Thursday, May 26, 2005

Money Tox - Pharma Walks

BrooklynDodger fumes at this inadequate study cluttering the abstractosphere [medline] with an uniformative null finding aimed at reducing concern for occupational exposures in the pharma industry. The Dodger hopes reviewers and editors will step up to responsibility in this matter, because careers are built on paper counts, not necesarily paper content.

The Dodger has posted the abstract in full below, so as to allow the gentle reader to note what was put in the full text but left out here.

Other than drawing null conclusions from an inadequate study, the investigators display a new trick to enhance and then ignore the healthy worker effect. By mixing salaried with hourly workers [or, professional, executive, administrative and clerical with operatives and service personnel] for cohort analyses, and then ignoring them for exposure response analyses, the investigators are able to sweep any possible work related associations under the rug.

The paper is published with no attribution of funding or statement on conflict in the end matter. The text notes "In 1995, the company commissioned us to design and conduct the study." Obviously there's a conflict which should be disclosed in the end matter: the degree of conflict would be reduced were the investigators selected by competitive application and independent peer review.

The Dodger notes that the same PI has published on a cohort from Merck, and wonders if the paper here is an offshoot of the larger other study. The Dodger can't resist noting the proud history of disclosing health issues in Merck's core business: Merck's Vioxx Liability Could Reach $38 Billion - a result of taking Vioxx were to receive an average-sized legal settlement,Evans wrote in a research report, Merck's liability could reach $38 billion. 2004/12/03/1203automarketscan13.html The Dodger will revisit the 50 year study in a later post.

J Occup Environ Med. 2004 Feb;46(2):161-6.

A 50-year historical cohort mortality study of workers in a pharmaceutical plant.

Dolan DG, Youk AO, Marsh GM, Buchanich JM.Merck & Company, Inc, Whitehouse Station, NJ 08889-0200, USA.

Having digressed, the Dodger returns to the study at hand.

The fatal flaw of this investigation was that only 112 or 5.6% of the total cohort was identified as deceased. They disclose and all cause SMR of 64 for the whole cohort, which approaches the world record low SMR. The SMR is a product of short observation, any employment related effect would have to come from comparing groups within the cohort, and with 112 cases total there's little change to do exposure response. However, comparing the part of the cohort who worked in production areas to those who didn't, the all-cause SMRs are 46 vs. 71 for production areas [which, the Dodger suspects, includes engineers and supervisors]. The Dodger would bet 50 cents that the increased mortality suffered by compared to salaried in this cohort achieved statistical significance.

The main finding in this study designed not to find anything, is 4 cases of non-Hodgkins Lymphoma, SMR = 307, marginally significant in the production cohort. This is compared to no cases of this disease in the office cohort. For exposure response analyses, the office cohort was excluded from the zero exposure group, attenuating measures of association. All these victims cam from among workers with at least 10 years of employment, and the SMR = 584 among those with 20 years since hire [a surrogate for latency from first exposure] was statistically significant.

The investigators say "we developed individual worker-level, multi-exposure profiles for all plant exposure agents with available air monitoring data (acetone, aceto-nitrile, di-methyl formamide, ethyl acetate, ethylene di-chloride, isopropyl alcohol, methlyene chloride, tri-chloroethane, and toluene)." Sounds like cranking the cash register on a grab bag of solvents, average exposure under 10 ppm, median exposures zero for several. MC, EDC and DMF are of some interest but not at these levels. The only shred of scientific merit in this effort would be effects of the product, but the product or class of product is not disclosed. Are they making birth control pills, viagra, anti-neoplastic agents, or are they packaging Tums?

Bottom line and new rules.

1. Abstracts should state whether study was adequate to address the goal of the study. In this case, the study was inadequate to support a null hypothesis that chemical exposures in the facility were not causing increase mortality.

2. Null results from inadequate studies should have no weight in risk assessment.

3. Associations revealed by inadequate studies should have some weight in risk assessment; these associations should not be discounted by failure to observe an exposure response relationship.

4. Publishing an SMR of salaried and hourly employees combined misleads the literature.

Ann Epidemiol. 2005 Feb;15(2):112-22.

Mortality patterns among workers in a US pharmaceutical production plant.

Marsh GM, Youk AO, Esmen NA, Buchanich JM.Department of Biostatistics, Graduate School of Public Health, University of Pittsburgh, Pittsburgh, PA 15261, USA.

PURPOSE: To examine mortality among workers in a pharmaceutical production plant and to address community concerns about 1980 to 1990 increases in local county cancer mortality rates. METHODS: Subjects were 1999 workers with some full-time employment during the period between 1970 and 1996. We identified deaths through the year 2000 and reconstructed exposures to nine chemical agents with available exposure measurements. Data analyses included standardized mortality ratios (SMRs) and time trends in local cancer mortality rates. RESULTS: We observed deficits in deaths from all causes combined, all cancers combined, and most cause of death categories examined. Male workers with potential plant exposure had excesses in deaths from all lymphatic-hematopoietic tissue cancers (LHTC), in particular non-Hodgkin's lymphoma (NHL), and respiratory system cancers (RSC) that were larger among long-term workers, but the pattern of findings suggested the excesses were probably not related to occupational factors at the plant. The increase in local county cancer mortality rates was simply the upward cycle of a periodic trend that peaked in 1990 and returned to 1980 levels in 2000. CONCLUSIONS: With the possible exceptions of LHTC, in particular NHL, and RSC, this study provided no evidence of elevated total or cause-specific cohort mortality risks. It does not appear that plant factors played a role in the 1980 to 1990 increases in local county cancer mortality rates.

No comments: