Printer Friendly

Applying the 2003 beers update to elderly Medicare enrollees in the Part D program.

Introduction

Inappropriate prescription medication use in the elderly can result in intense medical and safety problems. Elderly patients are more susceptible to medication-induced health complications, such as depression, falls, hip fractures, and confusion, because of their poorer health status, a greater potential to receive multiple medications, and differences in how the body absorbs, metabolizes, and eliminates a medication (Bootman, Harrison, & Cox, 1997; Hanlon et al., 1997; Shrank, Polinski, & Avorn, 2007) The Beers Criteria for potentially inappropriate medication use in older adults have been used widely to identify medications that may be inappropriately prescribed for older adults (Beers, 1997; Fick et al., 2003; Blackwell, Ciborowski, Baugh, & Montgomery, 2008). Several studies indicate that medications meeting the Beers Criteria, known as Beers drugs, are particularly harmful to the elderly (U.S. General Accounting Office 1995; Blackwell, Ciborowski, Baugh, & Montgomery, 2008). Blackwell et al(2008) estimated that approximately 47% of elderly Medicare enrollees who were dually enrolled in Medicare and Medicaid received a potentially inappropriate medication. (1)

In 2006, all Medicare beneficiaries were given the opportunity to participate in Medicare Part D, while the low income, elderly and disabled dual enrollees were automatically placed in Part D (The Henry J. Kaiser Family Foundation, 2011). Medicare is administered directly by the federal government, using a uniform set of national guidelines and rules that apply to enrollees regardless of residence. This uniformity allows the Centers for Medicare & Medicaid Services (CMS) to monitor, on a continual basis, the use of Beers drugs in the Part D program. Conversely, prior to implementation of the Part D program, prescription drug coverage for elderly dual enrollees was a states-administered conglomeration of programs that, by its very nature, made it difficult to rule out differences in one state's drug coverage rules compared to another state. The Medicare Part D program presents an opportunity to examine the use of Beers drugs among both dual enrollees and non-dual enrollees.

Previous studies addressing Medicare elderly enrollees have relied on Medicaid data, which was limited to dual enrollees--a population comprised of many individuals who are sick and frail, with substantial health needs (The Henry J. Kaiser Family Foundation, 2004). Thus, findings from these prior studies may have disproportionately represented Beers drug use among the elderly. The results of this study can be compared to those of previous studies to determine whether non-dual Medicare enrollees have Beers drug utilization similar to dual enrollees.

By examining Medicare's national enrollment and claims data, following implementation of the Part D program, policy makers will have more information at their disposal to assess the impact of continued Beers drugs use among all Medicare-- dual and non-dual--Part D medication recipients. They can use our findings about Beers drug utilization, within and between the two groups, to formulate improved policies in the Part D program. Part D plans may use the information to take a more active role in assisting practitioners with patient safety through improved medication therapy decisions.

Our primary research question was to assess Beers drug use in the dual enrollee population compared to the non-dual enrollee population. Our second question was to assess which medication therapeutic category had the highest percentage of Beers drug use. Our third question was to explore the association between dual/non-dual enrollment status and Beers drug use in the elderly, controlling for the effects of age, gender, race/ethnicity, census region, and health status.

Methodology

Data

Calendar year 2007 data were obtained from the Medicare Enhanced Denominator file (or simply, Denominator file), which provided dual eligibility status, age, gender, and race/ethnicity data. The Medicare Part D prescription drug event (PDE) files provided the National Drug Code (NDC) for each prescription fill. The NDC was linked using the Medi-Span[R] therapeutic classification system to identify Beers medications and medication therapeutic categories (Wolters Kluwer Health, 2010). Lastly, CMS' Hierarchical Condition Category (CMS-HCC) prospective risk score model (i.e., the health risk adjuster) provided beneficiary risk scores.

Methods

We employed a cross-sectional design. Using the Denominator file, we selected beneficiaries (irrespective of medication use) who would have attained age 65 as of December 31, 2007, and who were enrolled in Medicare Part D for the entire 2007 calendar year, and identified those who were dual enrollees and those who were non-dual enrollees (Exhibit 1). Enrollees who did not have a full twelve months of participation in Medicare Part D in 2007 were not selected in order to more accurately measure the effects of dual eligibility. From the selected population, dual/non-dual enrollees having at least one medication fill within calendar year 2007 were retained, and comprised our study population for subsequent analysis.

