Document 2jgy9pvkeXbG6nz0Va7NXa1D5

FILE NAME Brakes BRK DATE 2016 DOC BRK250 DOCUMENT DESCRIPTION Journal Article - Factors Associated with Non- Compliance of Asbestos Occupational Standards in Brake Repair Workers Ann Occup Hyg 2016 1-16 10.1093 mew028 Factors Associated With compliance of Asbestos Occupational Standards in Brake Repair Workers Mar^>aFernanda Garc^>aFrank C. Curriero Margarita Giraldo Lorena M^'ndezP,,atrick N. Breysse Mauricio Dur^nC,,arlos A. Duque45 Mauricio Gonz^lezCarolina P^'rezPatricia Parada and Juan Pablo Bonilla,, Downlade from 1.Department 1.Department of Civil and Environmental Engineering Universidad de Los Andes Carrera 1 Este 19A ML328 Bogot^111711 Colombia http 2.Department ofEpidemiology Bloomberg School of Public Health Johns Hopkins University 615 N. Wolfe Street Baltimore MD 21205 USA, 3.Department of Environmental Health Sciences Johns Hopkins University Bloomberg School of Public Health 615 N. Wolfe Street Baltimore MD 21205 USA 4.Medical Department Fundaci^nNeumol^gicaColombiana Cra 13B 161-85 Bogot^110131 Colombia anbyg 5.Research Department Fundaci^nNeumol^gicaColombiana Cia 13B 161-85 Bogot^110131 Colombia Currently employed by the Centers for Disease Control and Prevention Author to whom correspondence should be addressed Tel 571-3394949 fax 571-332-4313 mail jramos@uniandes.edu.co Submitted 8 January 2016 revised 19 April 2016 revised version accepted 25 April 2016 oxfrdjunals org ABSTRACT by Asbestos and asbestos containing brake products are currently used in low- and income guest countries like Colombia Because brake products are distributed detached from their supports they on June require manipulation before installation which release fibers and expose workers Previous studies of our research group have documented exposures in excess of the widely accepted 0.1 cmexposure 8 guideline The aim of this study is to identify factors associated with compliance of the h time weighted average TWA 0.1 cmasbestos occupational limit among brake mechanics i.e. riveters 2016 Eighteen brake repair shops BRS located in Bogot^Colombia were sampled during 3 to 6 consecu- tive days for the entire work Personal and short personal samples were collected following NIOSH methods 7400 and 7402 Longitudinal based logistic regression models were used to deter- mine the association between the odds of exceeding the h TWA 0.1 cmasbestos occupational limit and variables such as type of tasks performed by workers workload number of products manipulated daily years of experience as riveters and shop characteristTihcesse models can be used to estimate the odds of being currently or historically overexposed when sampling data do not exist Since the information required to run the models can vary for both retrospective and current asbestos occupa- tional exposure studies three models were constructed with different information requirements The first model evaluated the association between the odds of compliance with variables related to the workload the second model evaluated the association between the odds of compliance with variables related to the manipulation tasks and the third model evaluated the association between the 'The Author 2016 Published by Oxford University Press on behaofltf he British Occupational Hygiene Society Page compliance of asbestos standards brake mechanics odds of compliance with variables related with both the type of tasks performed by workers and the workload Variables associated with the odds of compliance included conducting at least one manipulation activity with beveling and grinding of asbestos and asbestos containing brake products during the work shift the location of the worker in the shop during manipulation activities cleaning activities of the manipulation area the years of experience working as riveters and the number of asbestos and asbestos containing brake products manipulated daily These models could be useful for current and retrospective occupational studies in determining the odds of compliance of the asbestos occupational limit among brake mechanics KEYWORDS asbestos exposures brake mechanics compliance occupational standards INTRODUCTION Recent studies performed by our research group Asbestos is comprised of a group of minerals that have shown that brake mechanics are exposed to high because of their physical and chemical characteristics asbestos concentrations exceeding in some cases the have been extensively used in different types ofproducts US OSHA PEL and the Colombian standards Cely- Downlade including brake and transmission products ATSDR 2001 OMS 2006 Asbestos is classified into two main groups amphiboles and serpentines ATSDR 2001 Garc^>aet al 2012 Garc^>aet al 2015 These findings are in contrast with studies conducted in income countries Hickish and Knight 1970 Chrysotile is the only type of asbestos of the serpen- Yeung et al 1999 Weir et al 2001 Blake et al from tine group and is currently the type of asbestos with 2003 Paustenbach et al 2003 Paustenbach et al http the most extended use worldwide ATSDR 2001 2004 Blake et al 2006 Madl et al 2009 One of an hyg All types of asbestos are classified as carcinogenic to humans IARC 1998 2012 and asbestos is one of the most important reasons for this difference is that in Colombia asbestos containing brake products are the most important occupational carcinogens OMS distributed by the manufacturer detached from their oxfrdjunals 2006 Asbestos is related with both malignant and support requiring manipulation in the shops before malignant diseases ACCP 2012 IARC 2012 installation in the vehicles Garc^>aet al 2012 The most important asbestos exposures occur in Garc^>aet al 2015 Thus brake products are org occupational environments ACSH 2007 Wagner drilled countersinked riveted beveled and grinded by and Lemen 2008 where the manipulation of asbes- and a specific group of mechanics known as riveters tos containing products may result in excessive asbes- conducts these tasks Garc^>aet al 2012 Cely- guest tos exposures OIT 1986 Most of the legislation Garc^>aet al 2015 An additional manipulation task on and control activities regarding asbestos exposures observed include bonding the brake linings or brake June of brake mechanics and other occupational groups pads to the support in passenger vehicles In this case 8 started in the 1970s OSHA 2013 and in 1994 the riveters have to grind and bevel the brake products in 2016 US Occupational Safety and Health Administration order to remove excess glue Garc^>aet al 2012 OSHA issued the current standards for personal The same riveters that manipulate asbestos containing asbestos exposure in workplaces a personal exposure brake products may also manipulate asbestos