an experimental and epidemiological study. Indoor Air 1991; 2: 111-21. 18. Harrison J, Pickering CAC, Finnegan MJ, Austwick PKC. The sick building syndrome: further prevalence studies and investigation of possible causes. Indoor Air 1987; 2: 487-91. 19. Norback D, Edling C. Environmental, occupational, and personal factors related to the prevalence of sick building syndrome in the general population. Br J Indust Med 1991; 48: 451-62. 20. Hedge A, Burge PS, Robertson AS, Wilson S, Harris-Bass J. Work related illness in offices: a proposed model of the sick building syndrome. Environ Int 1989; 15: 143-58. 21. Rossi H. Environment Committee. Sixth report: indoor pollution. London: HM Stationery Office, 1991. 22. Department of the Environment. Indoor pollution. London: HM Stationery Office, 1991. 23. Robertson AS. Medical management of building related complaints and illnesses. In: Levy F, ed. Epidemiology and medical management of building-related complaints and illnesses. Oslo: NATO (in press).
Seeing the way forward for treatment of CMV retinitis Yet another trial of an antiviral drug in AIDS patients has been stopped prematurely by the data and safety monitoring board. What are the implications for clinical practice? The trial in question’ compared ganciclovir with foscamet for the treatment of cytomegalovirus (CMV) retinitis in patients with AIDS. Two-hundred-andforty patients with a first episode of retinitis were randomised to receive one drug as both induction (60 mg/kg foscamet three times daily or 5 mg/kg ganciclovir twice daily) and maintenance (90 mg/kg once daily foscamet or. 5 mg/kg once daily ganciclovir). Patients who manifested side-effects or whose retinitis progressed could be given the alternative drug. Both drugs initially brought the retinitis under control and kept the progression of retinitis suppressed for similar periods, but on average, patients receiving foscamet survived for significantly longer (12 months) than did those receiving ganciclovir (8 months). Consequently, the trial was stopped and an alert was issued to 40 000 physicians in the USA, advising that "foscamet may be better initial treatment than ganciclovir". If one assumes that the patients were well matched in the two treatment arms and that the deaths were not attributable to toxicity from either drug, can the reported effect on survival be biologically plausible when CMV retinitis is not itself life-threatening? It is possible that foscamet treatment suppressed CMV disease elsewhere (eg, in the gastrointestinal tract) and so prevented or delayed another common cause of death in AIDS patients.2Alternatively, foscamet may
1495
have preferentially inhibited a putative co-factor effect of CMV which has been reported to accelerate the rate of progression of human immunodeficiency virus (HIV) infection.3In either case, one has to ask why foscamet could control these CMV effects better than ganciclovir. Is it possible that some patients had received ganciclovir previously for CMV disease other than retinitis and so were more likely to have resistant strains?4 Finally, foscamet may have had a direct effect on HIV itself through inhibition of reverse
immunodeficiency syndrome. Arch Pathol Lab Med 1988; 112: 540-44. A, Lee CA, Cook DG, et al. Cytomegalovirus infection and progression towards AIDS in haemophiliacs with human immunodeficiency virus infection. Lancet 1989; ii: 63-65. 4. Erice A, Chou S, Biron KK, Stanat SC, Balfour HH, Jordan MC. Progressive disease due to ganciclovir-resistant cytomegalovirus in immunocompromised patients. N Engl J Med 1989; 320: 289-93. 5. Majumdar C, Abbotts J, Broder S, Wilson SH. Studies on the mechanism of human immunodeficiency virus reverse transcriptase. J Biol Chem 1988; 263: 15657-65. 6. Fischl MA, Richman DD, Grieco MH, et al. The efficacy of azidothymidine (AZT) in the treatment of patients with AIDS and AIDS-related complex. N Engl Med 1987; 317: 185-91. J 3. Webster
transcriptase.5
There are also reasons why the observed survival benefit of foscamet could be spurious. Zidovudine significantly delays death in AIDS patients,6yet this drug would be less well tolerated in those allocated to receive ganciclovir because both drugs are myelotoxic. Furthermore, patients in the trial who were intolerant of zidovudine could be given dideoxycytidine or dideoxyinosine, drugs whose efficacy has not been proven in controlled trials and whose efficacy by comparison with zidovudine is unknown. Clearly, rigorous statistical analyses of these confounding variables will be needed to ensure that the better survival was not attributable to unequal antiretroviral therapy in the two groups. However, even if this was shown to be the explanation for the results, the conclusion that foscamet is better than ganciclovir in clinical practice could still be correct if doses of antiretrovirals were indeed reduced because of
