Cigarette smoking accounts for 440,000 deaths in the U.S. each year and is the most preventable cause of morbidity and mortality (1). Quitting smoking at any age provides important health benefits and greater life expectancy (2). In terms of lives saved, quality of life, and cost-efficacy, treating smoking is considered one of the most important activities a clinician can do (3).
While the prevalence of cigarette smoking among U.S. adults has declined since the first Surgeon General’s report on smoking and health in 1964, rates remain elevated among psychiatric populations. The smoking prevalence among individuals with a current psychiatric illness is nearly double that of individuals without mental illness (41% versus 23%) and even higher among the seriously mentally ill and those with substance use disorders (4, 5). In California, the smoking rate for adults is among the lowest in the nation (17%), yet recent studies in northern California estimate the smoking prevalence at 28% among psychiatric outpatients and 45% among psychiatric inpatients (6—8). Psychiatric patients also tend to be heavy smokers, in the number of cigarettes smoked per day and in how deeply they inhale (9). It is estimated that 44% of the cigarettes sold in the U.S. are to the mentally ill (4).
Cigarette smokers with psychiatric and substance use disorders are at high risk for smoking-related deaths (10, 11). Rates of cardiovascular and respiratory diseases and cancer are higher than those of age-matched controls (12, 13), with smoking believed to be a major contributing cause. Complicating treatment, the hydrocarbons of tar in cigarettes cause increased metabolism of some antipsychotic (e.g., clozapine, haloperidol, olanzapine) and antidepressant (e.g., nortriptyline) medications, which may lead to inadequate dosing, subtherapeutic blood levels, increased cost, and possibly even some neuroleptic side effects (e.g., tardive dyskinesia) (14). Financially, the smoking burden may be particularly difficult for individuals with severe mental illness, who are likely to be on a low income (15). Socially, in areas where smoking is prohibited, heavy smokers may find it difficult to participate, leading to further isolation.
The U.S. Public Health Service recommends all patients be screened for tobacco use, advised to quit, and offered intervention (16). Evidence indicates the more intensive the cessation counseling, the greater its effectiveness. The guidelines also recommend use of pharmacotherapy with all smokers trying to quit, expect in special circumstances, with nicotine replacement therapy (NRT) and bupropion SR being first-line and nortriptyline and clonidine being second-line medications. An organizational framework for tobacco cessation intervention is the National Cancer Institute’s 5-A intervention for physicians to ask about tobacco use, advise smokers to quit, assess readiness to quit, assist in quit attempts, and arrange follow up (17, 18). Given the complicated relationship between mental illness and smoking, integration of cessation efforts within psychiatric care is encouraged (19—21). The American Psychiatric Association’s (APA) clinical guidelines for treating nicotine dependence recommend psychiatrists "assess the smoking status of all their patients…discuss interest in quitting…[and provide] explicit advice to motivate the patient to stop smoking" (19). A recent metaanalysis of randomized trials examining physician advice for smoking cessation indicated a significant treatment effect with an increase in the odds of patients quitting (22). Of 39 trials identified, however, not one was conducted in a psychiatric setting.
A study analyzing data from the 1992—1996 National Ambulatory Medical Care Survey may be the only published report on psychiatrists’ counseling for smoking cessation (23). Data were collected from independent practice settings. Of most concern, 23% of psychiatric visits had to be dropped from the analysis because patient smoking status was unknown. For patients identified as smokers (N=1610), psychiatrists reported offering cessation counseling at only 12% of visits. Diagnosis of Nicotine Dependence was not made at any visit, and NRT was never prescribed.
A primary barrier to delivering smoking cessation counseling may be lack of training. A number of tobacco cessation curriculums have been developed and a recent systematic review of the literature suggested training health professionals to provide tobacco cessation interventions had a measurable effect on professional performance including offering counseling, setting quit dates and follow up visits, distributing self-help materials, and recommending NRT (24). Of the 10 trials identified, however, none was conducted with mental health professionals. Another factor may be low prioritization of tobacco use as an issue of relevance for psychiatric practice. The mental health, addictions, and tobacco control communities have largely ignored tobacco dependence among smokers with psychiatric disorders (25). Other potential barriers to counseling include beliefs that patients would not be interested in or able to quit, that patients need to smoke to manage their psychiatric symptoms, and that cessation attempts may exacerbate patients’ symptoms and/or threaten recovery from other substances of abuse (26).
The extent to which psychiatry residency programs prepare their residents for identifying and treating nicotine dependence is unknown. The purpose of this study was to assess the need for and interest in tobacco cessation curricula in psychiatry residency training. We surveyed psychiatry residents on their knowledge, attitudes, and behaviors regarding interventions for treating tobacco dependence in clinical practice.