We initially report the population characteristics of dual status in addition to census region, gender, age group, race/ethnicity, and health status (i.e., as disease burden and prescription burden). Regional assignment was based on United States Census regions (U.S. Department of Economics and Statistics Administration, U.S. Census Bureau, 2010) using the Social Security Administration's (SSA) state code of residence of the beneficiary (ResDAC, 2009). State codes identifying beneficiary residence outside of the continental United States were classified as "other."

Race/ethnicity assignment was made using race code designations (ResDAC, 2009). We divided race/ethnicity into the groups of Black, Hispanic, White, and Other. Due to small numbers of beneficiaries in the categories coded in the Denominator file as Asian/Pacific Islander, American Indian/Alaska Native, Unknown, and Other, these beneficiaries were combined into our Other category.

To evaluate health status, we employed the two variables of disease burden and prescription burden. Disease burden was assigned to the study population using the CMS-HCC prospective risk score model for calendar year 2007 (U.S. Department of Health and Human Services, 2011). The model is a risk adjustment model used for Medicare Part C reimbursement. The CMS-HCC model, updated yearly, uses demographics and a diagnosis-based medical profile, captured during all inpatient and outpatient clinician encounters the previous year, to produce a health-based measure of future medical expenditures. The HCC score is normalized to the extent that 1.0 means a beneficiary is expected to have expenditures of the average Feefor-Service Medicare beneficiary. The model is used to predict cost of care for an individual in a given year, even though it does not explicitly reflect how sick an individual may be in a given year. For the study, we use the model as a proxy for disease burden with the realization of this limitation. We divided disease burden into the quartile groups, having scores of low (0-0.543), medium-low (0.544-0.842), medium-high (0.843-1.452), and high (1.453 and over).

We defined prescription burden as the calculated total number of original and refill prescriptions filled--a definition we use throughout this paper--per enrollee during calendar year 2007. We divided prescription burden into the quartile groups of low (0- 17), medium-low (18-35), medium-high (36-64), and high (65 and over).

Beers drug identification was based on the criteria addressed by the Fick et al. (2003) update. Beneficiaries who received any medication defined by Fick et al. (2003) as being potentially inappropriate, independent of diagnoses or conditions, were classified as having potentially inappropriate use. Because we did not capture dosing information from our data, dose limited medications identified by Fick et al. (2003) were not included in the analysis.

Analysis

Population characteristics were initially assessed followed by an individual- level logistic regression analysis. We then assessed the percentage of Beers prescription fills against all prescription fills based on medication therapeutic category. Population characteristics for the calendar year 2007 study population are reported as follows:

1) Enrollees are the unique Part D beneficiaries (with or without medication use).

2) Medication Recipients are unique Part D enrollees who received at least one prescription medication.

3) Beers Medication Recipients are unique Part D enrollees who received at least one Beers drug.

The logistic regression analyses were limited to medication recipients having a risk score in order to more accurately measure the effects of disease burden. To perform the logistic regression analyses, we first constructed an outcome variable that identifies whether or not an enrollee was a Beers medication recipient. Enrollee demographics (viewed as risk factors) were used as predictor variables. In conducting each logistic regression analysis, the null hypothesis for a common odds ratio analysis applied--there is no statistically significant difference among the groups under study in terms of receiving a Beers medication. Results for the odds ratio analyses were reported as odds ratios (OR) along with pvalues and 95% confidence intervals (CIs). Analyses of the data were performed using SAS[R] software (SAS Institute Inc., 2010). The SAS[R] PROC LOGISTIC procedure was utilized to perform the logistic regression analyses.

To assess the percentage of Beers prescription fills to all prescription fills by medication therapeutic category, we retained any prescription fill for all Beers and non- Beers medications belonging to one of the nine medication categories previously addressed by Blackwell et al (2008) as Beers categories. These nine categories were analgesics and anesthetics, cardiovascular agents, central nervous system drugs, endocrine and metabolic drugs, gastrointestinal agents, genitourinary products, hematological agents, neuromuscular drugs, and respiratory agents. We compared the number of Beers prescription fills to all prescription fills in each of the nine therapeutic drug categories to demonstrate which category had the highest Beers drug utilization.

Results

Population Characteristics

As shown in Exhibit 1, 17.7 million elderly Medicare Part D enrollees were in the program through calendar year 2007 as either a dual enrollee for the entire twelve months or a non-dual enrollee for the entire twelve months. These enrollees comprised beneficiaries with or without medication use during calendar year 2007. The largest proportion of enrollees were White (83%) females (63%), aged 65-74 years old (47%), who lived in the South region (35%). The 13.6 million non-dual enrollees (77% of total enrollees) outnumbered the dual enrollees (4.0 million) more than three to one.

