Timothy M. Dall, MS

Medicare Part D is a prescription drug program that provides seniors and disabled individuals enrolled in Medicare with outpatient drug coverage benefits. Part D has been shown to increase access to medicines and improve medication adherence; however, the effect of Part D on health outcomes has not yet been extensively studied. In this study, we used a published and validated Markov-based microsimulation model to quantify the relationships among medication use, disease incidence and severity, and mortality. Based on the simulation results, we estimate that since the implementation of Part D in 2006, nearly 200,000 Medicare beneficiaries have lived at least 1 year longer. Reductions in mortality have occurred because of fewer deaths associated with medication-sensitive conditions such as diabetes, congestive heart failure, stroke, and myocardial infarction. Improved access to medication through Medicare Part D helps patients improve blood pressure, cholesterol, and blood glucose levels, which in turn can prevent or delay the onset of disease and the incidence of adverse health events, thus reducing mortality.

In 2003, Medicare Part D, a prescription drug program providing outpatient drug coverage benefits to seniors and disabled individuals enrolled in Medicare, was signed into law. A large body of literature documents the positive impact that Part D has had on increasing access to medicines and improving medication adherence; however, the effect of Part D on health outcomes has not yet been extensively studied.

While life expectancy among the population 65 years and older has increased since 2006,

 4 studies that have examined Part D’s impact on mortality report mixed findings.

 Kaestner et al and Briesacher et al found no statistically significant association between Part D and mortality.

 Huh and Reif estimated that Part D reduced overall mortality among beneficiaries by 2.2% annually, with the effect driven primarily by a decrease in cardiovascular-related deaths.

 Similarly, Dunn and Shapiro found that Part D reduced cardiovascular-related mortality, but that mortality associated with noncardiovascular disease remained unchanged.

 Dunn and Shapiro estimate that between 21,800 and 25,500 individuals were still alive in mid-2007 because of Part D’s implementation in 2006. These 4 studies were limited to a narrow time frame of analysis or made strong assumptions regarding the continuation of historical trends pre- and post Part D. The goal of this study is to further explore Part D’s effect on mortality from 2006 to the present.

This study used a published and validated Markov-based microsimulation model to quantify the relationships among medication use, disease incidence and severity, and mortality. Information on the model design, data, assumptions, methods, and validation results have been published elsewhere.

 A microsimulation approach was chosen based on its ability to model changes in health and disease states longitudinally based on individual demographics (age, sex, race, ethnicity), biometrics (body mass index [BMI], systolic blood pressure [SBP], diastolic blood pressure [DBP], total cholesterol, high-density lipoprotein cholesterol [HDL-C], and glycated hemoglobin [A1C]), smoking status, and presence of approximately 30 diseases to predict mortality as an individual ages. The model incorporates the expected impact of medication use on changes in blood pressure, cholesterol, and A1C levels, and simulates the resulting implications on disease incidence and severity that in turn impact the cause and timing of beneficiary mortality.

Population We created a representative sample of the Medicare population using data from the 2005 to 2012 National Health and Nutrition Examination Survey (NHANES) files.

 The combined files contain data for 3621 Medicare beneficiaries 65 years and older, including demographic and biometric information required for use in the microsimulation model. Using NHANES sample weights to dictate the probability of selection, we used Monte Carlo simulation, which involves repeated random sampling of the population, to produce a nationally representative simulation sample of 100,000 observations representative of the estimated 22.5 million beneficiaries enrolled in Medicare Part D in 2006.

 We added 20,000 additional observations in each year of the simulation to represent new enrollees in Part D. This annual number of observations added to the simulation population reflected that net enrollment in Part D grew by approximately 6.5% annually between 2006 and 2014 after factoring in mortality, and the assumption that once enrolled in Part D beneficiaries did not drop Part D coverage unless death occurred.

Using this model, we simulated outcomes for this population under 2 scenarios over the 2006 to 2014 period:

An estimated 25% of Part D enrollees would not have had comprehensive drug coverage in the absence of Part D.

 Beneficiaries who gained comprehensive coverage under Part D increased their use of medication.

 For modeling purposes, we used findings from Zhang et al, which showed that drug possession rates increased by 29%, 31%, and 22%, respectively, for people diagnosed with hyperlipidemia, diabetes, and hypertension.

In this scenario, there were no changes in drug coverage among the Medicare population. As a result, there were no changes in medication use, and natural progression of disease and mortality was maintained.

Results for both scenarios were then scaled to simulate total Part D enrollment.

For each beneficiary, the previous year’s health status was used to predict the current year’s outcomes, with this process repeated through 2014 or until death. Prediction equations dictating the transition probabilities for various cardiovascular conditions (congestive heart failure [CHF], ischemic heart disease, stroke, myocardial infarction [MI]), diabetes, chronic kidney disease, renal failure, and other diseases modeled were obtained from published studies including the Framingham Heart Study, the United Kingdom Prospective Diabetes Study, and various clinical trials and observational studies.

 Annual changes in body weight were estimated based on national trends by age, sex, and current weight.

 Annual changes in DBP, SBP, HDL-C and total cholesterol were determined based on each beneficiary’s age, sex, and change in BMI,

 while changes in A1C were predicted based on BMI, age, and total cholesterol (if beneficiary did not have diabetes) and years since diagnosis and previous year’s A1C (if diabetic).

