Variation in Early Diffusion of Sacubitril/Valsartan and Implications for Understanding Novel Drug Diffusion

, , ,
The American Journal of Managed Care, December 2021, Volume 27, Issue 12

Rates of sacubitril/valsartan diffusion have been slow and there has been significant geographic variation, highlighting the importance of local prescribing patterns in early drug diffusion.

ABSTRACT

Objectives: In the PARADIGM-HF trial, sacubitril/valsartan demonstrated a 20% reduction in mortality and heart failure hospitalization compared with standard angiotensin-converting enzyme inhibitor therapy. Despite this and a class I indication, drug diffusion has been much slower than anticipated. This study aims to examine the variation in early diffusion of sacubitril/valsartan and describe the factors associated with high and low rates of early use.

Study Design: Annual, cross-sectional analyses between January 2016 and December 2018.

Methods: We created a nationally representative cohort of Medicare fee-for-service beneficiaries with heart failure with reduced ejection fraction fully enrolled in parts A, B, and D for at least 1 year between 2016 and 2018. Sacubitril/valsartan use was determined using National Drug Codes. We generated age, sex, and race–adjusted rates of sacubitril/valsartan prescribing by hospital referral region from 2016 to 2018. We also examined the factors associated with high and low rates of early use.

Results: Early use rates of sacubitril/valsartan were low: 1.9% in 2016, 3.3% in 2017, and 4.0% in 2018. Even after controlling for out-of-pocket co-payments, there was substantial geographic variation in early use, with most early use concentrated in the Northeast and South.

Conclusions: There has been substantial variation in the early diffusion of sacubitril/valsartan. In addition to drug cost, geographic prescribing patterns appear to play a major role in early drug diffusion.

Am J Manag Care. 2021;27(12):524-530. https://doi.org/10.37765/ajmc.2021.88791

_____

Takeaway Points

There has been substantial variation in the early diffusion of sacubitril/valsartan. Geographic prescribing patterns appear to play a major role in early drug diffusion.

  • Despite the success of sacubitril/valsartan in clinical trials, early diffusion has been slow and highly variable.
  • Even after controlling for drug costs, local prescribing patterns significantly predict early diffusion.
  • For health systems and health plans, local prescribing patterns play a major role in early drug diffusion and can serve as another tool to optimize the speed of diffusion of novel therapies into clinical practice.

_____

Sacubitril/valsartan was approved by the FDA on July 7, 2015, for the treatment of heart failure with reduced ejection fraction (HFrEF).1 In the PARADIGM-HF study, sacubitril/valsartan demonstrated a 21% relative reduction in hospitalizations for HFrEF and a 26% relative decrease in mortality compared with standard angiotensin-converting enzyme (ACE) inhibitor therapy.2 Despite this and a class I indication from the American Heart Association, the American College of Cardiology, and the Heart Failure Society of America,3 early reports have found lower-than-anticipated rates of early adoption,4 prompting much discussion in both cardiology and health policy circles. To date, only high out-of-pocket drug costs have been identified as a barrier to early drug diffusion.5-7

Based on work from other fields, we know that patient, provider, and practice characteristics can also affect early drug diffusion. For example, a recent study of the diffusion of new, similarly expensive antidiabetic drugs found that most early use was driven by a small group of “high prescribers” operating in specific practice settings.8 Variation in the use of sacubitril/valsartan for HFrEF and factors associated with early diffusion are less well understood. Because 80% of deaths from heart failure occur in individuals 65 years and older,9 the aim of this study is to explore the variation in the use of sacubitril/valsartan in eligible Medicare beneficiaries with HFrEF between 2016 and 2018, after controlling for out-of-pocket co-payments.

METHODS

This study was approved by the institutional review board at Dartmouth College. This study is also compliant with the Strengthening the Reporting of Observational Studies in Epidemiology reporting guideline for retrospective cohort studies.

Study Population

We used the complete 100% national sample of fee-for-service (FFS) patients insured by Medicare in 2016 to 2018. We first identified Medicare beneficiaries who were 65 years or older on January 1 of each year and had 12 months of continuous parts A and B coverage from the respective Medicare Beneficiary Summary File. We then searched their inpatient (Medicare Provider Analysis and Review [MedPAR]) and outpatient (Physician/Supplier Part B file and Outpatient file) claim records to identify diagnoses of HFrEF using the International Classification of Diseases, Tenth Revision codes I50.2x and I50.4x. Patients with at least 1 inpatient claim or at least 2 outpatient claims with these codes, at least 7 days apart in a given calendar year, were determined to have HFrEF. The date of the inpatient hospitalization or the date of the first HFrEF diagnosis was considered the index HFrEF date.

