Federal Practitioner

Diabetes mellitus (DM) is a national health crisis affecting > 38 million people (11.6%) in the United States.¹ American Indian and Alaska Native (AI/AN) adults are disproportionately affected, with a prevalence of 14.5%—the highest among all racial and ethnic groups.¹ Type 2 DM (T2DM) accounts for 90% to 95% of all DM cases and is a leading cause of morbidity and mortality due to its association with cardiovascular disease, kidney failure, and other complications.²

Maintaining glycemic control is important for managing T2DM and preventing microvascular and macrovascular complications.³ The cornerstone of diabetes self-management has been patient self-monitored blood glucose (SMBG) using finger-stick glucometers.⁴ However, SMBG provides measurements from a single point in time and requires frequent, painful, and inconvenient finger pricks, leading to decreased adherence.^5,6 These limitations negatively affect patient engagement and overall glycemic control.⁷

Continuous glucose monitors (CGMs) offer real-time, continuous glucose readings and trends.⁸ CGMs improve glycemic control and reduce hypoglycemic episodes in patients who are insulin-dependent.^9,10 Flash glucose monitors, a type of CGM that requires scanning to obtain glucose readings, provide similar benefits.¹¹ Despite these demonstrated advantages, research has primarily focused on insulin-dependent populations, leaving a significant gap in understanding the effect of CGMs on patients with T2DM who are not insulin-dependent.¹²

Given the high prevalence of T2DM among AI/AN populations and the potential benefits of CGMs, this study sought to evaluate the effect of CGM use on glycemic control and other health metrics in patients with non–insulin-dependent T2DM in an AI/AN population. This focus addresses a critical knowledge gap and may inform clinical practices and policies to improve diabetes management in this high-risk group.

Methods

A retrospective observational study was conducted using deidentified electronic health records (EHRs) from 2019 to 2024 at a federally operated outpatient Indian Health Service (IHS) clinic serving an AI/AN population in the IHS Portland Area (Oregon, Washington, Idaho). The study protocol was reviewed and deemed exempt by institutional review boards at Washington State University and the Portland Area IHS.

Study Population

This study included patients diagnosed with non–insulin-dependent T2DM, had used a CGM for ≥ 1 year, and had hemoglobin A_1c (HbA_1c) measurements within 4 months prior to CGM initiation (baseline) and within ± 4 months after 1 year of CGM use. For other health metrics, including blood pressure (BP), weight, low-density lipoprotein cholesterol (LDL-C), and estimated glomerular filtration rate (eGFR), this study required measurements within 6 months before CGM initiation and within 6 months after 1 year of CGM use. The baseline HbA_1c in the dataset ranged from 5.3% to > 14%.

Patients were excluded if they used insulin during the study period, had incomplete laboratory or clinical data for the required time frame, or had < 1 year of CGM use. The dataset did not include detailed information on oral DM medications; thus, we could not report or account for the type or number of oral hypoglycemic agents used by the patients. The IHS clinical applications coordinator compiled the dataset from the EHR, identifying patients who were prescribed and received a CGM at the clinic. All patients used the Abbott Freestyle Libre CGM, the only formulary CGM available at the clinic during the study period.

A 1-year follow-up endpoint was selected for several reasons: (1) to capture potential seasonal variations in diet and activity; (2) to align with the clinic’s standard practice of annual comprehensive diabetes evaluations; and (3) to allow sufficient time for patients to adapt to CGM use and reflect any meaningful changes in glycemic control.

All patients received standard DM care according to clinic protocols, which included DM self-management education and training. Patients met with the diabetes educator at least once, during which the educator emphasized making informed decisions using CGM data, such as adjusting dietary choices and physical activity levels to manage blood glucose concentrations effectively.

A total of 302 patients were initially identified. After applying exclusion criteria, 132 were excluded due to insulin use, and 77 were excluded due to incomplete HbA_1c data within the specified time frames (Figure 1). The final sample included 93 patients.

Measures

The primary outcome was the change in HbA_1c levels from baseline to 1 year after CGM initiation. Secondary outcomes included changes in weight, systolic and diastolic BP, LDL-C concentrations, and eGFR. For the primary outcome, HbA_1c values were collected within a grace period of ± 4 months from the baseline and 1-year time points. The laboratory’s upper reporting limit for HbA_1c was 14%; values reported as “> 14%” were recorded as 14.1% for data analysis, although the actual values could have been higher.

For secondary outcomes, data were included if measurements were obtained within ± 6 months of the baseline and 1-year time points. Patients who did not have measurements within these time frames for specific metrics were excluded from secondary outcome analysis but remained in the overall study if they met the criteria for HbA_1c and CGM use.

Statistical Analysis

Statistical analysis was performed using R statistical software version 4.4.2. Paired t tests were conducted to compare baseline and 1-year follow- up measurements for variables with parametric distributions. Wilcoxon signed-rank test was used for nonparametric data. A linear regression analysis was conducted to examine the relationship between baseline HbA_1c levels and the change in HbA_1c after 1 year of CGM use. Differences were considered significant at P < .05 set a priori. To guide future research, a posthoc power analysis was performed using Cohen’s d to estimate the required sample sizes for detecting significant effects, assuming a similar population.

Results

The study included 93 patients, with a mean (SD) age of 55 (13) years (range, 29-83 years). Of the participants, 56 were female (60%) and 37 were male (40%). All participants were identified as AI/AN and had non–insulin-dependent T2DM.

Primary Outcomes

A significant reduction in HbA_1c levels was observed after 1 year of CGM use. The mean (SD) baseline HbA1c was 9.5% (2.4%), which decreased to 7.6% (2.2%) at 1-year follow-up (Table 1). This difference represents a mean change of -1.86% (2.4%) (95% CI, -2.35 to -1.37; P < .001 [paired t test, -7.53]).

A linear regression model evaluated the relationship between baseline HbA_1c (predictor) and the change in HbA_1c after 1 year (outcome). The change in HbA_1c was calculated as the difference between 1-year follow-up and baseline values. The regression model revealed a significant negative association between baseline HbA_1c and the change in HbA_1c (Β = -0.576; P < .001), indicating that higher baseline HbA_1c values were associated with greater reductions in HbA_1c over the year. The regression equation was: Change in HbA_1c = 3.587 – 0.576 × Baseline HbA_1c

The regression coefficient for baseline HbA_1c was -0.576 (standard error, 0.083; t = -6.931; P < .001), indicating that for each 1% increase in baseline HbA_1c, the reduction of HbA_1c after 1 year increased by approximately 0.576% (Figure 2). The model explained 34.6% of the variance in HbA_1c change (R² = .345; adjusted R² = .338).

Secondary Outcomes

Systolic BP decreased by a mean (SD) -4.9 (17) mm Hg; 95% CI, -8.6 to -1.11; P = .01, paired t test). However, no significant change was observed for diastolic BP (P = .77, paired t test). Similarly, no significant changes were observed in weight, LDL-C concentrations, or eGFR after 1 year of CGM use. A posthoc power analysis indicated that the study was underpowered to detect smaller effect sizes in secondary outcomes. For example, sample size estimates indicated that detecting significant changes in weight and LDL-C concentrations would require sample sizes of 152 and 220 patients, respectively (Table 2).

Discussion

This study found a clinically significant reduction in HbA_1c levels after 1 year among AI/AN patients with non–insulin-dependent T2DM who used CGMs. The mean HbA_1c decreased 1.9%, from 9.5% at baseline to 7.6% after 1 year. This reduction is not only statistically significant (P < .001), it is clinically meaningful—even a 1% decrease in HbA_1c is associated with substantial reductions in the risk of microvascular complications.³ The magnitude of the HbA_1c reduction observed suggests CGM use may be associated with improved glycemic control in this high-risk population. By achieving lower HbA_1c levels, patients may experience improved long-term health outcomes and a reduced burden of DM-related complications.

Changes in oral DM medications during the study period may have contributed to the observed improvements in HbA_1c levels. While the dataset lacked detailed information on types or dosages of oral hypoglycemic agents used, adjustments in medication regimens are common in DM management and could significantly affect glycemic control. The inability to account for these changes results in an inability to attribute the improvements in HbA_1c solely to CGM use. Future studies should collect comprehensive medication data to better isolate the effects of CGM use from other treatment modifications.

Another factor that may have contributed to the improved glycemic control is the DM self-management education and training patients received as part of standard care. Patients met with diabetes educators at least once and learned how to use the CGM device and interpret the data for self-management decisions. This education may have enhanced patient engagement and empowerment, enabling them to make informed choices about diet, physical activity, and medication adherence. Studies have shown that DM self-management education can significantly improve glycemic control and patient outcomes.¹³ By combining the CGM technology with targeted education, patients may have been better equipped to manage their condition, contributing to the observed reduction in HbA_1c levels. Future studies should consider synergistic effects of CGM use and DM education when evaluating interventions for glycemic control.

The significant reduction in HbA_1c indicates CGM use is associated with improved glycemic control in non–insulin-dependent T2DM. The linear regression analysis suggests patients with poorer glycemic control at baseline experienced greater reductions in HbA_1c over the course of 1 year. This finding aligns with previous studies that have shown greater HbA_1c reductions in patients with higher initial levels when using CGMs. Yaron et al reported similar findings: higher baseline HbA_1c levels predicted more substantial improvements with CGM use in patients with T2DM on insulin therapy.¹⁴

This study contributes to existing research by examining the association between CGM use and glycemic control in patients with non– insulin-dependent T2DM within an AI/AN population, a group that has been underreported in previous studies. Most prior research has focused on insulin-dependent patients or populations with different ethnic backgrounds.12 By focusing on patients with non–insulin-dependent T2DM, this study highlights the broader applicability of CGMs beyond traditional use, showcasing their potential association with benefits in earlier stages of DM management. Targeting the AI/AN population addresses a critical knowledge gap, given the disproportionately high prevalence of T2DM and associated complications in this group. The findings of this study suggest integrating CGM technology into the standard care of AI/AN patients with non–insulin-dependent T2DM may be associated with improved glycemic control and may help reduce health disparities.

The modest decrease in systolic BP observed in this study may indicate potential cardiovascular benefits associated with CGM use, possibly due to improved glycemic control and increased patient engagement in self-management. However, given the limited sample size and exclusion criteria, the study lacked sufficient power to detect significant associations between CGM use and other secondary outcomes such as BP, weight, LDL-C, and eGFR. Therefore, the significant finding with systolic BP should be interpreted with caution.

The lack of significant changes in secondary outcomes may be attributed to the study’s limited sample size and the relatively short duration for observing changes in these parameters. Larger studies are needed to assess the full impact of CGM on these variables. The required sample sizes for achieving adequate power in future studies were calculated, highlighting the utility of our study as a pilot, providing critical data for the design of larger, adequately powered studies.

Limitations

The retrospective design of this study limits causal inferences. Moreover, potential confounding variables were not controlled, such as changes in medication regimens (other than insulin use), dietary counseling, or physical activity. Additionally, we could not account for the type or number of oral DM medications prescribed to patients. The dataset included only information on insulin use, without detailed records of other antidiabetic medications. This limitation may have influenced the observed change in glycemic control, as variations in medication regimens could affect HbA_1c levels.

Because this study lacked a comparator group, the effect of CGM use cannot be definitively isolated from other factors (eg, medication changes, dietary modifications, or physical activity). Moreover, CGM devices can be costly and are not universally covered by all insurance or IHS programs, potentially limiting widespread implementation. Policy-level restrictions and patient-specific barriers may also hinder feasibility in other settings.

The small sample size may limit the generalizability of the findings. Of the initial 302 patients, about 69% were excluded due to insulin use or incomplete laboratory data. A ± 4-month window was selected to balance data quality with real-world practices. Extending this window further (eg, ± 6 months) might have included more participants but risked diluting the 1-year endpoint consistency. The lack of statistical significance in secondary metrics may be due to insufficient power rather than the absence of an effect.

Exclusion of patients due to incomplete data may have introduced selection bias. However, patients were included in the overall analysis if they met the criteria for HbA_1c and CGM use, even if they lacked data for secondary outcomes. Additionally, the laboratory’s upper reporting limit for HbA_1c was 14%, with values above this reported as “> 14%.” For analysis, these were recorded as 14.1%, which may underestimate the true baseline HbA_1c levels and impact of the assessment of change. This occurred for 4 of the 93 patients included.

All patients used the Freestyle Libre CGM, which may limit the generalizability of the findings to other CGM brands or models. Differences in device features, accuracy, scanning frequency, and user experience may influence outcomes, and results might differ with other CGM technologies. The dataset did not include patients’ scanning frequency because this metric was not consistently included in the EHRs.

Conclusions

This study found that CGM use was significantly associated with improved glycemic control in patients with non–insulin-dependent T2DM within an AI/AN population, particularly among patients with higher baseline HbA_1c levels. The findings suggest that CGMs may be a valuable tool for managing T2DM beyond insulin-dependent populations.

Additional research with larger sample sizes, control groups, and extended follow-up periods is recommended to explore long-term benefits and impacts on other health metrics. The sample size estimates derived from this study serve as a valuable resource for researchers designing future studies aimed at addressing these gaps. Future research that expands on our findings by including larger, more diverse cohorts, accounting for medication use, and exploring different CGM technologies will enhance understanding and contribute to more effective diabetes management strategies for varied populations.

