IX.  Untapped Data

We mentioned at the outset four sources of data that had already been
gathered and that could be tallied and analyzed to obtain incidence
and prevalence data.

1.  National Heath Care Survey.  A possibly significant and as yet
untapped source of Parkinson's disease incidence and prevalence data
is the National Center for Health Statistics (NCHS), a branch of the
U.S. Center for Disease Control.  The NCHS conducts the ongoing
National Health Care Survey, which is actually a group of many
individual surveys.  More detail about these surveys can be found at
the NCHS web page
http://www.cdc.gov/nchswww/about/major/nhcs/nhcs.htm.

The surveys which contain data which is particularly relevant to our
question are the National Ambulatory Medical Care Survey (NAMCS) and
the National Nursing Home Survey (NNHS). The NNHS was discussed
above.  The NAMCS, which has been conducted annually, provides data
on visits to physicians, emergency rooms and outpatient clinics,
including physicians' diagnoses and medications.  The survey sample
consists of some 40,000 visits to about 3000 physicians.  Recently
Hospital clinic visits have been surveyed separately by means of the
new National Hospital Ambulatory Medical Survey (NHAMS).

The published reports of the surveys, which can be viewed on-line, do
not break down the results to show every disease. But the detail
files of the questionnaire response data contain individual patient
diagnoses.  These detail files, which are on CD-ROMs, can be
purchased from the National Technical Information Service.  Also,
major university libraries have copies of the CD-ROMs in their
collections, where they are available to student and faculty
researchers.  All that is needed is to tally and analyze the data.

Allowance must be made for the fact that the NAMCS data is about
visits, not patients.  There appears to be no patient identifier
which ties together different visits by the same patient.  Thus an
average visits per patient factor needs to be developed for each
diagnosis in order to use this data for computing prevalence.  As far
as incidence goes, there does not appear to be data to indicate that
a diagnosis is new.

2.  Pharmacy data.  Methodological studies of drug sales conclude
that drug sales are a good proxy indicator of prevalence (5).  The
principal Parkinson's drug levodopa and the dopamine agonists are
good identifiers of Parkinson's patients, because they are not
generally used for treating other conditions, and because most
people in treatment for Parkinson's disease take them.  Since
individual Parkinson's patients may be taking any number of
combinations of the available drugs, or even no drugs at all, sales
volume in itself is not useful.  But if it were possible through a
broadly representative survey to determine an average drug profile,
then total sales divided by average individual utilization would
approximate the number of patients in treatment.

Pharmacy customer file data would also be useful.  The ratio of the
number of all pharmacy customers receiving Parkinson's disease drug
prescriptions to the total of all customers receiving any drug
prescriptions in the same period could be determined.  This could
then be used in combination with estimates from the National Health
Interview Survey of the numbers of people in the U.S. who made
visits to physician offices and hospital clinics for whom drugs were
prescribed.  The tally of such people could be multiplied by the
ratio of pharmacy customers receiving Parkinson's disease drugs to all
pharmacy drug customers to give an estimate of the number of people
with Parkinson's disease who are in treatment.

It is worth mentioning in this connection that there are commercial
firms that determine and make available (for a cost) to the health
industry information on what is referred to as the "disease market,"
including incidence and prevalence.  One of these is Menlo Biomedical
Associates of Santa Cruz, Ca.  They maintain a website at
http://www.mbio.com for their clients, who may select diseases and
view the "market value."

3.  Insurance claims data.  Insurance claims data contains patient
identifiers and diagnoses.  It would be technically simple to tally
patients in the claims files by diagnosis.  Probably the third party
payers already produce or have the capability of producing such
totals.  What would be difficult is the administrative aspect of
assembling or even getting access to the data or obtaining an
extract thereof.

Of course this data does not include people who are uninsured.
However, there are estimates of the numbers of people who are
uninsured, and one would expect the prevalence of Parkinson's
disease among them to be similar.

4.  Medical records data.  A good example of a large medical facility
with an extensive patient population is the Northwest Division of
Kaiser Permanente.  Their databases contain the medical records of
400,000 people in northwestern Oregon and Southwestern Washington
State.  The Center for Health Research is an independent organization
which conducts epidemiological studies using the Division's
databases.  (We have seen no reference to any CHR studies about the
incidence and prevalence of Parkinson's disease.)  The CHR believes
that the medical records are strongly representative of the entire
population of that area.  For further information, see the CHR web
pages starting at http://www.kpnw.org/~chr.


