Patient Tobacco Use, Quit Attempts, and Perceptions of Healthcare Provider Practices in a Safety-Net Healthcare System

INTRODUCTION

Tobacco use continues to lead the nation as a preventable cause of morbidity and mortality.¹ Despite reductions over the past 3 decades in smoking among the nation's general population,² rates of tobacco use remain high among low-income, less-educated, minority, and under- and uninsured groups.³ Louisiana's smoking prevalence (22%) is higher than the national average (17%).⁴ Nationwide, smoking rates vary by insurance coverage: 16% of those covered by private insurance smoke, compared to 30% of public insurance enrollees (Medicaid and Medicare) and 32% of uninsured residents.⁵ In Louisiana, a large proportion (25%) of residents is uninsured; 34% of these smoke.⁶

Healthcare providers and delivery systems can impact population-level cessation rates through implementation of the US Public Health Service (USPHS) clinical practice guideline (CPG) Treating Tobacco Use and Dependence: 2008 Update that includes the 5A protocol: (1) ask about tobacco use, (2) advise all identified smokers to quit, (3) assess smokers' willingness to quit, (4) assist smokers in their quit attempt, and (5) arrange for follow-up contact. Furthermore, the CPG delineates standards for quality care, endorses the effectiveness of evidence-based treatments for tobacco use, and provides strategies for integrating screening and treatment into routine patterns of care. Nearly 70% of smokers visit a physician at least once a year, providing an opportunity for intervention.⁷ However, CPG implementation in primary care settings is less than optimal.⁸

Patients' perceptions of their care have become increasingly important to health systems as many seek to improve the quality and satisfaction with treatments and to provide patient-centered care.^9,10 When tobacco users receive treatment according to the CPG, they report higher satisfaction with overall healthcare received relative to untreated tobacco users.⁷ Provider treatment of tobacco use can be measured by patient surveys (eg, Consumer Assessment of Healthcare Providers and Systems, National Health Interview Survey), provider surveys, medical record reviews, and direct observation. However, limitations exist for each of these and results vary. While direct observation is the standard for assessing provider treatment, healthcare systems can benefit from precise, cost-effective, and practical approaches to obtaining patient perceptions of provider intervention.^11,12 Patient surveys are more accurate than chart audits for assessing chronic disease advice, information dissemination, and, in some instances, general health promotion.¹³

In 2002, to accompany an increase in the excise tax on cigarettes, the Louisiana State University School of Public Health (LSUSPH), in partnership with Louisiana's safety-net healthcare system, created the Tobacco Control Initiative (TCI). The TCI, described in detail elsewhere,¹⁴ employed a systems approach to facilitate implementation of the CPG in the LSU network of public hospitals. This report presents results from an assessment of patient tobacco use, quit attempts, and perceptions of provider treatment before (2004) and after (2010) the CPG implementation in Louisiana's safety-net healthcare system.

RESULTS

DISCUSSION

The survey findings suggest that integrating the USPHS CPG in a large public hospital system impacts patient tobacco use, quit attempts and methods for quitting, and perceptions of provider treatment behavior. In 2010, fewer respondents reported ever smoking and heavy smoking. While the decrease among ever smokers may be attributed to statewide prevention programs and media campaigns focused on averting smoking initiation, the decrease in those smoking 11 or more cigarettes per day may result from successful quit attempts either on their own or from utilizing cessation services in the LSU Health System. Since 2004, TCI has made free group behavioral counseling, self-help material, access to low-cost pharmacotherapy, and, more recently, access to free quit-line telephone counseling available to aid patient quit attempts. Compared to 2004, more respondents in 2010 reported smoking ≤10 cigarettes per day, daily smoking, and having smoked in the previous month. This trend suggests an emergence of differing levels of smokers and their cessation pathways: smokers who are able to quit and maintain abstinence, smokers who quit and relapse, and recalcitrant smokers.¹⁵ Relapsing and recalcitrant smokers are likely recycled in the health system¹⁶ and may need tailored services to improve quit attempts and abstinence rates.¹⁷ However, the smoking rate of survey participants in both 2004 and 2010 was higher than the rates reported for patients in other primary care settings^18,19 and for patients who are uninsured.²⁰

Patient respondents reported an increase in all of the provider treatment measures, except advice to quit, which remained unchanged. There was a slight increase in provider screening for tobacco use and significant increases in providers who assessed patient willingness to quit, assisted patients with their quit attempts, and arranged follow-up for patients after the clinic visit.

