This paper uses a new calendar design implemented in the Guatemalan Survey
of Family Health to analyze diarrheal and respiratory illness among childre
n. The calendar provides a much richer description of child illness and tre
atment behavior than do conventional data typically collected in health int
erview surveys. The resulting estimates reveal that Guatemalan children exp
erience high rates of diarrheal and respiratory illness and that these illn
esses often involve multiple symptoms that only partially overlap one anoth
er. In addition, estimates from the calendar demonstrate that the measureme
nt of illness frequency is fairly complex and that classification of illnes
s into distinct categories may not always be feasible or meaningful. Result
s regarding treatment behavior indicate that the vast majority of illnesses
are treated and that mothers almost always receive advice regarding their
children's illnesses from relatives. On the other hand, health providers ar
e sought for advice and treatment in only about one-third of illnesses. Whe
n families do seek providers, they are more likely to rely on biomedical on
es - especially pharmacists, doctors and personnel at health posts and cent
ers - than traditional practitioners. By far, the most common form of treat
ment - recommended by both relatives and providers - consists of readily av
ailable Western medicines. In contrast, the use of traditional remedies app
ears to be relatively low. (C) 1999 Elsevier Science Ltd. All rights reserv
ed.