Tobacco use is a leading preventable cause of death and extensive health-related economic losses (1). Tobacco use cessation efforts are highlighted by the 2008 US Public Health Service smoking cessation clinical practice guideline (2). Effective treatments exist, and providers influence smoking cessation rates when they encourage their patients to quit smoking (2), which is recommended for quality care delivery (3). Yet, a large gap is found between the existence of guidelines and recommendations and their implementation in clinical practice (4,5).

The clinical practice guideline emphasizes the need for systems-level approaches that can increase provider intervention and reduce smoking (2,6). It calls on administrators, insurers, and purchasers to provide leadership in implementing systems interventions that include provider training, resources, and feedback. It also recommends insurance coverage and physician reimbursement for tobacco dependence treatment (2). Such systems interventions are necessary for preventive medicine and chronic disease management (7).

Health plans, which offer specific products or packages of defined benefits to their purchasers (8,9), may be influenced to adopt systems-level strategies. Purchaser demand often drives what is covered by those products, and purchasers can ensure specific activities by requiring them in the contract, although they usually do not demand smoking cessation services (10,11). Additionally, screening activities can be considered a basic indicator of the quality of the health plan, as indicated by inclusion of measures of medical assistance with smoking cessation — advising smokers to quit, discussing medications, and discussing strategies — in the National Committee for Quality Assurance's Healthcare Effectiveness Data and Information Set (HEDIS) accreditation measures (12,13). The HEDIS measures may be an incentive to health plans to offer screening, cessation advice, and referral.

Health plans, in turn, may influence the behaviors of providers and enrollees. About 60% of Americans have employer-based health insurance, largely through managed care organizations (14,15). Twenty percent of privately insured adults are current smokers (16). Thus, private health plans have an important role for adoption and implementation of clinical practice guidelines. By implementing systems-level changes, health plans can influence providers to help their patients quit smoking. In addition, health plans have been shown to independently influence HEDIS quality measures, beyond actions of providers, for patients across multiple health maintenance organizations and provider groups (17).

We used a nationally representative survey of health plans to examine systems-level strategies for smoking cessation, considering screening and guideline distribution practices in particular, across a full product range offered by private health plans (health maintenance organizations [HMOs], point-of-service [POS] products, and preferred provider organizations [PPOs]). Prior studies have examined provider activities or patient knowledge or receipt of services, excluded some types of health plans, or were not nationally representative (4,6,7,10,18-22). We investigated how health plans are implementing systems-level approaches to smoking cessation.

Health plans were split fairly evenly among product types; slightly fewer identified their products as HMOs (29%) than POS products (36%) or PPOs (35%). Nearly three-fourths of products were in health plans that contracted with a managed behavioral health organization for their alcohol, drug, and mental health services. Most (86%) plans were for-profit.

Systems-level support for smoking cessation is not widespread among private health plans. Despite current clinical guidelines and recommendations, most plans do not require providers to screen for smoking, determine whether screening is done, or distribute relevant guidelines to providers.

Some studies based on enrollee reports have found high rates of identifying smokers (10,13,21), but our results based on health plans are consistent with a 2001 survey of physician organizations, which found few physicians are required to provide smoking cessation activities, and very few receive financial incentives from health plans (25). America's Health Insurance Plans reports that two-thirds of health plans have written guidelines for smoking cessation (10), but we found that few actually distribute them. The positive news from our study is that health plans that make an effort to screen for smoking report multiple smoking cessation activities, an indication of a multipronged approach to improve smoking cessation rates. Such a systemic approach is likely to improve provider behavior.

When techniques were used to encourage adherence to guidelines, we found that plans most often compared results among providers, but financial incentives were rarely used; this was similar to other findings (10). In other research, both provider feedback and financial incentives were related to significantly more smoking cessation activities reported by patients, and greater knowledge of and assistance in these activities by providers (25-27). Health plans that increase use of these techniques may improve their enrollees' smoking rates. However, even when health plans use some of these techniques, providers may be unaware of them, indicating room to improve adoption and implementation (28). Health plans could also step beyond the provider to offer, for example, incentives directly to enrollees, online smoking cessation programs, or programs within a broader wellness approach.