Annual mortality rates associated with the chronic diseases modeled, acute events such as MI and stroke, and “all other causes” were estimated from published sources and based on demographics, biometrics, smoking status, and presence of adverse medical conditions.

To simulate the impact of gaining comprehensive drug coverage under the Part D scenario, enrollees with high cholesterol (total cholesterol >240 mg/dL), high blood pressure (SBP >140 mm Hg or DBP ≥90 mm Hg), and high A1C levels (≥6.5%) were assumed to increase their medication use, resulting in improvements on these measurements. We performed sensitivity analyses using published 95% CIs for key parameters used in the model. We modeled that statins would reduce total blood cholesterol by 34.42 mg/dL (CI, 22.04-46.40), and anti-hypertensives would reduce SBP by 14.5 mm Hg (CI, 14.2-14.8) and DBP by 10.7 mm Hg (CI, 10.5-10.8).

 These reductions were assumed to occur entirely within the first year of therapy. Oral diabetes medications were assumed to reduce A1C by 1 percentage point (CI, 0.5-1.25) annually across every year of the analysis until diabetes control was reached at A1C of 7.5%.

 (Most patients with high A1C levels see their A1C levels fall to 7.5% or lower within 1 year.) Medication use and drug effect were assumed to remain unchanged in the status quo scenario.

A nationally representative sample of Medicare beneficiaries 65 years and older served as the baseline simulation population in this study. The initial (2006) sample was 55.5% female, with an average BMI of 28.4, and an average age of 73.6 years (

). Controlled or uncontrolled hypertension (74.4%) and hypercholesterolemia (58.6%), prediabetes (60.0%), and diabetes (25.1%) were highly prevalent among this population. The CDC reports similar characteristics for Medicare beneficiaries in 2006, but reports a greater proportion of individuals 85 years and older (14.0% vs 2.2% in our sample) and a lower prevalence of hypertension (62.8%).

In the status quo scenario, the prevalence of uncontrolled hypertension remained fairly constant at about 41% from 2006 to 2014. We observed a lower prevalence under the Part D scenario (ranging from 38.8% to 39.6%), which translates into an estimated 550,000 fewer beneficiaries with uncontrolled hypertension between 2006 and 2014 (

). Cumulative over this period, there were 180,000 fewer new cases of CHF, 210,000 fewer MIs, and 830,000 more people with tightly controlled diabetes (defined by A1C level <7%).

The results of the simulation suggest that about 198,800 Part D beneficiaries lived at least 1 year longer following the implementation of the program. The average increase in longevity was 3.3 years. On average, approximately 22,100 lives were saved annually between 2006 and 2014. The number of lives saved each year increased from 2006 to 2009, but declined in the later years as deaths that might have occurred sooner eventually occurred in later years (

Mortality reductions were primarily attributable to fewer deaths from medication-sensitive conditions, particularly diabetes and cardiovascular diseases (

). We estimate that between 2006 and 2014, total deaths associated with diabetes declined by 100,400, deaths associated with CHF declined by 53,100, deaths from stroke declined by 40,400, and deaths from MI declined by 25,100.

These mortality reductions may potentially be offset by complications arising from medications or polypharmacy in the population examined. While the classes of medications modeled in our analysis do pose risk for adverse events (eg, muscle toxicity and increased liver enzymes from statin use; hypotension, in general, from antihypertensive use; gastrointestinal issues from oral diabetes medication use), it is widely accepted that safety profiles are good for these medications and the drugs are well tolerated.

 The implicit modeling assumption is that patients experiencing complications from medication will work with their doctors to modify their medication regimens.

We conducted several sensitivity analyses in which we varied our assumptions about the impact of medication use among beneficiaries gaining comprehensive drug coverage as a result of Part D. Improving A1C levels by 0.5 to 1.25 percentage points while holding all else constant resulted in a range of approximately 181,200 to 205,000 lives saved over 9 years. Similar analyses varying the efficacy of lipid lowering drugs showed a range of approximately 196,400 to 201,400 lives saved, while the estimated range for antihypertensive drugs was 196,600 to 199,900 lives saved over the same 9-year period.

Our analysis suggests that implementation of Part D in 2006 prolonged life by at least 1 year for nearly 200,000 Medicare beneficiaries through lower mortality associated primarily with diabetes, CHF, stroke, and MI. The simulated impact on mortality was greatest in the initial years following program inception, a trend also observed in similar studies.

 The gradual rise in numbers of lives saved, followed by lower estimates in later years, may reflect the large number of individuals gaining comprehensive drug coverage in the initial years (with smaller numbers of beneficiaries enrolling in Part D in later years), as well as delayed onset of morbidity and mortality associated with better access to medicines under Part D.

The largest contribution to lives saved came from a reduction in deaths attributable to diabetes, CHF, MI, and stroke. Studies have shown that the use of medications to treat diabetes and cardiovascular conditions increased significantly in the post Part D period. Although estimates vary by source, the use of diabetes medicines increased by an estimated 3.7% to 17.9%.

Similarly, utilization rates for statins are reported to have increased by 13% to 22%, and antihypertensive utilization rates by 14% to 29%.

Other studies that have examined the impact of Part D on mortality have been confined to a narrow window of analysis, often focusing only on the effects of the program in the first year or two following implementation.

 The Briesacher et al study analyzed data through 4 years following Part D implementation, but made strong assumptions regarding the continuation of falling mortality trends from 2000 to 2010.

 The magnitude of our findings is generally consistent with other studies. We estimated a 1.2% decline in annual mortality over 9 years. Huh and Reif estimated a 2.2% decline in the first 2 years of Part D, similar to our finding that the mortality effect was higher in the first few years of Part D.

 Briesacher et al found a decline of approximately 1% in mortality, though their result is not statistically significant.

 Dunn and Shapiro estimated that 21,800 to 25,500 individuals with cardiovascular disease were still alive in mid-2007 because of Part D.

 We estimated that about 20,000 lives were saved in 2006 and 26,000 lives were saved in 2007; this number includes people with diabetes as well as cardiovascular disease.

Our study is unique in that it uses a microsimulation approach to estimate how improved access to pharmaceuticals that can lower blood pressure, cholesterol levels, and blood glucose levels can reduce the incidence of adverse health events that contribute to mortality. The strengths of using a microsimulation model include the ability to explain the pathway by which Part D benefits those participants who otherwise would have lower access to medications. This approach controls for patient health risk factors to isolate the timing and magnitude of medication use on reducing mortality.

Limitations are largely driven by data challenges. Part D beneficiaries who otherwise would not have comprehensive access to medications cannot directly be identified and observed, so this study used a microsimulation approach that uses published data to make assumptions regarding increased use of medications and effect on health outcomes.

The model does not take into account possible declines in medication persistence. That is, the documented increases in medication use tied to Part D (ie, 29%, 31%, and 22% increases in drug possession rates observed for patients diagnosed with hyperlipidemia, diabetes, and hypertension, respectively)

 might not be sustained over time, thus overestimating the mortality benefits of Part D.

 A more detailed discussion of the strengths and limitations of this microsimulation model are described in depth elsewhere.

Since the implementation of Part D in 2006, nearly 200,000 Medicare beneficiaries have lived at least 1 year longer, with an average 3.3 year increase in longevity. Reductions in mortality have occurred because of fewer deaths associated with medication-sensitive conditions such as diabetes, CHF, stroke, and MI. 

 IHS Life Sciences, Washington, DC (FC, TMD, APS).

Dr Chen, Mr Dall, and Ms Semilla report serving as consultants for PhRMA.

Concept and design (FC, TMD, APS); acquisition of data (FC, APS); analysis and interpretation of data (FC, TMD, APS); drafting of the manuscript (TMD, APS); critical revision of the manuscript for important intellectual content (TMD, APS); statistical analysis (FC, TMD, APS); administrative, technical, or logistic support (APS); and supervision (TMD).

April P. Semilla, MS, 1150 Connecticut Ave NW, Suite 401, Washington, DC 20036. E-mail: 

		Joyce GF, Goldman DP, Vogt WB, Sun E, Jena AB. Medicare Part D after 2 years. 

		Duggan M, Morton FS. The effect of Medicare Part D on pharmaceutical prices and utilization. National Bureau of Economic Research; 2008.

		Lichtenberg FR, Sun SX. The impact of Medicare Part D on prescription drug use by the elderly. 

		Yin W, Basu A, Zhang JX, et al. The effect of the Medicare Part D prescription benefit on drug utilization and expenditures. 

		Kaestner R, Long C, Alexander GC. Effects of prescription drug insurance on hospitalization and mortality: evidence from Medicare Part D. National Bureau of Economic Research; 2014.

		Liu FX, Alexander GC, Crawford SY, et al. The impact of Medicare Part D on out-of-pocket costs for prescription drugs, medication utilization, health resource utilization, and preference-based health utility. 

		Schneeweiss S, Patrick AR, Pedan A, et al. The effect of Medicare Part D coverage on drug use and cost sharing among seniors without prior drug benefits. 

		Ayyagari P, Shane D. Does insurance coverage improve health? evidence from Medicare Part D. College of Business at Illinois website. 

https://business.illinois.edu/nmiller/mhec/Ayyagari.pdf

. Published April 2014. Accessed June 24, 2015.

		Ketcham JD, Simon KI. Medicare Part D’s effects on elderly patients’ drug costs and utilization. 

http://www.cdc.gov/nchs/products/life_tables.htm

		Briesacher BA, Madden JM, Zhang F, et al. Did Medicare Part D affect national trends in health outcomes or hospitalizations? a time-series analysis. 

		Huh J, Reif J. Did Medicare Part D reduce mortality? Julian Reif website. 

http://www.julianreif.com/research/medicare_d_mortality.pdf

		Dunn A, Shapiro AH. Does Medicare Part D save lives? Federal Reserve Bank of San Francisco website. 

http://www.frbsf.org/economic-research/files/wp2015-04.pdf

		Dall TM, Storm MV, Semilla AP, et al. Value of lifestyle intervention to prevent diabetes and sequelae. 

		Su W, Huang J, Chen F, et al. Modeling the clinical and economic implications of obesity using microsimulation. 

		National Center for Health Statistics. Continuous NHANES Web Tutorial. CDC website. 

http://www.cdc.gov/nchs/tutorials/NHANES/index_continuous.htm

		Kroese DP, Taimre T, Botev ZI. Handbook of Monte Carlo Methods. Hoboken, NJ: John Wiley & Sons; 2011.

		Hoadley J, Summer L, Hargrave E, Cubanski J, Neuman T. Medicare Part D in its ninth year: the 2014 marketplace and key trends, 2006-2014. Kaiser Family Foundation website. 

http://kff.org/report-section/medicare-part-d-inits-ninth-year-section-1-part-d-enrollment-and-plan-availability

		Afendulis CC, Chernew ME. State-level impacts of Medicare Part D. 

		Kaiser Family Foundation. Total number of Medicare beneficiaries-2006. Kaiser Family Foundation website. 

http://kff.org/medicare/state-indicator/total-medicare-beneficiaries

		Kaestner R, Khan N. Medicare Part D and its effect on the use of prescription drugs, use of other health care services and health of the elderly. 

		Madden JM, Graves AJ, Zhang F, et al. Cost-related medication nonadherence and spending on basic needs following implementation of Medicare Part D. 

		Zhang Y, Lave JR, Donohue JM, et al. The impact of Medicare Part D on medication adherence among older adults enrolled in Medicare-Advantage products. 

		Wilson PW, Anderson KM, Harris T, Kannel WB, Castelli WP. Determinants of change in total cholesterol and HDL-C with age: the Framingham Study. 

		Wilson PW, Bozeman SR, Burton TM, et al. Prediction of first events of coronary heart disease and stroke with consideration of adiposity. 

		26. D’Agostino RB, Wolf P, Belanger A, Kannel W. Framingham heart study: risk score profiles. Framingham Heart Study website. 

https://www.framinghamheartstudy.org/riskfunctions/cardiovascular-disease/10-year-risk.php

		National Institutes of Health. The Framingham study: an epidemiological investigation of cardiovascular diseases: section 34: some risk factors related to the annual incidence of cardiovascular disease and death using pooled repeated biennial measurements: Framingham heart study, 30 year followup. Biologic Specimen and Data Repository Information Coordinating Center website. 

https://biolincc.nhlbi.nih.gov/static/studies/fhsc/30-Year_Followup_(Section_34).pdf

		28. Clarke PM, Gray AM, Briggs A, et al. A model to estimate the lifetime health outcomes of patients with type 2 diabetes: the United Kingdom Prospective Diabetes Study (UKPDS) Outcomes Model (UKPDS no. 68). 

		Nichols GA, Reinier K, Chugh SS. Independent contribution of diabetes to increased prevalence and incidence of atrial fibrillation. 

		Hippisley-Cox J, Coupland C. Predicting the risk of chronic kidney disease in men and women in England and Wales: prospective derivation and external validation of the QKidney Scores. 

		Neter JE, Stam BE, Kok FJ, Grobbee DE,Geleijnse JM. Influence of weight reduction on blood pressure: a meta-analysis of randomized controlled trials. 

		Heianza Y, Arase Y, Fujihara K, et al. Longitudinal trajectories of HbA1c and fasting plasma glucose levels during the development of type 2 diabetes: the Toranomon Hospital Health Management Center Study 7 (TOPICS 7). 

		Schaufelberger M, Swedberg K, Köster M, Rosén M, Rosengren A. Decreasing one-year mortality and hospitalization rates for heart failure in Sweden; Data from the Swedish Hospital Discharge Registry 1988 to 2000. 

		CDC, National Center for Health Statistics. Underlying cause of death 1999-2010. 

		Taylor F, Huffman MD, Macedo AF, et al. Statins for the primary prevention of cardiovascular disease. 

Articles

Reductions in Mortality Among Medicare Beneficiaries Following the Implementation of Medicare Part D

Medicare Part D has had important implications for patient outcomes and treatment costs among beneficiaries with congestive heart failure (CHF). This study finds that improved medication adherence associated with expansion of drug coverage under Part D led to nearly $2.6 billion in reductions in medical expenditures annually among beneficiaries diagnosed with CHF and without prior comprehensive drug coverage, of which over $2.3 billion was savings to Medicare. Further improvements in adherence could potentially save Medicare another $1.9 billion annually, generating upwards of $22.4 billion in federal savings over 10 years.

Congestive heart failure (CHF) is the most common reason for hospitalization among the elderly, accounting for one-fifth of all admissions.

 Consequently, beneficiaries with CHF are among the most costly to Medicare; they represent 14% of the population, but account for 43% of Medicare Part A and B spending.

 More than 3.5 million Part D enrollees were diagnosed with CHF in 2010.

Extensive evidence demonstrates that medications can be used to effectively treat and manage CHF by slowing disease progression, allaying symptoms, and reducing use of medical services such as hospitalizations.

 However, adherence to CHF medications remains suboptimal, suggesting the potential for improvements in outcomes and medical cost savings. For example, Roebuck and colleagues found that patients with CHF who were adherent to medications experienced 5.7 fewer hospital days and $8881 in reduced medical expenditure annually compared with patients who were not adherent.

 Another study found that each 10% increase in the use of CHF medications was associated with reductions in 3-year Medicare Part A and B expenditures of $510 to $923 (2006 dollars).

