- Email: [email protected]
The Integrated Gateway Model: A catalytic approach to behavior change
International Journal of Gynecology and Obstetrics 130 (2015) E62–E68
Contents lists available at ScienceDirect
International Journal of Gynecology and Obstetrics journal homepage: www.elsevier.com/locate/ijgo
SUPPLEMENT ARTICLE
The Integrated Gateway Model: A catalytic approach to behavior change Hilary M. Schwandt a,b,⁎, Joanna Skinner a, Adel Takruri c, Douglas Storey a a b c
Center for Communication Programs, Johns Hopkins Bloomberg School of Public Health, Baltimore, USA Fairhaven College of Interdisciplinary Studies, Western Washington University, Bellingham, USA Department of Population, Family, and Reproductive Health, Johns Hopkins Bloomberg School of Public Health, Baltimore, USA
a r t i c l e
i n f o
Keywords: Behavior change Egypt Gateway behavior Gateway factor Gateway moment Health communication Nigeria Reproductive health
a b s t r a c t Objectives: To develop and test an Integrated Gateway Model of behaviors and factors leading to subsequent positive reproductive, maternal, and child health behaviors. Methods: A secondary analysis was conducted using previously published household survey data collected from men (n = 5551; 2011) and women (n = 16 144; 2011) in Nigeria and women in Egypt (n = 2240; 2004–2007). The number of health behaviors each potential gateway behavior predicted was assessed by multivariate regression, adjusting for potential confounders. The influence of gateway factors on gateway behaviors was tested via interaction terms. Gateway behaviors and factors were ranked by the number of health outcomes predicted, both separately and synergistically. Results: The key gateway behavior identified in both datasets was spousal communication about family planning, whereas the key gateway factor was exposure to family planning messages. Conclusions: The model could facilitate innovative research and programming that in turn might promote cascades of positive behaviors in reproductive, maternal, and child health. © 2015 Published by Elsevier Ireland Ltd. on behalf of International Federation of Gynecology and Obstetrics. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
1. Introduction Low-income countries face a multitude of health problems including malaria, HIV/AIDS, problematic pregnancies and deliveries, neonatal mortality, tuberculosis, and other communicable and chronic diseases. Many of these health problems, if not all, are directly affected by the health behaviors of individuals, families, and communities. To increase the efficiency of behavioral change programs, while ensuring that scarce resources are used effectively, it is advantageous to understand how one behavior might influence future lifestyle choices. Such behaviors are referred to as gateway behaviors. The concept of gateway behaviors has existed since at least the 1970s [1]. The term is most often used to describe the role of tobacco, alcohol, or marijuana use as introductory or ‘gateway’ drugs to other substances, such as heroin. Contextual factors that influence how the process works have gained growing attention; for example, the gateway might not be the actual substance abuse but rather the social interactions that surround it [2]. In the past decade, research has shifted focus to the facilitating power of a positive health behavior or factor that can catalyze other positive health behaviors. The gateway approach draws on the notion that successful change in one behavior creates self-efficacy to make change ⁎ Corresponding author at: Fairhaven College of Interdisciplinary Studies, Western Washington University, 516 High Street MS9118, Bellingham, WA 98229, USA. Tel.: +1 360 650 2948; fax: +1 360 650 3766. E-mail address: [email protected] (H.M. Schwandt).
in other behaviors [3]. To date, most research in this area has been conducted on healthy eating and physical activity in high-income countries [4,5], with few studies exploring the gateway concept in other health areas, especially among low-income countries. 1.1. The Integrated Gateway Model We developed the Integrated Gateway Model shown in Fig. 1 to provide a framework that can be used by researchers and policy makers in the design of effective behavior change programs. In this model, three gateway concepts—behaviors, factors, and moments—are presented as a set of interrelated components. The gateway behavior is an action initiated by the individual, which takes place within a gateway moment. Gateway moments refer to key transitional points in life (e.g. menarche, marriage, or first birth) when individuals or families could be particularly receptive to new information and motivated to make positive health changes. These life transitions might be developmental, situational (such as pregnancy or marriage), health-related, or illness-related (such as learning HIV status) and are characterized by three elements: cognitive (perception of personal risk and outcome expectancies, personal efficacy, readiness to change, and awareness of resources); emotional (prompting of a strong affective response); and self-concept (redefining self-image or social role). During this time, one or more gateway factors operate to influence the gateway behavior. A gateway factor refers to the context, attributes, or conditions that facilitate behavior change and might have a positive or negative influence on downstream behaviors.
http://dx.doi.org/10.1016/j.ijgo.2015.05.003 0020-7292/© 2015 Published by Elsevier Ireland Ltd. on behalf of International Federation of Gynecology and Obstetrics. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
H.M. Schwandt et al. / International Journal of Gynecology and Obstetrics 130 (2015) E62–E68
Gateway Factors
Gateway Moment
Gender
(developmental, situational or health-related)
Exposure to information Community dialogue and action
Gender
Outcome Behaviors
Emotional
Integrated services Health-seeking experience
SelfConcept
Gateway Behavior
Health competence
Fig. 1. The Integrated Gateway Model.
Working together, the three gateway components lead to multiple outcome behaviors. If the outcome behaviors are themselves gateway behaviors, then a chain of behavioral changes is set in motion. Interventions can focus on any or all of these components to influence outcome behaviors. 1.2. Gateway behaviors Four gateway behaviors have been associated with other positive reproductive, maternal, and child health behaviors; namely, prenatal care, immunization, HIV testing, and interpersonal communication. Research has demonstrated strong links between prenatal care and future positive health behaviors, most notably on other service-level behaviors, such as child vaccination, well-child visits, institutional delivery, and trained assistance at delivery [6–8]. By contrast, the effect on individual behaviors is mixed. In the USA, the frequency and timing of initiation of prenatal care showed no, or unclear, effects on breastfeeding [7]. In Nigeria and China, however, women who experienced high intensity and early initiation of prenatal care were more likely to breastfeed their offspring than those women who received less intense and later prenatal care [9,10]. Prenatal care was also an appreciable predictor of contraceptive use in several countries [11–13]. In the USA, strategies to promote immunization have increased the uptake of other preventive health services [14], whereas child immunization provided as part of a composite index has been associated with contraceptive use in low-income countries [11,12]. Testing for HIV can improve interpersonal communication both with sexual partners and within the wider community [15,16]. An effect of HIV testing on consistent condom use has also been reported [15,16]; However, no effect has been found on the number of sexual partners [17]. The relationship between interpersonal communication and many other behaviors is well-documented. In Lao People’s Democratic Republic, spousal communication about breastfeeding was associated with high levels of exclusive breastfeeding at 6 months [18]. Other studies have demonstrated that interpersonal communication reduces violence against women [19], increases condom use [18], and increases uptake of HIV testing [15,16]. The positive effect on contraceptive use has also received considerable attention [20–22]. 1.3. Gateway factors Four gateway factors affect reproductive, maternal, and child health behaviors: health-seeking experience, integrated services, health competence, and community dialogue and action. Health-seeking experience might influence future behaviors. As individuals access one type of health service, they overcome barriers and so might be likely to seek out other services [11,13]. Increased access to information from healthcare providers can also explain this
E63
relationship [12]. However, the content of counseling makes this factor difficult to examine. In some studies, integrated services are thought to explain the gateway effect of prenatal care, maternal and child health care, and HIV testing on family planning service use [11–13]. A health competent society is one in which “individuals, communities, and institutions have the knowledge, attitudes, skills and resources needed to improve and maintain health” [23]. At the individual level, health competence is a gateway factor associated with multiple family health behaviors in Egypt. In South Africa, individuals with high health competence were more likely to have undergone an HIV test and used a condom at their last sexual encounter than those with low health competence [23]. Community dialogue and action around a specific issue, such as neonatal and maternal health, can produce changes in multiple behaviors [24]. For example, engaging Indian communities in shifting gender norms and promoting “evidence-based decision making” within families led to change in broad social structures, such as collective selfefficacy, social norms, and leadership [25]. 1.4. Gateway moments During life transitions, there is the potential for impacts on multidimensional health outcomes [26]; these “teachable moments” motivate individuals to spontaneously adopt risk-reducing health behaviors [27]. Research on teachable moments has predominantly considered how an individual’s immediate situation creates susceptibility or openness to persuasive arguments. Nevertheless, no studies have explicitly explored how such a moment creates opportunities to think about the future and to consider not only an immediate behavioral choice but a long-term course of action, a sequence of protective behaviors, even a change in life course. 1.5. Aim The aim of the present study was to evaluate the Integrated Gateway Model, with an emphasis on identifying and testing gateway behaviors and factors among men and women living in low-income countries. 2. Materials and methods An exploratory secondary analysis of the gateway concept was conducted using existing data from two household surveys. The datasets used in the present study were the Nigerian Urban Reproductive Health Initiative (NURHI) Measurement, Learning and Evaluation baseline data collected in 2011 [28] and the Egyptian longitudinal Minya Village Health Surveys (MVHS) data (2004–2007) [29,30]. The Egyptian data were included in the analysis as they provided the opportunity to examine the effects of health behaviors on downstream health outcomes among a panel of women. The Nigerian data were selected because future operations research on the gateway concept was already planned in this setting. 2.1. Data Collection The NURHI baseline household survey was conducted in six Nigerian cities (Abuja, Benin City, Ibadan, Ilorin, Kaduna, and Zaria) between October 4, 2010, and April 15, 2011 [28]. The survey focused primarily on reproductive health. Interviews were conducted among all women aged 15–49 years who were living in selected households in all six cities and with men aged 15–59 years who were resident in half of the selected households in four cities (Abuja, Ibadan, Ilorin, and Kaduna). A twostage sampling design was used. In the first stage, a random sample of clusters was selected for each city based on probability proportional to their population. In the second stage, 41 households were selected in each cluster to create a sample of approximately 3000 households per city. A total of 19 556 households were selected and interviews were successfully conducted at 16 935 of them, yielding a completion rate of 87%. Among the interviewed households, 16 957 women were eligible for individual interviews; 95% of them were successfully completed.
