Tweet
Systematic literature review to examine the evidence for the effectiveness of interventions that use theories and models of behaviour change: towards the prevention and control of communicable diseases (ECDC, July 31 2013, excerpt)
[Source: European Centre for Disease Prevention and Control (ECDC), full PDF document: (LINK). Excerpt, edited.]
TECHNICAL REPORT
Systematic literature review to examine the evidence for the effectiveness of interventions that use theories and models of behaviour change: towards the prevention and control of communicable diseases
Insights into health communication
www.ecdc.europa.eu
________
This systematic literature review was commissioned by the European Centre for Disease Prevention and Control (ECDC) as one of the outputs of the Framework Partnership Agreement Grant/2009/007 ?Establishing a programme for dissemination of evidence-based health communication activities and innovations on communicable diseases for country support in the EU and EEA/EFTA, 2009?12?, with a consortium of universities comprised of the Health Promotion Research Centre at the National University of Ireland Galway, as the lead coordinating centre, and the Institute for Social Marketing, University of Stirling, Scotland, and the University of Navarra Clinic, Pamplona, Spain.
The systematic literature review was produced by Kathryn Angus, Georgina Cairns, Richard Purves, Stuart Bryce, Laura MacDonald and Ross Gordon, Institute for Social Marketing, University of Stirling and the Open University.
The project was overseen by ?lla-Karin Nurm, Andrea W?rz, Piotr Wysocki, ECDC Communication Knowledge Group in the Public Health Capacity and Communication Unit.
Acknowledgements
The authors would like to thank Jennifer McKell for her assistance with screening studies and Diane Dixon for her assistance with document retrieval and in preparing the report.
Suggested citation: Angus K, Cairns G, Purves R, Bryce S, MacDonald L, Gordon R. Systematic literature review to examine the evidence for the effectiveness of interventions that use theories and models of behaviour change: towards the prevention and control of communicable diseases. Stockholm: ECDC; 2013.
Stockholm, July 2013 - ISBN 978-92-9193-438-6 - doi 10.2900/73931 - Catalogue number TQ-30-13-261-EN-C
? Copyright is held jointly by the members of the Translating Health Communications Project Consortium.
Reproduction is authorised, provided the source is acknowledged.
(...)
Executive summary
A systematic literature review assessed the effectiveness of interventions using theories and models of behaviour change to prevent or control communicable diseases relevant to Europe. The review was commissioned by the European Centre for Disease Prevention and Control (ECDC) and conducted by the Institute for Social Marketing of the University of Stirling and the Open University.
Purpose of the review
Behavioural or social theories and models are considered an important tool in effective behaviour change interventions and programmes. They can identify the changes that take place, explain and support change dynamics, identify key influences on outcomes and select those participants most likely to benefit. Their use in health intervention planning and management also improves the prospects for replication, modification and the scaling up of effective interventions and the learning that can be derived from practice whether the intervention is successful or unsuccessful.
There is a substantial body of published work evaluating interventions using behaviour change theories and models to prevent or control communicable diseases relevant to Europe. This systematic review of the evidence aimed to collate the targets of interventions ? diseases and disease groups, populations, types of behaviour for change ? and map the behaviour change theories and models used. A further objective was to look at what techniques and activities were used by interventions, via which health communication channels and in which settings. The review also sought to examine the effectiveness of the interventions and report on any evidence of effectiveness for relevant interventions and programmes based on theories and models of behaviour change. The analysis and findings are intended to provide a current status report on the evidence and shortcomings in relation to good practice, policy, learning and strategies.
Review methods
Systematic reviews are designed to be comprehensive, transparent and replicable and to minimise selection bias. These methods are intended to capture and synthesise research evidence to meet pre-specified research objectives.
Systematic reviews therefore follow a detailed protocol, specified in advance, and fully document all stages of the process.
To be included in this review, studies were required to evaluate an intervention, programme, strategy, action plan, national policy or campaign that aimed to change an individual?s or a community?s health behaviour; prevent and/or control a Europe-relevant communicable disease(s) and used a theory or model of behaviour change cited in Glanz, Rimer and Viswaneth?s 2008 handbook ?Health Behavior and Health Education: Theory, Research, and Practice? [1].
A search strategy was devised to identify relevant studies published in peer-reviewed journals or published as reports by governments or health agencies in English since January 2001. Searches were run in electronic databases of academic literature in the fields of public health, medicine, psychology and social sciences and on the websites of relevant organisations. Relevant study designs were outcome evaluations using experimental or quasiexperimental designs. The measured and reported data had to include a behavioural precursor, or a behavioural outcome.
Behaviour change interventions only aimed at preventing or controlling HIV/AIDS, and those only targeting risky sexual behaviour were excluded from the review.
A team of researchers assessed studies against the inclusion criteria, scored them for internal and external validity and summarised their data in tables. All of the 61 included studies were used in a narrative synthesis to answer the pre-specified research objectives. A narrative synthesis of best evidence, using only the 21 studies rated as high quality (in terms of validity) was used to address the evidence of effectiveness research objectives.
Review results
After the completion of criteria screening 61 studies were included in the review. Three-fifths of the studies were conducted in North America (n=38), nine studies came from Europe, four studies from Australia and four from Africa. Two studies came from the Middle East, two from southern Asia, and two from Puerto Rico. Just over half of the studies included were randomised controlled trials (n=31) and a third were before-and-after studies (n=21).
