Improving the ACGIH threshold limit value (TLV) process

Abstract

The intent of this manuscript is to elucidate needed improvements in American Conference of Governmental Industrial Hygienists (ACGIH) threshold limit value (TLV) deliberations. More broadly, irreproducibility and bias adversely impact the collection, interpretation, statistical analysis, presentation, and reporting of results in many fields. In 2012, Begley and Ellis reported that scientists at Amgen had attempted to confirm published findings related to research topics of possible interest to Amgen. Fifty-three papers were deemed "landmark" studies. The authors were "shocked" when scientific findings were confirmed in only 6 (11%) cases. Many studies have confirmed that the peer-reviewed literature in biomedicine is in the midst of an irreproducibility crisis. Compounding the irreproducibility crisis is the existence of a significant bias against the publication of negative results. In the toxicology setting, negative toxicity test results are infrequently published as compared with reports that a chemical possesses a particular toxicity in a given test. Despite these deficiencies, the ACGIH states that ". .. the TLV ®-CS Committee preferably relies on published, peer reviewed literature available in the public domain." The primarily academic studies published in the peer-reviewed literature upon which ACGIH relies to determine TLVs rarely report raw data not already statistically transformed that are thus incalculable. In contrast, consideration of unpublished studies funded by industry, the vast majority of which are good laboratory practice-conducted contract lab studies, is only acceptable to ACGIH if the data owner provides the raw data to third parties upon request. This asymmetry in both the source of data emphasized, and inability to independently statistically analyze findings reported in the published academic literature, introduces a strong skew toward reliance on unverifiable although published measurements in the TLV process. Since Occupational Safety and Health Administration (OSHA) recommends that workplaces rely on ACGIH TLVs and National Institute for Occupational Safety and Health recommended exposure limits rather than older OSHA permissible exposure limit values to optimize worker safety, ACGIH should adopt a more transparent and science-based process.

142 ACGIH Threshold Limit Values® (TLV®s) established from 2008-2018 lack consistency and transparency

“…Recently, we addressed the lack of transparency in the ACGIH TLV process. 14 The current analysis extends the previous study by examining every TLV set during the 10-year interval between 2008 and 2018.…”

“…Compounding the already extremely problematic situation, many studies conducted over the last several years have demonstrated that a large but not exactly determinable percentage of the results in the peer-reviewed academic literature are not reproducible in another laboratory attempting to employ the original protocol. 14 Based on analyses conducted in different fields of study, a reasonable estimate would be that about 40-50% of the biomedical literature is not replicable. 14,[44][45][46][47][48] ACGIH states that the primary resource upon which it relies in setting TLVs is this same body of peer-reviewed academic literature in the midst of a replicability crisis.…”

“…14 Based on analyses conducted in different fields of study, a reasonable estimate would be that about 40-50% of the biomedical literature is not replicable. 14,[44][45][46][47][48] ACGIH states that the primary resource upon which it relies in setting TLVs is this same body of peer-reviewed academic literature in the midst of a replicability crisis. Given the enormous number of workers whose safety is affected by the TLV process (Table 3), and the hundreds of billions of dollars of commerce involved (Table 3), the opaqueness of the current TLV setting process should be replaced by a protocol that describes the selection of studies, numerical values extracted for consideration, statistical methods employed, actual statistical calculations, raw data analyzed, and logic employed.…”

“…In addition, a well-established bias against the publication of negative results can additionally skew evaluations of the toxicological profile of a chemical toward overestimate of hazard. 14 In this analysis, we evaluated the transparency and consistency of the data relied upon by ACGIH to set TLVs for 145 substances or groups of substances between the years 2008 and 2018. Table 1…”

“…Despite the serious adverse consequences of an inaccurately determined TLV, the current TLV determination process is primarily dependent on an academic peer-reviewed literature in the midst of an irreproducibility crisis not yet fully explained or addressed. 14 Also, a significant number of studies relied upon by ACGIH are not published and the data are not accessible to the public, thereby further exacerbating deficiencies in the TLV process. In addition, a well-established bias against the publication of negative results can additionally skew evaluations of the toxicological profile of a chemical toward overestimate of hazard.…”