Case study: Predictive analytics helped Sonic Healthcare cut costs and boost profits

Philip Chen, MD, PhD

Chief Strategy Officer, Sonic Healthcare USA

Case study: Predictive analytics helped Sonic Healthcare cut costs and boost profits

6 October 2020

Quick Takes

  • Many patients with chronic diseases go a year or more without seeing their doctors and they often end up in the hospital as a result

  • Predictive analytics can be used to identify these patients so they can be encouraged to follow up with their physicians before expensive exacerbations occur

  • This case study demonstrates that implementing such a strategy can potentially lead to substantial cost savings and better patient care

Many patients with chronic diseases go a year or more without seeing their doctors and they often end up in the hospital as a result. In this business model case study, Dr. Chen explains how he uses predictive analytics to find those missing patients and make sure they follow up with physicians before expensive exacerbations occur. 

His innovative approach led to a financial agreement with an Accountable Care Organization – and a significant percentage of the savings that came about as a result.

The challenge: driving expenses down while keeping the quality of patient care high

Dr. Philip Chen, the Chief Strategy Officer of Sonic Healthcare USA, was charged with finding a way for a regional payer to cut costs among its 75,000 patients.

The problem: underdiagnosed patients were leading to unexpected expenditures

When he began scrutinizing insurance data, Dr. Chen discovered a curious fact. Among the top 1% of the payer’s most expensive patients, two-thirds had spent next-to-nothing the previous year. 

“Why did those patients jump from no expenditure one year to a high expenditure the next year?” he recalls asking himself. 

The answer quickly emerged.

“It wasn’t because they were healthy,” he recalls. “It was because they had a problem, but they weren’t seeing doctors to get it taken care of – and all of a sudden, they showed up and required a hospitalization.”

One example: Dr. Chen estimates that 26% of diabetic patients had not seen their doctors for over a year. “They have no expenditure for that year, or a very low expenditure,” he says, “and then all of a sudden, they’re the ones getting into trouble.” 

The plan: identify the most expensive patients – and intervene earlier

To tackle the problem, Dr. Chen – an expert in predictive analytics in healthcare – worked with his team to develop a list of the 16 diseases they felt could bring about the greatest cost savings through proper interventions. 

They were the “conditions that if we catch early, we can keep them from jumping into that top tier of expenditures,” he says. “We sorted that group out based on claims data, and we lined them up from the most expensive per capita to the least.”

Diabetes and chronic kidney disease patients were among those near the top of the list. 

Dr. Chen then turned to lab data to “find all the so-called time bombs before they showed up in the hospital.”

The insights: after lab tests, patients weren’t following up to receive an official diagnosis

Soon came another revelation. A shockingly high number of patients had actually tested positive for chronic diseases – but had never been officially diagnosed. Why? Because they never followed up with their doctors after their lab tests.

“In the primary care setting, about 17% of the patients do not have a diagnosis of diabetes, even though the laboratory has results showing that the patient has diabetes,” says Dr. Chen. “The patient may or may not have any symptoms, so they don’t feel it’s an urgent situation. So they don’t come back to see the doctor. Those laboratory results are just sitting in the doctor’s chart and no one is actively dealing with them.”

The key then was to get patients back to their doctors after their lab tests, “because once they show up, they are managed fairly well,” he says.

The execution: an automated system to encourage testing and follow-up

Case study: Predictive analytics helped Sonic Healthcare cut costs and boost profits

To give patients the nudge they needed, Dr. Chen and Sonic Healthcare developed an automated system that contacted patients via phone calls, text messages, and emails. 

“We actually record their own doctor’s voice to tell them that they need to come back for a follow-up visit,” he says. “During that same phone call or text message, if they want to make an appointment right there and then, we will transfer them to the appointment center for their doctor’s office.”

The automated system also encourages patients to get needed tests done before they go back to their doctors.

44% of patients made appointments in response to the automated reminders

“We tell them: The doctor would like to see you, but before you see the doctor, please stop in the lab and get the work done,’” says Dr. Chen. “So, instead of two visits to the doctor, where the patient has to go get the order for the lab test, go to the lab for the test, and then come back to the doctor later to get the result and get treated, we can combine those two doctor visits into one.

