• unlimited access with print and download
    $ 37 00
  • read full document, no print or download, expires after 72 hours
    $ 4 99
More info
Unlimited access including download and printing, plus availability for reading and annotating in your in your Udini library.
  • Access to this article in your Udini library for 72 hours from purchase.
  • The article will not be available for download or print.
  • Upgrade to the full version of this document at a reduced price.
  • Your trial access payment is credited when purchasing the full version.
Buy
Continue searching

Validity of Roger's diffusion of treatment innovation for pregnant smokers

ProQuest Dissertations and Theses, 2009
Dissertation
Author: Kalpana Ramiah
Abstract:
The current research study aimed at developing an instrument to measure perceived attributes of using an innovation in a public health system. In this study, the innovation was a behavioral program (Smoking Cessation and Reduction In Pregnancy Treatment -- SCRIPT) to help pregnant mothers quit smoking and the public health system is the West Virginia (WV) Right From the Start Project. WV has a very high smoking rate amongst their pregnant women (27.1%) compared to the national rate (10.2%). The Designated Care Coordinators (DCCs) provide the SCRIPT program to the RFTS clients to help them quit smoking. A measurement tool was developed to measure the perceived attributes of using the SCRIPT program by the DCCs. This instrument (SCRIPT Adoption Survey - SAS) was developed to address the five constructs of Rogers Diffusion of Innovation model: Relative Advantage, Compatibility, Complexity, Observability, and Trialability. The instrument was tested for its validity and reliability. Forty three items were administered to the DCCs. Twenty Eight of these 43 items were found to be reliable and valid. The internal consistency of the full scale was 0.926 and test retest reliability was 0.756 (p<0.0001). The internal consistency of the sub-scales ranged from 0.667 to 0.879. The test-retest reliability of the sub-scales ranged from 0.599 to 0.737. The lowest reliability was found in the Observability domain. All items had a factor loading of greater than 0.4 (0.43-0.81). Model Fit statistics were not conclusive; however, the Akaike Information Criterion (AIC) was smallest in the hypothesized model. Correlation between SAS score and implementation rate measured by Program Implementation Index (PII) was 0.336 (p=0.0028); which will explain 11.29% of variation in the implementation rate. Trialability had the highest correlation of 0.391 (p=0.0004) explaining 15.29% of variance in the data.

ix Table of Contents Dedication ................................................................................................. iv Acknowledgments ....................................................................................... v Abstract of Dissertation ............................................................................. vii Table of Contents ....................................................................................... ix List of Figures ............................................................................................xv List of Tables............................................................................................ xvi Chapter 1: Introduction ................................................................................ 1 1.1 Problem ............................................................................................. 1 1.2 Public Health Response....................................................................... 2 1.3 Diffusion of Public Health Programs ................................................... 3 1.4 Dissemination of Evaluation ............................................................... 6 1.5 Chapter Overview ............................................................................... 7 1.6 Research Question .............................................................................. 8 Chapter 2: Literature Review ....................................................................... 9 2.1 Overview of Smoking ......................................................................... 9 2.2 Health Effects of Smoking During Pregnancy ....................................10 2.3 Smoking Behavior Trends ..................................................................12 2.4 Behavioral Interventions for Pregnant Smokers ..................................15 2.5 An Innovation for The Right From The Start (RFTS) Program ...........17 2.6 SCRIPT Procedures ...........................................................................18 2.7 Right From The Start (RFTS) Program ..............................................19 2.8 SCRIPT Dissemination Organizational Development.........................22

