Big is NOT better |

I always suspected this, but the only proof I had was anecdotal. Obviously, someone else decided that there was anecdotal evidence to merit a true study.

Dr. Kevin Erickson (Baylor College of Medicine/Rice, but he’s also a paid consultant for Acumen LLC) headed this study, with associates Drs. B. Zhao, J. Niu, and W.C. Winkelmayer (all from Baylor), Dr. V. Ho (Baylor/Rice), and J. Bhattacharya and G.M. Chertow (Stanford). The results were published in JAMA (Journal of the American Medical Association).

The title might be ambiguous (Association of Hospitalization and Mortality Among Patients Initiating Dialysis With Hemodialysis Facility Ownership and Acquisitions), but the results confirm what many of us have thought occurred over the decades.

First, a little history. Back in the 1980s, there were more than 3000 dialysis centers across the USA. There were two big for-profit chains (which were Fresenius and DaVita before name changes and coalesces), a few non-profit chains, a bunch of smaller (10 to 30 unit) groups, and a bunch of unaffiliated groups.

The government encouraged the merger of these units. Of course, they were discrete about it- but the logic was that if only one, two, or even three big chains existed, it would be harder to exert political pressure on HHS (Health and Human Services) to not lower the reimbursement rate or keep it constant. When there were a plethora of independent units, each complaining to their representative and senator, the political pressure on HHS would be unnerving. Once there were fewer independents and only two big chains (as I’ve reported, more than 90% of the dialysis business is locked up by the two behemoths, Fresenius and DaVita), then only a few senators and congressman would be involved- and HHS could better withstand the politicians’ heat.

We always thought that once a chain took over a clinic, with the “improvements” (lower professional staff per patient, uniform choices in equipment and modalities) that were imposed, patient care suffered. As I said, we noted anecdotal results, but had no hard or fast data. (Imagine how the chains would react if we proposed such a study. This particular study was funded by NIDDK [National Institutes for Diabetes and Digestive and Kidney Diseases) of the NIH, which had the clout necessary to effect the study.)

The analysis employed a cohort study, one that examined the difference in differences (DID) of health outcomes over time. The researchers compared the results from in-center treatment that were acquired (to be a chain, as defined in this study, meant 200 facilities under a single aegis or more), with independent facilities located nearby. Two of the primary outcomes they examined were patient mortality (the risk of death during the first 12 months of dialysis) and the number of days the patient had to be maintained in a hospital (hospital days per patient days).

The study term extended almost 15 years (January 2001 through September 2015). Almost 175K patients’ (174905 total, 79705 were women) data- from 3 years before and 3 years after acquisition- were included. Data analysis began in March 2017 and was completed by the end of 2018.

The patient mix included Hispanics (24409 or 14%), Blacks (61815 or 35.3%), and Native Americans (1247 or 0.7%). The rest of the patients (105272 or 60.2%) were Caucasian. The mean age for all the patients was 65.

Before an acquisition, mortality was 10% lower than when the chain took over the facility, with hospitalization rates of 2.9 fewer days per patient year for the same units. When comparing non-acquired to acquired clinics, mortality and hospitalization rates were also lower.

The researchers do note that independent units that demonstrated higher costs and better outcomes were the more likely units to be acquired by the chain. And, to minimize the effects any takeover can effect with the disruptions in care due to the acquisition, the study employed a 3 year follow-up after the acquisition.

But, this study demonstrates that our suspicions were clearly in order.