Evaluating Hospital Performance Using a Multilevel Analysis Model in the Hospitals of Tabriz City During the Years 1396-1400

Abstract

Abstract Introduction: Accurate and high-quality evaluation of healthcare system performance is essential for improving outcomes and ensuring equitable, adequate, efficient, and effective services. Objective: The aim of this study was to evaluate hospital performance using a multilevel analysis model in the hospitals of Tabriz city from 2017 to 2021. Materials and Methods: This research was a retrospective, multi-method study conducted on 1905 patients admitted to 15 hospitals in Tabriz over five years. The predictive power of patient, physician, and hospital-level variables on selected hospital performance indicators was determined using simple, two-level, and multilevel analysis models. The final model was developed using expert panel input. Results: Results from the single-level analysis showed that age, insurance type, and disease type had significant relationships with the length of stay and leave against medical advice. For the hospital mortality indicator, age and certain diseases were associated with increased mortality risk. In the two-level analysis, patient-level variables remained associated with length of stay and mortality risk. Additionally, physician specialization and type of collaboration were significantly related to the length of stay. In the three-level model, hospital specialization and the number of active beds at the hospital level had significant relationships with the length of stay. No significant relationships were found at the physician level. At the patient level, age, gender, insurance type, and disease type impacted the length of stay and leave against medical advice. Age and disease type were also related to mortality risk. Finally, a model was designed and presented to identify the factors influencing hospital performance. Conclusion: The findings of this study revealed that hospital performance is influenced by a combination of various factors at all three levels including patient, physician, and hospital. Hospital performance cannot be predicted by considering only the factors from a single level. The comprehensive model presented can be used for more accurate hospital performance evaluations.

Keywords: Performance evaluation, hospital, multilevel analysis model