Author's response to reviews Title: Can air pollution affect tear film stability? A cross-sectional study in the aftermath of an explosion accident Authors: Bente E Moen (bente.moen@isf.uib.no) Dan Norbäck (dan.norback@medsci.uu.se) Gunilla Wieslander (gunilla.wieslander@medsci.uu.se) Jan V Bakke (jan.bakke@atil.no) Nils Magerøy (nils.mageroey@helse-bergen.no) Jens-Tore Granslo (jens-tore.granslo@helse-bergen.no) Ågot Irgens (agot.irgens@helse-bergen.no) Magne Bråtveit (magne.bratveit@isf.uib.no) Bjørg E Hollund (bjoerg.eli.hollund@helse-bergen.no) Tor B Aasen (tor.aasen@helse-bergen.no) Version: 2 Date: 14 October 2010 Author's response to reviews: see over
Editor BMC-series Journals BioMed Central Floor 6, 236 Gray s Inn Road London, WC1X 8HL UK Bergen 14 th Oct 2010 MS: 1774845061371779: Can air pollution desiccate the eyes? A cross sectional study in the aftermath of an explosion accident. Bente E. Moen, Dan Norbäck, Gunilla Wieslander, jan V. Bakke, Nils Magerøy, Jens Tore Granslo, Ågot Irgens, Magne Bråtveit, Bjørg Eli Hollund and Tor B. Aasen. Thank you for the opportunity to revise this manuscript. We are grateful for all the comments, and have followed them all carefully. The changes are marked with red in the original manuscript. On behalf of all the authors, I present the responses to the reviewer's reports: The reviewers responses are in italic and I have marked our response to the reviewers comments with R. 1. Reviewer: Otto Schmut The major question is, why NIBUT and SBUT were shorter in the exposed workers group only and why no significant relationships were observed in woman, generally more affected with dry eyes. R: This is an important issue. We do not know the exact answer, but it might have to do with differences in work. In the exposed group of workers, the male workers have occupations that work more outside than the females ones who work in offices. We have added information about this fact in the results, page 9, 2 nd para. The difference in these findings might also be caused by a difference in number, as the number of female workers is low compared to the male number, causing less statistical power. We have included more text in the discussion about the gender differences, page 11, 2nd para. There might be other factors accountable fort he reduction of NIBUT and SBUT in exposed workers. An analysis of the correlation of shorter NIBUT and SBUT and smoking would be of interest, since at the first glance it is obvious that the group of exposed workers comprises more present smokers. This would be rather an explanation for reduction of BUT than analysis of the explosion consequences after such a long period of time. R :It is correct that there are more daily smokers among the exposed workers, as shown in Table 1. For this reason we added smoking in the statistical regression model. This is described in the results. However, we 1
understand that this was not clearly explained, so we have now added more information about the regression analysis in Statistical analyses, page 8, 1th para. The article does not answer the question why fire fighting, cleaning up the area did not influence the results. Table missing. R: We have now added a paragraph in the Results about these workers, page 8.The reason why we did not include these workers is explained here, it was that their exposure was very short. Most of them stayed there 1-2 days only, and left the area afterwards. The examination took place 1.5 years after the explosion. The other participants were located in the polluted area also after the explosion, either living or working there, and had quite another exposure situation. We have removed all other confusing information about this special group of workers. There is so much more speculation than hard facts. The autors fair-minded mention the obvious lacks of the study in their discussion. The only positive result seems to me the last paragraph of the DISCUSSION that tear film examinations in studies of health problems associated with pollution should be performed in future. However, these studies must be better planned than the presented study. All together the Conclusions in the Abstract are not in accordance with the results presented. R: We have shortened the conclusion to make the relation between the results better. Abstract: Conclusions: Line 4: tear film examinations should be performed R:The conclusion is now shortened. Background: Line 13: 0.0 0.5 mg/m3 Line 14: compounds Line 21: were R: This part of the background is now changed. Examinations: Line 18: fluorescein R: This has been corrected. Tables should be shortened and well arranged using abbreviations e.g. UI for unexposed inhabitants; EI for exposed inhabitants; EW for exposed workers. R:The abbreviations have been included in the tables and the explanations to them have been included in the table texts. 2. Reviewer: Britt-Marie Marie Larsson Reviewer's report: 1. As the authors already have stated, the study would have benefited greatly from a more detailed assessment of exposure levels in the ambient air. Preferably performed in connection with studies of health effects. However, some measurements have been done. Why are not any mean or median levels presented? If possible, give reference levels and/or hygienic exposure limits to put the stated levels in a context. R: The mean values were not given, as several of these samples had levels =0. We have rewritten this part of Introduction page 4 and hope the 2
