The Effectiveness of a Comprehensive Wellness Assessment on Medication Adherence in a Medicare Advantage Plan Diabetic Population : Journal of Healthcare Management

Secondary Logo

Journal Logo

RESEARCH ARTICLES

The Effectiveness of a Comprehensive Wellness Assessment on Medication Adherence in a Medicare Advantage Plan Diabetic Population

Guerard, Barbara DSc; Omachonu, Vincent PhD; Perez, Blake; Sen, Bisakha PhD

Author Information
Journal of Healthcare Management: March-April 2018 - Volume 63 - Issue 2 - p 132-141
doi: 10.1097/JHM-D-16-00034
  • Free

Abstract

INTRODUCTION

The objective of this study was to assess whether a comprehensive wellness assessment (CWA) program helps improve medication adherence for oral diabetic medications (ODMs), statins, and angiotensin-converting enzyme inhibitors or angiotensin II receptor blockers (ACE/ARBs) in the Medicare Advantage (MA) plan diabetic population.

The issue of medication nonadherence has generated significant interest because of its complexity from both cost and outcomes perspectives. Estimates indicate that half of the 3.2 billion prescriptions written annually in the United States are not taken as prescribed, especially among patients with asymptomatic chronic conditions (Zolnierek & Dimatteo, 2009). By some estimates, medication nonadherence accounts for approximately $100 billion in annual healthcare spending due to hospital admissions that are medication related (McDonnell & Jacobs, 2002; Patel, Piette, Resnicow, Kowalski-Dobson, & Heisler, 2016; Roebuck, Liberman, Gemmill-Toyama, & Brennan, 2011). With the need for overall healthcare cost reduction an ongoing challenge, the impact on Medicare beneficiaries' clinical outcomes of medication nonadherence, though difficult to quantify, must be ascertained.

As part of the goal to improve beneficiaries' outcomes, the Centers for Medicare & Medicaid Services (CMS) uses a five-star quality rating system to measure Medicare beneficiaries' experience with their health plans and the overall U.S. healthcare system. CMS included adherence to ODMs, statins, and ACE/ARBs among the triple-weighted measures informing the quality-rating program for MA plans. The measures are based on the percentage of members who fill prescriptions often enough to cover at least 80% of the days they should be taking the medication. This metric is referred to as the proportion of days covered. CMS's assignment of a triple weight to these measures underscores the importance to MA plans of investigating policies and interventions that improve adherence in these classes of medications.

Previous research has found that health plan initiatives, such as 90-day prescriptions, copayments of less than $10, and mail-order prescription programs, contribute to improved adherence in diabetic patients (Schmittdiel, et al., 2015). Previous research by the current authors using this MA plan's data found that adherence to ODMs among dual Medicare- and Medicaid-eligible Special Needs Plan (SNP) enrollees improved after undergoing a CWA (Guerard, Omachonu, Perez, & Sen, 2016). Details of the CWA are provided in the authors' previous research. Briefly, the main element is a detailed evaluation performed by a nurse practitioner (NP), which also affords the enrollee the opportunity to speak with a licensed clinical social worker and a clinical pharmacist. The CWA includes a thorough medication review and education on the importance of medication adherence. CWAs may be completed in any of a variety of settings on the basis of the enrollees' preference, such as the enrollee's home, one of the organization's service center locations, or in the enrollee's primary care office. The overarching goals are to improve patient outcomes and keep the enrollee active and informed about his or her own health management.

This study builds on the previous research of Guerard et al. (2016). Specifically, it considers the association between CWA and medication adherence for both an organization's entire diabetic population and segments of that population as stratifies according to various characteristics.

METHODS AND EMPIRICAL APPROACH

The authors conducted a retrospective panel study using administrative claims data and member month-level data. The study data cover 2010–2015. The researchers restricted their sample to MA members who received a new diagnosis of diabetes since 2010, thereby allowing for a maximum follow-up period of 5 years. The intervention of interest is whether the member received a CWA in the preceding 12 months of the study, which is operationalized as a binary outcome of 1 if the member underwent a CWA anytime within the past 12 months and 0 otherwise.