con- limit US OSHA PEL of 0.1 cmfor an h time taining brake products It is important to clarify that weighted average TWA based on phase contrast blowing the brake drums with compressed air is not microscopy equivalent PCM asbestos concen- a common practice of brake mechanics in Colombia tration and an excursion limit US OSHA EL of 1 / This study identifies factors associated with personal cmfor a min exposure period PCM asbestos asbestos concentrations that exceed the h TWA of concentration OSHA 1994a 2013 In Colombia 0.1 cmasbestos occupational limit among riveters i.e. an upper income country asbestos expo- sure standard is based on the American Conference that manipulate asbestos containing brake products of Governmental Industrial Hygienists ACGIH threshold limit value TLV of 0.1 cmfor TWA of h MSPS 2011 METHODS Since 2010 several sampling campaigns were conducted in 18 brake repair shops BRS located in Factors associated with compliance of asbestos Page 3 of 16 Bogot^Colombia Nine of these shops service passenger vehicles and nine service heavy vehicles Results of these sampling campaigns were used to longer or shorter personal samples were collected during manipulation activities Each sampling day had one blank sample and each shop had one background establish the factors associated with the odds of non- sample collected during one night of the sampling days compliance of the h TWA occupational standard Detailed activity diaries were filled with all the activities Ethics Committees from the Universidad de Los Andes performed by workers sampled during the entire sam- and Fundaci^nNeumol^gicaColombiana approved pling campaigns Shop's owners and workers sampled the sampling campaigns and the informed consent were informed of the results of the sampling campaigns signed by the riveters that participated in this study Samples were analyzed by Forensic Analytical Laboratories Inc Forensic Analytical Laboratories Sampling campaigns Hayward CA USA an American Industrial Hygiene The identification of the shops sampled was initially Association AIHA accredited laboratory The analy- done using the phone book and in a subsequent stage sis of the samples was performed on a Philips CM12 of the project by walking through the zones of Bogot^ transmission electron microscopy TEM Philips where BRS are clustered to identify additional shops Eindhoven The Netherlands Fibers with length that could potentially participate in the project This second strategy has been performed since 2012 and 5 mwere counted at a magnification of 2500 For fibers close to 5 ...min length fiber dimension was Downlade from includes the application of a structured question- determined by using a magnification of 10 000. The naire adapted from the OSHA Medical Questionnaire accelerating voltage used was 100 keV and an Energy OSHA 1994b to characterize the working conditions Dispersive ray EDXA Evex NanoAnalysis System http of those riveters of the shops who agreed to voluntarily participate in the project The information collected was IV Princeton NJ USA was also used Forensic Analytical Laboratories personal communication anhyg used to identify and select BRS that currently perform Equivalent PCM asbestos concentra- oxfrdjunals manipulation activities of brake products BRS were also classified according to their workload based on the average number of sets of asbestos and asbestos tions were calculated multiplying the phase contrast microscopy PCM fiber concentration by the fraction of asbestos fibers to total fibers determined by containing products that they manipulate daily Transmission Electron Microscopy TEM NIOSH org , Sample collection and analysis was similar to those 1994a 1994b To determine regulatory compliance by guest described in previous studies of our research group with the h TWA of 0.1 cmasbestos occupational Garc^>aet al 2012 Garc^>aet al 2015 limit the h TWA PCM personal asbestos con- on June Salazar et al 2014 Shops were sampled during 3 to centration was estimated for each worker When the 6 consecutive days for the entire work Personal work exceeded h the worst period 8 2016 samples were collected following NIOSH methods 7400 NIOSH 1994a and 7402 NIOSH 1994b sampling windows were used according with the rec- ommendations of the US OSHA Technical Manual Air samples were collected on mm MCE filters OSHA 2008 When the work was shorter 0.45 ...mpore size mounted on 50mm conducting than h i.e. due to overloaded samples or because extension cowls cassettes SKC Preloaded Cassette the work ended early or the worker left the shop SKC using AIRCheck XR5000 pumps SKC Eighty early the h TWA asbestos concentration was cal- Four PA USA that operated at a flow rate of 2LPM culated assuming 0 cmfor the remaining time Standard procedures were applied to measure flow Additionally when the PCM concentration was below rates at the beginning and at the end of each sampling the limit of detection LOD one half of the LOD was window using Defender 510 High or 520 High BIOS used for calculation purposes This last procedure was International Calibrators MesaLabs Butler NJ USA applied because of the results of a sensitivity analysis Personal samples were collected on riveters dur- performed by our research group in a previous study ing the entire work All workers sampled were which showed no important differences in the TWA instructed to keep their activities and working proce- concentration estimated using 0 cmor half the dures as usual Thirty minutes short personal sam- LOD when samples are below the LOD Garc^>a ples were collected during manipulation activities and et al 2015 Page . compliance of asbestos standards brake mechanics Additionally bulk asbestos content of 18 duplicate linings pads or blocks from 12 of the most common brake product's brands that have been used at the shops sampled was estimated following EPA 93-116 PLM method EPA 1993 by an AIHA accredited laboratory Forensic Analytical Laboratories Hayward CA USA performed during each sample collection the location of the worker during manipulation activities and the number of asbestos and asbestos containing products manipulated daily could be identified Since information on all the covariates proposed are not always available especially for retrospective expo- sure assessments three models were evaluated accord- ing with the type of information that may be obtained Data analysis Three longitudinal based logistic regression models and would be available for analysis The first model Model Workload assumes that the information were used to identify variables significantly associated with the odds of compliance of the h TWA 0.1 cmasbestos occupational limit according with the type of information available to