E64
H.M. Schwandt et al. / International Journal of Gynecology and Obstetrics 130 (2015) E62–E68
A total of 5962 men were also eligible for an individual interview, of which 93% were successfully completed. Study procedures, consent forms, and questionnaires were approved by both the Nigerian Health Research Ethics Committee, Abuja, Nigeria, and the institutional review board of the University of North Carolina at Chapel Hill, USA. The MVHS on reproductive health, maternal and child health, and infectious disease were conducted in the Minya governorate of Upper Egypt in July and August of 2004 and again in November and December of 2007 using the same multistage random sampling technique and sampling frame used by the national Demographic and Health Surveys in 2003 and 2005 [31,32]. In the first stage, seven villages were selected and the number of households and the total population of each village estimated. In the second stage, each village was divided into 10 − 13 segments of approximately 1000 households each. A list of the households in each segment was then prepared. In the third stage, 35 households were selected from each segment by systematic random sampling. Within each household, ever-married women aged 15–49 years were eligible to participate in the survey. A total of 2241 women were sampled, of whom 2240 were interviewed, representing a 99.9% response rate. Individuals from the original 2004 sample were re-interviewed in 2005 and 2007 using the same questionnaires. Study procedures, consent forms, and questionnaires were approved by the Human Research Protection Program, Tulane University, New Orleans, USA, and the institutional review board, Johns Hopkins Bloomberg School of Public Health, Baltimore, USA. 2.2. Data analysis Multivariate regression analyses were conducted with the NURHI data to examine the association between each potential gateway behavior and all measured health outcomes, accounting for the complex sample design and controlling for age, education, marital status, parity, religion, wealth, and city. Next, the effect of each gateway factor, as well as the interaction between the gateway factor and gateway behavior, were tested to discern whether the synergistic effects on each outcome were greater than the sum of each individually. All analyses were executed in Stata/SE version 12 (Stata, College Station, TX, USA) with 0.05 as the cut-off P value for significance. Multivariate regression analyses examining the association between each potential gateway behavior with all measured health outcomes were conducted using the MVHS data, controlling for age, education, parity, and working status. The gateway behaviors and factors were then ranked by the number of outcomes each predicted. All analyses were conducted using Stata/SE version 12 (Stata, College Station, TX, USA) with 0.05 as the cut-off P value for significance. 3. Results 3.1. Nigerian Urban Reproductive Health Initiative, 2011 All potential gateway behaviors were tagged according to the recognized definition and available data (Box 1). The gateway behaviors identified were then ranked by the average number of health outcomes each behavior predicted. Statistically significant associations with the largest number of positive health outcomes for men and women were: (1) spousal communication about family planning (number of outcomes significantly predicted in women’s dataset: 14; men’s: 7); (2) recommended family planning to family or friends (women: 10; men: 7); (3) ever use of modern family planning (women: 10; men: 6); (4) currently using modern family planning (women: 8; men: 5); and (5) partner had accompanied wife or child to the health facility (women: 10; men: 5) (Tables 1 and 2). From the ranked list, the top five gateway behaviors for women and men were then assessed for effect modification with each potential gateway factor (Box 1). The main gateway factors for both sexes were: (1) knowledge of modern contraceptive methods; (2) exposure
Box 1 Potential gateway behaviors and factors identified and included in the analysis of women’s and men’s Nigerian Urban Reproductive Health Initiative baseline household survey data, 2011. Potential gateway behaviors (men and women) • • • • •
Ever use of modern family planning Current use of modern family planning Family planning use at first ever sexual intercourse Spousal communication about family planning Ever recommended family planning to family or friends (condoms only for males) • Joint decision making for family planning • Partner has accompanied wife or child to a health facility Additional potential gateway behaviors (women only) • • • • • • •
Joint decision making about size of family Prenatal care visit (among currently pregnant women) Medically assisted delivery Delivery in a health facility Child health visit Maternal health visit HIV testing
Potential gateway factors (men and women) • • • • • • • • • • • • • •
Knowledge of modern contraceptive methods Knowledge of the fertile period Perceptions about community family planning use and norms Attitudes toward condom use Membership in a social group/club/organization Family planning discussion, use, or encouragement of use among social network members Attitudes in favor of males prohibiting their female partner’s mobility Tolerance of domestic violence Gender equitable attitudes towards household decisionmaking Gender equitable attitudes toward family planning use Aligned fertility preferences Exposure to family planning in the media (print, radio, or television) Exposure to leaders speaking publically about family planning Access to contraceptives
Additional potential gateway factors (women only) • • • •
Attitudes toward covert family planning use Gender equitable attitudes about finances Quality of family planning service interactions Family planning service integration
to leaders speaking publicly about family planning; (3) aligned fertility preferences; and (4) perceptions about community family planning norms.
3.2. Minya Village Health Surveys, 2004–2007 All potential gateway behaviors in 2004 were identified and examined through bivariate associations with outcome behaviors in 2007 (Box 2). The ranked gateway behaviors that predicted positive health outcomes were: (1) the mother breastfed within 24 hours of delivery; (2) spousal communication about family planning; (3) the mother checked her health with a healthcare provider after the most recent delivery; and (4) the most recent neonate was delivered in a health
H.M. Schwandt et al. / International Journal of Gynecology and Obstetrics 130 (2015) E62–E68
E65
Table 1 Multivariate logistic regression analysis of the four key potential gateway behaviors identified among women in the Nigerian Urban Reproductive Health Initiative baseline survey, 2011.a,b Behavioral outcomes
Family planning use Ever use of modern family planning Current use of modern family planning Long-acting and permanent method use (LAPM) Ever use of Emergency Contraception (EC) Use of family planning at first ever sex Family planning communication Spousal communication about family planning Recommended family planning to family and friends Joint family planning decision-making process Joint family size decision-making process Maternal and child health Prenatal care visit Assisted delivery Delivery in a health facility Child health visit Maternal health visit Other Partner accompanies wife or child to health facility HIV test
Spousal communication about family planning
Ever use of family planning
Recommended family planning to family and friends
Current use of family planning
Partner accompanies wife and/or child to the health facility
7.1e 6.8e 2.6e NS 1.8e
NA NA NA NA 5.3e
8.3e 4.0e 1.4c 1.9e 1.3c
NA NA NA 2.6e 2.0e
1.3e 1.2c NS NS NS
NA 3.9e 6.2e 1.8e
7.0e 8.4e 6.6e 1.2c
3.9e NA 2.4e 1.2c
6.9e 4.0e 18.1e 1.2d
1.7e NS 1.3d 1.3e
1.5c 1.5e 1.4d 1.3e 1.4e
NS 1.7e 1.4d NS 1.2d
NS NS NS 1.3d NS
NA 1.4d NS OD OD
NS 1.5e 1.7e 1.5e 1.4e
1.7e 1.4e
1.3e 1.6e
NS 1.8e
1.2c NS
NA 1.4e
Abbreviations: NA, not applicable; NS, not significant, OD, significant finding in the negative direction. a Values given as Odds ratios. b Multivariate logistic regression analysis with 0.05 as the cut-off P value for significance. All analyses were controlled for age, education, marital status, parity, residence, and wealth (not shown). c P ≤ 0.05. d P ≤ 0.01. e P ≤ 0.001.
facility, which predicted eight, six, six and five behavioral outcomes, respectively (Table 3). The influence of each potential gateway factor was tested by introducing them as interaction terms in the regression models (Box 2). This analysis indicated that gateway factors positively influenced gateway behaviors. For example, discussing family planning with a spouse in 2004 was related to numerous positive health behaviors in 2007. Women who discussed family planning with their spouse in 2004 and who received family planning advice from a health worker exhibited increased likelihood of family planning use in 2007 than women who discussed family planning with their spouse in 2004 but did not receive family planning advice from a health worker. This observation indicated a synergistic effect of spousal and health worker communication on subsequent family planning use.