text is more informative now: Twelve measurements of mercaptanes were made in the ambient air a few meters from the explosion area six weeks after the explosion. Mercaptanes were not detected in five of the samples while the other samples showed mercaptane levels of 1-2 percent of the Norwegian limit values for working atmosphere. Hydrocarbon compounds were measured in the air as well, and toluene and xylene were detected at similar low levels in an area close to the tanks. These hydrocarbons were also measured at longer distance from the tanks and were not detectable some hundreds meters away. 2. Where were these measurements performed? In the direct vicinity to the explosion site and/or in the residential area? R:This is now described as cited in your point 1, on page 4 in the manuscript. It is very difficult to describe ore, as the details do not exist. 3. How large land area (and/or sea) was actually affected by the contaminated oil products? We have added some more information on this issue in the beginning of the Background part, page 4, telling that the Black smoke spread over several hundred meters north of the harbor area, as the wind blew in this direction. 4. The response rate was 76 % in the exposed group and 59 % in the unexposed. These figures should be presented in the Result section. R: We have moved this information to the Results section, page 8. Did the authors analyse this discrepancy? Since there were no differences between the groups regarding allergy, pets, moisture at home, age and medication, a possible selection bias cannot be excluded. People with symptoms might be more eager to participate. R: We have added a paragraph in the Discussion about this topic, page 11. 5. Table 2. Since non-parametric methods are used for the statistical analysis, the results are not assumed to be normally distributed. Why does the authors present both mean and median? In order to make the table easier to access, it should sufficient to present median and interquartile range (percentiles). R: We have deleted mean and SD, and kept min-max values. Interquartile ranges were difficult to calculate with the skewed distribution. 6. The number of subjects in each exposure group (men and women) should be given in the table. I assume that the lack of significance regarding NIBUT among women/exposed workers is due to a small sample size. R: The numbers are now given in the table 1. In Discussion page 11, the low number of females is now mentioned. Minor comments: 1. What is the prevalence of eye symptoms in the normal population? This information has been added in the Introduction, page 4, by adding one more reference. 3
2. Were the different occupations equally distributed in the three exposure groups? R: Yes, and we have added some information about this in the Results page 9, 2 nd para. 3. The three exposure groups could be more thoroughly described. In page 6, line 1 it says all persons working or living in an area less than 6 km from the explosion site at the time of the explosion were asked to participate in this cross-sectional study. Was it possible that some of the participants had moved after the accident but still were included in the study? R: 22 persons did move the past year.this is now presented in the Results, page 9, para 1. None moved to the explosion area. 4. A total of 519 persons were included in the study, are contact lens wearers (30 persons) included in this figure? R: Contact lens wearers were not included in the examination of tear film. This is stated in Methods on page 7. 5. Page 11, line 3. Sentence Also, the direct association.., rephrase the sentence, the meaning is somewhat unclear. R: The sentence is removed, as it was not important. 3. Reviewer: Hamid Hosseini Abstract: 1-The following two statements mentioned in abstract are not similar with regard to inclusion criteria and should be clarified. (The present study reports examinations of tear film break up time in the population, as many of the persons in the area complained of sore eyes.) (All persons working or living in an area less than six km from the explosion site were invited to the study together with a similar number of persons matched on age and sex living more than 20 km away.) R: It is correct that we had written this in an inexact manner, and we have revised the information in the Abstract. 2- please revise following sentence: Totally, 734 were examined, response 76 percent. R:The sentence has now been lengthened and more details are given in the abstract. 3- please replace (ocular microscopy) with slit lamp biomicroscopy. R: We did not use a slit lamp. The name of the tear scope and name and address of producer is given in detail in Methods page 7. 4- please revise following statement in abstract conclusion: We suggest that tear film examination are performed in future studies to evaluate health effects from air pollution. R: The conclusion has been shortened to be more precise. Background: 5- The study has been performed in an industrial area (with another company of this type as well as a quarry, cement industry and different transport companies. Is there any information available about other industrial pollution in the area? Does duration of work in contaminated area has been recorded and analyzed? 4
