A new study suggests that some seemingly "modest" tweaks to the way an oncology practice operates could help realize millions of dollars of savings.
The study released Tuesday indicated that reducing the use of imaging services, getting patients discharged quicker and checking up on those people sooner could result in dramatic savings even as spending on chemotherapy drugs increased significantly at the same time.
The study of five oncology practices found that treatments for 810 patients that were projected to cost a total of more than $98 million actually ended up costing less than $65 million—a 34 percent reduction, according to an article published in the Journal of Oncology Practice.
That's despite the fact the actual spending on chemotherapy drugs used on the patients was nearly $21 million, as opposed to the $7.5 million that had been predicted on the outset of the project.
Read MoreObamacare's next BIG threat
One of the article's authors, oncologist Lee Newcomer, said that while he had expected some savings from the study focused on the common cancers of the breast, colon and lungs, it was not to the extent that was realized over the three-year project conducted by his employer, insurance giant UnitedHealthcare.
"I thought we'd see something," said Newcomer, a group senior vice president at UnitedHealthcare. "In our world, if we see a 5 percent decrease in expenses, that's a big day. But to see a 34 percent reduction was really shocking and a pleasant surprise."
"If every practice did that across the country, we'd have billions of dollars in savings," said the article's co-author Bruce Gould, an oncologist whose Northwest Georgia Oncology Centers participated in the study.
However, what actually lead to the cost savings isn't clear.
"The source of the cost savings is enigmatic," the paper said, noting that it was impossible to quantify to what degree the savings resulted from either the financial incentives or data sharing that were the cornerstones of the study.
However, Newcomer said whatever was leading to a reduction in costs didn't decrease the overall outcomes for the patients involved in the study.
"Quality actually improved," Newcomer said. "The number of hospitalizations went down."
UnitedHealthcare noted the findings come as the costs for cancer therapy as estimated by the National Cancer Institute are expected to rise to $173 billion annually by 2020, compared with $104 billion in 2006. UnitedHealthcare said cancer therapy now accounts for 11 percent of its spending commercial health plan benefits and is expected to rise in the future.
The study relied on two key changes in the way the oncology centers got paid by UnitedHealthcare for treating patients, and how that treatment was tracked by data.
Each practice agreed to accept a lump-sum payment from UnitedHealthcare for a cancer patient after the first visit for a single episode of care or treatment for their condition. One patient could have multiple episodes of care. The practices also agreed to be paid for chemo drugs at the average sales price rate.
Lastly, the practices agreed to share their clinical data with the insurer, and to meet annually to review data on cost and quality outcomes.
Gould said the study, which called for practices to share data with UnitedHealthcare and hence each other, was "beneficial" because "for the first time we got feedback from a payer, and we were able to get a better understanding how much things really cost."
"I was able to take that feedback and make some modest adjustments in how we did" things," Gould said.
Gould identified two such adjustments.
The first was to hire a so-called "mid-level," whose job was to coordinate care for patient, "and expedite admissions and discharges."
The second adjustment came after Gould realized that several physicians were having patients undergo imaging services, such as CT scans, PET scans and X-rays at a significantly higher rate than their peers at the 21-physician practice, and at doctors at the other oncology practices.
Read MoreGay parents = healthier kids?
"I'd say that 19 of the physicians were doing it right, but there were two or three high-volume physicians who were driving up the the cost of care," Gould said.
Gould said that when he spoke to those heavy users of radiological services, "they tended to be a bit defensive, and tried to justify that their patients were different and having symptoms that justified the radiological intervention."
While the doctors weren't ordered to scale back use of those services, "it served as notice that someone was looking over their shoulder," Gould said.
Another example of how data played a role in changing behavior at the practices came at the Kansas City Cancer Centers, which dropped out of the study after about a year, when it was purchased by the University of Kansas.
Newcomer said that until the study began, doctors at the Kansas City practice were unaware of the fact that the patients they had discharged from the hospital after treatment were waiting up to 30 days or more to get a follow-up appointment, because of the heavy volume of patients at their office.
"Guess what? They got sick again and got readmitted to the hospital," he said.
Yet, once the practice identified that problem "their hospital rate just dropped like a stone." A new system was established where discharged patients got a follow-up visit within 48 hours.
Newcomer added that the study underscored the idea that "the things that can be done in medicine to lower costs are not terrible complex."
"There is inefficiency and there is opportunity to reduce that cost," he said.
—By CNBC's Dan Mangan