Our study was conducted with residents from five psychiatry residency programs in northern California. Residency lists provided by program training directors defined the recruitment pool. The survey was mailed and/or emailed to the 155 identified residents. A cover letter explained the purpose of the survey and requested voluntary participation. Survey completion was considered consent to participate. This study was approved by the appropriate Institutional Review Boards.
A three-page self-report survey assessed participants’: 1) knowledge of smoking rates, health effects, and treatments (8 items); 2) attitudes regarding clinical interventions for treating nicotine dependence (10 items); 3) engagement in smoking cessation counseling practices with their patients (i.e., the 5-A’s) (6 items); 4) perceived confidence in their ability to counsel patients to quit smoking (6 items) with a rating of overall ability (1 item); 5) adequacy of prior training on smoking cessation (5 items); 6) interest in further training to help patients quit smoking (1 item); and 7) respondent characteristics (e.g., gender, level of training, advanced degrees, smoking status). Knowledge scores were calculated as the percent correct. Mean total scores were calculated for attitudes, behaviors, and confidence. Many of the items were used previously to evaluate the Rx for Change curriculum with more than 3,000 students of medicine, pharmacy, and nursing. The scales demonstrated good internal consistency previously (27) and with the current sample (Cronbach alpha’s ranged from 0.79 to 0.84). The full measure is available upon request from the authors.
Participants were instructed to complete the survey individually, which was anticipated to take less than 15 minutes. Respondents were asked to provide their name and address on a detachable coupon, for $5.00 reimbursement, and to return it by mail separately from the completed survey, thereby maintaining anonymity. Self-addressed, stamped return envelopes were provided. Nonresponders were sent a second survey.
Descriptive analyses (means, frequencies) were used to summarize residents’ survey responses. Correlations tested associations among the constructs.
Surveys were completed by 105 residents (68% response rate). The sample was 60% female; 19% held graduate degrees (M.P.H., Ph.D.) in addition to their medical training. Respondents were in their first (23%), second (22%), third (29%), or fourth (26%) year of residency training. Identified areas of specialized interest were psychotherapy (68%), biological psychiatry (61%), child psychiatry (23%), other (15%), and the addictions (13%). Respondents’ smoking status was 11% current, 17% former, and 72% never. There were no differences in smoking status by gender, year of training, advanced degree status, or area of specialty (all p’s > 0.05). The sample was representative of the residency recruitment pool with respect to gender (χ2= 0.38, df=1, p=0.535), year of training (χ2= 0.45, df=3, p=0.929), and residency site χ2= 4.45, df=4, p=0.349).
Knowledge, Attitudes, and Behaviors
Respondents averaged 54% correct (SD=17; range: 12.5% to 100%) on the knowledge items. Endorsement of negative attitudes toward smoking cessation counseling in psychiatric practice averaged 2.1 (SD=0.6) on a 5-point scale ranging from 1 "strongly disagree" to 5 "strongly agree." Responses by item are summarized in t1. F1 shows respondents’ engagement in the 5-A’s intervention for smoking cessation. The percent reporting often or always engaging in the 5-A’s was: 58% ask, 29% advise, 17% assess, 18% assist, and 13% arrange follow up. Knowledge scores and engagement in the 5-A’s were significantly correlated (r = 0.21, p=0.034); neither was associated with the attitudes scale (p’s> 0.10). Males (r = 0.27, p=0.005) and current tobacco users (r = 0.29, p=0.002) endorsed more negative attitudes toward counseling patients to quit smoking. Knowledge scores and reported behaviors did not differ significantly by respondent characteristics (p’s > 0.10).
Confidence and Overall Ability
Confidence ratings for tobacco cessation counseling averaged 3 (SD=0.6) on a 5-point scale from 1 "not at all" to 5 "extremely" confident; 76% rated their overall ability to help patients quit using tobacco as fair or poor. Higher confidence ratings were associated with greater perceived overall ability (r = 0.58, p<0.001), knowledge (r = 0.27, p=0.006), and engagement in the 5-A’s (r = 0.38, p<0.01). Greater perceived overall ability was significantly associated with engagement in the 5-A’s (r = 0.42, p<0.001), but not knowledge scores (r = 0.18, p=0.073).