The ranked gateway factors were: (1) comprehensive content of the prenatal care visit (as defined in Box 2); (2) a health worker gave the woman family planning advice; (3) a health worker gave the woman breastfeeding advice; (4) the mother received supplemental vitamin A in the first 2 months after pregnancy; and (5) the mother was exposed to family planning messages. 4. Discussion The results of the present study supported the Integrated Gateway Model. Positive behaviors either preceded, or were associated with, a cascade of downstream positive outcomes. Some of the relationships were directly related to the original gateway behavior, whereas others seemed unrelated. For example, spousal communication about family planning
Table 2 Multivariate logistic regression analysis of the four key potential gateway behaviors identified among men in the Nigerian Urban Reproductive Health Initiative baseline survey, 2011.a,b Behavioral outcomes
Family planning use Ever use of modern family planning Current use of modern family planning Long-acting and permanent method use (LAPM) Family planning use at first ever sex Family planning communication Spousal communication Recommended family planning to family and friends Joint family planning decision-making process Other Accompanied wife and/or child to health facility
Spousal communication about family planning
Recommended condoms to family and friends
Ever use of family planning
Current use of family planning
Accompanied wife and/or child to health facility
8.0 e 7.4 e 7.0 e 4.1 e
17.4 e 4.9 e 1.9 c 1.4 c
NA 13.0 e NA 7.0 e
12.3 e NA NA 1.7 e
2.4 e 1.8 e NS NS
NA 3.8 e 6.3 e
3.8 e NA 2.7 e
8.1 e 18.7 e 6.6 e
7.4 e 4.9 e NA
2.4 e 2.5 e 1.5 d
2.3 e
2.5 e
2.4 e
1.8 e
NA
Abbreviations: NA, not applicable; NS, not significant. a Values given as Odds ratios. b Multivariate logistic regression analysis with 0.05 as the cut-off P value for significance. All analyses were controlled for age, education, marital status, parity, residence, and wealth (not shown). c P ≤ 0.05. d P ≤ 0.01. e P ≤ 0.001.
E66
H.M. Schwandt et al. / International Journal of Gynecology and Obstetrics 130 (2015) E62–E68
Box 2 Potential gateway behaviors and factors identified and included in the analysis of women’s Egyptian longitudinal Minya Village Health Surveys data, 2004–2007. Family planning related behaviors • Spousal communication about family planning • Talked to anyone about birth spacing • Current use of family planning Prenatal care behaviors • Number of visits • Timing of first visit • Took iron tablets during pregnancy Delivery • Medically assisted delivery • Delivered in a health facility Postnatal care behaviors • • • • • • •
Woman visited a health worker during the postnatal period Timing of postnatal visit Postnatal visit in a medical facility Mother took postnatal supplementary vitamin A Timing of child health check up Child health checkup in a medical facility Duration of postpartum abstinence
Breastfeeding • Initiated within 24 h of delivery • Only breast milk given to the newborn; no other fluids given during the first 24 h • Mother breastfed for at least 6 mo Hygienic behaviors • Mother made sure children washed their hands every morning, before and after meals, and after using the bathroom • Mother washed her hands every morning, before and after meals, and after using the bathroom • Daughters were not circumcised • Never reused syringes • Hygienic precautions taken when handling birds Composite indices • • • • •
Prenatal care index Postnatal care index Breastfeeding index Avian influenza index Overall index of all positive health behaviors
Gateway Factors • Media exposure to family planning messages • Content of the prenatal visit regarding the mother: ○ Given a health card ○ Height, weight, and blood pressure measured ○ Urine and blood samples collected • Intention to use family planning • Intention to circumcise daughters • Mother received supplemental vitamin A in the first 2 months after pregnancy • Tetanus immunization given during pregnancy • A health worker offered the woman family planning advice • A health worker gave woman breast feeding advice • Mother was exposed to family planning messages
was associated not only with the use of contraceptives but also with behaviors such as hand washing, early initiation of breastfeeding (MVHS), and undergoing HIV testing (NURHI). Health workers appeared to play a substantial role in many gateway behaviors and factors. For example, delivery at a health facility; prenatal and postpartum visits with a health worker; content of prenatal care; and health worker advice on family planning and breastfeeding were all gateway behaviors and factors in the MVHS dataset. These findings indicated that services offered by healthcare providers had a catalytic effect on many gateway behaviors. Prenatal care, immunization, and other types of contact with health workers possibly operate as gateway factors. That is, by accessing one type of health service, opportunities to improve other service-related behaviors emerge, either through the breakdown of user barriers to care or through direct provider referrals to other health services. This situation would have implications for the impact of some types of service use as a gateway behavior depending on the mode of delivery. For example, whether vaccination is received in a clinic or during a national immunization day could make a difference, as the latter approach might not provide the opportunity for appreciable contact with a health facility or a healthcare professional. Some behaviors and factors that do not necessitate direct contact with healthcare providers emerged as potential gateways in the present study. For example, interpersonal communication within social networks about family planning—communicating with a spouse about family planning (NURHI and MVHS) or recommending family planning to friends or family (NURHI)—were identified as promising gateways for numerous future positive behaviors. These behaviors included family planning use, joint decision-making, prenatal care, and HIV testing. Interpersonal communication is generally a powerful influence on behavior because it (1) extends the reach of interventions through social networks and (2) moderates or mediates most behavioral decisions through social influence and the self-reflection it encourages. The present study had several strengths. The NURHI sample size was large and representative of urban Nigeria; data were available for both men and women of reproductive age; and it was possible to explore issues related to family planning in more detail than is often possible. The longitudinal design of the MVHS was a major strength as the same women were followed over a period of 3 years. The MVHS also included a variety of family health behaviors as well as reproductive behaviors. The present study also had some limitations. Although populationbased, the MVHS data were drawn from only one of the 23 Egyptian governorates. Furthermore, the selected governorate was in a predominantly rural area. The main limitation of the NURHI was that the data were collected at a single time point and it was not possible to empirically determine causal order. In addition, the data were focused on topics related to family planning; therefore, the ability to examine the potential of gateway behaviors to predict behaviors outside of reproductive health was limited. Unfortunately, it was not possible to assess the effect of gateway moments on health behaviors using the two datasets. In addition, the synergistic effects of gateway moments, behaviors, and factors, could not be determined. Nevertheless, the findings of the present analysis suggested that it would be useful to test whether the observed behaviors are more likely to occur around a gateway moment, such as first pregnancy. Furthermore, it can be hypothesized that gateway behaviors and gateway factors—when they coincide at a gateway moment—would have more influence on subsequent behaviors than if they occurred outside the context of a gateway moment. The identified gateway behaviors and factors suggest that behavioral change programs should focus on intervening either before women become pregnant or early in pregnancy. Such an approach might result in maximum impact on future health for both women and for their offspring. In conclusion, the results of the present study could help to guide investment decisions and improve programmatic impact across a wide range of family health initiatives. Indeed, the Integrated Gateway
H.M. Schwandt et al. / International Journal of Gynecology and Obstetrics 130 (2015) E62–E68
E67
Table 3 Multivariate regression analysis of the four key potential gateway behaviors identified among women in the longitudinal Minya Village Health Survey, Egypt, 2004 and 2007.a,b 2007 Behavioral outcomes
Spousal family planning communication
Breastfeeding within 24 h of delivery
Postnatal health worker visit
Delivery at a health facility
Current use of modern family planning Spousal communication about family planning Talked about spacing pregnancies Delivery at a health facility Postnatal care via health worker visit Postnatal care at a health facility Initiated breastfeeding within 24 h of delivery Neonate not given anything other than breast milk within 24 h of birth Hand washing of the mother Overall indices Behavioral healthc Family planningc Prenatal carec
2.10 f 1.44 f NS NS NS NS 2.10 f NS 1.34 e
1.83 f 1.72 f 1.57 d NS NS NS 2.37 f 1.76 e NS
NS NS NS 2.50 f 1.50 e NS NS NS 2.10 d
NS NS 1.39 d 5.10 f NS 2.40 d NS NS 2.25 d
0.03 f 0.06 f NS
0.04 f 0.09 f 0.12 e
0.04 f 0.04 e 0.09 d
NS NS 0.10 d
Abbreviation: NS, not significant. a Values given as Odds ratios and Regression Coefficients. b Multivariate logistic and linear regression with 0.05 as the cut-off P value for significance. All analyses were controlled for age, education, parity, and employment (not shown). c Continuous predictors d P ≤ 0.05. e P ≤ 0.01. f P ≤ 0.001.