The outcomes of interest were monthly adherence to ODMs, ACE/ARBs, and statins. For each medication, monthly adherence was operationalized using a binary indicator of 1 if prescription fills for that medication covered at least 80% of the days in the month and 0 otherwise. Monthly adherence was calculated using a metric that closely parallels the metric used by CMS to calculate annual adherence. Specifically:

  1. Only the member months following the month of the first fill of the year were included.
  2. The researchers assumed that the member received appropriate medications during acute inpatient or observation days, and these days were subtracted from the numerator and denominator. Member months that were completely associated with acute inpatient and/or observation stay days were excluded from the measure.
  3. In the case of ODMs, all member months were excluded for years during which a member filed any claim for a fill of insulin, per the CMS approach for calculating ODMs adherence.

Multivariable Poisson models estimated the binary indicator of “adherent” for each medication. The Poisson model was preferred over the more conventional logit model, as the researchers were interested in how a CWA influences relative risk of adherence for SNP enrollees. Odds ratios from logit models provide a poor approximation of relative risk ratios except for situations in which the outcome of interest is rare (McNutt, Wu, Xue, & Hafner, 2003). Misinterpretation of odds ratios as risk ratios have sometimes led to substantial distortions of scientific findings (Schwartz, Woloshin, & Welch, 1999). When the outcome of interest is relatively common, the Poisson model provides a more accurate estimate of the relative risk ratio than the logit model does (McNutt et al., 2003). Additional controls included in the multivariate Poisson models were SNP status, gender, race, age, and the presence of chronic conditions frequently associated with diabetes, including end-stage renal disease, chronic obstructive pulmonary disease, congestive heart failure, hypertension, chronic kidney disease, and cancer.

Further models were estimated after stratifying the sample using specific member characteristics. These included SNP versus non-SNP status, presence of three or more chronic diseases versus fewer than three, and age group (≤65 years, 66–80 years, and >80 years). Following the general conventions in the health science and health service research literature, statistical was set at a level of .05, which corresponds to a t-test statistic of 1.96.

Results are reported as an incidence rate ratio (IRR). An IRR higher than 1 indicates a greater likelihood of attaining the outcome of being adherent, while an IRR of less than 1 indicates a lower likelihood of adherence.

RESULTS

Descriptive statistics are presented in Table 1.These measures are based on pooled member-month data, so each member has multiple observations recorded. Results reveal that the proportion of observations whereby the member underwent a CWA in the previous 12 months is 13%. The proportion of observations from SNP members is 27%, from minority members is 46%, and from female members is 56%. The average age in the pooled sample is 68.49 years.

T1-11
Table 1:
Descriptive Statistics for Intervention and Control Variables

Table 2 shows the rate of monthly adherence for members on statins, ACE/ARBs, or ODMs medication. These statistics are presented for the full sample as well as by SNP status, chronic condition, and age group. Rate of adherence to statins is 69%, to ACE/ARBs is 73%, and to ODMs is 71%. Non-SNP members have slightly higher rates of adherence than SNP members, those with three or more chronic conditions have marginally higher adherence rates than those with fewer than three chronic conditions, and older members show somewhat higher adherence rates than younger members.

T2-11
Table 2:
Rate of Monthly Adherence to Statins, ACE/ARBs, and Oral Diabetes Medications for Full Sample and Subsamples

Multivariate Poisson results are presented in Tables 3–5. Results for the full sample show that a CWA visit in the previous 12 months is significantly associated with greater adherence to statin medication than absence of such a visit (IRR: 1.022, t-test: 2.51) and oral diabetes medication (IRR: 1.032, t-test: 3.00); a CWA visit is not significantly associated with adherence to ACE/ARB medication (IRR: 1.009, t-test: 1.09). Results from analysis of the stratified population segments reveal noticeable variations among the subgroups. When stratifying by SNP status, the authors found that a CWA is significantly associated with higher adherence among SNP members for statin medication (IRR: 1.056, t-test: 4.0), ACE/ARBs (IRR: 1.034, t-test: 2.69), and oral diabetes medication (IRR: 1.068, t-test: 4.21). In no cases did a CWA have a statistically significant association with better medication adherence among non-SNP members. Results from stratification by chronic disease and by age provide mixed results across medication types. For example, among those members with three or more chronic diseases, a CWA is significantly associated with improved adherence to statin medication (IRR: 1.031, t-test: 2.87) but not for the other medications. Among members under age 65, CWA is also significantly associated with higher adherence to statin medication (IRR: 1.046, t-test: 2.37) and oral diabetes medication (IRR: 1.051, t-test: 2.25). However, researchers uncovered less evidence of a statistical relationship between a CWA and better medication adherence for members who are 65 to 80 years old or who are 80 years or older.