The Congressional Budget Office (CBO) more broadly recognizes this link between medication use and spending on other medical services. Last fall, the CBO announced a change to the methodology used to estimate the impact of legislation affecting the volume of drug use in Medicare. Specifically, the CBO now assumes that each 1% increase in the number of prescription drugs filled is associated with a 0.2% decline in spending for other healthcare services.

 However, this estimate may be conservative, particularly in the case of CHF, as the CBO also acknowledges that the offsetting effects are generally larger for specific conditions that are more sensitive to drug use compared with the broader population. Despite the known benefits of medications to manage CHF, research shows that treatment for this life-threatening and costly disease remains suboptimal, with large percentages of patients not receiving or adhering to essential medications.

 A significant driver of poor medication use is lack of insurance coverage, which leads patients to not fill prescriptions or to skip medication due to cost. Prior to implementation of Part D in 2006, approximately 30% of Medicare beneficiaries lacked creditable drug coverage, representing a critical gap in care for the elderly population. While we cannot fully discern the extent to which undertreatment occurred, nearly 1 in 5 Medicare beneficiaries with CHF were not treated with medications commonly recommended for this condition.

 Part D expanded the Medicare benefit with the goal of addressing financial barriers to medication access and adherence. By the end of Part D’s inaugural year, more than half of previously uninsured beneficiaries had obtained coverage through Part D and the overall coverage rate reached 90%.

Research across a wide variety of conditions has linked Part D to reductions in out-of-pocket spending and improvements in adherence for beneficiaries with no or limited prior drug coverage.

One study found that Part D was associated with approximately $1200 in reduced medical spending among those who previously had limited drug coverage, resulting in overall savings of $13 billion in the first full year of the program.

The objective of our study was to estimate the impact of Part D and improved use of CHF medications on savings in medical expenditures.

Study Data and Methodology: Part D Participation, Adherence, and Medical Expenditures

We developed a simulation model that quantifies the impact of Part D on adherence and medical expenditures for Medicare beneficiaries with CHF by combining information from the literature with our own analytic methodology.

The analysis focused on Part D enrollees with CHF who had limited or no drug coverage prior to Part D. Based on CHF age-sex prevalence rates and US Census Bureau population projections, we estimated that 3.7 million Part D enrollees were diagnosed with CHF in 2012, including 1.6 million with prior limited or no drug coverage.

 Estimates of healthcare utilization, average medical costs, and age distribution of patients with CHF were generated from the 2006 to 2010 files of the Medical Expenditure Panel Survey (MEPS). Medicare beneficiaries diagnosed with CHF were identified in MEPS using 

The International Classification of Diseases, 9th Revision

 codes 428.xx (n = 962 patients). Per capita hospitalizations, emergency department visits, ambulatory visits, and home health days were estimated by sex and 5-year age intervals. The average length of stay per hospitalization and medical expenditures, by care delivery setting and age, were estimated using data from the 2010 Nationwide Inpatient Sample (n = 71,147 CHF hospitalizations). Hospitalization charges were converted to costs using published hospital-specific cost-to-charge ratios in 2012 dollars based on the medical component of the Consumer Price Index.

We quantified the impact of Part D on prescription medications used to treat CHF based on the peer-reviewed literature. According to 1 study, beneficiaries without prescription insurance prior to Part D filled an average of 6 more CHF prescriptions annually post Part D.

 Other studies showed that Part D increased cardiovascular-related medication use for individuals without prior coverage as well as those whose prior coverage was subject to a quarterly benefit cap, typically $150 or $350.

 From analysis of the 2005 to 2006 Medicare Current Beneficiary Survey and published sources, we estimated that 37.1% of enrollees were uninsured prior to Part D coverage, 3.6% had limited coverage (substantially less generous than Part D), 3.3% had moderate coverage (somewhat less generous than Part D), and the remaining 56% had prior comprehensive coverage.

 Combining information from the published literature, the average weighted increase in medication use attributed to Part D for those without prior comprehensive coverage was 19.6%. We used the CBO offset estimate which assumes that nonprescription medical expenditures will decline by 0.2% for each 1% increase in prescription use. To calculate net savings, we subtracted the increased medication expenditures associated with improved adherence.

Despite improvements in medication use, many Part D enrollees with CHF remain nonadherent to medications used to treat CHF based on the proportion of days covered at 80% or greater.

To explore Part D’s potential to generate better outcomes and increase savings, we also simulated the economic implications of improving adherence to 80%.

Our findings indicated that the mean annual medical expenditure was $33,247 per CHF beneficiary. Using the CBO offset estimate, this suggests a $66.52 annual decrease in medical expenditures per 1% increase in prescription drug use (or 0.2% of total annual cost). Our analysis of spending using the MEPS suggests that 88.3% of savings would be recognized by Medicare.

Extrapolating our findings to the Part D population with CHF who had no prior or limited medication coverage, our analysis demonstrated that the expansion of coverage under Part D reduced mean annual nondrug medical expenditure by $1827 per beneficiary in 2012 (

). The magnitude of the estimated impact was greater among the previously uninsured population ($2050 savings), and somewhat smaller for patients with previous limited and moderate coverage ($773 and $465, respectively). Overall, these savings translated to over $2.9 billion in reduced medical expenditures. Accounting for the increase in medication spending, we estimated that the net effect of overall savings exceeded $2.6 billion, of which over $2.3 billion was savings to Medicare.