x

2.9 Diffusion of Innovation (DOI) Theory ...............................................22 2.10 Characteristics of an Innovation .......................................................23 2.10.1 Relative Advantage ..............................................................26 2.10.2 Compatibility .......................................................................27 2.10.3 Complexity ...........................................................................28 2.10.4 Trialability............................................................................29 2.10.5 Observability ........................................................................30 2.10.6 Additional Constructs ...........................................................30 2.11 Perception of an Innovation vs Using an Innovation .........................32 2.12 Existing Measures for Diffusion of Innovation .................................32 2.13 Development of The SCRIPT ADOPTION SURVEY (SAS) ..........37 Chapter 3: Methods ....................................................................................39 3.1 Statistical Packages ............................................................................40 3.2 Phase One: SCRIPT Adoption Scale (SAS) Development ..................40 3.2.1 SAS Constructs .........................................................................40 3.2.2 Item Creation ............................................................................43 3.2.3 SAS Scale .................................................................................45 3.3 Phase Two:Instrument Administration and Psychometric Analysis.....46 3.3.1 Instrument Administration .........................................................46 3.3.2 Population Characteristics .........................................................47 3.3.3 Data Preparation ........................................................................47 3.4 Reliability ..........................................................................................48

xi 3.4.1 Internal Consistency ..................................................................48 3.4.2 Test-Retest Reliability ...............................................................49 3.5 Construct and Criterion Validity ........................................................50 3.5.1 Construct Validity .....................................................................50 3.5.1.1 Model Identification ............................................................52 3.5.1.2 Parameter Identification .......................................................53 3.5.1.3 Parameter Estimation ...........................................................53 3.5.1.4 Data-Model Fit ....................................................................53 3.5.1.5 Alternate Model Comparison ...............................................57 3.5.1.6 Possible Model Refit ...........................................................58 3.5.2 Criterion Validity ......................................................................62 3.6 Research Questions ............................................................................64 Chapter 4: Results.......................................................................................65 4.1Draft SAS ...........................................................................................65 4.2 Sample Description ............................................................................66 4.3 Univariate Analysis ............................................................................67 4.4 Psychometric Analysis .......................................................................71 4.4.1 Internal Consistency ..................................................................71 4.4.1.1 Relative Advantage .............................................................71 4.4.1.2 Compatibility ......................................................................73 4.4.1.3 Complexity ..........................................................................74 4.4.1.4 Observability .......................................................................75

xii 4.4.1.5 Trialability...........................................................................76 4.4.2 Factor Analysis .........................................................................77 4.5 Finalizing the Items of Scale ..............................................................84 4.5.1 Reliability .................................................................................84 4.5.1.1 Relative Advantage .............................................................84 4.5.1.2 Compatibility ......................................................................85 4.5.1.3. Complexity .........................................................................86 4.5.1.4. Observability ......................................................................87 4.5.1.5. Trialability ..........................................................................87 4.5.2 Validity .....................................................................................88 4.6 Test Retest Reliability ...................................................................... 103 4.7 Predictive Analysis .......................................................................... 104 4.8 SAS Score ....................................................................................... 107 4.9 Summary of Analysis ....................................................................... 107 Chapter 5: Discussion ............................................................................... 108 5.1 Overview ......................................................................................... 108 5.2 SAS Development............................................................................ 109 5.3. SAS Score....................................................................................... 114 5.4. Designated Care Coordinators (DCCs) Characteristics .................... 115 5.5 Reliability ........................................................................................ 116 5.5.1 Internal Consistency ................................................................ 117 5.5.2 Test-Retest Reliability ............................................................. 118 5.5.3 Hypothesis One ....................................................................... 121

