The advent of Medicare administrative contractors has emphasized the importance of ensuring that charge capture is consistent and accurate for the professional and technical components of care. This affects many areas, but arguably, none more greatly than outpatient diagnostic and interventional-radiology services. Hospitals and physicians encounter many challenges in trying to accomplish this task.
Code assignment for the physicians (the professional component) is typically a much simpler process than for the hospital (technical) side. Most radiology practices use either manual coding, in which a coder reviews the report and assigns diagnosis and procedure codes, or natural-language processing, in which the dictated report is analyzed by a computer and codes are assigned based on documentation. In most cases, the radiologist’s code selection is based on the dictated report; if a service was done, but not documented, it will not be coded and submitted to the insurance company for payment. On the hospital side, if the service was done, it will typically be assigned charges, regardless of whether it is documented (depending on the charge structure of the facility).
Historically, hospitals have had difficulty maintaining a high rate of accuracy in code reporting for outpatient procedures because of the complexity of the charge-capture process in the radiology department. There are typically multiple computer systems, multiple departments and personnel, and (many times) a poor or nonexistent feedback process.
In some instances, it is appropriate for the hospital and radiologist to submit different codes; however, these typically account for a very small percentage of hospitals’ total outpatient-radiology claims. In some instances, the payor will instruct the two parties to report different codes, and these instructions should be followed. For example, for an injection into the sacroiliac joint with anesthetic, the hospital is instructed to report G0260, but the physician is to report 27096.
In most cases, if the hospital’s codes are not the same as the radiologist’s codes, one party’s codes are wrong. Incorrect coding might mean incorrect payment and incorrect charges to the patient. This situation might also represent a compliance risk for one or both parties.
Discrepancies between hospital and radiologist coding can lead to claims review for either or both parties. For example, if the radiologist reports procedure codes or dates of service that differ from those reported by the hospital, a third-party payor might request a copy of the record from one or both parties. This situation will delay payment and increase costs. The more frequent and drastic the discrepancies are, the higher the risk of audit by Medicare or other payors. This can result in inconvenience, at best, or financial penalties, at worst.
Both sides can benefit from taking proactive steps to evaluate the extent of any coding mismatches and to correct problems in coding accuracy. While there are many approaches to such a problem, Coding Strategies, Inc, Powder Springs, Georgia, recommends these steps:
- initiate a dialogue,
- evaluate the extent of the problem,
- identify the causes,
- develop an action plan, and
- implement ongoing monitoring.
Making changes of any magnitude is always a challenge. It is usually best to start with informal discussions between the radiologists’ practice administrator and the hospital administration. Most will welcome this conversation. It is important to emphasize the potential benefits to both parties, as well as to the patient.
Next, both parties must collectively determine the best approach to comparing charge data (and, ultimately, the coding processes and resulting accuracy). The two primary methods used to approach this problem are the comparison of aggregate charge data and the comparison of individual charges for specific patient accounts. If aggregate charge data are evaluated, comparison of CPT® utilization over a one- to three-month period is recommended.
There are several key pieces of information that can be gleaned from evaluating utilization. For example, you can see how many times each party made a charge using a particular procedure code. The numbers will not match exactly, since there are sometimes delays in charge entry, dictation, and so on, but they should be relatively close. If there are significant differences in code distribution and utilization that