Propensity score-matching analyses on the effectiveness of integrated prospective payment program for patients with prolonged mechanical ventilation

•We provide evidence that patients of prolonged mechanical ventilation have a higher survival rate but have a lower weaning rate, relatively higher hospital days, ventilation days, and medical costs under integrated delivery system with prospective payment program.•We conducted 1:1 propensity-matching analyses using a nationwide database to investigate the effect of an integrated delivery system with a prospective payment program (IPP) on patients with prolonged mechanical ventilation (PMV).•Factors associated with successful weaning are integrated prospective payment program participation (HR = 0.84), age, income, presence of other catastrophic illness, comorbidity, presence of ventilator-associated complication, and healthcare organization level. An integrated delivery system with a prospective payment program (IPP) for prolonged mechanical ventilation (PMV) was launched by Taiwan's National Health Insurance (NHI) due to the costly and limited ICU resources. This study aimed to analyze the effectiveness of IPP and evaluate the factors associated with successful weaning and survival among patients with PMV. Taiwan's NHI Research Database was searched to obtain the data of patients aged ≥17 years who had PMV from 2006 to 2010 (N=50,570). A 1:1 propensity score matching approach was used to compare patients with and without IPP (N=30,576). Cox proportional hazards modeling was used to examine the factors related to successful weaning and survival. The related factors of lower weaning rate in IPP participants (hazard ratio [HR]=0.84), were older age, higher income, catastrophic illness (HR=0.87), and higher comorbidity. The effectiveness of IPP intervention for the PMV patients showed longer days of hospitalization, longer ventilation days, higher survival rate, and higher medical costs (in respiratory care center, respiratory care ward). The 6-month mortality rate was lower (34.0% vs. 32.9%). The death risk of IPP patients compared to those non-IPP patients was lower (HR=0.91, P