xiii 5.6 Validity ............................................................................................ 122 5.6.1 Confirmatory Factor Analysis ................................................. 122 5.6.1.1 Relative Advantage ........................................................... 123 5.6.1.2 Compatibility .................................................................... 123 5.6.1.3 Complexity ....................................................................... 124 5.6.1.4 Observability ..................................................................... 124 5.6.1.5 Trialability......................................................................... 125 5.6.2 Data Model Fit ........................................................................ 125 5.6.2.1 Model Fit Statistics of Hypothesized Model ..................... 126 5.6.2.2 Comparison of Hypothesized and Alternate Models........... 126 5.6.3 Hypothesis Two ...................................................................... 127 5.7. Hypothesis Three: Predictive Analysis ............................................ 130 5.8 Strengths and Limitations................................................................. 132 5.9 Recommendations for Future Research and the SCRIPT Program .... 134 References ................................................................................................ 137 Appendices ............................................................................................... 146 Appendix A: WV SCRIPT Policy and Management Committee ............ 146 Appendix B: Initial 43 item Scale ......................................................... 149 Appendix C: Frequency of Responses ................................................... 153 Appendix D: Covariance Matrix for Initial 43 Items .............................. 156 Appendix E: Items Sent to the External Expert Reviewers ..................... 163 Appendix F: Correlation of Relative Advantage Domain (9-items) ....... 167 Appendix G: Correlation of Compatibility Domain (4-items) ................ 167

xiv Appendix H: Correlation of Complexity Domain (6-items) ................... 168 Appendix I: Correlation of Observability Domain (4-items) .................. 168 Appendix J: Correlation of Trialability Domain (5-items) ..................... 169

xv List of Figures

Figure 1: Regional RFTS Divisions ............................................................ 21 Figure 2: Process to Establish Face and Content Validity.............................. 45 Figure 3: Hypothesized Five Factor Model................................................... 56 Figure 4: Alternate Four Factor Model (Holloway)....................................... 59 Figure 5: Alternate Three Factor Model (Pankratz)....................................... 60 Figure 6: Alternate Four Factor Model (Tornatzky and Klein) ...................... 61 Figure 7: Hypothesized Five Factor Model with 28 Items ............................. 97 Figure 8: Alternate Four Factor Model (Holloway) with 28 Items ................. 98 Figure 9: Alternate Three Factor Model (Pankratz) with 28 Items ................. 99 Figure 10: Alternate Four Factor Model (Tornatzky and Klein) with 28 Items100 Figure 11: Change in Correlation between Hypothesized Model and Alternate Model #1 ............................................................................. 120 Figure 12: Change in Correlation between Hypothesized Model and Alternate Model #2 ............................................................................. 120 Figure 13: Final Measurement Model for SAS ........................................... 129

xvi List of Tables

Table 1: Smoking Prevalence (in %) in the U.S. .......................................... 14 Table 2: Core Perceived Attributes of Innovation ......................................... 41 Table 3: Sample PII Calculation .................................................................. 63 Table 4: Length of SAS ............................................................................... 66 Table 5: Response to SAS by Region ........................................................... 67 Table 6: Univariate Analysis for Relative Advantage ................................... 68 Table 7: Univariate Analysis for Compatibility ............................................ 69 Table 8: Univariate Analysis for Complexity .............................................. 69 Table 9: Univariate Analysis for Observability ............................................. 70 Table 10: Univariate Analysis for Trialability .............................................. 70 Table 11: Relative Advantage -- 14 Items ................................................... 72 Table 12: Compatibility -- 8 Items .............................................................. 73 Table 13: Complexity -- 8 Items .................................................................. 74 Table 14: Observability -- 7 Items ................................................................ 75 Table 15: Trialability -- 6 items ................................................................... 76 Table 16: Internal Consistency after Preliminary Analysis --38 Items ........... 77 Table 17: Factor Loadings of--38-Items ....................................................... 78 Table 18: Model Fit Statistics -- 31-items ..................................................... 81 Table 19: Factor Loadings -- 31-items (Hypothesized 5-factor Model) ......... 82 Table 20: Reliability Estimates of Reliability Domain--9 Items .................... 85 Table 21: Reliability Estimates of Compatibility Domain--6 Items ............... 86 Table 22: Reliability Estimates of Complexity Domain--6 Items................... 86

