With millions of Americans joining the ranks of those with health insurance under the Affordable Care Act, payers need to become even more vigilant in controlling costs.
The concept of "big data" offers promise in unlocking information that can help keep costs in check, but it's important to know what costs you may not be aware of or tracking, but are still incurring. For many payers, the four biggest costs that are not obvious are:
· Avoidable inpatient admissions
· High non-generic Rx costs
· High inpatient re-admissions
· Lack of data quality, causing re-work and delays
While data alone cannot determine which hospital admissions are necessary and which are preventable, the Agency for Healthcare Research and Quality (AHRQ) has established specific Prevention Quality Indicators (PQIs) that are clues to those diagnoses that may have been avoidable.
Using software that can track these diagnoses helps payers uncover those primary care providers with high rates of admission, and then more closely look at the reasons for those admissions. It is not always the case that high admission rates indicate poor quality of care. Providers treating a larger number of high-risk patients often have admission rates that are higher than those providers treating a less acute population.
AHRQ identified the following 16 admission diagnoses as potentially unnecessary.
· Chronic obstructive pulmonary disease (COPD)
· Low birth weight
· Bacterial pneumonia
· Urinary tract infection
· Adult asthma
· Pediatric asthma
· Congestive heart failure (CHF)
· Diabetes short-term complication
· Diabetes long-term complication
· Lower-extremity amputation among patients with diabetes
· Uncontrolled diabetes
· Angina without procedure
· Perforated appendix
· Pediatric gastroenteritis
These PQIs offer a place to begin for measuring quality of care. However, they are just screening tools and not final determinants of how well a provider delivers outpatient care.
High non-generic Rx costs
The ongoing education of providers and patients as to the merits of generic medication made a huge impact on providers' prescribing habits. A study published in JAMA Internal Medicine reports that 63 percent of providers rarely or never prescribe brand-name medication when a generic is available.
The bad news is that the other 37 percent, which is about 286,000 providers nationwide, do favorably respond to patients who ask for a brand name medication. Many new medications do not have a generic equivalent as they are patent protected, but older generic medications may do just as well in managing a patient as a brand new medicine.
The study, performed by the Mongan Institute for Health Policy and Harvard-affiliated Massachusetts General Hospital, found that those providers who prescribed costly brand medications over generics had certain things in common: Physicians in practice for more than 30 years and those who received sponsored food and beverages in their practices.
While researchers couldn't prove a cause-and-effect between industry marketing and prescription practices, the correlation exists. The problem often facing payers is determining which brand name medications have older, equally-effective generic counterparts.
Since the provider drives prescription costs, a possible solution is incentivizing providers to prescribe generic medications. A study at Stanford University found that capitated patients generally receive generic medications, while PPO patients are likely to get branded drugs if they ask for them. This suggests that getting providers to share in the risks of prescription costs helps to cut brand name prescribing.
High inpatient readmissions
Medicare and Medicaid have shone a spotlight on preventable hospital readmissions within 30 days. Hospitals are held accountable for these readmissions and are financially penalized by CMS when they occur.
Most large payers also impose penalties on hospitals for readmissions, but smaller payers with limited networks often do not or cannot capture the data that allows for analyzing readmissions. They may not have software tools that allow them the longitudinal view to see that a patient was admitted for a heart attack, discharged, and then readmitted for an unrelated accident two weeks later, and therefore they may not see that this individual wasn't readmitted due to poor quality of care. To gain that view, payers need to acquire software that is advanced enough to determine, by diagnosis, whether a readmission is avoidable. Armed with accurate data and analysis, negotiations regarding penalties become more likely to favor the payer.
Lack of data quality causing re-work and delays
The push is on for Electronic Health Records (EHRs). The problem is not so much lack of quality data as it is lack of the ability to send and receive data in formats that are understood by all. Hospitals use certain systems geared to their internal needs, as do laboratories, imaging centers, provider offices, payers, and many other constituents in the health care system. Data from each system is not inherently compatible with data from other systems.
However, there is an easy, affordable fix for this problem. HL7-compliant interface engines convert data from one EHR to data that is understood by other systems. When contracting with providers, payers that insist that providers have interface exchange engines should see a reduction in repetitive tests and in patient care delays.
The amount of data available to the health care industry, including payers, is huge and is growing.
Gregory Berg is associate vice president for research and outcomes at AxisPoint Health.