Twenty-one of the studies scored ≥75% for overall validity in the assessment of study quality. The lowest overall validity score was 33% and the highest 100%, with a mean overall validity score of 68%.
The studies were categorised into six groups according to the disease or infection targeted. The behaviour change intervention in 34 studies targeted more than one disease group and four of the six groups were targeted by at least ten interventions: ?Respiratory tract infections?, ?Vaccine-preventable diseases and invasive bacterial infections?, ?Blood-borne diseases and sexually transmitted infections (STIs)? and ?Food- and waterborne diseases and zoonoses?.
Behaviour change theories and models were used in the studies either to inform the behaviour change intervention or programme, or to design or evaluate the intervention. The models and theories used most often were those that model individuals? health-related behaviour. Models of interpersonal health behaviour were the next most frequently used, with community and group models and theoretical planning frameworks for health promotion used less often for disease prevention or control. Nearly one third of the studies (n=18) used multiple theories or models and two studies mentioned new models synthesised from multiple theories.
All the studies shared the general aim of prevention and/or control of communicable diseases, however they differed in their specific aims (e.g. target population, settings, outcome measures and quality). We synthesised the narrative data of the studies to meet the pre-specified research objectives. These objectives are used as the framework for reporting the results below. Headings (a) to (g) below map the current use of interventions applying theories and models of behaviour change to prevent or control communicable diseases. Headings (h) to (j) assess the evidence of effectiveness where the theories and models are used. The review identifies some promising areas of practice that are recommended for further development and testing for effectiveness. The review also highlights gaps in the evidence base.
a) Communicable diseases and disease groups targeted by theory-based interventions
The 61 studies included were categorised into six groups according to the disease or infection targeted by the intervention. In 34 studies the intervention targeted more than one disease group. The ?Respiratory tract infections? group was the target in 28 studies, focussing on influenza, tuberculosis, upper- and lower-respiratory tract infections and acute respiratory infection. The ?Vaccine preventable diseases and invasive bacterial infections? group was targeted in 19 studies (mostly frequently influenza but also measles and pneumococcal disease), as was the ?Blood-borne diseases and STIs? group (including human papillomavirus infection, hepatitis B, hepatitis C and HIV). Ten studies targeted diseases from the ?Food- and waterborne diseases and zoonoses? group (including hepatitis A, E.coli and non-specific gastrointestinal infections and diarrhoea).
The ?Antimicrobial resistance and healthcare-associated infections? group was the intervention target in four studies (for methicillin-resistant Staphylococcus aureus, rotavirus infection and general nosocomial infections). Finally, diseases from the ?Emerging and vector-borne diseases? group were the target of three studies (including tick-borne Lyme disease, ehrlichiosis and babesiosis, and schistosomiasis).
b) Theory or model used to inform the intervention or programme
Thirty studies used behaviour change theories or models to inform the intervention. The models and theories used most often to inform the intervention or programme were those that model individuals? health-related behaviour.
Studies did not provide much detail beyond stating that a theory or model informed the intervention or evaluation. Nineteen studies were informed by the Health Belief Model, taking into account perceived beliefs of the target audiences for each intervention. Nine studies were informed by individual motivational factors ? four using the Theory of Reasoned Action and five using the Theory of Planned Behavior.
Models of interpersonal health behaviour examine the influence of social relations. Thirteen studies were informed by Social Cognitive Theory, or its earlier form of Social Learning Theory.
Three studies used community and group models of health behaviour. Two of these used the Diffusion of Innovations model to spread an intervention amongst a social group. One study used a community organisation model, the Locality Development Model.
Finally, theoretical frameworks for planning health promotion were used in two studies to inform the intervention. The PRECEDE?PROCEED model was mentioned by one study, and the Behavioral Ecological Model by another.
The majority of studies mentioned a single theory or model only. Ten studies used more than one theory to inform the intervention. Most of the studies that used the Theory of Reasoned Action and/or the Theory of Planned Behavior used a second theory or model. One of the studies combined three theories to create a new theory.
c) Theory or model used in the intervention design
Thirty-one studies used behaviour change theories or models in the intervention design or evaluation. The models and theories used most often in the intervention design or its evaluation were those that targeted individuals? health-related behaviour. Interventions in seven studies were informed by the Health Belief Model, using or measuring perceived beliefs.
Nine studies were informed by individual motivational factors ? two by the Theory of Reasoned Action, six by the Theory of Planned Behavior and one using constructs from both theories in an Integrated Behavioral Model.
The Common Sense Model, which emphasises the role of an individual's emotions in decision-making, was used in one study. Use of the Transtheoretical Model, also referred to as the Stages of Change Model, was mentioned in eight studies, and one study applied it for both the intervention design and its evaluation. One study used another model based on stages, the Precaution Adoption Process Model.
Models of interpersonal health behaviour examine the influence of social interactions. The intervention design of one study was informed by Social Cognitive Theory, another by the Transactional Model of Stress and Coping and a third by the Extended Parallel Process Model. Three studies used the community and group model of health behaviour, Diffusion of Innovations, to design an intervention which could be spread among a social group.
Finally, theoretical frameworks for planning health promotion were used by four studies. The PRECEDE?PROCEED model was applied in three studies to design and evaluate an intervention, and one study used the Social Ecological Model.
The majority of studies made reference to a single theory or model. Eight studies used more than one theory to develop or evaluate the intervention: seven combined constructs from two theories and one combined three theories. Two of these studies combined constructs to create a new theory. Many of the studies did not provide much detail beyond stating that a theory or model informed the intervention design or evaluation.
d) Populations targeted/ segmentation into sub-populations
The interventions in the 61 studies targeted a wide range of populations. Over one-quarter of the interventions targeted more than one population. End-users were the population targeted by most of the interventions.
Population types ranged from parents and children (infants to adolescents), college students and soldiers and new immigrants, to at-risk groups of infected individuals, injecting drug users and pregnant women. Intermediary populations, targeted for the prevention or control of communicable diseases, were predominantly defined as healthcare workers or general practitioners/family doctors (in 17 studies) and food industry workers and school staff (in one study each).
Only four studies segmented their target populations into sub-populations and delivered separate interventions to each. These studies obtained mixed results: two were successful in terms of changing all their sub-populations? behaviour and two were unsuccessful. Another four studies used stage of change models (e.g. the Transtheoretical Model and the Precaution Adoption Process Model) in the design of their intervention to categorise which stages the study subjects were at before the intervention started, and to tailor the intervention to that stage for that study subject. Thus different respondents received suitably tailored and segmented interventions in these four studies, however outcome data were not reported at this level of detail.
e) Types of behaviour targeted
Immunisation or vaccination uptake was the most frequently targeted behaviour for change in the 61 included studies reviewed. Attempts to alter behaviour favouring immunisation were the focus of 27 studies.
Thirteen of these studies specifically targeted the adoption of disease prevention behaviour and increasing awareness and knowledge surrounding the disease as a means of increasing vaccination uptake. Five of these studies targeted parents? attitudes towards vaccinating their child as a means of increasing vaccination rates among children.
Improved hand hygiene was the second most commonly targeted behaviour featuring in 13 of the studies. Six of these studies featured interventions which included promoting the use of waterless hand sanitiser, while four studies aimed to improve hand hygiene by implementing a new programme of hygienic practices or food-safety training. One of these hand hygiene studies was among six which aimed to improve food preparation practices.
Five or fewer studies examined interventions aimed at one of the following behaviour types: addressing the sharing or re-use of injecting drugs equipment; interventions encouraging medicine regimen adherence; reducing antibiotic use and prescription; improving respiratory hygiene behaviour; reducing unsafe sexual behaviour; the uptake of health screening; modification of injecting drug preparation practices and avoiding tick bites, safe tick removal and learning to recognise the symptoms of Lyme disease.
Just under a quarter of the studies reviewed targeted more than one behaviour for change with a view to preventing or controlling communicable diseases (n=14). For example, some interventions aimed at controlling influenza targeted hand hygiene improvement alongside vaccination uptake. Other interventions addressing the reuse of injecting drug equipment also targeted the curbing of unsafe sexual behaviour.
f) Health communication channels, activities and settings used
Health communication was not an inclusion criterion for studies, however all but two described some health communication channel or activity.
Many studies used multiple communication channels or activities in their intervention (148 channels in 61 studies; an average of 2.4 channels per intervention). The most common activity was classes or lectures occurring in just over a third of studies (n=21). One-to-one conversation (or instruction), posters and leaflets were each used by just over a quarter of the interventions. Between and eight and twelve interventions used one of the following channels: educational hand-outs, letters, promotional items (e.g. pens, fridge magnets) or films/DVDs/videotapes.
Seventy-six different settings were coded across the 61 studies, reflecting a tendency for interventions to target only one setting for the behaviour change intervention. The most common settings for interventions were healthcare settings (n=24). Eleven interventions were set in hospitals (four of these in a specific ward), 11 interventions in health centres or doctors? practices (including one intervention that used hospital and health centre settings) and three interventions were set in drug treatment services. Educational settings were the second most frequent setting, used for 15 interventions, from pre-school to tertiary education. Other settings included study participants? own homes (n=10) and other types of accommodation (e.g. residential, nursing or care homes/sheltered housing, military barracks or homeless shelters) (n=5).
Examples of less common settings in the review included: drug treatment services, ?the streets?, cinemas and ferryboats.
Only twelve studies were conducted in more than one setting. These studies tended to have interventions targeting hard-to-reach populations such as homeless people or injecting drugs users, or studies where the intervention was targeted at segmented populations (e.g. family doctors in their surgery and parents of young children in their own homes).
Digital communications were used as channels to communicate behaviour change in nine studies: eight studies (all published between 2008 and 2010) described interventions that used websites as a communication channel and one intervention, from a 2006 study, used email. Of the eight interventions using websites, four could be described as an online setting: three provided online training and in a fourth study, the evaluated intervention was an educational website.
g) Applicability of the theory/model in the evaluation
None of the evaluations in the studies assessed the applicability of the theory or model. Some studies made statements about the applicability of a theory or their intervention, but did not evaluate it. For example, authors were explicit about how theoretical constructs were applied to their intervention, or described similar studies that used the same theory, or simply stated that a certain theory was the most applicable. Other studies discussed applying their intervention to different settings or populations. Those that used behaviour change theory constructs in the evaluation tool sometimes gave an indication of a survey?s predictive validity and testretest reliability.
h) Extent to which the health behaviour change objective of the intervention/ programme was met
Of the 21 studies graded as high quality (≥75% overall validity), over half reported that the intervention had been successful in significantly changing the behaviour of its participants. Interventions in seven studies aimed at changing behaviour to improve a target population?s engagement with healthcare services reported significant behaviour changes. Changed behaviour included increased immunisation or vaccination uptake, increased health screenings attendance and reduced use and prescription of antibiotics. Parents were targeted in their own home setting in six of the studies: five studies reported a significant change in behaviour and only one study reported no significant change in behaviour. The evidence indicated that individual-level behaviour theories ? Health Belief Model, Theory of Reasoned Action and the Theory of Planned Behavior ? and interpersonal behaviour theories ? Social Cognitive Theory and the Transactional Stress and Coping Model ? were associated with positive outcomes.