xvii Table 23: Reliability Estimates of Observability Domain--5 Items ................ 87 Table 24: Reliability Estimates of Trialability Domain--5 Items ................... 88 Table 25: Factor Loading for 29 Items ......................................................... 89 Table 26: Model Fit Statistics for 29-item Scale ........................................... 91 Table 27: Factor Loadings for 28 Items ........................................................ 92 Table 28: Model Fit Statistics for 28 Item Scale ........................................... 96 Table 29: Internal Consistency of the Hypothesized 5-Factor Scale............. 101 Table 30: Internal Consistency for Alternative Model 1: Holloway ............. 102 Table 31: Internal Consistency for Alternative Model 2: Pankratz ............... 102 Table 32: Internal Consistency for Alternative Model 3: Tornatzky ............ 102 Table 33: Test Retest Reliability for Hypothesized 5-Factor ....................... 103 Table 34: Test Retest Reliability for Alternative Model 1: Holloway .......... 103 Table 35: Test Retest Reliability for Alternative Model 2: Pankratz ............ 104 Table 36: Test Retest Reliability for Alternative Model 3: Tornatzky .......... 104 Table 37: Predictive Analysis for Hypothesized 5-factor Model.................. 105 Table 38: Predictive Analysis for Alternative Model 1: Holloway .............. 105 Table 39: Predictive Analysis for Alternative Model 2: Pankratz ................ 106 Table 40: Predictive Analysis for Alternative Model 3: Tornatzky .............. 106 Table 41: Summary of SAS Score ............................................................. 107

1

Chapter 1: Introduction

1.1 Problem

Smoking is the main cause of multiple medical conditions such as lung cancer, heart diseases, stroke, etc. (International Agency for Research on Cancer [IARC], 1986; U.S. Department of Health and Human Services [U.S.DHHS], 2004). Smoking during pregnancy is the primary cause of a large number of adverse perinatal outcomes including: low birth weight, infant mortality, pre-term labor, spontaneous abortion, stillbirth, intrauterine growth retardation, small for gestational age, cleft lip with or without cleft palate, placenta previa, placental abruption, and premature rupture of membrane (American College of Obstetrics and Gynecology [ACOG], 2005; IARC, 1986; U.S.DHHS, 2004). Smoking during pregnancy doubles the risk of having a low birth weight (LBW) baby, premature rupture of membranes, placental abruption, and placental previa (Martin et al., 2006). In 2004, the Surgeon General’s Report noted that if all women in the U.S. stopped smoking during pregnancy, there would be an estimated 11% decrease in stillbirths and 5% reduction in newborn deaths (U.S.DHHS, 2004). One report has indicated that daily fetal exposure to carbon monoxide (CO), nicotine, and other tobacco smoke chemicals and carcinogens is worse than cocaine-related fetal exposures (Slotkin, 1998). In 2006, 34% of women of childbearing age (18-44 years) smoked in West Virginia (WV) compared to the national average of 22.4% (Centers for Disease Control and Prevention [CDC], 2008a). This high smoking rate among women of childbearing

2

age translates into a high smoking rate during pregnancy in WV. The national average for smoking during pregnancy is 10.2%, while in WV it is 27.1%; highest in the U.S. Among Medicaid-eligible pregnant women in WV, the self-reported rate of smoking at the onset of prenatal care has been approximately 46.6% for the last five years, which is more than double the national average of 18.6% (CDC, 2009). The recent increase in cigarette taxes is expected to help with reduction in cigarette use. However, due to the current economic conditions, one of the rival hypotheses is: low employment rates and increases in stress due to economic hardship may increase the use of cigarettes. The reward pathway associated with nicotine and dopamine release may be seen as a stress relieving mechanism. More longitudinal research is needed to understand the impact of increased taxes and dire economic conditions on cigarette consumption, specifically smokers with low income. However, public health response needs to be strengthened to control and reduce smoking during tough economic conditions.

