Okay. By now, you know what an All Payer Claims Database is, what types of data releases are available, and that any data released from the CO APCD must be used to benefit Coloradans. This is an excellent start to understanding what CIVHC does as administrator of the CO APCD but it probably doesn’t really illustrate what a cool resource the database is and the incredible projects that it can inform.
Don’t worry, it took us a while to understand too. In fact, we’re still learning the magnitude of what it can do and identify new uses every day.
Claims data is a tricky thing to get your head around, as it isn’t as neatly packaged as other datasets. There are certain things you can learn from it and other things that you can’t and the dividing line between the two can be less than intuitive.
One of the ways we describe the information in the CO APCD is: “it contains whatever is included on a health insurance claim.” Clear as mud, right?
We thought so. To help everyone (including us) understand what claims data – and more specifically, the CO APCD - contains and how it can be used, CIVHC created a guide to the capabilities of the CO APCD. This handy list explains what information is usually not found in claims data and what is currently available in the CO APCD. There is also a column that details what could be found in claims data if changes are made to legislation or data submission guidelines.
So, what can the CO APCD do?
Claims data only contains the information necessary for a provider to file a claim with an insurance company and be reimbursed for services rendered. This information is passed along via a special set of codes on forms that specify diagnoses, treatments, lab tests, and prescriptions. Claims have no clinical data because it is not necessary for the insurance company to know the height, weight, blood type, or medical history of a patient in order to reimburse a provider.
Health insurance claims identify which provider a patient saw, what health insurance company the patient has, and how much both the insurance company and the patient paid for a service. These bits of data can be quite useful when analyzed, as the Colorado Optometric Association (COA) learned in 2014.
When COA requested data from the CO APCD in 2013, Colorado Medicaid vision codes had not been reviewed in over 40 years. Medicaid reimbursed optometrists roughly 19 percent of care expenses, making it financially difficult for many doctors to accept Medicaid patients. Some of those who did accept patients had to restrict the number they treated, causing an access-to-care problem, especially in rural Colorado.
To combat access issues and increase options for Medicaid patients, COA used CO APCD Medicaid price and utilization data to calculate the financial impact a reimbursement increase would have on the overall Medicaid budget, the number of optometrists willing to accept patients covered by Medicaid, and access to vision care services across the state.
Upon completion of the analysis, COA presented the findings to the Department of Health Care Policy and Financing (HCPF). They were able to demonstrate that the low number of Medicaid-participating vision care providers was due to their inability to recover the minimum chair cost (price per patient per hour).
Based on data presented, HCPF made a recommendation to the Governor’s Office of State Planning and Budget to increase the reimbursement rates for certain vision codes. The increased rates became effective July 1, 2014 and the number of vision care providers participating in Medicaid has increased, expanding access to services in underserved areas including rural Colorado.
More people getting vision care is always a good thing; but, what about people with chronic conditions? How can CO APCD data improve care for them if there is no clinical information in an insurance claim?
That’s where health care’s special language of codes comes in. In order to reimburse providers accurately, the insurance company needs to know what services were provided to each patient. Associated with each hospital admittance, treatment, and prescription, there is a code that tells the insurance company how much to reimburse the provider and the patient’s benefit plan determines how much of the bill the patient pays. By analyzing these codes, it is possible to improve care, as one university student is trying to do.
Logic dictates that if a patient fully understands their hospital discharge instructions then they would be less likely to be readmitted. However, there is little formal data about the correlation between effective discharge information from the patient’s perspective and hospital readmission rates. A Regis University student evaluated CO APCD data alongside patient survey information from the Hospital Consumer Assessment of Healthcare Providers and Systems (HCAHPS) to see if such a relationship exists.
Hospital-specific 2009-2013 CO APCD data from across all lines of business (Commercial, Medicaid, and Medicare) was analyzed compared to responses to two specific HCAHPS care transitions survey questions for Colorado hospitals:
- Whether they were given information about what to do during their recovery at home.
- Where they ranked themselves when asked if they understood their care when they left the hospital.
The results of the study showed that across all diagnoses, higher rankings on the HCAHPS care transitions composites were associated with lower 30-day all cause readmission rates. Hospitals can use this information to build the case to enhance their own targeted interventions and ensure discharge instructions are communicated effectively to support healthy transitions of care.
Pretty cool, huh? The CO APCD is being used to make people’s lives better and these are only two examples of custom data fulfillments we’ve made since 2013. The Colorado Medical Price Compare Showcase has many more illustrations of the amazing ways claims data is improving care, lowering costs, and creating a healthier Colorado. And that’s just custom data…we haven’t even started talking about the value of the data available for free on our website.