Effectiveness of a Pregnancy Smoking Intervention

Abstract

Despite the known dangers of pregnancy smoking, rates remain high, especially in the rural, Southern United States. Interventions are effective, but few have been developed and tested in regions with high rates of pregnancy smoking, a culture that normalizes smoking, and a hard-to-reach prenatal population. The goals were to describe a smoking cessation intervention, the Tennessee Intervention for Pregnant Smokers program, and examine the impact on quit rates compared to usual care. Additionally we sought to examine reduction in smoking levels and number of quit attempts related to the intervention and finally to examine the impact of the intervention on birth outcomes. Intervention and historical control group participants, all smokers at entry to prenatal care, were recruited from five medical practices providing prenatal care in rural, South-Central Appalachia. The intervention, an expanded 5A’s (Ask, Advise, Assess, Assist, Arrange) model, was delivered by trained health educators. Over 28% of intervention group women quit smoking, compared to 9.8% in the control group. Two thirds of intervention group women significantly reduced smoking by delivery, with 40%+ attempting to quit at least once. Compared to controls, intervention group women saw significantly better birth outcomes, including newborns weighing 270g more and 50% less likely to have a neonatal intensive care unit admission. Among intervention group participants, those who quit smoking had significantly better birth outcomes than those who did not quit smoking. Findings point to the potential for appropriately tailored pregnancy smoking interventions to produce substantial improvements in birth outcomes within populations with health disparities.

Keywords

pregnancy smoking prenatal care rural health smoking cessation intervention

The negative impact of smoking during pregnancy is well recognized. Prenatal smoke exposure is linked to many deleterious pregnancy and birth outcomes and is a significant factor in later health and developmental problems including decreased growth, asthma and other respiratory conditions, and cognitive and behavioral problems, including deficits in executive function and conduct disorders (Bastra, Hadders-Algra, & Neeleman, 2003; Espy, Fang, Johnson, Stopp, & Wiebe, 2011; Fingerhut, Kleinman, & Kendrick, 1990; Rogers, 2008; U.S. Department of Health and Human Services [USDHHS], 2001). In the short term, women who smoke during pregnancy are at substantially elevated risk for low birth weight (LBW) babies (USDHHS, 2001). Nationally, smoking accounts for 20% of LBW births (USDHHS, 2001), and infants born to women who smoke are on average 200 g lighter than those born to nonsmokers (Chiolero, Bovet, & Paccaud, 2005). In fact, smoking is recognized as the most modifiable risk factor for poor birth outcomes (Cnattingius, 2004).

Nationally, 11% of pregnant women in the United States smoke (March of Dimes [MOD], 2014). However, in some regions of the country, the rates are substantially higher. In the “tobacco country” of rural Central and Southern Appalachia, rates can be triple the national averages. In recent studies and surveillance reports, pregnancy smoking rates in this region ranged from 25% to 49% (Bailey & Jones Cole, 2009b; Mullett, Britton, John, & Hamilton, 2010; Tong, Jones, Dietz, D’Angelo, & Bombard, 2009), with rates highest among women living below the poverty line in the most rural counties (Bailey & Jones Cole, 2009b). Consequently, rural South-Central Appalachia also has high rates of poor birth outcomes. Approximately 14% of babies born in the South are preterm, and 9% are LBW, rates 10% higher than national averages (MOD, 2014). The picture in the most rural and economically depressed areas of the South is particularly striking. In those Southern states with the highest percentage of residents living in rural areas of Appalachia (West Virginia, Tennessee, Kentucky, Alabama, and Mississippi), rates of LBW and preterm birth are 20% to 45% higher than national averages (MOD, 2014).

Many interventions to address pregnancy smoking are available. A comprehensive updated systematic review of pregnancy smoking cessation interventions was recently published (Chamberlain et al., 2013). Covering 86 trials involving 29,000+ women, the report revealed that the majority of interventions significantly reduced smoking rates (by approximately 6% compared to usual care according to a previous version of the review: Lumley, Oliver, Chamberlain, & Oakley, 2004). Studies that also included birth outcomes provided evidence for effects, including an overall 18% reduction in LBW. Higher intensity interventions (multiple components lasting 15+ minutes/session) were more effective than less involved interventions; however, after a certain point increased duration and frequency did not increase effectiveness. Components of the most effective interventions were focused on behavior change and included counseling, feedback, education, and incentives. The use of health educators was also noted as effective (Lumley et al., 2009). The authors concluded that interventions need to be appropriate for the targeted population, especially when the focus is on women in poverty with little support (Lumley et al., 2004). They also noted that infant birth outcomes need to be examined related to intervention effectiveness and that smoking cessation should be biochemically verified (Lumley et al., 2004). Indeed, the U.S. Preventive Services Task (USPSTF) force recently reaffirmed that the benefits of tobacco cessation intervention in pregnant women remain well established (USPSTF, 2009).

Providing smoking cessation interventions with pregnant women in rural, Southern Appalachia presents many challenges, and it is unclear the level to which they may be effective. Given the high rates of pregnancy smoking in this region, the potential benefits of intervention are substantial. However, many practical, professional, and cultural barriers exist. In this rural mountainous region, many women must travel great distances for prenatal care and other health services. Two thirds of the counties in Central and South-Central Appalachia do not have a single prenatal care provider, resulting in long drives to receive even the most basic care (Appalachian Regional Commission, 2014). A recent study showed that over 20% of pregnant women in Northeast Tennessee had to travel an hour or more to their delivery hospital (Bailey, McCook, & Chaires, 2014). Consequently, the rates of adequate prenatal care utilization are low in the region (MOD, 2014). Thus, even attempts to incorporate smoking cessation interventions into services pregnant women are already receiving are likely to miss many of the women who need them the most. Professional issues additionally interfere with addressing smoking as part of prenatal care in this region, as many health care providers in Appalachia perceive their patients as fatalistic or as passively accepting their own poor health and unwilling to follow health care recommendations (Shell & Tudiver, 2004). A survey of obstetricians in Appalachia revealed that only one quarter always asked pregnant patients about smoking and only a subset of these offered quit assistance or arranged for follow-up (Bailey & Jones Cole, 2009a). Finally, the culture of the region provides a significant barrier to addressing pregnancy smoking. Several studies conducted in the rural South have shown that due to isolation, lack of support, and fear of the unknown, lower income people in this region report low levels of motivation toward improving their health (Murimi & Harpel, 2010). Coupled with the tobacco producing history of the region and the normalization of cigarette smoking (Macnee & McCabe, 2004), affecting change related to pregnancy smoking in this region is challenging.

The goal of this project was to address four separate objectives. First, we sought to describe the components of a smoking cessation intervention, the Tennessee Intervention for Pregnant Smokers (TIPS) program, targeting low-income rural pregnant women with known high rates of smoking. Second, we examined the impact of the intervention on quit rates compared to usual care. Third, we described additional changes in smoking behaviors of women receiving the intervention, including reduction in smoking levels and number of quit attempts. Finally, we examined the impact of the TIPS intervention on birth outcomes.

Method

Participants

Study participants were recruited from five prenatal practices in rural South-Central Appalachia, all of which accepted Medicaid patients with no quota limits, resulting in a fairly low average socioeconomic status in each practice’s population of patients. The practices represented the spectrum of prenatal care options in the region including a university-affiliated obstetrics practice, a private obstetrics practice, a university-affiliated family medicine practice, a nurse-managed general practice, and a nurse-managed women’s health practice. As previously reported, the pregnancy smoking rate in the region from which the participants were drawn exceeds 25% (Bailey & Jones Cole, 2009b).

Intervention participants were recruited over a 4-year period from January 2008 to December 2011. All women entering prenatal care at the five practices during the intervention period who were current cigarette smokers were eligible. Because evaluation of the intervention involved the collection of research data, potential intervention participants had to also be willing to participate in the institutional review board–approved research interviews. At the first prenatal visit, the research data collection study and the intervention were explained by project staff as research and assistance related to pregnancy health behaviors, and those patients who were interested completed informed consent. Of the 1,685 eligible women, 1,595 (94.7%) agreed to participate in both the research and intervention (see below for additional information about later intervention refusers). To insure that all participating women had an opportunity to receive the intervention, women whose pregnancy ended prior to 20 weeks or who delivered less than 4 weeks after entering prenatal care were excluded from the study sample. Thus, the final intervention sample included 1,486 women (Figure 1).

Figure 1.

Summary of participant flow and outcome.

The control group received usual care as part of their prenatal visits. Usual care related to smoking varied by practice but included at minimum initial assessment of smoking status and brief advice to quit. In addition, many providers offered tips for smoking cessation, provided referrals for assistance, and checked about smoking at subsequent visits. Given the practical limitations related to randomization within small practices and to insure that none of the control group participants received any component of the intervention, which was part of systematic efforts region-wide, a historical control group was used. All women who entered prenatal care at the five practices prior to the intervention period, from January to December 2006, and were smokers at the first prenatal visit, were initially eligible to become part of the historical cohort. Of the 502 initially eligible women, those whose pregnancy ended prior to 20 weeks, or who delivered less than 4 weeks after entering prenatal care, were excluded, as was the case for the intervention group. Thus, the final control group sample included 461 women (Figure 1).

Procedures

Study procedures were approved by the affiliated university institutional review board and involved informed consent by participants. Participating women completed the research interview at or within 2 weeks of their first prenatal visit. A second research interview occurred during the third trimester. Research interviews included demographic assessment, screening for mental health and social issues, and detailed smoking and substance use surveys. As part of the interviews, exhaled carbon monoxide (CO) levels were checked using a CO monitor. In addition, urine samples were analyzed for cotinine. At delivery, prenatal medical charts and hospital delivery charts were reviewed for pregnancy and birth outcomes. Participants received a small monetary incentive for completing each research interview.

The Intervention

The intervention was designed to help pregnant women quit smoking by the end of the second trimester (Cliver et al., 1995; Ohmi, Hirooka, & Mochizuki, 2002) and remain smoking-free to delivery and beyond. It was offered and administered by health educators (i.e., dedicated staff as recommended by the USPSTF, 2009) with a minimum of 40 hours of training in brief intervention methodology, motivational interviewing, and the specifics of using a 5A’s (Ask, Advise, Assess, Assist, Arrange) model to address smoking in pregnancy. The formal education of the health educators ranged from bachelor’s degrees in social science related fields to master’s degrees in psychology and social work. They were housed in space within the prenatal practices and met with patients individually in either an exam room, private office, or counseling room.

A 5A’s model (American College of Obstetricians and Gynecologists [ACOG], 2005) was used at the initial contact and included asking about current smoking, advising quitting (including information about health risks), and assessing willingness to make a quit attempt in the next 30 days. The 5A’s model was chosen for the current intervention due to its published evidence-based effectiveness and recommendation by the ACOG for use with pregnant smokers (ACOG, 2005). For those who indicated a willingness to quit of 2 or less on a scale of 1 to 10, the health educator used education and motivational interviewing (Miller & Rollnick, 2002) in order to increase willingness to quit. If this was not successful, patients were permitted to decline participation at this point. When attempts to increase motivation were successful, or if the woman initially indicated a willingness to quit of 3 or more, the health educator moved into a brief intervention. This intervention, which used an expanded 5A’s model and involved motivational interviewing, was personalized to the needs of each patient. In all cases it involved addressing benefits and barriers to smoking cessation, and building knowledge, skills, self-efficacy, and resources to promote lasting behavior change. In particular, gaps between current behavior, personal goals, and health recommendations were highlighted.

Depending on patient needs, specific personalized assistance included practical tips for changing thinking and the environment leading up to quitting, education about withdrawal and how to manage it, use of a cessation contract and/or written action plan, assistance with practical needs (food, housing, transportation), stress relief advice, counseling or referral to counseling for stress and/or mental health issues, information about/referral to the state-administered smoking quitline, information about online resources, cessation education, assistance and referrals for family members, and enhancing motivation to attend all recommended prenatal visits. In addition, all intervention participants were offered a self-help book designed and published as part of the project (reference removed for peer review). The book was discussed and relevant parts highlighted by the health educators at the initial intervention session, with participants being encouraged to read and use it on their own. At the conclusion of the initial intervention session, participants were informed that the health educator would check in with them to provide additional intervention sessions during three additional prenatal visits (which typically occur monthly in the first and second trimester of pregnancy). Participants were given contact information for the health educator and encouraged to call or schedule additional appointments if desired. Initial intervention sessions, exclusive of research data collection, lasted from 15 to 30 minutes. Subsequent sessions with the health educators used the same expanded 5A’s model and lasted from 10 to 20 minutes. Fidelity checks occurred throughout the project to insure that the health educators were following the 5A’s model and tailoring the intervention to the needs and preferences of each patient.

Measures and Variables

The primary variable of interest was group membership, with participants classified as part of the intervention group or part of the historical control group.

Smoking status at entry to prenatal care was assessed solely via self-report, as those who deny smoking are unlikely to agree to a cessation intervention regardless of biochemical test results. However, smoking status by third trimester was assessed via self-report at final research interview, exhaled CO reading at final research interview, urinary cotinine at final prenatal visit, and self-report at hospital admission for delivery. A participant was considered to still be smoking if she indicated smoking via self-report, had a positive urinary cotinine (>50 ng/ml), or had a positive CO reading (>4 ppm; Bailey, 2013). Amount of smoking at entry to prenatal care and at delivery was assessed via self-report and CO levels during the research interviews. Number of quit attempts (smoking-free at least 24 hours) was assessed via self-report at each intervention and research contact and totaled over pregnancy.

Participant background and pregnancy information were obtained from both prenatal medical chart review and a demographic questionnaire at the first research interview and included age, race, education, medical insurance status, and reproductive history. Timing and frequency of prenatal care visits were recorded, and the Adequacy of Prenatal Care Utilization Index (Alexander & Kotelchuck, 1996) was calculated.

Birth outcome information was obtained from newborn delivery records including newborn health status (live birth vs. fetal demise, type of nursery care [neonatal intensive care unit vs. regular], and 1- and 5-minute Apgar scores), gestational age at delivery (based on ultrasound dating), birth size (weight, length, head circumference), and type of delivery (vaginal vs. cesarean).

Data Analysis

All analyses were performed using SPSS Version 20. Descriptive analyses examined the components of the intervention most used. Intervention and control groups were compared on background factors, smoking outcomes, and birth outcomes using t tests and chi-square analysis. As the two groups did not differ significantly (p > .05) on any of the background factors, subsequent controlled analyses were not performed. Changes in smoking behavior of intervention participants were evaluated descriptively and via paired t tests using a significance value of p < .05. The impact of successful cessation for intervention participants was assessed by comparing those who quit by delivery with those who continued to smoke through to delivery. Regression analysis (multivariate and logistic) was used to compare these two groups on birth outcomes (again using p < .05), controlling for significant background differences.

Results

Of the 1,486 women who consented to participate in the pregnancy health behavior research and smoking cessation intervention offered as part of their prenatal care, 90% (n = 1,332) actively participated (i.e., were willing to consider some health behavior change) in at least one intervention session with a health educator, 83% completed at least two sessions, and 21% received four or more counseling sessions. While all intervention session participants received an expanded 5A’s session and motivational interviewing, other intervention components were tailored to participant needs and requests (Table 1). By far the most provided/used service was information and assistance related to secondhand smoke exposure, followed by the program-designed self-help book. In addition, more than one third of intervention participants wanted information and assistance related to relapse prevention, a referral to the state quitline, referrals to other outside agencies, and assistance and information for family members who smoked. More than 30% of participants received practical assistance related to housing, food, transportation, or government services. At least one in six participants wanted and received assistance related to stress management or mental health issues, while fewer than 1 in 50 needed and were willing to receive information or assistance related to intimate partner violence.

Table 1.

Frequency of Utilization of Tailored Intervention Components/Services (N = 1,486).

Component/service	Frequency of utilization (%)
Secondhand smoke information and assistance	78.3
Program-designed self-help book	57.0
Relapse prevention information and assistance	40.7
Referral to the state quitline	38.0
Assistance and information for family members	33.7
Practical assistance (food, housing, transportation, government services)	31.4
Counseling/assistance/information about stress relief	16.9
Counseling/referral for mental health issues	16.8
Assistance related to intimate partner violence	1.4

Background characteristics by participant group (intervention or control) are presented in Table 2. The average participant was younger than 25 years, was Caucasian, possessed only a high school diploma, and had income low enough to qualify for government medical insurance assistance (i.e., state Medicaid). The intervention participants and control group members did not differ significantly on any of the available background variables.

Table 2.

Background Characteristics for Intervention and Control Group Participants.

Characteristic	Intervention group (n = 1,486)	Historical control group (n = 461)	t/χ²	p
Age (years)	24.2	24.5	1.21	.227
Race (% Caucasian)	93.8	93.6	0.42	.690
Education (years)	12.1	12.2	0.19	.843
Insurance (% Medicaid/none)	87.6	86.8	0.32	.763
Number of previous pregnancies	2.6	2.7	1.78	.075
Number of live births	1.2	1.2	0.36	.722

The intervention and control groups differed significantly on smoking status at delivery. Of those offered the intervention, 28.1% quit smoking by the end of the second trimester and remained smoking-free to delivery. In contrast, only 9.8% of the control group, who received usual care only, quit smoking prior to the third trimester and remained smoking-free to delivery (χ² = 59.2, p < .001).

Smoking reduction and quit attempts were also tracked for intervention participations, with a significant reduction in the amount of pregnancy smoking found. Among the participants who were not successful in quitting smoking during pregnancy (n = 1,068), 63.9% reported a 50% or more reduction in the number of cigarettes smoked/day. For these women the average number of cigarettes smoked/day at conception was 17.7 and at delivery was 8.6 cig/day (t = 47.3, p < .001). Exhaled CO readings supported these data, with the average CO reading at entry to prenatal care of 10.9 dropping to 6.0 at the end of pregnancy (t = 26.4, p < .001) for those who continued to smoke. Finally, 41.5% reported at least one pregnancy quit attempt lasting longer than 24 hours.

The effectiveness of the TIPS program was also examined in relation to pregnancy outcomes. Compared to the historical control group, intervention group participants were significantly less likely to experience a fetal or neonatal death and significantly more likely to deliver babies who were heavier, were longer, and with higher 5-minute Apgar scores. Specifically, infants of women in the intervention group were 270 g heavier. In addition, the infants of intervention participants were more than 50% less likely to be admitted to the neonatal intensive care unit (Table 3). Finally, intervention group participants were nearly 50% less likely to have less than adequate prenatal care utilization.

Table 3.

Outcomes by Participant Group.

Outcome	Intervention group (n = 1,486)	Historical control group (n = 461)	t/χ²	p
Quit smoking by end of second trimester (%)	28.1	9.8	59.2	<.001
Birth weight (g)	3,063	2,793	7.27	<.001
Low birth weight (%)	13.6	24.9	30.51	<.001
Birth length (inches)	19.4	18.8	6.18	<.001
Gestational age at delivery (weeks)	36.8	36.8	0.01	.940
Apgar score: 1-minute	7.7	7.7	0.36	.718
Apgar score: 5-minute	8.8	8.6	3.14	.002
Nursery assignment (%neonatal intensive care unit)	7.6	19.5	61.50	<.001
Fetal/neonatal demise (%)	0.8	2.6	21.11	<.001
Prenatal care utilization (% less than adequate)	27.4	41.6	38.96	<.001

The impact of quitting smoking during pregnancy on birth outcomes was examined among the intervention participants, controlling for background and pregnancy differences between those who quit smoking and those who did not. Compared to intervention participants who did not quit smoking, those who quit and were smoking-free at delivery had larger babies, with significantly heavier birth weights (difference of more than 200 g), longer lengths, and greater head circumferences. In fact, the risk of delivering a LBW baby dropped in half for women who quit smoking. In addition, those who quit smoking had babies with significantly shorter delivery hospital stays and were 80% less likely to experience a fetal/neonatal death (Table 4). Finally, those who successfully quit smoking were significantly more likely to have had adequate prenatal care utilization.

Table 4.

Adjusted Birth Outcomes by Cessation Status for Intervention Participants.

Outcome	Quit smoking (n = 418)	Continued smoking (n = 1,068)	t/Wald	p
Birth weight (g)	3,216	3,012	7.27	<.001
Low birth weight (%)	7.4	14.4	4.44	.032
Birth length (inches)	19.7	19.2	4.20	<.001
Head circumference (in)	13.3	13.1	2.47	<.001
Gestational age at delivery (weeks)	38.6	38.4	0.62	.533
Preterm delivery (%)	9.8	13.8	2.01	.089
Apgar score: 1-minute	7.7	7.6	1.28	.200
Apgar score: 5-minute	8.8	8.7	0.36	.72
Nursery assignment (% neonatal intensive care unit)	7.9	8.6	1.63	.144
Fetal/neonatal demise (%)	0.2	1.0	2.83	.046
Prenatal care utilization (% less than adequate)	22.9	33.3	4.01	.039

Note. Results from final step of multiple or logistic regression analysis controlling for all significant confounders: maternal age, maternal education, maternal marital status, maternal insurance status, and pregnancy marijuana use. Nonsignificant confounders not included in the model: maternal race, number of previous pregnancies, number of live deliveries, and pregnancy alcohol use.

Discussion

The results of the current study demonstrate the effectiveness of a pregnancy smoking intervention implemented as part of prenatal care in a rural, Southern population with high rates of smoking and poor birth outcomes. The biochemically verified quit rate of 28% for the TIPS program exceeds rates described in the published literature for similar interventions (Chamberlain et al., 2013). In addition, the improvements in birth outcomes, including a 270-g increase in birth weight compared to those who received usual care, are similar to or exceed effects described for other interventions (Chamberlain et al., 2013).

The current findings point to some of the intervention components that may be important to include with similar populations. Having dedicated staff who were trained in smoking cessation interventions, who used motivational interviewing that included a 5A’s model, and structured the intervention around each patient’s willingness to quit was a standard of this program. In addition, many of the individualized components were implemented for a large percentage of participants including tailored self-help materials, assistance related to secondhand smoke, involvement of family members, assistance with practical needs, and counseling for mental health concerns. Having health educators who assessed environmental, social, mental health, and motivational factors and then tailored the intervention for each participant permitted the most effective use of time-limited encounters while maximizing benefits, and suggests an effective model for future intervention and research efforts. In addition, since the intervention was provided as part of prenatal care in a population with high rates of inadequate utilization, an important component of intervention sessions was to increase motivation for adequate prenatal care attendance, which not only increased the likelihood of additional smoking cessation intervention sessions but also provided additional health benefits from receiving the recommended prenatal care. Indeed, compared to the control group with only 58% adequate prenatal care utilization, 73% of the intervention group had adequate prenatal care utilization.

The current study and intervention program are not without limitations, and these should be addressed in future work. The use of a historical control group, instead of a current control group as part of a randomized controlled design, could have introduced additional differences between the two groups. Because the control group received usual care, consent was not required, resulting in all eligible patients included. While study refusal rates were low among intervention group participants (5%), and all women were encouraged to participate even if they had no intention of quitting, exclusion of some women from the intervention group could have biased the results favoring the intervention group. In addition, possible changes in pregnancy smoking over time as a result of increased public awareness and increased cigarette taxes may have led to differences between intervention and control groups, again possibly increasing quit rates among intervention group participants. While a time difference of only a couple of years is unlikely to have a large effect, this possibility cannot be ruled out. The current intervention, with the use of dedicated grant-funded health educators, was costly to implement and may not be sustainable in community practice. However, the cost analysis done as part of the final report for this project showed that for every $1 spent on the TIPS program, $4.75 was saved in newborn hospital costs alone (with longer terms benefits likely even more substantial), with a net benefit newborn cost savings of $7.5 million over the 5-year program (TIPS, 2014). Clearly if start-up money can be found, investing in efforts such as the TIPS program will pay off in the long run, both financially and in improved community health, and may have increased likelihood of sustainability as insurers become more willing to reimburse for such services as reports like this one continue to be published.

Footnotes

Acknowledgements

The author would like to thank the TIPS staff, the student workers, the participating medical practices, and most of all the participating women.

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Funding for this project was provided by the State of Tennessee Department of Health in a grant to Dr. Beth A. Bailey.

References

Alexander

G. R.

Kotelchuck

(1996). Quantifying the adequacy of prenatal care: A comparison of indices. Public Health Reports, 111, 408-418.

American College of Obstetricians and Gynecologists. (2005). Smoking cessation during pregnancy: Committee opinion. Obstetrics & Gynecology, 106, 883-889.

Appalachian Regional Commission. (2014). Reports, maps, and data. Retrieved from http://www.arc.gov/research/RegionalDataandResearch.asp

Bailey

B. A.

(2013). Using expired air carbon monoxide to determine smoking status during pregnancy: Preliminary determination of an appropriately sensitive and specific cut-point. Addictive Behavior, 38, 2547-2550.

Bailey

B. A.

Jones Cole

(2009a). Are obstetricians following best-practice guidelines for addressing pregnancy smoking? Results from Northeast Tennessee. Southern Medical Journal, 102, 894-899.

Bailey

B. A.

Jones Cole

(2009b). Rurality and birth outcomes: Findings from Southern Appalachia and the potential role of pregnancy smoking. Journal of Rural Health, 25, 141-149.

Bailey

B. A.

McCook

J. G.

Chaires

(2014). The burden of early term births in rural Appalachia. Southern Medical Journal, 107, 624-629.

