Translation and psychometric validation of the Mental Illness: Clinicians Attitudes Scale (MICA-4) to assess attitudes of primary care physicians in Pakistan

This study successfully translated and psychometrically validated the Mental Illness: Clinicians Attitudes Scale (MICA-4) into Urdu for use among primary care physicians in Pakistan, identifying a three-factor structure that supports its cautious application despite some limitations in internal consistency and cross-cultural stability.

The Gist

Imagine you are trying to fix a broken machine, but the people who are supposed to be the mechanics (the doctors) are afraid of the machine, think it's cursed, or believe it's broken beyond repair. If the mechanics don't believe the machine can be fixed, they won't try very hard to fix it, and the machine stays broken.

This is exactly what happens with mental health care in many places, including Pakistan. Many patients don't get the help they need because their doctors, who are the first people they see, might have negative attitudes or "stigma" toward mental illness. They might think, "Mental illness isn't a real sickness," or "These patients are dangerous."

To fix this, we first need to measure how the doctors feel. But you can't just ask them, "Do you have bad feelings?" You need a ruler that is accurate and fits the specific shape of the country you are measuring.

The Story of the "Ruler" (The MICA-4 Scale)

The researchers in this paper were trying to use a famous measuring tool called the MICA-4. Think of the MICA-4 as a universal ruler designed in the UK to measure how doctors feel about mental illness. It has 16 marks (questions) on it.

However, you can't just take a ruler made for the UK and try to measure a Pakistani doctor's hand with it. The language is different, the culture is different, and the way people think about "madness" or "illness" is different. If you use the ruler without adjusting it, the measurements will be wrong.

So, the team led by Dr. Noor ul Ain Muneeb decided to translate and rebuild this ruler specifically for Pakistani primary care doctors.

The Process: How They Built the New Ruler

1. Translation (The Language Swap):
   They took the original English questions and had three experts translate them into Urdu. But it wasn't just a word-for-word swap. They had to make sure the feeling of the question stayed the same.

   • Analogy: Imagine translating a joke. If you translate the words literally, it might not be funny anymore. You have to find the equivalent joke that makes people in the new country laugh the same way. They did this with the questions to ensure the doctors understood them clearly.

2. The "Test Drive" (Cognitive Interviews):
   Before using the ruler on everyone, they asked 15 doctors to "test drive" the questions. They asked, "What do you think this question means?"

   • Result: They found some questions were confusing, so they tweaked the wording slightly to make them clearer, like polishing a foggy window.

3. The Big Experiment (The Two Groups):
   They split the doctors into two groups to test the ruler:

   • Group 1 (191 doctors): They asked them the questions to see what patterns emerged. It was like looking at a pile of puzzle pieces to see how they fit together naturally.
   • Group 2 (329 doctors): They used this group to confirm that the pattern they found in Group 1 was real and not just a fluke.

The Surprise: The Ruler Changed Shape

When they looked at the data, they found something interesting. The original ruler was designed to have 5 sections (factors). But when they used it in Pakistan, the pieces didn't fit that way.

Instead, the Pakistani doctors' answers naturally grouped into 3 distinct sections:

1. Views: How doctors see their role and whether they think mental health is a "real" job.
2. Stereotypes: The fixed, often negative beliefs they hold (e.g., "People with mental illness can never recover").
3. Stigma: The fear and desire to keep distance (e.g., "I wouldn't want to work with someone who has a mental illness").

The "Broken Pieces":
Just like a puzzle where some pieces are warped and don't fit anywhere, they found that 4 specific questions (items 6, 9, 12, and 13) were confusing or didn't measure anything useful in this context.

• Analogy: Imagine a ruler where the "5-inch" mark is actually at 4.5 inches, and the "10-inch" mark is at 11 inches. If you use it, your measurements are wrong. The researchers had to cut out those 4 bad marks and create a new, cleaner 12-mark ruler.

What Did They Learn?

• The Ruler Works (Mostly): The new, shortened Urdu version of the MICA-4 is a good tool for measuring attitudes in Pakistan. It fits the cultural "hand" of Pakistani doctors.
• No Magic Demographics: They expected that younger doctors or female doctors might have better attitudes, but they found that everyone had similar levels of stigma. This suggests that the problem isn't about age or gender; it's a systemic issue in how medical training is done in Pakistan.
• One Size Doesn't Fit All: The biggest lesson is that a "universal" ruler doesn't work everywhere. Different cultures hold different beliefs. In Pakistan, the fear of mental illness is deeply tied to cultural and religious beliefs (like thinking it's a spiritual issue rather than a medical one), which changes how the questions land.

The Bottom Line

This paper is a guidebook for anyone trying to improve mental health care in Pakistan. It says: "Don't just copy-paste tools from the West. Adapt them, test them, and if some parts don't fit, cut them out."

By creating a culturally accurate ruler, the researchers can now accurately measure how much stigma exists among doctors. Once we can measure the problem accurately, we can finally start fixing it—by training doctors to understand that mental illness is real, treatable, and that their patients deserve the same care as anyone with a broken leg.

Technical Summary

1. Problem Statement

• Context: Mental disorders constitute a significant global burden, yet over 90% of individuals in Low- and Middle-Income Countries (LMICs) lack access to evidence-based care. In Pakistan, Primary Care Physicians (PCPs) are often the first and only point of contact for patients, yet they frequently lack mental health training and hold stigmatizing attitudes.
• The Gap: Negative attitudes among healthcare providers (HCPs) act as a barrier to care, leading to delayed help-seeking and poor treatment adherence. While tools like the Community Attitudes Toward the Mentally Ill (CAMI) exist, they are designed for the general public, are lengthy, and lack validity for clinical populations.
• Objective: The study aimed to translate, culturally adapt, and psychometrically validate the Mental Illness: Clinicians' Attitudes Scale (MICA-4) specifically for Primary Care Physicians (PCPs) in Pakistan. The goal was to create a robust, culturally tailored instrument to measure stigma and negative attitudes in this specific demographic.

