Author: Rei Mitsuyama, MS2, Brown University

Editor: David Petrov, MD, PGY-3, Allegheny Health Network – Temple University School of Medicine 


Lowering readmission rates is one of the top priorities of US healthcare reform. Early readmission after acute care is associated with lower quality of care in the index admission and results in significant increases in health care costs [1]. For these reasons, these data are used by organizations such as the US Hospital Readmission Reduction Program to determine reimbursements with the goal of improving quality and reducing costs. 

While data on readmission rates are available for a wide variety of surgical procedures, there are currently limited data on readmissions following interventional radiology (IR) procedures. Furthermore, it is unclear whether metrics used to describe non-IR procedures are applicable to IR procedures. To address this, it is first necessary to identify the time frame during which most readmissions occur and whether these time periods represent a population at risk for early mortality. Determining which procedures are associated with early readmissions and if being readmitted increases 90-day mortality may provide insight into where there is room for quality improvement. 

This retrospective cohort analysis investigated readmission rates following IR procedures, timing of readmissions, and the association between readmissions and 90-day mortality [2]. 

Study Population 

This study was conducted at a single tertiary care center between June 2008 and May 2013. Data were collected from 3653 procedures in 3189 consecutive patients who received 1 of 12 selected IR procedures: (lower extremity angioplasty, arteriovenous fistulagram, port catheter placement, transjugular intrahepatic portosystemic shunt (TIPS) placement, transjugular and percutaneous liver biopsy, vascular embolization other than uterine fibroid embolization, inferior vena cava (IVC) filter placement, percutaneous cholecystostomy, percutaneous transhepatic biliary drainage (PTBD), percutaneous feeding tube placement, and primary urinary drainage). These procedures were chosen because they may be used to treat the cause of index admission; they are high-volume procedures at the authors’ institution; and they can be categorized by CPT codes [2]. 202 patients died during index admission and thus were not included in this study. Patient information including the procedure performed was collected for the 2897 who survived, and subsequent readmissions were recorded. 

Data Collection and Stats 

The primary outcome of interest was first readmission up to 180 days following index admission, where index admission is defined as the admission during which the selected procedure was performed. Time-to-readmission (TTR) was recorded as no readmission, 0-7 days, 8-30 days, 31-60 days, 61-90 days, or 91-180 days. Readmission rates were calculated for each TTR for each of the 12 selected procedures. 30-day readmission rates > 15% were considered to be high. This is based on the fact that the 20 non-IR procedures with the highest all-cause 30-day readmission rates for Medicare beneficiaries have rates between 15.1 and 29.1% [3]. Finally, 90-day mortality was determined for all selected procedures. Adjusted odds ratios (AORs) were calculated relative to patients receiving the same IR procedure who were not readmitted. 

Results and Outcomes 

1,479 patients (40.5%) did not require readmission and the readmission rate was highest between 8 and 30 days (17.8%). 30-day readmission rates were high (>15%) for all procedures, with the highest rates seen in port catheter placement (50.5%), transjugular liver biopsy (43.4%), PTBD (38.5%), percutaneous cholecystostomy (31.9%), and TIPS (31.3%). The 1-year mortality rate was 26.8% (855 of 3,189). Although 90-day mortality rates were greater in readmitted patients, the association with TTR varied widely between procedures. Using an AOR, 30-day readmission was associated with increased 90-day mortality for 5 of the 12 studied procedures: lower extremity angioplasty (AOR 3.19; P=.02), vascular embolization (AOR 10.01; P <.001), IVC filter placement (AOR 2.98; P = .01), percutaneous liver biopsy (AOR 2.86; P <.001), and primary urinary drainage (AOR 3.09; P =.01). 


The 30-day readmission rates for all procedures studied were as high as readmission rates for the 20 procedures with the highest readmission rates in the Medicare population (> 15%) [3]. This suggests that a 30-day threshold is relevant to inpatient IR procedures. However, IR procedures are often performed on patients whose diseases or illnesses have an intrinsically high baseline readmission rates. Similarly, IR procedures are often performed in conjunction with complex surgical procedures which have high readmission rates in themselves. Readmission for any reason was correlated with increased 90-day mortality for all but one of the studied procedures. This study did not investigate causality or preventability of readmission or early death. Thus, determining the etiology of readmissions and whether they are preventable could provide an opportunity to decrease both readmissions and mortality. While the relationship between procedures, readmissions, and mortality is highly complex, this study identifies several areas for future research. 


  • Ashton CM, Del Junco DJ, Souchek J, Wray NP, Mansyur CL. The association between the quality of inpatient care and early readmission: a meta-analysis of the evidence. Med Care 1997;35:1044–1059. doi: 
  • Sarwar A, Zhou L, Chakrala N, Brook OR, Weinstein JL, Rosen MP, Ahmed M. The Relevance of Readmissions after Common IR Procedures: Readmission Rates and Association with Early Mortality. J Vasc Interv Radiol 2017;28:629–636. doi: 
  • Weiss AJ, Elixhauser A, Steiner C. Readmissions to U.S. hospitals by procedure, 2010. 2013. Available at: