Hello,

The Spring 1999 issue of NPF's Parkinson Report, pp. 16-17 contains
an article which is relevant to the present discussion.  The author,
Dr. Abraham Lieberman, indicates the need for a survey of all 50 U.S.
states to determine a reasonably accurate count of the number of
PWPs.  To illustrate the potential benefits of such a count, Dr.
Lieberman used pharmaceutical sales figures for Sinemet to construct
a graph showing the percent of each state's population on Sinemet.
The graph shows a significantly greater prevalence of PD in the farm
states N. Dakota, S. Dakota, Nebraska, Minnesota and Iowa.

In the article Dr. Lieberman does not discuss the accuracy of the
sales figures as representing numbers of PWPs, much less as a
measure of the number of people taking Sinemet.  The assumptions
regarding an average Sinemet dose per PWP were not mentioned.

Dr. Lieberman indicates that the pharmaceutical sales figures need
further confirmation by the survey he proposes.  He does not
now commit to any particular estimate of total PWPs in the U.S., such
as 1 million or 1.5 million, because surveys which have been done to
date covered limited areas that may not represent the whole.  If
anyone has any better data or a greater basis of confidence, I would
be very interested to learn about it.

I was curious whether there was any correlation between the Sinemet
sales numbers and certain other statistics.  I took the chart from
the article and loaded it into a spread sheet, estimating as best I
could an additional decimal place for bars in the graph which ended
between the scale lines.  I next loaded in the 1990 U. S. census
population by state and computed an absolute number of people on
Sinemet by state.  Based on this, the total people on Sinemet in 1990
was 847,242. Multiplying this total by the ratio of a current
estimate of U.S. population to the 1990 census count (270,000,000 :
248,709,873), gives a current total of 919,768 people on Sinemet.  I
do not know the margin of error, because, as I said above, I do not
know how Sinemet sales numbers translate into numbers of PWPs taking
Sinemet.  I can send all the detail on request to anyone interested,
but I want to next try to obtain further info on the Sinemet sales
data.

Whenever an the estimate of the number of PWPs in the U.S. is cited
outside a strict epidemiological context, the assumptions behind
that estimate -- the definition of PD, whether estimates include
people who are pre-symptomatic, people who are symptomatic and not
diagnosed, people misdiagnosed as PWPs, as well as the statistical
margin of error -- are usually omitted.  Then the estimate is
repeated by people who are unaware of all the omitted assumptions,
and a very fuzzy number thereby gets taken to be a solid figure.

Phil Tompkins
Hoboken NJ
age 61/dx 1990