The medical technology community knows ethylene oxide (EtO) as an essential sterilisation tool. According to the Advanced Medical Technology Association (AdvaMed), the world’s largest association representing manufacturers of medical
technologies: “EtO is the most efficient and effective means of sterilisation available. In fact, for many products, EtO is the only acceptable method of sterilisation.”
Given the importance of EtO in the medical technology community, the draft document evaluating the carcinogenicity of EtO prepared by the US Environmental Protection Agency’s (EPA) National Center for Environmental Assessment (NCEA) released in
September 2006 is the cause of significant concern.
Many in the medical technology community believe the draft assessment greatly overestimates potential human health hazard from inhalation exposure to very low levels of EtO. NCEA has estimated that 0.33 parts per trillion (ppt) and 0.36ppt,
respectively, pose a one-in-a-million excess cancer risk from EtO occupational and residential exposures. Based on the many highly critical public comments on the draft assessment received to date, EtO producers and users alike believe that NCEA’s
draft cancer assessment document is seriously flawed and fails to provide a sound scientific basis on which to assess potential health risks from exposures to EtO.
Before focusing on the draft assessment itself, it is important to understand the context in which the assessment evolved. EPA is currently reassessing the carcinogenicity of EtO to update its file on IRIS (see box, p40). Once EPA completes an IRIS
file on a chemical, the values are available online and are relied upon extensively by other federal, state and regional regulators for a wide range of standard-setting purposes. IRIS values are used by state regulators to set environmental standards and
Chemical manufacturers and users are usually as engaged as the IRIS administrative process allows them to be to ensure that IRIS values are fully reflective of accurate and complete science, and fully consistent with prevailing legal standards. Since
risk managers in federal and state agencies and elsewhere in the private sector rely on IRIS values to assess risk, the manufacturers of EtO and its principal users, including the medical technology community, have been deeply involved in reviewing the
EPA intends that IRIS values identify the hazard posed by a particular chemical substance, not its risk. To quantify risk, EPA acknowledges that situational exposure information would need to be factored into the calculus. The IRIS values are crucial,
however, and significantly influence the outcome of the risk calculation.
The IRIS process is notoriously long and complicated, involving numerous internal and external peer review processes that typically span several years. The process has also been the subject of considerable debate, and EPA has tried to improve the IRIS
process over the years to make it more transparent and defensible. Nonetheless, many in the industry believe the process is not as transparent as it needs to be, given its significance, and that NCEA scientists are often overly conservative in
characterising the potential health hazards posed by chemical substances.
CAUSE FOR CONCERN
There is much disquiet regarding the draft assessment, and many of the comments submitted have been justifiably critical of NCEA’s scientific lapses. Many in the industry, including EtO producer AdvaMed, have strongly objected to NCEA’s
draft EtO cancer assessment. The comments offered by the American Chemistry
Council Ethylene Oxide/Ethylene Glycols Panel outline in considerable detail the reasons supporting its view that the SAB should reject the draft assessment.
Panel members are either manufacturers of most of the EtO produced in the US or they are major users of it, and the panel is the successor organisation to the EOIC. The panel’s comments are extensive and the brief summary below focuses only on
its main science, policy and legal objections to the draft assessment. In the panel’s view, key flaws and deficiencies in NCEA’s draft cancer assessment are:
- Failure to meet required legal standards. The draft assessment fails to meet the rigorous legal standard of quality required under the Information Quality Act and the EPA Guidelines or Carcinogen Risk Assessment. It also fails to use all
available epidemiologic data, inappropriately bases its evaluation on summaries of statistics rather than primary source information and excludes certain data.
- Unjustified departures from past EPA risk assessment practices. The draft assessment inexplicably departs from past EPA risk assessment practices in material respects, which result in overstatements of potential cancer risks from low levels of EtO exposure. For example, EPA assumed an 85-year lifetime rather than the more traditional 70-year life expectancy and relied on the lower bound of the point of departure rather than the best estimate when using human data. These and other decisions were not explained nor are they scientifically justified. They also lead to overstatements of potential risk from exposure to low levels of EtO.
- Flawed statistical analysis, which greatly overstates potential risk. EPA’s assessment is based on invalid linear regressions on odds ratios rather than the more traditional and appropriate review of available individual subject data. EPA also neglected to include all available epidemiologic data, among other statistical deficiencies.
- Exclusion of data. The cancer assessment inexplicably excludes important data, including data on males in a particular study, thus precluding a valid weight-of-evidence approach.
- Failure to pass simple reality checks. For example, the NCEA estimate of less than 1ppt EtO, which EPA asserts poses a one-in-a-million excess cancer risk, is two to three orders of magnitude below the endogenous level of EtO produced naturally in humans.
- Computational errors, which skew results. NCEA’s slope and standard error are calculated incorrectly and the method of incorporating an age-dependant adjustment factor is mathematically incorrect. These errors overestimate the potential
human health risk.
- Failure to inform decision-makers of the potentially adverse public health implications of the draft assessment. A key concern for the medical technology community is the draft assessment’s total failure to inform risk managers in the government and private sectors of the potentially adverse public health implications of the draft assessment, which identify the most conservative estimates ever proposed for EtO by any government body. Reliance on overly conservative risk estimates could adversely impact EtO’s use and availability at a time when public health experts believe controlling infections is more important than ever.
For these and many other reasons set forth in its comments, the panel has urged NCEA to substantially revise the draft assessment to address the many deficiencies and flaws.
EtO manufacturers and the medical products community are optimistic that the SAB will concur with the many adverse comments submitted on the draft assessment and urge EPA’s NCEA to revise the draft assessment significantly to correct these flaws and deficiencies.
Because the SAB is merely an advisory body, however, EPA is not required to accept its findings. It is important, therefore, that our industry monitors this regulatory initiative and continues to urge EPA to use the best science in developing its final assessment. This will help ensure that the medical technology community will be able to continue to use EtO to provide safe, effective, and sterile medical devices to combat infection and protect patient health.