To determine comorbidities and health care utilization in the year prior, we then limited the study population to beneficiaries with at least 1 year of FFS coverage prior to their index HFrEF date. We determined sociodemographic characteristics including age, sex, race, dual Medicare/Medicaid eligibility, geographic region,10,11 and zip code tabulation area–level estimates of socioeconomic status12 using the Medicare Master Beneficiary Summary File. We determined HF severity using the MedPAR file and comorbidities using the Chronic Conditions Data Warehouse.13 Finally, to examine drug use, we limited the study population to patients who were fully enrolled in Part D for at least 1 year after their HFrEF index date.

Sacubitril/Valsartan Exposure

Sacubitril/valsartan use was determined using National Drug Codes (eAppendix A [eAppendices available at ajmc.com]). Any beneficiary with at least 1 fill for sacubitril/valsartan in the calendar year was included. All others were considered eligible but not initiated on sacubitril/valsartan therapy. We report age, sex, and race (ASR)–adjusted rates of sacubitril/valsartan use by hospital referral region (HRR).

For all beneficiaries who received sacubitril/valsartan, we used Medicare Part D to determine out-of-pocket co-payments. The amount paid out of pocket by the beneficiary is determined by Medicare based on the 2018 4-phase Part D plan, which includes (1) an initial $405 deductible, (2) a standard coverage period until total drug costs reach $3750, (3) a coverage gap with 35% brand-name and 44% generic cost sharing until out-of-pocket costs reach $5000, and (4) catastrophic coverage with 5% cost sharing thereafter.14,15 Because Part D considers all other drug costs that contribute toward a patient’s deductible, the out-of-pocket cost per fill is an accurate reflection of what each beneficiary pays for sacubitril/valsartan. We standardized fill costs to 30 days for ease of comparison.

Statistical Analyses

We generated ASR-adjusted rates of sacubitril/valsartan prescribing by HRR in each year between 2016 and 2018 and then pooled across years. We used this to create heat maps of diffusion and to determine high-/low-prescribing HRRs. For the top- and bottom-prescribing HRRs in each year, we determined the mean (SD) out-of-pocket beneficiary co-payment per 30-day fill in each year. We then used multivariable logistic regression and a year-region indicator to explore how the impact of geographic region on likelihood of receiving sacubitril/valsartan changed over the 3 years of the study. All analyses were performed between May 2020 and July 2020 using SAS version 9.4 (SAS Institute).

RESULTS

In 2016, we identified 4182 Medicare beneficiaries who received sacubitril/valsartan. In 2017, we identified 7690 beneficiaries who received sacubitril/valsartan (an 83% relative increase from 2016), and by 2018 we identified 9594 beneficiaries who received sacubitril/valsartan (a 25% relative increase from 2017). The mean age of beneficiaries who received sacubitril/valsartan was 76 years, compared with 80 years among those who did not receive sacubitril/valsartan (Table 1). Sixty-five percent of beneficiaries who received sacubitril/valsartan were men, compared with 51% of those who did not. Geographically, 23% of all early sacubitril/valsartan use was in the South Atlantic region, followed by the Middle Atlantic region at 17%. More than 66% of sacubitril/valsartan use was in urban areas. Rates of HF hospitalizations in the year prior were similar between beneficiaries who received sacubitril/valsartan and those who did not (18.5% vs 18.9%). Rates of β-blocker use were similar between those who did and did not receive sacubitril/valsartan (69% vs 67%). More than 21% of patients who received sacubitril/valsartan were also receiving a mineralocorticoid receptor antagonist, compared with only 16% of those who did not receive sacubitril/valsartan. Among those who did not receive sacubitril/valsartan, 53% received an ACE inhibitor or angiotensin receptor blocker (ARB). Coronary artery disease was more common among those who received sacubitril/valsartan (91% vs 85%; P < .001), whereas chronic kidney disease was less common (66% vs 71%; P < .001).

There was substantial geographic variation in the ASR-adjusted rates of sacubitril/valsartan use among eligible beneficiaries with HFrEF between 2016 and 2018 (Figure 1). In 2016, there were only pockets of high use rates in the Mid-Atlantic and northern Texas/Oklahoma. In 2017, additional areas of high use could be seen in the Southeast and Appalachia. By 2018, use across the Northeast and South was high and the previously seen areas of high use in the Southwest, Southeast, and Appalachia spread to surrounding areas, illustrating the slow but steady diffusion.