1.2 Public Health Response

In response to this harmful trend, the WV Department of Health and Human Resource (WV DHHR) and the George Washington University School of Public Health and Health Services (GWU SPHHS) established a partnership, with funding support from the National Cancer Institute (NCI), to disseminate and evaluate an evidence-based cessation program for pregnant smokers. The Smoking Cessation and Reduction In Pregnancy Treatment (SCRIPT) Program has been recommended by American College

3

of Obstetricians and Gynecologists (ACOG) and Agency for Healthcare Research and Quality (AHRQ) as an effective evidence-based behavioral treatment for smoking cessation among pregnant women. Currently, the WV Right From the Start (RFTS) program policy states that the SCRIPT program should be provided to its clients at their regular home visits. The SCRIPT program is delivered by Designated Care Coordinators (DCCs) who are registered nurses or registered social workers.

1.3 Diffusion of Public Health Programs

Studies and reviews in the past two decades have emphasized the need for dissemination, diffusion, and adoption of “Best Practice” tobacco treatment programs by systems and clinical practices (Curry, Fiore, Tracy Orleans, & Keller, 2002; Orleans, 2001; Solberg et al., 1996; Windsor et al., 1993). Diffusion is a process in which an innovation is communicated through certain channels over time among the members of a social system (Rogers 2003). In the current context, adoption is implementation of a program or intervention with fidelity. Finally, dissemination is active spreading the intervention and its benefits. These concepts are further elaborated and operationlized in Chapters Two and Three. In the recent Clinical Practice Guidelines for Tobacco Treatment released by AHRQ, the Guidelines Panel recommended that future research should focus on the “effectiveness of different types of counseling, behavioral therapies, and motivational interventions (e.g., physiological feedback of adverse impacts, quitting benefits) for

4

pregnant women in general and in high prevalence populations (Fiore et al., 2008, p. 165)." Although the RFTS program policy is to provide the SCRIPT Program to its clients, rigorous evaluation of the level of adoption, implementation, and behavioral impact has not been conducted. Studies suggest that a typical reason for replication failure of a health service program is lack of evaluation data (Goldman, 2003). Along with replication failure, there remains a gap between evidence and practice for clinical and preventive services (Institutes of Medicine [IOM], 2001). Many health services programs with demonstrated efficacy in single trials have not been adopted and implemented by systems of care (Furano, Jucovy, Racine, & Smith, 1995; Goldman, 2003; Rogers, 2003). Quality improvement methods using ongoing evaluation data and systems change based on empirical evidence and insight may be a way to reduce this gap. The challenges of transitioning from the decision to adopt an innovation to the consistent use of the innovation (implementation) include lack of sustained leadership support, inadequate resources allocated for implementation, insufficient staff training and time, failure to develop and use a measurement and feedback system, and cultural resistance to change (Alexander, Weiner, Shortell, & Baker, 2007; Klein & Sorra, 1996). Counte and Meurer (2001) estimated that less than 40% of health care initiatives transition from adoption to implementation. Nearly 50 years ago, Everett Rogers presented a rubric to understand multiple issues that affect adoption of an innovation. Innovation is an idea, practice, or object that is perceived as new. Rogers introduced the theory of Diffusion of Innovation (DOI) in his book in 1962. This theory tries to explain how, why, and at what rate a new idea spreads

5

in a social system. In some cases, diffusion of an innovation takes time to be adopted and implemented. For example, in 1601, Captain James Lancaster discovered that administering lime juice to sailors prevented death due to scurvy. This simple technique, however, was not adopted officially by the British Board of Trade until 1865. In another setting, a mechanical peanut churning machine introduced by the United Nations Development Programme (UNDP) in Mali took less than a decade to spread to a number of villages. The DOI theory hypothesizes that a large variation in adoption and implementation rate could be dependent on the following: perceived attributes of innovation (relative advantage, compatibility, complexity, trialability, and observability), the type of innovation-decision (optional, collective, or authoritative), communication channel (mass media or interpersonal), nature of the social system, and extent of the change agent’s promotion effect. Dissemination of an innovation within an organization depends on the perception of the innovation by the leadership and regular users, the characteristics of the individuals who may adopt the change, and contextual and managerial factors within the organization (Berwick, 2003). Of the array of salient predictor variables, the perceived attributes of the innovation explain the majority of variance in the rate of adoption (Rogers, 2003). Ostlund (1974), based on two of his studies, concluded that perceived attributes of innovation was a stronger predictor of adoption than the characteristics of the innovator, i.e., innovators versus non-innovators. Based on the Rogers diffusion of innovation model, Berwick (2003) listed the following five characteristics of innovations that influence adoption in health care settings: 1) the perceived benefits of the change, 2) observability of the innovation, 3) compatibility of the change with the current

