Psychometric evaluation of the NoMAD instrument in cancer care settings: assessing factorial validity, measurement invariance, and differential item functioning.

Abstract

Background: The Normalization MeAsure Development (NoMAD) questionnaire is used to assess implementation processes based on Normalization Process Theory (NPT). However, its psychometric properties have not been extensively evaluated. This study aimed to examine the factorial validity, internal consistency, and measurement invariance at both scale and item levels of the NoMAD across three hybrid effectiveness-implementation studies determining the impact of routine symptom surveillance and guideline-based symptom management interventions in ambulatory oncology care settings.

Methods: A cross-sectional survey was conducted with 328 healthcare personnel (74.% clinicians) participating in the Improving the Management of SymPtoms during And following Cancer Treatment (IMPACT) Research Consortium between 2019 and 2024. Confirmatory factor analysis (CFA) tested the hypothesized four-factor structure (coherence, cognitive participation, collective action, reflexive monitoring). Internal consistency was assessed with McDonald's omega and Cronbach's alpha coefficients (> 0.70 acceptable). Measurement invariance was tested across research centers, professional roles, and years in current roles using multi-group CFA. Model fit was defined by standard fit indices (Comparative Fit Index (CFI) and Tucker-Lewis Index (TLI) values ≥ 0.95, Root Mean Square Error of Approximation (RMSEA) values ≤ 0.06, and Standardized Root Mean Square Residual (SRMR) values ≤ 0.08. Differential item functioning (DIF) was evaluated using ordinal logistic regression and item response theory methods (ΔR2 ≥ 0.02 indicative of meaningful DIF).

Results: The four-factor model demonstrated good fit to the data (CFI = 0.97, TLI = 0.96, RMSEA = 0.06, SRMR = 0.05). All factor loadings were statistically significant (p < 0.001), ranging from 0.606 to 0.871. Internal consistency was satisfactory for all four constructs (Omega range: 0.789-0.864, Cronbach's alpha range: 0.782-0.863). The NoMAD exhibited configural, metric, and scalar invariance across research centers, roles, and years in the current role. One item ("The staff agree that the intervention is worthwhile") showed uniform DIF across healthcare systems (ΔR2 = 0.047), but no DIF was found by role or years in the current role.

Conclusions: This study supports the factorial validity, internal consistency, and measurement invariance of the NoMAD across three oncology implementation efforts. The presence of DIF in one item provides an opportunity for refinement in this healthcare context. Researchers and practitioners can use the NoMAD to assess and compare implementation processes, informing the development and evaluation of implementation strategies.

Trial registration: (ClinicalTrials.gov ID NCT03850912, NCT03892967, NCT03988543).