Both the studies that proposed their own new theoretical model achieved their health behaviour goals. Eight of the 21 studies graded as high quality reported no evidence of effect and did not exhibit any pattern in the intervention targets to explain the lack of success. A broader range of theories and models was associated with the studies reporting a lack of significant results. Four of the studies used individual-level behaviour change theories ? the Health Belief Model, the Theory of Reasoned Action and the Transtheoretical (Stages of Change) Model ? to evaluate the intervention. The other studies mentioned that the same individual-level behaviour change theories informed the interventions. One study was based on a planning model (Social Ecological model) and another used the Diffusion of Innovations community model in its intervention design.
i) Evidence for effective interventions and programmes using theories/models of behaviour change to prevent communicable diseases
Of the 21 high quality studies (≥75% overall validity), the prevention (or prevention and control) of communicable disease was an aim of the intervention in 15 of them. Nine of these studies were considered successful in achieving their aim to change a particular behaviour type with a view to improving the prevention of communicable diseases.
Six were considered unsuccessful, in that they identified little or no change in the target behaviour. Among the 15 prevention studies, there was no comparative evidence available to determine whether using the theory made the intervention effective or not. However, by mapping the studies to their theoretical bases we compared those with an effective intervention (i.e. a successful outcome) to those without.
between the successful and unsuccessful prevention interventions. The main differences were that two successful studies used two theories of interpersonal health behaviour which take account of how an individual?s environment interacts with their health behaviour, and one study with an unsuccessful intervention used a community change model. However, an examination of how the theories were used revealed that the five interventions claiming to apply a theory or model in the intervention design were either significantly effective in changing preventive behaviour (four studies) or changed the behaviour insignificantly but in a positive direction (one study). Those studies which stated they used a theory or model to evaluate the intervention accounted for two successful preventive behaviour change interventions and four unsuccessful behaviour change interventions. In these cases the theory or model was not unsuccessful and acted as a useful tool, providing insight via the measured theoretical constructs.
j) Evidence for effective interventions and programmes using theories/models of behaviour change to control communicable diseases
Among the 21 high quality studies (≥75% overall validity), the control (or prevention and control) of communicable disease was an aim of the intervention in eight studies. Six of these were considered successful in achieving their aim, to change a particular behaviour with a view to improving the control of communicable diseases. Two were considered unsuccessful, in that they found little or no change in the target behaviour. Within the eight studies, there was no comparative evidence available to determine whether using the theory made the intervention effective or not. However, by mapping the studies to their theoretical bases we were able to compare those with an effective intervention (i.e. a successful outcome) to those without.
There were no substantive differences between the health behaviour change theories or models selected and the successful or unsuccessful control of communicable disease interventions. However, examining how theories were used, one of the main differences is that the two interventions stating that they used a theory/model in the intervention design were both significantly effective in changing infection control and preventive behaviour. The rest of the studies with interventions aimed at controlling communicable disease, both the successful and unsuccessful, only mention a theory or model in relation to their intervention, without specific details on its application.
Strategic implications and recommendations
Current practice and evidence of effectiveness
The majority of studies included in this review were informed by theories or models of individual-level behaviour change. There was a lack of evaluative evidence on interpersonal and community-level theories and how these can be used to inform behaviour change interventions. There was also little evidence on interventions which built and tested new theories.
Only one study was identified that evaluated the cost-effectiveness of theory-based interventions. Individual-level research of health behaviour and outcomes is likely to be more costly to conduct than ecological studies, yet as outlined above these theories intended to modify individual-level behaviour remain the most commonly applied.
The majority of interventions were communications-based and were tentative in their use of new media, using it in a more traditional manner.
There was a clear focus on the end-user as the target for the intervention (adults, parents and children, or adolescents). A smaller volume of identified evidence targeted health professionals (healthcare workers and general practitioners) or other intermediaries.
Recommendations for future research
As outlined above, theories intended to modify individual-level behaviour remain the most commonly applied. Policy could correct the current skewing of the evidence by recommending more ecological-level change research.
This could mean more community level theory and/or a policy of a pre-condition that research planning positions individual level theory into wider social scale planning frameworks.
There is a need for an approach to encourage and support research reporting to go beyond simply describing the theoretical model and to measure and test the key variables and how these influence and are influenced by health behaviour.
A number of more recent intervention studies used new media for online training and health promotion.
Technology-driven change and opportunities should trigger reflection on appropriate theories to inform and shape future intervention design, monitoring and evaluation.
Shared learning is perhaps more relevant and achievable than finding out which theory works best. The evidence supports the fact that more than one theory can be effective in achieving the desired impact. An initiative to collect good-practice case studies, could provide a focus for shared learning, practice-led information exchange and practice-informed policy development. This would complement the research evidence and may be more achievable than building a research-led knowledge base.
