Cultural Adaptation and Psychometric Assessment of Acceptability, Appropriateness, and Feasibility Measures for Family Planning Interventions Among Midwives and Nurses in Ghana

This study successfully culturally adapted and psychometrically validated the Acceptability, Appropriateness, and Feasibility measures for evaluating postpartum family planning interventions among Ghanaian midwives and nurses, finding that while the adapted tools are reliable, their abridged versions demonstrate superior construct validity compared to expanded versions.

The Gist

Imagine you are a chef who has perfected a famous soup recipe in Paris. The recipe is famous for measuring three things: Taste (do people like it?), Fit (does it work with local ingredients?), and Practicality (can you actually cook it in a busy kitchen?).

Now, you want to serve this soup in a bustling kitchen in Accra, Ghana. You can't just copy-paste the Paris recipe. The ingredients are different, the heat is different, and the chefs have different habits. If you hand them the Paris instructions without changing anything, they might say, "This doesn't make sense here," or "We can't do this with our tools."

This paper is about re-mixing that recipe for the Ghanaian kitchen and then testing if the new instructions actually work.

The Cast of Characters

• The Chefs: Midwives and family planning nurses in Ghana. They are the ones doing the hard work of talking to new mothers about birth control.
• The Soup (The Intervention): Two things they are trying to use:
  1. One-on-One Counseling: Sitting down and talking to a mother about her options.
  2. The MEC App: A mobile phone app that helps the nurse check if a specific birth control method is safe for a specific mother (based on World Health Organization rules).
• The Measuring Cups (The Tools): The researchers used three specific questions to measure the soup:
  1. Acceptability: "Do you like this? Is it comfortable?"
  2. Appropriateness: "Does this fit our job? Is it the right tool for the job?"
  3. Feasibility: "Can we actually do this without breaking a sweat?"

The Problem

The original "Measuring Cups" were made in the US and Europe. The researchers suspected that if they just handed those cups to the Ghanaian nurses, the measurements might be off. Maybe the word "feasible" means something slightly different to a nurse in a busy Ghanaian clinic than it does to a nurse in a quiet American hospital.

The Process: How They Fixed It

Step 1: The Taste Test (Qualitative Phase)
Before handing out the survey, the researchers gathered 18 nurses for a group chat (Focus Group Discussion). They didn't just ask, "Is this good?" They asked, "What does 'good' mean to you?"

• The Analogy: They realized that for these nurses, "Acceptable" didn't just mean "I like it." It meant, "I feel comfortable doing it, and I have the tools to do it."
• The Result: They took the original questions and translated them into the local dialect of work. They added new questions about comfort, offline access (since internet can be spotty), and how the app fits into a busy schedule.

Step 2: The Big Test (Quantitative Phase)
They gave the new, improved survey to 150 nurses who had actually used the app and the counseling method.

• The Goal: They wanted to see if the new questions were reliable. If a nurse said "Yes" to one question, did they say "Yes" to the others? Did the questions actually measure three different things, or were they all just asking the same thing in different ways?

The Results: Did the New Recipe Work?

1. The "Internal Consistency" (The Soup Tastes the Same)
The new survey was very reliable. When nurses answered, they were consistent. The "Taste," "Fit," and "Practicality" scores were high and stable. It was like a well-seasoned soup; every spoonful tasted the same.

2. The "Factor Analysis" (The Structure Holds Up)
The researchers ran complex math to see if the questions grouped together correctly.

• The Twist: The original 4-question version (from the US) didn't fit the Ghanaian data well. It was like trying to fit a square peg in a round hole.
• The Fix: The new, expanded version (with more questions) fit perfectly. It captured the reality of the Ghanaian nurses much better.

3. The "Predictive Validity" (Does it Predict the Future?)
This is the most important part. They asked: "Do the nurses who liked the new survey questions also actually want to use these tools?"

• The Answer: Yes!
  • Nurses who rated the counseling as "Acceptable" and "Feasible" said, "I want to work in a place where we do this."
  • Nurses who rated the App as "Feasible" said, "No, this app doesn't slow me down."
  • The Metaphor: The measuring cups were accurate. If the cup said "Good," the nurse actually thought it was good.

4. The "Abridged vs. Expanded" Surprise
Here is the tricky part. The researchers found that while the long version of the survey (with many new questions) was very detailed and reliable, the short version (the original 4 questions, slightly tweaked) was actually better at proving that the three concepts (Taste, Fit, Practicality) were truly different from each other.

• The Analogy: The long version was like a 20-page menu that described the soup in great detail. The short version was a simple 3-word tagline. The tagline was punchier and clearer for distinguishing the flavors, even though the menu had more information.

The Big Takeaway

This paper tells us that you cannot just copy-paste tools from rich countries to developing ones. You have to listen to the local workers, tweak the language, and test it again.

• Good News: The adapted tools work great in Ghana. They help us understand if new health tools (like apps) will actually be used by nurses.
• Lesson Learned: Sometimes, adding more questions makes the survey more detailed, but sometimes, sticking to the core, simple questions (the "abridged" version) gives you the clearest picture of what's really happening.

In short: The researchers took a foreign measuring cup, reshaped it to fit the Ghanaian hand, and proved that it now measures the success of family planning interventions accurately. This helps ensure that when new health tools are rolled out, they are actually helpful to the people using them.

Technical Summary

1. Problem Statement

Implementation research relies on accurate measurement of early outcomes—specifically acceptability, appropriateness, and feasibility—to determine whether an intervention will be adopted and sustained. However, instruments developed in high-income settings (such as the Acceptability of Intervention Measure [AIM], Intervention Appropriateness Measure [IAM], and Feasibility of Intervention Measure [FIM] by Weiner et al.) often lack cross-cultural validity when applied to different health systems.

In the context of postpartum family planning (PPFP) in Ghana, factors such as unique counselling routines, provider workloads, and service organization mean that unadapted measures may fail to capture provider experiences accurately. Furthermore, the integration of digital tools, specifically the World Health Organization (WHO) Medical Eligibility Criteria (MEC) mobile application, requires validation of how providers perceive these technologies within a low-resource, high-volume clinical setting. Without culturally adapted and psychometrically validated tools, researchers cannot reliably assess the implementation success of PPFP interventions in this context.

2. Methodology

The study employed a sequential exploratory mixed-methods design involving two distinct phases:

Phase 1: Qualitative Cultural Adaptation

• Setting: Three hospitals in the Greater Accra region (Achimota, Shai Osudoku, Pentecost).
• Participants: 18 midwives and family planning nurses across three Focus Group Discussions (FGDs).
• Process:
  • Researchers explored how providers interpreted the constructs of acceptability, appropriateness, and feasibility in their local clinical vocabulary.
  • Framework Analysis was used to code data, identifying themes that extended beyond the original AIM/IAM/FIM items.
  • Item Generation: New items were drafted based on recurring themes (e.g., "comfort," "offline capability," "simplicity") and reviewed by clinical leads for clarity and redundancy.
  • Outcome: An expanded survey instrument was created. For one-on-one counselling, the tool included 12 acceptability, 7 appropriateness, and 8 feasibility items. For the MEC App, it included 13, 7, and 7 items respectively.

Phase 2: Quantitative Psychometric Validation

• Setting: Three different hospitals where the MEC App was field-tested (Maamobi General, Ga West Municipal, Nsawam Government).
• Participants: 150 midwives and family planning nurses who had used the MEC App during routine postpartum counselling.
• Data Collection: Electronic surveys via REDCap using a 5-point Likert scale (0–4).
• Statistical Analysis:
  • Reliability: Cronbach's alpha for internal consistency.
  • Factorial Validity: Exploratory Factor Analysis (EFA) and Simultaneous Confirmatory Factor Analysis (SCFA) using Structural Equation Modeling (SEM) with Maximum Likelihood estimation.
  • Validity Checks: Predictive validity (Pearson correlations with preference questions), Convergent validity (Average Variance Extracted [AVE] ≥ 0.50, Composite Reliability [CR] ≥ 0.70), and Discriminant validity (Square root of AVE > inter-construct correlations).

