# Cross-cultural adaptation: South African Gestational Diabetes Mellitus Knowledge Questionnaire

## TL;DR

This study translated and adapted a gestational diabetes knowledge questionnaire for use in South Africa and tested its reliability and validity in three local languages.

## Contribution

The study provides a cross-culturally adapted GDM knowledge questionnaire in English, Afrikaans, and isiXhosa with initial psychometric evaluation.

## Key findings

- The three SA-GDMKQ versions showed good face and content validity.
- IsiXhosa version had the highest internal consistency (Cronbach’s alpha of 0.621).
- All versions showed a statistically significant positive correlation between total scores.

## Abstract

Many serious adverse events associated with gestational diabetes mellitus (GDM) can be mitigated by timely glucose control during pregnancy, achieved through education and lifestyle choices. The aim of this study was to translate and cross-culturally adapt and test the preliminary internal consistency and test-retest reliability of the South African English, Afrikaans and isiXhosa versions of the GDM Knowledge Questionnaire (GDMKQ).

A prospective, observational study was conducted at a high-risk antenatal clinic in South Africa. Pregnant women ≥ 18 years with GDM were consecutively sampled. Semantic equivalence between the original and adapted versions was assessed. Face and content validity, internal consistency and test–retest reliability were evaluated.

The three SA-GDMKQ versions demonstrated good face and content validity. For internal consistency, Cronbach’s alpha values were 0.534 for the Afrikaans version, 0.434 for the English version and 0.621 for the isiXhosa version. Test–retest reliability found kappa (standard error [s.e.]) values ranged between −0.03 (0.18) and 0.89 (0.13) for the English version, between −0.07 (0.18) and 0.53 (0.13) for the Afrikaans version and between 0.28 (0.18) and 0.87 (0.17) for the isiXhosa version. All versions of the SA-GDMKQ had a statistically significant (p < 0.001) positive linear correlation between the total scores.

The English, Afrikaans and isiXhosa SA-GDMKQ versions were found to be feasible and easy to comprehend, although lower internal consistency and test–retest reliability were displayed. Further validation of the psychometric properties of the English, Afrikaans and isiXhosa versions of the SA-GDMKQ among larger sample groups is however warranted.

This study adds to the knowledge around developing and using culturally appropriate questionnaires and outcome measures in research and clinical practice.