PainRelief.com Interview with:
Deepak Kumar, PT, PhD
Assistant Professor, Physical Therapy
Assistant Professor, BU School of Medicine
Director, Movement & Applied Imaging Lab

PainRelief.com:  What is the background for this study?  What are the main findings?

Response: We investigated the association of physical therapy interventions with long-term opioid use in people who undergo total knee replacement surgery.   For people with advanced osteoarthritis, total knee replacement is the only option. The number of total knee replacement surgeries has been increasing and is expected to rise exponentially over the next few years with an aging population and rising rates of obesity. However, up to a third of patients continue to experience knee pain after this surgery. Also, a significant proportion of people become long-term opioid users after total knee replacement. Reliance on opioids may reflect a failure of pain management in these patients. Given that physical therapy interventions are known to be effective at managing pain due to knee osteoarthritis, we wanted to study whether physical therapy before or after surgery may reduce the likelihood of long-term opioid use.

We used real-world data from insurance claims for this study. In our cohort of about 67,000 patients who underwent knee replacement between 2001-2016, we observed that, receiving physical therapy within 90 days before surgery or outpatient physical therapy within 90 days after surgery were both related to lower likelihood of long-term opioid use later. We also observed that initiating outpatient physical therapy within 30 days and 6 or more sessions of physical therapy were associated with reduced likelihood of long-term opioid use compared to later initiation or fewer PT sessions, respectively. However, we did not see an association between type of physical therapy. i.e., active (e.g., exercsise) vs. passive (e.g., TENS) and long-term opioid use.

Importantly, most of our findings were consistent for people who had or had not used opioids previously. We also were able to account of a larger number of potential factors that could confound these associations because of the large sample size. However, there are limitations to our work. Since we only had access to insurance claims data but not to health records, we are unable to make any inferences about association of physical therapy with pain or quality of life, etc.