Diabetes mellitus (DM) is a national health crisis affecting > 38 million people (11.6%) in the United States.¹ American Indian and Alaska Native (AI/AN) adults are disproportionately affected, with a prevalence of 14.5%—the highest among all racial and ethnic groups.¹ Type 2 DM (T2DM) accounts for 90% to 95% of all DM cases and is a leading cause of morbidity and mortality due to its association with cardiovascular disease, kidney failure, and other complications.²

Maintaining glycemic control is important for managing T2DM and preventing microvascular and macrovascular complications.³ The cornerstone of diabetes self-management has been patient self-monitored blood glucose (SMBG) using finger-stick glucometers.⁴ However, SMBG provides measurements from a single point in time and requires frequent, painful, and inconvenient finger pricks, leading to decreased adherence.^5,6 These limitations negatively affect patient engagement and overall glycemic control.⁷

Continuous glucose monitors (CGMs) offer real-time, continuous glucose readings and trends.⁸ CGMs improve glycemic control and reduce hypoglycemic episodes in patients who are insulin-dependent.^9,10 Flash glucose monitors, a type of CGM that requires scanning to obtain glucose readings, provide similar benefits.¹¹ Despite these demonstrated advantages, research has primarily focused on insulin-dependent populations, leaving a significant gap in understanding the effect of CGMs on patients with T2DM who are not insulin-dependent.¹²

Given the high prevalence of T2DM among AI/AN populations and the potential benefits of CGMs, this study sought to evaluate the effect of CGM use on glycemic control and other health metrics in patients with non–insulin-dependent T2DM in an AI/AN population. This focus addresses a critical knowledge gap and may inform clinical practices and policies to improve diabetes management in this high-risk group.

Methods

A retrospective observational study was conducted using deidentified electronic health records (EHRs) from 2019 to 2024 at a federally operated outpatient Indian Health Service (IHS) clinic serving an AI/AN population in the IHS Portland Area (Oregon, Washington, Idaho). The study protocol was reviewed and deemed exempt by institutional review boards at Washington State University and the Portland Area IHS.

Study Population

This study included patients diagnosed with non–insulin-dependent T2DM, had used a CGM for ≥ 1 year, and had hemoglobin A_1c (HbA_1c) measurements within 4 months prior to CGM initiation (baseline) and within ± 4 months after 1 year of CGM use. For other health metrics, including blood pressure (BP), weight, low-density lipoprotein cholesterol (LDL-C), and estimated glomerular filtration rate (eGFR), this study required measurements within 6 months before CGM initiation and within 6 months after 1 year of CGM use. The baseline HbA_1c in the dataset ranged from 5.3% to > 14%.

Patients were excluded if they used insulin during the study period, had incomplete laboratory or clinical data for the required time frame, or had < 1 year of CGM use. The dataset did not include detailed information on oral DM medications; thus, we could not report or account for the type or number of oral hypoglycemic agents used by the patients. The IHS clinical applications coordinator compiled the dataset from the EHR, identifying patients who were prescribed and received a CGM at the clinic. All patients used the Abbott Freestyle Libre CGM, the only formulary CGM available at the clinic during the study period.

A 1-year follow-up endpoint was selected for several reasons: (1) to capture potential seasonal variations in diet and activity; (2) to align with the clinic’s standard practice of annual comprehensive diabetes evaluations; and (3) to allow sufficient time for patients to adapt to CGM use and reflect any meaningful changes in glycemic control.

All patients received standard DM care according to clinic protocols, which included DM self-management education and training. Patients met with the diabetes educator at least once, during which the educator emphasized making informed decisions using CGM data, such as adjusting dietary choices and physical activity levels to manage blood glucose concentrations effectively.

A total of 302 patients were initially identified. After applying exclusion criteria, 132 were excluded due to insulin use, and 77 were excluded due to incomplete HbA_1c data within the specified time frames (Figure 1). The final sample included 93 patients.

Measures

The primary outcome was the change in HbA_1c levels from baseline to 1 year after CGM initiation. Secondary outcomes included changes in weight, systolic and diastolic BP, LDL-C concentrations, and eGFR. For the primary outcome, HbA_1c values were collected within a grace period of ± 4 months from the baseline and 1-year time points. The laboratory’s upper reporting limit for HbA_1c was 14%; values reported as “> 14%” were recorded as 14.1% for data analysis, although the actual values could have been higher.

For secondary outcomes, data were included if measurements were obtained within ± 6 months of the baseline and 1-year time points. Patients who did not have measurements within these time frames for specific metrics were excluded from secondary outcome analysis but remained in the overall study if they met the criteria for HbA_1c and CGM use.

Statistical Analysis

Statistical analysis was performed using R statistical software version 4.4.2. Paired t tests were conducted to compare baseline and 1-year follow- up measurements for variables with parametric distributions. Wilcoxon signed-rank test was used for nonparametric data. A linear regression analysis was conducted to examine the relationship between baseline HbA_1c levels and the change in HbA_1c after 1 year of CGM use. Differences were considered significant at P < .05 set a priori. To guide future research, a posthoc power analysis was performed using Cohen’s d to estimate the required sample sizes for detecting significant effects, assuming a similar population.

Results

The study included 93 patients, with a mean (SD) age of 55 (13) years (range, 29-83 years). Of the participants, 56 were female (60%) and 37 were male (40%). All participants were identified as AI/AN and had non–insulin-dependent T2DM.

Primary Outcomes

A significant reduction in HbA_1c levels was observed after 1 year of CGM use. The mean (SD) baseline HbA1c was 9.5% (2.4%), which decreased to 7.6% (2.2%) at 1-year follow-up (Table 1). This difference represents a mean change of -1.86% (2.4%) (95% CI, -2.35 to -1.37; P < .001 [paired t test, -7.53]).

A linear regression model evaluated the relationship between baseline HbA_1c (predictor) and the change in HbA_1c after 1 year (outcome). The change in HbA_1c was calculated as the difference between 1-year follow-up and baseline values. The regression model revealed a significant negative association between baseline HbA_1c and the change in HbA_1c (Β = -0.576; P < .001), indicating that higher baseline HbA_1c values were associated with greater reductions in HbA_1c over the year. The regression equation was: Change in HbA_1c = 3.587 – 0.576 × Baseline HbA_1c

The regression coefficient for baseline HbA_1c was -0.576 (standard error, 0.083; t = -6.931; P < .001), indicating that for each 1% increase in baseline HbA_1c, the reduction of HbA_1c after 1 year increased by approximately 0.576% (Figure 2). The model explained 34.6% of the variance in HbA_1c change (R² = .345; adjusted R² = .338).

Secondary Outcomes

Systolic BP decreased by a mean (SD) -4.9 (17) mm Hg; 95% CI, -8.6 to -1.11; P = .01, paired t test). However, no significant change was observed for diastolic BP (P = .77, paired t test). Similarly, no significant changes were observed in weight, LDL-C concentrations, or eGFR after 1 year of CGM use. A posthoc power analysis indicated that the study was underpowered to detect smaller effect sizes in secondary outcomes. For example, sample size estimates indicated that detecting significant changes in weight and LDL-C concentrations would require sample sizes of 152 and 220 patients, respectively (Table 2).

Discussion

This study found a clinically significant reduction in HbA_1c levels after 1 year among AI/AN patients with non–insulin-dependent T2DM who used CGMs. The mean HbA_1c decreased 1.9%, from 9.5% at baseline to 7.6% after 1 year. This reduction is not only statistically significant (P < .001), it is clinically meaningful—even a 1% decrease in HbA_1c is associated with substantial reductions in the risk of microvascular complications.³ The magnitude of the HbA_1c reduction observed suggests CGM use may be associated with improved glycemic control in this high-risk population. By achieving lower HbA_1c levels, patients may experience improved long-term health outcomes and a reduced burden of DM-related complications.

Changes in oral DM medications during the study period may have contributed to the observed improvements in HbA_1c levels. While the dataset lacked detailed information on types or dosages of oral hypoglycemic agents used, adjustments in medication regimens are common in DM management and could significantly affect glycemic control. The inability to account for these changes results in an inability to attribute the improvements in HbA_1c solely to CGM use. Future studies should collect comprehensive medication data to better isolate the effects of CGM use from other treatment modifications.

Another factor that may have contributed to the improved glycemic control is the DM self-management education and training patients received as part of standard care. Patients met with diabetes educators at least once and learned how to use the CGM device and interpret the data for self-management decisions. This education may have enhanced patient engagement and empowerment, enabling them to make informed choices about diet, physical activity, and medication adherence. Studies have shown that DM self-management education can significantly improve glycemic control and patient outcomes.¹³ By combining the CGM technology with targeted education, patients may have been better equipped to manage their condition, contributing to the observed reduction in HbA_1c levels. Future studies should consider synergistic effects of CGM use and DM education when evaluating interventions for glycemic control.

The significant reduction in HbA_1c indicates CGM use is associated with improved glycemic control in non–insulin-dependent T2DM. The linear regression analysis suggests patients with poorer glycemic control at baseline experienced greater reductions in HbA_1c over the course of 1 year. This finding aligns with previous studies that have shown greater HbA_1c reductions in patients with higher initial levels when using CGMs. Yaron et al reported similar findings: higher baseline HbA_1c levels predicted more substantial improvements with CGM use in patients with T2DM on insulin therapy.¹⁴

This study contributes to existing research by examining the association between CGM use and glycemic control in patients with non– insulin-dependent T2DM within an AI/AN population, a group that has been underreported in previous studies. Most prior research has focused on insulin-dependent patients or populations with different ethnic backgrounds.12 By focusing on patients with non–insulin-dependent T2DM, this study highlights the broader applicability of CGMs beyond traditional use, showcasing their potential association with benefits in earlier stages of DM management. Targeting the AI/AN population addresses a critical knowledge gap, given the disproportionately high prevalence of T2DM and associated complications in this group. The findings of this study suggest integrating CGM technology into the standard care of AI/AN patients with non–insulin-dependent T2DM may be associated with improved glycemic control and may help reduce health disparities.

The modest decrease in systolic BP observed in this study may indicate potential cardiovascular benefits associated with CGM use, possibly due to improved glycemic control and increased patient engagement in self-management. However, given the limited sample size and exclusion criteria, the study lacked sufficient power to detect significant associations between CGM use and other secondary outcomes such as BP, weight, LDL-C, and eGFR. Therefore, the significant finding with systolic BP should be interpreted with caution.

The lack of significant changes in secondary outcomes may be attributed to the study’s limited sample size and the relatively short duration for observing changes in these parameters. Larger studies are needed to assess the full impact of CGM on these variables. The required sample sizes for achieving adequate power in future studies were calculated, highlighting the utility of our study as a pilot, providing critical data for the design of larger, adequately powered studies.

Limitations

The retrospective design of this study limits causal inferences. Moreover, potential confounding variables were not controlled, such as changes in medication regimens (other than insulin use), dietary counseling, or physical activity. Additionally, we could not account for the type or number of oral DM medications prescribed to patients. The dataset included only information on insulin use, without detailed records of other antidiabetic medications. This limitation may have influenced the observed change in glycemic control, as variations in medication regimens could affect HbA_1c levels.

Because this study lacked a comparator group, the effect of CGM use cannot be definitively isolated from other factors (eg, medication changes, dietary modifications, or physical activity). Moreover, CGM devices can be costly and are not universally covered by all insurance or IHS programs, potentially limiting widespread implementation. Policy-level restrictions and patient-specific barriers may also hinder feasibility in other settings.

The small sample size may limit the generalizability of the findings. Of the initial 302 patients, about 69% were excluded due to insulin use or incomplete laboratory data. A ± 4-month window was selected to balance data quality with real-world practices. Extending this window further (eg, ± 6 months) might have included more participants but risked diluting the 1-year endpoint consistency. The lack of statistical significance in secondary metrics may be due to insufficient power rather than the absence of an effect.

Exclusion of patients due to incomplete data may have introduced selection bias. However, patients were included in the overall analysis if they met the criteria for HbA_1c and CGM use, even if they lacked data for secondary outcomes. Additionally, the laboratory’s upper reporting limit for HbA_1c was 14%, with values above this reported as “> 14%.” For analysis, these were recorded as 14.1%, which may underestimate the true baseline HbA_1c levels and impact of the assessment of change. This occurred for 4 of the 93 patients included.

All patients used the Freestyle Libre CGM, which may limit the generalizability of the findings to other CGM brands or models. Differences in device features, accuracy, scanning frequency, and user experience may influence outcomes, and results might differ with other CGM technologies. The dataset did not include patients’ scanning frequency because this metric was not consistently included in the EHRs.

Conclusions

This study found that CGM use was significantly associated with improved glycemic control in patients with non–insulin-dependent T2DM within an AI/AN population, particularly among patients with higher baseline HbA_1c levels. The findings suggest that CGMs may be a valuable tool for managing T2DM beyond insulin-dependent populations.

Additional research with larger sample sizes, control groups, and extended follow-up periods is recommended to explore long-term benefits and impacts on other health metrics. The sample size estimates derived from this study serve as a valuable resource for researchers designing future studies aimed at addressing these gaps. Future research that expands on our findings by including larger, more diverse cohorts, accounting for medication use, and exploring different CGM technologies will enhance understanding and contribute to more effective diabetes management strategies for varied populations.

Diabetes mellitus (DM) is a national health crisis affecting > 38 million people (11.6%) in the United States.¹ American Indian and Alaska Native (AI/AN) adults are disproportionately affected, with a prevalence of 14.5%—the highest among all racial and ethnic groups.¹ Type 2 DM (T2DM) accounts for 90% to 95% of all DM cases and is a leading cause of morbidity and mortality due to its association with cardiovascular disease, kidney failure, and other complications.²

Maintaining glycemic control is important for managing T2DM and preventing microvascular and macrovascular complications.³ The cornerstone of diabetes self-management has been patient self-monitored blood glucose (SMBG) using finger-stick glucometers.⁴ However, SMBG provides measurements from a single point in time and requires frequent, painful, and inconvenient finger pricks, leading to decreased adherence.^5,6 These limitations negatively affect patient engagement and overall glycemic control.⁷

Continuous glucose monitors (CGMs) offer real-time, continuous glucose readings and trends.⁸ CGMs improve glycemic control and reduce hypoglycemic episodes in patients who are insulin-dependent.^9,10 Flash glucose monitors, a type of CGM that requires scanning to obtain glucose readings, provide similar benefits.¹¹ Despite these demonstrated advantages, research has primarily focused on insulin-dependent populations, leaving a significant gap in understanding the effect of CGMs on patients with T2DM who are not insulin-dependent.¹²

Given the high prevalence of T2DM among AI/AN populations and the potential benefits of CGMs, this study sought to evaluate the effect of CGM use on glycemic control and other health metrics in patients with non–insulin-dependent T2DM in an AI/AN population. This focus addresses a critical knowledge gap and may inform clinical practices and policies to improve diabetes management in this high-risk group.