Model is currently being tested in a 3-year operations research project in Nigeria using a field experimental design. The model, therefore, represents a promising avenue for innovative research and programming to identify strategic and cost-effective interventions for public health that promote a catalytic and long-lasting effect on behavior change.
[9] [10] [11]
Acknowledgments The Bill & Melinda Gates Foundation funded the Nigeria Urban Reproductive Health Initiative (NURHI) baseline survey through the Measurement, Learning & Evaluation (MLE) project at the University of North Carolina, as well as all the secondary data analysis presented in this paper. The Minya Village Health Surveys were funded by the US Agency for International Development through the Health Communication Partnership (HCP) project. This analysis, NURHI, and HCP projects were all led by the Johns Hopkins Center for Communication Programs at the Bloomberg School of Public Health. The views expressed in this paper do not necessarily reflect those of the funding bodies.
[12]
[13]
[14]
[15] [16]
Conflict of interest
[17]
The authors declare no conflict of interest regarding the material discussed in the present manuscript.
[18]
References
[20]
[1] Sheppard CW, Gay GR, Smith DE. The changing patterns of heroin addiction in the Haight-Ashbury subculture. J Psychedelic Drugs 1971;3(2):22–31. [2] Ginzler JA, Cochran BN, Domenech-Rodríguez M, Cauce AM, Whitbeck LB. Sequential progression of substance use among homeless youth: an empirical investigation of the gateway theory. Subst Use Misuse 2003;38(3–6):725–58. [3] Prochaska JJ, Spring B, Nigg CR. Multiple health behavior change research: an introduction and overview. Prev Med 2008;46(3):181–8. [4] Dutton GR, Napolitano MA, Whiteley JA, Marcus BH. Is physical activity a gateway behavior for diet? Findings from a physical activity trial. Prev Med 2008;46(3): 216–21. [5] Tucker M, Reicks M. Exercise as a gateway behavior for healthful eating among older adults: an exploratory study. J Nutr Educ Behav 2002;34(Suppl. 1):S14–9. [6] Stevens-Simon C, Kelly LS, Singer D. Pattern of prenatal care and infant immunization status in a comprehensive adolescent-oriented maternity program. Arch Pediatr Adolesc Med 1996;150(8):829–33. [7] Reichman NE, Corman H, Noonan K, Schwartz-Soicher O. Effects of prenatal care on maternal postpartum behaviors. Rev Econ Househ 2010;8(2):171–97. [8] Sugathan, KS, Mishra V, Retherford RD. Promoting Institutional Deliveries in Rural India: the Role of Antenatal-Care Services. National Family Health Survey Subject
[19]
[21]
[22] [23]
[24]
[25]
[26]
Reports Number 20. http://www.eastwestcenter.org/sites/default/files/filemanager/ Research_Program/NFHS_Subject_Reports/subj-20.pdf. Published 2001. Accessed April 27, 2015. Agho KE, Dibley MJ, Odiase JI, Ogbonmwan SM. Determinants of exclusive breastfeeding in Nigeria. BMC Pregnancy Childbirth 2011;11:2. Nwaru BI, Wu Z, Hemminki E. Infant care practices in rural China and their relation to prenatal care utilisation. Glob Public Health 2011;6(1):1–14. Ahmed S, Mosley WH. Simultaneity in the use of maternal-child health care and contraceptives: evidence from developing countries. Demography 2002;39(1): 75–93. Hotchkiss DR, Rous JJ, Seiber EE, Berruti AA. Is maternal and child health service use a causal gateway to subsequent contraceptive use?: A multi-country study. Popul Res Policy Rev 2005;24(6):543–71. Zerai A, Tsui AO. The relationship between prenatal care and subsequent modern contraceptive use in Bolivia, Egypt and Thailand. Afr J Reprod Health 2001;5(2): 68–82. Shefer AM, Fritchley J, Stevenson J, Lyons B, Friedman R, Hopfensperger D, et al. Linking WIC and immunization services to improve preventive health care among low-income children in WIC. J Public Health Manag Pract 2002;8(2):56–65. van der Straten A, King R, Grinstead O, Serufilira A, Allen S. Couple communication, sexual coercion and HIV risk reduction in Kigali, Rwanda. AIDS 1995;9(8):935–44. Desgrées-Du-Loû A, Brou H, Djohan G, Becquet R, Ekouevi DK, Zanou B, et al. Beneficial effects of offering prenatal HIV counselling and testing on developing a HIV preventive attitude among couples. Abidjan, 2002–2005. AIDS Behav 2009;13(2): 348–55. Denison JA, O'Reilly KR, Schmid GP, Kennedy CE, Sweat MD. HIV voluntary counseling and testing and behavioral risk reduction in developing countries: a metaanalysis, 1990–2005. AIDS Behav 2008;12(3):363–73. Putthakeo P, Ali M, Ito C, Vilayhong P, Kuroiwa C. Factors influencing breastfeeding in children less than 2 years of age in Lao PDR. J Paediatr Child Health 2009;45(9):487–92. Naved RT, Persson LA. Factors associated with physical spousal abuse of women during pregnancy in Bangladesh. Int Fam Plan Perspect 2008;34(2):71–8. Storey D, Boulay M, Karki Y, Heckert K, Karmacharya DM. Impact of the integrated Radio Communication Project in Nepal, 1994–1997. J Health Commun 1999;4(4):271–94. Vaughan PW, Rogers EM. A staged model of communication effects: evidence from an entertainment-education radio soap opera in Tanzania. J Health Commun 2000; 5(3):203–27. Bawah AA. Spousal communication and family planning behavior in Navrongo: a longitudinal assessment. Stud Fam Plann 2002;33(2):185–94. Storey JD, Kaggwa E, Harbour C. Research Brief: Pathways to health competence for sustainable health improvement: Examples from South Africa and Egypt. Center for Communication Programs Johns Hopkins Bloomberg School of Public Health. http://ccp.jhu.edu/documents/StoreyKaggwaHarbour.pdf. Published May 2008. Accessed May, 2012. Harkins T, Drasbek C, Arroyo J, McQuestion M. The health benefits of social mobilization: experiences with community-based Integrated Management of Childhood Illness in Chao, Peru and San Luis, Honduras. Promot Educ 2008;15(2):15–20. Kumar V, Kumar A, Das V, Srivastava NM, Baqui AH, Santosham M, et al. Communitydriven impact of a newborn-focused behavioral intervention on maternal health in Shivgarh, India. Int J Gynecol Obstet 2012;117(1):48–55. Kaiser MM, Kaiser KL, Barry TL. Health effects of life transitions for women and children: a research model for public and community health nursing. Public Health Nurs 2009;26(4):370–9.
E68
H.M. Schwandt et al. / International Journal of Gynecology and Obstetrics 130 (2015) E62–E68
[27] McBride CM, Emmons KM, Lipkus IM. Understanding the potential of teachable moments: the case of smoking cessation. Health Educ Res 2003;18(2): 156–70. [28] Measurement, Learning and Evaluation (MLE) Project. 2010-2011 Baseline Household Survey Report for the Nigerian Urban Reproductive Health Initiative. https://www. urbanreproductivehealth.org/sites/mle/files/nurhi_baseline_household_survey.pdf. Published 2012. Accessed April 27, 2015. [29] El-Zanaty F, El-Ghazaly N, Meekers D, Armanious D. 2004. Menya village health survey. [Unpublished report]. Cairo: Communication for Healthy Living Project.
[30] El-Zanaty F, Shehata M, El-Said M, Storey D, Meekers D. 2008. Menya village health survey. [Unpublished report]. Cairo: Communication for Healthy Living Project. [31] El-Zanaty F, Way A. Egypt Interim Demographic and Health Survey. Cairo Egypt: Ministry of Health and Population [Egypt], National Population Council, El-Zanaty and Associates, and ORC Macro; 2004. [32] El-Zanaty F, Way A. Egypt Demographic and Health Survey 2005. Cairo, Egypt: Ministry of Health and Population, National Population Council, El-Zanaty and Associates, and ORC Macro; 2006.