2. Methodology

• Study Design: A cross-sectional design utilizing non-probability sampling.
• Participants: Two independent samples of PCPs were recruited:
  • Sample 1 (n=191): Used for Exploratory Factor Analysis (EFA).
  • Sample 2 (n=329): Used for Confirmatory Factor Analysis (CFA).
• Translation Protocol:
  • Followed WHO guidelines for translation.
  • Forward Translation: Conducted by three bilingual mental health professionals (two clinical psychologists, one psychiatrist).
  • Expert Review: An expert panel resolved discrepancies to ensure conceptual equivalence.
  • Back-Translation: Independently performed by a bilingual clinical psychologist to verify consistency with the original English version.
  • Cognitive Interviews: Conducted with 15 PCPs to assess item comprehension and clarity, leading to minor linguistic refinements (e.g., adding parenthetical examples for clarity without altering meaning).
• Statistical Analysis:
  • Software: SPSS v.29 and R (Lavaan package).
  • EFA: Principal Axis Factoring with Promax rotation to determine the underlying factor structure.
  • CFA: Diagonally Weighted Least Squares (DWLS) estimator was used to handle ordinal data and non-normality. Model fit was assessed using CFI, TLI, SRMR, and RMSEA.
  • Reliability & Validity: Assessed via Cronbach's alpha, McDonald's omega ($\omega$), Average Variance Extracted (AVE), and Heterotrait–Monotrait (HTMT) ratios for discriminant validity.

3. Key Results

• Factor Structure:
  • Initial attempts to replicate the original 5-factor model or previously reported 3-factor models failed to yield admissible solutions.
  • EFA revealed a three-factor solution explaining ~30% of the variance.
  • Item Reduction: Items 6, 9, 12, and 13 were removed due to weak loadings (< .40), cross-loadings, or low item-total correlations.
  • Final 12-Item Scale Structure:
    1. Views (7 items): Reflects clinicians' general evaluative perspectives on mental illness as a profession and their comfort in interacting with patients (e.g., "Working in mental health is respectable").
    2. Stereotypes (5 items): Represents generalized beliefs and disclosure concerns (e.g., "People with severe mental illness can never recover," fear of admitting one's own illness).
    3. Stigma (3 items): Captures social distancing and perceived threat (e.g., "People with severe mental illness are dangerous"). Note: This factor showed lower reliability due to having only 3 items initially, and further analysis suggested 2 items were retained for the final stable model in some contexts, though the abstract mentions 3.
• Model Fit (CFA):
  • The 3-factor model demonstrated good fit: CFI = .958, TLI = .946, SRMR = .085.
  • RMSEA = .108 was elevated but interpreted with caution due to the DWLS estimator and limited degrees of freedom.
• Reliability:
  • Views: $\omega = .70$ (Adequate).
  • Stereotypes: $\omega = .74$ (Adequate).
  • Stigma: $\omega \approx .53$ (Lower reliability, attributed to the small number of items).
• Validity:
  • Convergent Validity: Modest (AVE values ranged from .40 to .44).
  • Discriminant Validity: Confirmed via HTMT ratios, despite a high correlation between "Views" and "Stereotypes" ($r = .86$).
• Demographics: No significant differences in attitudes were found across gender, age, education, or years of experience, suggesting pervasive stigma across all PCP demographics in Pakistan.

4. Key Contributions

• Cultural Adaptation: Successfully translated and adapted the MICA-4 for the Pakistani context, addressing the lack of validated tools for HCPs in LMICs.
• Structural Refinement: Demonstrated that the MICA-4 factor structure is not universal. The study provided evidence that a 3-factor model (Views, Stereotypes, Stigma) is more appropriate for the Pakistani context than the original 5-factor model, necessitating the removal of specific items (6, 9, 12, 13) to achieve stability.
• Methodological Insight: Highlighted the necessity of "limited, evidence-based refinement" (item removal/modification) when validating scales across diverse cultures to ensure measurement stability, rather than rigidly adhering to the original item set.
• Baseline Data: Provided the first psychometrically validated baseline for assessing stigma among PCPs in Pakistan, a critical step for designing targeted training interventions.

5. Significance and Implications

• Clinical Practice: The validated Urdu MICA-4 can now be used by policymakers and mental health organizations in Pakistan to screen PCPs for stigmatizing attitudes and evaluate the efficacy of mental health training programs.
• Policy: The findings underscore the urgent need for integrating mental health training into pre-service medical education and continuing professional development in Pakistan.
• Global Context: The study contributes to the broader literature on cross-cultural validation of stigma scales, reinforcing the idea that the dimensionality of stigma instruments is context-sensitive. It suggests that for the MICA-4 to be effective globally, researchers must be open to structural modifications based on local cultural and healthcare system realities.
• Limitations: The authors acknowledge the use of non-probability sampling (limiting generalizability) and the modest reliability of the "Stigma" subscale, suggesting future iterations should expand this subscale.

Conclusion: The study concludes that the Urdu MICA-4 is a valid tool for assessing clinician attitudes in Pakistan, provided it is used in its refined 12-item, 3-factor form. It emphasizes that while conceptual equivalence is vital, structural flexibility is often required to achieve psychometric robustness in diverse cultural settings.