In 2012 the classification of FA as a risk factor for cancer in humans [9
] was a game changer in many anatomy and pathology departments, regarding the use of FA for preservation purposes. Recent evaluations of published literature have confirmed a causal link between FA exposure and nasopharyngeal tumors and leukaemia [11
]. Human biomonitoring studies have shown that anatomy and pathology workers with an average full-shift exposure of several hundred µg/m3
have a statistical significant enhanced level of chromosomal aberrations compared to hospital controls [15
]. This biomarker has been linked to an increased cancer risk [23
]. In a recent study among students in Brazil, the baseline frequency of micronucleated cells of buccal epithelial tissue was significantly increased (twofold after one month and threefold after 3.5 months) following 30–90 h of exposure to FA during human anatomy classes [24
In the Radboudumc the anatomy teaching and research facilities had been renewed in 2009 and the exposures assessment in 2012 was the first occasion of a systematic assessment of exposure. As the outcome indicated that the conditions were not sufficiently in compliance with current OELs, the workers and the department management decided to adopt some technical improvements and changed work practices. This study describes how these changes translated in FA exposures of employees and also of instructors and students over a period of 6 years.
4.1. Air Sampling at Fixed Locations
FA emissions from building materials [21
] offer a challenge when trying to reduce the background because of the narrow exposure margin of little more than one order of magnitude between normal indoor background levels and the OEL for a full shift. This is in particular a challenge if spills over a long time of use have caused deposits of solid polymerized formaldehyde (polyFA) that represent a persistent source of recurrent emissions because every time the floor is wetted e.g., during cleaning, some of the polymer deposit will dissolve and may contribute to FA vapors becoming airborne. In the present study we observed a background in a non-occupied teaching room which was fairly low (1.5 and 4.1 µg/m3
) and within a range that may also be attributed to FA emissions from building materials.
In 2014, on average, a higher concentration was observed in TR-2 (69.8 µg/m3
) compared to TR-3 (10.9 µg/m3
). The room dimensions are different but when calculating the installed and effective ventilation per section table the infrastructure is comparable (see Table 1
). A more obvious explanation for a higher background of FA in TR-2 is offered by the type of specimens used at the time of taking the air measurements. In TR-2 torsos were used that have many internal cavities that may cause draining of formalin, which resulted in a concentration of 124.6 µg/m3
in far field, while on the other three days concentrations of 50.3, 52.8 and 71.8 µg/m3
were observed at the same location [26
]. In 2017, for both rooms the concentrations were low. On those days both small and large specimens were used, including complete cadavers and torsos.
In the embalming room in 2012 the average background of FA was almost 50% of the 8 h OEL with 2 out 10 measurements indicating exceedance of this workplace standard. In 2014 this situation was improved by 63% (Table 3
). This improvement is attributed to introduction of an LEV system at the work bench were anatomical specimens were flushed to remove formalin (T-3).
For the storage room the observed average FA concentrations indicated exceedance of the OEL by twofold both in 2012 and 2014. At that time for all tasks performed in this room respiratory protection was required. This was feasible because only workers had access to this room and they were spending limited time for tasks specified in this report as ‘refilling’ storage tanks with formalin, ‘take out’ and ‘place back’ anatomical specimens in storage tanks after use. The high background is explained by the availability of 75 storage tanks that all represent potential sources of leakage, the availability of the 37% FA concentrate storage, the lift to take out and place back specimens from tanks and the storage cabinet for hoses that are used to fill tanks with formalin solutions. Most interventions were implemented in this room: leak prevention of storage tank (T-1), improvement of the down flow ventilation (DFV) system (T-4) and optimizing the storage system (O-1).
Within anatomy laboratories storage facilities have been identified as a location with high exposures. Higashikubo and co-workers [27
] reported average pre-shift FA concentrations of 450 µg/m3
in Japanese anatomy facilities and reported an increase of FA exposure with installed storage capacity. Leakage of loosely sealed containers was identified as the primary source of this background. For a reduction of the background concentrations in the storage facility to a level compliant with the OEL, the required air exchange rate would have to be increased to at least 10 per hour. As long as this cannot be achieved the workers will have to wear personal protective equipment. A reduction of the 8 h TWA exposure was achieved by a restriction of the exposure duration as a result of the optimization of storage of the complete cadavers and anatomical specimens. A DFV system was installed at the location where the storage tanks are opened and the specimens are heaved from the tanks by use of the lift. An intrinsic limitation of this system is related to the principle of DFV on a location were workers and some equipment (such as the electric engine of the lift) produce heat which results in air flow in the opposite direction than the airflow of the DFV system due to convection powered by temperature differences. This effect is aggravated by the relative low room air temperature of 19 °C. Smoke testing in 2012 and 2014 showed that a downward air flow was not achieved, even in an unoccupied setting. In addition, when workers are using the facility, movements of the workers cause a turbulent airflow which further decreases the efficiency of the ventilation system.