“That way, we don’t get into the same situation where the doctor sees them, orders tests, the test results come back, and they never come back to the doctor.”

The program achieved an astounding 44% response rate, says Dr. Chen, adding that the typical response rate to automated phone calls is in the single digits or low teens. 

The payoff: mutually-beneficial success for Sonic Healthcare and the ACO

By helping to generate more complete and accurate information about patient populations, Dr. Chen’s strategy provided a potentially huge boost for Accountable Care Organizations (ACOs).

ACOs are groups of doctors, hospitals, and other healthcare providers who work together to coordinate the best possible care for their Medicare patients. The goal of this coordinated care is to ensure that patients get the right care at the right time, while avoiding unnecessary and potentially costly duplication of services or medical errors.

“ACOs have financial benchmarks, especially the Medicare ACOs,” Dr. Chen explains. “Say an ACO has 10,000 patients, and they’re risk-adjusted. With 10,000 patients, we expect that group to spend, say, $120 million a year. But if that group spends less than $120 million – say, $110 million – then the government keeps half of the difference and the ACO keeps the other half.”

The key is that that hypothetical $120 million benchmark is based on the conditions and risks of the ACO’s patients. When patients aren’t properly diagnosed, the benchmark skews lower. 

“If a patient doesn’t have a diagnosis, Medicare thinks that patient is normal,” Dr. Chen says. “Therefore, that patient is only worth, say, $6,000 a year, like an average-risk patient. But if you put that diagnosis code on, then, all of a sudden, that patient may be worth $12,000 a year for diabetes, or $21,000 a year for chronic kidney disease. Now, that benchmark of $120 million can be $130 million, even if you don’t do anything, just by having the proper diagnosis for those patients.”

Armed with insurance records that identified the most expensive patient types, lab data that showed who the riskiest patients were, and a system that helped get those patients the care they needed, Sonic Healthcare was able to negotiate an agreement with an ACO.

Sonic Healthcare would serve as the preferred laboratory for all providers in the ACO, would integrate its IT systems at no cost to the providers, and would be entitled to a portion of the savings it helped generate.

The payoff came when Dr. Chen and his team at Sonic Healthcare used their data and their automated patient engagement system to:

  • Get patients properly coded, which raised the ACO’s benchmark
  • Get patients to a doctor before expensive exacerbations occur, which lowered expenditures 

The final result: A substantially bigger piece of the pie for the ACO to keep at the end of the year, with a portion of the shared savings going to Sonic Healthcare. 

Thanks to predictive analytics, shared savings skyrocketed from $12.8 million to $26 million in just two years.

The ACO went from a shared savings of around $12.8 million to roughly $26 million in just two years. And the program has been so successful that Sonic Healthcare has already reached a predetermined payout ceiling. “The cut that we receive will pay for everything we invested, plus continued funding for the future expansion of the program,” says Dr. Chen.

The takeaway: predictive analytics can potentially lead to substantial cost savings and better care

After being tasked with cutting costs, Dr. Chen’s data-driven program got to the bottom of the problem and led to a win-win solution. The ACO achieved substantial cost savings, while more patients received a proper diagnosis and the crucial care they desperately needed. 

Philip Chen, MD, PhD is a board-certified pathologist and the Chief Strategy Officer at Sonic Healthcare USA. Dr. Chen has been practicing pathology and running laboratories in commercial, non-profit, and academic settings. He founded Cognoscenti Health Institute in 2002, which was acquired by Sonic Healthcare. He also ran the clinical laboratories across two health systems: The University of Miami, and Jackson Health Systems in South Florida. Throughout his career, Dr. Chen has been exploring ways to practice laboratory medicine outside of the laboratories. In recent years, he has been focusing on using clinical and financial informatics to deliver value-based services and constructing new, value- and outcome-based reimbursement arrangements. Dr. Chen obtained his MD and PhD degrees from the University of Alabama at Birmingham and received pathology training at Brigham and Women’s Hospital and Harvard Medical School.