Of the 17.7 million elderly enrollees (Exhibit 1), approximately 16.6 million or 93% received at least one medication in calendar year 2007. The largest percentage of recipients were White (83%) females (64%), aged 65-74 years old (47%), who lived in the South region (35%), who were of medium-high disease burden (26%) and medium-low prescription burden (25%). Three times as many non-dual beneficiaries received medications compared to dual beneficiaries who received medications (12.7 million and 3.8 million, respectively). Dual beneficiaries were approximately equal to non-dual beneficiaries regarding the percentage of medication recipients to enrollees (94% and 93%, respectively).

Of the 16.6 million elderly medication recipients in calendar year 2007, 5.9 million or 35.5% received Beers medication (Exhibit 1). The Northeast region had the lowest percentage of Beers recipients (15%) whereas the South region had the highest (40%). Two and one-half times as many non-dual beneficiaries as dual beneficiaries received a Beers medication (72% and 28%, respectively). Female gender, elderly age (i.e., 65 to 74 year old elderly), White race/ethnic origin, high disease burden, and high prescription burden were characteristics found associated with Beers use.

Logistic Regression Analysis

Exhibit 2 presents the results of the individual-level odds ratio analysis to assess the likelihood of a medication recipient to receive a Beers drug based on dual enrollment status, controlling for the effects of age, gender, race/ethnicity, census region, and health status. We found that dual enrollee medication recipients were just slightly more likely to receive a Beers medication compared to non-dual enrollee recipients (OR 1.023, 95% CI 1.020-1.026), holding all other independent variables constant. Medication recipients residing in any of the remaining regions were more likely to be a Beers medication recipient compared to the Northeast region. Female gender, high disease burden, and high prescription burden were also associated with a higher likelihood of receiving a Beers medication compared to their respective counterparts. Counter to our population characteristic findings, we found that increased elderly age (i.e., 75 to 84 age group and 85+ age group) had a decreased likelihood of receiving a Beers medication compared to the younger elderly (i.e., the 65-74 age group). Hispanics were found to be more likely compared to Whites (OR 1.100, 95% CI 1.093-1.107) to receive a Beers medication, whereas Blacks were slightly less likely to receive a Beers medication compared to White beneficiaries (OR 0.960, 95% CI 0.956-0.963).

Beers Prescriptions

There were 36.2 million Beers prescriptions filled for our study population (Exhibit 3). Non-dual beneficiaries received a larger number of Beers prescriptions compared to dual beneficiaries (22.8 million and 13.4 million, respectively). Within group, non-dual enrollees also received a higher percentage of Beers prescriptions compared to dual enrollees (5.2% and 4.8%, respectively). Genitourinary products had the highest Beers use within medication therapeutic category among both dual and non-dual enrollees (21.1% and 19.9%, respectively).

Discussion

Our findings can be grouped into three areas. First, the likelihood of Beers medication use among non-duals approximates that of duals for our study population. Second, characteristics associated with the receipt of a Beers medication in our study population include Hispanic origin, younger age, female gender, poor health status, and residence outside of the U.S.' Northeast region. Third, our findings support previous findings regarding genitourinary products having the highest within therapeutic category use.

When modeling the probability of receiving a Beers medication in our regression model, based on a prioriindependent variables, we found that duals had only a slightly greater likelihood of receiving a Beers medication compared to non-duals for our study population. This finding adds to the literature by suggesting that Beers medication use among non-duals approximates that of duals for our study population. This finding is important in that it suggests that Part D plans can use similar tools/techniques to reduce Beers use among non-duals and duals.

Our findings also indicate an association between some demographic variables and the likelihood of receiving a potentially inappropriate medication. Such variables include Hispanic origin, female gender, residence in the South region, advanced age, and poor health status.

Literature addressing the prescribing of Beers medications based on racial/ethnic origin has been inconclusive. Piecoro, Browning, Prince, Ranz, and Scutchfield (2000) and Zhan et al. (2001) suggest that Black beneficiaries are at a lower risk for receiving a potentially inappropriate medication compared to other ethnic groups. Blackwell et al (2008) suggests that inappropriate medication prescribing in Blacks more closely approximates that of Whites. Differences among these findings may be attributed to the different versions of the Beers list used when conducting the analysis. The Piecoro et al. (2000) and Zhan et al (2001) studies were conducted prior to the release of the Fick et al(2003) update to the previous Beers list, whereas the Blackwell et al. (2008) study used the Fick et al(2003) update. Our current findings agree with the Blackwell etals (2008) findings. Using the Fick etal(2003) update, we found that Blacks were slightly less likely to receive a Beers medication compared to Whites. We also found Hispanics to be more likely to receive a Beers medication compared to Whites, which also agrees with Blackwell et al.'s (2008) previous finding. Thus, the gap may be closing regarding the differential use of Beers medications based on race/ethnic origin between Whites and Blacks, when comparisons are made using the Fick et al(2003) update, but continues to exist between Whites and Hispanics.

