Occupational disease predictors in the nickel pyrometallurgical production: a prospective cohort observation

Study design, cohort construction and its occupational profile

This was a prospective observation of all enlisted employees of a nickel pyrometallurgical production site of one of the leading Russian nickel producers, situated in the Russian High North. The company reported production of 166,265 tons of nickel (approximately 10% of the total world production) in 2019. The company is one of the largest nickel producers in the world and sustains full nickel production cycle at various sites, including ore mining, ore dressing, pyrometallurgical processing and nickel refining by electrolysis and carbonyl methods. We earlier analyzed and described occupational claims of workers at another nickel production site of the same company, employed in the nickel electrolytic production [10]. Unlike electrolysis, usually considered the final stage of ready product production, pyrometallurgical ore processing stands midway between mining and final product. Because occupational exposures, physical site location and technologies were basically different throughout the production cycle, cohorts of electrolysis workers presented earlier, and the current pyrometallurgical processing cohort were analyzed separately.

This study received approval from the Committee of Bioethics of the Northwest Public Health Research Center and was conducted in accordance with the relevant guidelines and regulations. At the annual screening described below (mandatory for employees under Russian law), all employees signed a written informed consent to participate and to have the obtained data used for research purposes. There were no workers who refused to participate.

Once constructed by the end of annual screening in 2007, the study cohort was fixed and followed for subsequent years till 2021. All enlisted at the pyrometallurgical site employees were referred to the annual screening by the official Order of the human resources department. Therefore, annual screening of 2007 was considered as time 0 for this observational study. We tracked all new cases of occupational disease (compensation) claims leading to contract discontinuation, as well as all other reasons of drop-out annually using workers’ personal IDs. All cases of occupational disease (compensation) claims were obtained from the central database of the Kola branch of the Northwest Public Health Research Center, which also acted as the Provincial Center of Occupational Disease, legally in charge of occupational claims verification.

In total, workers from 54 workplaces of pyrometallurgical processing site were included and grouped into 16 occupational groups. As of the legislation in occupational health in 2007, work conditions assessment in the industry was part of workplace attestation performed by the governmentally certified organizations, such as Kola branch of the Northwest Public Health Research Center once every five years. In addition, governmentally certified in-house analytical laboratory was in charge of routine work conditions assessment four times a year. Comprehensive exposure assessment as part of such work conditions assessment implied classification of hazard grade in each workplace depending on the fold-range of OEL exceedance [11]. Of note, out of five occupational factors (Table 1), fixed OEL existed only for four occupational factors, whereas OELs for chemicals varied depending on a specific substance. The current classification assumed four hazard grades: (1) optimal (below OEL); (2) acceptable (below or equal to OEL); (3) hazardous, with four subgrades (exceeding OEL in all sub-grades with the corresponding fold-range); (4) dangerous (above OEL and with the fold-range higher than in 3.4) (Table 1).

Table 1 Hazard grading classification with regard to OEL exceedance

Compared to hazard grading classification with regard to OEL exceedance, the methodology to define hazard grades with regard to labor intensity and workplace microclimate was more complex and included dynamic and static workload, motions stereotype, lifting and moving weight and work posture. Hazard grading as of the labor intensity was defined as a function of intellectual, sensor and emotional load, as well as their monotony. Manual exertion of nickel production workers, as reported by the company, was usually within the existing recommendations, differential for males and females, which considered 14 measured indicators. These indicators included single weight to lift (below 30 kg); repeated weight to lift (below 15 kg); the overall hourly weight moved from the working surface (below 870 kg); the overall hourly weight moved from the floor (below 435 kg); the overall count of repeated local movements during the shift for palm fingers (below 40,000); the overall count of repeated regional movements during the shift for arm muscles (below 20,000).

In addition, workplace microclimate assessment was based on the temperature, humidity, air flow and work intensity with four corresponding hazard grades [11]. Such approach of workplace assessment allowed to mold fifteen occupational groups, including smelters (group 1), metalworkers (group 2), electricians (group 3), crane operators (group 4), burners (group 5), welders (group 6), metal cleaners (group 7), foremen (group 8), riggers (group 9), crusher operators (group 10), batchers (group 11), gas purification operators (group 12), converter operators (group 13), flotation operators (group 14), mil operators (group 15), and in addition a versatile group of “others” (group 16). The greatest OEL exceedance was noted for chemical substances and low fibrogenic aerosols, which made the greatest contribution to the overall grade along with labor severity. In general, smelters, burners, crusher and converter operators were graded as most hazardous workplaces with 3.4-4 hazard grades, as shown in Table 2.

