Memorandum

To:		Patricia Mercer, EPA

From:		Maribelle Rodríguez and Earl Harris

Date:		September 21, 2007

Subject:	Analysis 6 under Technical Directive 1 of Work Assignment
Base-25, Contract No. EP-W-07-003

The above-referenced technical directive asks for clarification on the
types of operations of both research labs and clinical labs in teaching
and research hospitals.  EPA would like to know if clinical labs and
research labs can be distinguished from one another based on whether
they conduct routine medical procedures or clinical trials or other
types of medical research.

  In communications with you, ICF agreed to call a number of medical
associations (e.g., Association of American Medical Colleges) and
colleges and universities (CUs) to answer the threshold question: 
“Can labs in a hospital be separated into two distinct groups:  those
that do routine tests and those that do clinical trials, or do labs
normally do some of both?  And can research labs be separated from both
possible types of clinical labs either physically or via definitions?”
 

This memorandum lays out the findings of our research.  Section I
attempts to answer the threshold question stated above.  Section II
briefly outlines a very preliminary framework for extending Subpart K
regulations to research labs, if EPA chooses to do so.  Section III lays
out some possible next steps to evaluate the straw approach further.

I.	Distinguishing Between Clinical Labs and Research Labs

	ICF contacted a number of lab representatives (e.g., EH&S staff at CUs)
to learn if clinical labs and research labs can be separated into two
distinct groups, and if research labs are separated physically from
clinical labs.  Our research focused specifically on labs that are
located “within the doors of the hospital.”  Labs located outside
the hospital (e.g., in a separate building on campus, such as a research
institute) were not a focus.  ICF’s conversations are summarized
below.

I.1	Summary of Conversations

	CU Hospitals 

Weill Hospital of Cornell University, New York Presbyterian Hospital. 
Erik Talley, EH&S Director, said that Weill College has 1) clinical-only
labs; 2) research-only labs; and 3) labs that perform both research and
clinical work.  He was unable to quantify the labs in these categories,
but said the research-only labs are the smallest in number.  In
particular, he said that the research-only labs conduct basic research
(e.g., examination of cell growth), clinical trials, and other research
(e.g., fine-tuning diagnostic techniques).  He generally considers a
“research lab” to be a lab that does experimentation usually under a
grant, e.g., from the National Institutes of Health (NIH).  He cautioned
that there may be collaboration among different types of labs, e.g.,
sharing of personnel and equipment.  For example, a physician in a
clinical lab may buy a piece of equipment that the principal
investigators in a research lab may want to use, and they will work out
an arrangement to share the equipment.  Therefore, it is not possible to
say in all cases that a particular lab does only clinical work or
research.  

University of Washington Medical Center and Harborview Medical Center. 
Doug Gallucci and Sheila Lockwood of the EH&S staff said that their two
medical centers have 1) clinical-only labs; 2) research-only labs; 3)
labs that perform both research and clinical work; and 4) teaching labs
(i.e., labs devoted to teaching students on such things as use of
equipment and proper health care procedures; these labs are linked to
academic programs).  All of them can generate hazardous waste.  Examples
of work in their research-only labs include such things as
experimentation to improve diagnostic testing processes, research to
improve facility design/layout (e.g. to improve indoor air in hospital
settings), as well as clinical trials.  

Ohio State University Hospitals.  Michael St. Clair, EH&S Director, said
that the university has four medical centers.  These centers have 1)
clinical-only labs; 2) research-only labs; and 3) labs that perform both
research and clinical work.  For example, one of the hospitals contains
a cancer institute that performs research on cells for cancer treatment.

University of Illinois Medical Center.  Cynthia Klein Banai of the EH&S
staff said that its medical center has both clinical and research labs. 
However, the research labs are very rare (e.g., roughly 1% of al labs in
the center).  



Emory University Hospital.  Scott Thomaston of EH&S staff said that its
hospital contains only clinical labs, to his knowledge.  It does not
contain labs that are dedicated to research.  The research labs are
located in other buildings on the campus.

Municipal Hospital with Affiliation to CU

Hartford Hospital in Connecticut.  Jim Baio, Director of Environmental
Services, said that, to his knowledge, the hospital has clinical labs. 
It also has some labs that perform surgical research on animals as part
of a training program.  His hospital and other municipal hospitals in
the state have agreements with the University of Connecticut to accept
residents on a rotational basis.  He generally agreed that hazardous
wastes from clinical labs are routine.  They vary when the hospital
establishes a new program (e.g., a new inoculation program to address an
emerging disease).  This may give rise to a new hazardous waste stream.