We also found that female gender is associated with an increased likelihood of potentially inappropriate prescribing relative to males for our study population. This finding agrees with previous studies performed prior to the Fick update (Zhan et al, 2001; Piecoro et al, 2000; Meredith et al., 2001; Fick et al., 2001) as well as following the Fick update (Blackwell et al.,2008).

Regarding region of residence, previous research addressing characteristics associated with a decreasedlikelihood of receiving inappropriate medications may include living in the Northeast, whereas an increasedlikelihood of receiving inappropriate medications may include living in the South (Mort & Aparasu, 2000; Rothberg et al, 2008). Our findings agree. Prescribing practices by region do not appear to have changed since earlier studies. Prescribing behavior in the Northeast region continues to remain less problematic, for which further exploration may prove most fruitful.

Studies addressing the impact of age on Beers medication use have been inconclusive. One study found the odds of inappropriate prescribing increased with age when assessing only psychotropics (Mort & Aparasu, 2000); others have found the odds decreased with age when assessing total medication use (Piecoro et al., 2000; Rothberg et al., 2008). By assessing total medication use, we also found that as age increased, the likelihood of receiving a Beers medication decreased. This finding suggests that prescribers continue to avoid prescribing potentially inappropriate medications to the eldest of the elderly.

Health status has been assessed in previous studies and is considered to be a potential risk factor for inappropriate prescribing--as health status decreases, the risk of inappropriate prescribing may increase (Chin et al., 1999; Zhan et al., 2001; Gallagher, Barry, Ryan, Hartigan, & O'Mahony, 2008). When assessing health status based on disease burden and prescription burden, we also found a potential difference between the use of Beers medication among those with poor health status compared to those with better health status.

Regarding the mix of Beers medications by therapeutic category, we found that for both duals and non-duals, genitourinary agents had the highest percentage of Beers prescriptions filled to total prescriptions by therapeutic category. In a similar study using 2003 state Medicaid data for dual eligibles, a pre-Part D study, Blackwell et al(2008) reported genitourinary products as having the highest percentage. Thus, our finding suggests that prescribing practices for genitourinary products have not changed since the previously reported work.

Limitations

First, Part D event data do not capture all medications provided to elderly enrollees, thereby possibly resulting in undercounts of numbers of prescriptions and payments. For example, prescription medications provided to enrollees during a hospital stay are not captured, because they are specifically excluded by statute. Also, prescription medications paid for by other payers (for example, the Department of Veterans Affairs) or the enrollee are not captured. Furthermore, variation in coverage determinations by Part D sponsors exists.

Second, findings based on our study population may not be representative of the population as a whole. We cannot generalize our findings to the Medicare population as a whole, since we studied only elderly medication recipients in 2007 having either twelve months of dual coverage or twelve months of non-dual coverage.

Third, medications believed to be problematic in our study population may be considered appropriate by the prescribing practitioner for a particular patient on a case-by-case basis. In this instance, prescribing the Beers medication is not indicative of lesser quality care.

Fourth, we construed the CMS-HCC risk scores as a proxy for patient disease burden, although they were originally developed as a measure of capitated payment prediction in Medicare Part C. However, these scores have been used previously as a measure of disease burden (Blackwell, Baugh, Montgomery, Ciborowski, & Levy, 2011) and compare favorably with the Charlson and Elixhauser methods as risk adjusters for mortality (Li, Kim, & Doshi, 2010).

Conclusion

The Part D program offers an opportunity for incorporation of the Beers criteria into current tools, such as formularies, utilization tools (e.g., prior authorization), and medication therapy management programs that may be further developed to assist in ensuring appropriate prescribing for older patients. Following the program's implementation, CMS has been monitoring the use of Beers medications in the Part D program (C. Tudor, Ph.D., Director, Medicare Drug Benefit and C & D Data Group, Center for Medicare & Medicaid Services, personal communication, October 5, 2010). With uniformity in program administration, a better understanding of risk factors associated with the prescribing of Beers medications in the Part D population can now be acquired. Our finding that poor health status is associated with a higher likelihood of Beers use is problematic. Future studies assessing this phenomenon, particularly focusing on a specific disease state/clinical condition, appear warranted in order to provide policymakers and Part D plans with additional information. Given that there are medication alternatives for most patients, our study findings should provide additional information for policy makers as they continue to monitor the Part D program.