(...)
-
------
TECHNICAL REPORT
Systematic literature review to examine the evidence for the effectiveness of interventions that use theories and models of behaviour change: towards the prevention and control of communicable diseases
Insights into health communication
www.ecdc.europa.eu
________
This systematic literature review was commissioned by the European Centre for Disease Prevention and Control (ECDC) as one of the outputs of the Framework Partnership Agreement Grant/2009/007 ?Establishing a programme for dissemination of evidence-based health communication activities and innovations on communicable diseases for country support in the EU and EEA/EFTA, 2009?12?, with a consortium of universities comprised of the Health Promotion Research Centre at the National University of Ireland Galway, as the lead coordinating centre, and the Institute for Social Marketing, University of Stirling, Scotland, and the University of Navarra Clinic, Pamplona, Spain.
The systematic literature review was produced by Kathryn Angus, Georgina Cairns, Richard Purves, Stuart Bryce, Laura MacDonald and Ross Gordon, Institute for Social Marketing, University of Stirling and the Open University.
The project was overseen by ?lla-Karin Nurm, Andrea W?rz, Piotr Wysocki, ECDC Communication Knowledge Group in the Public Health Capacity and Communication Unit.
Acknowledgements
The authors would like to thank Jennifer McKell for her assistance with screening studies and Diane Dixon for her assistance with document retrieval and in preparing the report.
Suggested citation: Angus K, Cairns G, Purves R, Bryce S, MacDonald L, Gordon R. Systematic literature review to examine the evidence for the effectiveness of interventions that use theories and models of behaviour change: towards the prevention and control of communicable diseases. Stockholm: ECDC; 2013.
Stockholm, July 2013 - ISBN 978-92-9193-438-6 - doi 10.2900/73931 - Catalogue number TQ-30-13-261-EN-C
? Copyright is held jointly by the members of the Translating Health Communications Project Consortium.
Reproduction is authorised, provided the source is acknowledged.
(...)
Executive summary
A systematic literature review assessed the effectiveness of interventions using theories and models of behaviour change to prevent or control communicable diseases relevant to Europe. The review was commissioned by the European Centre for Disease Prevention and Control (ECDC) and conducted by the Institute for Social Marketing of the University of Stirling and the Open University.
Purpose of the review
Behavioural or social theories and models are considered an important tool in effective behaviour change interventions and programmes. They can identify the changes that take place, explain and support change dynamics, identify key influences on outcomes and select those participants most likely to benefit. Their use in health intervention planning and management also improves the prospects for replication, modification and the scaling up of effective interventions and the learning that can be derived from practice whether the intervention is successful or unsuccessful.
There is a substantial body of published work evaluating interventions using behaviour change theories and models to prevent or control communicable diseases relevant to Europe. This systematic review of the evidence aimed to collate the targets of interventions ? diseases and disease groups, populations, types of behaviour for change ? and map the behaviour change theories and models used. A further objective was to look at what techniques and activities were used by interventions, via which health communication channels and in which settings. The review also sought to examine the effectiveness of the interventions and report on any evidence of effectiveness for relevant interventions and programmes based on theories and models of behaviour change. The analysis and findings are intended to provide a current status report on the evidence and shortcomings in relation to good practice, policy, learning and strategies.
Review methods
Systematic reviews are designed to be comprehensive, transparent and replicable and to minimise selection bias. These methods are intended to capture and synthesise research evidence to meet pre-specified research objectives.
Systematic reviews therefore follow a detailed protocol, specified in advance, and fully document all stages of the process.
To be included in this review, studies were required to evaluate an intervention, programme, strategy, action plan, national policy or campaign that aimed to change an individual?s or a community?s health behaviour; prevent and/or control a Europe-relevant communicable disease(s) and used a theory or model of behaviour change cited in Glanz, Rimer and Viswaneth?s 2008 handbook ?Health Behavior and Health Education: Theory, Research, and Practice? [1].
A search strategy was devised to identify relevant studies published in peer-reviewed journals or published as reports by governments or health agencies in English since January 2001. Searches were run in electronic databases of academic literature in the fields of public health, medicine, psychology and social sciences and on the websites of relevant organisations. Relevant study designs were outcome evaluations using experimental or quasiexperimental designs. The measured and reported data had to include a behavioural precursor, or a behavioural outcome.
Behaviour change interventions only aimed at preventing or controlling HIV/AIDS, and those only targeting risky sexual behaviour were excluded from the review.
A team of researchers assessed studies against the inclusion criteria, scored them for internal and external validity and summarised their data in tables. All of the 61 included studies were used in a narrative synthesis to answer the pre-specified research objectives. A narrative synthesis of best evidence, using only the 21 studies rated as high quality (in terms of validity) was used to address the evidence of effectiveness research objectives.
Review results
After the completion of criteria screening 61 studies were included in the review. Three-fifths of the studies were conducted in North America (n=38), nine studies came from Europe, four studies from Australia and four from Africa. Two studies came from the Middle East, two from southern Asia, and two from Puerto Rico. Just over half of the studies included were randomised controlled trials (n=31) and a third were before-and-after studies (n=21).
Twenty-one of the studies scored ≥75% for overall validity in the assessment of study quality. The lowest overall validity score was 33% and the highest 100%, with a mean overall validity score of 68%.
