Special Chat Virapongse, MD Barbara S. Martineau
Research
#{149} Scott
Emerson, MD, PhD V. Staab, MD
#{149} Edward
Resources
Research resources in academic radiology were investigated by analyzing the responses to a survey from 72 North American institutions. The questionnaire addressed five general areas: department size, departmental resources committed to research, availability of research training, research quality control, and research productivity. The highest correlates of grant productivity included measures of departmental resources committed to research, for example, space devoted to research, size of research budget, and full-time employee support for engineers, physicists, and chemists. In a regression model, measures of the number of engineers employed by a department, the number of attending staff, and the number of training lectures given by engineers were found to be most highly associated with dollar value of grant support. The average level of research resources available at responding institutions was generally low, despite a seemingly strong desire to do quality research. This is evidenced by a strong sentiment among respondents in favor of research training and quality control of research. Index terms: Radiology partmental management ologists,
research
Radiology
#{149} Special
1990;
I From K.C.P.L.,
deand
radi-
reports
and
of Radiology Biostatistics
(CV., (SE.),
of Florida College of Medicine, From the 1989 RSNA annual September 5. 1989; accepted
vember
7. Address
Watson
Clinic,
Lakeland
radiologists.
175:247-251
the Departments B.S.M.. E.V.S.)
University Gainesville. ing. Received
and
#{149} Radiology
reprint
requests
Radiology
Hills
Blvd.
to CV.,
Department. P0
Box
meetNo1600
95000,
Lakeland,
RSNA, See this
also issue.
I
the
editorial
by
Brody
(pp
37-38)
in
Radiology’
N academic
radiology, research is a major departmental objective and a necessity for the promotion of its staff. However, under the current cost-cutting climate, research mesources may be dwindling. How badby this reduction has affected adademic radiology remains speculative; however, questions are already being asked as to how research is currently being conducted in academic radiology (1,2). What are the key resource requirements needed to enhance proper research? Should research training and quality control be performed at the institutional level? In an attempt to answer these questions we sent out a survey to academic madiobogy chairmen throughout North America.
MATERIALS
AND
METHODS
A questionnaire was mailed to chairmen of 135 academic radiology departments identified by the Association of University The survey
Radiologists addressed
(AUR) listing. five important
an-
eas related to research: (a) departmental characteristics-name, location, size of physical space, number of examinations performed yearly, number of staff radiologists,
number
of fellows; search-funds mitted
of residents,
and
(b) resources (excluding
to research,
number
committed grants)
research
to recorn-
funds
as per-
centage located devoted support
of total budget, physical space alto research, equipment time to research purposes, research staff not funded by grants, seed
money
available
for
pilot
projects,
aver-
age research time provided for faculty (as well as ranking factors determining research time allocation, such as research productivity, seniority, and administrative duties); (c) research training-existence of departmental research training program, opinion regarding the need for such a program for residents and faculty, funds committed to research training, hours of lectures on research techniques, of
lectures
research
1990
C. P. Li, MD
#{149}
in Academic
types
FL 33804-5000.
King
lows, search
electives
and
on
research
for
the attendance
training
courses
Report
manship workshop given by the AUR; (d) research quality control-opinion as to the need for quality control, who performs it, and the number of hours devoted to it; and (e) research productivitydollar amount of gross grant support obtamed per year, total number of principal investigators,
and
the
number
of clinical
radiologists serving as principal investigators. For the majority of questions, the respondent was asked to characterize his department according to categories (eg, number of residents between zero and 10, 1 1 and
20,
etc),
although
explicit
values
were solicited for some questions (eg, percentage of FTE support for physicists). Chairmen were not asked for separate mesponses for nuclear medicine and radiation therapy. To quantify bias due to nonresponse to the survey, all articles published in Radiology and AIR for the 5 years 1984-1988 were classified according to the institutional affiliation of the first author. These journals were chosen due to their covenage of a wide spectrum of the radiologic subspecialties. For each of the 135 institutions
identified
resenting published
by
Statistical
AUR,
number journals
a count
rep-
of articles was obtained.
Analysis
The mined
frequency for each
for
respondents
all
the
the total in these
of response of the survey combined,
was deterquestions as well
as
within categories defined by the amount of grant support received by the departments. The Pearson correlation coefficient was computed to determine the conrelation between each of the variables and the variable measuring dollar amount of grant support. To assess the most important factors associated with the amount of grants received by a radiology department, a regression model was explored that modeled the level of grant support as a function of the variables related to departmental size, research resources, research training, and quality
techniques,
residents
and
of faculty such
as the
fel-
at regrants-
Abbreviations: versity
Radiologists.
AUR FTE
Association
of Uni-
full-time
employee.
247
Table 1 Number of Publications per Institution Classified by Response the Survey
to
Response
No Response
0-9 10-19 20-39 40-99 100
24 15 16 12 4
25 13 14 8 4
71*
64
*
One
the name
respondent
(of the 72) failed
of the institution
and
of 69 Institutions
Relative
to the Number
No.
No.of Publications
Total
Table 2 Matrix Distribution and Grant Acquisition Amount
*
Staff
Staff
of Grant
Funding
to provide
of Faculty
of Faculty
(X
60
>3,000 1,000-3,000 500-1,000 250-500 100-250
0 0 0 2 1
0 2 0 2 8
0 2 3 4 2
1 1 2 1 1
0 1 2 0 0
2 0 1 0 0
0 0 0 1 1
.74). Thus, there do not appear to be significant
248
Radiology
#{149}
Institutional
Research
Funds
100-200
(X $1,000)
Committed
200-300
Research
300-400
400-500
>500
>3,000
1
0
0
0
0
2
2
2
1
1
0
0
500-1,000 250-500 100-250
1 3 5
1 1 6
2
1
2
2
1
1
1
0
0
1 2 1
1
1
0
0
0
Three
27
institutions
failed
to respond
to questions
concerning
grant
funding.
4
Matrix Distribution Grant Acquisition
of
66
Institutions
Relative
to Available
Research
Amount of Grant Funding(X$l,000)*
RESULTS
due