T3-11
Table 3:
Multivariate Poisson Results for Association Between Adherence to Statins and CWA Visit in Past 12 Months, Full Sample and Subsamples
T4-11
Table 4:
Multivariate Poisson Results for Association Between Adherence to ACE/ARBs and CWA Visit in Past 12 Months, Full Sample and Subsamples
T5-11
Table 5:
Multivariate Poisson Results for Association Between Adherence to Oral Diabetes Medication and CWA Visit in Past 12 Months, Full Sample and Subsamples

We conducted further sensitivity analyses using a CWA visit in the prior 0 to 3 months, 3 to 6 months, and 6 to 12 months as the indicator of interest in place of the single indicator. Results remained consistent in terms of which subgroups showed the strongest associations between a CWA and improved adherence.

DISCUSSION AND RECOMMENDATIONS

Physician office visits are often short in length and may not give the provider the opportunity to discuss medication adherence issues. As a result, the prescribing clinicians may not be aware of their patient's medication adherence status. Nonadherence to medications is a common problem in clinical practice. This issue is especially prevalent among people with asymptomatic chronic conditions such as hypertension, diabetes, and hyperlipidemia (Pladevall, et al., 2004). Other studies have demonstrated that interventions such as care coordination; education; and collaborative, team-based care for diabetic and hyperlipidemia patients resulted in increased medication adherence(Viswanathan et al., 2012).

In addition to performing a complete member assessment, the opportunity for enrollees of this MA plan to undergo a CWA is intended to facilitate, through a 1-hour visit, an atmosphere in which patients feel comfortable enough to report their concerns about and side effects from taking certain medications. The study's subsample analyses show that CWAs are most strongly associated with improvements in adherence for SNP members as well as MA members under the age of 65—who are typically eligible for this program due to disability status. While the causal link is not obvious, the authors offer some speculations. SNP patients are less likely than non-SNP members to have been insured or to have had a regular source of care before they achieved dual-eligible status. This population may have cognitive challenges and thus may benefit from detailed guidance about medication adherence and the importance of self-engagement in their own health, which is provided during the CWA. In addition, a higher likelihood of race concordance may be at play between SNP enrollees and the NPs or clinical social workers administering the CWA (Schoenthaler, Schwartz, Wood, & Stewart, 2012). In general, evidence indicates that patients may be more satisfied with practitioner interactions when interacting with an NP than with a physician, which may be due to perceived improved communication and comprehension leading to increased adherence (Roblin, Becker, Adams, Howard, & Roberts, 2004). Everett et al. (2013) report that older patients with diabetes on panels with physician assistants or NPs experienced the same or better results for most outcomes than those with physician-only care.

This team-based approach of including the organizations' NPs has demonstrated encouraging results in enrollee satisfaction and enhanced education for members as well as improvement in medication adherence rates among the most vulnerable populations. On the basis of these findings, the most strategic and cost-effective approach may be to target this intervention, as it relates to medication adherence, to the most vulnerable populations rather than attempt to target the full MA membership.

The CWA program implementation was one component of the study organizations' redesign from a traditional utilization management structure to a team-based, patient-centered medical home model. Multilevel factors could have affected a successful implementation. These included, but were not limited to, organizational factors related to market forces, the healthcare environment in which the organizations operate, reimbursement for MA plans, and an overall goal of healthcare cost reduction. Provider- and enrollee-level factors also require consideration. The organizations' market is in an area in which the NP's role with providers is not as mature as in other areas of the country. Enrollee-level factors such as patient characteristics, health-related beliefs and motivation, and a general distrust of the healthcare system were present. While the organization consistently demonstrate a culture of innovation, a more rigorous analytic approach to assessing these factors is necessary as the healthcare environment becomes increasingly complex. Spaulding, Kash, Johnson, and Gamm (2015) propose a powerful approach involving the use of a valid and reliable tool to assess an organization's capacity for major change. This tool could prove valuable in the context of our study.

This study has some limitations. It uses the organizations' claims data, which rely on providers to submit complete claims. The sample was confined to enrollees who have been with the MA plan for at least 5 years following their diabetes diagnosis and may not be generalized to members who stay in the plan for shorter periods of time. Some self-selection as to which enrollees actually choose to undergo a CWA may have occurred, which may have led to omitted variable bias in the results indicating an association between CWA and adherence. Therefore, caution must be exercised when making causal inferences. Furthermore, these results pertain to one MA plan in the Southeast region of the United States, and its results may not be generalizable across all MA plans. Finally, this study only considers the association of CWA with medication adherence for medications that are triple weighted in CMS quality program measures. While not the focus of this study, additional research is required to provide a comprehensive picture of the overall effectiveness and cost benefit of programs such as CWAs.