ganciclovir toxicity. This study shows, once again, how controlled trials in AIDS patients can identify therapeutic advantages more rapidly than can anecdotal clinical experience; the results also show that unexpected benefits (eg, survival) may emerge. The reported survival benefit has been described only in patients with CMV retinitis, so the trial results have no relevance for AIDS patients without this complication. Furthermore, neither foscamet nor ganciclovir is the ideal drug for CMV retinitis since both agents have side-effects and have to be given by intravenous infusion. Other drugs with anti-CMV activity are clearly needed; meanwhile, results of trials of oral ganciclovir are awaited. What about the ethics of stopping the trial prematurely and issuing an alert to physicians many months before they can hope to see the detailed analyses presented in a peer-reviewed journal? We believe that the correct balance has been struck here between the science and art of medicine. The scientific data will be published in due course and will be examined in great detail to determine whether the initial conclusions of the investigators can be supported. Until then, the researchers have given their colleagues a medical opinion about the best means of treating a serious infection which, like all medical opinions, does not come with a guarantee. 1. Foscavir’s survival benefit in CMV retinitis. SCRIP Oct 30, 1991: 22. 2. Klatt EC, Shibata D. Cytomegalovirus infection in the acquired
Should trials carry
a
health
warning?
Wilt thou forgive that sinne which I did shunne A yeare, or two: but wallowed in, a score?- JOHN DONNE
Preventive cardiology has rightly been seen as the ultimate answer to coronary artery disease. Nevertheless, it is important that attempts at riskfactor reduction are subjected to the rigours of clinical trials. The first convincing demonstration of this principle came from the World Health Organisation clofibrate trial.12In this study, the success of the lipid-lowering agent in decreasing plasma cholesterol and reducing cardiac events and cardiac mortality in a high-cholesterol group was counterbalanced by an increase in all-cause mortality. This unexpected outcome has given rise to much speculation but as yet
wholly convincing explanation. Superficially, the report by Strandberg and colleagues in Helsinki3 of an increase in long-term mortality in the intervention group after a five-year multifactorial primary prevention trial in middle-aged no
suggests a rerun of the clofibrate controversy. The immediate temptation is to pour it into the already large pool of existing intervention trial data, which overall shows benefit from selective risklowering strategies.4,5 Yet on closer inspection the differences between Strandberg’s results and those of the clofibrate trial and other multifactorial interventions6 are much more striking than the similarities. In the Finnish study, participants were randomised either to an intensive risk-factor-lowering strategy-including control of hypercholesterolaemia (principally with probucol or clofibrate), blood pressure reduction (propranolol and/or diuretics), and regular counselling--or to a control group. The controls received "usual health care". Although there was a substantial reduction in the estimated coronary artery disease risk score, at the end of the five-year trial period there was no significant difference in cardiovascular or all-cause mortality between the groups.’ However, after the trial ended the cumulative mortality curves of the intervention and control groups began to diverge, and at ten-year follow-up there was a marginally significant increase in total mortality and a highly significant increase in cardiovascular deaths and deaths associated with violence or accident in the intervention group. Simultaneously, the trialists noted that there was a men
that allergy is responsible for most of the features. Ventilation engineering and architecture do not have a strong tradition of human-based research, and building services management research is in its infancy. In the UK, a parliamentary select committee is looking into the health effects of the indoor environment.21 Meanwhile, the Government, despite being the largest employer of office workers in Britain, is resistant to change. Thus Government buildings have more problems man those in the private sector,6 yet the Department of the Environment has decided that the recommendations for better building service design and management should be left to "those setting professional standards", and "consideration" should be given as to whether further research is required. Recommendations for improved design and maintenance standards were not supported.22 The solutions include treating the workers’ environment as seriously as that of computers, returning control of the environment to individual workers (ventilation engineering standards require only 80% of workers to be satisfied), and providing better training, careers, and supervision for building services workers, who should relate to the occupants of the building as well as to the owners. There are many lessons for the design of new buildings. Occupational health services have given low priority to office workers in the past; a recent NATO expert group recommended occupational surveillance of office workers with regular questionnaires.23 Employers would then know which of their buildings were sick or healthy, could make informed management decisions when new buildings were required, and know where remedial action should be taken. 1. Franck C.