While Part D reduced rates of undertreatment among beneficiaries with CHF, adherence to therapy remains suboptimal. Improving adherence to a target 80% of days covered would produce additional savings of approximately $2.2 billion ($600/patient) via reductions in healthcare utilization, of which $1.9 billion would be savings to Medicare (

The increasing prevalence of CHF among the Medicare population highlights the importance of initiatives to manage the high medical costs associated with this disease. Evidence consistently demonstrates a strong connection between better medication use and reductions in other medical expenditures. One challenge with studies reviewing the relationship between prescription drugs and outcomes is that they typically consider only short-term implications. For example, much of the published literature surrounding the impact of Part D relies on data from 2006 to 2008, when the Part D program was still in its infancy. While these studies generally demonstrate short-term savings, we anticipate evidence of the value of improved coverage and the potential for savings to grow over time. In addition, most patients with CHF have multiple comorbidities, including coronary artery disease, hypertension, and diabetes, which can also be managed by many of the same therapies used for CHF.

Over the next decade, millions of new beneficiaries will become eligible for Medicare, making efforts to appropriately manage disease and optimize medication use important policy tools for improving beneficiary health and controlling costs. Based on current age-sex CHF prevalence rates, US Census Bureau population projections, and Part D participation rates, the number of beneficiaries in Part D with CHF is projected to increase by about 26% between 2013 and 2022—suggesting that the number of Part D beneficiaries with CHF could grow to about 4.77 million by 2022.

Our analysis suggests that over this 10-year period, Medicare will realize over $26.9 billion in savings, driven by reductions in Part A and B expenditures associated with improved CHF medication adherence following the implementation of Part D. Going forward, reaching recommended levels of adherence (proportion of days covered of 80% or higher) among Part D enrollees with CHF could potentially double estimated savings, yielding roughly an additional $22.4 billion in federal savings from 2013 through 2022.

 IHS Healthcare & Pharma, Washington, DC (TDB, TMD, PDG, APS).

Mr Blanchard, Mr Dall, Mr Gallo, and Ms Semilla report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this supplement.

Concept and design (TMD); analysis and interpretation of data (TMD, PDG, APS); drafting of the manuscript (TDB, TMD, APS); critical revision of the manuscript for important intellectual content (TDB, TMD); statistical analysis (TMD, PDG); administrative, technical, or logistic support (TDB, PDG, APS); and supervision (TMD).

 Timothy M. Dall, MS, 1150 Connecticut Ave NW, Ste 401, Washington, DC 20036. E-mail: 

		Emory Healthcare. Heart Failure Statistics. 

 [serial online]. Published 2013. Accessed April 2, 2013.

		Linden A, Adler-Milstein J. Medicare disease management in policy context. 

		3. Medication therapy management in a chronically ill population: interim report. Centers for Medicare & Medicaid Services website. 

http://innovation.cms.gov/Files/reports/MTM-Interim-Report-01-2013.pdf

		Sokol MC, McGuigan KA, Verbrugge RR, Epstein RS. Impact of medication adherence on hospitalization risk and healthcare cost. 

		Boswell KA, Cook CL, Burch SP, Eaddy MT, Cantrell CT. Associating medication adherence with improved outcomes: a systematic literature review. 

		Roebuck MC, Liberman JN, Gemmill-Toyama M, Brennan TA. Medication adherence leads to lower health care use and costs despite increased drug spending. 

		Heart Failure Society of America; Lindenfeld J, Albsert NM, Boehmer JP, et al. HFSA 2010 Comprehensive Heart Failure Practice Guideline. 

		Lopert R, Shoemaker JS, Davidoff A, et al. Medication adherence and Medicare expenditure among beneficiaries with heart failure. 

		Offsetting effects of prescription drug use on Medicare’s spending for medical services. Congressional Budget Office website. 

http://www.cbo.gov/sites/default/files/cbofiles/attachments/43741-MedicalOffsets-11-29-12.pdf

		Chernew ME, Shah MR, Wegh A, et al. Impact of decreasing copayments on medication adherence within a disease management environment. 

		Management of chronic heart failure: a national clinical guideline. Scottish Intercollegiate Guidelines Network website. 

		Safran DG, Strollo MK, Guterman S, Li A, Rogers WH, Neuman P. Prescription coverage, use and spending before and after Part D implementation: a national longitudinal panel study. 

		Polinski JM, Kilabuk E, Schneeweiss S, Brennan T, Shrank WH. Changes in drug use and out-of-pocket costs associated with Medicare Part D implementation: a systematic review. 

		2012 National Population Projections. United States Census Bureau website. 

http://www.census.gov/population/projections/data/national/2012.html

		Donohue JM, Zhang Y, Lave JR, et al. The Medicare drug benefit (Part D) and treatment of heart failure in older adults. 

		Schneeweiss S, Patrick AR, Maclure M, Dormuth CR, Glynn RJ. Adherence to statin therapy under drug cost sharing in patients with and without acute myocardial infarction: a population-based natural experiment. 

		Zhang Y, Lave JR, Donohue JM, Fischer MA, Chernew ME, Newhouse JP. The impact of Medicare Part D on medication adherence among older adults enrolled in Medicare-Advantage products. 

		Doshi JA, Polsky D. Drug benefit generosity and essential medication use among Medicare-eligible retirees. 