3. Key Results

A. Reliability (Internal Consistency)

The adapted measures demonstrated strong internal consistency for both interventions:

• One-on-One Counselling: Cronbach's $\alpha$ ranged from 0.82 to 0.90.
• MEC App: Cronbach's $\alpha$ ranged from 0.86 to 0.93.
• These values were comparable to or higher than the original measures reported by Weiner et al.

B. Factorial Validity (Model Fit)

• Initial Models: The original 4-item structure (3 items per construct) showed poor fit ($\chi^2/df > 3$, CFI < 0.80) for both interventions.
• Adapted Models: After minor modifications (correlating error terms based on modification indices), the expanded adapted models achieved acceptable fit:
  • Counselling: CFI = 0.96, TLI = 0.95, RMSEA = 0.06.
  • MEC App: CFI = 0.88, TLI = 0.87, RMSEA = 0.08.
• The adapted models supported a unidimensional structure for each construct (acceptability, appropriateness, feasibility).

C. Predictive Validity

Significant associations were found between implementation outcome scores and provider preferences:

• Higher scores on counselling acceptability/appropriateness/feasibility correlated positively with the preference to work in facilities offering one-on-one counselling ($r = 0.36–0.49, p < 0.01$).
• Higher scores on MEC App feasibility correlated negatively with the belief that "computer applications slow down work" ($r = -0.22, p < 0.01$).

D. Convergent and Discriminant Validity

• Expanded Versions: While the expanded scales showed high reliability, they failed to consistently meet criteria for convergent and discriminant validity (specifically regarding AVE and MSV thresholds).
• Abridged Versions: When restricted to the original item counts (4 items for Acceptability; 3 items each for Appropriateness and Feasibility), the measures demonstrated robust convergent and discriminant validity for both interventions.
  • Example: For the MEC App, the abridged version showed CR values of 0.80–0.87 and AVE values of 0.57–0.70.

4. Key Contributions

1. Contextual Adaptation Framework: The study provides a replicable model for adapting implementation outcome measures in Low- and Middle-Income Countries (LMICs), demonstrating that qualitative insights (e.g., defining "feasibility" as "doable" or "offline-capable") are essential for semantic equivalence.
2. Psychometric Evidence in LMICs: It offers the first empirical evidence validating the AIM, IAM, and FIM among midwives in Ghana, confirming that these constructs are applicable to both traditional counselling and digital health interventions (mHealth).
3. Measurement Strategy Insight: A critical finding is the trade-off between content expansion and construct validity. While adding items increased cultural relevance and reliability (Cronbach's $\alpha$), it diluted the statistical distinctiveness of the constructs. The study concludes that abridged versions (retaining the original item count) offer superior construct validity compared to expanded versions in this specific context.
4. Digital Readiness Validation: The successful validation of the MEC App measures suggests that midwives in Ghana possess the digital literacy and readiness to evaluate mobile health tools using standard implementation science frameworks.

5. Significance

This research addresses a critical gap in implementation science by providing validated tools for measuring early implementation outcomes in resource-constrained settings.

• For Practitioners: It confirms that the WHO MEC App is perceived as acceptable, appropriate, and feasible by Ghanaian midwives, supporting its scale-up.
• For Researchers: It highlights the necessity of balancing contextual adaptation with measurement stability. The finding that abridged versions often outperform expanded versions in validity tests suggests that researchers should prioritize preserving the original measurement structure over adding numerous context-specific items unless those items are rigorously tested for construct distinctiveness.
• Policy Impact: The validated tools enable more accurate monitoring and evaluation of postpartum family planning programs, facilitating data-driven decisions to improve maternal and child health outcomes in Ghana and similar settings.

Limitations Noted: The study relied on self-reported data, lacked an independent validation sample for cross-validation, and did not include client perspectives. Future research should assess test-retest reliability and include patient-centered outcomes.