Associations

Association of American Medical Colleges.  Steven Heinig, Senior
Research Fellow, said that hospitals vary greatly in regard to their
design and operations.  He said that some hospitals have clinical-only
labs and research-only labs, as well as labs that do both.  He said that
the Clinical Laboratory Improvement Amendments (CLIA) applies to
laboratories that do clinical testing but not to research.  The purpose
of the CLIA program is to ensure quality laboratory testing.  A clinical
laboratory must be accredited by an accrediting organization (e.g.,
College of American Pathologists) and obtain a CLIA certificate.  If it
conducts only simple procedures, it will obtain a CLIA waiver.  He
agreed that CLIA may be a useful way to distinguish a clinical lab for
purposes of the EPA rule.  

College of American Pathologists.  Donna Meyer said that some hospitals
may have research-only labs in part because of the tax consequences.  If
the hospital is tax exempt, it may want to segregate their
patient-related care from their research, if the research is not covered
under the tax exemption (e.g., research that is funded by a private
company).  It may devote floors or a wing of its hospital for research.
This helps with cost accounting and other aspects of compliance with tax
law.  Ms. Meyer believes this is very common in CUs.

I.2	Summary

Based on our discussions, several observations can be made:

CU hospitals have a variety of labs that may be physically separate
locations:  1) clinical-only labs; 2) research-only labs; 3) teaching
labs; and 4) labs that perform a combination of these (e.g., clinical
and research activities).   All of them may generate hazardous waste. 
In particular, research-only labs perform a wide range of research,
including basic research (e.g., human genome research) and applied
research (e.g., clinical trials, therapeutic interventions).  Lab
representatives gave several reasons why research-only labs may be
physically separated from clinical labs in a hospital:

Some hospitals find it makes sense to keep researchers together in one
part of the hospital and patient-care workers in another part.  

Keeping them physically separated can lead to greater efficiencies.  For
example, principal investigators performing related research projects
can collaborate more easily.  The same is true for clinical workers in
handling and processing patient specimens.

Keeping them separate aids in tracking costs separately for
reimbursement purposes.  For example, some patient-related costs are
reported to, and reimbursed by, Medicaid and Medicare or other source
(e.g., patient or insurer), while research-only costs may be reimbursed
by a granting source, such as the NIH.  

Clinical labs are generally regulated differently than research-only
labs.  Keeping them separate can aid in compliance oversight by
laboratory compliance staff.

Some hospitals simply evolved to include physically separate spaces for
clinical labs and research labs, i.e., because of historical factors. 
For example, the original design of a hospital may have resulted in the
labs being physically separated, or an organization may have donated
funds to establish a wing of a hospital devoted to a particular type of
research. 

Some hospitals segregate research-only labs from the rest of the
hospital because of the tax implications, as suggested by a
representative of the College of Pathologists, above.

All of the CU lab representatives, except for one, stated that, in
general, research labs satisfy the proposed rule’s applicability
criteria better than clinical labs.  In particular, they believe that,
everything else being equal, research labs have:

Higher number of waste streams and greater variability in waste type. 
This is because of the diversity in experiments being conducted in
research labs.  Clinical labs, on the other hand, normally have standard
procedures and tests that are performed.  Test procedures in clinical
labs, generally, are subject to greater and more uniform regulation
(e.g., CLIA) and oversight (e.g., oversight by hospital management and
CLIA-accrediting organization), which serve to help standardize their
processes and make waste generation patterns more stable.  Research labs
may be subject to regulation (e.g., NIH-stipulated requirements under a
grant agreement to comply with CLIA), but some types of research are
excluded from CLIA and driven more by research-specific protocols.

Smaller quantities of individual waste.  Research laboratories normally
have fewer routine processes that rely on the same chemicals and
instruments than do clinical labs.  Because clinical labs perform tests
repeatedly and in higher volume (e.g., the same diagnostic tests
performed for all patients getting a particular examination), they
generate larger quantities of some wastes.

Greater student involvement.  The representatives stated that clinical
labs often rely more on hospital staff to perform the tests, whereas
research labs more often involve students.

Note:  These representatives emphasized that it is impossible to make an
absolute statement that research labs meet the applicability criteria
better than clinical labs.  Lab- and hospital-specific factors (e.g.,
size and number of labs, number of employees, types/number of test
procedures, characteristics/diagnostic needs of patients) vary greatly,
making generalized comparisons blurry.

