The most award winning
healthcare information source.
TRUSTED FOR FOUR DECADES.
Hospitals depend so much on outcomes data to determine quality, but one researcher says most hospital don’t even know their outcomes. That leads them to make critical decisions based on faulty information, says Donald Fry, MD, executive vice president for clinical outcomes management with MPA Healthcare Solutions in Chicago, and adjunct professor of surgery at Northwestern University Feinberg School of Medicine.
Fry’s conclusion is an outgrowth of his recent research in the Journal of Bone & Joint Surgery, in which he found that risk-adjusted complication and readmission rates vary widely between hospitals where joint replacement surgery is performed. In addition, patients are twice as likely to suffer an adverse outcome after they are discharged than while they are still in the hospital. (For more on the study, see the story in this issue.)
Fry says the study results confirmed his belief that hospitals often don’t know their own outcomes. In too many cases, he says, neither the hospital nor the surgeons know about deaths after discharge or readmissions to another hospital. In his recent research, inpatient major complications of care and deaths represented less than 50% of the adverse outcomes occurring across the entire continuum of care, Fry notes.
“The spectrum for adverse outcomes has to extend into the postoperative period. The era of ever-declining lengths of stay means that more complications of care are not recognized or declared until the patient is in the post-discharge period,” he says. “Hospitals have very limited knowledge, if any, about 90-day post-discharge deaths that are not readmitted.”
Hospitals could gain a better idea of true outcomes with a database holding encrypted patient identifiers so individuals could be followed after discharge for admissions to other hospitals, and for deaths, Fry suggests. If a state database is not possible, the problem might be addressed with a system in which hospitals in a community communicate with one another when patients are admitted to one facility after being treated in another, he suggests.
Medicare’s move toward bundled payments makes it imperative for hospitals to obtain accurate outcomes data, he says.
“Hospitals have to know the results of their care because the hospitals and clinicians under bundled payments will sustain substantial financial penalties when their patients have excessive rates of readmission to the hospital,” Fry says. “You can’t fix a problem if you don’t even know that it is happening.”
Hospitals can unfairly benefit from the lack of true outcomes data, Fry notes, though it is not intentional. By not including some post-discharge data, the hospital’s quality of care can seem higher than it is, he says.
“They’re not being malicious; it’s not some sinister plot to avoid the realities of your quality of care,” Fry says. “But I do think it’s a problem when you declare victory and there’s still five minutes left in the first half. That is what exists with our current measurements of inpatient care.”
There are multiple reasons that hospitals aren’t working with valid outcomes data, says Alan Cudney, RN, MBA, a principal healthcare consultant with SAS, a consulting company based in Cary, NC. Lack of analytic maturity is a key reason, along with how data typically are compartmentalized, disorganized, and difficult to access, he says. Many healthcare organizations do not have a strategy for organizing and managing data, or subsequent analytics with that data, he says.
“The process of preparing data for cross-functional analytics is cumbersome and poorly managed, and the culture does not support, empower, and reward data exploration and collaboration to create real improvements in care and service,” Cudney says. “There is a lack of appropriate analytic tools and a lack of modern analytic tools that can prepare data for analytics, as well as run complex modeling and queries.”
The selection of analytic tools often is based on localized needs and is not aligned with an enterprise strategy, he says. Siloed use of analytic tools makes it more difficult to perform complex analytics across the organization, he says.
“Value-based care is shifting the cost-benefit equation for leaders and clinicians at care delivery organizations. The ability to analyze care and outcomes across the continuum is becoming a capability that is necessary for success and even long-term viability,” Cudney says.
Outcome measures are only as good as the data available to compute them, says George Dealy, vice president of healthcare applications at Dimensional Insight, a data analytics company in Burlington, MA. However, with increased adoption of electronic health records (EHRs) and standardization of clinical and quality data, there’s more opportunity than ever before to work with measurements that have the potential to help improve care, he says.
In the near term, healthcare providers should be able to refine their approaches and develop the associated competencies, to work with the most useful and reliable measures that exist today, he says. As interoperability of healthcare information continues to improve, additional measures — such as those that require information from multiple health systems — will also become increasingly available and practical. Examples of these measures include those related to readmission rates and complications of care, which require information from multiple settings of care, potentially across several organizations, he says.
“Government entities are able to help. The example of the difficulty identifying deaths outside of care settings points to the fact that mortality information is not effectively shared as well as it could be,” he says. “Both the federal government and state governments track this information closely and could potentially share it, assuming the appropriate level of confidentiality and privacy was adhered to.”
Financial Disclosure: Author Greg Freeman, Editor Jill Drachenberg, Editor Dana Spector, Nurse Reviewer Fameka Leonard, AHC Editorial Group Manager Terrey L. Hatcher and Consulting Editor Patrice Spath, report no consultant, stockholder, speaker’s bureau, research, or other financial relationships with companies having ties to this field of study.