With adjustment for several important and easily measured demographic, health, and prescription drug use covariates, Beers drug use appears to be as common among non-dual enrollees as it is among dual enrollees in the Part D program. New Part D drug utilization policies that apply to all beneficiaries might be enacted to reduce Beers drug use. With this knowledge, Part D plans may be able to take a more active role in assisting practitioners with patient safety through improved medication therapy decisions.

doi:http://dx.doi.org/10.5600/mmrr.002.02.a01

Correspondence

Steven A. Blackwell, Ph.D., J.D. Center for Medicare & Medicaid Innovation, Centers for Medicare & Medicaid Services, 7500 Security Boulevard, WB-06-05, Baltimore, MD 21244-1850, SBlackwell@cms.hhs.gov, Tel: (410) 786-6852, Fax: (410)786-1048

Appendix
Exhibit 4. Number of Medicare Part D Enrollees (1)
and Percent by Dual Eligibility Status, Gender, Age
Group, Origin, and Region, 2007. Age 65 and Over

Region                     Midwest                Northeast

Enrollment                          Non-                   NonStatus
Dual (2)   dual (3)    Dual (2)   dual (3)

Total Enrollees        715,288    3,136,893   834,093    2,462,599

Dual + Non-dual             3,852,181              3,296,692

Characteristic

Gender %
  Female                 73.6       62.0        71.7       61.9
  Male                   26.4       38.0        28.3       38.2
Age Group %
  65-74                  37.7       48.2        40.3       45.4
  75-84                  37.4       37.4        36.9       39.4
  85+                    24.9       14.5        22.9       15.3
Race/Ethnic Origin %
  Black                  15.3        4.0        15.5        6.5
  Hispanic               1.7         0.2        8.0         0.7
  Other                  5.0         1.0        11.3        2.1
  White                  77.9       94.8        65.2       90.7

Region                       South                  West

Enrollment                           Non-                   NonStatus
Dual (2)    dual (3)    Dual (2)   dual (3)

Total Enrollees        1,586,864   4,571,109   948,002    3,183,830

Dual + Non-dual             6,157,973              4,131,832

Characteristic

Gender %
  Female                 72.1        60.1        66.0       58.1
  Male                   27.9        39.9        34.0       41.9
Age Group %
  65-74                  43.3        51.7        44.2       48.4
  75-84                  37.4        35.9        38.6       37.3
  85+                    19.4        12.5        17.2       14.2
Race/Ethnic Origin %
  Black                  27.9         9.6        6.1         3.1
  Hispanic                8.5         1.3        14.3        1.9
  Other                   4.4         1.5        29.2        7.7
  White                  59.2        87.7        50.5       87.3

Region                       Other                  Total

Enrollment                          Non-                    NonStatus
Dual (2)   dual (3)   Dual (2)     dual (3)

Total Enrollees         2,885     286,203    4,087,132   13,640,634

Dual + Non-dual             289,088                17,727,766

Characteristic

Gender %
  Female                 60.4       57.9       70.9         60.4
  Male                   39.6       42.1       29.1         39.7
Age Group %
  65-74                  50.4       54.5       41.9         49.0
  75-84                  38.4       33.3       37.6         37.1
  85+                    11.2       12.2       20.5         13.8
Race/Ethnic Origin %
  Black                  9.3        6.8        18.1         6.2
  Hispanic               24.0       22.1        8.6         1.5
  Other                  19.1       4.4        11.7         3.0
  White                  47.6       66.7       61.7         89.3

Region

Enrollment
Status                    All

Total Enrollees        17,727,766

Dual + Non-dual

Characteristic

Gender %
  Female                  62.8
  Male                    37.2
Age Group %
  65-74                   47.4
  75-84                   37.2
  85+                     15.4
Race/Ethnic Origin %
  Black                   8.9
  Hispanic                3.2
  Other                   5.0
  White                   83.0

(1) A Medicare Part D enrollee is a beneficiary enrolled
for the entire 12 months in a Medicare Part D program
and who is eligible to receive Medicare Part D
prescription benefits whether or not the
individual received a prescription in 2007.

(2) Beneficiaries enrolled in Part D for the entire
twelve months as dual enrollees for calendar year 2007.

(3) Beneficiaries enrolled in Part D for the entire
twelve months as non-dual enrollees for calendar year 2007.

SOURCE: Medicare Part D Prescription Drug Event files,
Medicare Denominator file, and US Census Bureau.

