How are procedure statistics derived?

When searching for certain procedures, our site will sometimes include procedure information that we think is relevant to the search query. For example, a search query for "hip replacement" might show which providers have performed hip replacement surgery, and the search results page might list some basic statistics: procedure volume and price range.

The number of times a surgeon performs a procedure (known as procedure volume) has been found to be correlated with better outcomes: that is, surgeons who get more experience with a procedure tend to do a better job.

Price can be helpful to those who are paying cash, have a high deductible, or are simply interested in helping the system reduce healthcare expenses. The displayed price is an average of prices that a provider might have submitted to Medicare for a set of procedures and most closely represents the price a patient without insurance might be asked to pay. In practice, insurance companies generally pay significantly less than the submitted price. Note also that the price only represents compensation for a provider's services; frequently, there are other fees that may accompany the professional fee (e.g. facility fees or professional fees for other providers that are involved with a procedure, such as anesthesiologists).

Data such as procedure volume and pricing information is often derived from claims data (providers submit claims to insurance plans to be reimbursed for their procedures). Claims data is attractive because a single insurance company (known as a payer) can release millions of data points about hundreds of thousands of providers. However, claims data comes with limitations. The largest limitation is that the procedure volume and pricing information on our site comes from Centers for Medicare & Medicaid Services (CMS for short, also referred to as Medicare) and as such, does not include procedure information for other insurers. Hence, a provider might have performed a specific procedure many times, but if that provider has only a few Medicare patients, the procedure count will be severely underreported. Additionally CMS has a policy of suppressing data points where a provider has performed a procedure for fewer than eleven Medicare patients in a year (for privacy reasons). Another nuance is that procedure codes will often involve what is known as a modifier. For example, if someone assisted in a procedure, that provider would likely have a different modifier than the provider who performed most of the procedure and the two providers would likely charge different amounts. Medicare removes the procedure code modifiers, which hides some important differences in the procedure information. Yet another limitation is that CMS tends to be more than a year behind in making their data available -- as of August 2019, CMS has provided data for 2012 to 2017.

Aside from CMS limitations, there are other limitations related to insurance claims more generally. Providers use procedure codes to specify which procedures they are billing for. These procedure codes are developed by the industry primarily for billing purposes and differentiate between different amounts of work rather than between different procedures. For example, "calf or ankle tendon lengthening or shortening" might refer to four different procedures:

  1. 1. calf tendon lengthening
  2. 2. calf tendon shortening
  3. 3. ankle tendon lengthening
  4. 4. ankle tendon shortening

The committee that designed the codes for these procedures believe they require approximately the same amount of effort, so it grouped these four procedures together and assigned them two codes (only differentiated by the number of tendons involved). Groupings like this can cause some confusion when these billing codes are used to identify providers who have experience in a specific procedure. For example, if someone searches for "calf tendon lengthening," we will show the information that we have for "calf or ankle tendon lengthening or shortening" since the relevant codes do not differentiate among the four different procedures. It is possible that a provider has only performed calf tendon shortening and not calf tendon lengthening, in which case the displayed information is less relevant; however, the underlying claims data provide us no way of figuring out how much of each specific procedure the provider performed.

Note about year range: we have chosen to sum the number of procedures performed by a provider over the years that we have information for, but the price range is from the most recent year.

Even with all of those caveats, we believe that displaying procedure volume and price range can be helpful to prospective patients. If you have any questions or comments, please let us know.