develop the models available is related with the work load i.e. how many brake products were manipulated and some characteristics of the workers and shops The second model Model Work tasks assumed that the informa- Eight TWA PCM personal asbestos concen- tion available is related with the work tasks i.e. type of trations below or equal to 0.1 cmwere classified as Downlade in compliance and h TWA PCM personal asbes- tos concentrations exceeding 0.1 cmwere classified from as compliant The h TWA PCM asbestos manipulation performed and some characteristics of the workers and shops The third model Model 3Complete was developed for those cases in which the information regarding both work tasks and workload concentrations calculated for workers sampled were was available http an hyg approximated to three decimals and compared against the 0.100 cmvalue Samples collected during work with at least The description of the covariates evaluated and the model in which they were used are detailed in Table 1 Detailed descriptions of the covariates related with oxfrdjunals one manipulation activity or at least one dry clean- the work tasks i.e. shown in Table ) can be found ing activity of the manipulation area were analyzed by in previous studies of our research group Garc^>a both PCM and TEM allowing for the estimation of the et al 2012 Garc^>aet al 2015 PCM asbestos concentration and this information All models have the following logistic regression org was used for the analysis Because a worker could have format for identifying variables significantly associ- by more than one h TWA PCM personal asbestos concentration observations within workers could not be considered independent Stata 2015 Thus regres- sion inference utilized robust estimates of standard ated with the odds of compliance In OR non compliance ) = ^ o+ ^ +X ^ .X.. ^ guest on X June 8 errors to accommodate the within person correlation Where OR denotes the odds ratio of compliance UCLA 2015c This was accomplished clustering by relative to compliance X ) are the covariates 2016 riveter in the longitudinal logistic regression models so available based on the model type Table ) and pos- that estimation of the regression coefficient standard sible interactions and " 'sare their associated effects errors are adjusted for the lack independence between All covariates were evaluated and the model build- the h TWA PCME personal asbestos concentrations ing process started with those predictors statistically collected on the same riveter on different days significantly associated with the odds of compli- Regression covariates evaluated included the type ance in univariate models 1.e. each variable separately of tasks performed by the workers the workload one at a time Model fit and multicollinearity were defined as the number of asbestos or asbestos con- evaluated and selection of models was performed taining products manipulated daily disregarding the type of manipulation performed the shop character- according with the Akaike's Information Criterion AIC the individual and general results obtained istics 1.e. ventilation presence of control systems and and personal judgment UCLA 2015b Interactions type of vehicle serviced and the number of years of were also analyzed in the models The statistical analy- experience working as riveters Based on observations sis was conducted using the software Stata 12.1 and the detailed activity diaries the type of activities StataCorp College Station TX USA Factors associated with compliance of asbestos Page 5 of 16 Table 1. Description of the covariates used in the three models evaluated to establish the odds of compliance of the h TWA of 0.1 cmasbestos occupational limit Covariate Type of covariate Description of the covariate Model that included the covariate Number asbestos products manipulated daily Integer variable Represents the number of asbestos containing brake products i.e. brake linings brake pads or brake blocks manipulated daily without considering the type of manipulation performed Model Workload Model Complete Number asbestos Integer variable Represents the number of Model Workload products manipulated daily asbestos containing brake products i.e. brake linings brake Model Complete pads or brake blocks manipulated daily without considering the type of Downlade manipulation performed from Manipulation w beveling _Indicator variable Represents at least one Model Work tasks http and grinding of asbestos manipulation that included Model Complete drilling countersinking riveting and bonding asbestos anhyg containing brake products one Manipulation w oxfrdjunals Indicator variable or more manipulations = 1 no manipulations = ) Represents at least one Model Work tasks o beveling and grinding of asbestos manipulation that included Model Complete obryg drilling countersinking riveting and bonding asbestos guest containing brake products one on or more manipulations = 1 no June manipulations = 0 8 Manipulation w beveling Indicator variable Represents at least one Model Work tasks 2016 and grinding of asbestos manipulation that included beveling grinding and cutting with Model CompleteComplete electric saw asbestos containing brake products one or more manipulations = 1 no manipulations = ) These tasks release important amounts of fibers and may also include drilling countersinking riveting and bonding asbestos containing brake products Page of compliance of asbestos standards brake mechanics Table 1. Continued Covariate Type of covariate Description of the covariate Model that included the covariate Manipulation w beveling and grinding of asbestos Indicator variable Represents at least one manipulation that included beveling grinding and cutting with electric saw asbestos Model Work tasks Model Complete containing brake products one or more manipulations = , no manipulations = 0 These tasks release important amounts of fibers and may also include drilling countersinking riveting and bonding asbestos Dry cleaning activities the Dofwnrloadem manipulation area _Indicator variable containing brake products Represents at least one cleaning activity of the manipulation area performed under dry conditions Model Work tasks Model 3 Complete one or more cleaning activity = , no http cleaning = 0 Location of worker during Indicator variable Incorporate the location of the Model Workload anhyg manipulation activities worker during manipulation Model Work tasks oxfrdjunals activities either inside ofthe manipulation area X = ) or outside of the manipulation area X = 0 Model Complete Ventilation conditions the Indicator variable Represents the ventilation condition Model Workload org shop of the shop good X = ) or bad Model Work tasks by guest X = 0 This classification was based on the near field concept Cherrie Model Complete on and Schneider 1999 Tielemans June et al 2008 Cherrie et al 2011 If 8 manipulation equipment was within 2016 1 m of a window or door ventilation was considered good Presence of control system Indicator variable Represents the presence of local control ventilation in the manipulation equipment such as fans or extraction systems disregarding their performance presence of control system = 1 no control system = 0 Model Workload Model Work tasks Model Complete Experience of worker Continuous variable Represents the number of years of experience working as riveter Model Workload Model Work tasks Model Complete Factors associated with compliance of asbestos Page 7 of 16 Table 1. Continued Covariate Type of vehicle serviced in the shop Type of covariate Description of the covariate Model that included the covariate Indicator variable Represents the type of shop sampled passenger vehicles BRS X = 0 or heavy duty vehicles BRS X = ) If shop serviced both types of vehicles it was classified based on the type of vehicle most commonly served during sampling campaigns Model Workload Model - Work tasks Model Complete hr TWA asbestos concentrations for riveters sampled 90 rm o cc 05 _ fe} 0% Downlade 04 from conetraisasbetos03 1 htp./anhyg.oxfrdjournals 02 01 o PCM00 % 0 & < ae wt tenn 20000 00000 # 20 40 60 af, ye igor 80 bs 0000 0000 0000 00000000 TT 100 hr TWA PCM personal samples passenger vehicles BRS = heavy duty vehicles BRS | htp./anygoxfrdjulsorg Figure 1 Eight TWA PCM asbestos personal concentrations for workers sampled by RESULTS guest measurements five riveters had three repeated measure- on A total of 28 riveters were sampled Thirteen out the ments and two riveters had one measurement June twenty riveters were from the nine passenger vehi- Twenty h TWA PCM personal samples 8 2016 cles BRS sampled and the remaining 15 were from the nine heavy vehicles BRS sampled Forty h 24 exceeded both the US OSHA PEL and the Colombian standard of 0.1 cmSome of these con- TWA PCM asbestos exposures were calculated for centrations were up to six times the standard as can be passenger vehicles riveters and 60 were calculated for observed in Figure 1. Seventeen out of the twenty heavy vehicles riveters for a total of one hundred h TWA PCM asbestos concentrations above the three h TWA PCM personal asbestos concentra- occupational standard were collected on riveters from tions The number of riveters based on both the type of passenger vehicles BRS and the remaining eight were vehicle serviced and the number of repeated h TWA collected on riveters from heavy vehicles BRS PCME personal asbestos concentrations each riveter The only type of asbestos identified in personal samples had was For passenger vehicles BRS i.e. 13 riveters was chrysotile Owners and brake mechanics involved one riveter had six repeated measurements three riveters in the study were informed regarding the results of the had five repeated measurements two riveters had four sampling campaigns and we also explained to them repeated measurements one riveter had three repeated some preventive strategies they could implement to measurements five riveters had two repeated measure- reduce or eliminate asbestos exposure ments and one riveter had one measurement For heavy- The bulk asbestos analysis of 18 duplicate samples duty vehicles BRS i.e. 15 riveters three riveters had six repeated measurements five riveters had five repeated oflinings pads or blocks from 12 brands ofbrake products distributed in Bogot^is presented in Table 2. The Page 8 of 16 compliance of asbestos standards brake mechanics Table 2. Bulk asbestos content obtained for two linings pads or blocks of 18 product's brands of brake products distributed in Bogot^ Manufacturer Brand Type of brake product Asbestos content % Sample 1 Sample 2 1 1 Brake lining 10 10 2 Brake block 15 15 2 Brake lining 15 15 Brake block 10 10 Brake block ND ND 3 Brake lining ND ND Downlade 4 Brake lining 15 15 3 5 Brake liningb 10 10 Brake blockb 15 15 from Sheet for brake pads ND ND http 4 67 Brake lining 10 10 anhyg 67 Brake liningab 10 10 oxfrdjunals S 8 Brake liningab 15 15 Brake lining 10 10 org 9 Brake lining 10 10 by 6 10 Sheet for brake pads 10 10 guest on June 7 11 Roll for brake lining S 5 8 12 Brake liningh 15 15 8 ND detected 2016 Product distributed in Bogot^more than 20 years ago Label ofthe product does not specify that it contain asbestos Label of the product does not report the asbestos content following standard recommendations EPA 2011 OIT 1984 bulk asbestos analysis showed that only three brake products one brake lining one sheet that is cut into pads and one brake block from three different brands did not contain asbestos fibers 1.e. reported as nondetected For the other 15 brake products analyzed the chrysotile asbestos content ranged between 5 and 15 Asbestos content did not differ between duplicate samples As shown in Table 2 four brake linings from brands 1 7 and 8 that were used in the past but are no longer distributed were also analyzed i.e. purchased from old stocks of one of the BRS sampled Daily workload can be expressed in terms of the number of asbestos containing products manipulated daily ranging from 0 to 50 and the number of asbestos containing products manipulated daily ranging from 0 to 52. Workers experience ranged from 1 to 34 years Table 3 presents the results of the models evalu- ated including the longitudinal based univariate logistic regression models for each covariate analyzed and the three longitudinal based multivariate logistic regression models evaluated according with the type of information available Factors associated with non-compliance of asbestos Page 9 of 16 univartetheincluded sampled duro logistc EvI-OS -(ENTR4A-d) Covarites y Longitudinal S]Pour pop; logistic YO s[apour UOIsaZI 1D multivarite PU (6F7'0) PU 98 89 7 Model INSIZO] uorsai productsasbestos asbestos products Sse} %S6 OR value daily daily 1.1 daily 1.12 0.082 0.277 daily 1.08 value OR Complet Downlade (100) PU 1.0 ( 0.92-1.09 0.92-1.09 +'0 @6' b2'I from asbestos suo] products asbetos http beveling 2.15 0.239 %S6 NA VN VN 2.94 ) 0.68-12.65 VN WN VN VN Factors Ref peoAT bevling asociated grinding grinding ( uo 0.81 0.736 VN VN VN VN VN YN VN VN VN WN asociated productspure asbetos 7mamanniippuullaatteedd manipulatedmanipulatedmanipulated bbeevvellinigng grinding Ref NA PU PY (Surpurs3 grinding 0.00 NA 18 0 (00'0) PU (Z0'0) PU ) 4.19-20.7 48.6 with ( <900) PU with on 6.26 37.5 June compliance 0.0 1 9y} 2016 /Zuyaq 4.19-20.7 UT Ref ou) asbetos Ref grinding ou) 6.71 0.02 payejnduew 0.96 0.78-19.7 soze1IA0d Ref syonpord cJYleaOOning activties 3.3 syonpoid 0.52 ) < 