Abstract
Introduction
Patient satisfaction is a critical indicator of healthcare quality and a key driver for improvement in hospital services. This study aimed to assess the levels and regional variations in inpatient satisfaction across all 20 Ministry of Health (MOH) regional health clusters in Saudi Arabia.
Methods
A nationwide retrospective cross-sectional analysis was conducted using anonymized survey data from 90,867 inpatient respondents collected through the MOH Patient Experience Measurement Program. The dataset captured patient experiences across multiple domains, including overall care, staff hygiene, physician communication, waiting times, and pain management. As data were ordinal (five-point Likert scale), non-parametric Kruskal–Wallis
H
-tests were employed to compare median satisfaction scores across regions, supplemented by
post-hoc
Dunn's tests with Bonferroni adjustment. Effect sizes (epsilon-squared, ε
2
= H/(n−1)) were calculated for all nine domains, and a Holm–Bonferroni correction was applied across the family of omnibus tests. To identify the domains most strongly associated with the global rating of care, and to partition variation between hospitals and clusters, a proportional-odds ordinal regression and a three-level mixed-effects logistic model with respondents nested within hospitals within clusters were additionally fitted.
Results
Differences between clusters were statistically significant in every domain (all Holm-adjusted
p
2
= 0.002–0.020, with ε
2
Discussion
Because most of the modeled variation lies between hospitals rather than between clusters, quality-improvement effort is better targeted at facility level than at cluster level; clusters are accordingly described here as showing higher or lower observed ratings rather than as high- or under-performing. National healthcare policies should focus on standardizing protocols and sharing best practices to ensure equitable, high-quality patient care across all regions.