Methods

A retrospective observational study was conducted using deidentified electronic health records (EHRs) from 2019 to 2024 at a federally operated outpatient Indian Health Service (IHS) clinic serving an AI/AN population in the IHS Portland Area (Oregon, Washington, Idaho). The study protocol was reviewed and deemed exempt by institutional review boards at Washington State University and the Portland Area IHS.

Study Population

This study included patients diagnosed with non–insulin-dependent T2DM, had used a CGM for ≥ 1 year, and had hemoglobin A_1c (HbA_1c) measurements within 4 months prior to CGM initiation (baseline) and within ± 4 months after 1 year of CGM use. For other health metrics, including blood pressure (BP), weight, low-density lipoprotein cholesterol (LDL-C), and estimated glomerular filtration rate (eGFR), this study required measurements within 6 months before CGM initiation and within 6 months after 1 year of CGM use. The baseline HbA_1c in the dataset ranged from 5.3% to > 14%.

Patients were excluded if they used insulin during the study period, had incomplete laboratory or clinical data for the required time frame, or had < 1 year of CGM use. The dataset did not include detailed information on oral DM medications; thus, we could not report or account for the type or number of oral hypoglycemic agents used by the patients. The IHS clinical applications coordinator compiled the dataset from the EHR, identifying patients who were prescribed and received a CGM at the clinic. All patients used the Abbott Freestyle Libre CGM, the only formulary CGM available at the clinic during the study period.

A 1-year follow-up endpoint was selected for several reasons: (1) to capture potential seasonal variations in diet and activity; (2) to align with the clinic’s standard practice of annual comprehensive diabetes evaluations; and (3) to allow sufficient time for patients to adapt to CGM use and reflect any meaningful changes in glycemic control.

All patients received standard DM care according to clinic protocols, which included DM self-management education and training. Patients met with the diabetes educator at least once, during which the educator emphasized making informed decisions using CGM data, such as adjusting dietary choices and physical activity levels to manage blood glucose concentrations effectively.

A total of 302 patients were initially identified. After applying exclusion criteria, 132 were excluded due to insulin use, and 77 were excluded due to incomplete HbA_1c data within the specified time frames (Figure 1). The final sample included 93 patients.

Measures

The primary outcome was the change in HbA_1c levels from baseline to 1 year after CGM initiation. Secondary outcomes included changes in weight, systolic and diastolic BP, LDL-C concentrations, and eGFR. For the primary outcome, HbA_1c values were collected within a grace period of ± 4 months from the baseline and 1-year time points. The laboratory’s upper reporting limit for HbA_1c was 14%; values reported as “> 14%” were recorded as 14.1% for data analysis, although the actual values could have been higher.

For secondary outcomes, data were included if measurements were obtained within ± 6 months of the baseline and 1-year time points. Patients who did not have measurements within these time frames for specific metrics were excluded from secondary outcome analysis but remained in the overall study if they met the criteria for HbA_1c and CGM use.

Statistical Analysis

Statistical analysis was performed using R statistical software version 4.4.2. Paired t tests were conducted to compare baseline and 1-year follow- up measurements for variables with parametric distributions. Wilcoxon signed-rank test was used for nonparametric data. A linear regression analysis was conducted to examine the relationship between baseline HbA_1c levels and the change in HbA_1c after 1 year of CGM use. Differences were considered significant at P < .05 set a priori. To guide future research, a posthoc power analysis was performed using Cohen’s d to estimate the required sample sizes for detecting significant effects, assuming a similar population.

Results

The study included 93 patients, with a mean (SD) age of 55 (13) years (range, 29-83 years). Of the participants, 56 were female (60%) and 37 were male (40%). All participants were identified as AI/AN and had non–insulin-dependent T2DM.

Primary Outcomes

A significant reduction in HbA_1c levels was observed after 1 year of CGM use. The mean (SD) baseline HbA1c was 9.5% (2.4%), which decreased to 7.6% (2.2%) at 1-year follow-up (Table 1). This difference represents a mean change of -1.86% (2.4%) (95% CI, -2.35 to -1.37; P < .001 [paired t test, -7.53]).

A linear regression model evaluated the relationship between baseline HbA_1c (predictor) and the change in HbA_1c after 1 year (outcome). The change in HbA_1c was calculated as the difference between 1-year follow-up and baseline values. The regression model revealed a significant negative association between baseline HbA_1c and the change in HbA_1c (Β = -0.576; P < .001), indicating that higher baseline HbA_1c values were associated with greater reductions in HbA_1c over the year. The regression equation was: Change in HbA_1c = 3.587 – 0.576 × Baseline HbA_1c

The regression coefficient for baseline HbA_1c was -0.576 (standard error, 0.083; t = -6.931; P < .001), indicating that for each 1% increase in baseline HbA_1c, the reduction of HbA_1c after 1 year increased by approximately 0.576% (Figure 2). The model explained 34.6% of the variance in HbA_1c change (R² = .345; adjusted R² = .338).

Secondary Outcomes

Systolic BP decreased by a mean (SD) -4.9 (17) mm Hg; 95% CI, -8.6 to -1.11; P = .01, paired t test). However, no significant change was observed for diastolic BP (P = .77, paired t test). Similarly, no significant changes were observed in weight, LDL-C concentrations, or eGFR after 1 year of CGM use. A posthoc power analysis indicated that the study was underpowered to detect smaller effect sizes in secondary outcomes. For example, sample size estimates indicated that detecting significant changes in weight and LDL-C concentrations would require sample sizes of 152 and 220 patients, respectively (Table 2).

Discussion

This study found a clinically significant reduction in HbA_1c levels after 1 year among AI/AN patients with non–insulin-dependent T2DM who used CGMs. The mean HbA_1c decreased 1.9%, from 9.5% at baseline to 7.6% after 1 year. This reduction is not only statistically significant (P < .001), it is clinically meaningful—even a 1% decrease in HbA_1c is associated with substantial reductions in the risk of microvascular complications.³ The magnitude of the HbA_1c reduction observed suggests CGM use may be associated with improved glycemic control in this high-risk population. By achieving lower HbA_1c levels, patients may experience improved long-term health outcomes and a reduced burden of DM-related complications.

Changes in oral DM medications during the study period may have contributed to the observed improvements in HbA_1c levels. While the dataset lacked detailed information on types or dosages of oral hypoglycemic agents used, adjustments in medication regimens are common in DM management and could significantly affect glycemic control. The inability to account for these changes results in an inability to attribute the improvements in HbA_1c solely to CGM use. Future studies should collect comprehensive medication data to better isolate the effects of CGM use from other treatment modifications.

Another factor that may have contributed to the improved glycemic control is the DM self-management education and training patients received as part of standard care. Patients met with diabetes educators at least once and learned how to use the CGM device and interpret the data for self-management decisions. This education may have enhanced patient engagement and empowerment, enabling them to make informed choices about diet, physical activity, and medication adherence. Studies have shown that DM self-management education can significantly improve glycemic control and patient outcomes.¹³ By combining the CGM technology with targeted education, patients may have been better equipped to manage their condition, contributing to the observed reduction in HbA_1c levels. Future studies should consider synergistic effects of CGM use and DM education when evaluating interventions for glycemic control.

The significant reduction in HbA_1c indicates CGM use is associated with improved glycemic control in non–insulin-dependent T2DM. The linear regression analysis suggests patients with poorer glycemic control at baseline experienced greater reductions in HbA_1c over the course of 1 year. This finding aligns with previous studies that have shown greater HbA_1c reductions in patients with higher initial levels when using CGMs. Yaron et al reported similar findings: higher baseline HbA_1c levels predicted more substantial improvements with CGM use in patients with T2DM on insulin therapy.¹⁴

This study contributes to existing research by examining the association between CGM use and glycemic control in patients with non– insulin-dependent T2DM within an AI/AN population, a group that has been underreported in previous studies. Most prior research has focused on insulin-dependent patients or populations with different ethnic backgrounds.12 By focusing on patients with non–insulin-dependent T2DM, this study highlights the broader applicability of CGMs beyond traditional use, showcasing their potential association with benefits in earlier stages of DM management. Targeting the AI/AN population addresses a critical knowledge gap, given the disproportionately high prevalence of T2DM and associated complications in this group. The findings of this study suggest integrating CGM technology into the standard care of AI/AN patients with non–insulin-dependent T2DM may be associated with improved glycemic control and may help reduce health disparities.

The modest decrease in systolic BP observed in this study may indicate potential cardiovascular benefits associated with CGM use, possibly due to improved glycemic control and increased patient engagement in self-management. However, given the limited sample size and exclusion criteria, the study lacked sufficient power to detect significant associations between CGM use and other secondary outcomes such as BP, weight, LDL-C, and eGFR. Therefore, the significant finding with systolic BP should be interpreted with caution.

The lack of significant changes in secondary outcomes may be attributed to the study’s limited sample size and the relatively short duration for observing changes in these parameters. Larger studies are needed to assess the full impact of CGM on these variables. The required sample sizes for achieving adequate power in future studies were calculated, highlighting the utility of our study as a pilot, providing critical data for the design of larger, adequately powered studies.

Limitations

The retrospective design of this study limits causal inferences. Moreover, potential confounding variables were not controlled, such as changes in medication regimens (other than insulin use), dietary counseling, or physical activity. Additionally, we could not account for the type or number of oral DM medications prescribed to patients. The dataset included only information on insulin use, without detailed records of other antidiabetic medications. This limitation may have influenced the observed change in glycemic control, as variations in medication regimens could affect HbA_1c levels.

Because this study lacked a comparator group, the effect of CGM use cannot be definitively isolated from other factors (eg, medication changes, dietary modifications, or physical activity). Moreover, CGM devices can be costly and are not universally covered by all insurance or IHS programs, potentially limiting widespread implementation. Policy-level restrictions and patient-specific barriers may also hinder feasibility in other settings.

The small sample size may limit the generalizability of the findings. Of the initial 302 patients, about 69% were excluded due to insulin use or incomplete laboratory data. A ± 4-month window was selected to balance data quality with real-world practices. Extending this window further (eg, ± 6 months) might have included more participants but risked diluting the 1-year endpoint consistency. The lack of statistical significance in secondary metrics may be due to insufficient power rather than the absence of an effect.

Exclusion of patients due to incomplete data may have introduced selection bias. However, patients were included in the overall analysis if they met the criteria for HbA_1c and CGM use, even if they lacked data for secondary outcomes. Additionally, the laboratory’s upper reporting limit for HbA_1c was 14%, with values above this reported as “> 14%.” For analysis, these were recorded as 14.1%, which may underestimate the true baseline HbA_1c levels and impact of the assessment of change. This occurred for 4 of the 93 patients included.

All patients used the Freestyle Libre CGM, which may limit the generalizability of the findings to other CGM brands or models. Differences in device features, accuracy, scanning frequency, and user experience may influence outcomes, and results might differ with other CGM technologies. The dataset did not include patients’ scanning frequency because this metric was not consistently included in the EHRs.

Conclusions

This study found that CGM use was significantly associated with improved glycemic control in patients with non–insulin-dependent T2DM within an AI/AN population, particularly among patients with higher baseline HbA_1c levels. The findings suggest that CGMs may be a valuable tool for managing T2DM beyond insulin-dependent populations.

Additional research with larger sample sizes, control groups, and extended follow-up periods is recommended to explore long-term benefits and impacts on other health metrics. The sample size estimates derived from this study serve as a valuable resource for researchers designing future studies aimed at addressing these gaps. Future research that expands on our findings by including larger, more diverse cohorts, accounting for medication use, and exploring different CGM technologies will enhance understanding and contribute to more effective diabetes management strategies for varied populations.

Cardiovascular disease (CVD) is the leading cause of death among women in the United States.¹ Most CVD is due to the buildup of plaque (ie, cholesterol, proteins, calcium, and inflammatory cells) in artery walls.² The plaque may lead to atherosclerotic cardiovascular disease (ASCVD), which includes coronary heart disease, cerebrovascular disease, peripheral artery disease, and aortic atherosclerotic disease.^2,3 Control and reduction of ASCVD risk factors, including high cholesterol levels, elevated blood pressure, insulin resistance, smoking, and a sedentary lifestyle, can contribute to a reduction in ASCVD morbidity and mortality.² People with type 2 diabetes mellitus (T2DM) have an increased prevalence of lipid abnormalities, contributing to their high risk of ASCVD.^4,5

The prescribing of statins (3-hydroxy-3-methyl-glutaryl-coenzmye A reductase inhibitors) is the cornerstone of lipid-lowering therapy and cardiovascular risk reduction for primary and secondary prevention of ASCVD.⁶ The American Diabetes Association (ADA) and American College of Cardiology/American Heart Association (ACC/AHA) recommend moderate- to high-intensity statins for primary prevention in patients with T2DM and high-intensity statins for secondary prevention in those with or without diabetes when not contraindicated.^4,5,7 Despite eligibility according to guideline recommendations, research predominantly shows that women are less likely to receive statin therapy; however, this trend is improving. ^[6,8-11] To explain the sex differences in statin use, Nanna et al found that there is a combination of women being offered statin therapy less frequently, declining therapy more frequently, and discontinuing treatment more frequently.¹¹ One possibility for discontinuing treatment could be statin-associated muscle symptoms (SAMS), which occur in about 10% of patients.¹² The incidence of adverse effects (AEs) may be related to the way statins are metabolized.