Exhibit 5. Number of Medicare Part D Medication
Recipients (1) and Percent by Dual Eligibility Status,
Gender, Age Group, Origin, and Region, 2007. Age
65 and Over

                            Midwest                Northeast
Region
Enrollment                          Non-                   NonStatus
Dual (2)   dual (3)    Dual (2)   dual (3)

Total Recipients       672,029    2,915,818   786,330    2,290,672

Dual + Non-dual             3,587,847              3,077,002

Characteristic

Gender %
  Female                 74.8       62.9        72.7       62.6
  Male                   25.2       37.1        27.4       37.4
Age Group %
  65-74                  37.1       47.2        39.7       44.5
  75-84                  37.5       37.9        36.9       39.9
  85+                    25.5       14.9        23.4       15.6
Race/Ethnic Origin %
  Black                  15.1        3.7        15.2        6.1
  Hispanic               1.7         0.2        8.0         0.7
  Other                  4.9         1.0        11.3        2.0
  White                  78.3       95.1        65.5       91.2
Disease Burden (4) %
  High                   42.9       18.7        44.1       21.7
  MediumHigh
    31.9       22.8        32.7       25.0
  MediumLow
    19.5       25.6        19.0       24.7
  Low                    5.8        32.8        4.2        28.5
  Missing                0.0         0.0        0.0         0.0
Prescription
Burden (5) %
  High                   55.4       18.3       52 .8       15.5
  MediumHigh
    22.3       25.5        22.7       24.0
  MediumLow
    12.7       28.0        13.9       29.5
  Low                    9.6        28.1        10.6       31.0

                             South                  West
Region
Enrollment                           Non-                   NonStatus
Dual (2)    dual (3)    Dual (2)   dual (3)

Total Recipients       1,501,861   4,306,873   877,734    2,935,617

Dual + Non-dual             5,808,734              3,813,351

Characteristic

Gender %
  Female                 73.5        60.9        67.5        59
  Male                   26.5        39.1        32.5       41.0
Age Group %
  65-74                  42.8        51.2        43.6       47.5
  75-84                  37.5        36.2        38.8       37.9
  85+                    19.7        12.7        17.6       14.6
Race/Ethnic Origin %
  Black                  27.8         9.2        6.0         3.0
  Hispanic                8.4         1.3        13.8        1.8
  Other                   4.3         1.4        29.6        7.5
  White                  59.4        88.2        50.5       87.7
Disease Burden (4) %
  High                   42.2        20.2        35.4       18.5
  MediumHigh
    33.7        23.5        34.4       22.7
  MediumLow
    20.5        24.5        25.0       25.1
  Low                     3.6        31.8        5.2        33.7
  Missing                 0.0         0.0        0.0         0.0
Prescription
Burden (5) %
  High                   49.9        19.6        43.5       11.8
  MediumHigh
    24.4        27.7        23.8       21.2
  MediumLow
    14.8        27.7        17.2       30.3
  Low                    11.0        25.1        15.5       36.7

                             Other                  Total
Region
Enrollment                          Non-                    NonStatus
Dual (2)   dual (3)   Dual (2)     dual (3)

Total Recipients        5,033     262,557    3,842,987   12,711,537

Dual + Non-dual            267,590                16,554,524

Characteristic

Gender %
  Female                 60.8       59.3       72.2         61.2
  Male                   39.2       40.7       27.8         38.8
Age Group %
  65-74                  51.1       54.5       41.4         48.3
  75-84                  38.5       33.3       37.7         37.6
  85+                    10.4       12.2       21.0         14.2
Race/Ethnic Origin %
  Black                  7.8        6.5        18.0         5.9
  Hispanic               25.7       22.3        8.4         1.5
  Other                  21.0       4.1        11.7         2.9
  White                  45.5       67.1       61.9         89.8
Disease Burden (4) %
  High                   32.5       31.2       41.2         20.0
  MediumHigh
    34.4       31.6       33.3         23.6
  MediumLow
    27.0       20.9       21.1         24.9
  Low                    6.1        16.2        4.5         31.5
  Missing                0.0        0.1         0.0         0.0
Prescription
Burden (5) %
  High                   32.0       21.4       50.0         16.8
  MediumHigh
    24.5       27.4       23.5         25.0
  MediumLow
    21.3       24.6       14.8         28.6
  Low                    22.2       26.6       11.7         29.5

Region
Enrollment
Status                    All

Total Recipients       16,554,524

Dual + Non-dual

Characteristic

Gender %
  Female                  63.8
  Male                    36.2
Age Group %
  65-74                   46.7
  75-84                   37.6
  85+                     15.7
Race/Ethnic Origin %
  Black                   8.7
  Hispanic                3.1
  Other                   4.9
  White                   83.3
Disease Burden (4) %
  High                    24.9
  MediumHigh
    25.9
  MediumLow
    24.0
  Low                     25.3
  Missing                 0.0
Prescription
Burden (5) %
  High                    24.5
  MediumHigh
    24.7
  MediumLow
    25.4
  Low                     25.4

(1) A Medicare Part D recipient is a beneficiary
enrolled for the entire 12 months in a Medicare
Part D program and received at least one medication
in 2007.

