By applying IT at the front end for clinic visits, one healthcare system captured an additional $11 million in charges.
At a Glance
- Identifying missed opportunities for revenue requires examining front-end processes and how they might join with transactions generated during processes of care.
- Linking the initiation of a charge to an important process in the clinic procedures helps ensure capture of both professional and facility charges.
- Implementing a software program to automatically initiate correlated charges and measuring the change in charges and revenue are steps that provide a substantial revenue opportunity for healthcare organizations.
With almost 700,000 outpatient visits a year, leaders at Vanderbilt University Medical Center in Nashville, Tenn., knew some charges related to those visits were not being billed. The question was, how could the health system ensure that all clinic-related charges-both professional and facility-were captured and billed? The problem was resolved by using IT to link professional and facility charges and by automating the organization's manual charge capture process. The end result: Vanderbilt estimates it captured an additional $11 million in charges annually.
Payment for clinic visits is still primarily fee for service, making clinic charges an important revenue source for hospitals and health systems. IT projects present opportunities to improve revenue by identifying transactions with correlated information and integrating the related charges into the charge capture processes. However, merely installing IT is not sufficient. Identifying missed opportunities for revenue requires an examination of front-end processes and how they might join with transactions generated in the processes of care.
Vanderbilt examined ways to integrate charge capture with processes in its clinics. When Vanderbilt began the project in 2003, its physician practice included more than 1,200 full-time faculty members, and the facility payments for clinic patients totaled $133 million.
Vanderbilt's front-end charge capture process was dominated with manual charge forms used by data-entry staff to key data into billing systems. After completion of a redesign of professional charge capture processes the previous year, Vanderbilt's leaders requested a similar review of facility charge capture processes.
With the goal of streamlining data entry and capturing missing charges, Vanderbilt assigned a clinic charge redesign committee to recommend improvements to the clinic's processes. The focus was on the facility charges similar to those payable under the Medicare outpatient prospective payment system (OPPS). The ideal scenario was to link the initiation of the charge to an important process in the clinic procedures. For example, in cardiology, linking a physician's professional charges to an electronic patient checkout resulted in an immediate 39 percent decrease in the number of missing provider charge forms.
Identifying Charges Eligible for the Model
The committee started by examining the types of charges occurring in clinics, which included office professional procedures/treatments, evaluation and management (E&M) services, testing/ancillary services, and miscellaneous other services.
Because Vanderbilt's practice is provider-based, most professional services involve a facility service charge. For example, based on payment rates from 2003, Vanderbilt would have billed the Centers for Medicare & Medicaid Services (CMS) $24.89 and the patient $6.22 for the professional service "destroy benign lesion" (CPT 17000). At the same time, it would have billed CMS $36.41 and the patient $7.22 for the facility portion of the service. A nonprovider-based practice, on the other hand, would have combined the professional and facility charges to bill CMS $45.28 and the patient $11.32. Other facility charges that could be linked to important nonprofessional services in the visit include:
- Injection/administration fees for a group of clinic-administered drugs
- Drug charges for clinic-administered drugs dispensed from clinic pharmacy cabinets
- Infusion fees for patients checked into appointments for chemotherapy visits
- Venipuncture fees for patients with laboratory results on blood specimens
The initial project described here, however, was limited to investigating the creation of facility charges linked to professional services in the visit.
An investigation into whether the quantities billed for the facility and professional components matched found that often they did not (see Exhibit 1). The study group speculated that a huge revenue opportunity existed in defining the electronic business rules for when a facility fee for a professional service was needed and automatically creating the facility charge based on the professional charge. The committee chose the Medicare OPPS status indicators and ambulatory payment classification (APC) status (with minor adjustments for nonfederal payers) to define the business rule for generating facility charges based on the professional charge. Doing so helped Vanderbilt not only generate previously missed revenue, but also eliminate the time spent by clinic staff filling out facility charge forms and by data-entry staff keying the facility charges.
At the time, clinic facility charges were keyed by data-entry staff from charge forms completed by physicians or clinic staff in a process comparable to that used for professional charges. However, provider charge forms were always completed by the physicians, not by clinic staff. The processes also had different final billing system destinations. Professional charge forms were keyed into the professional billing system, and facility charge forms were keyed into the hospital/facility billing system. Both systems had data-entry staff who entered the charges. Thus, the service destroy benign lesion (CPT 17000) generated two forms and required two data-entry teams to capture professional and facility charges.
Linking facility charges to professional charges made sense because Vanderbilt physicians, like physicians in many group practices, have productivity targets and incentives associated with the relative value units based upon the common procedure codes (CPTs) logged as a result of professional charges. In addition, Vanderbilt has reports to identify encounters with missing charges on the professional side but not on the facility side. Thus, linking facility charges (poorly tracked) to professional charges (well tracked) was likely to both increase the number of charges captured and eliminate some data entry for the relevant charges.
The largest revenue opportunity lay in capturing missed charges. Rough estimates, although incomplete, indicated that facility charges were missing for 4 percent to 46 percent of professional charges,as shown in exhibit 1 . This was true of both the E&M services and the CPTs for other office professional services. E&M codes were later excluded from the study because CMS ruled that matching physician level of service was not appropriate for facility-level charging on E&M services. The preprogram estimate was based on a crude count generated prior to implementation from the charge "mirroring" logic performed for a subset of the clinics. These data indicated that a variance existed by specialty and by CPT/Healthcare Common Procedure Coding System (HCPCS) code.
Missing charges were even found in a prospective audit of the manual professional and technical charge forms for a physician in a clinic with "exemplary" administrative processes. For this physician, nine of 28 visits with professional charges had no facility charges where the business rule indicated there should have been charges.
Why were charges missed?
First, some physicians and staff may have forgotten to mark the charge. Similarly, some physicians may have viewed the facility charge as excessive or an unjustified expense for their patients. Alternatively, physicians may have felt the facility charge made Vanderbilt's hospital-based practice less competitive with community-based group practices. Early project implementation meetings bore out this assumption. There were anecdotal reports that staff thought the facility charge was unreasonable or excessive ("sticker shock"). For example, nurses in one clinic thought a $50 office visit charge for taking a patient's blood pressure was excessive. In such instances, they might have "forgotten" to charge the facility fees.
Second, there may have been lapses in communication over time about the need to create new chargeable items in master files. Examples of that possibility surfaced during the project implementation meetings, where concern was raised that new items may have been added to the professional chargemaster, but not to the facility chargemaster. Professional charge maintenance appeared to be timelier than facility charge maintenance, possibly because physicians derived a portion of their income from their professional billings. Thus, physicians might have advocated and lobbied for professional charge updates. There was no similar advocacy for additions to facility charges. In fact, procedures were often missing not only from the facility charge form, but also from the facility chargemaster. Given that some new items never appeared on the facility charge form, users might have assumed there was no facility charge for the new service.
Third, there may have been a desire to shelter specific patient populations from charges. For example, preliminary data indicated the pediatric clinics appeared to have a particularly high rate of missing facility charges. High rates of variability between clinics also may have been due to differences in staff education on the policy and method about when billing regulations require a facility charge.
Fourth, facility charge forms might have been overlooked because physicians postponed completing the professional charge form. In the targeted clinic scenarios, the facility charge could not be completed until the professional charge form was completed by the physician, establishing both the service the patient received and the diagnosis for the visit. So in clinics where the support staff (as opposed to the physician) completed the facility charge form, the clinic staff had to wait until the physician had completed the professional form before they could complete the facility form.
Initial Results and Benefits
To measure the program's impact, the committee measured the proportion of matching professional and facility charges for a 28-day period immediately before and after the program's implementation. The effect varied by specialty. Exhibit 2 shows the percentage of matched professional and facility billing for the preintervention and postintervention period by specialty. For example, facility billing increased significantly for neurology and rehabilitation, general pediatrics, and obstetrics/gynecology, whereas cardiology and hematology/oncology showed a statistically insignificant decrease in facility billing. The small ancillary area included in the study, neurodiagnostics, had the only significant decrease in facility billing. For neurodiagnostics, the negative finding suggests the ancillary area had a well-honed manual process and a confusing automated process, which might have been corrected over time.