Pharmacogenomic testing is free for veterans through the US Department of Veterans Affairs (VA) PHASER program, which offers information and recommendations for a panel of 11 gene variants. The panel includes genes related to common medication classes such as anticoagulants, antiplatelets, proton pump inhibitors, nonsteroidal anti-inflammatory drugs, opioids, antidepressants, and statins. The VA PHASER panel includes the solute carrier organic anion transporter family member 1B1 (SLCO1B1) gene, which is predominantly expressed in the liver and facilitates the hepatic uptake of most statins.^13,14 A reduced function of SLCO1B1 can lead to higher statin levels, resulting in increased concentrations that may potentially cause SAMS.^13,14 Some alleles associated with reduced function include SLCO1B1*5, *15, *23, *31, and *46 to *49, whereas others are associated with increased function, such as SLCO1B1 *14 and *20 (Appendix).¹⁵ Supporting evidence shows the SLCO1B1*5 nucleotide polymorphism increases plasma levels of simvastatin and atorvastatin, affecting effectiveness or toxicity. ¹³ Females tend to have a lower body weight and higher percentage of body fat compared with males, which might lead to higher concentrations of lipophilic drugs, including atorvastatin and simvastatin, which may be exacerbated by decreased function of SLCO1B1*5.¹⁵ With pharmacogenomic testing, therapeutic recommendations can be made to improve the overall safety and efficacy of statins, thus improving adherence using a patient-specific approach.^14,15

Methods

Carl Vinson VA Medical Center (CVVAMC) serves about 42,000 veterans in Central and South Georgia, of which about 15% are female. Of the female veterans enrolled in care, 63% identify as Black, 27% White, and 1.5% as Asian, American Indian/Alaska Native, or Native Hawaiian/Other Pacific Islander. The 2020 Veterans Chartbook report showed that female veterans and minority racial and ethnic groups had worse access to health care and higher mortality rates than their male and non-Hispanic White counterparts.¹⁶

The Primary Care Equity Dashboard (PCED) was developed to engage the VA health care workforce in the process of identifying and addressing inequities in local patient populations.¹⁷ Using electronic quality measure data, the PCED provides Veterans Integrated Service Network-level and facility-level performance on several metrics.¹⁸ The PCED had not been previously used at the CVVAMC, and few publications or quality improvement projects regarding its use have been reported by the VA Office of Health Equity. PCED helped identify disparities when comparing female to male patients in the prescribing of statin therapy for patients with CVD and statin therapy for patients with T2DM.

VA PHASER pharmacogenomic analyses provided an opportunity to expand this quality improvement project. Sanford Health and the VA collaborated on the PHASER program to offer free genetic testing for veterans. The program launched in 2019 and expanded to various VA sites, including CVVAMC in March 2023. This program has been extended to December 31, 2025.

The primary objective of this quality improvement project was to increase statin prescribing among female veterans with T2DM and/or CVD to reduce cardiovascular risk. Secondary outcomes included increased pharmacogenomic testing and the assessment of pharmacogenomic results related to statin therapy. This project was approved by the CVVAMC Pharmacy and Therapeutics Committee. The PCED was used to identify female veterans with T2DM and/or CVD without an active prescription for a statin between July and October 2023. A review of Computerized Patient Record System patient charts was completed to screen for prespecified inclusion and exclusion criteria. Veterans were included if they were assigned female at birth, were enrolled in care at CVVAMC, and had a diagnosis of T2DM or CVD (history of myocardial infarction, coronary bypass graft, percutaneous coronary intervention, or other revascularization in any setting).

Veterans were excluded if they were currently pregnant, trying to conceive, breastfeeding, had a T1DM diagnosis, had previously documented hypersensitivity to a statin, active liver failure or decompensated cirrhosis, previously documented statin-associated rhabdomyolysis or autoimmune myopathy, an active prescription for a proprotein convertase subtilisin/kexin type 9 inhibitor, or previously documented statin intolerance (defined as the inability to tolerate ≥ 3 statins, with ≥ 1 prescribed at low intensity or alternate-day dosing). The female veterans were compared to 2 comparators: the facility's male veterans and the VA national average, identified via the PCED.

Once a veteran was screened, they were telephoned between October 2023 and February 2024 and provided education on statin use and pharmacogenomic testing using a standardized note template. An order was placed for participants who provided verbal consent for pharmacogenomic testing. Those who agreed to statin initiation were referred to a clinical pharmacist practitioner (CPP) who contacted them at a later date to prescribe a statin following the recommendations of the 2019 ACC/AHA and 2023 ADA guidelines and pharmacogenomic testing, if applicable.^4,5,7 Appropriate monitoring and follow-up occurred at the discretion of each CPP. Data collection included: age, race, diagnoses (T2DM, CVD, or both), baseline lipid panel (total cholesterol, triglycerides, high-density lipoprotein, low-density lipoprotein), hepatic function, name and dose of statin, reasons for declining statin therapy, and pharmacogenomic testing results related to SLCO1B1.

Results

At baseline in July 2023, 77.8% of female veterans with T2DM were prescribed a statin, which exceeded the national VA average (77.0%), but was below the rate for male veterans (78.7%) in the facility comparator group.17 Additionally, 82.2% of females with CVD were prescribed a statin, which was below the national VA average of 86.0% and the 84.9% of male veterans in the facility comparator group.17 The PCED identified 189 female veterans from July 2023 to October 2023 who may benefit from statin therapy. Thirty-three females met the exclusion criteria. Of the 156 included veterans, 129 (82.7%) were successfully contacted and 27 (17.3%) could not be reached by telephone after 3 attempts (Figure 1). The 129 female veterans contacted had a mean age of 59 years and the majority were Black (82.9%) (Table 1).

Primary Outcomes

Of the 129 contacted veterans, 31 (24.0%) had a non-VA statin prescription, 13 (10.1%) had an active VA statin prescription, and 85 (65.9%) did not have a statin prescription, despite being eligible. Statin adherence was confirmed with participants, and the medication list was updated accordingly.

Of the 85 veterans with no active statin therapy, 37 (43.5%) accepted a new statin prescription and 48 (56.5%) declined. There were various reasons provided for declining statin therapy: 17 participants (35.4%) declined due to concern for AEs (Table 2).

From July 2023 to March 2024, the percentage of female veterans with active statin therapy with T2DM increased from 77.8% to 79.0%. For those with active statin therapy with CVD, usage increased from 82.2% to 90.2%, which exceeded the national VA average and facility male comparator group (Figures 2 and 3).¹⁷

Secondary Outcomes

Seventy-one of 129 veterans (55.0%) gave verbal consent, and 47 (66.2%) completed the pharmacogenomic testing; 58 (45.0%) declined. Five veterans (10.6%) had a known SLCO1B1 allele variant present. One veteran required a change in statin therapy based on the results (eAppendix).

Discussion

This project aimed to increase statin prescribing among female veterans with T2DM and/or CVD to reduce cardiovascular risk and increase pharmacogenomic testing using the PCED and care managed by CPPs. The results of this quality improvement project illustrated that both metrics have improved at CVVAMC as a result of the intervention. The results in both metrics now exceed the PCED national VA average, and the CVD metric also exceeds that of the facility male comparator group. While there was only a 1.2% increase from July 2023 to March 2024 for patients with T2DM, there was an 8.0% increase for patients with CVD. Despite standardized education on statin use, more veterans declined therapy than accepted it, mostly due to concern for AEs. Recording the reasons for declining statin therapy offered valuable insight that can be used in additional discussions with veterans and clinicians.

Pharmacogenomics gives clinicians the unique opportunity to take a proactive approach to better predict drug responses, potentially allowing for less trial and error with medications, fewer AEs, greater trust in the clinician, and improved medication adherence. The CPPs incorporated pharmacogenomic testing into their practice, which led to identifying 5 SLCO1B1 gene abnormalities. The PCED served as a powerful tool for advancing equity-focused quality improvement initiatives on a local level and was crucial in prioritizing the detection of veterans potentially receiving suboptimal care.

Limitations

The nature of “cold calls” made it challenging to establish contact for inclusion in this study. An alternative to increase engagement could have been scheduled phone or face-to-face visits. While the use of the PCED was crucial, data did not account for statins listed in the non-VA medication list. All 31 patients with statins prescribed outside the VA had a start date added to provide the most accurate representation of the data moving forward.

Another limitation in this project was its small sample size and population. CVVAMC serves about 6200 female veterans, with roughly 63% identifying as Black. The preponderance of Black individuals (83%) in this project is typical for the female patient population at CVVAMC but may not reflect the demographics of other populations. Other limitations to this project consisted of scheduling conflicts. Appointments for laboratory draws at community-based outpatient clinics were subject to availability, which resulted in some delay in completion of pharmacogenomic testing.

Conclusions

CPPs can help reduce inequity in health care delivery. Increased incorporation of the PCED into regular practice within the VA is recommended to continue addressing sex disparities in statin use, diabetes control, blood pressure management, cancer screenings, and vaccination needs. CVVAMC plans to expand its use through another quality improvement project focused on reducing sex disparities in blood pressure management. Improving educational resources made available to veterans on the importance of statin therapy and potential to mitigate AEs through use of the VA PHASER program also would be helpful. This project successfully improved CVVAMC metrics for female veterans appropriately prescribed statin therapy and increased access to pharmacogenomic testing. Most importantly, it helped close the sex-based gap in CVD risk reduction care.

Cardiovascular disease (CVD) is the leading cause of death among women in the United States.¹ Most CVD is due to the buildup of plaque (ie, cholesterol, proteins, calcium, and inflammatory cells) in artery walls.² The plaque may lead to atherosclerotic cardiovascular disease (ASCVD), which includes coronary heart disease, cerebrovascular disease, peripheral artery disease, and aortic atherosclerotic disease.^2,3 Control and reduction of ASCVD risk factors, including high cholesterol levels, elevated blood pressure, insulin resistance, smoking, and a sedentary lifestyle, can contribute to a reduction in ASCVD morbidity and mortality.² People with type 2 diabetes mellitus (T2DM) have an increased prevalence of lipid abnormalities, contributing to their high risk of ASCVD.^4,5

The prescribing of statins (3-hydroxy-3-methyl-glutaryl-coenzmye A reductase inhibitors) is the cornerstone of lipid-lowering therapy and cardiovascular risk reduction for primary and secondary prevention of ASCVD.⁶ The American Diabetes Association (ADA) and American College of Cardiology/American Heart Association (ACC/AHA) recommend moderate- to high-intensity statins for primary prevention in patients with T2DM and high-intensity statins for secondary prevention in those with or without diabetes when not contraindicated.^4,5,7 Despite eligibility according to guideline recommendations, research predominantly shows that women are less likely to receive statin therapy; however, this trend is improving. ^[6,8-11] To explain the sex differences in statin use, Nanna et al found that there is a combination of women being offered statin therapy less frequently, declining therapy more frequently, and discontinuing treatment more frequently.¹¹ One possibility for discontinuing treatment could be statin-associated muscle symptoms (SAMS), which occur in about 10% of patients.¹² The incidence of adverse effects (AEs) may be related to the way statins are metabolized.

Pharmacogenomic testing is free for veterans through the US Department of Veterans Affairs (VA) PHASER program, which offers information and recommendations for a panel of 11 gene variants. The panel includes genes related to common medication classes such as anticoagulants, antiplatelets, proton pump inhibitors, nonsteroidal anti-inflammatory drugs, opioids, antidepressants, and statins. The VA PHASER panel includes the solute carrier organic anion transporter family member 1B1 (SLCO1B1) gene, which is predominantly expressed in the liver and facilitates the hepatic uptake of most statins.^13,14 A reduced function of SLCO1B1 can lead to higher statin levels, resulting in increased concentrations that may potentially cause SAMS.^13,14 Some alleles associated with reduced function include SLCO1B1*5, *15, *23, *31, and *46 to *49, whereas others are associated with increased function, such as SLCO1B1 *14 and *20 (Appendix).¹⁵ Supporting evidence shows the SLCO1B1*5 nucleotide polymorphism increases plasma levels of simvastatin and atorvastatin, affecting effectiveness or toxicity. ¹³ Females tend to have a lower body weight and higher percentage of body fat compared with males, which might lead to higher concentrations of lipophilic drugs, including atorvastatin and simvastatin, which may be exacerbated by decreased function of SLCO1B1*5.¹⁵ With pharmacogenomic testing, therapeutic recommendations can be made to improve the overall safety and efficacy of statins, thus improving adherence using a patient-specific approach.^14,15

Methods

Carl Vinson VA Medical Center (CVVAMC) serves about 42,000 veterans in Central and South Georgia, of which about 15% are female. Of the female veterans enrolled in care, 63% identify as Black, 27% White, and 1.5% as Asian, American Indian/Alaska Native, or Native Hawaiian/Other Pacific Islander. The 2020 Veterans Chartbook report showed that female veterans and minority racial and ethnic groups had worse access to health care and higher mortality rates than their male and non-Hispanic White counterparts.¹⁶

The Primary Care Equity Dashboard (PCED) was developed to engage the VA health care workforce in the process of identifying and addressing inequities in local patient populations.¹⁷ Using electronic quality measure data, the PCED provides Veterans Integrated Service Network-level and facility-level performance on several metrics.¹⁸ The PCED had not been previously used at the CVVAMC, and few publications or quality improvement projects regarding its use have been reported by the VA Office of Health Equity. PCED helped identify disparities when comparing female to male patients in the prescribing of statin therapy for patients with CVD and statin therapy for patients with T2DM.