		Zhang Y, Wu SH, Fendrick AM, Baicker K. Variation in medication adherence in heart failure. 

The Economic Impact of Medicare Part D on Congestive Heart Failure

To share outcomes and lessons learned from an evaluation of disease management (DM) programs for asthma, congestive heart failure (CHF), and diabetes for TRICARE patients.

Multiyear evaluation of participants in voluntary, opt-out DM programs. Patientcentered programs, administered by 3 regional contractors, provide phone-based consultations with a care manager, educational materials, and newsletters. The study sample consisted of 23,793 asthma, 4092 CHF, and 29,604 diabetes patients with at least 6 months' tenure in the program.

Medical claims were analyzed to quantify program effect on healthcare utilization, medical costs, and clinical outcomes. Multivariate regression analysis with an historical control group was used to predict patient outcomes in the absence of DM. The difference between actual and predicted DM patient outcomes was attributed to the program. A patient survey collected data on program satisfaction and perceived usefulness of program information and services.

Modest improvements in patient outcomes included reduced inpatient days and medical costs, and (with few exceptions) increased percentages of patients receiving appropriate medications and tests. Annual per patient reductions in medical costs were $453, $371, and $783 for asthma, CHF, and diabetes program participants, respectively. The estimated return on investment was $1.26 per $1.00 spent on DM services.

Findings suggest that the DM programs more than pay for themselves, in addition to improving patient health and quality of life. Lessons learned in program design, implementation, effectiveness, and evaluation may benefit employers contemplating DM, DM providers, and evaluators of DM programs.

A multiyear evaluation of 3 TRICARE disease management (DM) programs for asthma, congestive heart failure, and diabetes looked at patient outcomes along 4 dimensions: utilization, financial, humanistic, and clinical.

For all 3 diseases, outcome measures moved in the direction anticipated -- that is, less emergency use and inpatient care, lower medical costs, and (with few exceptions) increased percentages of patients receiving appropriate medications and tests. 

The estimated return on investment was $1.26 per $1.00 spent on DM services. 

These DM programs more than pay for themselves, in addition to improving patient health and quality of life.

Providing patients with information to manage their medical conditions and encouraging positive change in health-related behavior can lead to improved health and avoidance of costly medical expenditures. The Disease Management Association of America identified asthma, congestive heart failure (CHF), and diabetes among the chronic diseases with the greatest potential for disease management (DM).

 There is no consensus, however, on whether DM provides the anticipated benefits. Some studies report positive return on investment (ROI),

 whereas others report negative ROI or inconclusive results.

 The Military Health System, through the TRICARE program administered by the TRICARE Management Activity (TMA), provides care to 9.2 million beneficiaries at an annual cost of $44.8 billion (FY 2009 budget).

 Among current beneficiaries under age 65 years residing in the United States, analysis of administrative records suggests that approximately 80,000 have asthma, 11,000 have CHF, and 225,000 have diabetes.

 TRICARE Management Activity contracted with its 3 regional managed care support contractors (MCSCs) to provide DM services to high-utilization patients with asthma and CHF beginning in September 2006, adding diabetes in June 2007. Although this voluntary, opt-out DM program is ongoing, the findings presented here are based on analysis of patients contacted through September 30, 2008, who had at least 6 months’ tenure in the program. The 6-month-tenure criterion was to allow sufficient time for the MCSCs to provide DM services. The September 2008 cutoff was to ensure 100% completeness of medical claims. The analysis sample consisted of 23,793 asthma, 4092 CHF, and 29,604 diabetes patients.

 TRICARE Management Activity identifies candidates for DM based on high utilization of disease-related services (emergency and outpatient visits, hospitalizations) and, for asthma, high use of short-acting rescue prescriptions. Patient names are forwarded to the MCSCs to verify eligibility and provide DM services. Some patients choose to fully participate and receive personalized telephonic counseling and educational mailings (referred to as “engaged” patients), whereas others decline personalized care but receive newsletters. Some patients opt out of the program.

 The DM content and services follow industry best practices, including (1) an initial 40- to 50-minute baseline assessment by telephone with a care manager; (2) monthly follow-up telephone calls to set goals, assess progress toward those goals, and educate patients about their conditions and self-management; (3) educational materials specific to the patient’s needs (eg, pamphlets, videos, cookbooks); and (4) newsletters and Internet-based materials. All services are educational and are intended to provide patients with disease-specific knowledge and self-management skills. Two MCSCs consider a patient to have “graduated” 12 months after receiving the baseline assessment, while the third considers a patient to have graduated when goals regarding disease understanding and self-management have been met. After graduation, patients continue to receive newsletters and biannual phone contact.

 Best practices for DM evaluation suggest measuring program effect along 4 dimensions: utilization, financial, clinical, and humanistic.

 Using an encrypted patient identifier, we electronically linked patient DM administrative records, medical claims, and a patient survey. Study protocols were approved by an institutional review board.

 Patient outcome metrics were selected using recommendations of the 2nd Annual Disease Management Outcomes Summit.

 Utilization metrics (emergency visits and hospital inpatient days) and financial metrics (medical expenditures) for asthma and CHF are disease specific. Claims are counted only when the disease is listed in the first or second diagnosis code position (first diagnosis position only for asthma-related hospitalization and emergency visits).

