As fax referral programs linking state quitlines to local health care providers are implemented in the U.S. nationwide, ask-advise-refer models supported by these programs have been promoted as a solution to enhance and simplify adoption of the 5As for improving tobacco treatment in primary care [3–7]. Findings from our study demonstrate that even such seemingly straightforward solutions for improving tobacco treatment must overcome significant organizational barriers in primary care practice to ensure full utilization and sustainability. Physicians and staff described numerous benefits of the chart stamp and Fax-to-Quit systems for providers and patients alike, but pointed out gaps in implementation that needed to be addressed to guarantee full integration into clinic and patient flow. The time-consuming process of referring patients to the Quitline combined with substantial patient resistance and limitations in information and care delivery systems raised questions about the long-term sustainability of the fax referral program in these clinics. Yet respondents identified several strategies for improving integration through tool simplification, enhanced teamwork and training, improved Quitline feedback, and patient education.
Adding new functions that are not well integrated into office routines can place a significant burden on overcommitted primary care practices. The relatively seamless integration of the chart stamp versus the fax referral form illustrated this difference. The chart stamp was included as part of the vital signs in the encounter form that was already used by MAs and physicians, and thus it did not require an additional step in the care process. While adding the stamp to the nurse and social worker forms was still needed to ensure full integration, respondents viewed the new chart stamp system as a helpful prompt that did not extend the visit time. In contrast, the fax referral form was cumbersome and the faxing process added a time-consuming new procedure to office routines. Staff suggested further simplification of the form to minimize respondent burden and modifications for patients with lower literacy levels to help in explaining the program to patients.
Web-based fax referral is an additional option that, if implemented efficiently, may increase referral rates. Sherman et al. found that transitioning from a multi-question consult referral that took one-minute to a simple web-based "two-click" referral, taking only seconds, in Veterans Health Administration clinics generated an overwhelming increase in smoker referrals for the California Smokers' Helpline . The NYS Fax-to-Quit program currently offers a web-based option for providers using an online template that is similar to the hard copy referral, but less than 2% of provider referrals currently come from this source (P. Bax, personal communication, June 18, 2009). As part of a larger academic institution, these community health clinics faced unique barriers to using the web to transfer patient information, including an electronic hospital firewall that prevented clinics from sending patient data to outside sources and privacy concerns related to the Health Insurance Portability and Accountability Act provisions.
Delivery system design
Variation in delivery of the fax referral among non-physician staff reflected lack of a consistent policy with clearly designated roles and formal accountability for tobacco referrals. Integrating a team-oriented approach for preventive care and treatment, particularly in resource-constrained settings, often requires substantive delivery system redesign as well as prioritizing among personnel at all levels [28–30]. The fax intervention may have been facilitated by enlisting visible internal support from top leadership at the ACRN to promote and assign responsibilities to specific staff for performing fax referrals. The chart system implementation was part of a broader high priority quality improvement initiative conducted across the entire ACRN and included formally redefining the MA role to conduct tobacco use screening. In contrast, the fax referral intervention engaged clinic-level leadership at the two study clinics to encourage staff to refer patients and training was provided, yet the Fax-to-Quit intervention was still viewed by staff and clinic leadership as another outside initiative among many, and official responsibilities for referral and faxing were not assigned. The lack of formal responsibility, limited priority, and slim staffing margins undermined efforts to spread referral tasks across non-physician staff. As a result, the responsibility for the Fax-to-Quit fell almost entirely on the physician. Similar to other studies [31, 32], doctors expressed challenges offering the time-consuming referral to every eligible smoker during brief clinical encounters.
In developing a teamwork approach, further efforts were needed to engage non-physician clinic staff at all levels through additional education on tobacco and the Fax-to-Quit program. While staff received a total of 90 minutes of training as part of the quality improvement initiative and the Fax-to-Quit intervention, employee turnover, patients' Quitline questions, and patients' personal issues that often arose during tobacco discussions highlighted the need for relatively comprehensive education with regular booster sessions [33, 34]. Such training should include basic motivational interviewing techniques, as well as instructions on brief counseling and referral techniques and explicit scripting of Quitline services to support the goal of the referral model, which is designed to cut down on the amount of time clinicians counsel patients by re-directing smokers to outside specialists [4, 35]. Trainings should also include front-line staff, who can be incorporated into the tobacco treatment team for such critical tasks as faxing, coordinating follow up, or re-filing progress reports. Research suggests that ongoing behavioral counseling training and improved knowledge of community resources among clinic staff may be particularly important for sustaining referral linkages between clinics and external organizations .