Contents lists available at ScienceDirect
International Journal of Gynecology and Obstetrics journal homepage: www.elsevier.com/locate/ijgo
SUPPLEMENT ARTICLE
The Integrated Gateway Model: A catalytic approach to behavior change Hilary M. Schwandt a,b,⁎, Joanna Skinner a, Adel Takruri c, Douglas Storey a a b c
Center for Communication Programs, Johns Hopkins Bloomberg School of Public Health, Baltimore, USA Fairhaven College of Interdisciplinary Studies, Western Washington University, Bellingham, USA Department of Population, Family, and Reproductive Health, Johns Hopkins Bloomberg School of Public Health, Baltimore, USA
a r t i c l e
i n f o
Keywords: Behavior change Egypt Gateway behavior Gateway factor Gateway moment Health communication Nigeria Reproductive health
a b s t r a c t Objectives: To develop and test an Integrated Gateway Model of behaviors and factors leading to subsequent positive reproductive, maternal, and child health behaviors. Methods: A secondary analysis was conducted using previously published household survey data collected from men (n = 5551; 2011) and women (n = 16 144; 2011) in Nigeria and women in Egypt (n = 2240; 2004–2007). The number of health behaviors each potential gateway behavior predicted was assessed by multivariate regression, adjusting for potential confounders. The influence of gateway factors on gateway behaviors was tested via interaction terms. Gateway behaviors and factors were ranked by the number of health outcomes predicted, both separately and synergistically. Results: The key gateway behavior identified in both datasets was spousal communication about family planning, whereas the key gateway factor was exposure to family planning messages. Conclusions: The model could facilitate innovative research and programming that in turn might promote cascades of positive behaviors in reproductive, maternal, and child health. © 2015 Published by Elsevier Ireland Ltd. on behalf of International Federation of Gynecology and Obstetrics. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
1. Introduction Low-income countries face a multitude of health problems including malaria, HIV/AIDS, problematic pregnancies and deliveries, neonatal mortality, tuberculosis, and other communicable and chronic diseases. Many of these health problems, if not all, are directly affected by the health behaviors of individuals, families, and communities. To increase the efficiency of behavioral change programs, while ensuring that scarce resources are used effectively, it is advantageous to understand how one behavior might influence future lifestyle choices. Such behaviors are referred to as gateway behaviors. The concept of gateway behaviors has existed since at least the 1970s [1]. The term is most often used to describe the role of tobacco, alcohol, or marijuana use as introductory or ‘gateway’ drugs to other substances, such as heroin. Contextual factors that influence how the process works have gained growing attention; for example, the gateway might not be the actual substance abuse but rather the social interactions that surround it [2]. In the past decade, research has shifted focus to the facilitating power of a positive health behavior or factor that can catalyze other positive health behaviors. The gateway approach draws on the notion that successful change in one behavior creates self-efficacy to make change ⁎ Corresponding author at: Fairhaven College of Interdisciplinary Studies, Western Washington University, 516 High Street MS9118, Bellingham, WA 98229, USA. Tel.: +1 360 650 2948; fax: +1 360 650 3766. E-mail address: [email protected] (H.M. Schwandt).
in other behaviors [3]. To date, most research in this area has been conducted on healthy eating and physical activity in high-income countries [4,5], with few studies exploring the gateway concept in other health areas, especially among low-income countries. 1.1. The Integrated Gateway Model We developed the Integrated Gateway Model shown in Fig. 1 to provide a framework that can be used by researchers and policy makers in the design of effective behavior change programs. In this model, three gateway concepts—behaviors, factors, and moments—are presented as a set of interrelated components. The gateway behavior is an action initiated by the individual, which takes place within a gateway moment. Gateway moments refer to key transitional points in life (e.g. menarche, marriage, or first birth) when individuals or families could be particularly receptive to new information and motivated to make positive health changes. These life transitions might be developmental, situational (such as pregnancy or marriage), health-related, or illness-related (such as learning HIV status) and are characterized by three elements: cognitive (perception of personal risk and outcome expectancies, personal efficacy, readiness to change, and awareness of resources); emotional (prompting of a strong affective response); and self-concept (redefining self-image or social role). During this time, one or more gateway factors operate to influence the gateway behavior. A gateway factor refers to the context, attributes, or conditions that facilitate behavior change and might have a positive or negative influence on downstream behaviors.
http://dx.doi.org/10.1016/j.ijgo.2015.05.003 0020-7292/© 2015 Published by Elsevier Ireland Ltd. on behalf of International Federation of Gynecology and Obstetrics. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
H.M. Schwandt et al. / International Journal of Gynecology and Obstetrics 130 (2015) E62–E68
Gateway Factors
Gateway Moment
Gender
(developmental, situational or health-related)
Exposure to information Community dialogue and action
Gender
Outcome Behaviors
Emotional
Integrated services Health-seeking experience
SelfConcept
Gateway Behavior
Health competence
Fig. 1. The Integrated Gateway Model.
Working together, the three gateway components lead to multiple outcome behaviors. If the outcome behaviors are themselves gateway behaviors, then a chain of behavioral changes is set in motion. Interventions can focus on any or all of these components to influence outcome behaviors. 1.2. Gateway behaviors Four gateway behaviors have been associated with other positive reproductive, maternal, and child health behaviors; namely, prenatal care, immunization, HIV testing, and interpersonal communication. Research has demonstrated strong links between prenatal care and future positive health behaviors, most notably on other service-level behaviors, such as child vaccination, well-child visits, institutional delivery, and trained assistance at delivery [6–8]. By contrast, the effect on individual behaviors is mixed. In the USA, the frequency and timing of initiation of prenatal care showed no, or unclear, effects on breastfeeding [7]. In Nigeria and China, however, women who experienced high intensity and early initiation of prenatal care were more likely to breastfeed their offspring than those women who received less intense and later prenatal care [9,10]. Prenatal care was also an appreciable predictor of contraceptive use in several countries [11–13]. In the USA, strategies to promote immunization have increased the uptake of other preventive health services [14], whereas child immunization provided as part of a composite index has been associated with contraceptive use in low-income countries [11,12]. Testing for HIV can improve interpersonal communication both with sexual partners and within the wider community [15,16]. An effect of HIV testing on consistent condom use has also been reported [15,16]; However, no effect has been found on the number of sexual partners [17]. The relationship between interpersonal communication and many other behaviors is well-documented. In Lao People’s Democratic Republic, spousal communication about breastfeeding was associated with high levels of exclusive breastfeeding at 6 months [18]. Other studies have demonstrated that interpersonal communication reduces violence against women [19], increases condom use [18], and increases uptake of HIV testing [15,16]. The positive effect on contraceptive use has also received considerable attention [20–22]. 1.3. Gateway factors Four gateway factors affect reproductive, maternal, and child health behaviors: health-seeking experience, integrated services, health competence, and community dialogue and action. Health-seeking experience might influence future behaviors. As individuals access one type of health service, they overcome barriers and so might be likely to seek out other services [11,13]. Increased access to information from healthcare providers can also explain this
E63
relationship [12]. However, the content of counseling makes this factor difficult to examine. In some studies, integrated services are thought to explain the gateway effect of prenatal care, maternal and child health care, and HIV testing on family planning service use [11–13]. A health competent society is one in which “individuals, communities, and institutions have the knowledge, attitudes, skills and resources needed to improve and maintain health” [23]. At the individual level, health competence is a gateway factor associated with multiple family health behaviors in Egypt. In South Africa, individuals with high health competence were more likely to have undergone an HIV test and used a condom at their last sexual encounter than those with low health competence [23]. Community dialogue and action around a specific issue, such as neonatal and maternal health, can produce changes in multiple behaviors [24]. For example, engaging Indian communities in shifting gender norms and promoting “evidence-based decision making” within families led to change in broad social structures, such as collective selfefficacy, social norms, and leadership [25]. 1.4. Gateway moments During life transitions, there is the potential for impacts on multidimensional health outcomes [26]; these “teachable moments” motivate individuals to spontaneously adopt risk-reducing health behaviors [27]. Research on teachable moments has predominantly considered how an individual’s immediate situation creates susceptibility or openness to persuasive arguments. Nevertheless, no studies have explicitly explored how such a moment creates opportunities to think about the future and to consider not only an immediate behavioral choice but a long-term course of action, a sequence of protective behaviors, even a change in life course. 1.5. Aim The aim of the present study was to evaluate the Integrated Gateway Model, with an emphasis on identifying and testing gateway behaviors and factors among men and women living in low-income countries. 2. Materials and methods An exploratory secondary analysis of the gateway concept was conducted using existing data from two household surveys. The datasets used in the present study were the Nigerian Urban Reproductive Health Initiative (NURHI) Measurement, Learning and Evaluation baseline data collected in 2011 [28] and the Egyptian longitudinal Minya Village Health Surveys (MVHS) data (2004–2007) [29,30]. The Egyptian data were included in the analysis as they provided the opportunity to examine the effects of health behaviors on downstream health outcomes among a panel of women. The Nigerian data were selected because future operations research on the gateway concept was already planned in this setting. 2.1. Data Collection The NURHI baseline household survey was conducted in six Nigerian cities (Abuja, Benin City, Ibadan, Ilorin, Kaduna, and Zaria) between October 4, 2010, and April 15, 2011 [28]. The survey focused primarily on reproductive health. Interviews were conducted among all women aged 15–49 years who were living in selected households in all six cities and with men aged 15–59 years who were resident in half of the selected households in four cities (Abuja, Ibadan, Ilorin, and Kaduna). A twostage sampling design was used. In the first stage, a random sample of clusters was selected for each city based on probability proportional to their population. In the second stage, 41 households were selected in each cluster to create a sample of approximately 3000 households per city. A total of 19 556 households were selected and interviews were successfully conducted at 16 935 of them, yielding a completion rate of 87%. Among the interviewed households, 16 957 women were eligible for individual interviews; 95% of them were successfully completed.