Table 2 Hazard grades of included occupations for each hazard and overall Annual screening

Annual screening was conducted in accordance with the Russian Federation Order of the Ministry of Health, which dictated panel composition, timing, legal assumptions, procedure to inform employer, and listed mandatory medical examinations and tests along with contraindications to work. Workers were referred to the annual screening to the Kola branch of the Northwest Public Health Research Center. Eight specialists of the panel, who screened health status, identified medical conditions and diseases and verified medical contraindications to work were reinforced by a pulmonologist, endocrinologist, gastroenterologist and urologist. Upon screening completion, an employee was expected to have a fitness certificate, whereas the medical profile was stored, and depersonalized data were transferred and analyzed afterwards. The panel should identify medical conditions and confirm diagnosis with routine (electrocardiography, spirometry, X-ray, audiometry) and specific ancillary examinations, including cold and vibration tests. Should an employee exhibit signs suspicious of a work-related disease, he or she is then referred to an advanced medical examination in the specialized facility. For workers employed in the nickel production industry, the Order mandates screening once a year; the screening protocol from 2007 did not change until the study completion. Diagnoses from the annual screening in 2007, at that time treated as general chronic conditions and not work-related yet, were analyzed as predictors for occupational disease claims in the current presentation along with other occupational and behavioral (smoking) variables.

Air nickel concentrations

Nickel chemical compounds in the breathing zone may act as a major occupational chemical hazard in those employed for nickel pyrometallurgic processing and may include metallic nickel, nickel oxides and sulfides, nickel compounds mixtures called stein and fine stein, nickel concentrate and agglomerate, cleaning device dust, nickel compounds aerosol and nickel with chromium. Air sampling was routinely performed by a certified in-house chemical laboratory in selected workplaces at least once every three months. As of the current legislation, sampling can only be performed by the certified laboratories and only in compliance with the nationwide protocol. This protocol implies area sampling in locations with close proximity to emission sources and locations with expected high exposure, such as electrolysis baths, blast furnaces, etc. This sampling yields records of the mean, minimal and maximal concentrations, as well the percent of samples exceeding OEL. In addition to routine sampling of cobalt, copper, lead, arsenic anhydrides, formaldehyde, sulfur dioxide, carbon monoxide, ammonia, sulfuric acid and overall dust, nickel was quantified in samples using polarographic method and ПУ-1 device. The air in selected locations was aspirated using А-01 or АМ-5 pumps. As of the current protocol, nickel, nickel oxides and sulfates were quantified in the dust samples precipitated on АФА-ВП20 filters from 120 l (pump speed 10 l/min), or from 1000 l of air (pump speed 35 l/min) for water-soluble compounds. We extracted data on routine nickel sampling data from the company records and averaged four quarterly concentrations in a given year for each specific workplace analyzed in this study. As of the current legislation, OEL for water-insoluble nickel compounds was 0.05 mg/m3 and 0.005 mg/m3 for nickel hydroaerosol.

Statistical analysis

The primary outcomes of interest were absolute number of cases of occupational disease claims overall and each year during the follow-up. These cases are also presented as relative measures of effect when divided by the total numbed of occupational disease claims and, alternatively, to the overall number of workers initially included in the cohort. All baseline demographic data as continuous variables, including age, years in service and air nickel concentrations were tested for normality using Shapiro-Wilk test. Binary variables, such as sex distribution, are expressed as N with the corresponding percent from the overall count. Because most continuous variables were non-normally distributed, we used nonparametric Mann-Whitney U-test to compare two groups and Kruskall-Wallis test for three or more groups. Binary variables between groups were tested using contingency tables and the corresponding χ2 test. If not shown otherwise, we reported medians with the associated interquartile range (IQR). Whenever data were normally distributed in the group, we reported means with standard deviations.

We first tested the difference in the major continuous and binary variables between sixteen occupational groups, as specified above, as of the annual screening in 2007, representing cross-sectional analysis. We then documented the new cases of occupational disease claims during the follow-up and reported the incidence rates as number of incident cases per 1000 workers per year for each specific occupational diagnosis. Air nickel concentrations, first available as continuous variables, were tested as a predictor for occupational disease claims (described below), but were recoded to a binary variable of high vs. low nickel exposure. To determine the cut-off level, we used receiver operating curves (ROC)-analysis, in which an air nickel cut-off level with the greatest (sensitivity + specificity) was obtained in addition to reported area under the curve (AUC) with its 95% confidence interval (CI) and the associated p-value. In the subsequent analyses, exposure to nickel was treated as a binary variable of high vs. low concentrations using the obtained cut-off value.

The secondary outcome of interest in the current analysis was the chance (probability) to obtain a confirmed occupational disease (compensation) claim, first overall, and then for a specific diagnosis, in crude, and then adjusted Cox regression models. Selected predictors were chosen from the annual screening in 2007, including occupational groups (one of fifteen groups, because group 16 was excluded from the later analyses), whereas the “time” variable in the regression model was the elapsed time since the start of employment (the overall work duration) to either fail, such as in case of an occupational disease claim, or censor, should this case not happen. These models reported hazard ratios (HR) with the corresponding 95% CI in the adjusted models as specified in each specific case or for a specific occupational diagnosis. Predictors for adjusted Cox regression models were chosen depending on the crude comparisons. Smoking in all presented models was included as a binary variable (yes/no); and the alternative analysis with pack-years did not alter the effect (data not presented). Among other predictors, nickel exposure and occupational groups were tested in the adjusted model to see whether their effects were independent of each other, despite some nickel exposure present in most groups. All tests were accomplished in NCSS 2021 (Utah, USA), and p-values below 0.05 were considered significant.

留言 (0)

沒有登入
gif