(2) Beneficairies enrolled in Part D for the entire
twelve months as dual enrollees for calendar year 2007.

(3) Beneficiaries enrolled in Part D for the entire
twelve months as non-dual enrollees for calendar year 2007.

(4) Disease burden based on hierarchical condition
category risk scores acquired from the Centers for
Medicare & Medicaid Services. Derivation based on
quartiles.

(5) Prescription burden based on total number of
prescriptions filled per enrollee during calendar
year 2007. Derivation based on quartiles.

SOURCE: Medicare Part D Prescription Drug Event
files, Medicare Denominator file, and US Census
Bureau.

Exhibit 6. Number of Medicare Part D Beers Medication
(1) Recipients (2) based on the 2003 Fick et al. List
(3) and Percent by Dual Eligibility Status, Gender,
Age Group, Origin, and Region, 2007. Age 65 and Over

Region                       Midwest               Northeast
Enrollment
Status                               Non-                  NonDual
                        (4)   dual (5)   Dual (4)   dual (5)

Total Recipients        266,321    896,928    275,215    595,704

Dual + Non-dual               1,163,249             870,919

Characteristic

Gender %
  Female                  77.8       69.3       74.6       67.6
  Male                    22.2       30.7       25.4       32.5
Age Group %
  65-74                   40.6       47.2       42.9       44.9
  75-84                   36.8       38.2       36.6       40.3
  85+                     22.7       14.6       20.5       14.8
Race/Ethnic Origin %
  Black                   15.4       3.8        14.6       6.2
  Hispanic                1.6        0.2        8.8        0.8
  Other                   4.4        0.8        11.0       1.7
  White                   78.6       95.2       65.6       91.3
Disease
Burden (6)
  High                    48.7       22.9       48.6       26.2
  MediumHigh
    30.5       24.0       31.5       25.7
  MediumLow
    16.2       23.9       16.2       22.7
  Low                     4.6        29.1       3.7        25.4
  Missing                 0.0        0.0        0.0        0.0
Prescription
Burden (7)
  High                    68.2       28.1       65.3       23.8
  MediumHigh
    19.1       30.5       19.9       29.6
  MediumLow
    8.5        26.1       9.9        28.8
  Low                     4.2        15.3       4.9        17.9

Region                       South                  West
Enrollment
Status                               Non-                   NonDual
                        (4)   dual (5)    Dual (4)   dual (5)

Total Recipients        723,836    1,643,208   373,637    965,582

Dual + Non-dual             2,367,044              1,339,219

Characteristic

Gender %
  Female                  76.8       67.6        70.5       67.0
  Male                    23.2       32.4        29.5       33.0
Age Group %
  65-74                   45.0       51.6        45.4       47.2
  75-84                   36.7       36.4        38.5       38.5
  85+                     18.3       12.1        16.1       14.4
Race/Ethnic Origin %
  Black                   26.7        9.0        6.3        3.0
  Hispanic                8.0         1.3        14.2       2.0
  Other                   3.7         1.2        27.3       5.6
  White                   61.6       88.5        52.2       89.4
Disease
Burden (6)
  High                    47.1       23.9        40.1       22.5
  MediumHigh
    32.4       24.4        33.7       24.0
  MediumLow
    17.5       23.0        21.7       23.7
  Low                     3.1        28.7        4.5        29.8
  Missing                 0.0         0.0        0.0        0.0
Prescription
Burden (7)
  High                    61.4       29.1        54.9       19.1
  MediumHigh
    22.6       32.3        23.5       28.1
  MediumLow
    10.9       25.1        13.8       31.5
  Low                     5.2        13.5        7.8        21.3

Region                       Other                  Total
Enrollment
Status                               Non-                   NonDual
                        (4)   dual (5)   Dual (4)    dual (5)

