‘These benefits not only assist the transition toward patient-focused care, they’re helping healthcare institutions reduce associated costs.’
Hospitals are just beginning to catch on to the promise of integrated data analytics to manage patient population health and measure treatment outcomes, among other predictive uses. These benefits not only assist the transition toward patient-focused care, they’re helping healthcare institutions reduce associated costs.
By discerning better treatments for patients through clinical analytics, treatments that otherwise would fail to produce the desired outcome are eliminated. “You can reduce both the hospital’s waste of resources and the patients’ waste of time, each of which presents positive financial implications for the healthcare provider,” explained Amy P. Abernethy, MD, PhD, director of the Center for Learning Healthcare in the Duke Clinical Research Center, and the director of the Duke Cancer Care Research Center.
While there is every reason to believe that data analytics, specifically its use to support clinical decision-making and personalized medicine, offers significant financial value, Abernethy noted that, at present, this remains a hypothesis. “This is a proof point that needs to be tested over time, but it’s a pretty reasonable hypothesis, nonetheless,” she says.
Abernethy is extremely knowledgeable about the use and promise of “big data” analytics in the healthcare industry. She heads up the CancerLinQ health information technology initiative sponsored by the American Society of Clinical Oncology (ASCO). To date, CancerLinQ has collected more than 170,000 de-identified medical records of breast cancer patients provided by oncology practices around the United States. At present, CancerLinQ is ensuring this data is of the highest quality; it then hopes to draw conclusive knowledge of specific treatment outcomes based on different types of patients.
[See also: Reevaluating capital spending strategies.]
Clifford Hudis, MD, ASCO president and chief of the Breast Cancer Medicine Service at Memorial Sloan-Kettering Cancer Center in New York City, agrees that data analytics offers tremendous hope for hospitals to develop more predictive clinical treatment conclusions. “Future research will depend to an ever greater extent on data analytics,” Hudis said. “Only three percent of adults with cancer currently participate in research studies. With data analytics, the information that already exists on patient populations can be a gold mine, insofar as the development of specific medicines for certain types of people.”
Healthcare standardization is good and variability is bad, he explained, as it leads to higher costs, lower efficiencies and reduced care quality. “Data analytics enforces more standardized care, guiding better treatment outcomes,” said Hudis. “At least, that’s our hypothesis.”
Analytics in action
Several hospitals are at the early stages of applying data analytics to both improve patient outcomes and reduce their internal costs. Schneck Medical Center in Seymour, Ind., for instance, is aggregating patient data by diagnosis to pinpoint inconsistencies in treatment patterns. “As we looked at patients with a specific diagnosis, some had significantly higher lab charges than others,” said Debbie Ridden, Schneck’s vice president of fiscal services. “We then drilled down through the data to learn which types of lab tests were ordered.”
In one particular case, the hospital discovered a single lab test that its physicians were utilizing inconsistently for some reason. Schneck’s pathologists subsequently educated the medical staff on how the particular test could be more effectively utilized in caring for patients.
The hospital has also leveraged data analytics to review causes of increased lengths of stay among inpatients. Five specific diagnoses were examined for length of stay, targeting the unexpected factors that led to these patients’ longer inpatient stays. The findings included a particular test ordered on a weekend that wasn’t critical but nonetheless kept the patient in the hospital, and a staffing situation where the patient needed to stay an extra day because it was a Sunday night and necessary staff would not come on duty until Monday.
This insight guided areas where Schneck’s evidence-based practice could be more effective. “Not only were we able to reduce the average length of stay, but an even better outcome is that our patients got well faster,” Ridden says. “Any time we can get them better faster is both good for them and for us from a cost standpoint.”
Penn Medicine Chester County Hospital also is wielding data analytics to its financial benefit. “As we and other hospitals move to fully electronic systems, there is this wealth of data that we are just beginning to mine,” said Ray Hess, the hospital’s vice president of information management. “Data becomes information, and information becomes intelligence. We’re moving from retrospective reviews of what happened to real time knowledge in action.”
Asked for a specific example, Hess cited the work being done to reduce hospital readmissions. “We’ve created a series of algorithms that help us identify specific patterns among certain patients that are predictors of their high risk for readmission,” he explained. “We then target interventions while they’re here in the hospital to reduce the possibility they will need to be readmitted.”
Is the data analysis actually reducing patient readmission rates? It’s too early to tell, Hess replied, but he is sanguine that patient treatment outcomes will improve along with the hospital’s financials. “We’ve been a showcase for this work and have spoken about it at the regional and national levels, but the truth is we just got it up and running,” he added.
There is other early evidence of hospital financial success stories from data analytics. Hudis cites a project at Memorial Sloan-Kettering employing data analytics to standardize the use of anti-nausea treatments by cancer patients after chemotherapy. “By putting everyone on a regimen of certain anti-nausea treatments, we no longer have to stock 50 different drugs; we can buy five drugs in much greater volumes, which then reduce the overall cost,” he explained. “Meanwhile, nurses can plan the treatment better based on this care, and patients can plan their lives knowing better how long it will take them to get this more specific regimen of drugs.
“There is a remarkable series of efficiencies that arise from standardization,” he added.
Abernethy concurs: “If a hospital learns that patients with the most complex chronic illnesses like emphysema and heart failure are costing a lot of money, data analytics in conjunction with mobile devices offer a way to get better at organizing and coordinating the related healthcare, from the physical level down through ancillary staff like physical therapists, nursing homes, residential facilities and even the local pharmacy.”
As time goes by, these and other efforts are likely to produce greater evidence supporting the efficacy of data analytics to reduce hospital costs.
This article was originally published on Healthcare Finance