Hartmann, Katherine E. MD, PhD1,2,3; Wechter, Mary E. MD, MPH1,2,3; Payne, Patricia CNM, MPH4; Salisbury, Kathryn1,6; Jackson, Renee D. MHA5; Melvin, Cathy L. PhD, MPH6
Smoking in pregnancy is linked to fetal and neonatal morbidity, including fetal demise, preterm birth, low birth weight, preterm premature rupture of membranes, placental abruption, and placenta previa. Antenatal smoking is also associated with neonatal and pediatric morbidity, such as sudden infant death syndrome, respiratory infections and asthma, otitis media, and poor school performance.1–11 Although knowledge of such risks is widespread among women and care providers, approximately 12% of women delivering in the United States report having smoked cigarettes during pregnancy. Costs of smoking-attributable pregnancy complications exceed $300 million each year in the United States.12–15
Smoking cessation intervention by prenatal care providers can decrease the number of pregnant women who smoke by 30–70%, and the intensity of the intervention directly affects the probability of cessation.14,16–19 Core recommendations for the conduct of smoking cessation intervention have been stable for more than 20 years. Resulting evidence-based clinical practice guidelines for smoking cessation intervention endorse five key components, termed the “5 A's”: 1) Ask about and document tobacco use status of all patients; 2) Advise all tobacco users to discontinue use at each care encounter; 3) Assess willingness to quit; 4) Assist with a cessation plan; and 5) Arrange follow-up to monitor progress and provide support.17 Although these guidelines have been proven effective and are time- and cost-efficient, the prevalence of prenatal smoking in North Carolina remains high (approximately 13.8%), suggesting a need for enhanced intervention by prenatal providers.20,21
Together, the Woman and Tobacco Coalition for Health, the North Carolina Division of Public Health–Women's Health Branch, the Center for Women's Health Research at University at North Carolina, and the American College of Obstetricians and Gynecologists, joined with the North Carolina Department of Health and Human Services to document the practice patterns and resource needs of prenatal care providers in North Carolina in regard to promoting smoking cessation. We believe such data can help identify high-priority targets to facilitate use of best practice prenatal smoking cessation interventions across the United States. Specifically, the objectives of this study were to measure the use of best practice intervention, including each of the “5 A's,” and to assess the relationship between best practice and current intervention resources, prior training in smoking cessation intervention, and barriers to providing intervention.
MATERIALS AND METHODS
Public access state licensure files and primary care association membership rolls were used to compile an exhaustive list of all 1,687 health professionals eligible to provide prenatal care in North Carolina. In all, 1,138 providers, including physicians (family medicine and obstetrics), midwives, nurse practitioners, and physician assistants, practiced in the state and had a complete address. We mailed a questionnaire to these providers in 2003–2004 requesting information on smoking cessation interventions and practice patterns. The three-page, 18-item survey required approximately 10 minutes to complete and return by mail and was sent at least three times to each eligible provider who did not reply to prior requests. Telephone calls, facsimiles, and use of regional provider networks supplemented our efforts to achieve a maximal response rate. Incorrect addresses were investigated through consultant databases, telephone directories, Web searches, and updates to licensure and professional society membership records. Questionnaires were excluded from analysis if the provider denied current provision of prenatal care or the questionnaire was incomplete for key questions. No incentive was offered for completion of the survey. A cover letter emphasized the importance of accurate reporting for the purpose of identifying resource needs.
Survey data were collected under the auspices of the North Carolina Department of Health and Human Services with approval granted through the University of North Carolina biomedical institutional review board. The questionnaire was adapted for this project from two prior surveys on physicians' smoking interventions and was designed to evaluate the individual components of providers' smoking cessation interventions.22,23 Providers were asked to respond to questions about their practice type and to estimate the proportion of smoking patients and that of Medicaid and uninsured patients in the practice. Providers were then asked to rate the frequency, on a Likert scale from 1=never to 5=always, with which they provided each of the “5 A's.” Use of all “5 A's” “usually” or “always” defined best practice intervention, the main outcome of interest in this survey.
We assessed factors thought to be potential predictors of best practice implementation. We requested information on practice tools for smoking cessation intervention. These tools were designated as material or counseling resources. Material resources included written questionnaires, chart flags or prompts, education and self-help materials, “prescriptions to quit,” pledge forms for a smoke-free environment, smoking diaries, tracking forms to monitor counseling, follow-up contact by calls or mailings, a method for documenting smoking interventions in the prenatal chart, and “other.” Counseling resources included on-site or referral options for providing additional individual or group counseling, staff familiar with providing smoking cessation intervention, telephone hot-line support, referral to Web-based smoking cessation programs, and “other” resources. Providers were asked about the frequency with which they prescribed nicotine replacement or bupropion. They were also asked if their practice had a written protocol to identify a staff member(s) responsible for implementing and documenting cessation interventions. Providers reported on their previous training in smoking cessation intervention, barriers to providing intervention, and areas in which they would like more training or resources. Finally, the influence of reimbursement on providers' willingness to provide smoking cessation intervention, from “none” to “very substantial,” was assessed using a Likert scale.
Returned surveys were coded and double-entered by staff unaware of the respondent's provider type. Missing values were identified and patterns in the missing observations were sought. Likert variables were recoded as dichotomous variables based on an inclusive cut point of at least “sometimes” for prescribing nicotine or bupropion, and at least “somewhat substantial [influence]” to define a positive role of reimbursement in willingness to provide best practice. Variables for both material and counseling resources were recoded into count variables, and their relationship to best practice was evaluated. Based on a significant χ2 test, followed by pairwise comparisons, these variables were further divided into categories at meaningful cut points. Dichotomous variables were also created to represent none or at least one for both material and counseling resources.
We performed descriptive statistics and univariable analysis. Using a χ2 test and unadjusted logistic regression, we assessed bivariable associations for the main outcome of best practice intervention and each independent variable in the survey. We also examined the relationship between provider type and each of the independent variables. We then generated a logistic regression model to reflect the relationship between the performance of best practice intervention and potential covariates in this survey. These included provider type, practice resources, principal barriers, previous training in smoking cessation, the influence of reimbursement on willingness to provide intervention, and the use of a written protocol to identify staff responsibilities for smoking cessation intervention. Best fit for logistic regression was determined by using the likelihood ratio test to compare nested models, maintaining variables that maximized the probability of the model to represent the data. We assessed the trend in the odds of best practice according to quantity of resources using disjoint indicators in a logistic regression model, comparing each category of resources with the reference category of no resources. Statistical analysis was performed with Stata 9.1 (StataCorp, College Station, TX).
Of the 1,138 eligible providers to whom questionnaires were mailed, 844 (74%) returned surveys. Of these, 232 did not currently provide prenatal care, and 63 questionnaires were incomplete for key items that define best practice. Thus, our study population consisted of 549 participants (Fig. 1). Fifty percent were obstetricians; 18%, midwives; 15%, family medicine physicians; 13%, nurse practitioners; and 4%, physician assistants. The most common practice setting was groups of two to 10 providers (62%), primarily in single-specialty groups. In aggregate, providers reported that 42% of their patients received Medicaid, and 10% had no form of payment for services. Providers estimated that 15% of their patients were smokers at the time of delivery.
Provider groups were comparable in their provision of each of the “5 A's” and overall use of best practice intervention (P≥.3). Providers almost universally reported that they “usually” or “always” ask their pregnant patients about smoking (98%) and advise patients to quit (100%). Seventy-four percent reported that they regularly assess willingness to quit; 62% consistently assist patients with making a plan to quit; and 37% arrange support for cessation efforts after an initial counseling session. Overall, 31% of providers use best practice intervention.
Nearly one half of providers (48%) reported having had no formal training in smoking cessation intervention. Only 20% of respondents reported a written protocol to document staff responsibilities involved in smoking cessation intervention and follow-up, with no significant difference among provider types (P=.1). More than 90% reported at least one material resource, with no difference by provider type (P=.9): 75% reported a method for documenting screening and smoking cessation counseling in the prenatal chart, 56% provided self-help materials, 41% used chart flags to trigger counseling, 22% wrote prescriptions to quit, and 19% reported using written questionnaires.
Provider groups differed significantly in terms of prescribing nicotine replacement and bupropion (P<.001). Obstetricians and family medicine physicians reported prescribing nicotine replacement (55% and 53%) significantly more often than nurse practitioners (28%) and physician assistants (17%). Obstetricians also prescribed nicotine replacement significantly more often than midwives (42%). Obstetricians and midwives were significantly more likely to prescribe bupropion (51% and 46%) than family medicine physicians (27%) and nurse practitioners (28%), and obstetricians only (51%) were significantly more likely to prescribe bupropion than physician assistants (35%).
More than 70% of respondents reported at least one option for providing smoking cessation counseling, with no difference among provider type (P=.6). The most commonly reported resources were clinic staff familiar with smoking cessation (32%), on-site individual counseling (31%), and referral for individual (26%) or group counseling (25%). Only 7% and 9%, respectively, reported referring pregnant patients to a hot line or quit line or a Web-based program for smoking cessation support.
The most commonly acknowledged barriers to best practice intervention were similar across provider types and included time constraints (71% of providers) and lack of patient interest (68% of providers). Time and patient interest were also most commonly chosen as the most important barriers to best practice. Over one third of providers (39%) acknowledged limited effectiveness of smoking intervention as a barrier to intervention. Nine percent listed lack of confidence in personal intervention skills as a barrier. Providers without formal training in smoking cessation intervention were more than three times as likely to report this barrier as those with training (P<.001). Increasing number of perceived barriers, not surprisingly, was associated with a significant decrease in the odds of best practice. Over half of providers reported that reimbursement was at least “somewhat” influential on their willingness to provide smoking cessation, with no difference between provider types (P=.9).
In a full model adjusted for provider type, training, at least one counseling resource, a written protocol, the influence of reimbursement, and limitations of time and patient interest, providers reporting at least one material resource (compared with none) were nearly 10 times as likely to provide best practice (odds ratio [OR] 9.6, 95% confidence interval [CI] 1.3–72.9). Having at least one counseling resource was associated with more than twice the adjusted odds of best practice, compared with having no counseling resources (OR 2.5, 95% CI 1.4–4.4). An increasing number of material resources, compared with no resources, was associated with an upward trend in the odds of best practice (P for trend <.001). Compared with having no counseling resources, providers with one and those with two or more counseling resources had approximately double and quadruple the odds of best practice intervention (P for trend <.001) (Table 1).
Certain practice patterns were also associated with best practice. Adjusted for all other variables in the model, having a written protocol to identify staff responsibilities was associated with more than twice the odds of best practice (OR 2.5, 95% CI 1.5–4.3). The practice of prescribing nicotine replacement or bupropion at least “sometimes,” although not included in the regression model, was associated with approximately twice the unadjusted odds of best practice (OR 2.0, 95% CI 1.4–3.0, and OR 1.8, 95% CI 1.2–2.6).
Neither of the two most commonly reported barriers, time and patient interest, was significantly associated with best practice. Attitudes about reimbursement were associated with best practice, however. Providers who reported that reimbursement was at least “somewhat” influential on their willingness to provide smoking cessation intervention were 40% less likely than their counterparts to provide best practice (OR 0.6, 95% CI 0.4–1.0).
After reducing the full model, three factors remained significantly, positively associated with provision of best practice. These factors are the possession of 1) at least one material resource, 2) at least one counseling resource, and 3) a written protocol to identify responsibilities. Reporting reimbursement to be at least “somewhat” influential remained negatively associated with the odds of best practice.
Smoking in pregnant women has well-documented, serious consequences with significant cost to mother, infant, the health care system, and society. Cessation during pregnancy can limit these adverse effects, and evidence shows physician intervention promotes smoking cessation. Consistent, coordinated use of the “5A's” of best-practice intervention is cost-effective and readily incorporated into practice as an essential part of comprehensive care for pregnant smokers.
This study describes a robust respondent pool of all prenatal providers in a state with comparatively high maternal smoking prevalence. One goal of this study was to identify aspects of smoking intervention services to target for improvement, particularly those likely to be susceptible to changes in practice patterns and provision of additional resources from state and federal initiatives. Results indicate that most providers accurately estimate the magnitude of the prenatal smoking problem. Yet, only one third of providers use all “5 A's” of best practice.
This study identifies three practice and organizational factors linked to providing best practice, namely having at least one material resource, at least one counseling resource, and a written protocol to guide implementation. The importance of these factors may measure the value of coordination and office supports in accomplishing the “5 A's.” Alternatively, rather than exerting causal influence, having these resources in place may simply reflect commitment to cessation intervention. Our results can only hypothesize interventions to be further investigated. Such interventions might include availability of basic resources such as patient education materials and counseling and office tools to track intervention and coordinate follow-up.
Toward providing these resources, proactive hot lines and Web-based interactive programs are increasingly available. States and stakeholder organizations, in concert, can drive awareness of these resources for wider use by prenatal providers. An example of an accessible referral source is the smoking cessation counseling hot line for women in North Carolina. Coordinated through the national network of tobacco quit lines at 1-800-QUIT-NOW, North Carolina's new pregnancy-specific feature, available since November 2005, is described in its associated Web site (www.QuitlineNC.com). For states without pregnancy-specific counseling, 1-800-QUIT-NOW leads women to the American Legacy Foundation's Great Start Quit Line designed specifically for pregnant women.24 (Additional material resources for smoking cessation intervention are available online at: www.helppregnantsmokersquit.org, sponsored by The National Partnership to Help Pregnant Smokers Quit.)
Beyond resource availability, the most commonly perceived barriers to intervention reflect the pressures of practice. At the core are a lack of time, insufficient staff support, and the absence of mandates that would make smoking cessation intervention a practice priority. Considering cessation intervention a quality-of-care index, tracked and expected by third-party payers, might promote change. The relevance of appropriate reimbursement merits further investigation, particularly given recent literature demonstrating improved quitting and cessation rates in the setting of enhanced Medicaid reimbursement.25
This survey, with a strong response rate (74%), provides a detailed description of practice patterns in a largely suburban–rural state. Results, however, may not be generalizable to states with different demographics. Characteristics of nonresponders were not available, and these providers may differ substantially from those represented. Like all self-reported data, the responses likely reflect the desire to provide socially desirable answers and, as such, may overestimate the level of performance and resources available. For that reason, we are confident that we can learn from the substantial component of respondents willing to report less than best practice patterns and few resources. Falling short of best practice occurs primarily by failure to assist with a plan to quit or to arrange follow-up, both activities that can depend on the availability of material and counseling resources. This survey suggests simple, practical steps in care settings that may hold promise for further reducing prenatal smoking.
1. Ananth CV, Savitz DA, Luther ER. Maternal cigarette smoking as a risk factor for placental abruption, placenta previa, and uterine bleeding in pregnancy. Am J Epidemiol 1996;144:881–9.
2. Armstrong BG, McDonald AD, Sloan M. Cigarette, alcohol, and coffee consumption and spontaneous abortion. Am J Public Health 1992;82:85–7.
3. DiFranza JR, Aligne CA, Weitzman M. Prenatal and postnatal environmental tobacco smoke exposure and children's health. Pediatrics 2004;113:1007–15.
4. Ekwo EE, Gosselink CA, Woolson R, Moawad A. Risks for premature rupture of amniotic membranes. Int J Epidemiol 1993;22:495–503.
5. Eskenazi B, Prehn AW, Christianson RE. Passive and active maternal smoking as measured by serum cotinine: the effect on birthweight. Am J Public Health 1995;85:395–8.
6. Harger JH, Hsing AW, Tuomala RE, Gibbs RS, Mead PB, Eschenbach DA, et al. Risk factors for preterm premature rupture of fetal membranes: a multicenter case–control study. Am J Obstet Gynecol 1990;163:130–7.
7. Raymond EG, Mills JL. Placental abruption. Maternal risk factors and associated fetal conditions. Acta Obstet Gynecol Scand 1993;72:633–9.
8. Sexton M, Hebel JR. A clinical trial of change in maternal smoking and its effect on birth weight. JAMA 1984;251:911–5.
9. Williams MA, Mittendorf R, Stubblefield PG, Lieberman E, Schoenbaum SC, Monson RR. Cigarettes, coffee, and preterm premature rupture of the membranes. Am J Epidemiol 1992;135:895–903.
10. Malloy MH, Kleinman JC, Land GH, Schramm WF. The association of maternal smoking with age and cause of infant death. Am J Epidemiol 1988;128:46–55.
11. Pollack HA. Sudden infant death syndrome, maternal smoking during pregnancy, and the cost-effectiveness of smoking cessation intervention. Am J Public Health 2001;91:432–6.
12. Ventura SJ, Hamilton BE, Mathews TJ, Chandra A. Trends and variations in smoking during pregnancy and low birth weight: evidence from the birth certificate, 1990-2000. Pediatrics 2003;111:1176–80.
13. Centers for Disease Control and Prevention (CDC). Annual smoking-attributable mortality, years of potential life lost, and economic costs—United States, 1995–1999. MMWR Morb Mortal Wkly Rep 2002;51:300–3.
14. Melvin C, Gaffney C. Treating nicotine use and dependence of pregnant and parenting smokers: an update. Nicotine Tob Res 2004;6 suppl:S107–24.
15. Adams EK, Melvin CL. Costs of maternal conditions attributable to smoking during pregnancy. Am J Prev Med 1998;15:212–9.
16. A clinical practice guideline for treating tobacco use and dependence: a US Public Health Service report. The Tobacco Use and Dependence Clinical Practice Guideline Panel, Staff, and Consortium Representatives. JAMA 2000;283:3244–54.
17. Fiore MC, Bailey WC, Cohen SJ, Dorfman SF, Goldstein MG, Gritz ER, et al. Treating tobacco use and dependence. Clinical practice guideline. Rockville (MD): U.S. Department of Health and Human Services; 2000. Available at: www.surgeongeneral.gov/tobacco/treating_tobacco_use.pdf
. Retrieved July 10, 2007.
18. Dolan-Mullen P, Ramirez G, Groff JY. A meta-analysis of randomized trials of prenatal smoking cessation interventions. Am J Obstet Gynecol 1994;171:1328–34.
19. Lumley J, Oliver SS, Chamberlain C, Oakley L. Interventions for promoting smoking cessation during pregnancy. Cochrane Database Syst Rev 2004;4: CD001055.
20. Avery M, Stallings W. Tobacco use among pregnant women in North Carolina: Predictors of smoking cessation during pregnancy. Results from the North Carolina Pregnancy Risk Assessment Monitoring System (PRAMS), 1997–2001. A State Center for Health Statistics Report. Raleigh (NC): N.C. Department of Health and Human Services; 2003.
21. Marks JS, Koplan JP, Hogue CJ, Dalmat ME. A cost-benefit/cost-effectiveness analysis of smoking cessation for pregnant women. Am J Prev Med 1990;6:282–9.
22. Hartmann KE, Espy A, McPheeters M, Kinsinger LS. Physicians taught as residents to conduct smoking cessation intervention: a follow-up study. Prev Med 2004;39:344–50.
23. Melvin CL, Tucker P. Measurement and definition for smoking cessation intervention research: the smoke-free families experience. Smoke-Free Families Common Evaluation Measures for Pregnancy and Smoking Cessation Projects Working Group. Tob Control 2000;9 suppl:III87–90.
24. Haviland L, Thornton AH, Carothers S, Hund L, Allen JA, Kastens B, et al. Giving infants a great start: launching a national smoking cessation program for pregnant women. Nicotine Tob Res 2004;6 suppl:S181–8.
25. Petersen R, Garrett JM, Melvin CL, Hartmann KE. Medicaid reimbursement for prenatal smoking intervention influences quitting and cessation. Tob Control 2006;15:30–4.