Case Study: Hospital readmissions

Case Study: Hospital readmissions#

The Hospital Readmissions Reduction Program, which is included in the Affordable Care Act (ACA), applies financial penalties to hospitals that have higher-than-expected readmission rates for targeted conditions.

Zuckerman, R. B., Sheingold, S. H., et al. (2016). Readmissions, observation, and the hospital readmissions reduction program. NEJM, 374(16), 1543-1551.

2-minute activity#

Write down one question you may be interested in answering:

  1. if you were a patient.

  2. if you were a hospital financial advisor.

  3. if you were an insurance company administrator.

Add your answers to https://padlet.com/yuchan2021/readmissions.

10-minute group activity#

Form a group of 2 to 3.

  1. Exchange your ideas about the questions you wrote down.

  2. Pick and/or consolidate to ONE questions to focus on.

  3. For each question, answer the following:

    • What kind of data do you need to answer this question?

    • How would you find such data?

Upload your answers in In-class exercise 1 on Canvas.

Total knee arthroplasty#

Suppose now we are only concerned about patients who have had a total knee replacement. We are given the following dataset descriptions for Michigan (State Inpatient Dataset) in 2012.

We are interested in answering three questions:

  1. How often are patients readmitted within 30/90 days after a total knee replacement, respectively?

  2. What are the major reasons for their readmissions?

  3. How much are the costs associated with their readmissions?

There are two important aspects here:

  1. Do we have the data? (Data availability)

    • In other words, can we answer these questions with the given data?

    • Which variables in the dataset are relevant?

  2. How do we answer the questions? (Data computability)

    • Outline the steps you would take, and be specific.

    • What are some challenges you anticipate?

State Inpatient Dataset#

The State Inpatient Dataset (SID) contains \(>99\%\) inpatient hospital records that occurred within the year 2012.

Data Set Name: MI_SIDC_2012_CORE  
Number of Observations: 1249805  
Total Record Length: 1178  
Total Number of Variables: 341

The variables are viewable in https://hcup-us.ahrq.gov/db/state/sidc/tools/filespecs/MI_SID_2012_CORE.loc.