Assessing the Influence of Innovation Characteristics on the Implementation Process of an Optimised Tuberculosis Diabetes Integrated Care Package (Opt TBD) -A Mixed Method Study Protocol

Study settings and design

The Opt-TBD care package intervention is being implemented at thirteen purposefully selected TB care facilities across the provinces of Khyber Pakhtunkhwa and Punjab in Pakistan, over a period of 18 months. These facilities include a combination of primary, secondary, and tertiary care settings to ensure that the intervention can be implemented and evaluated across different healthcare levels.

A fully integrated convergent parallel¹⁷ mixed-method design is employed, in which quantitative and qualitative data are collected and analyzed simultaneously. This approach enhances the depth and validity of the research findings by enabling the triangulation of results, thereby providing a more comprehensive and nuanced understanding of both the implementation process and contextual influences.¹⁸

Data collection method & tools

Qualitative and quantitative approaches will be employed to gather insights from the Opt TBD developers, TB patients with diabetes, DOTS facilitators, and medical officers. Table 1 below represents the scheduled timing of all data collection tools during the Opt TBD care package implementation. Tables 2 and 3 mentioned below, outline these methods and their respective participants. The qualitative data collection instruments, comprising three topic guides, were independently developed based on the CFIR model to ensure theoretical grounding and contextual relevance. The guides underwent critical review by a multidisciplinary panel, including a qualitative research expert, an implementation scientist, and a public health researcher, to enhance content validity and methodological rigor. In parallel, a quantitative survey instrument designed to assess perceived innovation characteristics was developed in alignment with Objective 3 of the study. This self-developed tool is conceptually grounded in the Consolidated Framework for Implementation Research (CFIR). An extensive review informed item generation from existing literature and was subsequently adapted and contextualized to reflect the components of the OPT-TBD care package. The instrument will undergo expert validation prior to its application in Objective 4 to ensure content validity, relevance, and clarity. The implementation outcomes questionnaire (IOQ), which is freely available online, will be utilized in the context of the OPT TBD care package to assess the progress and effectiveness of the implementation efforts within this study. This tool will be piloted first before its full application in Objective 4 along with the innovation characteristics instrument.

We will first detail the qualitative methods for objectives one, two, and four of the study. Next, we will describe the quantitative methods for objectives three and four. For all four objectives, the data collection will be based on the constructs of the first domain, which represent the innovation characteristics of the CFIR model.^16,19 We have adhered to the Medical Research Council guidance on developing and evaluating complex interventions, emphasising the significance of engaging all relevant stakeholders.²⁰ The first author will undertake primary data collection. To ensure efficient data collection across multiple study sites and adherence to project timelines, two trained research assistants will be engaged if additional logistical or operational support is needed.

Qualitative data

Qualitative data collection serves as the foundational component of this study, enabling an in-depth exploration of implementation experiences across three distinct stages. Guided by the CFIR framework, the inquiry specifically focused on the domain of ‘Innovation Characteristics’ to examine how perceived attributes of the OPT TBD care package influenced its implementation across the TB settings.

Data collection tool ( Table 2)

Topic guides for the three qualitative data collection stages will be systematically developed using the eight core constructs of the ‘Innovation Characteristics’ domain from the Consolidated Framework for Implementation Research (CFIR).^16,19 These constructs will serve as the analytical lens guiding data collection. In Stage 1, which involves interviews with intervention developers, the topic guide will encompass all eight constructs: intervention source, evidence strength and quality, adaptability, relative advantage, trialability, complexity, design and packaging, and cost. Stage 2 will focus specifically on trialability and complexity, explored through interviews to identify necessary adaptations and challenges before definitive implementation. In the final stage 4 of qualitative data collection, the topic guide will be designed to assess evidence strength and quality, adaptability, relative advantages, complexity, design and packaging, drawing on the lived experiences of implementers and end-users to evaluate the intervention’s perceived value and sustainability.

These guides will be reviewed and finalized with input from a qualitative expert, a public health expert, and implementation science experts. They will serve as a structured framework to capture in-depth insight into participants’ experiences, perceptions of the intervention, the contextual and operational factors influencing its delivery, as well as barriers and facilitators to scaling up. A total of three distinct topic guides will be developed, tailored to intervention developers, DOTS facilitators, Medical Officers (MOs) and TB diabetes patients.

Stage 1: Source, evidence-based innovation, adaptability, relative advantage, trialability, complexity, design & packaging and cost

This is the first phase of a study which aimed to explore intervention developers’ perceptions and experiences regarding the development of OPT TBD care packages based on the innovation characteristics of CFIR.