An unexpected finding was that even after the program was implemented, 37 percent of facility fees still went unbilled (at least at 28 days after implementation). Reasons the facility fee might still not be billed even after program implementation include:
- Substantially late billing of the professional fee such that the facility fee becomes unbillable (By policy, Vanderbilt does not rebill additional facility fees that were not posted at the time of the original bill.)
- Revolving accounts, such as those seen in hematology/oncology, where the late facility fees were billable but may not have passed billing system edits
- Lack of a facility system matching charge item (a maintenance issue)
- The programmatic charge write-off of facility fee charges
An example of the last category was found while collecting the data for the study. After identifying a large number of psychiatric procedures with no matching facility fees, the committee uncovered a large, erroneous programmatic write-off. Unfortunately, the write-off was discovered well
after the postimplementation study period, so it is reflected in the data. If the discrepancy had been noted at the time the program was implemented, the psychiatric procedure fees for 2003 could have been $2.8 million for the year (had they been billed).
The missing postimplementation facility fees also may have been a temporary artifact resulting from the recent implementation. Data collection started about five days after the program was implemented for each clinic. There could have been unresolved implementation issues, such as missing charge items, billing delays, or confusion over manual versus automated processes. However, the committee kept the time intervals before and after implementation close to the implementation date to reduce the probability of influences other than the program implementation contaminating the study results.
Based on the change in the proportion of facility charges billed (52 percent before and 63 percent after), the increase in facility fees charged after the intervention was estimated at $11 million (assuming Medicare payment rates). This amount was still less than the revenue opportunity of as much as $100 million if all eligible facility fees could have been recovered.
Other benefits included the elimination of five FTEs in data entry. Assuming the cost of one FTE was approximately $31,000 per year, Vanderbilt saved about $155,000 per year in data-entry costs. In addition, the time spent by the clinic or physician practice staffs in preparing and completing the form was eliminated, but these savings were not harvested from the clinics. Finally, the organization achieved some smaller savings in form creation and printing in a few clinics where the facility form was completely eliminated.
Keys to Success
Some of the lessons Vanderbilt learned from this initiative can help other healthcare organizations that want to achieve similar results. Taking a few key steps can help organizations capture their clinic charges accurately.
Measure results. The study group delayed so long in measuring the results of the initiative that a consultant recommended deinstalling the program. Without outcome measures, improvements cannot be assumed.
Look for charges left on the table. Consider "always" parallels between charges and transactions in other systems. For example, if there is a laboratory result on a specimen type of "blood," there is always a venipuncture charge. And if a one-hour appointment is checked in for a chemotherapy infusion, there is always a charge for chemotherapy infusion first hour (CPT 96410).
Perform ongoing audits. Leverage audits of professional bills by simultaneously auditing charges to ensure the revenue capture of the parallel transaction. The study group might have found the areas where the charge needed to be added to the master had they been randomly auditing the charge process.
Be prepared for some push-back. Patients and nonfederal payers may react negatively to corrected bills. For example, nonfederal payers cannot be charged less than what federal payers are billed. Before implementing a similar program, healthcare organizations should analyze rules and regulations for contracted rates and administrative costs for charging fees that would not be paid by nonfederal payers. Posting a high-level explanation of the change in billing on the web site may be helpful to patients.
The business rules surrounding billing are complex. Organizations would be well served to computerize the logic wherever possible. Even in the one Vanderbilt clinic where trained coders abstracted the documentation and completed the facility billing forms, some charges were "left on the table." Computerizing the logic can let clinicians focus on helping patients, while knowing that the correct amount of revenue will be collected.
Fern FitzHenry, PhD, RN, is an assistant professor, department of biomedical informatics, Vanderbilt University, Nashville, Tenn. (firstname.lastname@example.org).
Linda Martin is former compliance director, Vanderbilt University, Nashville, Tenn., and current vice president, corporate compliance and privacy officer, UT Medical Group, Inc., Memphis, Tenn. (email@example.com).
Johniene Doran is a health systems programmer analyst III, Vanderbilt University, Nashville, Tenn. (firstname.lastname@example.org).
The authors acknowledge the assistance of the design and implemenatation team, especially Lloyd Petteys, Cynthia Moe, and Cass Fagan for their work on the analysis, programming, and implementation of the system changes.
Publication Date: Tuesday, September 01, 2009