Why Population Health Management Strategies Require Both Clinical and Claims Data

My Folder



Health systems will never achieve population health management if they continue to use only claims data or clinical data to analyze patient populations. Neither type of data provides the necessary depth of information when used alone. The solution is to combine claims and clinical data. While this process won’t be easy, it’s necessary if providers and payers want to achieve the three goals of population health management: improved outcomes, increased patient safety, and decreased costs.


The concept of population health management has become intriguing for health systems and payers for two critical reasons. The first reason is to find a way to manage the escalating costs of treating chronic diseases in the United States. According to the Centers for Disease Control and Prevention (CDC), chronic diseases are the leading cause of death and disability in the U.S.1

“Health systems and payers can use population health management to control escalating costs and to accommodate the shift to pay-for-value.”

Consider these facts:

  • Seven out of 10 deaths among Americans each year are from chronic diseases.2
  • Heart disease, cancer, and stroke account for more than 50 percent of all deaths each year.3
  • Arthritis is the most common cause of disability, with nearly 19 million Americans reporting activity limitations.4
  • Diabetes continues to be the leading cause of kidney failure, non- traumatic lower-extremity amputations, and blindness among adults aged 20-74.5

The second reason health systems and payers are interested in population health management is because of the shift in government payment models from fee-for-service to pay-for-value. Led by the Centers for Medicare and Medicaid Services (CMS), the pay-for-value model offers financial incentives for disease prevention and management but penalizes poor outcomes. As a result, health systems are moving patients away from high-cost, acute settings to the most appropriate, lowest-cost settings — and understandably so.

CMS’s mandatory shift in reimbursement models for health systems is necessary because the fee-for-service model is unsustainable. In contrast, the pay-for-value model will significantly reduce the government’s spending on healthcare and curtail the growing budget deficits.

According to a report on the Clinically Appropriate & Cost-Effective Placement Project (CACEP) from the Alliance for Home Health Care Quality and Innovation, Medicare could reduce its spending over a 10-year period by $34.7 billion if patient care settings were shifted from facility-based care to home and community-based care. The report estimates an additional $100 billion could be saved if care delivery were restructured to be less wasteful and more effective. These additional reductions could be accomplished by using bundled payments and by adding a policy to reduce post-acute care spending (excluding the index hospitalization) by 7.5 percent.6

Providers and payers have powerful incentives to change with the new payment model. But to survive, they will need to access and analyze vast amounts of population data from many different outpatient settings. In particular, the following groups and programs need accurate data about the outcomes for various patient populations: Patient-Centered Medical Home (PCMH), Accountable Care Organizations (ACOs), and CMS’s Bundled Payments for Care Improvement initiative (BPCI).


Claims data is administrative data and includes information about patient demographics, billable charges, dates of service, diagnosis codes, procedure codes, insurance, and providers. This type of data is created after every patient encounter with a provider and is subsequently submitted to a payer (the government or an insurance company) for reimbursement.

Historically, payers and health systems have exclusively relied on claims data for analysis associated with population health management because of the following attributes:

  • Claims data is readily available. The requirements for payment to the provider ensure the data is entered into the claims system on a complete and timely basis.
  • Claims data spans a patient’s full continuum of care. The payer has a record of every encounter and every prescription filled unless a patient pays for services out of his or her own pocket.
  • Claims data is highly structured. Almost all the data must be captured in specific fields on standardized forms in order for claims to be approved and payment to be issued. Because of this consistent format, it’s relatively easy for other systems to consume.
Figure 1 – CMS 1450 requires claims data be entered into highly structured fields, making it easy to be used by other systems.

Figure 1 – CMS 1450 requires claims data be entered into highly structured fields, making it easy to be used by other systems.

By examining claims data, analysts gain information about the cost and utilization of a patient population across multiple care settings as well as information about the types of diagnoses and procedures performed. In specific, claims data provides insight into the following performance measures:

  • Mortality rates
  • Complications
  • Access to appropriate health services
  • Charges for care provided7

However, despite having some value, claims data has limited use for quality and cost improvements for the reasons indicated below.

  • Claims data lacks important clinical detail. Claims data is designed  to capture only as much detail as is required to determine payment. This type of data does not capture all of the clinical details of the patient care process. Although a collection of claims can include information about chronic conditions or history, a single claim does not unless it is relevant to the specific treatment or procedure. As a result, data critical to an accurate analysis may be missing from a particular claim. Likewise, a claim may not include all diagnoses and typically lacks critical data, such as lab results and medications.
  • Claims data is highly retrospective. There is a significant lag between the date of care or service and the date the claims data is available for analysis which is most often measured in months. Part of  this delay is due to the need to verify the accuracy of the coding that is the basis for establishing payment. With the speed at which new discoveries in population health management (and healthcare in general) are made, this type of lag time makes it difficult to deliver accurate, timely analysis.
  • Claims data does not provide insight into the actual process of care. The key to improving the quality of care is to understand the dynamics of the processes that are involved in the actual delivery  of care. Since claims data is a static summary of the diagnoses, procedures, and costs that are the result of these care processes, efforts to identify the root cause of any variations in care between patients is impossible.

Strengths — and Limitations — of Claims Data

In addition to claims data, health systems have access to another source of information about their patient populations: clinical data.

Unlike claims data, clinical data provides critical detail and insights into the processes used to deliver

Page 1 of 3
Previous Page
1 2 3