6

organizational culture and personal belief systems, 4) level of simplicity of the innovation, and 5) trialability of the innovation. Although the diffusion theory has been used in the health sector, limited attention has been paid to applying the theory to measuring the variables that affect the rate of adoption of new behavior change programs. Only qualitative studies have been conducted and published by successful programs and, unfortunately, there is less literature on programs that did not work. The paucity of empirical evidence indicates a need to develop a tool to measure the perceived attributes of an innovation in the public health arena. The theory of Diffusion of Innovation will be used to measure the level of dissemination and adoption of the SCRIPT Program in WV by the RFTS program and its DCCs. Since most public health programs end in the efficiency phase, a tool to demonstrate effectiveness on a larger scale would benefit public health researchers and programmers. A reliable and valid measurement tool (SCRIPT Adoption Scale – SAS) will also increase the adoption of the SCRIPT Program by other states.

1.4 Dissemination of Evaluation

Although a number of books are written about the importance and methods of program monitoring and evaluation, there is paucity in guidance to evaluate the dissemination of public health programs. As mentioned earlier, this may be due, in part, to lack of funding. This paucity can also be due to lack of scientific rigor in this area. Moreover, public health programs cover a wide range of topics and issues with varying

7

levels of complexity. For instance, measurement of the dissemination of a vaccination campaign would likely be much easier than a measurement of a behavioral change program since the key evaluation criterion of interest is the number of vaccinations administered, which is more easily quantifiable than behavior. Behavioral change programs, in general, are difficult to measure even on a small scale. Measurement tools specifically to measure dissemination and diffusion are imperative to strengthen public health systems both in the U.S. and around the world. Such evidence of effectiveness forms the necessary building blocks for the future evidence-based public health system and structure. In addition, the availability of funding for public health problems has always been much less than the need. Building strong scientific knowledge in evaluation of diffusion of public health programs in public health systems is imperative to make wise use of the scarce resources.

1.5 Chapter Overview

Diffusion studies can be categorized into three major areas: perceived attributes of an innovation, characteristics of adoption units, and subject-specific factors such as communication, leadership, social networks, and incentives. Chapter Two of this dissertation focuses on existing literature about the measurement of perceived attributes of innovation and its application to the dissemination evaluation of the SCRIPT Program. The summary of the literature will demonstrate the need for a measurement tool. It is best practice to establish validity and reliability when a new measurement tool is used with a new population. Validity represents how well a scale measures the

8

construct of interest, and reliability represent how well a scale provides consistent information if administered more than once. These measurement characteristics, reliability and validity, of the newly developed tool will be tested using factor analysis and other statistical methods. A detailed description of the method of developing the instrument, its administration, and statistical tests to measure the validity and reliability are presented in Chapter Three. The results of the development process and its psychometric testing are presented in Chapter Four. Chapter Five concludes with a discussion of the potential contribution to the fields of public health practice, theory, and research.

1.6 Research Question

In this dissertation, using the methods explained in Chapter Three, the following research questions are addressed:  Will the SCRIPT Adoption Scale demonstrate adequate reliability?  Will the SCRIPT Adoption Scale demonstrate adequate validity?  Is the perception of using an innovation associated with performance in using the SCRIPT innovation?

9

Chapter 2: Literature Review

This chapter presents a synopsis of current literature for smoking in pregnancy, its health effects, and a review of studies on diffusion of innovation theory and its uses in healthcare. The chapter concludes with a discussion of the need for studies of the perceived attributes of using an innovation in a public health system and the necessity of measuring these attributes.

