Drivers of perceived healthcare service quality among healthcare workers in TQM-implementing hospitals in conflict-affected yemen: a cross-sectional study

Abstract Background Healthcare systems in protracted conflicts face compounded pressures that challenge service quality. In 2021, Yemen’s Ministry of Public Health mandated the implementation of total quality management (TQM) across hospitals. This cross-sectional study examined the association between TQM implementation and perceived healthcare service quality from the provider’s perspective, framing the findings as hypothesis-generating rather than causal. Methods A total of 410 healthcare professionals from four TQM-implementing hospitals in Sana’a completed a validated questionnaire measuring five European Foundation for Quality Management (EFQM)-based TQM dimensions and five SERVQUAL-adapted provider proxy service quality dimensions. Multiple linear regression with robust standard errors (Huber–White), bootstrap mediation, and moderated regression were used in the analysis. Results Perceived TQM implementation and service quality were rated relatively high on a 7-point scale (overall TQM median = 5.48, IQR = 1.28; overall perceived service quality median = 5.82, IQR = 1.39). The multiple linear regression model was statistically significant (R² = 0.651, F(5, 404) = 102.22, p < .001). Processes (β = 0.458, 95% CI [0.269, 0.647], p < .001), policy and strategy (β = 0.283, 95% CI [0.129, 0.438], p < .001), and partnerships and resources (β = 0.144, 95% CI [0.041, 0.246], p = .006) were significant positive predictors of perceived service quality. Leadership commitment showed no significant direct association (β = 0.009, p = .873); post-hoc mediation indicated that its effect on service quality operated largely through processes (indirect effect = 0.506, 95% CI [0.412, 0.617], 82.4% mediated) and policy and strategy (indirect effect = 0.489, 95% CI [0.384, 0.605], 79.7% mediated). Human resources showed a marginally significant negative trend (β = −0.089, 95% CI [− 0.179, 0.001], p = .051). Gender significantly moderated four of the five relationships (all p < .05), with steeper TQM–quality slopes among female than among male staff (e.g., processes: β = 0.843, 95% CI [0.771, 0.915] vs. 0.625, 95% CI [0.515, 0.736]), whereas experience was not significant. Conclusions Process-oriented, policy-oriented, and partnership-oriented TQM dimensions were positively associated with perceived service quality in this conflict-affected setting, whereas leadership influence appeared indirect. These cross-sectional, provider-reported findings generate hypotheses for future longitudinal and mixed-methods research incorporating patient-reported outcomes and objective clinical indicators.