 Only a small portion of diabetes-related medical costs are directly for treatment of diabetes; the majority are associated with complications of diabetes (eg, neurologic symptoms, peripheral vascular disease, cardiovascular disease, renal complications, endocrine complications, ophthalmic complications) and higher use of medical care for general medical conditions.

 Therefore, for diabetes we report disease-related utilization and cost but focus on total utilization and cost (with claims for injury, pregnancy, congenital abnormalities, and malignant cancer excluded). All medical costs were adjusted to 2008 dollars using the medical component of the consumer price index.

 Clinical metrics were limited to those available in electronic records and include the percentage of patients receiving an exam or a pharmaceutical in the past year. Asthma metrics were (1) use of long-term controller medications (measured as the percentage of patients with at least 1 dispensed prescription for inhaled corticosteroids, nedocromil, cromolyn sodium, leukotriene modifiers, or methylxanthines); and (2) the percentage of patients with a documented spirometry test. For CHF we determined the percentage of patients with angiotensin-converting enzyme (ACE) inhibitor and beta-blocker prescription fills. Diabetes metrics were (1) the percentage of patients with at least 1 glycosylated hemoglobin (A1C) test, (2) the percentage of patients with at least 1 dilated retinal examination, and (3) the percentage of patients with at least 1 microalbuminuria or clinical albuminuria test.

 Humanistic metrics included patient self-assessment of change in quality of life, improved understanding of both the disease and self-management, and overall program satisfaction.

 Patient characteristics and demographics came from the Defense Enrollment Eligibility Reporting System. Healthcare utilization and expenditures came from TRICARE’s Medical Data Repository. These included care provided through TRICARE’s purchased care program and care provided at military treatment facilities. Information on DM services received and patient DM participation status came from the MCSC patient tracking system.

 A disease-specific questionnaire was mailed to patients 6 months after their initial baseline assessment to collect information on change in quality of life, perceived usefulness of the program and materials received, and overall program satisfaction. The survey questionnaire (available from the authors) is designed based on existing survey instruments such as the SF-36 short-form survey, Healthcare Effectiveness Data and Information Set Member Satisfaction Survey, and the 2006 Disease Management Association of America Participant Satisfaction Survey.

 A total of 73,156 patients were contacted by the MCSCs as of September 30, 2008. The analysis excluded 8969 patients who had opted out of the program and 4685 patients with fewer than 6 months of program tenure (who had insufficient time to complete the program). Also excluded from the analysis were patients whose complete medical records were no longer available because the patient became Medicare eligible (n = 434), was no longer eligible for TRICARE (n = 1272), or died (n = 25). Patients diagnosed with HIV/AIDS (n = 50) or end-stage renal disease (n = 232) were excluded, with the exception of diabetes patients first diagnosed with end-stage renal disease after becoming eligible for DM.

 As these are opt-out DM programs with eligibility determined by high utilization, using nonparticipants as the control group to measure program effect introduces the problem of selection bias, whereas a pre—post evaluation design introduces the problem of regression to the mean.

 To reduce these known biases, we identified a comparable sample of patients, referred to as the historical control group (HCG), and used their outcomes to predict outcomes for the DM population in the absence of a DM program.

 The HCG consists of 12,192 asthma, 3247 CHF, and 23,778 diabetes patients who in September 2004 met the same eligibility criteria used to identify candidates for DM in September 2006 and later. We estimated multivariate regressions using patient annual outcomes from October 2004 to September 2005 as the dependent variables. We used Poisson regression to model inpatient days and ambulatory visits, a generalized linear model with gamma distribution to model medical expenditures, and logistic regression to model receipt of tests and pharmaceuticals. Separate regressions were estimated for the Prime (managed care) and non-Prime (preferred provider organization) populations. Regression results are available from the authors upon request.

 Covariates included patient characteristics in September 2004 (age, sex, military sponsoring service [Army, Navy, Air Force, other], region [North, South, West]) and utilization between October 2003 and September 2004 (Charlson Comorbidity Index and disease-related inpatient days and ambulatory visits). For asthma we included the number of 30-day asthmarelated prescriptions. For CHF we included whether the patient used ACE inhibitors and beta-blockers. For diabetes we included whether the patient had an A1C test in the past year and used total rather than diabetes-specific utilization.

 Using regression coefficients estimated with the HCG, we predicted DM patient outcomes for the period following program eligibility based on patient characteristics when first eligible for DM and healthcare utilization patterns during the 12 months prior to DM. Differences between annualized observed and predicted outcomes for the DM population were attributed to DM.

DM Process and Patient Participation Outcomes

 Although the regression analysis controls for observed differences between the DM and HCG populations, the populations are relatively similar (

). The DM population tended to be slightly less healthy as measured by the Charlson Comorbidity Index, had higher rates of ambulatory visits, and had lower rates of emergency visits and inpatient days during the 12-month identification window.

Findings from TRICARE's disease management programs for asthma, congestive heart failure, and diabetes patients suggest that the programs more than pay for themselves.

Timothy M. Dall, MS

Articles

Reductions in Mortality Among Medicare Beneficiaries Following the Implementation of Medicare Part D

The Economic Impact of Medicare Part D on Congestive Heart Failure

Outcomes and Lessons Learned From Evaluating TRICARE's Disease Management Programs