Abstract
This cohort study analyzes quality incentives, requirements, and other measures in employment and payer contracts of primary care clinicians.
Introduction
Value-based contracts are popular for quality improvement in primary care despite mixed evidence of their effectiveness.1,2,3 Explanations for their underperformance include complexity of health care, misalignment of measures, and inadequate financial incentives.1,2,3 Another potential unexplored factor is volume of quality measures, especially if clinicians face multiple contracts featuring different quality measures and reporting requirements. We quantified the number and diversity of quality measures and value-based contracts faced by primary care physicians (PCPs).
Methods
We obtained employment contract data on PCPs continuously employed by an integrated health system from 2020 to 2022 along with payer contracts associated with their attributed patients. Patients who interacted with the health system in the previous 2 years were attributed to 1 PCP annually and linked to a payer contract based on their insurance plan at year end (eMethods in Supplement 1). The Providence Research Network Institutional Review Board approved the study and waived informed consent because it was not considered human participant research. We followed the STROBE reporting guidelines.
Payer contract data included type (commercial, Medicaid, or Medicare), incentivized process- or outcome-based quality measures, and Health Care Payment Learning & Action Network (HCPLAN) category.4 HCPLAN categories 2C, 3A, and 3B were considered value-based contracts.
We measured the number of unique value-based contracts and quality measures per physician-year based on their assigned patients. Quality measures were considered distinct if they referenced different conditions, and measures for the same condition were considered distinct if the value differed (eg, hemoglobin A1c <8% vs <9% [to convert to proportion of total hemoglobin, multiply by 0.01]). We conducted 2-sample t tests to assess changes in exposure to value-based contracts and quality measures across years; P < .05 indicated statistical significance. Results were robust to excluding PCPs with small panels.
Quality measure information was missing for 29% of value-based contracts with attributed patients; thus, we reported both mean number of contracts and mean number of contracts with nonmissing quality measures per physician-year. Analyses were run with R 4.4.0 (R Core Team).
Results
The 809 PCPs included (519 females [58.3%], 371 males [41.7%]) had a mean (SD) of 1308.71 (622.73) attributed patients and an increasing number of value-based contracts from 2020 to 2022 (9.39 to 12.26; P < .001) (Table). Contracts contained a mean (SD) of 10.24 (2.66) quality measures. Physicians faced a mean (SD) 57.08 (24.58) unique quality measures across 7.62 (5.08) value-based contracts (Table). Distinct measures per physician ranged from 0 to 103 (Figure). Medicare contracts had more quality measures on average than commercial or Medicaid contracts (13.42 vs 10.07 or 5..37). The mean (SD) number of quality measures in Medicare contracts increased significantly from 13.14 (4.72) in 2020 to 15.04 (3.99) in 2022 (P < .001) (Table).
Table. Changes in Value-Based Care Contracts and Quality Measures Over Timea.
| Mean (SD) | Difference between 2022 and 2020 | P valueb | ||||
|---|---|---|---|---|---|---|
| Overall | 2020 | 2021 | 2022 | |||
| PCPs, No.c | 890 | 890 | 890 | 890 | NA | NA |
| No. of patients in physician’s panel | 1308.71 (622.73) | 1272.17 (601.76) | 1336.39 (606.09) | 1317.58 (657.73) | 45.42 | .13 |
| Among contracts in a physician’s patient panel | ||||||
| No. of value-based contracts | 11.18 (5.70) | 9.39 (5.28) | 11.89 (5.43) | 12.26 (5.94) | 2.87 | <.001 |
| No. of value-based contracts with nonmissing quality measure data | 7.62 (5.08) | 7.79 (5.71) | 8.31 (4.66) | 6.74 (4.67) | −1.05 | <.001 |
| No. of unique quality measures | 57.08 (24.58) | 54.78 (24.75) | 64.08 (24.71) | 52.37 (22.67) | −2.41 | .03 |
| Among value-based contracts in a physician’s patient panel | ||||||
| Share of contracts by payer type, %d | ||||||
| Commercial | 49.50 | 55.02 | 48.64 | 44.65 | −0.10 | <.001 |
| Medicaid | 21.49 | 14.78 | 17.68 | 32.43 | 0.18 | <.001 |
| Medicare | 29.01 | 30.20 | 33.68 | 22.93 | −0.07 | <.001 |
| No. of measures per contracte | ||||||
| Overall | 10.24 (2.66) | 10.23 (2.14) | 10.38 (2.92) | 10.11 (2.85) | −0.12 | .31 |
| Commercial | 10.07 (2.29) | 9.92 (1.50) | 10.69 (2.43) | 9.58 (2.66) | −0.34 | .002 |
| Medicaid | 5.37 (1.00) | 4.54 (0.88) | 5.61 (0.53) | 5.70 (1.10) | 1.16 | <.001 |
| Medicare | 13.42 (4.72) | 13.14 (4.72) | 12.42 (4.92) | 15.04 (3.99) | 1.91 | <.001 |
Abbreviations: NA, not applicable; PCP, primary care physician.
Value-based contracts were Health Care Payment Learning & Action Network categories 2C, 3A, and 3B.
P value from a 2-sided t test.
The PCPs were employed by the health system for all 3 study years.
Denotes statistics among all value-based contracts.
Denotes statistics among value-based contracts with nonmissing quality metrics data.
Figure. Number of Unique Quality Measures Among Physicians' Primary Care Patient Contracts, by Year.
Sum of unique quality measure among value-based care contracts in primary care physicians’ patient panels (n = 890 physicians). A unique quality measure is, for example, share of patients with blood pressure under 140/90 mm Hg. If the same metric appeared in multiple contracts, it was counted once. Vertical lines represent the mean number of quality measures per physician: 54.8, 64.1, and 52.4 in 2020, 2021, and 2022 respectively.
Discussion
Previous research on value-based contracts suggests these models have not lived up to their potential.1,2,3 We found saturation of the quality measure environment as a possible explanation: average physicians were incentivized to meet 57.08 different quality measures annually.
Study limitations include estimates that were likely lower bounds on PCPs’ exposure to quality measures. Physicians often face additional quality measures in their employment contracts, but our data on contracts’ quality measures were incomplete. Percentage of missing data was lower in 2021, which may explain the larger number of unique quality measures that year. Additionally, the data source was an integrated health system with multiple payers; thus, findings may not generalize to other settings.
Value-based contracting is intended to incentivize care improvement, but it is unlikely a clinician or practice can reasonably optimize against 50 or more measures at a time. Increased use of such levers may also carry unintended consequences. Clarity and salience are crucial to changing behavior,5 and the burden of extraneous information and processes has been increasingly associated with adverse outcomes, such as physician burnout.6 As payers increasingly shift toward value-based contracts, additional research is needed to understand how their ubiquity affects their benefits and how such contracts can be scaled sustainably for clinical care.
eMethods. Sample and Attribution of Patients to Physicians
Data Sharing Statement
References
- 1.Milad MA, Murray RC, Navathe AS, Ryan AM. Value-based payment models in the commercial insurance sector: a systematic review. Health Aff (Millwood). 2022;41(4):540-548. doi: 10.1377/hlthaff.2021.01020 [DOI] [PubMed] [Google Scholar]
- 2.McWilliams JM. Pay for performance: when slogans overtake science in health policy. JAMA. 2022;328(21):2114-2116. doi: 10.1001/jama.2022.20945 [DOI] [PubMed] [Google Scholar]
- 3.Bond AM, Schpero WL, Casalino LP, Zhang M, Khullar D. Association between individual primary care physician merit-based incentive payment system score and measures of process and patient outcomes. JAMA. 2022;328(21):2136-2146. doi: 10.1001/jama.2022.20619 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.APM Framework. HCPLAN. 2017. Accessed September 19, 2023. https://hcp-lan.org/apm-framework/
- 5.Loewenstein G, Asch DA, Volpp KG. Behavioral economics holds potential to deliver better results for patients, insurers, and employers. Health Aff (Millwood). 2013;32(7):1244-1250. doi: 10.1377/hlthaff.2012.1163 [DOI] [PubMed] [Google Scholar]
- 6.DiGiorgio AM, Ehrenfeld JM, Miller BJ. Improving health care quality measurement to combat clinician burnout. JAMA. 2023;330(12):1135-1136. doi: 10.1001/jama.2023.15512 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eMethods. Sample and Attribution of Patients to Physicians
Data Sharing Statement