Variation in Annual Wellness Visit Use and Risk Scores Across Medicare Advantage Insurers

Abstract

This cross-sectional study assesses the use of Medicare’s Annual Wellness Visits among older patients with Medicare Advantage and traditional Medicare coverage.

Introduction

Medicare Advantage (MA) plans receive capitated, risk-adjusted payments depending on beneficiary health status.^1,2 One mechanism for documenting health status is the Medicare Annual Wellness Visit (AWV), a zero-cost-sharing preventive benefit introduced under the Patient Protection and Affordable Care Act in 2011.³ Although designed for early disease detection and prevention, AWVs’ role in documenting diagnoses that contribute to higher risk-adjusted payments has raised policy interest.^4,5,6 Studies show higher AWV use in MA than traditional Medicare (TM),² but it remains unclear whether this variation is systematic across MA insurers and whether AWV uptake in MA is associated with differences in measured risk. To evaluate the strategic use of AWVs, we examined insurer-level variation in AWV use and its association with Hierarchical Condition Category (HCC) risk scores.

Methods

We analyzed 2019 TM and MA Medicare claims for a 20% sample of beneficiaries 65 years or older (eMethods, eTable in Supplement 1). Claims were linked to Centers for Medicare & Medicaid Services Monthly Enrollment by contract/plan/state/county to identify MA insurers, including UnitedHealth Group (UHG), Humana, CVS Health, Anthem, WellCare Health Plans, and other plans. The Johns Hopkins Bloomberg School of Public Health Institutional Review Board deemed this cross-sectional study exempt from review and consent requirement because administrative data were used. We followed the STROBE reporting guideline.

The primary outcome was a dichotomous indicator of AWV uptake in 2019. The secondary outcome was the HCC score, capturing health risks relevant to risk adjustment. Our key explanatory variable was Medicare insurance type: TM or MA. We estimated an ordinary least squares model for both outcomes.

Two-sided P < .05 indicated statistical significance. Data analysis was performed from April to September 2025 using Stata 14.1 (StataCorp).

Results

Of 7 351 522 beneficiaries (mean [SD] age, 75.5 [7.3] years; 4 145 498 females [56.4%]), 38.8% had an AWV. Uptake was higher in MA (44.1%) than in TM (36.1%) but varied substantially across insurers, from 53.9% for UHG to 30.6% for Anthem (Figure).

Figure. Bar Chart of Annual Wellness Visits (AWVs) Across Major Medicare Insurance Plans in 2019.

The dashed horizontal line represents the overall mean uptake (38.8%) among Medicare beneficiaries in both Medicare Advantage and traditional Medicare (TM) fee-for-service plans.

MA enrollment was associated with increased probability of AWV uptake compared with TM (0.078 [95% CI, −0.019 to 0.176] percentage points; P < .10). However, after excluding UHG, the difference between MA and TM was no longer statistically significant (Table).

Table. Association Between Medicare Insurance Type and Annual Wellness Visit Uptake and HCC Score.

Variable	β-Coefficient (95% CI)
	TM fee-for-service and MA plans	UnitedHealth Group Inc	All other plans
Binarya
AWV uptake: yes or no, percentage point
TM enrollment	1 [Reference]	1 [Reference]	1 [Reference]
MA enrollment	0.078 (−0.019 to 0.176)b	0.178 (0.177 to 0.179)c	0.030 (−0.026 to 0.086)
Continuousd
HCC risk score, mean (SD)	1.145 (1.113)	1.129 (1.113)	1.121 (1.098)
MA enrollment	0.100 (0.071 to 0.129)c	0.134 (0.131 to 0.138)c	0.088 (0.072 to 0.105)c
AWV uptake	−0.045 (−0.048 to −0.041)c	−0.045 (−0.047 to −0.043)c	−0.046 (−0.048 to −0.043)c
MA AWV uptakeb	0.083 (−0.021 to 0.188)c	0.160 (0.155 to 0.165)c	0.022 (−0.005 to 0.050)b

Abbreviations: AWV, Annual Wellness Visit; HCC, Hierarchical Condition Category; MA, Medicare Advantage; TM, traditional Medicare.

^a The coefficient was interpreted as a change by percentage points. For example, in the first column, compared with TM, MA is associated with a 7.8 percentage point increase in the probability of using AWVs.

^b P < .10.

^c P < .001.

^d The coefficient was interpreted as a change in HCC score. For example, in the first column, AWV use among MA enrollees is associated with significantly higher HCC scores of 0.083.

The mean (SD) HCC score was 1.145 (1.113). AWV use among MA enrollees was associated with significantly higher HCC scores (0.083 [95% CI, −0.021 to 0.188]; P < .001), particularly for UHG (0.160 [95% CI, 0.155-0.165]; P < .001). In contrast, among all other plans, AWV uptake in MA was associated with a minor increase in HCC score (0.022 [95% CI, −0.005 to 0.050]; P < .001) (Table).

Discussion

We observed substantial insurer-level variation in AWV use among MA plans. The overall increase in AWV use in MA was largely driven by UHG, which had markedly higher uptake than the others.

While beneficiaries using AWVs generally have lower HCC risk scores, those in MA using AWVs had substantially higher risk scores than those in TM. When only UHG MA enrollees were compared with TM enrollees, the increase in risk scoring for those with AWVs doubled. These findings suggest that insurer-specific strategies, rather than uniform features of MA, contribute to observed differences in AWV adoption and associated risk scores. While AWVs can improve documentation for care planning, their use as an entry point for coding intensity raises concerns about payment accuracy.

This study is descriptive and cannot determine whether diagnoses documented during AWVs reflect better identification of underlying conditions, differences in coding practices, or other unmeasured factors. Limitations include the cross-sectional design, potential selection into MA, and limited adjustment for fine-grained geographic variation. As MA enrollment increases, regulatory oversight should ensure AWVs are used as clinical tools to enhance care rather than as financial instruments to inflate risk-adjusted payments. Future research should evaluate whether these visits lead to downstream value-based services and improved clinical outcomes.

Supplement 1.

eMethods

eTable. Sample Selection

Supplement 2.

Data Sharing Statement