  X.  Parkinson's Disease and the NHIS: A Serious Omission

In the ongoing National Health Interview Survey (NHIS), sample
households are surveyed by U.S. Census personnel, who ask a set of
health-related questions (30).  One category of data that the survey
collects is "chronic conditions." Interviewees are asked whether
they have any of a number of "conditions" from a "conditions list,"
which is a checklist component of the questionnaire.  Within the
checklist there is a sub-category of "conditions affecting the
nervous system."  The specific conditions which are listed in this
sub-category are "epilepsy; REPEATED seizures, convulsions or
blackouts; multiple sclerosis; migraine; FREQUENT headaches; and
neuralgia or neuritis."  (The bold capital letters occur on the
questionnaire form.)

We do not know why just these specific conditions are listed. But for
some reason, if not from pure oversight, Parkinson's disease is
missing from the checklist.

There is another opportunity to report Parkinson's disease in the
survey, but only as required by the context.  It is in the section
intended to count people having limitations on activities.
Interviewees are asked whether there is any impairment or health
problem 1) that prevents them from working or doing housework, 2)
that limits them in any way in any activities, and 3) that requires
assistance in personal care activities, and if so, what is the
impairment or problem in each of these instances.  The same question
is also asked in regard to children not being able to attend school.
Any "condition" or "impairment" may be given in answer to these
questions.  Several may be recorded in each case, although the
interviewees must designate one as the major cause of the limitation.
But such questions are apparently intended to indicate time or
productivity lost due to ill health, and not disease prevalence.

The National Center for Health Statistics recognizes the importance
of including specific conditions on its survey checklist.  In the
NCHS publication, "Prevalence of Selected Chronic Conditions: United
States, 1990-92," a description of data collection methods states,
"Methodological studies have shown that chronic conditions are
generally underreported in interview surveys...Methodological studies
have also indicated that inclusion of a checklist of descriptive
condition titles as part of the interview questionnaire increases the
probability that a respondent will recognize the terms and report
those of which the respondent is aware" (19).

If Parkinson's disease were on the checklist, the NHIS would be an
important source of data on prevalence of Parkinson's disease, and we
would have additional data to draw on to answer our question.

Including Parkinson's disease in the NHIS would also increase public
awareness.  There are a number of statistical reference sources on
health, chronic illness and aging, both government and commercially
published, which lack any mention of Parkinson's disease.  They
often cite the NHIS as the source of information, and they fail to
mention Parkinson's disease.  In this manner the omission gets
propagated.


                XI.  Conclusions

If we wanted to find the prevalence of many major chronic conditions
in the United States, such as multiple sclerosis, epilepsy, heart
disease, headaches, asthma, diabetes, anemia etc. we could look in
reference books or U.S. government publications on health statistics
and find this data.  Yet we cannot currently answer this question
about Parkinson's disease with any degree of certainty.

There are estimates that at least one million people in the U.S. are
affected by Parkinson's disease, and this number will grow as the
population ages. Yet when we turn on the T.V. or radio, or read the
newspapers or publications on chronic disease in the U.S., we rarely
find mention of Parkinson's disease.  When we talk to the politicians
about the inequitable government funding for Parkinson's disease
research, we find they have very little awareness of the impact of
this disease.

These are but a few of the reasons why we need to obtain and publish
better information on the incidence and prevalence of Parkinson's
Disease, as well as on the other epidemiological aspects of the
disease.  Epidemiological studies could not only help researchers
determine the number of PWP, but could also help identify
environmental, geographical, occupational, socioeconomic, dietary,
and health-related factors that might lead to preventative measures.

We have seen that existing epidemiological studies apply to only a
few areas of the U.S.  Generalizing from the study data to the U.S.
as a whole is uncertain.  The prevalence estimates of 1 or more
million cases in the U.S. appear to be based on more than the data
from the epidemiological surveys. Compensation for cases missed in
the surveys, in particular, rough estimates of numbers of people in
undiagnosed and pre-symptomatic stages of the disease, appear to be
included. These estimates have been produced by processes which have
not been documented.  There has been no comprehensive effort to
survey the U.S. as a whole.  To do this accurately would require an
effort greater in magnitude than any of the previous studies.