Most residents reported receiving none or inadequate tobacco cessation training in medical school (74%) or residency training (79%), as continuing medical education (97%) or on-the-job training (91%). Adequate training was associated with greater knowledge (r = 0.30, p=0.002), confidence (r = 0.50, p<0.001), perceived ability (r = 0.37, p<0.001), and engagement in the 5-A’s (r = 0.22, p=0.024). However, among those who rated previous training as adequate, ratings of confidence (mean = 3.3) and overall ability (mean = 2.4) were still relatively low (5-point scales), and nearly all residents (94%) reported moderate to high interest in learning more about helping their patients quit smoking. Males (r = —0.24, p=0.014), current tobacco users (r = —0.24, p=0.012), and residents with more negative attitudes toward counseling patients to quit smoking (r = —0.37, p<0.001) reported less interest in further training on smoking cessation.
Early (N=75) and late (N=30) responders were compared, where late responders were defined as returning their survey after the second mailing. Late responders may serve as proxies for nonrespondents providing a means for assessing survey representativeness (28). Comparisons indicated no differences between early and late responders on any of the measured variables (all p’s > 0.05).
This study may be the first to examine the extent to which psychiatry residents are prepared to intervene on nicotine dependence with their patients. The findings reveal low levels of knowledge, confidence, perceived ability, and smoking cessation interventions in clinical practice. Further, engagement in the 5-A’s is lower than that reported by clinicians in other medical specialties (29, 30).
Lack of training appears to be a major factor. The majority of respondents reported receiving none or inadequate training on tobacco-related interventions in medical school or residency training. While reports of adequate training were associated with greater knowledge, confidence, perceived ability, and engagement in the 5-A’s for smoking cessation, levels of confidence and perceived ability were still relatively low, and nearly all (94%) residents reported moderate to high interest in further training for helping their patients quit smoking. The findings demonstrate the need for and interest in tobacco cessation curricula in psychiatry residency training programs.
Negative attitudes toward counseling patients to quit smoking were infrequently endorsed overall, suggesting few perceived barriers to implementing cessation interventions in clinical practice. Negative attitudes, however, were more salient among males and residents who reported current tobacco use, and were associated with less interest in further training. For a subset of residents, addressing negative attitudes may be important for increasing receptiveness to training. While smoking rates are lower than found in the general population, significantly higher rates of tobacco use have been reported among psychiatry residents and psychiatrists in practice relative to other medical specialties (31, 32). Perhaps related, psychiatrists are less likely to treat tobacco dependence compared to other health care providers (31, 33). Training in tobacco cessation interventions may yield benefits both through encouraging cessation among psychiatrists as well as improving clinical practice.
Representativeness of the sample is unknown. Residents were recruited from five programs in northern California, a state with strong antitobacco policies. Nevertheless, residents’ knowledge, confidence, and behaviors were low. The response rate of 68% is good for a physician survey, and the sample was found to be representative of the recruitment pool. In the literature, response rates for mailed physician surveys average 54% to 61% (34, 35). Factors associated with higher response rates include use of shorter surveys, multiple mailings, stamped return envelopes, and monetary incentives, all elements used in this study (28). Comparison of early and late responders in the current sample revealed no significant group differences, allaying concerns with response bias. Further, nonresponse bias is believed to be less of an issue with physician surveys given the greater homogeneity in knowledge, training, attitudes, and behaviors relative to the general population (28).
Smokers with psychiatric disorders have been identified as a priority population (16, 19, 25, 36), and psychiatric treatment encounters provide an ideal but, as yet, untapped opportunity for treating this deadly addiction. Integration of smoking cessation services within psychiatric care is recommended given the specialized needs of mentally ill smokers (19—21). The high rates of tobacco use among the mentally ill and the resulting negative health, social, financial, and treatment consequences cannot be ignored. Without clinical intervention, however, levels of tobacco use are unlikely to change. A focus on training the next generation of psychiatrists may help ensure that changes in clinical practice are achieved and that tobacco interventions are delivered to this high risk group of smokers.
This study was supported by the National Institute on Drug Abuse (NIDA) San Francisco Treatment Research Center (grant #P-50 DA-09253); National Institute on Drug Abuse (NIDA) grants #R01 DA-02538, #R01 DA15732, and #T32 DA-07250; the National Institutes of Mental Health (NIMH) grant #R25 MH-060482; and a Postdoctoral Fellowship from the Tobacco-Related Disease Research Program (#11FT-0013).
The authors thank Marc Jacobs, M.D., Alan Louie, M.D., C. Barr Taylor, M.D., Craig Campbell, M.D., and David Goldberg, M.D. for allowing work with the residents in their programs.
The authors also thank Lindsay Fletcher for assistance with data management and Karen Hudmon, Dr.PH, MS, R.Ph and Robin Corelli, Pharm.D. for sharing their survey from the Rx for Change curriculum.