4.2. Personal Air Sampling
Inhalation exposure was assessed in the breathing zone of the exposed subjects. Over the period from 2012 to 2017 this gave guidance to efforts made to reduce the number and magnitude of emission sources as determinants of inhalation exposure. Two strategies were used: The first being an overall reduction of contamination of the section tables, floors and other surfaces to try and reduce general background exposure. The 8 h TWA OEL of 150 µg/m3
offers good guidance to achieve an overall lower exposure level but this OEL is only 1.5 times the WHO guidance for indoor concentrations for FA [28
]. The second strategy was targeting specific peak exposures arising from tasks identified by the anatomy workers as critical to FA emissions. For short-term exposures resulting from such tasks, guidance is provided by a 15 min OEL of 500 µg/m3
. This guidance is defined as time-weighted average, which means that the value may exceed 500 µg/m3
for some time, as long as (within the same time interval) this exceedance is compensated with periods of low(er) exposure. Therefore, it is possible that shorter exposures (‘peaks’) of several 1000 µg/m3
or even higher may still occur.
Direct reading equipment with a short response time is sometimes used to analyze the tasks real-time. However, for FA these instruments have many limitations for the use in anatomy and pathology setting due the cross sensitivity with other organic compounds. False positive response may occur due to methanol which is a constituent of formalin or due to ethanol and isopropyl alcohol used in healthcare facilities for skin and surface disinfection [20
]. In our study we used the DNPH method for air measurements with both 8 h and 15 min timeframes. This method also has its limitations but in active sampling it is possible to reach a sufficiently low sensitivity and good precision. Problems with high air humidity can occur [20
] but is not a problem in indoor settings.
Overall, the full shift personal air sampling results indicate a gradual improvement for both workers and students from 2012 to 2014 and from 2014 to 2017 (Table 4
and Figure 9
). Despite some changes of work practices, no changes were observed over the first three years (not in GM and also no in the number of measurements exceeding the OEL). This is most probably due to the situation in the storage room that was not showing a reduction in FA air concentrations (see Section 4.2
). The only change that was observed is a reduction of the P95
, bringing the concentration down from 407.9 µg/m3
to 252.8 µg/m3
but the frequency of non-compliance only decreased from 42.8% to 38.5%. For the students/instructors some improvement was observed, also in the P95
and in the number of exposure measurements exceeding the OEL that decreased from 60% to 40%. In the last measurements performed in 2017, all descriptors of the exposure in Table 4
indicated a substantial reduction, leading to a situation that is close to a well-controlled exposure situation. There is still some room for improvement as the GM could be further reduced from the observed overall GM of 28.8 µg/m3
which is close to 20% of the OEL to less than 10% of the OEL.
For the task-based measurements, the exposure of the most critical FA related work practices were (much) improved over the 6-year interval (Table 5
). The highest exposures related to ‘take out’ and ‘place back’ did not show an improvement (Figure 10
). For the task ‘refill’ there was no improvement either. On the contrary, the average exposure appears to be worse in 2014 compared to 2012. Because of the wide variability this is not a statistical significant increase, but it was not a reassuring finding for the workers. The high frequency of exceedance of the OEL for 15 min confirmed the necessity to keep on wearing respiratory protective equipment (full face mask with a class A1 organic vapor filter for FA, so called ‘A1 plus formaldehyde’, 3M Nederland, Delft, The Netherlands). As an extra precaution, the entrance to the room is equipped with a red flash light to indicate that the room may not be entered as long as the FA-related task is performed. The flash light is turned on manually by the worker who is performing the task.
A positive finding in 2014 was the observation of a reduction of the exposure related to the ‘flushing’ of specimens in the embalming room that was carried out using the containment with LEV. The average exposure related to this task translated in a tenfold decrease, leading to an initial frequency of exceeding the OEL from 66.7% in 2014 to 0% in 2017. The overall evaluation of task-based exposure shows a predicted probability of exceeding the 15 min OEL of 1.6%. A priority for improvement is the ventilation rate in TR-3 that was only 60% of the target value. The limited ventilation capacity would be more efficiently used when installing an LEV.
4.3. Strengths and Limitations of the Study
An obvious strength of this follow-up study was the worker’s participation in finding solutions to improve the technical infrastructure and (most importantly) the daily work practice. The change from 2012 to 2014 really showed one cannot rely on technological changes, only. Especially in the second stage (2014–2017) the workers themselves were able to further reduce exposure. There was good interaction and participation of the workers in identifying solutions to the critical points that were raised during the exposure studies. This process was well supported by occupational hygienists and ventilation technicians of the hospital.
Another strength of the approach was the choice for a reliable measurement methodology. The air sampling and in-house analysis of the FA-DNPH with a high standard of quality assurance based on including a complete calibration curve in each sample run.