Data collection, participants & sampling

A purposive maximum variation sampling technique will be employed to identify participants. We will select participants involved in the intervention’s development, ensuring representation by choosing one participant from each group at every stage, thereby enhancing the transferability and richness of findings. Selection will be based on the following criteria: the participant’s experience in intervention development, their role within the development team (e.g., lead, co-investigator, researcher), their professional background (such as clinician, public health expert), service users such as patients, and the type and setting of the intervention.²¹ A target of approximately 10 participants will be recruited, aiming to include at least one representative from each identified stakeholder group.

The data collection will be initiated through formal email invitations, which will include a participant information sheet that explicitly outlines the study’s purpose, confidentiality measures, audio recording procedures, and the voluntary nature of participation. Written consent form will be taken from all the participants of the study. Non-responders will receive two reminder emails over a two-week interval.

Interviews will be conducted via Zoom for remote participants and in person where feasible, depending on participants’ locations and availability. All interviews will follow a semi-structured format guided by a piloted interview guide mapped to relevant CFIR innovation constructs. Interviews will be audio recorded, transcribed verbatim, and supplemented by brief field notes. Demographic data will be systematically collected using a standardised form.

Data collection will span approximately 8 to 12 weeks, and all procedures will be documented in a study log.

Stage 2: Trialability & complexity

The trialability of an intervention is crucial, because it helps identify the operational challenges, contextual fit and complexity of its components at a small-scale level. Therefore, this stage of the study is conducted to understand the trialability and complexity of the OPT-TBD care package. Insights from this phase will guide the refinement of the intervention to enhance its adaptability for broader implementation, in line with CFIR innovation characteristics.

Data collection, participants and sampling

This stage of study involves data collection from two key participant groups: 1) Intervention deliverers such as TB health workers (DOTS facilitators and medical officers) and 2) Intervention recipients, consisting of TB patients diagnosed with diabetes and those who will receive OPT TBD care package.

The OPT TBD care package is initially implemented at small level only at four sites: three in KP and one in Punjab. We will conduct qualitative interviews with four DOTS facilitators, four MOs, and eight patients, ensuring a balanced representation of gender and settings and data saturation. The intervention will be delivered in three stages at the TB facility (at 0, 3, and 6 months). Therefore, qualitative semi-structured interviews will be held with the intervention deliverers (DOTs & MOs) and intervention recipients (patients) at the 6-month time point, after the complete delivery of the intervention. All participants will complete demographic forms, and there will be an information sheet containing all the details about the research study. Before starting the interview, participants will be asked to sign consent forms, and their interviews will be audio recorded. Data collection will span approximately 8–32 weeks, and all procedures will be documented in a study log.

Stage 4: Evidence-based innovation, adaptability, relative advantage, complexity, design and packaging

In the definitive implementation phase, collecting qualitative data is critical to gaining an in-depth understanding of the challenges faced by end users during the real-world delivery of the OPT-TBD care package. Specifically, this phase aims to explore perceptions related to the adaptability and relative advantage of the intervention based on the first hand experiences of both providers and recipients. To complement and contextualise the quantitative findings, a series of semi-structured interviews will be conducted at the six-month point, following the complete delivery of the care package across all implementation sites.

Data collection, participants and sampling

The definitive implementation will take place across 13 designated TB care facilities. To ensure site-level representation, one patient and one DOTS facilitator will be purposively selected from each site, resulting in a total of 26 interviews, ensuring the data saturation. These participants will include intervention deliverers (DOTS facilitators) and intervention recipients (patients with TB and diabetes).

Before the interviews, participants will complete a demographic information form and receive a detailed participant information sheet outlining the study’s purpose, procedures, risks, and benefits. Written informed consent will be obtained from all participants, including consent for audio recording. Face-to-face interviews will be conducted, digitally recorded, transcribed verbatim, and anonymized. Data collection is expected to occur over 8 to 32 weeks, depending on site readiness and participant availability.

Quantitative data

Objective 3: Innovation characteristics tool validation

To ensure methodological rigour and enhance the reproducibility of the innovation characteristics tool, a structured, two-phase process will be implemented to establish its content validity. The development of the tool will be informed by a comprehensive literature review of key implementation science frameworks, particularly the Consolidated Framework for Implementation Research (CFIR), which outlines core constructs such as evidence-based innovation, adaptability, trialability, relative advantage, and complexity. This review will guide the conceptual definition and operationalization of each construct within the specific context of the OPT-TBD care package. Based on this foundational work, an initial pool of items will be generated to capture the theoretical breadth of the selected constructs. The rationale for undertaking a formal content validity assessment is to ensure that the instrument adequately represents the intended construct domains and demonstrates contextual relevance across diverse healthcare implementation settings.

Data collection, participants and sampling

Content validity will be assessed using Lawshe’s Content Validity Ratio (CVR) methodology. A panel of 10–14 subject matter experts, including implementation researchers, public health professionals, TB/diabetes clinicians, and medical health professions education specialists, will be purposively selected based on their domain expertise. Experts will be contacted via email and invited to participate in two rounds of item evaluation using structured questionnaires administered through Google Forms.

In each round, experts will independently evaluate the clarity, relevance, and representativeness of each item using a 4-point ordinal scale (i.e., very clear, clear, somewhat clear, or not at all clear). CVR scores will be calculated for each item based on expert consensus, with items retained, revised, or discarded according to Lawshe’s critical value thresholds. Furthermore, the Item- and Content Validity Ratio (I-CVI and CVR) will be computed to further inform item-level decision-making. The tool will be iteratively refined after each round, based on both quantitative indices and qualitative expert feedback in the suggestion column. The entire expert validation and data collection process will be completed within a two-month timeframe. All procedures will be documented in a study log to ensure transparency. The end goal is to produce a psychometrically sound instrument capable of reliably measuring the innovation characteristics relevant to implementation outcomes in real-world health systems and to establish a way forward to collect data on objective 4 of this study.

Objective 4: Association between innovation characteristics & implementation outcomes

To evaluate how specific innovation characteristics influence the implementation process, a structured mixed-methods approach will be adopted during the definitive implementation phase of the refined OPT-TBD intervention. Quantitative and qualitative data will be collected concurrently from TB health workers and patients receiving the OPT TBD care package across all 13 implementation sites. Qualitative data have already been explained in the previous section, while the quantitative section is stated below.

Data collection tools ( Tables 2 & 3)

The measurement of innovation characteristics, including evidence-based innovation, adaptability, trialability, relative advantage, and complexity, will rely on an instrument developed and content-validated under Objective 3. Following pilot testing on 10% of the sample for clarity and feasibility, this tool will be deployed during the definitive implementation phase to collect data from both TB and diabetes patients, as well as TB health workers (DOTS facilitators and medical officers) involved in delivering the OPT-TBD care package.

On the other hand, implementation outcomes will be assessed using the implementation outcome questionnaire,²² which captures core constructs such as acceptability, adoption, appropriateness, and feasibility. These outcomes will provide quantitative insights into the perceived effectiveness and operational integration of the care package.

The fidelity of implementation will be assessed using a structured checklist developed under the POTENTIAL project, grounded in the conceptual framework for implementation fidelity.²³ This checklist will guide both observational assessments and audio-recorded fidelity monitoring. Because this research is nested within the broader POTENTIAL project, the same fidelity data will be utilised to ensure consistency, comparability, and traceability of implementation fidelity measures across sites and phases.

Data collection, participants and sampling

A census-based sampling strategy will be used to include all eligible participants. Quantitative data will be gathered from the entire cohort of almost 250 patients, as well as from all DOTS facilitators and MOs involved in delivering the intervention. Approximately 250 patients with confirmed co-diagnoses of tuberculosis and diabetes are expected to be recruited under the POTENTIAL project and will be provided with the OPT-TBD integrated care package as part of the planned intervention delivery.⁸ The census approach²⁴ is scientifically justified given the relatively small, accessible population size, allowing for maximal coverage, elimination of sampling bias, and detailed assessment of inter-individual variability in implementation responses. All participants will be informed through a detailed study information sheet and written informed consent will be taken before completing hard-copy questionnaires.

Fidelity to the intervention protocol will be rigorously assessed. The OPT-TBD intervention will be delivered in three stages (baseline, 3 months, and 6 months). Audio recordings of intervention delivery sessions will be collected at each stage from five districts, with two audio recordings per district (total n = 10). Three independent reviewers will score each recording on the checklist, with discrepancies resolved through consensus to ensure inter-rater reliability. To further validate fidelity, data from the POTENTIAL project⁸ will be leveraged.

All data collection activities, including survey administration, audio recordings, and fidelity assessments, will be completed within a six-month timeline, with detailed documentation of procedures maintained in a central study log to ensure reproducibility and auditability.

Data confidentiality

The confidentiality of all the personal data will be securely maintained throughout the study’s duration and afterwards. Only essential personal information will be collected and linked to participant IDs on a demographic sheet. The correspondence details will be deleted once they are no longer required. All data will be anonymised by giving participant IDs and will be stored at Khyber Medical University’s assigned locker to the first author. Audio recordings of all three phases of the interviews will be captured via an encrypted audio recorder.