The studies were categorised into six groups according to the disease or infection targeted. The behaviour change intervention in 34 studies targeted more than one disease group and four of the six groups were targeted by at least ten interventions: ?Respiratory tract infections?, ?Vaccine-preventable diseases and invasive bacterial infections?, ?Blood-borne diseases and sexually transmitted infections (STIs)? and ?Food- and waterborne diseases and zoonoses?.
Behaviour change theories and models were used in the studies either to inform the behaviour change intervention or programme, or to design or evaluate the intervention. The models and theories used most often were those that model individuals? health-related behaviour. Models of interpersonal health behaviour were the next most frequently used, with community and group models and theoretical planning frameworks for health promotion used less often for disease prevention or control. Nearly one third of the studies (n=18) used multiple theories or models and two studies mentioned new models synthesised from multiple theories.
All the studies shared the general aim of prevention and/or control of communicable diseases, however they differed in their specific aims (e.g. target population, settings, outcome measures and quality). We synthesised the narrative data of the studies to meet the pre-specified research objectives. These objectives are used as the framework for reporting the results below. Headings (a) to (g) below map the current use of interventions applying theories and models of behaviour change to prevent or control communicable diseases. Headings (h) to (j) assess the evidence of effectiveness where the theories and models are used. The review identifies some promising areas of practice that are recommended for further development and testing for effectiveness. The review also highlights gaps in the evidence base.
a) Communicable diseases and disease groups targeted by theory-based interventions
The 61 studies included were categorised into six groups according to the disease or infection targeted by the intervention. In 34 studies the intervention targeted more than one disease group. The ?Respiratory tract infections? group was the target in 28 studies, focussing on influenza, tuberculosis, upper- and lower-respiratory tract infections and acute respiratory infection. The ?Vaccine preventable diseases and invasive bacterial infections? group was targeted in 19 studies (mostly frequently influenza but also measles and pneumococcal disease), as was the ?Blood-borne diseases and STIs? group (including human papillomavirus infection, hepatitis B, hepatitis C and HIV). Ten studies targeted diseases from the ?Food- and waterborne diseases and zoonoses? group (including hepatitis A, E.coli and non-specific gastrointestinal infections and diarrhoea).
The ?Antimicrobial resistance and healthcare-associated infections? group was the intervention target in four studies (for methicillin-resistant Staphylococcus aureus, rotavirus infection and general nosocomial infections). Finally, diseases from the ?Emerging and vector-borne diseases? group were the target of three studies (including tick-borne Lyme disease, ehrlichiosis and babesiosis, and schistosomiasis).
b) Theory or model used to inform the intervention or programme
Thirty studies used behaviour change theories or models to inform the intervention. The models and theories used most often to inform the intervention or programme were those that model individuals? health-related behaviour.
Studies did not provide much detail beyond stating that a theory or model informed the intervention or evaluation. Nineteen studies were informed by the Health Belief Model, taking into account perceived beliefs of the target audiences for each intervention. Nine studies were informed by individual motivational factors ? four using the Theory of Reasoned Action and five using the Theory of Planned Behavior.
Models of interpersonal health behaviour examine the influence of social relations. Thirteen studies were informed by Social Cognitive Theory, or its earlier form of Social Learning Theory.
Three studies used community and group models of health behaviour. Two of these used the Diffusion of Innovations model to spread an intervention amongst a social group. One study used a community organisation model, the Locality Development Model.
Finally, theoretical frameworks for planning health promotion were used in two studies to inform the intervention. The PRECEDE?PROCEED model was mentioned by one study, and the Behavioral Ecological Model by another.
The majority of studies mentioned a single theory or model only. Ten studies used more than one theory to inform the intervention. Most of the studies that used the Theory of Reasoned Action and/or the Theory of Planned Behavior used a second theory or model. One of the studies combined three theories to create a new theory.
c) Theory or model used in the intervention design
Thirty-one studies used behaviour change theories or models in the intervention design or evaluation. The models and theories used most often in the intervention design or its evaluation were those that targeted individuals? health-related behaviour. Interventions in seven studies were informed by the Health Belief Model, using or measuring perceived beliefs.
Nine studies were informed by individual motivational factors ? two by the Theory of Reasoned Action, six by the Theory of Planned Behavior and one using constructs from both theories in an Integrated Behavioral Model.
The Common Sense Model, which emphasises the role of an individual's emotions in decision-making, was used in one study. Use of the Transtheoretical Model, also referred to as the Stages of Change Model, was mentioned in eight studies, and one study applied it for both the intervention design and its evaluation. One study used another model based on stages, the Precaution Adoption Process Model.
Models of interpersonal health behaviour examine the influence of social interactions. The intervention design of one study was informed by Social Cognitive Theory, another by the Transactional Model of Stress and Coping and a third by the Extended Parallel Process Model. Three studies used the community and group model of health behaviour, Diffusion of Innovations, to design an intervention which could be spread among a social group.
Finally, theoretical frameworks for planning health promotion were used by four studies. The PRECEDE?PROCEED model was applied in three studies to design and evaluate an intervention, and one study used the Social Ecological Model.