VA PHASER pharmacogenomic analyses provided an opportunity to expand this quality improvement project. Sanford Health and the VA collaborated on the PHASER program to offer free genetic testing for veterans. The program launched in 2019 and expanded to various VA sites, including CVVAMC in March 2023. This program has been extended to December 31, 2025.

The primary objective of this quality improvement project was to increase statin prescribing among female veterans with T2DM and/or CVD to reduce cardiovascular risk. Secondary outcomes included increased pharmacogenomic testing and the assessment of pharmacogenomic results related to statin therapy. This project was approved by the CVVAMC Pharmacy and Therapeutics Committee. The PCED was used to identify female veterans with T2DM and/or CVD without an active prescription for a statin between July and October 2023. A review of Computerized Patient Record System patient charts was completed to screen for prespecified inclusion and exclusion criteria. Veterans were included if they were assigned female at birth, were enrolled in care at CVVAMC, and had a diagnosis of T2DM or CVD (history of myocardial infarction, coronary bypass graft, percutaneous coronary intervention, or other revascularization in any setting).

Veterans were excluded if they were currently pregnant, trying to conceive, breastfeeding, had a T1DM diagnosis, had previously documented hypersensitivity to a statin, active liver failure or decompensated cirrhosis, previously documented statin-associated rhabdomyolysis or autoimmune myopathy, an active prescription for a proprotein convertase subtilisin/kexin type 9 inhibitor, or previously documented statin intolerance (defined as the inability to tolerate ≥ 3 statins, with ≥ 1 prescribed at low intensity or alternate-day dosing). The female veterans were compared to 2 comparators: the facility's male veterans and the VA national average, identified via the PCED.

Once a veteran was screened, they were telephoned between October 2023 and February 2024 and provided education on statin use and pharmacogenomic testing using a standardized note template. An order was placed for participants who provided verbal consent for pharmacogenomic testing. Those who agreed to statin initiation were referred to a clinical pharmacist practitioner (CPP) who contacted them at a later date to prescribe a statin following the recommendations of the 2019 ACC/AHA and 2023 ADA guidelines and pharmacogenomic testing, if applicable.^4,5,7 Appropriate monitoring and follow-up occurred at the discretion of each CPP. Data collection included: age, race, diagnoses (T2DM, CVD, or both), baseline lipid panel (total cholesterol, triglycerides, high-density lipoprotein, low-density lipoprotein), hepatic function, name and dose of statin, reasons for declining statin therapy, and pharmacogenomic testing results related to SLCO1B1.

Results

At baseline in July 2023, 77.8% of female veterans with T2DM were prescribed a statin, which exceeded the national VA average (77.0%), but was below the rate for male veterans (78.7%) in the facility comparator group.17 Additionally, 82.2% of females with CVD were prescribed a statin, which was below the national VA average of 86.0% and the 84.9% of male veterans in the facility comparator group.17 The PCED identified 189 female veterans from July 2023 to October 2023 who may benefit from statin therapy. Thirty-three females met the exclusion criteria. Of the 156 included veterans, 129 (82.7%) were successfully contacted and 27 (17.3%) could not be reached by telephone after 3 attempts (Figure 1). The 129 female veterans contacted had a mean age of 59 years and the majority were Black (82.9%) (Table 1).

Primary Outcomes

Of the 129 contacted veterans, 31 (24.0%) had a non-VA statin prescription, 13 (10.1%) had an active VA statin prescription, and 85 (65.9%) did not have a statin prescription, despite being eligible. Statin adherence was confirmed with participants, and the medication list was updated accordingly.

Of the 85 veterans with no active statin therapy, 37 (43.5%) accepted a new statin prescription and 48 (56.5%) declined. There were various reasons provided for declining statin therapy: 17 participants (35.4%) declined due to concern for AEs (Table 2).

From July 2023 to March 2024, the percentage of female veterans with active statin therapy with T2DM increased from 77.8% to 79.0%. For those with active statin therapy with CVD, usage increased from 82.2% to 90.2%, which exceeded the national VA average and facility male comparator group (Figures 2 and 3).¹⁷

Secondary Outcomes

Seventy-one of 129 veterans (55.0%) gave verbal consent, and 47 (66.2%) completed the pharmacogenomic testing; 58 (45.0%) declined. Five veterans (10.6%) had a known SLCO1B1 allele variant present. One veteran required a change in statin therapy based on the results (eAppendix).

Discussion

This project aimed to increase statin prescribing among female veterans with T2DM and/or CVD to reduce cardiovascular risk and increase pharmacogenomic testing using the PCED and care managed by CPPs. The results of this quality improvement project illustrated that both metrics have improved at CVVAMC as a result of the intervention. The results in both metrics now exceed the PCED national VA average, and the CVD metric also exceeds that of the facility male comparator group. While there was only a 1.2% increase from July 2023 to March 2024 for patients with T2DM, there was an 8.0% increase for patients with CVD. Despite standardized education on statin use, more veterans declined therapy than accepted it, mostly due to concern for AEs. Recording the reasons for declining statin therapy offered valuable insight that can be used in additional discussions with veterans and clinicians.

Pharmacogenomics gives clinicians the unique opportunity to take a proactive approach to better predict drug responses, potentially allowing for less trial and error with medications, fewer AEs, greater trust in the clinician, and improved medication adherence. The CPPs incorporated pharmacogenomic testing into their practice, which led to identifying 5 SLCO1B1 gene abnormalities. The PCED served as a powerful tool for advancing equity-focused quality improvement initiatives on a local level and was crucial in prioritizing the detection of veterans potentially receiving suboptimal care.

Limitations

The nature of “cold calls” made it challenging to establish contact for inclusion in this study. An alternative to increase engagement could have been scheduled phone or face-to-face visits. While the use of the PCED was crucial, data did not account for statins listed in the non-VA medication list. All 31 patients with statins prescribed outside the VA had a start date added to provide the most accurate representation of the data moving forward.

Another limitation in this project was its small sample size and population. CVVAMC serves about 6200 female veterans, with roughly 63% identifying as Black. The preponderance of Black individuals (83%) in this project is typical for the female patient population at CVVAMC but may not reflect the demographics of other populations. Other limitations to this project consisted of scheduling conflicts. Appointments for laboratory draws at community-based outpatient clinics were subject to availability, which resulted in some delay in completion of pharmacogenomic testing.

Conclusions

CPPs can help reduce inequity in health care delivery. Increased incorporation of the PCED into regular practice within the VA is recommended to continue addressing sex disparities in statin use, diabetes control, blood pressure management, cancer screenings, and vaccination needs. CVVAMC plans to expand its use through another quality improvement project focused on reducing sex disparities in blood pressure management. Improving educational resources made available to veterans on the importance of statin therapy and potential to mitigate AEs through use of the VA PHASER program also would be helpful. This project successfully improved CVVAMC metrics for female veterans appropriately prescribed statin therapy and increased access to pharmacogenomic testing. Most importantly, it helped close the sex-based gap in CVD risk reduction care.

Cardiovascular disease (CVD) is the leading cause of death among women in the United States.¹ Most CVD is due to the buildup of plaque (ie, cholesterol, proteins, calcium, and inflammatory cells) in artery walls.² The plaque may lead to atherosclerotic cardiovascular disease (ASCVD), which includes coronary heart disease, cerebrovascular disease, peripheral artery disease, and aortic atherosclerotic disease.^2,3 Control and reduction of ASCVD risk factors, including high cholesterol levels, elevated blood pressure, insulin resistance, smoking, and a sedentary lifestyle, can contribute to a reduction in ASCVD morbidity and mortality.² People with type 2 diabetes mellitus (T2DM) have an increased prevalence of lipid abnormalities, contributing to their high risk of ASCVD.^4,5

The prescribing of statins (3-hydroxy-3-methyl-glutaryl-coenzmye A reductase inhibitors) is the cornerstone of lipid-lowering therapy and cardiovascular risk reduction for primary and secondary prevention of ASCVD.⁶ The American Diabetes Association (ADA) and American College of Cardiology/American Heart Association (ACC/AHA) recommend moderate- to high-intensity statins for primary prevention in patients with T2DM and high-intensity statins for secondary prevention in those with or without diabetes when not contraindicated.^4,5,7 Despite eligibility according to guideline recommendations, research predominantly shows that women are less likely to receive statin therapy; however, this trend is improving. ^[6,8-11] To explain the sex differences in statin use, Nanna et al found that there is a combination of women being offered statin therapy less frequently, declining therapy more frequently, and discontinuing treatment more frequently.¹¹ One possibility for discontinuing treatment could be statin-associated muscle symptoms (SAMS), which occur in about 10% of patients.¹² The incidence of adverse effects (AEs) may be related to the way statins are metabolized.

Pharmacogenomic testing is free for veterans through the US Department of Veterans Affairs (VA) PHASER program, which offers information and recommendations for a panel of 11 gene variants. The panel includes genes related to common medication classes such as anticoagulants, antiplatelets, proton pump inhibitors, nonsteroidal anti-inflammatory drugs, opioids, antidepressants, and statins. The VA PHASER panel includes the solute carrier organic anion transporter family member 1B1 (SLCO1B1) gene, which is predominantly expressed in the liver and facilitates the hepatic uptake of most statins.^13,14 A reduced function of SLCO1B1 can lead to higher statin levels, resulting in increased concentrations that may potentially cause SAMS.^13,14 Some alleles associated with reduced function include SLCO1B1*5, *15, *23, *31, and *46 to *49, whereas others are associated with increased function, such as SLCO1B1 *14 and *20 (Appendix).¹⁵ Supporting evidence shows the SLCO1B1*5 nucleotide polymorphism increases plasma levels of simvastatin and atorvastatin, affecting effectiveness or toxicity. ¹³ Females tend to have a lower body weight and higher percentage of body fat compared with males, which might lead to higher concentrations of lipophilic drugs, including atorvastatin and simvastatin, which may be exacerbated by decreased function of SLCO1B1*5.¹⁵ With pharmacogenomic testing, therapeutic recommendations can be made to improve the overall safety and efficacy of statins, thus improving adherence using a patient-specific approach.^14,15

Methods

Carl Vinson VA Medical Center (CVVAMC) serves about 42,000 veterans in Central and South Georgia, of which about 15% are female. Of the female veterans enrolled in care, 63% identify as Black, 27% White, and 1.5% as Asian, American Indian/Alaska Native, or Native Hawaiian/Other Pacific Islander. The 2020 Veterans Chartbook report showed that female veterans and minority racial and ethnic groups had worse access to health care and higher mortality rates than their male and non-Hispanic White counterparts.¹⁶

The Primary Care Equity Dashboard (PCED) was developed to engage the VA health care workforce in the process of identifying and addressing inequities in local patient populations.¹⁷ Using electronic quality measure data, the PCED provides Veterans Integrated Service Network-level and facility-level performance on several metrics.¹⁸ The PCED had not been previously used at the CVVAMC, and few publications or quality improvement projects regarding its use have been reported by the VA Office of Health Equity. PCED helped identify disparities when comparing female to male patients in the prescribing of statin therapy for patients with CVD and statin therapy for patients with T2DM.

VA PHASER pharmacogenomic analyses provided an opportunity to expand this quality improvement project. Sanford Health and the VA collaborated on the PHASER program to offer free genetic testing for veterans. The program launched in 2019 and expanded to various VA sites, including CVVAMC in March 2023. This program has been extended to December 31, 2025.

The primary objective of this quality improvement project was to increase statin prescribing among female veterans with T2DM and/or CVD to reduce cardiovascular risk. Secondary outcomes included increased pharmacogenomic testing and the assessment of pharmacogenomic results related to statin therapy. This project was approved by the CVVAMC Pharmacy and Therapeutics Committee. The PCED was used to identify female veterans with T2DM and/or CVD without an active prescription for a statin between July and October 2023. A review of Computerized Patient Record System patient charts was completed to screen for prespecified inclusion and exclusion criteria. Veterans were included if they were assigned female at birth, were enrolled in care at CVVAMC, and had a diagnosis of T2DM or CVD (history of myocardial infarction, coronary bypass graft, percutaneous coronary intervention, or other revascularization in any setting).

Veterans were excluded if they were currently pregnant, trying to conceive, breastfeeding, had a T1DM diagnosis, had previously documented hypersensitivity to a statin, active liver failure or decompensated cirrhosis, previously documented statin-associated rhabdomyolysis or autoimmune myopathy, an active prescription for a proprotein convertase subtilisin/kexin type 9 inhibitor, or previously documented statin intolerance (defined as the inability to tolerate ≥ 3 statins, with ≥ 1 prescribed at low intensity or alternate-day dosing). The female veterans were compared to 2 comparators: the facility's male veterans and the VA national average, identified via the PCED.

Once a veteran was screened, they were telephoned between October 2023 and February 2024 and provided education on statin use and pharmacogenomic testing using a standardized note template. An order was placed for participants who provided verbal consent for pharmacogenomic testing. Those who agreed to statin initiation were referred to a clinical pharmacist practitioner (CPP) who contacted them at a later date to prescribe a statin following the recommendations of the 2019 ACC/AHA and 2023 ADA guidelines and pharmacogenomic testing, if applicable.^4,5,7 Appropriate monitoring and follow-up occurred at the discretion of each CPP. Data collection included: age, race, diagnoses (T2DM, CVD, or both), baseline lipid panel (total cholesterol, triglycerides, high-density lipoprotein, low-density lipoprotein), hepatic function, name and dose of statin, reasons for declining statin therapy, and pharmacogenomic testing results related to SLCO1B1.