Further information about working condition such as indoor Vs outdoor should be considered. R: We do agree, but there is no information available about the exposure caused by the present industry, neither indoor nor outdoor values. Methods: 6- different response rate in exposed and unexposed subjects is a weak point of this study and should be mentioned. R: This is mentioned in the Discussion, page 11, 2 nd para. 7-please mention inclusion and exclusion criteria (especially with regard to age) clearly. R: The criteria for participation is now clarified better in methods, page 6 for the whole study group (concerning also examinations not presented in the present article) and on page 7. This is also summarized in results page 8. Examinations 8- please provide more details about examination method, were examiners aware of exposure status of subjects? R: The examiners were not blinded during examinations, and this is described in Methods page 8, and also other additional information is given on page 7 9- please mention which eye of each subject has been included, if both eyes have been studied, in cases of different measurement which one has been recorded. R: Only one eye, the right one, was examined, page 7. 10- eyes with ocular surface problems such as pterygium, previous ocular surgery,... can affect tear BUT and should be excluded from the study. I think the examinations should be performed by an ophthalmologist to detect such cases and this point needs to be mentioned as a weak point of the study. The participants were asked for previous eye injuries, this is now mentioned on page 7, this was present only for one person, also mentioned on page 7. Conclusion 12-the main finding of the study is low tear BUT in men working in polluted area and normal tear BUT in peoples living in the area. Therefore these tests cannot be proposed as a suitable screening test in those exposed to chemical pollutions atleast based on results of current study. (because as the authors state abnormal BUT in workers may be due to other work related problems). 13-In tables 1 and 2 sums (total) are not correct. R: We do agree and have removed the statement from the conclusion. We have checked on the sums, and they are correct now. Be aware that in Table 1 there are not sums regarding age and smokers, but mean and percent. 4. Reviewer: Rohit Saxena 1. Title: Is a little severe to use the word desiccate considering the only change presented in the study is change in the BUT values. 5
R: We agree, and have changed the title accordingly to Can air pollution affect tear film stability? The purpose of this study was to assess the relationship between living or working close to the polluted area and tear film stability one and a half years after the explosion. 2. In this regard there is no information in the text to inform about the environmental differences between the area chosen to be close to the polluted area and that away from it. The first two references are not in published English literature so ate difficult to access. The authors should include a table comparing the environmental conditions between the two sites. R: We have added some more information in the introduction about the area and the exposure, page 4. However, it is not possible to describe more hard facts on this issue than we have already done, as very few measurements were performed. The Norwegian reports confirm this. 3. Details about specific malodorous sulphur products in the air would help to identify the suspected chemicals that have caused the observed changes. There are numerous studies documenting the deleterious effects of environmental pollution on the cornea and ocular surface. As this was a specific industrial accident it would help to pick up those chemicals that have a direct impact. R: Again, we are very sorry, but no more information is available about exposure in the air. We found it difficult to include studies of serious chemical accidents with severe cornea injuries, there were not eye injuries here that needed treatment, this was more like experiencing sore and irritated eyes. 4. A correlation between the SBUT and NIBUT for the groups would help to increase the value of these tests. R: We have performed a correlation analysis of SBUT and NIBUT, and they are significantly correlated, with r = 0.3 (Pearson correlation coefficient). We did not include this value in the text, as it is not optimal doing this kind of correlation analysis when the values are skewed. The discussion includes the limitations of the study some of which significantly affect the interpretation of the results. 5. The study has shown that the exposed workers had the most affected result. As the authors maintain this could be due to their nature of work. Attempt can be made to evaluate the exposure to toxins their routine work or to see the SBUT or NIBUT values in a similar group engaged in similar activity. R: We agree that this can be of interest in a future study, and this is stated in the discussion, page 12. All authors have read and accepted the revisions, and we hope that the manuscript now is acceptable for publication in Your Journal. Best regards, Bente E. Moen 6