0.63 0.32-9 s1ay10M SOJELIBAOT) soyaqse soyaqse Ref ) [< 0 I< 0 [< 0 T< 0 [< 0 31qeL oN, on ony on on oN Ag Page 10 of 16 Non-compliance of asbestos standards brake mechanics 19 ajdui0y %S6 L9-760 <CTI-<60 -(@nye4-d) S}POU! [apo Lolnoggiitsutdiicnal univariate mu7lt0ivaria!te HO UOTsaIZr ID d1)s1Z0] SSE} %S6 value during manipulation manipulation maUnipuOlation ~0.0(O3#900) PY Downlade manipulation from http conditons ProplimT 19%S6 TOb-F60 Poy -(2D]R4-d) Ref PU UO E19 STI-b60 /anhyg oxfrdjumals UOTINGUOD (SPI) jepoui ayy ur pastobas pO'T JO o1g/ $0 0 xe] by 51 O} ajyelraTun (anjes-d) doneusn 31STB0] s[apour 0.854 PT (STF) PU (S80) PU (8S+'0) (ZOT 0) PU pour ANP guest on worker years 0.75 worker HO 0.458 1.428 STI 70'1 970 jo 0.775 pay jas June 8, 0.775 [apous 1.04 IystBo] Guin ayy 2016 WI 0.26 TOJEMUPAQUSuipoz2panpour paseg pajenas you yng 05 regresion model (savad) wi19}sA [EUIPNBUC] japolu aij payen sa UI information lack 13yx0M repsent contribuon Peoqas selcted [O.JUOD dy} sajero9y) papa10J [oad] Jo JO JO aporya Aynp [eatq jOU yuasrda UOTE, aoutredxg sayelsoy) Jasuseg UOTEIOT 3duasg QouesyIUKsaoupyue eprisng * apisuy po r peg aWOS aun Jo Aavazy BPEL sijaa afqey, addy, [> oy, YN yuryg Factors associated with compliance of asbestos Page 11 of 16 Based on the univariate regression analysis the odds of compliance are 45 times higher with at least one manipulation with beveling and grinding ofasbestos containing products than when this activity is not performed not shown the value of the interaction term was no longer significant and the AIC coefficient indicated that the model was more parsimonious without it Thus this interaction was excluded from Model 1 value = 0.000 are seven times higher with at least one manipulation with beveling and grinding ofasbestos Model 2 which evaluated the association between compliance of the asbestos occupational limit and containing products than when this activity is not performed value = 0.002 and are three times higher with at least one dry cleaning activity of the manipulation area than when this activity is not performed value = 0.063 The odds of compliance are the work tasks performed by the workers sampled showed that the odds of compliance are 30 times higher with at least one manipulation with beveling and grinding of asbestos containing products than when this activity is not performed value = 0.001 and are eight times higher when the worker remains inside of the manipulation area than when the worker remains outside nine times higher when the worker remains inside of the manipulation area during manipulation activi- during manipulation activities value = 0.039 are ties than when he remains outside value = 0.040 1.1 times higher with each additional asbestos containing and these two associations are statistically significant products manipulated daily value = 0.082 are 1.1 Downlade times higher with each additional asbestos contain- ing products manipulated daily value = 0.093 and from are 77 times lower in heavy duty vehicles BRS than in The odds of compliance are four times higher with at least one manipulation with beveling and grinding of asbestos containing products than when this activity is not performed value = 0.096 and this association passenger vehicles BRS value = 0.102 is marginally significant The odds of compliance http an hyg Model 1 which was focused on the association between compliance of the asbestos occupa- are three times higher with at least one manipulation without beveling and grinding of asbestos contain- tional limit and the workload of the workers sampled ing products than when this activity is not performed oxfrdjunals showed that the odds ofcompliance are 1.1 times higher with each additional asbestos containing product manipulated daily disregarding the type of value = 0.149 and are two times higher with at least one dry cleaning activity of the manipulation area than when this activity is not performed value = 0.522 manipulation performed and this association is signif- although these two associations are not significant org / icant value = 0.004 The odds of compliance Several interactions were tested and a potential by guest increases six times if the worker remains inside of the effect modifier was identified in Model 2. Similarly to manipulation area during manipulation activities Model 1 interactions were initially tested with mod- on June than when he remains outside and this association els that only included two covariates and their inter- is marginally significant value = 0.059 Although action term The presence of a control system and at 8 2016 significant the odds of compliance increases 1.1 times with each additional asbestos contain- least one manipulation without beveling and grinding of asbestos containing products was identified as ing product manipulated daily disregarding the type of manipulation performed value = 0.277 and a significant and meaningful interaction Although the inclusion of the interaction term in Model 2 favored the increases 1.04 times per each additional year of experi- AIC coefficient it increased the robust standard errors ence working as riveter value = 0.428 and the 95 confidence intervals were wider For this Several interactions were tested and a potential reason the interaction term was excluded from Model 2 effect modifier was identified in Model 1. Interactions Model 3 which evaluated the association between were initially tested with models that only included two compliance of the asbestos occupational limit and covariates and their interaction term With these mod- both the workload and work tasks of the workers sam- els the presence of a control system and the effect of the number of asbestos containing products manipu- lated daily disregarding the type of manipulation performed was identified as a significant and meaningful interaction However when the interaction term and the two covariates were included in Model 1 i.e. results pled show that the odds of compliance are 49 times higher with at least one manipulation with beveling and grinding of asbestos containing products than when this activity is not performed value = 0.000 and this association is statistically significant The odds of compliance are seven times higher when the Page of . compliance of asbestos standards brake mechanics worker remains inside of the manipulation area during Additional models for the odds of compliance manipulation activities than when he remains out- were constructed i.e. not shown with detail informa- side value = 0.058 and this association is marginally tion regarding the number of manipulations and the significant Although significant the odds of non- tasks involved in each manipulation No significant compliance are