Results

At baseline in July 2023, 77.8% of female veterans with T2DM were prescribed a statin, which exceeded the national VA average (77.0%), but was below the rate for male veterans (78.7%) in the facility comparator group.17 Additionally, 82.2% of females with CVD were prescribed a statin, which was below the national VA average of 86.0% and the 84.9% of male veterans in the facility comparator group.17 The PCED identified 189 female veterans from July 2023 to October 2023 who may benefit from statin therapy. Thirty-three females met the exclusion criteria. Of the 156 included veterans, 129 (82.7%) were successfully contacted and 27 (17.3%) could not be reached by telephone after 3 attempts (Figure 1). The 129 female veterans contacted had a mean age of 59 years and the majority were Black (82.9%) (Table 1).

Primary Outcomes

Of the 129 contacted veterans, 31 (24.0%) had a non-VA statin prescription, 13 (10.1%) had an active VA statin prescription, and 85 (65.9%) did not have a statin prescription, despite being eligible. Statin adherence was confirmed with participants, and the medication list was updated accordingly.

Of the 85 veterans with no active statin therapy, 37 (43.5%) accepted a new statin prescription and 48 (56.5%) declined. There were various reasons provided for declining statin therapy: 17 participants (35.4%) declined due to concern for AEs (Table 2).

From July 2023 to March 2024, the percentage of female veterans with active statin therapy with T2DM increased from 77.8% to 79.0%. For those with active statin therapy with CVD, usage increased from 82.2% to 90.2%, which exceeded the national VA average and facility male comparator group (Figures 2 and 3).¹⁷

Secondary Outcomes

Seventy-one of 129 veterans (55.0%) gave verbal consent, and 47 (66.2%) completed the pharmacogenomic testing; 58 (45.0%) declined. Five veterans (10.6%) had a known SLCO1B1 allele variant present. One veteran required a change in statin therapy based on the results (eAppendix).

Discussion

This project aimed to increase statin prescribing among female veterans with T2DM and/or CVD to reduce cardiovascular risk and increase pharmacogenomic testing using the PCED and care managed by CPPs. The results of this quality improvement project illustrated that both metrics have improved at CVVAMC as a result of the intervention. The results in both metrics now exceed the PCED national VA average, and the CVD metric also exceeds that of the facility male comparator group. While there was only a 1.2% increase from July 2023 to March 2024 for patients with T2DM, there was an 8.0% increase for patients with CVD. Despite standardized education on statin use, more veterans declined therapy than accepted it, mostly due to concern for AEs. Recording the reasons for declining statin therapy offered valuable insight that can be used in additional discussions with veterans and clinicians.

Pharmacogenomics gives clinicians the unique opportunity to take a proactive approach to better predict drug responses, potentially allowing for less trial and error with medications, fewer AEs, greater trust in the clinician, and improved medication adherence. The CPPs incorporated pharmacogenomic testing into their practice, which led to identifying 5 SLCO1B1 gene abnormalities. The PCED served as a powerful tool for advancing equity-focused quality improvement initiatives on a local level and was crucial in prioritizing the detection of veterans potentially receiving suboptimal care.

Limitations

The nature of “cold calls” made it challenging to establish contact for inclusion in this study. An alternative to increase engagement could have been scheduled phone or face-to-face visits. While the use of the PCED was crucial, data did not account for statins listed in the non-VA medication list. All 31 patients with statins prescribed outside the VA had a start date added to provide the most accurate representation of the data moving forward.

Another limitation in this project was its small sample size and population. CVVAMC serves about 6200 female veterans, with roughly 63% identifying as Black. The preponderance of Black individuals (83%) in this project is typical for the female patient population at CVVAMC but may not reflect the demographics of other populations. Other limitations to this project consisted of scheduling conflicts. Appointments for laboratory draws at community-based outpatient clinics were subject to availability, which resulted in some delay in completion of pharmacogenomic testing.

Conclusions

CPPs can help reduce inequity in health care delivery. Increased incorporation of the PCED into regular practice within the VA is recommended to continue addressing sex disparities in statin use, diabetes control, blood pressure management, cancer screenings, and vaccination needs. CVVAMC plans to expand its use through another quality improvement project focused on reducing sex disparities in blood pressure management. Improving educational resources made available to veterans on the importance of statin therapy and potential to mitigate AEs through use of the VA PHASER program also would be helpful. This project successfully improved CVVAMC metrics for female veterans appropriately prescribed statin therapy and increased access to pharmacogenomic testing. Most importantly, it helped close the sex-based gap in CVD risk reduction care.

BERLIN — Re-treatment with an antiepidermal growth factor receptor (EGFR) agent is effective in patients with chemorefractory metastatic colorectal cancer (mCRC) with RAS and BRAF wild-type tumors confirmed on circulating tumor DNA (ctDNA), although the sequencing of therapy does not seem to matter, suggest overall survival results from the crossover trial PARERE.

The findings nevertheless indicate that anti-EGFR rechallenge with panitumumab may prolong progression-free survival (PFS) over the multiple kinase inhibitor regorafenib. This suggests that “the most pragmatic choice” would be to give the anti-EGFR before regorafenib, said study presenter Marco Maria Germani, MD, Department of Translational Research and New Technologies in Medicine and Surgery, University of Pisa, Pisa, Italy.

The caveat, however, is in patients who have an anti-EGFR interval since previously receiving the drugs of < 6 months. Those patients appeared to do better if they had regorafenib first and then anti-EGFR rechallenge.

Overall, Germani said that “since [trifluridine/tipiracil] plus bevacizumab is today the third-line standard of care” in this patient population, “anti-EGFR re-treatment might be considered after progression” on that combination.

Germani presented the research on October 18 at the European Society for Medical Oncology (ESMO) Annual Meeting 2025, which was simultaneously published in the Annals of Oncology.

Michel P. Ducreux, MD, PhD, head of the Digestive Cancer Committee at Gustave Roussy, Villejuif, France, and invited discussant for the results, said, despite the study being negative, it is “very important to continue to perform this kind of trial to evaluate the [ideal] sequence in the treatment of our patients.”

He continued that the secondary endpoints in the trial of PFS and objective response and disease control rates were “fairly in favor of the use of rechallenge before regorafenib, and in my opinion, this is really quite convincing.”

Ducreux, who was not involved in PARERE trail, also pointed to the sex difference seen in the study, which suggested that women responded much better to having anti-EGFR retreatment before regorafenib than did men.

Similar findings have been reported in a number of other trials, and previous work has suggested that there are sex differences in the pharmacokinetics of several anticancer drugs. However, while this is “very important,” he said that “we never consider it, because we are not able to really explain [it].”

Overall, he concluded that, on the basis of these results, he would agree with the notion that it is better to propose a rechallenge with anti-EGFR treatment as the fourth-line therapy in this patient population, before administering regorafenib.

Ducreux explained that, after a partial response, tumors acquire resistance to EGFR inhibitors through alterations and mutations that occur during treatment, via nongenetic mechanisms, and through treatment-induced selection for preexisting mutations.

Previous work has shown that mutations, such as in the RAS gene, are detectable early during EGFR inhibitor therapy, but that they then decay exponentially once the drugs are stopped, with the potential that tumors regain their sensitivity to them.

Germani said that this means that ctDNA-guided retreatment with anti-EGFR therapies is a “promising approach” in pretreated patients with RAS and BRAF wild-type mCRC, and that the sequencing of the drugs may be important. Indeed, the REVERCE trial showed that giving regorafenib followed by the anti-EGFR drug cetuzximab was associated with longer overall survival than the other way around in anti-EGFR medication-naive patients.

 

Methods and Results

For PARERE, the researchers enrolled patients aged at least 18 years with RAS and BRAF wild-type mCRC who were previously treated with a first-line anti-EGFR-containing regimen and had at least a partial response or stable disease for at least 6 months.

The patients were also required to have had at least one intervening anti-EGFR-free line of therapy, and to have previously received treatment with fluoropyrimidine, oxaliplatin, irinotecan, and anti-angiogenics. At least 4 months were required to have passed between the end of anti-EGFR administration and screening for the study.

In all, 428 patients were screened between December 2020 and December 2024, with 213 patients with RAS and BRAF wild-type mCRC, as detected on ctDNA, enrolled. They were randomized to panitumumab or regorafenib until first progression, followed by regorafenib, if they started on panitumumab, or panitumumab, if they started on regorafenib, until second progression.

The median age of the patients was 61 years among those who started on panitumumab and 64 years among those initially given regorafenib in the trial, and 63% and 57%, respectively, were male. The median number of prior lines of therapy was two in both groups, and 65% and 69%, respectively, had received pantitumumab as their first-line anti-EGFR.

Initial findings from the study presented at the 2025 ASCO Annual Meeting indicated that, after a median follow-up of 23.5 months, there was no significant difference in the median first PFS between the two treatment arms.

However, patients who started with panitumumab had a significant improvement in both the objective response and disease control rates (P < .001), as well as a signal for a potentially longer median second PFS, than those who started with regorafenib, particularly on the per-protocol analysis.

Presenting the overall survival results, Germani said that there was no significant difference between the groups on the intention-to-treat analysis, at a stratified hazard ratio of 1.13 (P = .440), or on the per-protocol analysis, at a hazard ratio of 1.07 (P = .730).

“We then ran a subgroup analysis,” he continued, “and we found out that an anti-EGFR-free interval before liquid biopsy shorter than 6 months was associated with less benefit from a panitumumab [first] sequence, which is biologically sound.”

It was also observed that women did significantly better when having panitumumab first, whereas men did not, for which “we do not have a clear biological explanation,” Germani added.

Confining the analysis to so-called “hyperselected” patients, who not only were RAS and BRAF wild type but also had no pathogenic mutations associated with anti-EGFR resistance, did not reveal any significant overall survival differences between the treatment groups.

However, Ducreux took issue with the way in which hyperselection, which is turning up more and more regularly in trials, is defined, as the choice of which mutations to include varies widely. He suggested that a consensus group be assembled to resolve this issue.

Looking more broadly, the researchers were able to show that, in this updated analysis, anti-EGFR re-treatment was superior to regorafenib regardless of the treatment sequence in terms of PFS, at 4.2 months vs 2.4 months (P = .103) when given first in the trial, and 3.9 months vs 2.7 months (P = .019) when given second in the trial, as well as in terms of objective response and disease control rates.

 

Adverse Events

In terms of safety, the results showed that, as expected, acneiform rash, fatigue, and hypomagnesemia were the most common adverse events associated with panitumumb, while those with regorafenib were fatigue, hand-foot skin reactions, and hypertension.

There were no notable differences in the number of patients receiving a post-study treatment nor in the post-study therapeutic choices, between the study arms.

The study was sponsored by GONO Foundation and partially supported by Amgen and Bayer. Germani declared having relationships with MSD and Amgen. Ducreux declared having relationships with Amgen, Bayer, BeiGene, Incyte, Jazz, Merck KGaA, Merck Serono, Merck Sharp & Dohme, Pierre Fabre, Roche, Servier, Keocyt, AbbVie, Abcely, Arcus, Bayer, BMS, Boehringer, GlaxoSmithKline, Sanofi, Scandion, and Zymeworks.

A version of this article first appeared on Medscape.com.

BERLIN — Re-treatment with an antiepidermal growth factor receptor (EGFR) agent is effective in patients with chemorefractory metastatic colorectal cancer (mCRC) with RAS and BRAF wild-type tumors confirmed on circulating tumor DNA (ctDNA), although the sequencing of therapy does not seem to matter, suggest overall survival results from the crossover trial PARERE.

The findings nevertheless indicate that anti-EGFR rechallenge with panitumumab may prolong progression-free survival (PFS) over the multiple kinase inhibitor regorafenib. This suggests that “the most pragmatic choice” would be to give the anti-EGFR before regorafenib, said study presenter Marco Maria Germani, MD, Department of Translational Research and New Technologies in Medicine and Surgery, University of Pisa, Pisa, Italy.

The caveat, however, is in patients who have an anti-EGFR interval since previously receiving the drugs of < 6 months. Those patients appeared to do better if they had regorafenib first and then anti-EGFR rechallenge.

Overall, Germani said that “since [trifluridine/tipiracil] plus bevacizumab is today the third-line standard of care” in this patient population, “anti-EGFR re-treatment might be considered after progression” on that combination.

Germani presented the research on October 18 at the European Society for Medical Oncology (ESMO) Annual Meeting 2025, which was simultaneously published in the Annals of Oncology.

Michel P. Ducreux, MD, PhD, head of the Digestive Cancer Committee at Gustave Roussy, Villejuif, France, and invited discussant for the results, said, despite the study being negative, it is “very important to continue to perform this kind of trial to evaluate the [ideal] sequence in the treatment of our patients.”

He continued that the secondary endpoints in the trial of PFS and objective response and disease control rates were “fairly in favor of the use of rechallenge before regorafenib, and in my opinion, this is really quite convincing.”

Ducreux, who was not involved in PARERE trail, also pointed to the sex difference seen in the study, which suggested that women responded much better to having anti-EGFR retreatment before regorafenib than did men.

Similar findings have been reported in a number of other trials, and previous work has suggested that there are sex differences in the pharmacokinetics of several anticancer drugs. However, while this is “very important,” he said that “we never consider it, because we are not able to really explain [it].”

Overall, he concluded that, on the basis of these results, he would agree with the notion that it is better to propose a rechallenge with anti-EGFR treatment as the fourth-line therapy in this patient population, before administering regorafenib.

Ducreux explained that, after a partial response, tumors acquire resistance to EGFR inhibitors through alterations and mutations that occur during treatment, via nongenetic mechanisms, and through treatment-induced selection for preexisting mutations.

Previous work has shown that mutations, such as in the RAS gene, are detectable early during EGFR inhibitor therapy, but that they then decay exponentially once the drugs are stopped, with the potential that tumors regain their sensitivity to them.

