Achieving Productivity Expectations Among General Internal Medicine Clinicians at an Urban Safety-Net Academic Medical Center

Urban safety-net medical centers traditionally have been viewed as unlikely to meet standard productivity expectations.(1) In recent years, a number of safety-net hospitals in large metropolitan areas have closed due to lack of financial viability.(2,3) Patient, clinician, and systems factors may contribute to these productivity and revenue challenges. Patients receiving care in urban safety-net settings have a high disease burden, as well as psychosocial complexity, including mental illness, disability, substance use disorders, and low health literacy.(4-6) As a result, clinicians spend a significant proportion of their time on uncompensated activities, such as care coordination. Thus, in a fee-for-service model, the efforts needed to care for psychosocially complex patients often are not billable. Moreover, the faculty in academic settings are predominantly part-time clinicians. They often focus most of their activities in areas outside of direct patient care, such as research, education, and public health. Further, the faculty who choose to practice in safety-net settings may be mission-driven rather than financially driven, and they may resist initiatives focused on improving revenue.

System-based factors also impede clinical productivity. Academic practices focus on research and educational initiatives and often have less administrative support than other practice models.(7) The lack of administrative support also increases the amount of clinician time that must be dedicated to uncompensated activities, such as returning patient phone calls and completing paperwork. Finally, a payer mix with a higher proportion of government-funded health plans (e.g., Medicaid) results in overall lower compensation rates compared with those in other practice settings.(7)

In response to increasing financial pressure on our institution to achieve productivity expectations, we implemented a multicomponent intervention to improve clinician relative value unit (RVU) productivity in a general internal medicine group at an urban safety-net academic medical center. RVUs are a measure of value used in the United States Medicare reimbursement formula for physician services,(8) and comprise part of the resource-based relative value scale (RBRVS). Although medical literature describes some productivity models at safety-net medical centers, we are unaware of programs similar to the one we describe here.(9,10)

Methods

Setting

The General Internal Medicine section is a large academic medical group based at Boston Medical Center, the largest safety-net hospital in New England. A total of 72 clinicians deliver care in the outpatient, inpatient, or both settings. For more than 10 years, clinicians have had individual RVU targets, the primary metric used to measure clinical productivity. Based on the complexity of each patient encounter, the clinician designates a current procedural terminology (CPT) code. The RBRVS assigns each CPT code an RVU value—the higher the complexity of the CPT code, the higher the RVU value.

Intervention

The multicomponent intervention implemented in 2012 included three major components: communication, coaching, and incentives/disincentives. We collected RVU productivity data over five years, from 2012 to 2016.

Communication

The Section Chief of General Internal Medicine outlined productivity expectations to clinicians. The section administrative staff members gave each clinician frequent feedback on his or her productivity progress over time in the form of a monthly dashboard (see Appendix). The dashboard outlined clinician productivity metrics including RVUs, percentage breakdown of CPT codes used (e.g., percentage of level 3 versus level 4 visits), clinical volume, and time in clinic (session counts).

Coaching

The medical director of the outpatient practice, who is also a practicing clinician in the medical group, functioned as the clinician coach. The coach held periodic group trainings that focused on clinicians’ coding. The group trainings reviewed appropriate use of CPT codes and improvement of medical record documentation to reflect the complexity of care provided. The coach emphasized an approach of “working smarter, not necessarily harder,” which encouraged clinicians to capture and appropriately credit what they were already doing. For example, in treating patients with diabetes mellitus, the coach trained the clinicians to document whether the patient was insulin-dependent or whether there were any renal, ophthalmic, or neurological manifestations of the disease. By documenting these factors, clinicians could use appropriate CPT codes reflecting these complexities. Through chart audits, the coach found that clinicians were delivering complex care, yet the CPT codes attributed to these encounters were lower than expected (that is, faculty were under-coding at times).

Faculty who were not achieving productivity expectations and those who requested support had individual sessions with the coach. During these coaching sessions, clinicians received individualized and targeted feedback on how to improve their productivity. The coach also provided instruction on efficiency, in both practice flow and utilization of the electronic health record (EHR). For example, the coach encouraged the use of documentation shortcuts (also known as “smart phrases”), which insert data or text into a note by typing short phrases. The coach introduced techniques such as the OHIO (only handle it once) principle for managing results and staff messages in the EHR. The coach also reviewed visit notes to look for opportunities to improve documentation. Finally, the coach provided education on payment models and emphasized that the viability of the general internal medicine group and medical center was dependent on achieving productivity expectations.

Incentives and Disincentives

Clinicians were eligible to receive financial incentives for productivity above expectations. Incentives were a dollar amount in the $10 to $20 range per RVU for each RVU achieved above the clinicians’ target. Clinicians who did not achieve 90% of their target RVUs were at risk for a salary reduction of up to 10%. However, if the entire general internal medicine group achieved their collective target, no individual would be at risk for salary reduction.

Results

We observed an 11% increase in RVU productivity over five years (Figure 1). In 2012, clinicians produced 3774 RVUs per FTE. Annual RVUs per FTE increased to 4205 by 2016. Multiple severe blizzards resulted in market reduction of clinical volume in the winter of 2015. In addition, we implemented a new EHR system in 2015, which likely impacted clinical productivity. Despite these factors, RVUs continued to increase in 2015. Approximately one-third of faculty received a clinical productivity bonus every year; no faculty received salary penalties during this period. The percentage of faculty who departed the practice trended downward over the five-year period (Figure 1).

