Skip to main content

How the Integration of Specialty Pharmacies Into ACOs Helps Lower Costs

November 12, 2020

By Julie Gould

apurv soni headshotIn a research letter published online in JAMA Network Open, study authors sought to better understand how the integration of specialty pharmacies in an Accountable Care Organization (ACO) impacts total medical expenditures.

“Use of integrated specialty pharmacies within ACOs optimizes medication adherence, increases care coordination with physicians, and reduces medication-related adverse events,” the study authors explained. “They may also decrease health care costs for patients because medication coordination and fulfillment could reduce adverse events and improve underlying conditions, which in turn decreases health care visits.”

To better understand how the cost of care changed among patients enrolled in an ACO who used an integrated specialty pharmacy, we spoke with Apurv Soni, BA, MD‑PhD candidate at University of Massachusetts Medical School. He explains why increased coordination and multidisciplinary elements of care is helpful. 

What existing data led you and your co‑investigators to conduct this research?

Some of this was based on clinical anecdotes and information that was provided by the stakeholders in the ACO, as well as in the hospital system, where they were suggesting the use of integrated specialty pharmacy was seeming to be cost‑effective, but there was a need for estimating the magnitude of savings through robust analysis..

Based on that, there was a coalition of investigators that was formed with representation from Shields Pharmacy,  which is the pharmacy provider that helps integrate the specialty pharmacy for ACO patients that are part of the UMass Medicare ACO. We had Brian Smith and Bill McElnea from Shields working with Drs. Tom Scornavaca, Eric Dickson, and Alice Shakman from the hospital side. My role—as someone who is a health science researcher—along with my mentor, Dr. David McManus, was to be the principal investigators for this study in terms of designing the study and analyze the data using rigorous statistical methodology to produce unbiased quantitative estimates for how total medical expenditure varied based on use of integrated specialty pharmacy.

Can you please briefly describe the study and its findings? Of those findings, were any of them particularly surprising?

The main objective of this study was to look at whether or not integrating specialty pharmacy is associated with cost savings for patients that are enrolled in the UMass Memorial Medicare Accountable Care Organization.

This is a subset of patient population that are Medicare‑eligible and are enrolled in Medicare, and then have also opted to be part of accountable care organization that UMass Memorial Hospital System, which is a network of hospitals in Central Massachusetts. It's the predominant health care provider in Central Massachusetts.

Among that population, we wanted to look at patients that are getting specialty care, whether or not those that enroll and sign up for integrated specialty pharmacy, have a higher cost savings in comparison to those that get their specialty medications from unintegrated pharmacy providers.

Integrated specialty pharmacy provides provisions to be able to do more care coordination between the physicians and the pharmacy team and helps increase awareness about appropriate medication usage and closer follow‑up by pharmacy.

We wanted to take a broader look at whether or not this intervention and this care model is associated with a reduced medical expenditure. We cast a pretty wide net and looked at total medical expenditure, and not parse it out based on reduction in adverse clinical outcomes or increased hospitalization, or anything like that.

Part of it was essentially first trying to look at whether or not there is any there there, and then start to look into what's the cause and mechanism by which integrated specialty pharmacy is associated with a cost savings.

One of the steps that we had to do to be able to do an appropriate comparison between these two groups is even them out. There was about 10% of population in the subset of those that were using specialty, or were part of specialty clinics, and were enrolled in a UMass Medicare ACO that were also using integrated specialty pharmacies.

10% of them, versus 90% of those that were using non‑integrated. What we ended up doing is we used matching on age, sex, and the level of care, which is assigned by the ACO based on how severe the disease that's being treated and what are the number of comorbidities, as well as social determinants of health that play into the overall complexity of care for these patients.

Once we did that matching, we were starting off at a pretty even starting point in 2016 total medical expenditure. Then we looked at what was the medical expenditure in 2017 and 2018? Among the group that did not use the integrated specialty pharmacy, it increased steadily over that three‑year period, which is not surprising.