Germani said that this means that ctDNA-guided retreatment with anti-EGFR therapies is a “promising approach” in pretreated patients with RAS and BRAF wild-type mCRC, and that the sequencing of the drugs may be important. Indeed, the REVERCE trial showed that giving regorafenib followed by the anti-EGFR drug cetuzximab was associated with longer overall survival than the other way around in anti-EGFR medication-naive patients.

 

Methods and Results

For PARERE, the researchers enrolled patients aged at least 18 years with RAS and BRAF wild-type mCRC who were previously treated with a first-line anti-EGFR-containing regimen and had at least a partial response or stable disease for at least 6 months.

The patients were also required to have had at least one intervening anti-EGFR-free line of therapy, and to have previously received treatment with fluoropyrimidine, oxaliplatin, irinotecan, and anti-angiogenics. At least 4 months were required to have passed between the end of anti-EGFR administration and screening for the study.

In all, 428 patients were screened between December 2020 and December 2024, with 213 patients with RAS and BRAF wild-type mCRC, as detected on ctDNA, enrolled. They were randomized to panitumumab or regorafenib until first progression, followed by regorafenib, if they started on panitumumab, or panitumumab, if they started on regorafenib, until second progression.

The median age of the patients was 61 years among those who started on panitumumab and 64 years among those initially given regorafenib in the trial, and 63% and 57%, respectively, were male. The median number of prior lines of therapy was two in both groups, and 65% and 69%, respectively, had received pantitumumab as their first-line anti-EGFR.

Initial findings from the study presented at the 2025 ASCO Annual Meeting indicated that, after a median follow-up of 23.5 months, there was no significant difference in the median first PFS between the two treatment arms.

However, patients who started with panitumumab had a significant improvement in both the objective response and disease control rates (P < .001), as well as a signal for a potentially longer median second PFS, than those who started with regorafenib, particularly on the per-protocol analysis.

Presenting the overall survival results, Germani said that there was no significant difference between the groups on the intention-to-treat analysis, at a stratified hazard ratio of 1.13 (P = .440), or on the per-protocol analysis, at a hazard ratio of 1.07 (P = .730).

“We then ran a subgroup analysis,” he continued, “and we found out that an anti-EGFR-free interval before liquid biopsy shorter than 6 months was associated with less benefit from a panitumumab [first] sequence, which is biologically sound.”

It was also observed that women did significantly better when having panitumumab first, whereas men did not, for which “we do not have a clear biological explanation,” Germani added.

Confining the analysis to so-called “hyperselected” patients, who not only were RAS and BRAF wild type but also had no pathogenic mutations associated with anti-EGFR resistance, did not reveal any significant overall survival differences between the treatment groups.

However, Ducreux took issue with the way in which hyperselection, which is turning up more and more regularly in trials, is defined, as the choice of which mutations to include varies widely. He suggested that a consensus group be assembled to resolve this issue.

Looking more broadly, the researchers were able to show that, in this updated analysis, anti-EGFR re-treatment was superior to regorafenib regardless of the treatment sequence in terms of PFS, at 4.2 months vs 2.4 months (P = .103) when given first in the trial, and 3.9 months vs 2.7 months (P = .019) when given second in the trial, as well as in terms of objective response and disease control rates.

 

Adverse Events

In terms of safety, the results showed that, as expected, acneiform rash, fatigue, and hypomagnesemia were the most common adverse events associated with panitumumb, while those with regorafenib were fatigue, hand-foot skin reactions, and hypertension.

There were no notable differences in the number of patients receiving a post-study treatment nor in the post-study therapeutic choices, between the study arms.

The study was sponsored by GONO Foundation and partially supported by Amgen and Bayer. Germani declared having relationships with MSD and Amgen. Ducreux declared having relationships with Amgen, Bayer, BeiGene, Incyte, Jazz, Merck KGaA, Merck Serono, Merck Sharp & Dohme, Pierre Fabre, Roche, Servier, Keocyt, AbbVie, Abcely, Arcus, Bayer, BMS, Boehringer, GlaxoSmithKline, Sanofi, Scandion, and Zymeworks.

A version of this article first appeared on Medscape.com.

BERLIN — Re-treatment with an antiepidermal growth factor receptor (EGFR) agent is effective in patients with chemorefractory metastatic colorectal cancer (mCRC) with RAS and BRAF wild-type tumors confirmed on circulating tumor DNA (ctDNA), although the sequencing of therapy does not seem to matter, suggest overall survival results from the crossover trial PARERE.

The findings nevertheless indicate that anti-EGFR rechallenge with panitumumab may prolong progression-free survival (PFS) over the multiple kinase inhibitor regorafenib. This suggests that “the most pragmatic choice” would be to give the anti-EGFR before regorafenib, said study presenter Marco Maria Germani, MD, Department of Translational Research and New Technologies in Medicine and Surgery, University of Pisa, Pisa, Italy.

The caveat, however, is in patients who have an anti-EGFR interval since previously receiving the drugs of < 6 months. Those patients appeared to do better if they had regorafenib first and then anti-EGFR rechallenge.

Overall, Germani said that “since [trifluridine/tipiracil] plus bevacizumab is today the third-line standard of care” in this patient population, “anti-EGFR re-treatment might be considered after progression” on that combination.

Germani presented the research on October 18 at the European Society for Medical Oncology (ESMO) Annual Meeting 2025, which was simultaneously published in the Annals of Oncology.

Michel P. Ducreux, MD, PhD, head of the Digestive Cancer Committee at Gustave Roussy, Villejuif, France, and invited discussant for the results, said, despite the study being negative, it is “very important to continue to perform this kind of trial to evaluate the [ideal] sequence in the treatment of our patients.”

He continued that the secondary endpoints in the trial of PFS and objective response and disease control rates were “fairly in favor of the use of rechallenge before regorafenib, and in my opinion, this is really quite convincing.”

Ducreux, who was not involved in PARERE trail, also pointed to the sex difference seen in the study, which suggested that women responded much better to having anti-EGFR retreatment before regorafenib than did men.

Similar findings have been reported in a number of other trials, and previous work has suggested that there are sex differences in the pharmacokinetics of several anticancer drugs. However, while this is “very important,” he said that “we never consider it, because we are not able to really explain [it].”

Overall, he concluded that, on the basis of these results, he would agree with the notion that it is better to propose a rechallenge with anti-EGFR treatment as the fourth-line therapy in this patient population, before administering regorafenib.

Ducreux explained that, after a partial response, tumors acquire resistance to EGFR inhibitors through alterations and mutations that occur during treatment, via nongenetic mechanisms, and through treatment-induced selection for preexisting mutations.

Previous work has shown that mutations, such as in the RAS gene, are detectable early during EGFR inhibitor therapy, but that they then decay exponentially once the drugs are stopped, with the potential that tumors regain their sensitivity to them.

Germani said that this means that ctDNA-guided retreatment with anti-EGFR therapies is a “promising approach” in pretreated patients with RAS and BRAF wild-type mCRC, and that the sequencing of the drugs may be important. Indeed, the REVERCE trial showed that giving regorafenib followed by the anti-EGFR drug cetuzximab was associated with longer overall survival than the other way around in anti-EGFR medication-naive patients.

 

Methods and Results

For PARERE, the researchers enrolled patients aged at least 18 years with RAS and BRAF wild-type mCRC who were previously treated with a first-line anti-EGFR-containing regimen and had at least a partial response or stable disease for at least 6 months.

The patients were also required to have had at least one intervening anti-EGFR-free line of therapy, and to have previously received treatment with fluoropyrimidine, oxaliplatin, irinotecan, and anti-angiogenics. At least 4 months were required to have passed between the end of anti-EGFR administration and screening for the study.

In all, 428 patients were screened between December 2020 and December 2024, with 213 patients with RAS and BRAF wild-type mCRC, as detected on ctDNA, enrolled. They were randomized to panitumumab or regorafenib until first progression, followed by regorafenib, if they started on panitumumab, or panitumumab, if they started on regorafenib, until second progression.

The median age of the patients was 61 years among those who started on panitumumab and 64 years among those initially given regorafenib in the trial, and 63% and 57%, respectively, were male. The median number of prior lines of therapy was two in both groups, and 65% and 69%, respectively, had received pantitumumab as their first-line anti-EGFR.

Initial findings from the study presented at the 2025 ASCO Annual Meeting indicated that, after a median follow-up of 23.5 months, there was no significant difference in the median first PFS between the two treatment arms.

However, patients who started with panitumumab had a significant improvement in both the objective response and disease control rates (P < .001), as well as a signal for a potentially longer median second PFS, than those who started with regorafenib, particularly on the per-protocol analysis.

Presenting the overall survival results, Germani said that there was no significant difference between the groups on the intention-to-treat analysis, at a stratified hazard ratio of 1.13 (P = .440), or on the per-protocol analysis, at a hazard ratio of 1.07 (P = .730).

“We then ran a subgroup analysis,” he continued, “and we found out that an anti-EGFR-free interval before liquid biopsy shorter than 6 months was associated with less benefit from a panitumumab [first] sequence, which is biologically sound.”

It was also observed that women did significantly better when having panitumumab first, whereas men did not, for which “we do not have a clear biological explanation,” Germani added.

Confining the analysis to so-called “hyperselected” patients, who not only were RAS and BRAF wild type but also had no pathogenic mutations associated with anti-EGFR resistance, did not reveal any significant overall survival differences between the treatment groups.

However, Ducreux took issue with the way in which hyperselection, which is turning up more and more regularly in trials, is defined, as the choice of which mutations to include varies widely. He suggested that a consensus group be assembled to resolve this issue.

Looking more broadly, the researchers were able to show that, in this updated analysis, anti-EGFR re-treatment was superior to regorafenib regardless of the treatment sequence in terms of PFS, at 4.2 months vs 2.4 months (P = .103) when given first in the trial, and 3.9 months vs 2.7 months (P = .019) when given second in the trial, as well as in terms of objective response and disease control rates.

 

Adverse Events

In terms of safety, the results showed that, as expected, acneiform rash, fatigue, and hypomagnesemia were the most common adverse events associated with panitumumb, while those with regorafenib were fatigue, hand-foot skin reactions, and hypertension.

There were no notable differences in the number of patients receiving a post-study treatment nor in the post-study therapeutic choices, between the study arms.

The study was sponsored by GONO Foundation and partially supported by Amgen and Bayer. Germani declared having relationships with MSD and Amgen. Ducreux declared having relationships with Amgen, Bayer, BeiGene, Incyte, Jazz, Merck KGaA, Merck Serono, Merck Sharp & Dohme, Pierre Fabre, Roche, Servier, Keocyt, AbbVie, Abcely, Arcus, Bayer, BMS, Boehringer, GlaxoSmithKline, Sanofi, Scandion, and Zymeworks.

A version of this article first appeared on Medscape.com.

Repeated and frequent hospitalizations—sometimes referred to as the revolving door phenomenon— are a particular risk for patients during the first month after discharge. Early psychiatric readmission is a standard indicator of adverse outcomes. However, the results 

The quality of previous care has long been thought to be a driver of readmission. If that’s the case, a 2025 study suggests that on average veterans received high-quality inpatient psychiatric services at Veterans Health Administration (VHA) facilities across the nation and that may have been key to keeping readmissions down. Analyzing data from 88,954 veterans who received care at VHA Inpatient Mental Health (IMH) services, the researchers found a “relatively low” rate of readmission within 30 days: 7.1% compared with 8% to 31% of other psychiatric patients in the US. With 40,220 unique patients receiving IMH care per year on average between October 2019 and September 2022, a 7.1% readmission rate means > 2800 30-day readmissions annually.

Research has found that veterans who receive care at the VA have better outcomes than those treated in the private sector. Part of that has to do with practitioners who understand the unique needs of their patients. Veterans may have posttraumatic stress disorder or multiple diagnoses, such as depression, panic disorder, and a substance use disorder. Their mental health issues may also coexist with physical health problems, such as traumatic brain injuries due to explosions.

“If you’re trained at the VA, you learn something important about veteran mental health care that you’ll never get if you’re trained someplace else,” Rodney R. Baker, PhD, retired mental health director and chief of psychology for the South Texas VA Health Care System, said recently. Community clinicians may not know how to collect and incorporate information about a patient’s military history, including details about deployments, combat exposure, injuries, military sexual trauma, and unit culture. They may also lack expertise in navigating the transition between military and veteran life, now considered a critical adjustment period.

“This is a unique population,” said Conwell Smith, the American Psychological Association’s deputy chief of military and veteran policy. “Sending veterans out to the community without requiring that mental health care providers understand them is concerning.”

IMH services aim to stabilize mental health crises and improve veterans’ functioning through patient-centered, evidence-based, and recovery-oriented approaches shown to reduce readmission rates. Treatment generally involves a minimum of 4 hours of interdisciplinary, therapeutic programming each day. And upon discharge, the inpatient care team facilitates the patient’s transition to appropriate outpatient services.

Follow-up care, particularly during the first 30 days, has proved critical in reducing readmissions. In studies that have analyzed postdischarge interventions (psychoeducation, mentoring, community-based hospital treatment, use of continuous follow-up and compulsory community treatment), all found fewer hospitalizations when compared to a control group, or a smaller number of admissions after the intervention. 

Mental health care for veterans should be provided by experienced practitioners—but those practitioners are leaving VA. According to the VA Office of Inspector General, 57% of medical centers report a shortage of psychologists. And according to the VA’s monthly Workforce Dashboard, the VHA lost 234 psychologists in the first 9 months of 2025. The VA has also announced plans to cut 30,000 jobs by the end of the year and impose caps on staff at every medical center.

“This approach locks in permanent VA understaffing just as demand for mental health services is projected to continue growing through 2030,” said Russell Lemle, PhD, a clinical psychologist and senior policy analyst for the Veterans Healthcare Policy Institute. “The private sector can’t fill this gap either—over a third of Americans live in areas already facing mental health professional shortages. That’s not taking care of our veterans.