Eye symptoms and signs in buildings with indoor climate problems ("office eye syndrome"). Acta Ophthalmol 1986; 64: 306-11.
2. Franck C, Skov P. Foam at inner eye canthus in office workers, compared with an average Danish population as control group. Acta Ophthalmol 1989; 67: 61-68. 3. Franck C, Skov P. Evaluation of two different questionnaires used for diagnosing ocular manifestations in the sick building syndrome on the basis of an objective test. Indoor Air 1991; 1: 5-11. 4. Burge PS, Finnegan M, Horsfield N, et al. Occupational asthma in a factory with a contaminated humidifier. Thorax 1985; 40: 248-54. 5. Burge PS, Robertson AS, Hedge A. Validation of self-administered questionnaire in the disgnosis of sick building syndrome. Indoor Air 1990; 1: 575-80. 6. Burge PS, Hedge A, Wilson S, Harris Bass J, Robertson AS. Sick building syndrome; a study of 4373 office workers. Ann Occup Hyg 1987; 31: 493-504. 7. Anderson I, Lundquist GR, Proctor DF. Human perception of humidity under four controlled conditions. Arch Environ Health 1973; 26: 22-27. 8. Burge PS, Jones P, Robertson AS. Sick building syndrome; environmental comparisons of sick and health buildings. Indoor Air 1990; 1: 479-83. 9. Skov P, Valbjørn O. The sick building syndrome in the office environment; the Danish town hall study. Environ Int 1987; 13: 339-44. 10. Skov P, Valbjørn O, Pedersen BV Influence of personal characteristics, job-related factors and psychological factors on the sick building syndrome. Scand J Work Environ Health 1989; 15: 286-95 11. Skov P, Valbjørn O, Pedersen BV. Influence of indoor climate on the sick building syndrome in an office environment. Scand J Work Environ Health 1990; 16: 363-71. 12. Finnegan MJ, Pickering CAC, Burge PS. Sick building syndrome; prevalence studies. Br Med J 1984; 289: 1573-75.
13. Harrison
J, Pickering CAC, Faragher EB, Austwick PKC. An investigation of the relationship between microbial and particulate indoor air pollution and the sick building syndrome. Indoor Air 1990; 1:
149-54 14. Preller L, Zweers T, Boleij JSM, Brunekreef B. Gezondheidsklachten en klachten over het binnenklimaat in kantoorgebouwen. DirectoraatGeneraal van de Arbeid 1990; S 83. 15. Mendell MJ, Smith AH. Consistent pattern of elevated symptoms in air-conditioned office buildings: a reanalysis of epidemiological studies. Am J Publ Health 1990; 80: 1193-99. 16. Jaakkola JJK, Heinonen OP, Seppanen 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. Environ Int 1989; 15: 163-68. 17. Jaakkola JJK, Heinonen OP, Seppanen O. Mechanical ventilation in office buildings and the sick building syndrome: an experimental and epidemiological study. Indoor Air 1991; 2: 111-21. 18. Harrison J, Pickering CAC, Finnegan MJ, Austwick PKC. The sick building syndrome: further prevalence studies and investigation of possible causes. Indoor Air 1987; 2: 487-91. 19. Norback D, Edling C. Environmental, occupational, and personal factors related to the prevalence of sick building syndrome in the general population. Br J Indust Med 1991; 48: 451-62. 20. Hedge A, Burge PS, Robertson AS, Wilson S, Harris-Bass J. Work related illness in offices: a proposed model of the sick building syndrome. Environ Int 1989; 15: 143-58. 21. Rossi H. Environment Committee. Sixth report: indoor pollution. London: HM Stationery Office, 1991. 22. Department of the Environment. Indoor pollution. London: HM Stationery Office, 1991. 23. Robertson AS. Medical management of building related complaints and illnesses. In: Levy F, ed. Epidemiology and medical management of building-related complaints and illnesses. Oslo: NATO (in press).
Seeing the way forward for treatment of CMV retinitis Yet another trial of an antiviral drug in AIDS patients has been stopped prematurely by the data and safety monitoring board. What are the implications for clinical practice? The trial in question’ compared ganciclovir with foscamet for the treatment of cytomegalovirus (CMV) retinitis in patients with AIDS. Two-hundred-andforty patients with a first episode of retinitis were randomised to receive one drug as both induction (60 mg/kg foscamet three times daily or 5 mg/kg ganciclovir twice daily) and maintenance (90 mg/kg once daily foscamet or. 5 mg/kg once daily ganciclovir). Patients who manifested side-effects or whose retinitis progressed could be given the alternative drug. Both drugs initially brought the retinitis under control and kept the progression of retinitis suppressed for similar periods, but on average, patients receiving foscamet survived for significantly longer (12 months) than did those receiving ganciclovir (8 months). Consequently, the trial was stopped and an alert was issued to 40 000 physicians in the USA, advising that "foscamet may be better initial treatment than ganciclovir". If one assumes that the patients were well matched in the two treatment arms and that the deaths were not attributable to toxicity from either drug, can the reported effect on survival be biologically plausible when CMV retinitis is not itself life-threatening? It is possible that foscamet treatment suppressed CMV disease elsewhere (eg, in the gastrointestinal tract) and so prevented or delayed another common cause of death in AIDS patients.2Alternatively, foscamet may
1495
have preferentially inhibited a putative co-factor effect of CMV which has been reported to accelerate the rate of progression of human immunodeficiency virus (HIV) infection.3In either case, one has to ask why foscamet could control these CMV effects better than ganciclovir. Is it possible that some patients had received ganciclovir previously for CMV disease other than retinitis and so were more likely to have resistant strains?4 Finally, foscamet may have had a direct effect on HIV itself through inhibition of reverse
immunodeficiency syndrome. Arch Pathol Lab Med 1988; 112: 540-44. A, Lee CA, Cook DG, et al. Cytomegalovirus infection and progression towards AIDS in haemophiliacs with human immunodeficiency virus infection. Lancet 1989; ii: 63-65. 4. Erice A, Chou S, Biron KK, Stanat SC, Balfour HH, Jordan MC. Progressive disease due to ganciclovir-resistant cytomegalovirus in immunocompromised patients. N Engl J Med 1989; 320: 289-93. 5. Majumdar C, Abbotts J, Broder S, Wilson SH. Studies on the mechanism of human immunodeficiency virus reverse transcriptase. J Biol Chem 1988; 263: 15657-65. 6. Fischl MA, Richman DD, Grieco MH, et al. The efficacy of azidothymidine (AZT) in the treatment of patients with AIDS and AIDS-related complex. N Engl Med 1987; 317: 185-91. J 3. Webster
transcriptase.5