three times higher with at least one dry associations were found Considering that this level cleaning activity ofthe manipulation area than when this of detail information is difficult and in retrospective activity is not performed value = 0.249 and increase studies almost impossible to obtain the three models with each additional asbestos containing product presented in the current study require information manipulated daily disregarding the type of manipulation that is reasonable to obtain for both current and retro- performed value = 0.994 and with each additional spective studies which favors these models year of experience working as riveter value = 0.775 Three different models were constructed offer- Both interaction terms previously identified in ing different options depending on the information Model 1 and 2 were evaluated by separate in Model 3 required to run the models The results when analyz- i.e. results not shown In both cases the AIC coeffi- ing the workload Model ) show that both the num- cient favored the model with the interactions but nei- ber of asbestos and asbestos containing products ther was included because they increased the robust standard errors and the 95 confidence intervals were manipulated daily increase the odds of compliance The results when analyzing the work tasks per- Downlade from wider with the terms Additionally when the interac- formed by workers Model 2 suggest that tasks that http tions were included the odds of compliance for the covariate number asbestos products decreased i.e. OR can potentially release important amounts of fibers are highly associated with the odds of compliance 1.e. an hyg below ) per each additional asbestos containing prod- uct manipulated daily disregarding the type of manipu- manipulation ofbrake products that include beveling grinding and cutting Furthermore this associa- lation performed which is a counter intuitive behavior tion was observed not only with asbestos containing oxfrdjunals DISCUSSION This study analyzed factors that could be associated with products but also when asbestos containing products were manipulated This is something also found in Model 1 with the number of asbestos containing the odds of compliance of the h TWA 0.1 cm products and is something we have reported in previ- org asbestos occupational limit in riveters exposed to chry- ous studies Garc^>aet al 2012 Garc^>aet al by guest sotile asbestos that work in BRS in Bogot^Among the 2015 Salazar et al 2014 Two probable reasons why variables tested the most unportant ones that were asso- the odds of compliance of the asbestos occupa- on June ciated with the odds of compliance were the types tional limit could be associated with the manipulation of tasks performed during the work and the loca- of brake products that do not contain asbestos include 8 2016 tion of the worker during manipulation activities The number of products manipulated daily 1.e. work- poor hygiene of the manipulation equipment and lack of maintenance of the shops Poor hygiene and lack load showed significant associations in the longitudinal based multivariate logistic regression models of maintenance could result in the accumulation of asbestos fibers on different surfaces and the subse- One of the potential uses of the models constructed in the current study is to determine the risk of com- quent resuspension of the fibers during other activities such as the manipulation of brake products that do not pliance experienced by workers in countries or situations in which conducting sampling campaigns is not possible Using the models presented in this study compliance of the h TWA 0.1 cmasbestos occupational limit could contain asbestos Another explanation could be that asbestos containing products were not labeled correctly which may create a false sense of safety in the workers that manipulate these products be estimated by determining the type of manipulations performed on asbestos and asbestos containing brake products and the number of asbestos and nonasbestos containing brake products manipulated daily This information could be obtained using occupational The analysis of asbestos content of brake products confirmed the presence of chrysotile asbestos in a range between 5-15 in most of the brake products used in the shops sampled and no other asbestos type was identified As can be observed in Table 2 the asbes- questionnaires or observations during work tos content of brands distributed in the past does not Factors associated with compliance of asbestos Page 13 of 16 differ from the asbestos content of some of the brands compliance increased i.e. significantly currently distributed Most of the brands analyzed are manufactured in Colombia except for Brand 11 which Since the models were adjusted for the workload higher efficiency of experienced mechanics in per- has the lowest asbestos content % Furthermore the forming the manipulations which may result in asbestos content does not seem related with the type higher exposures is already considered in the analysis of brake product lining block or sheet for pads It is important to highlight that brands 2 and 5 i.e. manufacturer 2 and 3 respectively have both asbestos and asbestos containing products This seems to contradict one of the arguments of the asbestos industry against an asbestos ban and is the lack of safer substi- Experienced mechanics having higher odds of noncompliance contradicts the findings of other occupational groups where it has been observed that less years of experience can result in poor skills to perform the work tasks which may result in higher occupational exposures Andjelkovich et al 1977 Delzell tutes Furthermore since some of the brake industry is et al 1982 already manufacturing asbestos containing brake Several authors have recognized the importance ofthe products a transition to asbestos substitutes in a future workplace characteristics in the estimation of personal asbestos ban seems to be feasible in the short term exposures Cherrie and Schneider 1999 Tielemans Downlade The results suggest that the manipulation equip- ment used and the type of manipulation performed et al 2008 Cherrie et al 2011 Ventilation conditions of the shops are one of these characteristics The from i.e. with or without grinding beveling and cutting longitudinal based univariate logistic regression model http tasks have a stronger association with the risk of non- compliance than the number of asbestos containing performed with the ventilation conditions of the shops showed a significant decrease of the odds of non- products manipulated daily Furthermore the odds of compliance in shops with good ventilation compared to compliance are higher when asbestos containing