In 2016, the highest ASR-adjusted use rate was in Bryan, Texas (ASR rate, 68.0 per 1000 eligible HFrEF beneficiaries) (Figure 2). The mean (SD) out-of-pocket cost per 30-day fill in the top 5 prescribing HRRs was $56.84 ($11.27). By comparison, the lowest (nonzero) prescribing HRR was Boston, Massachusetts (ASR rate, 7.9 per 1000 eligible HFrEF beneficiaries). The mean (SD) out-of-pocket cost per 30-day fill in the 5 lowest-prescribing HRRs was $66.02 ($19.37).

By 2018, the highest-prescribing HRR was Santa Rosa, California (ASR rate, 160.6 per 1000 eligible HFrEF beneficiaries, a 135% increase in use compared with the highest-prescribing HRR in 2016). The mean (SD) out-of-pocket cost per 30-day fill in the top 5 prescribing HRRs in 2018 was $56.57 ($11.77) (Figure 2). The lowest (nonzero) prescribing HRR was Madison, Wisconsin (ASR rate, 13.7 per 1000 eligible HFrEF beneficiaries, a 73% increase in use compared with the lowest-prescribing HRR in 2016). The mean (SD) out-of-pocket cost per 30-day fill in the 5lowest-prescribing HRRs was $75.09 ($14.93).

To examine the diffusion of sacubitril/valsartan over time and the impact of regional practice patterns on drug diffusion, we calculated the odds of receiving sacubitril/valsartan based on year and region in Table 2. Among eligible patients with HFrEF, the odds of receiving sacubitril/valsartan increased with each passing year, but the increase from 2016 to 2017 (odds ratio [OR], 1.70 in 2017 vs 2016; 95% CI, 1.64-1.76; P < .001) was larger than the increase from 2017 to 2018 (OR, 2.06 in 2018 vs 2016; 95% CI, 1.98-2.14; P < .001). Regionally, the lowest-prescribing region was the Midwest and the highest was the Northeast (OR, 1.45 vs Midwest; 95% CI, 1.40-1.50; P < .001).

In Table 3, we examine the impact of geographic region, by year, on sacubitril/valsartan diffusion, again after controlling for patient characteristics and out-of-pocket costs. Here we found that initially in 2016, there was no statistically significant difference between the use rate of sacubitril/valsartan in the Midwest (reference region and lowest-prescribing HRR) and the West (OR, 1.11; 95% CI, 1.00-1.22; P = .054). In the Northeast and South, use rates of sacubitril/valsartan were higher (OR, 1.52; 95% CI, 1.40-1.66; P < .001 in the Northeast; OR, 1.40; 95% CI, 1.30-1.51; P < .001 in the South). Between 2016 and 2017, the rate of sacubitril/valsartan use increased significantly in all regions, but the largest proportional increase was observed in the West (OR, 1.11 in 2016; OR, 2.12 in 2017) such that by the end of 2017, the rates of sacubitril/valsartan use were more similar across the Northeast (OR, 3.18), South (OR, 2.84), and West (OR, 2.58) compared with the rate in the reference region (Midwest 2016).

DISCUSSION

We detected significant geographic variation in the diffusion of sacubitril/valsartan since its approval in July 2015. Despite a class I recommendation encouraging providers to initiate it in eligible patients and switch patients already on ACE inhibitors/ARBs to it,3 the early rates of sacubitril/valsartan use among eligible beneficiaries with HFrEF were low. Although the rate of sacubitril/valsartan use among eligible beneficiaries with HFrEF increased 83% between 2016 and 2017, it increased only 25% between 2017 and 2018. Much of this early increase between 2016 and 2017 occurred in the West. Importantly, we found use rates of other guideline-directed medical therapies for HF, such as β-blockers and mineralocorticoid receptor antagonists, and preadmission ACE inhibitor/ARB use and postdischarge ACE inhibitor/ARB use in the non–sacubitril/valsartan group, similar to previously published rates among Medicare beneficiaries.16-18

On average, beneficiaries who received sacubitril/valsartan were younger and more likely to be male and White. This has been observed previously in work by Sangaralingham et al using commercial claims.19 This work extends the finding to Medicare beneficiaries with Part D and notes that, on average, those who received sacubitril/valsartan were younger and “healthier” than the average Medicare beneficiary with HFrEF.20 In addition, this work adds information about geographic prescribing patterns, finding that early sacubitril/valsartan use was concentrated in the Northeast and South, with areas of high use in Appalachia, Arizona, and the Mid-Atlantic.