E64
H.M. Schwandt et al. / International Journal of Gynecology and Obstetrics 130 (2015) E62–E68
A total of 5962 men were also eligible for an individual interview, of which 93% were successfully completed. Study procedures, consent forms, and questionnaires were approved by both the Nigerian Health Research Ethics Committee, Abuja, Nigeria, and the institutional review board of the University of North Carolina at Chapel Hill, USA. The MVHS on reproductive health, maternal and child health, and infectious disease were conducted in the Minya governorate of Upper Egypt in July and August of 2004 and again in November and December of 2007 using the same multistage random sampling technique and sampling frame used by the national Demographic and Health Surveys in 2003 and 2005 [31,32]. In the first stage, seven villages were selected and the number of households and the total population of each village estimated. In the second stage, each village was divided into 10 − 13 segments of approximately 1000 households each. A list of the households in each segment was then prepared. In the third stage, 35 households were selected from each segment by systematic random sampling. Within each household, ever-married women aged 15–49 years were eligible to participate in the survey. A total of 2241 women were sampled, of whom 2240 were interviewed, representing a 99.9% response rate. Individuals from the original 2004 sample were re-interviewed in 2005 and 2007 using the same questionnaires. Study procedures, consent forms, and questionnaires were approved by the Human Research Protection Program, Tulane University, New Orleans, USA, and the institutional review board, Johns Hopkins Bloomberg School of Public Health, Baltimore, USA. 2.2. Data analysis Multivariate regression analyses were conducted with the NURHI data to examine the association between each potential gateway behavior and all measured health outcomes, accounting for the complex sample design and controlling for age, education, marital status, parity, religion, wealth, and city. Next, the effect of each gateway factor, as well as the interaction between the gateway factor and gateway behavior, were tested to discern whether the synergistic effects on each outcome were greater than the sum of each individually. All analyses were executed in Stata/SE version 12 (Stata, College Station, TX, USA) with 0.05 as the cut-off P value for significance. Multivariate regression analyses examining the association between each potential gateway behavior with all measured health outcomes were conducted using the MVHS data, controlling for age, education, parity, and working status. The gateway behaviors and factors were then ranked by the number of outcomes each predicted. All analyses were conducted using Stata/SE version 12 (Stata, College Station, TX, USA) with 0.05 as the cut-off P value for significance. 3. Results 3.1. Nigerian Urban Reproductive Health Initiative, 2011 All potential gateway behaviors were tagged according to the recognized definition and available data (Box 1). The gateway behaviors identified were then ranked by the average number of health outcomes each behavior predicted. Statistically significant associations with the largest number of positive health outcomes for men and women were: (1) spousal communication about family planning (number of outcomes significantly predicted in women’s dataset: 14; men’s: 7); (2) recommended family planning to family or friends (women: 10; men: 7); (3) ever use of modern family planning (women: 10; men: 6); (4) currently using modern family planning (women: 8; men: 5); and (5) partner had accompanied wife or child to the health facility (women: 10; men: 5) (Tables 1 and 2). From the ranked list, the top five gateway behaviors for women and men were then assessed for effect modification with each potential gateway factor (Box 1). The main gateway factors for both sexes were: (1) knowledge of modern contraceptive methods; (2) exposure
Box 1 Potential gateway behaviors and factors identified and included in the analysis of women’s and men’s Nigerian Urban Reproductive Health Initiative baseline household survey data, 2011. Potential gateway behaviors (men and women) • • • • •
Ever use of modern family planning Current use of modern family planning Family planning use at first ever sexual intercourse Spousal communication about family planning Ever recommended family planning to family or friends (condoms only for males) • Joint decision making for family planning • Partner has accompanied wife or child to a health facility Additional potential gateway behaviors (women only) • • • • • • •
Joint decision making about size of family Prenatal care visit (among currently pregnant women) Medically assisted delivery Delivery in a health facility Child health visit Maternal health visit HIV testing
Potential gateway factors (men and women) • • • • • • • • • • • • • •
Knowledge of modern contraceptive methods Knowledge of the fertile period Perceptions about community family planning use and norms Attitudes toward condom use Membership in a social group/club/organization Family planning discussion, use, or encouragement of use among social network members Attitudes in favor of males prohibiting their female partner’s mobility Tolerance of domestic violence Gender equitable attitudes towards household decisionmaking Gender equitable attitudes toward family planning use Aligned fertility preferences Exposure to family planning in the media (print, radio, or television) Exposure to leaders speaking publically about family planning Access to contraceptives
Additional potential gateway factors (women only) • • • •
Attitudes toward covert family planning use Gender equitable attitudes about finances Quality of family planning service interactions Family planning service integration
to leaders speaking publicly about family planning; (3) aligned fertility preferences; and (4) perceptions about community family planning norms.
3.2. Minya Village Health Surveys, 2004–2007 All potential gateway behaviors in 2004 were identified and examined through bivariate associations with outcome behaviors in 2007 (Box 2). The ranked gateway behaviors that predicted positive health outcomes were: (1) the mother breastfed within 24 hours of delivery; (2) spousal communication about family planning; (3) the mother checked her health with a healthcare provider after the most recent delivery; and (4) the most recent neonate was delivered in a health
H.M. Schwandt et al. / International Journal of Gynecology and Obstetrics 130 (2015) E62–E68
E65
Table 1 Multivariate logistic regression analysis of the four key potential gateway behaviors identified among women in the Nigerian Urban Reproductive Health Initiative baseline survey, 2011.a,b Behavioral outcomes
Family planning use Ever use of modern family planning Current use of modern family planning Long-acting and permanent method use (LAPM) Ever use of Emergency Contraception (EC) Use of family planning at first ever sex Family planning communication Spousal communication about family planning Recommended family planning to family and friends Joint family planning decision-making process Joint family size decision-making process Maternal and child health Prenatal care visit Assisted delivery Delivery in a health facility Child health visit Maternal health visit Other Partner accompanies wife or child to health facility HIV test
Spousal communication about family planning
Ever use of family planning
Recommended family planning to family and friends
Current use of family planning
Partner accompanies wife and/or child to the health facility
7.1e 6.8e 2.6e NS 1.8e
NA NA NA NA 5.3e
8.3e 4.0e 1.4c 1.9e 1.3c
NA NA NA 2.6e 2.0e
1.3e 1.2c NS NS NS
NA 3.9e 6.2e 1.8e
7.0e 8.4e 6.6e 1.2c
3.9e NA 2.4e 1.2c
6.9e 4.0e 18.1e 1.2d
1.7e NS 1.3d 1.3e
1.5c 1.5e 1.4d 1.3e 1.4e
NS 1.7e 1.4d NS 1.2d
NS NS NS 1.3d NS
NA 1.4d NS OD OD
NS 1.5e 1.7e 1.5e 1.4e
1.7e 1.4e
1.3e 1.6e
NS 1.8e
1.2c NS
NA 1.4e
Abbreviations: NA, not applicable; NS, not significant, OD, significant finding in the negative direction. a Values given as Odds ratios. b Multivariate logistic regression analysis with 0.05 as the cut-off P value for significance. All analyses were controlled for age, education, marital status, parity, residence, and wealth (not shown). c P ≤ 0.05. d P ≤ 0.01. e P ≤ 0.001.
facility, which predicted eight, six, six and five behavioral outcomes, respectively (Table 3). The influence of each potential gateway factor was tested by introducing them as interaction terms in the regression models (Box 2). This analysis indicated that gateway factors positively influenced gateway behaviors. For example, discussing family planning with a spouse in 2004 was related to numerous positive health behaviors in 2007. Women who discussed family planning with their spouse in 2004 and who received family planning advice from a health worker exhibited increased likelihood of family planning use in 2007 than women who discussed family planning with their spouse in 2004 but did not receive family planning advice from a health worker. This observation indicated a synergistic effect of spousal and health worker communication on subsequent family planning use.