The majority of studies made reference to a single theory or model. Eight studies used more than one theory to develop or evaluate the intervention: seven combined constructs from two theories and one combined three theories. Two of these studies combined constructs to create a new theory. Many of the studies did not provide much detail beyond stating that a theory or model informed the intervention design or evaluation.
d) Populations targeted/ segmentation into sub-populations
The interventions in the 61 studies targeted a wide range of populations. Over one-quarter of the interventions targeted more than one population. End-users were the population targeted by most of the interventions.
Population types ranged from parents and children (infants to adolescents), college students and soldiers and new immigrants, to at-risk groups of infected individuals, injecting drug users and pregnant women. Intermediary populations, targeted for the prevention or control of communicable diseases, were predominantly defined as healthcare workers or general practitioners/family doctors (in 17 studies) and food industry workers and school staff (in one study each).
Only four studies segmented their target populations into sub-populations and delivered separate interventions to each. These studies obtained mixed results: two were successful in terms of changing all their sub-populations? behaviour and two were unsuccessful. Another four studies used stage of change models (e.g. the Transtheoretical Model and the Precaution Adoption Process Model) in the design of their intervention to categorise which stages the study subjects were at before the intervention started, and to tailor the intervention to that stage for that study subject. Thus different respondents received suitably tailored and segmented interventions in these four studies, however outcome data were not reported at this level of detail.
e) Types of behaviour targeted
Immunisation or vaccination uptake was the most frequently targeted behaviour for change in the 61 included studies reviewed. Attempts to alter behaviour favouring immunisation were the focus of 27 studies.
Thirteen of these studies specifically targeted the adoption of disease prevention behaviour and increasing awareness and knowledge surrounding the disease as a means of increasing vaccination uptake. Five of these studies targeted parents? attitudes towards vaccinating their child as a means of increasing vaccination rates among children.
Improved hand hygiene was the second most commonly targeted behaviour featuring in 13 of the studies. Six of these studies featured interventions which included promoting the use of waterless hand sanitiser, while four studies aimed to improve hand hygiene by implementing a new programme of hygienic practices or food-safety training. One of these hand hygiene studies was among six which aimed to improve food preparation practices.
Five or fewer studies examined interventions aimed at one of the following behaviour types: addressing the sharing or re-use of injecting drugs equipment; interventions encouraging medicine regimen adherence; reducing antibiotic use and prescription; improving respiratory hygiene behaviour; reducing unsafe sexual behaviour; the uptake of health screening; modification of injecting drug preparation practices and avoiding tick bites, safe tick removal and learning to recognise the symptoms of Lyme disease.
Just under a quarter of the studies reviewed targeted more than one behaviour for change with a view to preventing or controlling communicable diseases (n=14). For example, some interventions aimed at controlling influenza targeted hand hygiene improvement alongside vaccination uptake. Other interventions addressing the reuse of injecting drug equipment also targeted the curbing of unsafe sexual behaviour.
f) Health communication channels, activities and settings used
Health communication was not an inclusion criterion for studies, however all but two described some health communication channel or activity.
Many studies used multiple communication channels or activities in their intervention (148 channels in 61 studies; an average of 2.4 channels per intervention). The most common activity was classes or lectures occurring in just over a third of studies (n=21). One-to-one conversation (or instruction), posters and leaflets were each used by just over a quarter of the interventions. Between and eight and twelve interventions used one of the following channels: educational hand-outs, letters, promotional items (e.g. pens, fridge magnets) or films/DVDs/videotapes.
Seventy-six different settings were coded across the 61 studies, reflecting a tendency for interventions to target only one setting for the behaviour change intervention. The most common settings for interventions were healthcare settings (n=24). Eleven interventions were set in hospitals (four of these in a specific ward), 11 interventions in health centres or doctors? practices (including one intervention that used hospital and health centre settings) and three interventions were set in drug treatment services. Educational settings were the second most frequent setting, used for 15 interventions, from pre-school to tertiary education. Other settings included study participants? own homes (n=10) and other types of accommodation (e.g. residential, nursing or care homes/sheltered housing, military barracks or homeless shelters) (n=5).
Examples of less common settings in the review included: drug treatment services, ?the streets?, cinemas and ferryboats.
Only twelve studies were conducted in more than one setting. These studies tended to have interventions targeting hard-to-reach populations such as homeless people or injecting drugs users, or studies where the intervention was targeted at segmented populations (e.g. family doctors in their surgery and parents of young children in their own homes).
Digital communications were used as channels to communicate behaviour change in nine studies: eight studies (all published between 2008 and 2010) described interventions that used websites as a communication channel and one intervention, from a 2006 study, used email. Of the eight interventions using websites, four could be described as an online setting: three provided online training and in a fourth study, the evaluated intervention was an educational website.
g) Applicability of the theory/model in the evaluation
None of the evaluations in the studies assessed the applicability of the theory or model. Some studies made statements about the applicability of a theory or their intervention, but did not evaluate it. For example, authors were explicit about how theoretical constructs were applied to their intervention, or described similar studies that used the same theory, or simply stated that a certain theory was the most applicable. Other studies discussed applying their intervention to different settings or populations. Those that used behaviour change theory constructs in the evaluation tool sometimes gave an indication of a survey?s predictive validity and testretest reliability.
h) Extent to which the health behaviour change objective of the intervention/ programme was met
Of the 21 studies graded as high quality (≥75% overall validity), over half reported that the intervention had been successful in significantly changing the behaviour of its participants. Interventions in seven studies aimed at changing behaviour to improve a target population?s engagement with healthcare services reported significant behaviour changes. Changed behaviour included increased immunisation or vaccination uptake, increased health screenings attendance and reduced use and prescription of antibiotics. Parents were targeted in their own home setting in six of the studies: five studies reported a significant change in behaviour and only one study reported no significant change in behaviour. The evidence indicated that individual-level behaviour theories ? Health Belief Model, Theory of Reasoned Action and the Theory of Planned Behavior ? and interpersonal behaviour theories ? Social Cognitive Theory and the Transactional Stress and Coping Model ? were associated with positive outcomes.