The Clinical Laboratory Improvement Amendments (CLIA) could be helpful
to EPA in differentiating between clinical and research laboratories for
purposes of extending Subpart K to research labs only.  CLIA is codified
in the regulations of the Department of Health and Human Services under
Title 42 of the CFR.  The purpose of CLIA is to establish quality
standards for laboratory testing to ensure the accuracy, reliability and
timeliness of patient test results.  A laboratory is subject to CLIA if
it meets the definition of “laboratory” at 42 CFR 493.2, except as
otherwise specified: 

Laboratory means a facility for the biological, microbiological,
serological, chemical, immunohematological, hematological, biophysical,
cytological, pathological, or other examination of materials derived
from the human body for the purpose of providing information for the
diagnosis, prevention, or treatment of any disease or impairment of, or
the assessment of the health of, human beings. These examinations also
include procedures to determine, measure, or otherwise describe the
presence or absence of various substances or organisms in the body.
Facilities only collecting or preparing specimens (or both) or only
serving as a mailing service and not performing testing are not
considered laboratories.”   

If an entity performs even one of these tests, it is considered under
CLIA to be a laboratory and must register with the CLIA program as
specified.  A laboratory subject to CLIA must, among other things, apply
for and obtain a certificate from the CLIA program that corresponds to
the complexity of the tests performed.  For example, a laboratory that
performs only waived tests must obtain a Certificate of Waiver.  A
laboratory that performs specific microscopy procedures during the
course of a patient’s visit must obtain a Certificate for Provider
Performed Microscopy (PPM) Procedures.  A laboratory must obtain a
Certificate of Accreditation (COA) on the basis of the laboratory’s
accreditation by an accreditation organization approved by the Centers
for Medicare and Medicaid Services (CMS).  This type of certificate is
issued to a laboratory that performs nonwaived (moderate and/or high
complexity) testing.  The laboratory must comply with the requirements
and procedures (e.g., fees, surveys, quality standards) laid out in the
CLIA regulations. 

Note, however, that 42 CFR 493.3(b) sets forth an exception from CLIA
for “research laboratories that test human specimens but do not report
patient specific results for the diagnosis, prevention or treatment of
any disease or impairment of, or the assessment of the health of
individual patients.”  CMS representatives said that the exception can
generally be interpreted to exclude research from regulation if:

The research involves blind specimens (i.e., the specimens do not
include patient-identifying information, such as patient names); and

Test results from the research (e.g., clinical trial results) are not
returned to the patients or physicians for the patients’ care.  
Rather, test results should be applied for generalized purposes (e.g.,
statistical interpretation).  CLIA representatives pointed out that a
clinical trial could be subject to CLIA, depending of how it is
conducted: 

A clinical trial that involves non-blind specimens and/or whose results
are returned to the patient or physician for the patients’ care would
meet the definition of “laboratory” under CLIA.

Clinical trials that involve blind specimens and whose results are not
returned to the patient or physician for the patients’ care would be
excluded from CLIA regulation.  Note, however, that some granting
organizations (e.g., NIH and Food and Drug Administration) stipulate
that a grantee’s clinical trial must comply with CLIA, even if CLIA
does not otherwise apply.

In summary, the CLIA regulations could provide a useful mechanism to
distinguish clinical labs and research labs under Subpart K.  Labs
meeting the “laboratory” definition at 42 CFR 493.2 are subject to
CLIA regulation.  These labs can be categorized as clinical labs. 
Research laboratories that satisfy the exception at 42 CFR 493.3(b) are
excluded from CLIA.  These labs can be categorized as non-clinical labs.
 In its preamble and rule text, EPA could distinguish between clinical
labs and research labs on the basis of the CLIA lab definition and
exception, respectively.

II.	 Straw Approach for Extending Subpart K to Research Labs

	This section lays out a very preliminary (i.e., straw) approach for
extending Subpart K to research labs.  The straw approach is meant only
to provoke discussion as one of several options that EPA may want to
consider.  More research is needed to flesh out the approach in full. 
Additional research also may be needed to determine if inclusion of
research labs under Subpart K is appropriate in any case.

II.1 	Outline of Straw Approach

 

EPA could consider taking the following steps to integrate research labs
into the scope of Subpart K, if the Agency decides that it is desirable
to do so (i.e., based on further research and consideration as needed):

Explain in the preamble to the final rule 1) the differences between
clinical and research labs in CU hospitals in regard to waste generation
patterns (e.g., greater waste variability from research labs); and 2)
readily available data that demonstrate why research labs should be
included under Subpart K but not clinical labs.

Discuss CLIA briefly in the preamble, indicating that labs meeting the
definition of “laboratory” at 42 CFR 493.2 (or the comparable
definition if approved state programs) are considered clinical labs and
are therefore ineligible for Subpart K.  EPA should also indicate that
research labs are excluded from CLIA as specified.  Related to this, EPA
will need to clarify certain issues.  For example, are research labs
that conduct clinical trials eligible for Subpart K, if the trials
involve patient-specific testing and are therefore subject to CLIA?

