Shop Talk

Discharge to Home Sweet Home after Joint Replacement

Today we’ll look at a paper I spotted in the newsletter of the Foundation for Physical Therapy Research (FPTR). It’s an article by Amit Kumar and colleagues on discharge to home after joint replacement surgery. The piece got a shout-out because one of the co-authors is a past recipient of grant funding from the FPTR. We’re using the paper to talk about strengths and limitations of retrospective studies.

What’s the Kumar paper about?

Kumar’s group wanted to know if discharge to home after hip and knee replacement surgery was associated with the amount of rehabilitation delivered during acute hospitalization. They examined the Medicare records of 640,901 patients, figured out whether they were discharged to home or someplace else, and quantified the amount of rehabilitation they got in the hospital. They found that patients with hip replacement who received >65 minutes/day of rehabilitation were 1.76 times as likely to be discharged home as those who got <40 minutes/day. Patients with knee replacements were 1.38 times as likely to be discharged home if they got >68 minutes/day as compared to <45 minutes per day. The authors concluded that amount of rehabilitation is associated with discharge home.

Quick knowledge check:

1. Your grandpa is having his hip replaced.
He wants to go home after surgery.
How much inpatient rehabilitation will you advocate for?

A) >65 minutes/day

B) <40 minutes/day

2. Grandpa’s surgery is over.
He is complaining about the therapists who keep asking him to exercise.
What do you say?

A) You’re right, Grandpa. Therapy is a real bother. You should take a nap.

B) Do the therapy, Gramps.
People who get at least 65 minutes of daily therapy after hip replacement are nearly twice as likely to go home as those who get fewer than 40 minutes.

Pros and cons of retrospective studies:

One of the major strengths of retrospective studies is sample size. Big samples are good because they are more likely than small samples to represent the population. Kumar et al. were able to examine over 600,000 cases because they were willing to use data that already existed. Imagine collecting 600,000 samples of anything all by yourself!

But to get the big sample, they gave up control. For example, they could not dictate how much therapy was provided. Amount was sorted out after the fact by placing patients in to 3 groups. For patients with hip replacement, the groups were <40 minutes (low), 40-65 minutes (medium), and >65 minutes (high), which means that some patients in the low group could have received only 1 minute less than some in the medium group. Perhaps the association between rehabilitation and discharge to home would be even stronger if there had been large distinctions between groups, like 15, 45 and 90 minutes.

On the other hand, the inability to control the data set and how it was collected may lead us to inflate the value of rehabilitation. Though the authors do not claim it, one interpretation of their result is that therapy helps get people home, and more is better. But I noticed that presurgical living status (home, not home) was one variable the authors could not control. I can imagine a scenario where patients living at home prior to surgery get more therapy than those not living at home. A therapist might reason that those living at home are likely to return home and should get as much therapy as possible to achieve that goal. Those not living at home might get less therapy, given that they are expected to return to a residential facility where the demands for independence will be lower than at home. If this were the case, one could argue that discharge to home is determined by presurgical living status, not amount of therapy.

With respect to this study, we may never know. Like many retrospective studies, this one relinquished control of this and other variables in exchange for a huge data set that was waiting to be mined for insight into an important question.

Before you go, here’s a note for students and early career professionals.

On Monday, March 28, at 8 pm EST, the FPTR is hosting a free virtual event called, "You've Got the Job, What's Next?" It’s part of the VCU-Marquette Challenge, and it’s meant to help graduating PTs and PTAs prepare for their first clinical jobs. Here’s the link to register: 

Thanks for reading. As always, you can learn more by checking out the resources below.

1. Read the whole article on amount of therapy and discharge to home by Kumar et al. at 

Here’s the citation.

Amit Kumar, MPH, PhD, Indrakshi Roy, MS, PhD, Meghan Warren, PT, MPH, PhD, Stefany D Shaibi, DPT, Maximilian Fabricant, DPT, Jason R Falvey, PT, PhD, Amit Vashist, MD, Amol M Karmarkar, PhD, Impact of Hospital-Based Rehabilitation Services on Discharge to the Community by Value-Based Payment Programs after Joint Replacement Surgery, Physical Therapy, 2022;, pzab313,

2. Check out the newsletter of the FPTR at

3. The answers to the knowledge check are
1) A
2) B

March Edition

Mar. 19, 2022

If you appreciate what we’re doing, please consider supporting our work -