We offer a few suggestions to remedy this situation.  There are
surely many others.

1.  Data gathered by the National Center for Health Statistics which
identifies people with Parkinson's disease by diagnosis needs to be
tallied and analyzed.

2.  Studies of the pharmaceutical companies' sales of Parkinson
drugs, and also tallies of pharmacy customers taking Parkinson drugs
could be useful as surrogate measures of Parkinson's disease
prevalence.  Perhaps this is an area of study that the pharmaceutical
companies could become involved in.

3.  Numbers of patients in files of insurers and large health care
organizations that contain diagnosis data could be tallied and
analyzed.

4.  The NCHS should include Parkinson's disease on it's checklist of
chronic conditions for the National Health Interview Survey. By
doing so, it is more likely that Parkinson's disease data will appear
in the NCHS's relevant survey reports, and subsequently in other
publications on health statistics.

Also, the Morris K. Udall Act for Parkinson's Research and Education
includes provisions for a nationwide Parkinson's disease data system
and a Parkinson's Disease Information Clearinghouse. The Udall Act
must be fully funded by Congress in order for these facilities to be
implemented. An information system currently exists for cancer data,
called SEER - the Surveillance, Epidemiology and End Results Program.
It is run by the National Cancer Institute and collects "data on a
routine basis from designated population-based cancer registries in
various areas of the country.  Trends in cancer incidence, mortality
and patient survival in the U.S., as well as many other studies, are
derived from this data bank" (31).  A system such as this one could
be created for Parkinson's Disease.


                      References

1.  Office of Scientific and Health Reports.  "Parkinson's
Disease: Hope Through Research," NIH Publication No. 94-139.
Bethesda, Md.: National Institute of Neurological Disorders
and Stroke. 1994.  Also Internet URL:
http://www.ninds.nih.gov/healinfo/disorder/parkinso/pdhtr.htm

2. Fact Sheet: Parkinson's Disease. San Francisco: Family
Caregiver Clearinghouse, 1996. Internet URL:
http://www.caregiver.org/factsheets/parkinsonsC.html

3. Rajput A.H. "Frequency and Cause of Parkinsons Disease."
Canadian Journal of Neurological Sciences. 19 (1992): 103-
107.

4.  Lanska, D.J. "Comparison of Utilization of Sinemet and
Parkinson's Disease Mortality as Surrogate Indicators of
Parkinson's Disease in the United States." Journal of the
Neurological Sciences 145 (January 1997):105-8.

5.  Strickland D., J.M. Bertoni, and R.F. Pfeiffer.
"Descriptive Epidemiology of Parkinson's Disease Through
Proxy Measures."  Canadian Journal of Neurological Sciences
23 (Nov. 1996):279-84.

6.  Svenson, L.W., G.H. Platt, and S.E. Woodhead.
"Geographic Variations in the Prevalence Rates of Parkinson's
Disease in Alberta."  Canadian Journal of Neurological
Sciences 20 (Nov 1993): 307-311.

7.  Roman, G.C., Z.X. Zhang, and J.H. Ellenbeg. "The
Neuroepidemiology of Parkinson's Disease." Etiology of
Parkinson's Disease. Eds. J.H. Ellenberg, W.C. Koller, and
J.W. Langston. New York: Marcel Dekker Inc, 1995.

8.  Rajput A.H., K.P. Offord, C.M. Beard, and L.T. Kurland.
"Epidemiology of Parkinsonism: Incidence, Classification and
Mortality."  Annals of Neurology 16 (Sept.1984): 278-282.

9.  Rajput A.H. "Frequency and Cause of Parkinson's
disease." Canadian Journal of Neurological Sciences 19 Suppl.
(Feb. 1992): 103-107.

10.  Tanner, C.M. and S.M. Goldman. "Epidemiology of
Parkinson's Disease."  Neurologic Clinics 14 (May 1996):
317-335.

11.  Zhang, Z.X. and G.C. Roman. "Worldwide Occurrence of
Parkinson's Disease: An Updated Review."  Neuroepidemiology
12 (1993): 195-208.