There are also reasons why the observed survival benefit of foscamet could be spurious. Zidovudine significantly delays death in AIDS patients,6yet this drug would be less well tolerated in those allocated to receive ganciclovir because both drugs are myelotoxic. Furthermore, patients in the trial who were intolerant of zidovudine could be given dideoxycytidine or dideoxyinosine, drugs whose efficacy has not been proven in controlled trials and whose efficacy by comparison with zidovudine is unknown. Clearly, rigorous statistical analyses of these confounding variables will be needed to ensure that the better survival was not attributable to unequal antiretroviral therapy in the two groups. However, even if this was shown to be the explanation for the results, the conclusion that foscamet is better than ganciclovir in clinical practice could still be correct if doses of antiretrovirals were indeed reduced because of
ganciclovir toxicity. This study shows, once again, how controlled trials in AIDS patients can identify therapeutic advantages more rapidly than can anecdotal clinical experience; the results also show that unexpected benefits (eg, survival) may emerge. The reported survival benefit has been described only in patients with CMV retinitis, so the trial results have no relevance for AIDS patients without this complication. Furthermore, neither foscamet nor ganciclovir is the ideal drug for CMV retinitis since both agents have side-effects and have to be given by intravenous infusion. Other drugs with anti-CMV activity are clearly needed; meanwhile, results of trials of oral ganciclovir are awaited. What about the ethics of stopping the trial prematurely and issuing an alert to physicians many months before they can hope to see the detailed analyses presented in a peer-reviewed journal? We believe that the correct balance has been struck here between the science and art of medicine. The scientific data will be published in due course and will be examined in great detail to determine whether the initial conclusions of the investigators can be supported. Until then, the researchers have given their colleagues a medical opinion about the best means of treating a serious infection which, like all medical opinions, does not come with a guarantee. 1. Foscavir’s survival benefit in CMV retinitis. SCRIP Oct 30, 1991: 22. 2. Klatt EC, Shibata D. Cytomegalovirus infection in the acquired
Should trials carry
a
health
warning?
Wilt thou forgive that sinne which I did shunne A yeare, or two: but wallowed in, a score?- JOHN DONNE
Preventive cardiology has rightly been seen as the ultimate answer to coronary artery disease. Nevertheless, it is important that attempts at riskfactor reduction are subjected to the rigours of clinical trials. The first convincing demonstration of this principle came from the World Health Organisation clofibrate trial.12In this study, the success of the lipid-lowering agent in decreasing plasma cholesterol and reducing cardiac events and cardiac mortality in a high-cholesterol group was counterbalanced by an increase in all-cause mortality. This unexpected outcome has given rise to much speculation but as yet
wholly convincing explanation. Superficially, the report by Strandberg and colleagues in Helsinki3 of an increase in long-term mortality in the intervention group after a five-year multifactorial primary prevention trial in middle-aged no
suggests a rerun of the clofibrate controversy. The immediate temptation is to pour it into the already large pool of existing intervention trial data, which overall shows benefit from selective risklowering strategies.4,5 Yet on closer inspection the differences between Strandberg’s results and those of the clofibrate trial and other multifactorial interventions6 are much more striking than the similarities. In the Finnish study, participants were randomised either to an intensive risk-factor-lowering strategy-including control of hypercholesterolaemia (principally with probucol or clofibrate), blood pressure reduction (propranolol and/or diuretics), and regular counselling--or to a control group. The controls received "usual health care". Although there was a substantial reduction in the estimated coronary artery disease risk score, at the end of the five-year trial period there was no significant difference in cardiovascular or all-cause mortality between the groups.’ However, after the trial ended the cumulative mortality curves of the intervention and control groups began to diverge, and at ten-year follow-up there was a marginally significant increase in total mortality and a highly significant increase in cardiovascular deaths and deaths associated with violence or accident in the intervention group. Simultaneously, the trialists noted that there was a men
