Multivariate evaluation Table 2 revealed lower odds of receiving advice among persons who were younger, men, African American, uninsured, healthier on the SF-12 subscales, lower users of health care services, lighter smokers, or who lacked a usual care location or used an emergency room or other location. The odds of receiving advice were increased for persons with military health insurance and those who received their care in hospital clinics.
Discussion
Less than 50% of the surveyed patients reported that they had been counseled to quit smoking by a physician during the 12-month period. This finding is similar to earlier published reports,2,11-13 indicating that little progress has been made toward achieving national health goals for 20009,20 and performance standards for health plans.21 Since by 1996 and 1997, the years the CTS was administered, ample evidence supported use of brief motivational counseling,6-9 it is unfortunate that rates of physicians’ cessation counseling advice had not improved substantially compared with earlier reports. Further, the low rate of physician cessation advice we observed likely indicates that more intensive cessation services were even less frequently offered. For example, the benefits of pharmacotherapy had been well documented by the time the CTS was administered,22 but it seems unlikely that medications for cessation would have been recommended to respondents who reported they did not receive cessation advice.
We observed significant associations between lower rates of physicians’ cessation advice and several patient factors. First, lower rates of physician counseling were reported by persons with the lowest current burden of smoking-related illness, including those who were younger, had better perceived health status, and had lower smoking intensity. Although greater attention to persons with end-organ damage is understandable, cessation advice for primary prevention, such as to younger persons, might yield the greatest long-term benefit. Second, men were less likely to report receiving cessation advice, even after adjusting for their lower utilization of health care services. This finding, consistent with observations indicating that men are much less likely than women to receive preventive care,23 suggests that approaches to helping men become more effective consumers of health care need to be developed.24 Third, respondents who lacked a usual care location or used an emergency room or other location were less likely to report they received cessation advice, supporting approaches to health care delivery that promote continuity of care. Fourth, persons who lacked health insurance were less likely to report they received cessation advice independent of other factors. Given that such persons frequently face myriad obstacles to accessing health care, including financial barriers to pharmacotherapy for cessation, efforts aimed at increasing the provision of cessation counseling services to these persons are warranted. Finally, the lower rates of physician counseling among African Americans, independent of other factors, demands attention, especially because African Americans have been targeted by cigarette manufacturers, advertisers, and merchants.24-28 Efforts to increase cessation counseling provided to African Americans who smoke might help reduce disparities in health status.
Limitations
Our study is subject to several limitations. It is difficult to assess the reliability and validity of patient reports, and it is possible that recall of physicians’ advice to quit could deteriorate as the interval between administration of the survey and the clinical encounter lengthened. However, past studies have supported the validity of patient report of physicians’ cessation advice.12,17 Also, although we were not able to account for the time interval since the last visit, we were able to adjust for respondents’ total number of clinical encounters in 12 months. Given that respondents had roughly 4 visits on average, we believe the magnitude of recall bias would be low.
Although it is possible that the observed relationships could be the result of confounding by unmeasured or incompletely measured factors, the richness of our data source allowed us to control for a wide range of potential confounders. The inferences that can be drawn from these results are further strengthened by the survey design and the recency of the data.
Many physician-level and system-level factors could not be evaluated, so future exploration of contextual factors is warranted. What specific features of hospital clinics, health maintenance organizations, or military coverage account for higher rates of cessation advice? Might physicians who practice in such settings have greater access to systems that help identify smokers, thereby improving cessation interventions?
Conclusions
Although the 1996 AHCPR smoking cessation guideline9 gives counseling its highest recommendation, our findings suggest that physicians are performing poorly on this measure, with little improvement since the early 1990s. Many physicians might benefit from being trained in the use of brief counseling techniques that have been found effective.9,10 System changes designed to promote cessation should be implemented in clinical settings. For example, physicians are more likely to provide cessation advice when patients’ smoking status is routinely identified in the medical record.29,30,31 Funds from tobacco settlements could provide the funding for system-level interventions but reportedly are infrequently being applied toward smoking cessation efforts.32