Objectives-To determine rates of patient reported and physician documented
counselling; to identify predictors of each report; and to identify the imp
act of each report on smoking cessation attempts after discharge from the h
Design-Stickers on subjects' charts prompted physicians to give brief smoki
ng cessation counselling to patients in the hospital. Patients reported cou
nselling received and quit attempts in a phone interview conducted 7-18 day
s after discharge. Rates of counselling and correlations were calculated, a
nd multivariate analysis identified predictors of patient report, physician
documentation, and quit attempts.
Setting-Four hospitals in the Minneapolis/St Paul metropolitan area.
Subjects-682 hospital patients who had smoked more than 100 cigarettes in t
heir lifetime and had smoked in the last three months.
Results-71.0% of patients reported counselling, and physicians documented c
ounselling in the charts of 46.2% of patients (correlation = 0.15, kappa =
0.13). Patient report was predicted by specific hospital, belief that their
hospitalisation was smoking related, diagnosis of a smoking related diseas
e, and physician documentation of counselling. Physician documentation was
predicted by female patient, specific hospital, longer hospital stay, and m
arginally predicted by smoking related disease. Quit attempts were predicte
d by patient report of counselling, but not physician documentation.
Conclusions-Physicians document counselling in the hospital at a lower rate
than patients report it, and the correlation between reports is very low,
making an accurate assessment of true rates of counselling difficult. While
it is important to increase physician documentation, it is even more impor
tant to increase patient recall, as this is the only report that predicts a