CONCLUSION

Identifying methods of improving patients' medication adherence for chronic conditions remains a complex issue. The research in this field needs to be continued, as previous studies do not reveal consistent improvements in either interventions to improve adherence or clinical outcomes (Nieuwlaat et al., 2014). Further study on the effectiveness of the CWA, analyzing other expected outcomes, and using the entire population of enrollees is called for in directing future strategic initiatives.

References

Everett, C., Thorpe, C., Palta, M., Carayon, P., Bartels, C., & Smith, M. A. (2013). Physician assistants and nurse practitioners perform effective roles on teams caring for Medicare patients with diabetes. Health Affairs, 32(11), 1942–1948.
Guerard, B., Omachonu, V., Perez, B., & Sen, B. (2016). Improvement in oral diabetes medication adherence after a comprehensive wellness assessment in a dual eligible population. Population Health Management, 19(6), 405–413.
McDonnell, P. J., & Jacobs, M. R. (2002). Hospital admissions resulting from preventable adverse drug reactions. Annals of Pharmacotherapy, 36(9), 1331–1336.
McNutt, L.-A., Wu, C., Xue, X., & Hafner, J. P. (2003). Estimating the relative risk in cohort studies and clinical trials of common outcomes. American Journal of Epidemiology, 157(10), 940–943.
Nieuwlaat, R., Wilczynski, N., Navarro, T., Hobson, N., Jeffery, R., Keepanasseril, A., & Jack, S. (2014). Interventions for enhancing medication adherence. Cochrane Database of Systematic Reviews, 11. doi:10.1002/14651858.CD000011.pub4
Patel, M. R., Piette, J. D., Resnicow, K., Kowalski-Dobson, T., & Heisler, M. (2016). Social determinants of health, cost-related non-adherence, and cost-reducing behaviors among adults with diabetes: Findings from the National Health Interview Survey. Medical Care, 54(8), 796–803. doi:10.1097/MLR.0000000000000565
Pladevall, M., Williams, L. K., Potts, L. A., Divine, G., Xi, H., & Lafata, J. E. (2004). Clinical outcomes and adherence to medications measured by claims data in patients with diabetes. Diabetes Care, 27(12), 2800–2805.
Roblin, D. W., Becker, E. R., Adams, E. K., Howard, D. H., & Roberts, M. H. (2004). Patient satisfaction with primary care: Does type of practitioner matter? Medical Care, 42(6), 579–590.
Roebuck, M. C., Liberman, J. N., Gemmill-Toyama, M., & Brennan, T. A. (2011). Medication adherence leads to lower health care use and costs despite increased drug spending. Health Affairs (Millwood), 30(1), 91–99.
Schmittdiel, J. A., Nichols, G. A., Dyer, W., Steiner, J. F., Karter, A. J., & Raebel, M. A. (2015). Health care system-level factors associated with performance on Medicare STAR adherence metrics in a large, integrated delivery system. Medical Care, 53(4), 332.
Schoenthaler, A. M., Schwartz, B. S., Wood, C., & Stewart, W. F. (2012). Patient and physician factors associated with adherence to diabetes medications. The Diabetes Educator, 38(3), 397–408.
Schwartz, L. M., Woloshin, S., & Welch, H. G. (1999). Misunderstandings about the effects of race and sex on physicians' referrals for cardiac catheterization. The New England Journal of Medicine, 341(4), 279.
Spaulding, A., Kash, B. A., Johnson, C. E., & Gamm, L. (2015). Organizational capacity for change in health care: Development and validation of a scale. Health Care Management Review, 42(2), 151–161.
Viswanathan, M., Golin, C. E., Jones, C. D., Ashok, M., Blalock, S. J., Wines, R. C., & Lohr, K. N. (2012). Interventions to improve adherence to self-administered medications for chronic diseases in the United States: A systematic review. Annals of Internal Medicine, 157(11), 785–795.
Zolnierek, K. B., & Dimatteo, M. R. (2009). Physician communication and patient adherence to treatment: A meta-analysis. Medical Care, 47(8), 826–834.
© 2018 Foundation of the American College of Healthcare Executives