The study had a focus on formaldehyde and did not consider co-exposure to methanol from the (37%/10% methanol/FA stock). Also the study did not include evaluations of skin contact and potential uptake by skin exposure (Figure 11
). We did not include any short (<1 min) exposure ‘peak’ measurements; we restricted ourselves to measurements with a duration of 15 min. This was related to the lack of access to reliable measurement principle that could be used to apply real-time observations of FA exposure. Another reason for using the DNPH measurement over 15 min intervals is the definition of the established OEL for short-term exposure.
The number of task-based measurements was lower in 2017 (n = 12) compared to 2012 (n = 19) and 2014 (n = 21). This could lead to an underestimation of exposure measured by the number of non-compliant measurements (29 in 2012, 33 in 2014 and 0 in 2017). As the measurements have a log-normal distribution, when increasing the number of measurements the probability of finding an occasionally higher exposure increases as well. As a precaution to potential underestimation of exceedance of the OEL we calculated the probability based on the variability in the 12 task-based personal air sampling data of 2017. This resulted in a predicted probability of exceedance of the OEL of 1.6%. Regarding the suggestion that peak exposures may be of particular relevance for an increased cancer risk, future surveillance measurements are recommended.
4.4. Interpretation of Results in the Context of Published Literature
As was stated in one of the early occupational hygiene studies ‘Each gross anatomy facility is a unique environment’ [1
]. In those days solutions of 5% were regularly used, whereas other studies proposed to use alternative formulations of preserving solutions, reducing FA contents to 0.5–0.75% [29
]. In our laboratory the strategy is to keep the anatomical specimens in 1.9–2.2 vol % of FA for long-term storage but flush materials thoroughly during 24 to 48 h prior to use for research and teaching purposes.
The highest concentration of 126.6 µg/m3
was measured in the small section room when the thorax preparations were on display. This is in accordance with earlier reports on dissection of body cavities or deep structures when exposures were higher compared to dissection of superficial structures such as extremities [2
]. Of thirteen occupational job titles reported in a national survey in Italy, the exposure of medical doctors was reported to be the highest with a geometric median of 375 µg/m3
with 43% of the measurements in the healthcare sector exceeding 250 µg/m3
Klein and co-workers suggested that planning and constructing a large-scale teaching facility with LEV installed at dissection tables leads to working conditions with consistent low exposures, below 125 µg/m3
]. Room ventilation is a general requirement but is not sufficient. Klein and co-workers observed that a two-fold increase of the air exchange rate did not result in a reduction of exposure in the breathing zone [30
]. When calculating emission factors, it can be demonstrated that it is more efficient to use LEV instead of room ventilation [33
]. The situation in the embalming room in the present study showed that a tenfold reduction in the breathing zone can be achieved by use of LEV [35
]. LEV systems installed in dissection tables is current practice [36
] and commercially available ducted grossing stations have been evaluated as effective systems [33
], some equipped with LEV [39
] or with LEV and with UV-powered photocatalytic filters to decompose FA [30
Like in some other studies we found that both instructors and students in anatomy teaching are expected to have similar exposure levels as the workers [2
]. Vohra [40
] reported exposures in instructors to be higher compared to students. In a Thai gross anatomy laboratory mean (±sd) FA concentrations in the breathing zone were reported to be much higher (616 ± 116 µg/m3
) than in the current study, which resulted in a range of symptoms including burning eyes and burning nose [41
]. Other studies reported lower exposures [30
Peak exposures are expected to be most relevant for employed staff. Students who do internships or are hired to do preparation work may also be exposed to peaks, depending on the type of research they are doing. Especially for long-term research and preparation assignments a well-ventilated room is not sufficient to prevent exposure in excess of current OELs. LEV should be installed and personal protective equipment provided along with good instructions of how to use them. In our study, removing skin from a cadaver was identified as a task associated with high exposures up to almost tenfold the OEL of 8 h. Skin incisions and subsequent release from subcutaneous adipose tissue were identified as a high emission sources, especially in embalmed female cadavers [26
The potential of FA to cause allergic responses is well known. In animal studies repeated dermal contact with 4% solutions of FA in water, induced allergic responses in mice over a period of two weeks [42
]. The allergic potency of direct skin contact is also supported by (limited) human data implicating FA as a causal factor in allergic contact dermatitis [6
]. Therefore, in addition to prevention of inhalation, also skin protection should be or become a priority in gross anatomy facilities. In the Radboudumc facility the workers may need to consider to introduce long sleeve garment (Figure 11
Klein et al. [31
] observed a downward trend from original anatomy/pathology laboratories (1996–1999) to ‘enhanced’ original laboratories (2000–2002) and new lab (2003–2012) by use of 3 h personal measurements. The current study shows that in the same infrastructure a similar trend of exposure reduction could be achieved for fixed, full-shift and task-based measurements.