That's consistent with data that we know at the national level. The health care costs are increasing at a slow but steady level. What was surprising, and part of our major finding, is those patients that were enrolled—there were about 100 patients enrolled in the integrated specialty pharmacy. Those patients' health care total medical expenditure from 2016, '17, and '18 did not increase. It stayed flat, or if anything, it decreased marginally. One thing I would want to point out, though, is statistical significance at 95% levels was not achieved. A big part of the reason, because the trends are pretty consistent and strong in terms of the difference between the two groups, but we are limited by the sample size.

This is a single‑center study where it's within the UMass Health Care System We had limited sample size of patients who were enrolled in integrated specialty pharmacy, around 100, and we matched   We were reassured by the directionality and how consistent it was from one year to the next year.

What are some of the possible real‑world applications of these findings in clinical practice?

I think this speaks to increased coordination and creating multidisciplinary elements of care is helpful. Especially in the context of accountable care organizations where there is a greater bandwidth to be able to do more integration is important. Especially when it comes to caring for patients that require specialty care.

These have people that require specialists to prescribe medications for them and follow up with them more regularly. When you integrate pharmacy and how they are getting medications, making sure that they are taking their medications properly, that's associated with reduced medical costs for the patient and for the system.

The next steps for this is to then look at: “well, savings are good, but is it associated with an improvement in clinical outcomes?” That's part of the investigation that we're looking at. The other part of it is are these estimates stable across different health care systems.

As I mentioned, this is part of only UMass patients. What we are trying to work on is seeing if this difference that we observed is consistent across different systems, and working with the Shields Pharmacy, as well as other service providers to see if we can replicate part of these studies in other organizations.

If this is borne out by other studies, then I think the implications of this is we need to start looking at integrating as much care between pharmacy and the consultants as possible so that we're increasing the adherence as well as the fidelity of the treatment plans.

Do you and your co‑investigators intend to expand upon this research?

Yes, and I don't want to get too ahead in front of the investigations that are going on, but we are planning on doing investigations on two threads. One is expanding the setting of where this study was done.

This aspect is to facilitate similar analysis by other care providers who have also integrated specialty pharmacy for their patient population, to ascertain if they are also experiencing the same kind of net savings.

Then we are planning on looking at whether or not clinical outcomes are improved when you integrate specialty pharmacy. At the patient-facing end, that's the most important part, is being able to identify whether or not this integration is associated with improvement in clinical outcomes.

Is there anything else that you would like to add pertaining to your research or the findings?

I think an important part of this is that this was a combination of the academic side of medicine at our institution, with expertise in biostatistics and program analysis, and the clinical side of things with expertise in policy implementation, and defining and structuring different kinds of care models.

More collaboration between the academic side and the clinical providing side is important and can allow for expansion of effective care models in different settings.

About Mr Soni:

Apurv Soni is a MD‑PhD candidate at University of Massachusetts Medical School. He finished his doctoral training in epidemiology and biostatistics with a focus on looking at health disparities in global and local settings.

Clinically, his focus has been on internal medicine. He personally is in postdoctoral training in internal medicine starting next year. Overall, his clinical research has also evolved more recently into looking at clinical outcomes and alternate care models.

Mr Soni’s previous background was on international health. He’s had a lot of community‑based studies that had been performed in India, and he has always been interested in looking at what different care models can be implemented to have a value‑based proposition for providing care efficiently.


Soni A, Smith BS, Scornavacca T, et al. Association of Use of an Integrated Specialty Pharmacy With Total Medical Expenditures Among Members of an Accountable Care Organization. JAMA Netw Open. 2020;3(10):e2018772. Published 2020 Oct 1. doi:10.1001/jamanetworkopen.2020.18772

Agree or disagree with an article? Share your professional thoughts on an article you read.

Your Name
18 + 1 =
Solve this simple math problem and enter the result. E.g. for 1+3, enter 4.
Back to Top