Why are health plans not doing more in this arena, when it has been shown that health plan strategies can independently affect activities such as screening (17)? Lack of provider compliance is cited as a barrier to increased activities by health plans (10). We found that slightly more health plans verify screening than require screening, perhaps in response to this perceived barrier of poor compliance. Moreover, a health plan requirement may be considered a burden for providers (14) or not realistically enforceable. Health plans also report specific systems-level barriers to addressing smoking cessation, including competing priorities, issues with data and reporting systems, and lack of resources, staffing, or funding (10,29). Health plans may assume that responsibility lies elsewhere, such as with providers or the public health system, despite the call for systems-level approaches that include insurers (2). Purchasers largely do not demand smoking cessation interventions as part of their health care packages (10,11), yet purchaser demand can drive health plan activities. Health plans may doubt the business case for smoking cessation activities: 61% of plans report delayed return on investment as a barrier (10). Yet, the literature suggests that tobacco use cessation activities are cost-effective for both health plans and employers and provide a reasonable return on investment (2,30,31).

Despite overall low rates of required screening for smoking and distribution of written guidelines, HMOs were significantly more likely than other products to report each activity. The original HMO mission often highlighted prevention as well as treatment (32,33); thus, smoking cessation activities may fit better there than in other types of health plans. It is also perhaps easier to conduct these activities within the constraints of the more managed provider network used by most HMOs. That is, HMO providers might be more aware of the health plan rules and guidelines compared with providers in a loosely networked PPO who participate in multiple health plans. HMOs may also be more likely to emphasize smoking cessation activities because they are measured as part of HEDIS (12) and thus may be viewed as valuable to accreditation; HEDIS did not apply to PPOs at the time of the survey. In 2003, HEDIS showed that 69% of smokers who saw a physician were advised to quit, yet only 36% discussed specific strategies with their physician (13). Although not directly comparable to our findings of health plan activities, because they are ascertained via enrollee surveys and do not identify the systemic approaches that health plans may use, HEDIS measures may improve with a systems focus on smoking cessation by health plans.

We note several potential limitations. First, this survey focused on alcohol, drug, and mental health services and may have been completed by someone more familiar with these services and less familiar with general medical services, such as a representative of a separately contracted managed behavioral health organization. If smoking cessation is in the purview of general medical services (ie, those provided by PCPs) then some respondents may not have been aware of those activities, as indicated by missing data, and we have potentially underestimated the prevalence of cessation activities. However, even if all missing data had been a positive response, the prevalence of smoking cessation activities by health plans would still be low. Second, this is a survey of health plans, not providers, so it is unknown how successfully providers in these plans are screening their patients and how this varies by health plan requirements. Third, although these data from 2003 are still the most recently available about health plans, changes over time may have occurred. Fourth, screening is only the first step in a range of recommendations that lead to smoking cessation.

Even if all health plans required screening, verified that their providers did it, and distributed treatment guidelines, health plans should encourage and monitor smoking cessation activities and follow-up (29), moving beyond calls for improved delivery of services by providers (34,35). Such activities could include requiring that providers not only ask their patients whether they smoke but also advise them to quit and then provide assistance with quitting through either medication or quitline services. Health plans should pay for smoking cessation services, which is an incentive to providers to offer them. Billing codes specific to tobacco cessation counseling were added to the Healthcare Common Procedure Coding System in 2005 and Current Procedural Terminology in 2008. The addition of the codes to these standard billing code systems used by insurers and providers allows providers to be paid for smoking cessation services and gives plans a way to indicate their coverage of these services. However, it is unclear how many health plans accept these codes for reimbursement. Unless the codes are adopted across plans, providers, who largely accept multiple types of insurance and may not know the insurer for any given patient (34), may be unlikely to use the codes and to offer smoking cessation services for which they are not reimbursed.

These data from the 2003 benefit year, the most recently available national data on private health plans, are still relevant to consider how systems-level interventions can affect providers' ability to change behavior in their patients. Our findings of limited systems-level efforts by health plans to promote smoking cessation activities by their providers suggest several conclusions. First, systems approaches to smoking cessation still need to be enhanced and adopted (2,18,36). Second, purchasers and participants need to demand systems-level smoking cessation activities (35). Third, research should investigate effective incentives and techniques that health plans could use to change behavior among providers and enrollees. Fourth, a variety of smoking cessation strategies, as highlighted by Orleans (35), should continue to be encouraged. Future research should consider how health plan activities interact with activities of other systems, such as the public health system, and of providers themselves. It should also address health plans' concern with return on investment as a barrier to these activities (34). In the meantime, by encouraging health plans to focus on smoking cessation at the systems level, further inroads can be made to reduce the burden of smoking on enrollees, employers, and health plans themselves.