Abstract
BACKGROUND: Despite the central role of resilience in promoting mental well-being, little is known about the resilience process among healthcare workers in East Africa, partly due to the lack of adequately standardised and psychometrically sound assessment tools for our context. This study provides the first psychometric evaluation of the Swahili version of the Connor-Davidson Resilience Scale 10 (CD-RISC 10) among nurses/midwives and community health volunteers (CHVs) in Tanzania, potentially informing the accuracy of measuring resilience in this setting. Additionally, we examine factors associated with self-reported resilience as measured by the CD-RISC 10 scale. METHODS: We used cross-sectional data on nurses/midwives (n = 950) and CHVs (n = 355). Cronbach's alpha and McDonald's omega were used to assess internal consistency. The one-factor structure of the scale was confirmed using confirmatory factor analysis (CFA). Gender and age invariance properties were evaluated using multi-group CFA. Item response theory was used to estimate discrimination and threshold parameters. Multivariable linear regression models were fitted to identify the factors associated with resilience. RESULTS: The scale had acceptable internal consistency (Cronbach's alpha and McDonald's omega > 0.70), was unidimensional, and was gender and age-invariant for nurses/midwives and CHVs. The scale measured resilience with high precision across a wide range of resilience continuum. There were statistically significant differences in sociodemographic variables between nurses/midwives and CHVs. Stronger mental well-being, as measured by the WHO-5 item scale, higher work engagement, higher perceived social support, and the male gender were associated with higher resilience levels, while higher perceived stress and burnout were associated with lower resilience levels among nurses/midwives. Higher work experience and mental well-being were associated with higher resilience levels, while higher burnout was associated with lower resilience levels among CHVs. CONCLUSION: The Swahili CD-RISC 10 scale demonstrates preliminary validity and reliability and may therefore be used to accurately measure resilience and assess the effectiveness of interventions aiming to improve resilience among nurses/midwives and CHVs in Tanzania. Sociodemographic and psychological measures may have different effects on resilience across different cadres, necessitating the development of cadre-specific interventions. More studies are required to validate the scale and confirm our findings across other Swahili-speaking countries and settings. CLINICAL TRIAL NUMBER: Not applicable.