The ranked gateway factors were: (1) comprehensive content of the prenatal care visit (as defined in Box 2); (2) a health worker gave the woman family planning advice; (3) a health worker gave the woman breastfeeding advice; (4) the mother received supplemental vitamin A in the first 2 months after pregnancy; and (5) the mother was exposed to family planning messages. 4. Discussion The results of the present study supported the Integrated Gateway Model. Positive behaviors either preceded, or were associated with, a cascade of downstream positive outcomes. Some of the relationships were directly related to the original gateway behavior, whereas others seemed unrelated. For example, spousal communication about family planning
Table 2 Multivariate logistic regression analysis of the four key potential gateway behaviors identified among men in the Nigerian Urban Reproductive Health Initiative baseline survey, 2011.a,b Behavioral outcomes
Family planning use Ever use of modern family planning Current use of modern family planning Long-acting and permanent method use (LAPM) Family planning use at first ever sex Family planning communication Spousal communication Recommended family planning to family and friends Joint family planning decision-making process Other Accompanied wife and/or child to health facility
Spousal communication about family planning
Recommended condoms to family and friends
Ever use of family planning
Current use of family planning
Accompanied wife and/or child to health facility
8.0 e 7.4 e 7.0 e 4.1 e
17.4 e 4.9 e 1.9 c 1.4 c
NA 13.0 e NA 7.0 e
12.3 e NA NA 1.7 e
2.4 e 1.8 e NS NS
NA 3.8 e 6.3 e
3.8 e NA 2.7 e
8.1 e 18.7 e 6.6 e
7.4 e 4.9 e NA
2.4 e 2.5 e 1.5 d
2.3 e
2.5 e
2.4 e
1.8 e
NA
Abbreviations: NA, not applicable; NS, not significant. a Values given as Odds ratios. b Multivariate logistic regression analysis with 0.05 as the cut-off P value for significance. All analyses were controlled for age, education, marital status, parity, residence, and wealth (not shown). c P ≤ 0.05. d P ≤ 0.01. e P ≤ 0.001.
E66
H.M. Schwandt et al. / International Journal of Gynecology and Obstetrics 130 (2015) E62–E68
Box 2 Potential gateway behaviors and factors identified and included in the analysis of women’s Egyptian longitudinal Minya Village Health Surveys data, 2004–2007. Family planning related behaviors • Spousal communication about family planning • Talked to anyone about birth spacing • Current use of family planning Prenatal care behaviors • Number of visits • Timing of first visit • Took iron tablets during pregnancy Delivery • Medically assisted delivery • Delivered in a health facility Postnatal care behaviors • • • • • • •
Woman visited a health worker during the postnatal period Timing of postnatal visit Postnatal visit in a medical facility Mother took postnatal supplementary vitamin A Timing of child health check up Child health checkup in a medical facility Duration of postpartum abstinence
Breastfeeding • Initiated within 24 h of delivery • Only breast milk given to the newborn; no other fluids given during the first 24 h • Mother breastfed for at least 6 mo Hygienic behaviors • Mother made sure children washed their hands every morning, before and after meals, and after using the bathroom • Mother washed her hands every morning, before and after meals, and after using the bathroom • Daughters were not circumcised • Never reused syringes • Hygienic precautions taken when handling birds Composite indices • • • • •
Prenatal care index Postnatal care index Breastfeeding index Avian influenza index Overall index of all positive health behaviors
Gateway Factors • Media exposure to family planning messages • Content of the prenatal visit regarding the mother: ○ Given a health card ○ Height, weight, and blood pressure measured ○ Urine and blood samples collected • Intention to use family planning • Intention to circumcise daughters • Mother received supplemental vitamin A in the first 2 months after pregnancy • Tetanus immunization given during pregnancy • A health worker offered the woman family planning advice • A health worker gave woman breast feeding advice • Mother was exposed to family planning messages
was associated not only with the use of contraceptives but also with behaviors such as hand washing, early initiation of breastfeeding (MVHS), and undergoing HIV testing (NURHI). Health workers appeared to play a substantial role in many gateway behaviors and factors. For example, delivery at a health facility; prenatal and postpartum visits with a health worker; content of prenatal care; and health worker advice on family planning and breastfeeding were all gateway behaviors and factors in the MVHS dataset. These findings indicated that services offered by healthcare providers had a catalytic effect on many gateway behaviors. Prenatal care, immunization, and other types of contact with health workers possibly operate as gateway factors. That is, by accessing one type of health service, opportunities to improve other service-related behaviors emerge, either through the breakdown of user barriers to care or through direct provider referrals to other health services. This situation would have implications for the impact of some types of service use as a gateway behavior depending on the mode of delivery. For example, whether vaccination is received in a clinic or during a national immunization day could make a difference, as the latter approach might not provide the opportunity for appreciable contact with a health facility or a healthcare professional. Some behaviors and factors that do not necessitate direct contact with healthcare providers emerged as potential gateways in the present study. For example, interpersonal communication within social networks about family planning—communicating with a spouse about family planning (NURHI and MVHS) or recommending family planning to friends or family (NURHI)—were identified as promising gateways for numerous future positive behaviors. These behaviors included family planning use, joint decision-making, prenatal care, and HIV testing. Interpersonal communication is generally a powerful influence on behavior because it (1) extends the reach of interventions through social networks and (2) moderates or mediates most behavioral decisions through social influence and the self-reflection it encourages. The present study had several strengths. The NURHI sample size was large and representative of urban Nigeria; data were available for both men and women of reproductive age; and it was possible to explore issues related to family planning in more detail than is often possible. The longitudinal design of the MVHS was a major strength as the same women were followed over a period of 3 years. The MVHS also included a variety of family health behaviors as well as reproductive behaviors. The present study also had some limitations. Although populationbased, the MVHS data were drawn from only one of the 23 Egyptian governorates. Furthermore, the selected governorate was in a predominantly rural area. The main limitation of the NURHI was that the data were collected at a single time point and it was not possible to empirically determine causal order. In addition, the data were focused on topics related to family planning; therefore, the ability to examine the potential of gateway behaviors to predict behaviors outside of reproductive health was limited. Unfortunately, it was not possible to assess the effect of gateway moments on health behaviors using the two datasets. In addition, the synergistic effects of gateway moments, behaviors, and factors, could not be determined. Nevertheless, the findings of the present analysis suggested that it would be useful to test whether the observed behaviors are more likely to occur around a gateway moment, such as first pregnancy. Furthermore, it can be hypothesized that gateway behaviors and gateway factors—when they coincide at a gateway moment—would have more influence on subsequent behaviors than if they occurred outside the context of a gateway moment. The identified gateway behaviors and factors suggest that behavioral change programs should focus on intervening either before women become pregnant or early in pregnancy. Such an approach might result in maximum impact on future health for both women and for their offspring. In conclusion, the results of the present study could help to guide investment decisions and improve programmatic impact across a wide range of family health initiatives. Indeed, the Integrated Gateway
H.M. Schwandt et al. / International Journal of Gynecology and Obstetrics 130 (2015) E62–E68
E67
Table 3 Multivariate regression analysis of the four key potential gateway behaviors identified among women in the longitudinal Minya Village Health Survey, Egypt, 2004 and 2007.a,b 2007 Behavioral outcomes
Spousal family planning communication
Breastfeeding within 24 h of delivery
Postnatal health worker visit
Delivery at a health facility
Current use of modern family planning Spousal communication about family planning Talked about spacing pregnancies Delivery at a health facility Postnatal care via health worker visit Postnatal care at a health facility Initiated breastfeeding within 24 h of delivery Neonate not given anything other than breast milk within 24 h of birth Hand washing of the mother Overall indices Behavioral healthc Family planningc Prenatal carec
2.10 f 1.44 f NS NS NS NS 2.10 f NS 1.34 e
1.83 f 1.72 f 1.57 d NS NS NS 2.37 f 1.76 e NS
NS NS NS 2.50 f 1.50 e NS NS NS 2.10 d
NS NS 1.39 d 5.10 f NS 2.40 d NS NS 2.25 d
0.03 f 0.06 f NS
0.04 f 0.09 f 0.12 e
0.04 f 0.04 e 0.09 d
NS NS 0.10 d
Abbreviation: NS, not significant. a Values given as Odds ratios and Regression Coefficients. b Multivariate logistic and linear regression with 0.05 as the cut-off P value for significance. All analyses were controlled for age, education, parity, and employment (not shown). c Continuous predictors d P ≤ 0.05. e P ≤ 0.01. f P ≤ 0.001.