Previous work by DeJong et al using Medicare formulary and pricing files to estimate out-of-pocket co-payments suggested that the high out-of-pocket costs, which they estimated to be $57 per 30-day fill, were likely a significant barrier to sacubitril/valsartan diffusion.14,21 In this study, we confirm and extend those findings. By using Part D data to directly measure the out-of-pocket costs per 30-day fill per beneficiary, we are able to calculate a more accurate assessment of out-of-pocket costs. Moreover, we are able to observe the difference in cost between high- and low-prescribing HRRs (pooled 3-year difference of $11.65 per 30-day fill) and note that the difference in out-of-pocket costs between high- and low-prescribing HRRs effectively doubled over the 3-year time period ($9.18 per 30-day fill in 2016 to $18.52 per 30-day fill in 2018). In addition, having 3 years of cost data allows us to observe that mean out-of-pocket costs in the high-prescribing HRRs did not change ($56.84 per 30-day fill in 2016 vs $56.57 per 30-day fill in 2018) but that the mean out-of-pocket costs in the low-prescribing HRRs increased approximately 14% over the same time period ($66.02 per 30-day fill in 2016 vs $75.09 per 30-day fill in 2018).

Importantly, however, even after controlling for cost, geographic prescribing patterns remained significantly associated with likelihood of receiving sacubitril/valsartan. Beneficiaries in the Northeast and South were 26% to 45% more likely to receive sacubitril/valsartan than beneficiaries living in the Midwest between 2016 and 2018. We also found that adoption lagged in the West in 2016, but between 2016 and 2017, the rate of increase in the West outpaced all other regions, such that by the end of 2017, the odds of receiving sacubitril/valsartan were much more similar in the Northeast, South, and West. Use rates in the Midwest remained the lowest.

Many possible explanations exist for these significantly different regional trends. First, as the cost/co-payment data from this study and others demonstrate, regions with lower use rates have higher co-payments, and high co-payments are known to be associated with slower drug diffusion.22 However, even after controlling for co-payment, we found significant variation in sacubitril/valsartan use by geographic region. Second, it is possible that the pharmaceutical marketing efforts were more heavily targeted toward one region vs another. For example, after seeing unexpected low use rates in the West in 2016, given the population of eligible beneficiaries in that area, it is possible that enhanced marketing efforts directed at the West explain some of the large increase in sacubitril/valsartan use between 2016 and 2017.

Next, we know from CDC data that the average patient with HFrEF in the Midwest tends to be “healthier,” defined by lower rates of comorbidities, hospitalizations, and HF mortality, compared with other regions.23 These “healthier” patients might exacerbate provider inertia, making them more reluctant to risk a transition in therapy to a new, but more expensive, drug in a stable patient. We also know from prior work that provider familiarity with a drug is associated with increased use.8,24,25 It is therefore possible that more providers in the Northeast and South were exposed to sacubitril/valsartan earlier—perhaps in clinical trials—such that by the time it was FDA approved, they were more comfortable initiating it. Relatedly, prior work examining the diffusion of similarly expensive, novel antidiabetic agents found that much of the early use was driven by a few high-prescribing providers.8 Given the overall low rates of use of sacubitril/valsartan, a few aggressive prescribers in a geographic area could increase the overall rate of use in that area and, in doing so, expose their colleagues to the novel therapy, simultaneously decreasing regional provider inertia.

Limitations

The primary limitation of this study is the use of claims data and their lack of clinical granularity. Therefore, we are not able to comment on appropriateness of drug use. Second, this study is limited to FFS Medicare beneficiaries with Part D pharmaceutical coverage. Caution should be used in extrapolating these results to other populations. Third, we do not have access to data regarding the marketing efforts of pharmaceutical companies and where they deployed advertising resources over this time period. Fourth, we have information about only filled prescriptions. We do not know how many prescriptions were written for sacubitril/valsartan and never filled due to cost or other prohibitive factors. Although we expect that reasons for not filling a new prescription for sacubitril/valsartan should be randomly distributed, if a specific reason was more common in a given area, this could introduce bias. Because approximately 25% of patients who received sacubitril/valsartan only had 1 fill, we compared baseline characteristics between beneficiaries who received only 1 fill and more than 1 fill and found that there are clinically significant differences (eAppendix B). Finally, clinical contraindications to sacubitril/valsartan are poorly coded in claims data. Therefore, we were unable to exclude these patients. This may cause the rate of uptake to appear lower than expected, but because only a small number of individuals have true clinical contraindications to sacubitril/valsartan and/or ACE inhibitors/ARBs, we would not expect this to alter the study’s findings.