Bastra

Hadders-Algra

Neeleman

(2003). Effect of antenatal exposure to maternal smoking on behavior problems and academic achievement in childhood: Prospective evidence from a Dutch birth cohort. Early Human Development, 75, 21-33.

Chamberlain

O’Mara-Eves

Oliver

Caird

J. R.

Perlen

S. M.

Eades

S. J.

Thomas

(2013). Psychosocial interventions for supporting women to stop smoking in pregnancy (Review). Cochrane Database of Systematic Reviews, (10), CD001055.

10.

Chiolero

Bovet

Paccaud

(2005). Association between maternal smoking and low birth weight in Switzerland: The EDEN study. Swiss Medicine Weekly, 135, 525-530.

11.

Cliver

S. P.

Goldenberg

R. L.

Cutter

G. R.

Hoffman

H. J.

Davis

R. O.

Nelson

K. G.

(1995). The effect of cigarette smoking on neonatal anthropometric measurements. Obstetrics & Gynecology, 85, 625-630.

12.

Cnattingius

(2004). The epidemiology of smoking during pregnancy: Smoking prevalence, maternal characteristics, and pregnancy outcomes. Nicotine & Tobacco Research, 6(Suppl. 2), S125-S140.

13.

Espy

K. A.

Fang

Johnson

Stopp

Wiebe

S. A.

(2011). Prenatal tobacco exposure: Developmental outcomes in the neonatal period. Developmental Psychology, 47, 153-156.

14.

Fingerhut

L. A.

Kleinman

J. C.

Kendrick

J. S.

(1990). Smoking before, during, and after pregnancy. American Journal of Public Health, 80, 541-544.

15.

Lumley

Chamberlain

Dowswell

Oliver

S. S.

Oakley

Watson

(2009). Interventions for promoting smoking cessation during pregnancy (Review). Cochrane Database of Systematic Reviews, (3), CD001055.

16.

Lumley

Oliver

S. S.

Chamberlain

Oakley

(2004). Interventions for promoting smoking cessation during pregnancy (Review). Cochrane Database of Systematic Reviews, (4), CD001055.

17.

Macnee

C. L.

McCabe

(2004). The transtheoretical model of behavior change and smokers in Southern Appalachia. Nursing Research, 53, 243-250.

18.

March of Dimes. (2014). PeriStats. Retrieved from http://www.marchofdimes.com/peristats/

19.

Miller

W. R.

Rollnick

(2002). Motivational interviewing: Preparing people for change (2nd ed.). New York, NY: Guilford Press.

20.

Mullett

M. D.

Britton

C. M.

John

Hamilton

C. W.

(2010). WV birth score: Maternal smoking and drugs of abuse. West Virginia Medical Journal, 106, 16-20.

21.

Murimi

M. W.

Harpel

(2010). Practicing preventive health: The underlying culture among low-income rural populations. Journal of Rural Health, 26, 373-382.

22.

Ohmi

Hirooka

Mochizuki

(2002). Fetal growth and the timing of exposure to maternal smoking. Pediatrics International, 44, 55-59.

23.

Rogers

J. M.

(2008). Tobacco and pregnancy: Overview of exposure and effects. Birth Defects Research, 84, 1-15.

24.

Shell

Tudiver

(2004). Barriers to cancer screening by rural Appalachian primary care providers. Journal of Rural Health, 20, 368-373.

25.

Tennessee Intervention for Pregnant Smokers. (2014). Five year progress report. Retrieved from http://www.etsu.edu/tips/TIPS%205%20year%20Progress%20Summary.pdf

26.

Tong

V. T.

Jones

J. R.

Dietz

P. M.

D’Angelo

Bombard

J. M.

(2009). Trends in smoking before, during, and after pregnancy: Pregnancy Risk Assessment Monitoring System (PRAMS), United States, 31 sites, 2000-2005. Morbidity and Mortality Weekly Review, 58, 1-29.

27.

U.S. Department of Health and Human Services. (2001). Women and smoking: A report of the Surgeon General. Rockville, MD: National Center for Chronic Disease Prevention and Health Promotion, Office of Smoking and Health.

28.

U.S. Preventive Services Task Force. (2009). Counseling interventions to prevent tobacco use and tobacco-caused disease in adults and pregnant women: U.S. Preventive Services Task Force reaffirmation recommendation statement. Annals of Internal Medicine, 150, 551-555.