The aim of this review is to evaluate the effect of TPB-based interventions on changing health behaviour among population with chronic diseases in low health-literacy settings. This study also aims at finding out which types of interventions were used, the time-frame of such interventions, the modes of delivery and the settings in which these interventions were delivered. This protocol has been developed in accordance with the Preferred Reporting Items for Systematic review and Meta-Analysis Protocols (PRISMA-P) guidance and draws on the Cochrane Handbook for Systematic Reviews of Interventions guidelines for developing a protocol .
The following study types will be considered for inclusion: interventional studies with a control arm including randomised controlled trials, quasi-experimental studies, community-based cluster randomised trials and controlled before and after studies. Case-control studies, cohort studies, reviews, case reports, case series and animal studies will be excluded.
Participants will be adults over 18 years of age with any chronic disease . Healthy population and pregnant women will be excluded.
Any educational or health intervention used on individuals or groups that documented the use of the constructs of TPB i.e. attitude towards the behaviour, subjective norms and perceived behavioural control for changing health behaviour. From a scoping search of literature, the following terms for identifying TPB will be used – behavioural beliefs, normative beliefs, control beliefs, motivation to comply and perceived power and behavioural intention, besides attitude towards the behaviour, subjective norms and perceived behavioural control.
The control groups will have either 1) any health or educational intervention not based on any psychological theory, 2) health education based on psychological theory other than TPB or 3) treatment as usual without any education. The same group may also be used as control before the TPB-based intervention was delivered to the group and the outcomes evaluated.
The primary outcome will be change in health behaviour which will include preventive behaviours, adherence to treatment and care seeking. Adherence has been defined by World Health Organisation (WHO) as “the extent to which a person’s behaviour – taking medication, following a diet, and/or executing lifestyle changes, corresponds with agreed recommendations from a health care provider”.
Secondary outcomes will include the constructs of TPB which were important influences to health behaviour change and moderators of these effects, including type of intervention, the time-frame of such interventions, the types of health literacy settings, the mode of delivery, the type of providers and satisfaction among patients to the intervention.
Animal studies, studies on health behaviour change which do not mention TPB or other psychological theories and studies undertaken on healthy individuals with a purely health promotion focus will be excluded.
The following databases will be searched for relevant studies: MEDLINE, EMBASE, COCHRANE library, Psych INFO, Web of Science, Scopus, CINAHL, ProQuest databases (ProQuest Sociology and ProQuest Social Sciences), Global Index Medicus, Bibliography of Asian Studies and IndMed. We will also search the grey literature through Open Grey and The Grey Literature Report. Search strategies will be developed for all the databases.
The databases will be searched for relevant studies and all such records will be exported to the Endnote Library for screening, de-duplication and overall management of the records. All the studies will be screened by two independent reviewers and any disagreement will be resolved by discussion, and if necessary will be resolved by arbitration by a third reviewer. A similar process will be followed for full screening of full-text studies. Multiple publications of the same study utilising the same data set will be taken as one study. In case of missing data efforts will be made to contact the authors to request the data. If the full text article for a particular study title or abstract is not freely available through our library resources, external request through an inter-library loan will be made; if this isn’t successful the authors will be approached to provide the full text article. Studies arriving after the cut-off date will be not included but their titles will be mentioned along with the reasons for non-inclusion.
Data extraction and management
A data extraction form will be developed and standardised; it will be piloted and revised before the start of the review. The data extraction form will be adapted from the Cochrane Collaboration data collection form of randomised controlled trials (RCTs) and non-RCTs intervention review . Data extraction will be performed by two independent reviewers and any discrepancies will be resolved by discussion; if still disagreement persists it will be arbitrated by a third reviewer.
Quality and risk of bias for all potential studies will be evaluated using the Cochrane Collaboration’s Tool for Assessing the Risk of Bias, RoB 2 . As per the guidelines in the Cochrane Handbook of Systematic Reviews of Interventions, the studies will be assessed as per standard criteria and will be labelled as ‘low’, ‘unclear’ or ‘high’ risk of bias.
All the characteristics of included studies will be presented in a tabular form with the description of study design, type of disease, type of intervention used, no. of groups involved, outcomes and methods of assessment and risk of bias in each study. The fidelity of the studies will also be assessed, i.e. the number of people completing the intervention and reasons for non-compliance. We will do a narrative synthesis of the studies to draw the conclusions. Whether or not we will do a pooled quantitative estimation (meta-analysis) will depend on the type and heterogeneity of the studies. We will follow the guidance provided in the handbook of Cochrane systematic review to evaluate the heterogeneity. As we expect a high level of heterogeneity, we will apply random effect model. Perhaps, when I-2 statistics is more than 50% we will not do meta- analysis. We intended to explore the reason for heterogeneity with the following subgroup analysis – low-health literacy, types of intervention and modes of delivery.
We intend to do a sensitivity analysis by excluding high risk of bias studies to evaluate the robustness of the overall pooled estimate.
Grading of overall strength of evidence
We will attempt to apply the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach  for creating a summary of findings table. The GRADEpro Guideline Development Tool (GDT), an online tool will be used for either importing the data from Revman or manually adding numerical data into the software for each outcome. The GRADE approach has five domains; risk of bias, inconsistency, indirectness, imprecision and publication bias to evaluate the certainty of evidence for randomised studies. The four levels of evidence of GRADE will be applied, i.e. very low, low, moderate and high.
Registration and reporting
The study is registered with the University of York Centre for Reviews and Dissemination International prospective register of systematic reviews (PROSPERO) with registration no. CRD42018104890. The review will be reported in accordance with the PRISMA guidelines for reporting of systematic reviews. Any amendments to the protocol made during the process of the systematic review will be reported in the final report with rationale for such modifications.
The review team
The search for databases and retrieval of studies, screening, data extraction and quality assessment will be performed independently by BP and RK. BP will work under the guidance of MD. RK is an experienced reviewer working with Cochrane Collaboration – South Asia and well experienced in conducting systematic reviews. MD, DW or RI will arbitrate any disagreements in the review process and will provide field expertise in synthesising the data. DW and RI are experienced researchers; DW and DM have a substantial experience in undertaking systematic reviews.