This study compares ordered logit and ordered probit regression models in identifying patient-level factors that influence his/her satisfaction with quality of health care. Patient satisfaction is a desirable outcome of a health system and is the most common metric used to assess the quality of healthcare provided. The data used was collected from public and private/faith-based health facilities in Kenya. The target population was the patients exiting the health facilities after receiving healthcare. The outcome variable was measured on a five-point ordinal scale: Very Satisfied, Somewhat Satisfied, Neutral, Somewhat Dissatisfied and Very Dissatisfied. The sociodemographic factors considered as influencing the outcome measure are the patient’s age, gender, education level and employment status as well as facility type and ownership of health facility. The study demonstrated that, provided that the proportional odds assumption is met, the ordered logit, and the ordered probit regression models produce relatively similar results. It also highlighted that patient’s age, gender, education level and employment status are significant in predicting his/her satisfaction with quality of healthcare received. Since the two models considered produce similar results, either of them can be used to model patient satisfaction data.

show more