Background. Community-oriented primary care (COPC) requires the develo
pment of practical tools if it is to be carried out. A previous study
demonstrated a practical approach to carrying out one portion of a com
munity assessment for COPC using a few representative health indicator
s. Objective. To determine the validity of this process elsewhere, we
tested whether these findings were generalizable to other settings and
to the same setting a decade later. Method. In a cross-sectional stud
y design, data on 18 health indicators were collected for census tract
s in two target areas and for the entire state of Ohio, USA, for 1990.
A factor analysis was performed to identify factors underlying the he
alth indicators in the three areas examined. Results. Two underlying f
actors, termed age and poverty, were present in all locations and over
time. Each factor was defined by core indicators and a cluster of ass
ociated indicators. Conclusions. These results suggest that one part o
f a COPC community assessment can be done by selecting very few indica
tors. The distribution of indicators of age and income explains the va
riability of most of the health related indicators studied. These fact
ors are stable over time and location. A community assessment should i
nclude indicators which, at a minimum, provide information on these tw
o factors.