Figure 1. Yearly relative value unit (RVU) production and clinician attrition rate. Total RVUs per year were combined, as was total FTE. Average RVU per FTE was calculated. Faculty in their first year of practice, clinicians with extended leaves, fellows, and residents were excluded. Attrition rate was defined as percent of providers leaving the practice per year.

Discussion

After implementing a multicomponent intervention to improve clinician productivity, we observed a sustained increase in RVUs over a five-year period among general internal medicine clinicians practicing at an urban safety-net academic medical center. This program may have enabled our practice to remain solvent in an era when other safety-net medical centers were closing for financial reasons.(2,3) The increase in productivity followed implementation of periodic dashboard reporting, use of an internal productivity coach, group education sessions, targeted individual interventions, delineation of expectations, and engagement of faculty in the financial mission of the institution.

We believe acquiring the skills needed to achieve productivity expectations allowed clinicians to gain a sense of control and empowerment that helped motivate them to achieve those expectations. The coach was a practicing member of the practice and served as a role model. As a clinical leader, he took a supportive but firm stance in delineating expectations while providing faculty with the tools and support they needed to achieve productivity targets.

In tandem with increased productivity, we observed a decline in the number of departing faculty over the five-year period.

We implemented a new EHR in 2015, which likely impacted clinical productivity, but RVUs continued to increase in 2015. Even though we scheduled fewer patients for one month at the beginning of the new EHR implementation (11 rather than 12 patients per four-hour session), the expected yearly productivity target remained the same. Two providers used scribes during this period, and an EHR “super user” in the group provided coaching on efficient EHR use. These supports may have contributed to the continued increase in clinical productivity during this challenging transition period.

In tandem with increased productivity, we observed a decline in the number of departing faculty over the five-year period. This decline was surprising, given that increasing productivity pressure can contribute to the development of clinician burnout,(11) which can impact faculty retention. We did not collect data on clinician burnout; it is unclear, therefore, if the decrease in departing faculty is a proxy for lower levels of burnout. It is possible that the sense of control the clinicians gained in our model may have reduced the risk of burnout.

Our analysis has limitations. We do not have RVU productivity data prior to the intervention to compare the trend in productivity over time. It is also unclear whether there was a general trend of increased RVU productivity with the increased emphasis on coding nationwide. There may have been unmeasured confounders that contributed to the increase in productivity. For instance, if clinicians who left the practice during the intervention period suffered higher rates of burnout and were chronically underperforming below the productivity target, the productivity would appear to have increased after they departed. Finally, we did not have a control group of clinicians who did not receive interventions to increase productivity.

Our program may have limited generalizability. RVU-based payment models may be replaced by other models such as accountable care organizations, which employ global payment rather than fee-for-service payment. Nevertheless, many of the intervention components used in our productivity program could be applicable to other payment models.

In conclusion, despite the barriers to achieving clinical productivity in an urban safety-net environment, our program demonstrated that productivity can increase while preserving faculty retention.

References

1. Duck E, DeLia D, Cantor JC. Primary care productivity and the health care safety net in New York City. J Ambul Care Manage. 2001;24(1):1-14.

2. Whiteis DG. Hospital and community characteristics in closures of urban hospitals, 1980-87. Public Health Rep. 1992;107:409-416.

3. Williams D, Hadley J, Pettengill J. Profits, community role, and hospital closure: an urban and rural analysis. Med Care. 1992;30:174-187.

4. Liebschutz J, Saitz R, Brower V, et al. PTSD in urban primary care: high prevalence and low physician recognition. J Gen Intern Med. 2007;22:719-726. doi:10.1007/s11606-007-0161-0.

5. Lorant V, Deliege D, Eaton W, Robert A, Philippot P, Ansseau, M. Socioeconomic inequalities in depression: a meta-analysis. Am J Epidemiol. 2003;157(2):98-112.

6. Olfson M, Mechanic D. Mental disorders in public, private nonprofit, and proprietary general hospitals. Am J Psychiatry. 1996;153:1613-1619. doi:10.1176/ajp.153.12.1613.

7. Zaman OS, Cummings LC, Laycox S. America’s safety net hospitals and health systems, 2010. Results of the Annual NAPH Hospital Characteristics Survey. National Public Health and Hospital Institute. 2012. https://essentialhospitals.org/wp-content/uploads/2013/12/NPH214.pdf . Accessed October 15, 2018.

8. Reinhardt UE. The little-known decision-makers for Medicare physicians’ fees. 2010. https://economix.blogs.nytimes.com/2010/12/10/the-little-known-decision-makers-for-medicare-physicans-fees . Accessed May 11, 2018.

9. Scoggins CR, Crockett T, Wafford L, Cannon RM, McMasters KM. Improving clinical productivity in an academic surgical practice through transparency. J Am Coll Surg, 2013;217:46-51. doi:10.1016/j.jamcollsurg.2013.01.066.

10. Stewart MG, Jones DB, Garson AT. An incentive plan for professional fee collections at an indigent-care teaching hospital. Acad Med. 2001;76:1094-1099.

11. Linzer M, Visser MR, Oort FJ, Smets EM, McMurray JE, de Haes HC; Society of General Internal Medicine Career Satisfaction Study Group. Predicting and preventing physician burnout: results from the United States and the Netherlands. Am J Med. 2001;111(2):170-175.

Appendix

Appendix continues below