Model is currently being tested in a 3-year operations research project in Nigeria using a field experimental design. The model, therefore, represents a promising avenue for innovative research and programming to identify strategic and cost-effective interventions for public health that promote a catalytic and long-lasting effect on behavior change.
[9] [10] [11]
Acknowledgments The Bill & Melinda Gates Foundation funded the Nigeria Urban Reproductive Health Initiative (NURHI) baseline survey through the Measurement, Learning & Evaluation (MLE) project at the University of North Carolina, as well as all the secondary data analysis presented in this paper. The Minya Village Health Surveys were funded by the US Agency for International Development through the Health Communication Partnership (HCP) project. This analysis, NURHI, and HCP projects were all led by the Johns Hopkins Center for Communication Programs at the Bloomberg School of Public Health. The views expressed in this paper do not necessarily reflect those of the funding bodies.
[12]
[13]
[14]
[15] [16]
Conflict of interest
[17]
The authors declare no conflict of interest regarding the material discussed in the present manuscript.
[18]
References
[20]
[1] Sheppard CW, Gay GR, Smith DE. The changing patterns of heroin addiction in the Haight-Ashbury subculture. J Psychedelic Drugs 1971;3(2):22–31. [2] Ginzler JA, Cochran BN, Domenech-Rodríguez M, Cauce AM, Whitbeck LB. Sequential progression of substance use among homeless youth: an empirical investigation of the gateway theory. Subst Use Misuse 2003;38(3–6):725–58. [3] Prochaska JJ, Spring B, Nigg CR. Multiple health behavior change research: an introduction and overview. Prev Med 2008;46(3):181–8. [4] Dutton GR, Napolitano MA, Whiteley JA, Marcus BH. Is physical activity a gateway behavior for diet? Findings from a physical activity trial. Prev Med 2008;46(3): 216–21. [5] Tucker M, Reicks M. Exercise as a gateway behavior for healthful eating among older adults: an exploratory study. J Nutr Educ Behav 2002;34(Suppl. 1):S14–9. [6] Stevens-Simon C, Kelly LS, Singer D. Pattern of prenatal care and infant immunization status in a comprehensive adolescent-oriented maternity program. Arch Pediatr Adolesc Med 1996;150(8):829–33. [7] Reichman NE, Corman H, Noonan K, Schwartz-Soicher O. Effects of prenatal care on maternal postpartum behaviors. Rev Econ Househ 2010;8(2):171–97. [8] Sugathan, KS, Mishra V, Retherford RD. Promoting Institutional Deliveries in Rural India: the Role of Antenatal-Care Services. National Family Health Survey Subject
[19]
[21]
[22] [23]
[24]
[25]
[26]
Reports Number 20. http://www.eastwestcenter.org/sites/default/files/filemanager/ Research_Program/NFHS_Subject_Reports/subj-20.pdf. Published 2001. Accessed April 27, 2015. Agho KE, Dibley MJ, Odiase JI, Ogbonmwan SM. Determinants of exclusive breastfeeding in Nigeria. BMC Pregnancy Childbirth 2011;11:2. Nwaru BI, Wu Z, Hemminki E. Infant care practices in rural China and their relation to prenatal care utilisation. Glob Public Health 2011;6(1):1–14. Ahmed S, Mosley WH. Simultaneity in the use of maternal-child health care and contraceptives: evidence from developing countries. Demography 2002;39(1): 75–93. Hotchkiss DR, Rous JJ, Seiber EE, Berruti AA. Is maternal and child health service use a causal gateway to subsequent contraceptive use?: A multi-country study. Popul Res Policy Rev 2005;24(6):543–71. Zerai A, Tsui AO. The relationship between prenatal care and subsequent modern contraceptive use in Bolivia, Egypt and Thailand. Afr J Reprod Health 2001;5(2): 68–82. Shefer AM, Fritchley J, Stevenson J, Lyons B, Friedman R, Hopfensperger D, et al. Linking WIC and immunization services to improve preventive health care among low-income children in WIC. J Public Health Manag Pract 2002;8(2):56–65. van der Straten A, King R, Grinstead O, Serufilira A, Allen S. Couple communication, sexual coercion and HIV risk reduction in Kigali, Rwanda. AIDS 1995;9(8):935–44. Desgrées-Du-Loû A, Brou H, Djohan G, Becquet R, Ekouevi DK, Zanou B, et al. Beneficial effects of offering prenatal HIV counselling and testing on developing a HIV preventive attitude among couples. Abidjan, 2002–2005. AIDS Behav 2009;13(2): 348–55. Denison JA, O'Reilly KR, Schmid GP, Kennedy CE, Sweat MD. HIV voluntary counseling and testing and behavioral risk reduction in developing countries: a metaanalysis, 1990–2005. AIDS Behav 2008;12(3):363–73. Putthakeo P, Ali M, Ito C, Vilayhong P, Kuroiwa C. Factors influencing breastfeeding in children less than 2 years of age in Lao PDR. J Paediatr Child Health 2009;45(9):487–92. Naved RT, Persson LA. Factors associated with physical spousal abuse of women during pregnancy in Bangladesh. Int Fam Plan Perspect 2008;34(2):71–8. Storey D, Boulay M, Karki Y, Heckert K, Karmacharya DM. Impact of the integrated Radio Communication Project in Nepal, 1994–1997. J Health Commun 1999;4(4):271–94. Vaughan PW, Rogers EM. A staged model of communication effects: evidence from an entertainment-education radio soap opera in Tanzania. J Health Commun 2000; 5(3):203–27. Bawah AA. Spousal communication and family planning behavior in Navrongo: a longitudinal assessment. Stud Fam Plann 2002;33(2):185–94. Storey JD, Kaggwa E, Harbour C. Research Brief: Pathways to health competence for sustainable health improvement: Examples from South Africa and Egypt. Center for Communication Programs Johns Hopkins Bloomberg School of Public Health. http://ccp.jhu.edu/documents/StoreyKaggwaHarbour.pdf. Published May 2008. Accessed May, 2012. Harkins T, Drasbek C, Arroyo J, McQuestion M. The health benefits of social mobilization: experiences with community-based Integrated Management of Childhood Illness in Chao, Peru and San Luis, Honduras. Promot Educ 2008;15(2):15–20. Kumar V, Kumar A, Das V, Srivastava NM, Baqui AH, Santosham M, et al. Communitydriven impact of a newborn-focused behavioral intervention on maternal health in Shivgarh, India. Int J Gynecol Obstet 2012;117(1):48–55. Kaiser MM, Kaiser KL, Barry TL. Health effects of life transitions for women and children: a research model for public and community health nursing. Public Health Nurs 2009;26(4):370–9.
E68
H.M. Schwandt et al. / International Journal of Gynecology and Obstetrics 130 (2015) E62–E68
[27] McBride CM, Emmons KM, Lipkus IM. Understanding the potential of teachable moments: the case of smoking cessation. Health Educ Res 2003;18(2): 156–70. [28] Measurement, Learning and Evaluation (MLE) Project. 2010-2011 Baseline Household Survey Report for the Nigerian Urban Reproductive Health Initiative. https://www. urbanreproductivehealth.org/sites/mle/files/nurhi_baseline_household_survey.pdf. Published 2012. Accessed April 27, 2015. [29] El-Zanaty F, El-Ghazaly N, Meekers D, Armanious D. 2004. Menya village health survey. [Unpublished report]. Cairo: Communication for Healthy Living Project.
[30] El-Zanaty F, Shehata M, El-Said M, Storey D, Meekers D. 2008. Menya village health survey. [Unpublished report]. Cairo: Communication for Healthy Living Project. [31] El-Zanaty F, Way A. Egypt Interim Demographic and Health Survey. Cairo Egypt: Ministry of Health and Population [Egypt], National Population Council, El-Zanaty and Associates, and ORC Macro; 2004. [32] El-Zanaty F, Way A. Egypt Demographic and Health Survey 2005. Cairo, Egypt: Ministry of Health and Population, National Population Council, El-Zanaty and Associates, and ORC Macro; 2006.