By LEONARD D’AVOLIO

The waiting room of New England Baptist Hospital evokes a flood of memories. My father has just been wheeled into the operating theater, prompting me to reflect on my previous visits here.

A while back, Dr. Scott Tromanhauser, the Chief Medical Officer at this facility, enlisted my help to improve outcomes for total knee replacement surgeries. Alarmingly, statistics reveal that nearly 20% of thes operations do not achieve favorable results. A key area for enhancement is reducing needless surgeries.

This issue may appear straightforward from an outsider’s perspective; however, within the intricate framework of U.S. healthcare systems, many surgical centers fail to evaluate whether their procedures lead to improved patient outcomes. Initiatives aimed at lowering surgical volumes are frequently enough perceived as harmful to business interests.

Together with Dr. Tromanhauser, we proposed our idea to his Board of Directors.

“What if,” we proposed, “we could monitor one-year post-operative results for every total knee replacement? By sharing this details with our surgeons for the first time ever, we would gain valuable insights into patient recovery rates. With enough data available, we could provide personalized outcome predictions during pre-operative consultations and empower patients with essential knowledge for making informed medical decisions.”

The board was receptive to our vision despite recognizing it might lead to fewer surgeries—especially those deemed unnecessary—and acknowledged it could strengthen negotiations with insurers. Ultimately they agreed that it was ethically responsible.

This initiative echoes back over a century when Dr. Ernest Codman introduced his “End Results System” as a surgeon at Massachusetts General Hospital in 1905. he advocated that hospitals shoudl systematically collect data before and after every procedure for at least one year—to assess success rates and investigate failures by asking “why?” Codman believed patients deserved access to this information; without it how could outcomes improve or informed decisions be made?

A century later—just miles from where Codman’s ideas originated—we sought to bring his vision into reality.

The institution had begun collecting outcome data under Scott’s leadership; we gathered all participating surgeons for an initial review session where identities were anonymized as Surgeon A through C—but Scott and I knew who they were.

Their reactions were revealing; those who expressed doubts about what they observed tended to be among the lowest performers while dr. Carl Talmo—the strongest proponent of utilizing outcome data—emerged as having superior post-operative results.

This prompted us toward proposing a pilot project designed modernizing Codman’s concept by leveraging historical results effectively predicting future patient outcomes.

An iPad request was developed featuring a validated machine learning model capable of forecasting each patient’s likelihood of improvement—or lack thereof—one year following surgery based on their responses during intake assessments prior entering examination rooms where surgeons would collaboratively discuss findings with them as part shared decision-making processes.