“Unless actions are taken quickly to reverse the trend, its mental health services could easily diminish substantially within 10 to 20 years.”

Repeated and frequent hospitalizations—sometimes referred to as the revolving door phenomenon— are a particular risk for patients during the first month after discharge. Early psychiatric readmission is a standard indicator of adverse outcomes. However, the results 

The quality of previous care has long been thought to be a driver of readmission. If that’s the case, a 2025 study suggests that on average veterans received high-quality inpatient psychiatric services at Veterans Health Administration (VHA) facilities across the nation and that may have been key to keeping readmissions down. Analyzing data from 88,954 veterans who received care at VHA Inpatient Mental Health (IMH) services, the researchers found a “relatively low” rate of readmission within 30 days: 7.1% compared with 8% to 31% of other psychiatric patients in the US. With 40,220 unique patients receiving IMH care per year on average between October 2019 and September 2022, a 7.1% readmission rate means > 2800 30-day readmissions annually.

Research has found that veterans who receive care at the VA have better outcomes than those treated in the private sector. Part of that has to do with practitioners who understand the unique needs of their patients. Veterans may have posttraumatic stress disorder or multiple diagnoses, such as depression, panic disorder, and a substance use disorder. Their mental health issues may also coexist with physical health problems, such as traumatic brain injuries due to explosions.

“If you’re trained at the VA, you learn something important about veteran mental health care that you’ll never get if you’re trained someplace else,” Rodney R. Baker, PhD, retired mental health director and chief of psychology for the South Texas VA Health Care System, said recently. Community clinicians may not know how to collect and incorporate information about a patient’s military history, including details about deployments, combat exposure, injuries, military sexual trauma, and unit culture. They may also lack expertise in navigating the transition between military and veteran life, now considered a critical adjustment period.

“This is a unique population,” said Conwell Smith, the American Psychological Association’s deputy chief of military and veteran policy. “Sending veterans out to the community without requiring that mental health care providers understand them is concerning.”

IMH services aim to stabilize mental health crises and improve veterans’ functioning through patient-centered, evidence-based, and recovery-oriented approaches shown to reduce readmission rates. Treatment generally involves a minimum of 4 hours of interdisciplinary, therapeutic programming each day. And upon discharge, the inpatient care team facilitates the patient’s transition to appropriate outpatient services.

Follow-up care, particularly during the first 30 days, has proved critical in reducing readmissions. In studies that have analyzed postdischarge interventions (psychoeducation, mentoring, community-based hospital treatment, use of continuous follow-up and compulsory community treatment), all found fewer hospitalizations when compared to a control group, or a smaller number of admissions after the intervention. 

Mental health care for veterans should be provided by experienced practitioners—but those practitioners are leaving VA. According to the VA Office of Inspector General, 57% of medical centers report a shortage of psychologists. And according to the VA’s monthly Workforce Dashboard, the VHA lost 234 psychologists in the first 9 months of 2025. The VA has also announced plans to cut 30,000 jobs by the end of the year and impose caps on staff at every medical center.

“This approach locks in permanent VA understaffing just as demand for mental health services is projected to continue growing through 2030,” said Russell Lemle, PhD, a clinical psychologist and senior policy analyst for the Veterans Healthcare Policy Institute. “The private sector can’t fill this gap either—over a third of Americans live in areas already facing mental health professional shortages. That’s not taking care of our veterans.

“Unless actions are taken quickly to reverse the trend, its mental health services could easily diminish substantially within 10 to 20 years.”

Repeated and frequent hospitalizations—sometimes referred to as the revolving door phenomenon— are a particular risk for patients during the first month after discharge. Early psychiatric readmission is a standard indicator of adverse outcomes. However, the results 

The quality of previous care has long been thought to be a driver of readmission. If that’s the case, a 2025 study suggests that on average veterans received high-quality inpatient psychiatric services at Veterans Health Administration (VHA) facilities across the nation and that may have been key to keeping readmissions down. Analyzing data from 88,954 veterans who received care at VHA Inpatient Mental Health (IMH) services, the researchers found a “relatively low” rate of readmission within 30 days: 7.1% compared with 8% to 31% of other psychiatric patients in the US. With 40,220 unique patients receiving IMH care per year on average between October 2019 and September 2022, a 7.1% readmission rate means > 2800 30-day readmissions annually.

Research has found that veterans who receive care at the VA have better outcomes than those treated in the private sector. Part of that has to do with practitioners who understand the unique needs of their patients. Veterans may have posttraumatic stress disorder or multiple diagnoses, such as depression, panic disorder, and a substance use disorder. Their mental health issues may also coexist with physical health problems, such as traumatic brain injuries due to explosions.

“If you’re trained at the VA, you learn something important about veteran mental health care that you’ll never get if you’re trained someplace else,” Rodney R. Baker, PhD, retired mental health director and chief of psychology for the South Texas VA Health Care System, said recently. Community clinicians may not know how to collect and incorporate information about a patient’s military history, including details about deployments, combat exposure, injuries, military sexual trauma, and unit culture. They may also lack expertise in navigating the transition between military and veteran life, now considered a critical adjustment period.

“This is a unique population,” said Conwell Smith, the American Psychological Association’s deputy chief of military and veteran policy. “Sending veterans out to the community without requiring that mental health care providers understand them is concerning.”

IMH services aim to stabilize mental health crises and improve veterans’ functioning through patient-centered, evidence-based, and recovery-oriented approaches shown to reduce readmission rates. Treatment generally involves a minimum of 4 hours of interdisciplinary, therapeutic programming each day. And upon discharge, the inpatient care team facilitates the patient’s transition to appropriate outpatient services.

Follow-up care, particularly during the first 30 days, has proved critical in reducing readmissions. In studies that have analyzed postdischarge interventions (psychoeducation, mentoring, community-based hospital treatment, use of continuous follow-up and compulsory community treatment), all found fewer hospitalizations when compared to a control group, or a smaller number of admissions after the intervention. 

Mental health care for veterans should be provided by experienced practitioners—but those practitioners are leaving VA. According to the VA Office of Inspector General, 57% of medical centers report a shortage of psychologists. And according to the VA’s monthly Workforce Dashboard, the VHA lost 234 psychologists in the first 9 months of 2025. The VA has also announced plans to cut 30,000 jobs by the end of the year and impose caps on staff at every medical center.

“This approach locks in permanent VA understaffing just as demand for mental health services is projected to continue growing through 2030,” said Russell Lemle, PhD, a clinical psychologist and senior policy analyst for the Veterans Healthcare Policy Institute. “The private sector can’t fill this gap either—over a third of Americans live in areas already facing mental health professional shortages. That’s not taking care of our veterans.

“Unless actions are taken quickly to reverse the trend, its mental health services could easily diminish substantially within 10 to 20 years.”

Hi, everyone. I’m Dr Kenny Lin. I am a family physician and associate director of the Lancaster General Hospital Family Medicine Residency, and I blog at Common Sense Family Doctor.

The receding of the pandemic and the understandable desire to return to normalcy has made COVID-19 vaccines a lower priority for many of our patients. However, family physicians should keep in mind that from October 1, 2024, to September 6, 2025, COVID-19 was responsible for an estimated 3.2 to 4.6 million outpatient visits, 360,000 to 520,000 hospitalizations, and 42,000 to 60,000 deaths.

In a previous commentary, I discussed the worsening disconnect between the evidence supporting the effectiveness and safety of vaccinations and increasing reluctance of patients and parents to receive them, fueled by misinformation from federal health agencies and the packing of the Advisory Committee on Immunization Practices (ACIP) with vaccine skeptics. Since then, Secretary of Health and Human Services (HHS), Robert F. Kennedy, Jr, has fired Dr Susan Monarez, his handpicked director of the CDC. This caused three senior CDC officials to resign in protest and precipitated further turmoil at the embattled agency. 

The FDA has approved 3 updated COVID-19 vaccines targeted to currently circulating strains: an mRNA vaccine from Moderna (Spikevax) for those aged 6 months or older; an mRNA vaccine from Pfizer/BioNTech (Comirnaty) for those aged ≥ 5 years; and a protein subunit vaccine from Novavax (Nuvaxovid) for those aged ≥ 12 years. However, approvals restricting the scope of these approvals to certain high-risk groups, combined with the ACIP’s recent decision to not explicitly recommend them for any group, have complicated access for many patients.

Medical groups, including the American Academy of Pediatrics (AAP), the American Academy of Family Physicians (AAFP), and the American College of Obstetricians and Gynecologists (ACOG), have published their own recommendations (Table). Of note, in opposition to the FDA and ACIP, the AAP and AAFP strongly recommend routine vaccination for children aged 6 to 23 months because they have the highest risk for hospitalization. The AAFP and ACOG both recommend COVID-19 vaccination in pregnancy to protect the pregnant patient and provide passive antibody protection to their infants up to 6 months of age. The Vaccine Integrity Project’s review of 12 safety studies published since June 2024 found that mRNA vaccines were not associated with increases in any adverse maternal or infant outcomes and had a possible protective effect against preterm birth.

In my previous commentary, 70% of Medscape readers indicated that they would follow vaccination recommendations from AAP even if they differed from CDC guidance. Administering vaccines outside of FDA labeling indications (i.e., “off label”) typically requires a physician’s prescription, which will almost certainly reduce COVID-19 vaccine uptake in children and pregnant patients, given that most people received these shots in pharmacies during the 2024-25 season. CVS and Walgreens, the country’s two largest pharmacy chains, are requiring physician prescriptions or waiting for ACIP guidance to make the new vaccines available in many states. However, an increasing number of states have implemented executive orders or passed legislation to permit pharmacists to provide vaccines to anyone who wants them. For example, the Pennsylvania State Board of Pharmacy voted unanimously to issue guidance that would allow pharmacists to administer any vaccines recommended by AAFP, AAP, or ACOG.

Erosion of vaccine uptake could easily worsen the burden of illness for our patients and the health system. Navigating the unnecessarily complex landscape of COVID-19 vaccines will be challenging, but it remains worthwhile.
 

Kenneth W. Lin, MD, MPH, Associate Director, Department of Family Medicine, Lancaster General Hospital, Lancaster, Pennsylvania, has disclosed the following relevant financial relationships: Serve(d) as a director, officer, partner, employee, advisor, consultant, or trustee for: UpToDate; American Academy of Family Physicians; Archdiocese of Washington; Association of Prevention Teaching and Research.

A version of this article appeared on Medscape.com.

Hi, everyone. I’m Dr Kenny Lin. I am a family physician and associate director of the Lancaster General Hospital Family Medicine Residency, and I blog at Common Sense Family Doctor.

The receding of the pandemic and the understandable desire to return to normalcy has made COVID-19 vaccines a lower priority for many of our patients. However, family physicians should keep in mind that from October 1, 2024, to September 6, 2025, COVID-19 was responsible for an estimated 3.2 to 4.6 million outpatient visits, 360,000 to 520,000 hospitalizations, and 42,000 to 60,000 deaths.

In a previous commentary, I discussed the worsening disconnect between the evidence supporting the effectiveness and safety of vaccinations and increasing reluctance of patients and parents to receive them, fueled by misinformation from federal health agencies and the packing of the Advisory Committee on Immunization Practices (ACIP) with vaccine skeptics. Since then, Secretary of Health and Human Services (HHS), Robert F. Kennedy, Jr, has fired Dr Susan Monarez, his handpicked director of the CDC. This caused three senior CDC officials to resign in protest and precipitated further turmoil at the embattled agency. 

The FDA has approved 3 updated COVID-19 vaccines targeted to currently circulating strains: an mRNA vaccine from Moderna (Spikevax) for those aged 6 months or older; an mRNA vaccine from Pfizer/BioNTech (Comirnaty) for those aged ≥ 5 years; and a protein subunit vaccine from Novavax (Nuvaxovid) for those aged ≥ 12 years. However, approvals restricting the scope of these approvals to certain high-risk groups, combined with the ACIP’s recent decision to not explicitly recommend them for any group, have complicated access for many patients.

Medical groups, including the American Academy of Pediatrics (AAP), the American Academy of Family Physicians (AAFP), and the American College of Obstetricians and Gynecologists (ACOG), have published their own recommendations (Table). Of note, in opposition to the FDA and ACIP, the AAP and AAFP strongly recommend routine vaccination for children aged 6 to 23 months because they have the highest risk for hospitalization. The AAFP and ACOG both recommend COVID-19 vaccination in pregnancy to protect the pregnant patient and provide passive antibody protection to their infants up to 6 months of age. The Vaccine Integrity Project’s review of 12 safety studies published since June 2024 found that mRNA vaccines were not associated with increases in any adverse maternal or infant outcomes and had a possible protective effect against preterm birth.

In my previous commentary, 70% of Medscape readers indicated that they would follow vaccination recommendations from AAP even if they differed from CDC guidance. Administering vaccines outside of FDA labeling indications (i.e., “off label”) typically requires a physician’s prescription, which will almost certainly reduce COVID-19 vaccine uptake in children and pregnant patients, given that most people received these shots in pharmacies during the 2024-25 season. CVS and Walgreens, the country’s two largest pharmacy chains, are requiring physician prescriptions or waiting for ACIP guidance to make the new vaccines available in many states. However, an increasing number of states have implemented executive orders or passed legislation to permit pharmacists to provide vaccines to anyone who wants them. For example, the Pennsylvania State Board of Pharmacy voted unanimously to issue guidance that would allow pharmacists to administer any vaccines recommended by AAFP, AAP, or ACOG.

Erosion of vaccine uptake could easily worsen the burden of illness for our patients and the health system. Navigating the unnecessarily complex landscape of COVID-19 vaccines will be challenging, but it remains worthwhile.