Clinical information systems
The lack of internal clinical information systems to provide timely information on referred smokers was a severe limitation to the sustainability of the referral program. Although Quitline progress reports provided individual-level information on referred smokers' status, the sites had no internal tracking system and limited resources to follow up with smokers the Quitline was unable to contact or who had set a quit date and received medications. The clinics needed in-house real-time information on the tobacco treatment status of smoking patients to adequately track referred patients. Since the time of this study, some ACRN clinics have transitioned to an electronic medical record (EMR) system. While EMRs can help practices track and monitor the delivery of tobacco services to patients , similar issues must still be addressed, including the development of efficient templates in the electronic record for documenting tobacco screening and intervention, policies delegating responsibility for data entry and follow-up, and integration into current clinical routines [24, 36, 37, 38]. Sites without EMRs can implement patient registries as a potential option for managing smoking patients . A simple paper-based registry to incorporate data from separate sources (e.g., patient charts, fax referral forms, Quitline progress reports, prescriptions) can be effective if designed with minimal response burden in mind and a staff member is identified for updating and maintenance .
Tighter integration between the clinic and the Fax-to-Quit program was needed to make use of the external information provided by the Quitline to clinics on their smoking patients. Most staff and physicians reported uncertainty on the specifics of the Fax-to-Quit program. While more intensive and continuous training may partially remedy this problem, further publicity efforts and outreach may be needed. The Quitline has worked with NYS-funded cessation centers, and increasingly, with provider organizations and health care plans to publicize the Fax-to-Quit program. In addition to these general efforts, targeted outreach to specific clinic networks and clinic staff at all levels could improve knowledge of this valuable program. Information from the Quitline could also be improved through the use of summary reports to individual clinicians on the progress of all smokers referred, which can reduce clinics' paperwork and faxing burden and would be particularly helpful for sites without internal systems to aggregate information on smoking patients.
Providers also reported seemingly low rates of Quitline contact of patients. Studies of fax referral programs in health care settings report quitline rates of reaching patients ranging from approximately 40-70% [9–16]. The reach rate in this study for patients referred from the two clinics during the intervention period was approximately 41% (NYS DOH, unpublished data, 2007), on the low end compared to these studies and slightly lower than the average NYS Quitline reach rate of 53% for fax referrals (NYS DOH, unpublished data, 2009). Our rates were similar to Mahabee-Gittens et al., who reported reach rates of 42% among a transient patient population fax referred through emergency departments . Further research and collaboration with providers to develop new strategies for contacting highly mobile populations may be needed to improve patient connections. To improve contact rates, experts also recommend clinical targeting of only those smokers who are ready to quit . Lastly, Quitlines must be adequately resources and sustainable,  utilizing funding from multiple sources if necessary,  to manage increases in demand following the introduction of fax referral systems and to implement outreach campaigns to providers, as discussed above.
According to providers, patient resistance and unfamiliarity with the Fax-to-Quit program was a significant barrier to referral in this study. This was in sharp contrast to recent reports from the national Prescription for Health study, which found that patients responded enthusiastically to new health behavior change resources . Our findings point to the need for further outreach in medical and other settings tailored to diverse patient populations. Direct-to-consumer marketing on the Fax-to-Quit program, available either in the clinic waiting room or through other outside venues, can enhance the appeal of and demand for quitline services, particularly among disadvantaged groups [41, 42]; such marketing will also benefit clinicians by reducing the amount of time needed to explain the program to patients. Further, increased consumer demand for quitline programs may have a synergistic effect: research indicates that providers are more likely to offer referrals if they believe the referral source is perceived as valuable by patients .
Data from this study come from clinicians and staff members of two health clinics serving a predominantly low-income, ethnic minority community in a large urban city in the U.S. To the extent that respondents and practices differ from the larger population of primary care practices and their employees in the U.S. and other countries in approaches to tobacco use management, our findings may have limited generalizability. The care components of the CCM model, however, can be applied across clinical settings. While specific care process changes needed for effective fax referral systems may depend somewhat on site leadership as well as system and population characteristics, consistent themes such as time, staff and resource needs for building and sustaining tobacco referral infrastructures [31, 33, 38] are likely to be relevant across primary care practices in developed countries' health care settings. Further, recommendations for improvements to quitline interactions with health care providers may apply to fax referral programs in other states or locales. Lastly, the nature of interviews and focus groups provides information only on aspects of clinical systems of which participants were aware; clinic observations or patient interviews might reveal additional characteristics of clinic systems and patient interactions that may impact systems implementation.