poorly ventilated shops OR 0.4 P = 0.425 However /anhyg.oxfrdjuls products are manipulated compared to the manipula- tion of asbestos containing products as expected For all three models Model 1 2 3 the location of the worker during manipulation activities was the classification of the ventilation conditions was based on qualitative assessment of specific characteristics of the shops such as the distance of the manipulation equip- ment to doors or windows and not quantitative assess- org marginally or significantly associated with the odds of ment of the air exchange rates The AIC indicated that compliance This covariate was included because including this covariate did not contribute to the models by some personal samples collected during manipu- Regarding other workplace characteristic the results guest lation activities contained asbestos fibers which could of the longitudinal based univariate logistic regression on June indicate that asbestos fibers could remain suspended model suggest that the presence of control systems 1.e. in the air for long periods of time and migrate to other that in most cases included a general extraction system 8 2016 areas of the shop Thus workers that stay or work at the proximity of the manipulation equipment during non- does not guarantee the protection ofworkers Although the objectives of this study did not include an evalua- manipulation activities could be also at risk of asbes- tion of the performance of the existing control systems tos exposure This is supported by inhalation exposure it was observed that in some occasions workers did not studies that evaluate both near- and field compart- use them and it seemed i.e. just by observation that ments Cherrie and Schneider 1999 Tielemans et al in some instances the control system did not redirect 2008 Cherrie et al 2011 De Vocht et al 2013 Dry cleaning activities of the manipulation area which can suspend fibers had a significant associa- tion with the odds of compliance using a longitudi- and extract the fibers released during the manipulation processes This was corroborated by the results of the longitudinal based univariate logistic regression model which showed OR values above 1 although in this case nal based univariate logistic regression model OR 3.33 P = 0.063 The importance of dry cleaning activities in predicting the odds of compliance of the asbestos occupational limit was also observed in models 2 and 3 One unexpected result was that as the years of experience of the mechanic increased the odds of the association was not statistically significant Additionally the association between the odds of compliance and the type of vehicle serviced i.e. passenger versus heavy duty showed an OR of 0.235 in the longitudinal based univariate logistic regression model value = 0.102 Thus the type Page 16 compliance of asbestos standards brake mechanics of vehicle serviced does not seem to have an influ- Ciencia Tecnolog^>a e Innovaci^n Colciencias ence on the odds of compliance Moreover the Contract 760-2013 project code 120456933509 convergence of the models evaluated was bad when and grant 567-2012 for doctoral students from this covariate was included because it increased the the same institution Financial support was also robust standard errors and made the confidence obtained from the Department of Civil and intervals very wide Environmental Engineering and the School of To the best of our knowledge a sampling cam- Engineering from Universidad de Los Andes and paign of this size to understand the risk of elevated Fundaci^nMAPFRE asbestos exposures of brake mechanics has never been done in low- and income countries However ACKNOWLEDGEMENTS as authors we recognize that a larger sample could improve the results in the longitudinal based logistic regression models with the cluster analysis applied in P.N.B. is serving in his personal capacity The opinions expressed in this article are the author's own and do not reflect the views of the Centers for Disease this study UCLA 2015a Control the Department of Health and Human Downlade As we have explained in previous studies Cely- Garc^>aet al 2012 Garc^>aet al 2015 Salazar al 2014 the results ofthe models evaluated confirm Services and the US government We thank the owners administrative staff brake mechanics and rivet- ers from the BRS sampled We also acknowledge the that the physical form some of these brake products are from distributed i.e. detached from the support results in http an unavoidable manipulation of asbestos containing undergraduate students who collected information in the zones where BRS are clustered and to Juliana Uribe and Natalia Salazar for their assistance during an hyg brake products at the shops making the controlled or safe use difficult if not impossible to achieve the sampling campaigns Additionally we acknowledge Mar^>aGrau P^'rezfor her advice in the statistical Finally it is also important to highlight once analysis Finally we want to thank Alonso Rojas for his oxfrdjunals again the unacceptable working conditions of these workers Out of the one hundred three h TWA PCM personal asbestos concentrations esti- help during the sampling campaigns The authors have no conflict of interest to declare org mated 24 exceeded the Colombian standard and REFERENCES guest the US OSHA PEL of 0.1 cmasbestos occupa- by ACCP 2012 Chapter 18. Pneumoconiosis In Kamp DW tional limit Furthermore 15 of the samples were editor American College of Chest Physicians Pulmonary in compliance but above the USA OSHA PEL action Medicine Board Review 26th edn Glenview IL American on June level of 0.050 cmThe highest exposures were College of Chest Physicians pp 259-83 observed in three riveters that had high workloads ACSH 2007 Asbestos exposure How risky is it Available 8 2016 and that worked in two poorly ventilated passenger vehicles BRS who also bond and grind the brake products to remove excess glue and in two riveters from another poorly ventilated passenger vehicle at http://www.acsh.org/docLib/20071015_Asbestos.pdf Accessed 5 April 2011 Andjelkovich D TaulbeeJSymons M et al 1977 Mortality of rubber workers with reference to work experience J Occup Med 19 397 405 BRS with a medium workload In this last shop ATSDR 2001 Toxicology profile for asbestos Atlanta GA asbestos fibers accumulated over the floor and the Agency for Toxic Substances and Disease Registry U.S manipulation equipment Similarly to the conclu- Department of Health and Human Services Available at sions of previous studies conducted by our research http://www.atsdr.cdc.gov/toxprofiles/tp61http://www.atsdr.cdc.gov/toxprofiles/tp61 pdf Accessed group with auto mechanics these workers experi- ence high asbestos occupational personal