Consider revising the definition of “laboratory” at 40 CFR 262.200
to indicate that clinical labs are not included under it. 

II.2	Analysis of Straw Approach

Pros

Relies on the existing regulations of CLIA to distinguish between
“clinical laboratory” and “research laboratory.”  This relieves
EPA from the need to create a new distinction.  Creating a RCRA-specific
distinction could be unnecessarily confusing and create an additional
compliance burden for lab managers.

Is relatively simple to comply with, relative to approaches that create
new distinctions.  Laboratories (e.g., physicians, ES&H staff) should
already be familiar with the CLIA-related terms and applicability
requirements.  

Relies on a relatively tangible, straight-forward mechanism for state
enforcement personnel to confirm if a lab qualifies as a research lab
for purposes of Subpart K, i.e., if the lab has a CLIA certificate or
waiver, it is not eligible for Subpart K. 

Cons

Enables all research labs that are excluded from CLIA to invoke Subpart
K.  EPA may find this too accommodating an approach and desire further
research to identify those particular types of research labs that best
meet the Subpart K eligibility criteria.  ICF is uncertain, however, if
this research would be possible within EPA’s timeframe for rule
finalization.

May require some clarifications (e.g., in the preamble or regulations)
to resolve potential compliance issues, e.g., regarding whether labs
conducting non-blind clinical trials qualify as “research labs”
under Subpart K.

Summary

	It appears that the benefits of the straw approach are that it is
relatively straightforward and relies on existing regulatory
distinctions regarding clinical and research labs.  Familiarity with
CLIA may make it easier for labs to determine their eligibility under
Subpart K, in comparison with a new set of RCRA distinctions.  The main
drawback of the straw approach is that it has not been fully fleshed out
and some questions remain.  

III.	Next Steps

	Following are some possible steps in the near term:

Continue to evaluate the straw approach (e.g., to resolve the questions
raised in this memorandum and by stakeholders), as well as other viable
approaches.  

Consider if hard data are needed (and exist) to substantiate extending
Subpart K to research labs but not clinical labs (e.g., based on waste
generation patterns).  Based on conversations with EH&S staff, it may
not be easy to collect hazardous waste data on research labs and
clinical labs separately. 

Consider if EPA requires further research to identify the particular
types of research labs that best meet the Subpart K eligibility
criteria, so that the rule can be applied solely to them.

  It is possible that the other CUs contacted also have teaching labs. 
These representatives did not mention such labs, but they could have
overlooked them.  They were not a focus of our conversations.

 Erik Talley of Cornell University disagreed with Ms. Meyer’s
explanation regarding the tax incentives for segregating research labs. 
ICF has not resolved this disagreement.

  Ibid.

 The exception was Sheila Lockwood of University of Washington.  She
believes that lab-specific factors are too variable (e.g., types of test
procedures and chemicals used) to say that research labs satisfy the
criteria better than clinical labs.  

  CLIA applies to laboratories seeking payment under the Medicare and
Medicaid programs.

 42 CFR 493.3(b) establishes some exceptions from CLIA regulation (e.g.,
for forensics labs, research labs).  

 As defined by CLIA, waived tests are categorized as “simple
laboratory examinations and procedures that have an insignificant risk
of an erroneous result.”

 There are six CMS-approved accreditation organizations.

 Among other things, the application for certification requests the
physical location (i.e., building, floor, suite) of the laboratories
covered by the certification.  Hence, it can be assumed generally that
the approximate physical location of a clinical lab is spelled out.

 Text in this paragraph is paraphrased from the document, “Clinical
Laboratory Improvement Amendments: How to Obtain a CLIA Certificate.” 
It is available at:
http://www.cms.hhs.gov/CLIA/downloads/HowObtainCLIACertificate.pdf.

 Conversations with Donna Phillips-DiMaria of the Delaware State Public
Health Authority; Joanne Sparhawk of the California Department of Health
Services; and Judy Yost, Director, Division of Laboratories, Centers for
Medicare and Medicaid Services (CMS).

 In speaking with CMS staff, ICF learned that labs in CMS-approved
states are not subject to Federal CLIA.  They are subject to the
state’s program.  However, state programs incorporate the Federal
definition.  This should be confirmed.

  There are exceptions to this.   For example, some research-only labs
voluntarily obtain a CLIA certificate to demonstrate to granting
institutions that their research adheres to CLIA standards.  In such
cases, this would need to be explained to the enforcement personnel.

  PAGE  9 