CONCLUSIONS

Between 2016 and 2018, the overall rate of sacubitril/valsartan use among eligible Medicare beneficiaries with HFrEF was low and highly variable. Areas with low use rates had higher co-payments than areas with high rates of early use. However, even after controlling for costs, geographic prescribing patterns had a significant impact on drug diffusion. Most early use was concentrated in the Northeast and South. The rate of increase in the West between 2016 and 2018 outpaced that of other regions such that by the end of 2018, use rates were more similar in the Northeast, South, and West, with the highest rates observed in the Northeast and the lowest in the Midwest. This variation highlights opportunities to expand the use of sacubitril/valsartan among eligible beneficiaries with HFrEF and highlights the importance of local prescribing patterns in managing the diffusion of novel therapies into clinical practice.

Author Affiliations: Heart and Vascular Center, Dartmouth-Hitchcock Medical Center (LG), Lebanon, NH; The Dartmouth Institute for Health Policy and Clinical Practice, Geisel School of Medicine at Dartmouth (LG, AK, ST, AMA), Hanover, NH.

Source of Funding: This work is supported by the National Heart Lung and Blood Institute, grant K23HL142835.

Author Disclosures: The authors report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.

Authorship Information: Concept and design (LG, AMA); acquisition of data (LG, ST); analysis and interpretation of data (AK, AMA); drafting of the manuscript (LG, AK, ST); critical revision of the manuscript for important intellectual content (LG, ST); statistical analysis (LG, AK, AMA); provision of patients or study materials (LG); obtaining funding (LG); administrative, technical, or logistic support (LG, AK, ST); and supervision (ST).

Address Correspondence to: Lauren Gilstrap, MD, MPH, Dartmouth-Hitchcock Medical Center, 1 Medical Center Rd, Lebanon, NH 03766. Email: lauren.g.gilstrap@dartmouth.edu.

REFERENCES

1. Fala L. Entresto (sacubitril/valsartan): first-in-class angiotensin receptor neprilysin inhibitor FDA approved for patients with heart failure. Am Health Drug Benefits. 2015;8(6):330-334.

2. McMurray JJ, Packer M, Desai AS, et al; PARADIGM-HF Investigators and Committees. Angiotensin-neprilysin inhibition versus enalapril in heart failure. N Engl J Med. 2014;371(11):993-1004. doi:10.1056/NEJMoa1409077

3. Yancy CW, Jessup M, Bozkurt B, et al; Writing Committee Members. 2016 ACC/AHA/HFSA focused update on new pharmacological therapy for heart failure: an update of the 2013 ACCF/AHA guideline for the management of heart failure: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines and the Heart Failure Society of America. Circulation. 2016;134(13):e282-e293. doi:10.1161/CIR.0000000000000435

4. Luo N, Fonarow GC, Lippmann SJ, et al. Early adoption of sacubitril/valsartan for patients with heart failure with reduced ejection fraction: insights from Get With the Guidelines–Heart Failure (GWTG-HF). JACC Heart Fail. 2017;5(4):305-309. doi:10.1016/j.jchf.2016.12.018

5. Fonarow GC, Navar AM, Yancy CW. Impediments to implementing guideline-directed medical therapies. JAMA Cardiol. 2019;4(8):830-831. doi:10.1001/jamacardio.2019.2276

6. DeJong C, Kazi DS, Dudley RA, Chen R, Tseng CW. Assessment of national coverage and out-of-pocket costs for sacubitril/valsartan under Medicare Part D. JAMA Cardiol. 2019;4(8):828-830. doi:10.1001/jamacardio.2019.2223

7. Sumarsono A, Vaduganathan M, Ajufo E, et al. Contemporary patterns of Medicare and Medicaid utilization and associated spending on sacubitril/valsartan and ivabradine in heart failure. JAMA Cardiol. 2019;5(3):336-339. doi:10.1001/jamacardio.2019.4982