exposures and because of this they are at excessive risk of developing asbestos related diseases 5 April 2011 Blake CL Dotson GS Harbison RD 2006 Assessment of an- borne asbestos exposure during the servicing and handling of automobile containing gaskets Regul Toxicol Pharmacol 45 214 22 Blake CL Van Orden DR Banas~-kM et al 2003 Airborne FUNDING asbestos concentration from brake changing does not This study was funded with a grant from the Colombian Departamento Administrativo de exceed permissible exposure limit Regul Toxicol Pharmacol 38 58-70 Factors associated with compliance of asbestos Page 15 of 16 Cely Garc^>a MF S^nchez M Breysse PN et al 2012 Personal exposures to asbestos fibers during brake maintenance of passenger vehicles Ann Occup Hyg 56 NIOSH 1994a 7400 Asbestos and other fibers by PCM Available at http://www.cdc.gov/niosh/docs/2003-154/ 7400.pdf Accessed February 2010 985-99 NIOSH 1994b 7402 Asbestos by TEM Available at http Garc^>aMF Duque CA M Dur^n et al 2015 Personal exposure to asbestos and respiratory health of www.cdc.gov/niosh/docs/2003-154/pdfs/7402.pdf Accessed February 2010 heavy vehicle brake mechani] Ecxpso Sci Environ Epidemiol 25 26-36 OIT 1984 Seguridad en la Utilizaci^ndel Anuanto Ginebra Suiza Organizaci^nInternacional del Trabajo Available at CherriJeW Maccalman L W Fransman et al 2011 Revisiting the effect of room size and general ventilation on the rela- http://aragon.ugt.org/slaboral/normativa/europea/am~-anto.pdf Accessed 5 April 2011 tionship between near- and field air concentrations Ann OIT 1986 Convenio 162 de la OIT sobre la utilizaci^n del Occup Hyg 55 1006 15 asbesto en condiciones de seguridad Ginebra Suiza Instituto Cherrie JW Schneider T. 1999 Validation of a new method Nacional de Seguridad e Higiene en el Trabajo Organizaci^n for structured subjective assessment of past concentrations Internacional del Trabajo Available at //www.insht Ann Occup Hyg 43 235-45 Contenidos De Vocht F Northage C Money C et al 2013 The future 162delaoitsobreutiliza- of exposure assessment perspectives from the X2012 ciondelasbestoencondiciones.pdf Accessed 5 April 2011 Conference Ann Occup Hyg 57 280 5 Delzell E Andjelkovich D Tyroler HA 1982 A control Downlade study of employment experience and lung cancer among from rubber workers Am J Ind Med 3 393 404 OMS 2006 Eliminaci^nde las enfermedades relacionadas con asbestos Available at http://apps.who.int/iris/bit- WHO_SDE_OEH_06.WHO_S0 DE_OEH_06.3 03_spa W_ HO_SDE_OEs H_06.03_sp pa WHO_SDEa _OEH_06.03_spa pdf Accessed 2 May 2011 EPA 1993 EPA Method for the OSHA 1994a Occupational Safety and Health Standards Toxic http Determination of Asbestos in Bulk Building Materials and Hazardous Substances Standards 29CFR Available at http://www.epa.gov/region1/info/test- 1910-1001 Available at htps:/ww.osha.gov/pls/oshaweb/ https://www.osha.gov/pls/oshaweb/ EPA_600R93116_bulk_asbestos owadisp.show_document part1.pdf 9995 Accessed November 2013 /anhyg.oxfrdjuls EPA 2011 40 CFRPart763 Subpart I -- Prohibitionofthe Manufacture Importation Processing and Distribution in Commerce of Certain Contianing products Labeling Requirements Available at www gpo.gov/fdsys/pkg/CFR-2011-title gpo.gov/fdsys/pkg/CFR-2011-title CFR- OSHA 1994b Regulations Standards CFR Medical questionnaires mandatory Available at www osha.gov/pls/oshaweb/owadisp.show_document?p_ STANDARDS Accessed September 2011 OSHA 2008 Personal sampling for air contaminants OSHA title40 part763 I.pdf Accessed 27 Technical Manual OTM Section II Chapter 1. Available by guest October 2015 at https://www.osha.gov/dts/osta/otm/otm_u/otm_i_1 Hickish DE Knight KL 1970 Exposure to asbestos during order_sampling_media Accessed July 2013 on brake maintenance Ann Occup Hyg 13 17 21 OSHA 2013 TODO SOBRE LA OSHA 3171-01R IARC 1998 Asbestos Actinolite Amosite Anthophyllite Available at http://www.osha.gov/Publications/osha3173 June Chrysotile Crocidolite Tremolite Supplement 7 1987 pdf Accessed June 2013 8 Available at h_ttp://www.inchem.org/documents/iarc/ Paustenbach DJ Finley BL Lu ET et al 2004 Environmental 2016 asbestos.html Accessed 24 September 2013 and occupational health hazards associated with the pres- IARC 2012 IARC Monographs on the Evaluation of ence of asbestos in brake linings and pads 1900 to pre- Carcinogenic Risks to Humans Asbestos chrysotile sent a state review J Toxicol Environ Health amosite crocidolite tremolite actinolite and anthophyl- 7 33 110 lite Volume 100C Available at //monographs.iarc Paustenbach DJ Richter RO Finley BL et al 2003 An evalu- Monographs Accessed 31 ation of the historical exposures of mechanics to asbestos in October 2015 brake dust Appl Occup Environ Hyg 18 786 804 Madl AK Gaffney SH BalzeJrL et al 2009 Airborne asbes- tos concentrations associated with heavy equipment brake Salazar N Garc^>aMF Breysse PN et al 2014 Asbestos exposure among transmission mechanics in automotive removal Ann Occup Hyg 53 839 57 repair shops Ann Occup Hyg 59 292-306 MSPS 2011 Ministerio de Salud y Protecci^nSocial 2011 Stata 2015 vce_options estimators Available Resoluci^n007 de 2011 polra cual se adopta el Reglamento de Higiene y Seguridad del Crisotilo y otras fibras de uso at http://www.stata.com/manuals Accessed 6 October 2014 similar Available at htp:/fondoriesgoslaborales.gov.co/ http://fondoriesgoslaborales.gov.co/ Normatividad Tielemans E Schneider T Goede H et al 2008 Conceptual model for assessment of inhalation exposure defining mod- pdf Accessed November 2011 ifying factors Ann Occup Hyg 52 577-86 Page 16 compliance of asbestos standards brake mechanics UCLA 2015a UCLA Statistical Consulting Group Analyzing Correlated Clustered Data Available at //www.ats.ucla cpsu.htm Accessed 10 November 2014 UCLA 2015b UCLA Statistical Consulting Group Lesson 3 Logistic Regression Diagnostics Available at http www.ats.ucla.edu/stat/stata/webbooks/logistic/chapter3/ www.ats.ucla.edu/stat/stata/webbooks/logistic/chapter3/ statalog3.htm Accessed 10 November 2014 UCLA 2015c UCLA Statistical Consulting Group Regression with Stata Chapter Beyond OLS Available at http://www.ats.ucla.edu/stat/stata/webbooks/reg/chapstatareg4.htm Accessed 10 November 2014 Wagner GR Lemen R. 2008 Asbestos In Heggenhougen K editor International Encyclopedia of Public Health San Diego CA Academic Press pp 238 45 Weir FW Tolar G Meraz LB. 2001 Characterization of vehicular brake service personnel exposure to airborne asbestos and particulate Appl Occup Environ Hyg 16 1139-46 Yeung P Patience K Apthorpe L et al 1999 An Australian study to evaluate worker exposure to chrysotile in the automotive service industry Appl Occup Environ Hyg 7 449-58 Downlade from http / an hyg.oxfordjournals /anhyg.oxfrdjuls org by guest on June 8 2016