Data analysis plan

In this research study, a fully integrated convergent parallel mixed method design (QUAL (dc) QUAL (dc)🡪QUANT (dc) + QUAL (dc) 🡪Analysis 🡪Integration of data)¹⁷ will be used to assess the influence of innovation characteristics on the implementation process of an optimised TBD care package. The parallel collection and analysis of quantitative and qualitative data for a better understanding of the causal relationship between innovation characteristics and the implementation process will be performed. Using both methods help confirm that the results converge to the same conclusion.

Qualitative analysis

Audio-recorded interviews, from all three stages, will be transcribed word-for-word to capture every detail accurately. The data from these interviews will be analysed via thematic framework analysis. This approach combines both inductive and deductive methods. The process begins with familiarising ourselves with the data and developing a framework. Following this, we proceed with indexing the data, organising it into charts, and finally interpreting the findings. Microsoft Excel will be utilised to assist with these stages of data management and analysis.

Quantitative analysis

Objective 3: Content validity analysis

To assess the content validity of the innovation characteristics measurement tool, a systematic analysis will be conducted using Microsoft Excel. Two rounds of expert review will be carried out, involving a panel of subject matter experts who will evaluate each item for relevance, clarity, and representativeness. The Item-Level Content Validity Index (I-CVI) and the Content Validity Ratio (CVR) will be computed for each item following established guidelines. An I-CVI threshold of ≥0.78.²⁵ will be used to indicate excellent content validity, as recommended for tools evaluated by more than six experts. For CVR, the minimum acceptable value was determined based on Lawshe’s critical values; for example, a CVR threshold of ≥0.49 was applied when 15 experts participated.²⁵ Items falling below these thresholds were either revised for improved clarity and alignment with construct definitions or removed from the instrument. The analysis will be conducted in Excel using binary coding (1 = essential, 0 = not essential) for CVR and relevance ratings for I-CVI, ensuring transparency and reproducibility of scoring. This rigorous process ensured that the final tool demonstrated high content validity for use in evaluating innovation characteristics in healthcare implementation settings.

Objective 4:

The data will be collected during stage 4, coinciding with the definitive implementation of the intervention’s Opt TBD integrated care package.

Variables:

Independent variables: The independent variables in this study include, Evidence-based innnovation, Relative Advantage, Adaptability, Trialability and Complexity. Each of these variables is measured consistently across stage 4, ensuring a standardised approach to data collection.

Dependent variables: The process domain of CFIR has a total of four constructs: planning, engaging, execution, and reflection & evaluation. In this research study, we will use proxy measures to assess these constructs by utilising the Implementation Outcomes Questionnaire (2021).³⁰ According to this questionnaire and aligned with the objectives of our study using CFIR, the measure for planning will be appropriate, for engaging it will be adoption and acceptability, for execution it will be feasibility and fidelity (with fidelity being measured separately through a checklist), and the measure for evaluation will be penetration.

Analysis Plan: Initially, percentages, means, and standard deviations will provide a foundational understanding of the data distribution. Following this, inference analysis will be employed, utilising ANOVA for mean comparison and Post Hoc tests to further differentiate between the variables. The significance of associations will be determined using the P-value. Finally, regression analysis will be performed, employing Multinomial Logistic Regression, to assess the associations between exposure and outcomes, adjusting for a set of covariates. All the statistical data will be analysed using the software STATA version 14.0.

Fidelity: The Fidelity index scale will be assessed as frequencies and percentages using 95% of confidence intervals (CIs), aimed at the delivery of innovation in terms of a 3-point Likert scale that is 0-Missed, 1- Partial, and 2 complete. A higher Likert value indicates that innovation is being delivered according to the specifications as designed in terms of adherence or adaptability, quality, and quantity.

Integration of data

To ensure optimal results and achieve convergent validity between the quantitative and qualitative methods, data integration will be employed.²⁶ The data from phase four will be integrated with findings from phase one, which involved qualitative interviews with intervention developers. This integration of data aims to provide a comprehensive understanding of the intervention’s success. The results from phase three will validate the adherence to the innovation domain during the intervention’s development, suggesting that it exhibits lower complexity and higher acceptability and adoption rates among patients. Furthermore, these findings will confirm that interventions designed in alignment with the eight constructs of the innovation domain are more likely to achieve higher adoption and penetration rates within the healthcare system.

Dissemination

The findings from this study will contribute to the primary author’s doctoral dissertation and will be presented during their thesis defence at the university. Additionally, a summary report of the results will be compiled and shared with the TB programs in KP and Punjab, as well as with study participants upon request. The results will also be submitted for consideration in an international peer-reviewed journal.

Study status

OPT TBD care package is in definitive implementation stage at the moment.