12.  Kessler, I.I. "Epidemiologic Studies of Parkinson's
Disease: A Community-based Survey." American Journal of
Epidemiology 96 (Oct. 1972): 242-254.

13.  Haerer, A.F., D.W. Anderson, and B.S. Schoenberg.
"Survey of Major Neurologic Disorders in a Biracial United
States Population: the Copiah County Study."  Southern
Medical Journal 80 (1987): 339-343.

14.  Morens, D.M., et al. "Epidemiologic Observations on
Parkinson's Disease: Incidence and Mortality in a Prospective
Study of Middle-aged Men." Neurology 46 ( Apr. 1996):
1044-1050.

15.  Mayeux, R., et al. "A Population-based Investigation of
Parkinson's Disease With and Without Dementia: Relationship
to Age and Gender." Archives of Neurology 49 (May 1992):
492-497.

16.  Svenson, Lawrence G., et. al.  "Geographic Variations in
the Prevalence Rates of Parkinsons Disease in Alberta."
Canadian Journal of Neurological Sciences. 1993:20:307-11.

17.  "Age Data" web page. Washington, D.C.: U.S. Bureau of the
Census.  Internet URL:
http://www.census.gov/population/www/socdemo/age.html.

18.  Schnappert, S. M. Office Visits to Neurologists: United
States, 1991-2. Washington D.C.: National Center for Health
Statistics, 1995. Internet URL:
http://www.cdc.gov/nchsswww/data/ad267.pdf.

19.  Collins, J.G.  Prevalence of Selected Chronic
Conditions: United States, 1990-1992. National Center for
Health Statistics. Vital and Health Statistics 10(194). 1997.
Internet URL:  http://www.cdc.gov/nchswww/data/sr10_194.pdf.

20.  Mitchell, S.L., et al. "The Epidemiology, Clinical
Characteristics, and Natural History of Older Nursing Home
Residents With a Diagnosis of Parkinson's Disease."  Journal
of the American Geriatric Society 44 (Apr. 1996): 394-399.

21.  Duncker A. And S. Greenberg.  A Profile of Older
Americans, 1997. Washington, D.C.: American Association of
Retired Persons, 1997.

22.  Strahan, G.  An Overview of Nursing Homes and Their
Current Residents: Data from the 1995 National Nursing Home
Survey.  Washington D.C.: National Center for Health
Statistics, 1997. Internet URL:
http://www.cdc.gov/nchsswww/data/ad280.pdf.

23.  Duvoisin, R. Parkinson's Disease: A Guide for Patient
and Family. New York: Raven Press, 2nd ed., 1984. 3rd ed.,
1992.

24.  Hutton, J.T. and R. L. Dippel. Caring for the Parkinson
Patient. Buffalo, N.Y.: Prometheus Books, 1989.

25.  Lieberman, A.N. and F.L. Williams. Parkinson's Disease:
The Complete Guide for Patients and Caregivers. New York:
Simon & Schuster, 1993, p. 21.

26.  Koller, W.C. "When Does Parkinson's Disease Begin?"
Neurology 42,suppl.4 (1992): 27-32.

27.  Koller, W.C, et al. "Does a Long Preclinical Period
Occur in Parkinson's Disease?"  Geriatrics 46,suppl. 1
(1991):8-16.

28.  Tasmar, the Parkinsons Disease Drug: Interview with Dr.
W.C.Koller, Dr. Arnold Sterman (Medical Director on TASMAR
for Hoffman-LaRoche), and Rosemary Paul, a patient of Dr.
Koller's. 1998.  Web page.  Thrive Online. Internet URL:
http://www.thriveonline.com/health/edge/tasmar2.html.

29.  Senter, R. J.  Analysis of Data.  Glenview, Ill.:
Scott, Foresman and Company, 1969.

30.  Adams, P.F. and M.A. Marano.  Current Estimates From the
National Health Interview Survey, 1994.  Washington D.C.:
National Center for Health Statistics, 1995. Internet URL:
http://www.cdc.gov/nchsswww/data/10_193-1.pdf through
www.cdc.gov/nchsswww/data/10_193-5.pdf.

31.  About SEER (1998).  Web page.  National Cancer
Institute. Internet URL:
http://www-seer.ims.nci.nih.gov/AboutSEER.html