8. Gilstrap LG, Blair RA, Huskamp HA, Zelevinsky K, Normand SL. Assessment of second-generation diabetes medication initiation among Medicare enrollees from 2007 to 2015. JAMA Netw Open. 2020;3(5):e205411. doi:10.1001/jamanetworkopen.2020.5411

9. Bui AL, Horwich TB, Fonarow GC. Epidemiology and risk profile of heart failure. Nat Rev Cardiol. 2011;8(1):30-41. doi:10.1038/nrcardio.2010.165

10. Census regions and divisions of the United States. US Census Bureau. Accessed July 1, 2018. https://www2.census.gov/geo/pdfs/maps-data/maps/reference/us_regdiv.pdf

11. Rural-urban commuting area codes. United States Department of Agriculture.Updated August 17, 2020. Accessed June 1, 2020. https://www.ers.usda.gov/data-products/rural-urban-commuting-area-codes.aspx

12. American Community Survey (ACS). US Census Bureau. Accessed September 7, 2018. https://www.census.gov/programs-surveys/acs/

13. Chronic Conditions Data Warehouse. Accessed June 15, 2020. https://www2.ccwdata.org/web/guest/home/

14. DeJong C, Kazi DS, Dudley RA, Chen R, Tseng CW. Assessment of national coverage and out-of-pocket costs for sacubitril/valsartan under Medicare Part D. JAMA Cardiol. 2019;4(8):828-830. doi:10.1001/jamacardio.2019.2223

15. An overview of the Medicare Part D prescription drug benefit. Kaiser Family Foundation. Updated October 13, 2021. Accessed February 22, 2021. https://www.kff.org/medicare/fact-sheet/an-overview-of-the-medicare-part-d-prescription-drug-benefit/

16. Gilstrap L, Skinner JS, Gladders B, et al. Opportunities and challenges of claims-based quality assessment: the case of postdischarge beta-blocker treatment in patients with heart failure with reduced ejection fraction. Circ Cardiovasc Qual Outcomes. 2020;13(3):e006180. doi:10.1161/CIRCOUTCOMES.119.006180

17. Loop MS, van Dyke MK, Chen L, et al. Low utilization of beta-blockers among Medicare beneficiaries hospitalized for heart failure with reduced ejection fraction. J Card Fail. 2019;25(5):343-351. doi:10.1016/j.cardfail.2018.10.005

18. Inampudi C, Parvataneni S, Morgan CJ, et al. Spironolactone use and higher hospital readmission for Medicare beneficiaries with heart failure, left ventricular ejection fraction <45%, and estimated glomerular filtration rate <45 ml/min/1.73 m2. Am J Cardiol. 2014;114(1):79-82. doi:10.1016/j.amjcard.2014.03.062

19. Sangaralingham LR, Sangaralingham SJ, Shah ND, Yao X, Dunlay SM. Adoption of sacubitril/valsartan for the management of patients with heart failure. Circ Heart Fail. 2018;11(2):e004302. doi:10.1161/CIRCHEARTFAILURE.117.004302

20. Khera R, Pandey A, Ayers CR, et al. Contemporary epidemiology of heart failure in fee-for-service Medicare beneficiaries across healthcare settings. Circ Heart Fail. 2017;10(11):e004402. doi:10.1161/CIRCHEARTFAILURE.117.004402

21. Prescription drug plan formulary, pharmacy network, and pricing information files. CMS. Updated January 5, 2021. Accessed March 15, 2021. https://www.cms.gov/Research-Statistics-Data-and-Systems/Files-for-Order/NonIdentifiableDataFiles/PrescriptionDrugPlanFormularyPharmacyNetworkandPricingInformationFiles

22. Gilstrap LG, Mehrotra A, Bai B, Rose S, Blair RA, Chernew ME. National rates of initiation and intensification of antidiabetic therapy among patients with commercial insurance. Diabetes Care. 2018;41(8):1776-1782. doi:10.2337/dc17-2585

23. Interactive Atlas of Heart Disease and Stroke. CDC. March 13, 2020. Accessed July 2, 2020. https://www.cdc.gov/dhdsp/maps/atlas/index.htm

24. Anderson TS, Lo-Ciganic WH, Gellad WF, et al. Patterns and predictors of physician adoption of new cardiovascular drugs. Healthc (Amst). 2018;6(1):33-40. doi:10.1016/j.hjdsi.2017.09.004

25. Lublóy Á. Factors affecting the uptake of new medicines: a systematic literature review. BMC Health Serv Res. 2014;14:469. doi:10.1186/1472-6963-14-469