2.1 Overview of Smoking

Smoking causes lung cancer (IARC, 1986; IARC, 2002; U.S.DHHS, 2004) and other major diseases including heart disease, chronic obstructive pulmonary disease, and stroke (IARC, 1986; IARC, 2002). Tobacco smoke contains more than 4,000 chemicals, including carcinogens and tumorogens. Some of the chemicals produced during cigarette combustion include: aniline, arsenic, formaldehyde, hydrogen cyanide, propylene glycol, toluene, benz(a)pyrene, phenol, cadmium, benzene, etc. (Hoffmann & Hoffmann, 1997). In the U.S. from 1997- 2001, one in five premature deaths (438,000 deaths) each year was attributed to smoking (CDC, 2005; CDC, 2007). Smoking also causes a very heavy burden to society in the form of excess healthcare expenditures, estimated at $96 billion per year in direct medical expenses and $97 billion in lost productivity (CDC, 2007).

10 2.2 Health Effects of Smoking During Pregnancy

Smoking is especially harmful to both the mother and fetus during pregnancy, and has been strongly associated with multiple adverse perinatal outcomes including ectopic pregnancy, spontaneous abortion, preterm delivery, placental previa, placental abruption, premature rupture of membranes, stillbirth, LBW, perinatal and neonatal mortality, cleft lip with or without cleft palate, and sudden infant death syndrome (SIDS) (ACOG, 2005; Andres & Day, 2000; Burguet et al., 2004; CDC, 2002; CDC, 2004; Kyrklund-Blomberg, Granath, & Cnattingius, 2005; Lee & Silver, 2001). More recent studies link smoking during pregnancy to deficits in attention and auditory processing in children (Fried, Watkinson, & Siegel, 1997; Fried, Watkinson, & Gray, 2003; Landgren, Kjellman, & Gillberg, 1998), postnatal infection in preterm neonates (Jeppesen, Nielsen, Ersbøll, & Valerius, 2008), and childhood obesity and overweight (Mendez, Torrent, Ferrer, Ribas- Fitó, & Sunyer, 2008; Widerøe, Vik, Jacobsen, & Bakketeig, 2003). In the long term, studies have shown that babies of pregnant smokers are also more likely to become smokers (Buka, Shenassa, & Niaura, 2003), have increased risk of adolescent-onset drug dependence for girls whose mothers smoked 10 or more cigarettes almost daily during pregnancy (Weissman, Warner, Wickramaratne, & Kandel, 1999), and present with severe antisocial behaviors (Fergusson, Woodward, & Horwood, 1998; Wakschlag et al., 1997; Wakschlag, Pickett, & Leventhal, 2000; Wakschlag & Hans, 2002; Wakschlag, Pickett, Cook Jr., Benowitz, & Leventhal, 2002). The biological mechanism through which smoking impacts fetal development is unclear, but it is believed that nicotine and carbon monoxide (CO) from cigarette

