Implementation of large volume paracentesis in an academic generalist setting to reduce cost of care

From the 2018 HVPAA National Conference

Cristin Colford (University of North Carolina Chapel Hill School of Medicine), Christina Mcmillan (University of North Carolina Chapel Hill School of Medicine), Lisa Iyer (University of North Carolina Chapel Hill School of Medicine), Scott Keller (University of North Carolina Chapel Hill School of Medicine), Meghan Black (University of North Carolina Chapel Hill School of Medicine), E. Allen Liles (University of North Carolina Chapel Hill), John Vargas (University of North Carolina Chapel Hill School of Medicine), Darren Dewalt (University of North Carolina Chapel Hill School of Medicine)

Background

Patients with refractory ascites suffer poor quality of life and poor outcomes1. These patients often present to the emergency department acutely symptomatic from increasing ascites and are admitted for bedside therapeutic large volume paracentesis (LVP). This procedure can be performed safely in the outpatient setting.2

Objectives

The UNC Internal Medicine Clinic (IMC) has a same day access during weekday hours staffed by faculty and resident physicians.We proposed that performance of LVP in this clinic setting could reduce utilization of resources and cost in our institution.

Methods

LVP is a core procedure in internal medicine residency training.  Many faculty members have practiced primarily in the outpatient setting and had not performed or supervised residents performing the procedure in many years. Ultrasound is increasingly utilized when performing LVP. We conducted faculty development sessions to increase faculty comfort with LVP using video review and hands on simulation using ultrasound and paracentesis kits. Identification of eligible patients came from established patients of the IMC practice who had ED and short hospital stays for LVP and through referral from the inpatient procedure service and the hepatology group.  Nursing staff were trained on the standard work needed to safely and efficiently complete an LVP.  Scheduling templates were developed and administrative staff trained to appropriately schedule patients needing the procedure.  The cost of implementing the process for LVP in clinic was primarily driven by the purchase an ultrasound for $30,000. Additional costs were in training time and administrative time for development of standard work processes. Estimation of cost was determined by reviewing professional and hospital charges and receipts for LVP procedures performed in the hospital on patients during the time of the study who also had a 0-2 day hospital stay.

Results

38 therapeutic LVPs were performed on 17 unique patients in the clinic in the first year of the project. There was one minor complication of a small abdominal wall hematoma that required no additional intervention. The average number of LVPs per individual patient was 2.3.  Four of the 17 patients were previously established patients of our practice. Five of the patients are now deceased, 2 are enrolled in hospice, 3 had procedures such as drains or shunts that reduced need for LVP, 2 are receiving routine LVPs in radiology or hepatology clinic, and the remaining 5 were lost to follow up. Performing one procedure in clinic resulted in receipts of $242 compared to $5,780 per 0-2 day hospitalization including an LVP. The estimated cost saving for the 38 LVPs performed in clinic is $210,444.

Conclusion

Performing an LVP in the academic general medicine clinic setting has allowed for increased flexibility of scheduling and has shown significant cost savings. We chose to be conservative in cost estimation by using receipts, rather than total charges. There are other savings that could be considered such as the opportunity cost of lost bed days. The cost savings would be seen by the health care system, not our individual practice, but has been enough to cover the cost of implementation in the first year of the program.

Implications for the Patient

Patients served by this program have high morbidity and mortality. The procedure was performed in the clinic with no major complications and is less expensive and more convenient for patients. Given the patient outcomes, the next step for this program may be to consider palliative care referral.