These increases may indicate that TCI's effort to galvanize the clinical and patient-level cessation interventions by the LSU Health System were effective, and this systems-based approach may serve as a model for future statewide intervention efforts to decrease patient tobacco use through the systematic implementation and evaluation of the USPHS CPG for treating tobacco use by healthcare providers. Identification and documentation of tobacco users were obtained from nursing assessment forms, followed by TCI referral forms, and finally from electronic medical records with prompts and reminders for intervention. Systemwide adoption of a tobacco treatment policy, training to improve provider intervention skills, and provider feedback on clinical performance via an electronic dashboard occurred throughout the system. These multilevel cessation interventions, including counseling and medication, have been successful in improving adherence to the USPHS CPG for treatment of tobacco use in primary care settings.²¹ Although few studies have reported patient perceptions of provider adherence to the 5A approach (ask, advise, assess, assist, arrange) in primary care settings, overall our results were more favorable than those found by others in similar settings^11,18,22,23 and in studies of the uninsured.⁸

Respondents reported an increase in quit attempts, and significantly more reported awareness of assistance to help them quit. These increases suggest that TCI's efforts to ensure that all providers were trained to identify and discuss best-practice treatment options with their patients were effective. In addition, TCI employed dedicated tobacco specialists to coordinate cessation services and to support clinicians at each facility. The present results regarding quit attempts were less favorable than results reported in primary care settings¹⁹ but higher than in studies among the uninsured.⁸

Between 2004 and 2010, no change was seen in respondents who reported making a quit attempt on their own (cold turkey), and fewer respondents reported making a quit attempt by using a nicotine patch or Zyban. This finding suggests efforts to increase quit attempts using evidence-based strategies were not equally effective. However, the decrease in use of a patch or Zyban may also be explained by increased access to and availability of varenicline (Chantix) starting in 2006,²⁴ an option that was not on the 2004 survey.

Study limitations should be noted. One limitation is that patient responses were based on their previous clinic visit (ie, surveys were conducted before their current clinic visit). This delayed measurement and reliance on recall may overestimate performance.¹² A second limitation is that the results were based on a self-report method. Further, several commonly used methods to assess provider cessation counseling have limitations. Patient surveys may under-²³ or overestimate,^25,26 provider surveys may overestimate,^23,27 paper medical charts may underestimate,^25,26,28,29 and electronic medical records may underestimate all items on the 5A protocol, except for asking about tobacco use.¹¹ Direct observation, the ideal method, is burdensome and costly.²⁹ Healthcare systems can benefit from precise, cost-effective, and practical approaches to assessing treatment for tobacco use, and this type of onsite survey distribution proved to be an effective strategy for data collection in Louisiana's public hospital system.

CONCLUSION

This study—which examined patient tobacco use, quit attempts and methods, and perceptions of physician behavior to treat tobacco use among primary care patients in Louisiana's safety-net healthcare system—found positive changes between 2004 and 2010. During this time frame, practice guidelines for the treatment of tobacco use and dependence were implemented. Obviously, these results do not indicate cause and effect because this study was a nonrandomized and noncontrolled observation. Other extraneous factors may have promoted improvements in many of the items measured; however, some differences are quite large and statistically significant. Therefore, these results overall indicate a positive trend in patient behaviors and perceptions of provider care after implementation of the 5A clinical protocol. These observations have implications for eliminating tobacco use at the population level, especially among ethnic/racial minorities, those of low socioeconomic status, and the under- and uninsured. Future studies should examine the effects of implementation experimentally in a randomized-controlled trial and compare patient responses to clinician (electronic) documentation. Furthermore, tailored interventions that target relapsed and recalcitrant smokers and the use of evidence-based strategies during quit attempts are warranted. Decreasing tobacco use is a top objective of Healthy People 2020.³⁰ Improving the quality of treatment through a systems-based approach may impact the disproportionate use of tobacco in especially disparate populations.