Both the studies that proposed their own new theoretical model achieved their health behaviour goals. Eight of the 21 studies graded as high quality reported no evidence of effect and did not exhibit any pattern in the intervention targets to explain the lack of success. A broader range of theories and models was associated with the studies reporting a lack of significant results. Four of the studies used individual-level behaviour change theories ? the Health Belief Model, the Theory of Reasoned Action and the Transtheoretical (Stages of Change) Model ? to evaluate the intervention. The other studies mentioned that the same individual-level behaviour change theories informed the interventions. One study was based on a planning model (Social Ecological model) and another used the Diffusion of Innovations community model in its intervention design.
i) Evidence for effective interventions and programmes using theories/models of behaviour change to prevent communicable diseases
Of the 21 high quality studies (≥75% overall validity), the prevention (or prevention and control) of communicable disease was an aim of the intervention in 15 of them. Nine of these studies were considered successful in achieving their aim to change a particular behaviour type with a view to improving the prevention of communicable diseases.
Six were considered unsuccessful, in that they identified little or no change in the target behaviour. Among the 15 prevention studies, there was no comparative evidence available to determine whether using the theory made the intervention effective or not. However, by mapping the studies to their theoretical bases we compared those with an effective intervention (i.e. a successful outcome) to those without.
between the successful and unsuccessful prevention interventions. The main differences were that two successful studies used two theories of interpersonal health behaviour which take account of how an individual?s environment interacts with their health behaviour, and one study with an unsuccessful intervention used a community change model. However, an examination of how the theories were used revealed that the five interventions claiming to apply a theory or model in the intervention design were either significantly effective in changing preventive behaviour (four studies) or changed the behaviour insignificantly but in a positive direction (one study). Those studies which stated they used a theory or model to evaluate the intervention accounted for two successful preventive behaviour change interventions and four unsuccessful behaviour change interventions. In these cases the theory or model was not unsuccessful and acted as a useful tool, providing insight via the measured theoretical constructs.
j) Evidence for effective interventions and programmes using theories/models of behaviour change to control communicable diseases
Among the 21 high quality studies (≥75% overall validity), the control (or prevention and control) of communicable disease was an aim of the intervention in eight studies. Six of these were considered successful in achieving their aim, to change a particular behaviour with a view to improving the control of communicable diseases. Two were considered unsuccessful, in that they found little or no change in the target behaviour. Within the eight studies, there was no comparative evidence available to determine whether using the theory made the intervention effective or not. However, by mapping the studies to their theoretical bases we were able to compare those with an effective intervention (i.e. a successful outcome) to those without.
There were no substantive differences between the health behaviour change theories or models selected and the successful or unsuccessful control of communicable disease interventions. However, examining how theories were used, one of the main differences is that the two interventions stating that they used a theory/model in the intervention design were both significantly effective in changing infection control and preventive behaviour. The rest of the studies with interventions aimed at controlling communicable disease, both the successful and unsuccessful, only mention a theory or model in relation to their intervention, without specific details on its application.
Strategic implications and recommendations
Current practice and evidence of effectiveness
The majority of studies included in this review were informed by theories or models of individual-level behaviour change. There was a lack of evaluative evidence on interpersonal and community-level theories and how these can be used to inform behaviour change interventions. There was also little evidence on interventions which built and tested new theories.
Only one study was identified that evaluated the cost-effectiveness of theory-based interventions. Individual-level research of health behaviour and outcomes is likely to be more costly to conduct than ecological studies, yet as outlined above these theories intended to modify individual-level behaviour remain the most commonly applied.
The majority of interventions were communications-based and were tentative in their use of new media, using it in a more traditional manner.
There was a clear focus on the end-user as the target for the intervention (adults, parents and children, or adolescents). A smaller volume of identified evidence targeted health professionals (healthcare workers and general practitioners) or other intermediaries.
Recommendations for future research
As outlined above, theories intended to modify individual-level behaviour remain the most commonly applied. Policy could correct the current skewing of the evidence by recommending more ecological-level change research.
This could mean more community level theory and/or a policy of a pre-condition that research planning positions individual level theory into wider social scale planning frameworks.
There is a need for an approach to encourage and support research reporting to go beyond simply describing the theoretical model and to measure and test the key variables and how these influence and are influenced by health behaviour.
A number of more recent intervention studies used new media for online training and health promotion.
Technology-driven change and opportunities should trigger reflection on appropriate theories to inform and shape future intervention design, monitoring and evaluation.
Shared learning is perhaps more relevant and achievable than finding out which theory works best. The evidence supports the fact that more than one theory can be effective in achieving the desired impact. An initiative to collect good-practice case studies, could provide a focus for shared learning, practice-led information exchange and practice-informed policy development. This would complement the research evidence and may be more achievable than building a research-led knowledge base.
(...)
-
------