11 combustion alters fetal development and produces specific teratogenic effects on the nervous system and neural development (Benowitz, 1998; Slotkin, 1998). Cyanide from combustion and inhalation contributes to impaired fetal growth and increased fetal morbidity and mortality. CO contributes to reduced oxygen supply to the fetus (fetal hypoxia) by combining with the hemoglobin. Nicotine also reduces blood flow to the fetus by constricting the blood vessels of the umbilical cord and uterus, reducing the amount of blood in the fetal cardiovascular system (U.S.DHHS, 2004). Reduced blood supply causes growth retardation, neurobehavioral deficits in infants, and has been linked to Sudden Infant Death Syndrome (SIDS). Fetal and infant exposure to tobacco smoke also reduces babies’ lung function (U.S.DHHS, 2004). Infants whose mothers smoked before and after birth are at three to four times greater risk for SIDS (U.S.DHHS, 2004). LBW is a leading cause of infant death (Martin et al., 2006; U.S.DHHS, 2004). Smoking during pregnancy resulted in an estimated 776 infant deaths annually during 2000–2004 (CDC, 2008b). Salihu et al. (2004) estimated that smoking is responsible for 30% of small for gestational age babies, 10% of preterm babies and 5% of infant deaths. In the same year, the Surgeon General’s Report noted that if all women who smoke in the U.S. stop smoking during pregnancy, there would be an estimated 11% decrease in stillbirths and 5% reduction in newborn deaths (U.S.DHHS, 2004). Smoking during pregnancy doubles the risk of having a LBW baby. In 2004, 11.9% of LBW babies were born to smokers as compared with only 7.2% for non-smokers (Martin et al., 2006). Although statistically pregnant smokers eat more than pregnant non-smokers do, their babies weigh less than babies of nonsmokers. This weight deficit is smaller if smokers quit early in their pregnancy (U.S.DHHS, 2004). MacArthur and Knox (1988) reported

12 that mothers who quit smoking by 16 weeks of gestation have normal birth weight babies. Rush and Cassano (1983), furthermore, noted that mothers who quit smoking by 28 weeks had a mean birth weight higher than those who smoked during the entire pregnancy, although the difference was not significant. Li and colleagues (1993), found that reducing cotinine by 50% or more improves birth weight of infants by approximately 96 grams. The costs of neonate complications attributed to smoking based on the CDC’s Smoking-Attributable Mortality, Morbidity, and Economic Costs (SAMMEC) for a pregnant smoking prevalence of 10.2% is $353 per smoker and $138 million for Smoking-Attributable Neonatal Expenditures for the U.S. per year (CDC, 2009). Miller and colleagues (2001) reported in 2001 that the sum of additional costs attributable to smoking for all conditions in the first year after birth ranged from $1,142 to $1,358 per smoking pregnant woman. In this study, the cost for various conditions had a wide range from $23,697 for placental abruption to $428 for respiratory infection for the infant. In 2003, Windsor and colleagues adjusted for inflation and estimated that the excess total smoking-attributable healthcare cost at birth and for the first year for major adverse perinatal conditions was $2,500 per smoking pregnant woman.

Full document contains 187 pages
Abstract: The current research study aimed at developing an instrument to measure perceived attributes of using an innovation in a public health system. In this study, the innovation was a behavioral program (Smoking Cessation and Reduction In Pregnancy Treatment -- SCRIPT) to help pregnant mothers quit smoking and the public health system is the West Virginia (WV) Right From the Start Project. WV has a very high smoking rate amongst their pregnant women (27.1%) compared to the national rate (10.2%). The Designated Care Coordinators (DCCs) provide the SCRIPT program to the RFTS clients to help them quit smoking. A measurement tool was developed to measure the perceived attributes of using the SCRIPT program by the DCCs. This instrument (SCRIPT Adoption Survey - SAS) was developed to address the five constructs of Rogers Diffusion of Innovation model: Relative Advantage, Compatibility, Complexity, Observability, and Trialability. The instrument was tested for its validity and reliability. Forty three items were administered to the DCCs. Twenty Eight of these 43 items were found to be reliable and valid. The internal consistency of the full scale was 0.926 and test retest reliability was 0.756 (p<0.0001). The internal consistency of the sub-scales ranged from 0.667 to 0.879. The test-retest reliability of the sub-scales ranged from 0.599 to 0.737. The lowest reliability was found in the Observability domain. All items had a factor loading of greater than 0.4 (0.43-0.81). Model Fit statistics were not conclusive; however, the Akaike Information Criterion (AIC) was smallest in the hypothesized model. Correlation between SAS score and implementation rate measured by Program Implementation Index (PII) was 0.336 (p=0.0028); which will explain 11.29% of variation in the implementation rate. Trialability had the highest correlation of 0.391 (p=0.0004) explaining 15.29% of variance in the data.