Total Recipients         2,015     135,634    1,641,024   4,237,056

Dual + Non-dual             137,649                5,878,080

Characteristic

Gender %
  Female                  64.0       62.9       75.2        67.7
  Male                    36.0       37.1       24.9        32.3
Age Group %
  65-74                   51.6       55.8       44.0        48.9
  75-84                   37.8       32.8       37.1        37.7
  85+                     10.6       11.4       18.8        13.5
Race/Ethnic Origin %
  Black                   6.7        6.2        18.1         6.1
  Hispanic                27.5       22.9        8.6         1.9
  Other                   17.8       3.7        10.4         2.3
  White                   48.0       67.2       62.9        89.8
Disease
Burden (6)
  High                    38.0       35.9       46.0        24.1
  MediumHigh
    34.6       32.6       32.2        24.7
  MediumLow
    22.7       18.6       18.0        23.1
  Low                     4.7        12.9        3.8        28.1
  Missing                 0.0        0.1         0.0         0.0
Prescription
Burden (7)
  High                    42.2       28.5       61.7        25.8
  MediumHigh
    26.4       31.3       21.8        30.6
  MediumLow
    19.1       23.5       11.0        27.2
  Low                     12.4       16.8        5.6        16.4

Region
Enrollment
Status
                           All

Total Recipients        5,878,080

Dual + Non-dual

Characteristic

Gender %
  Female                  69.8
  Male                    30.3
Age Group %
  65-74                   47.5
  75-84                   37.5
  85+                     15.0
Race/Ethnic Origin %
  Black                    9.4
  Hispanic                 3.7
  Other                    4.6
  White                   82.3
Disease
Burden (6)
  High                    30.2
  MediumHigh
    26.8
  MediumLow
    21.7
  Low                     21.3
  Missing                  0.0
Prescription
Burden (7)
  High                    35.8
  MediumHigh
    28.1
  MediumLow
    22.7
  Low                     13.4

(1) Beers, M. H., Ouslander, J. G., Rollingher,
I., Reuben, D. B., Brooks, J., and Beck, J. C.
(1991). Explicit criteria for determining
inappropriate medication use in nursing home
residents. Archives of
Internal Medicine, 151(9),1825-1832. Legend
drugs considered inappropriate based on dose
were excluded from this analysis.

(2) A Medicare Part D Beers medication recipient
is a Part D enrollee who received at least one
Beers drug in 2007.

(3) Fick, D. M., Cooper, J. W., Wade, W. E.,
Waller, J. L., Maclean, J. R., and Beers, M. H.
(2003). Updating the Beers Criteria for Potentially
Inappropriate Medication Use in Older Adults. Archives of
Internal Medicine. 163, 2716-2724.

(4) Beneficiaries enrolled in Part D for the entire
twelve months as dual enrollees for calendar year 2007.

(5) Beneficiaries enrolled in Part D for the entire
twelve months as non-dual enrollees for calendar year 2007.

(6) Disease burden based on hierarchical condition
category risk scores acquired from the Centers for
Medicare & Medicaid Services. Derivation based on
quartiles.

(7) Prescription burden based on total number of
prescriptions filled per enrollee during calendar
year 2007. Derivation based on quartiles.

SOURCE: Medicare Part D Prescription Drug Event
files, Medicare Denominator file, and US Census
Bureau.


Acknowledgements

The authors received input and guidance from the following individuals in the development of this article (in alphabetical order by last name): Bill Clark, Renee Mentnech, and Noemi Rudolph. The article was substantially improved by the contributions of these individuals.

Financial Disclosure

This research was funded internally within the Centers for Medicare & Medicaid Services (CMS). The views and opinions expressed in this article are those of the authors and do not necessarily reflect the views of CMS.

References

Beers, M. H. (1997). Explicit criteria for determining potentially inappropriate medication use by the elderly. Archives of Internal Medicine., 157(14), 1531-1536. PubMed

Blackwell, S. A., Baugh, D. K., Montgomery, M. A., Ciborowski, G. M., & Levy, J. M. (2011). National study of injury-increasing analgesics in the elderly. International Journal of Pharmaceutical and Healthcare Marketing, 5(1), 25-42. http://dx.doi.org/10.1108/17506121111121569

Blackwell, S. A., Ciborowski, G. M., Baugh, D. K., & Montgomery, M. A. (2008). Applying the 2003 Beers update to Medicaid/Medicare enrollees. International Journal of Pharmaceutical and Healthcare Marketing, 2(4), 257-272. http://dx.doi.org/10.1108/17506120810922312

Bootman, J. L., Harrison, D. L., & Cox, E. (1997). The health care cost of drug- related morbidity and mortality in nursing facilities. Archives of Internal Medicine, 15718), 2089-2096. PubMed http://dx.doi.org/10.1001/archinte.1997.00440390083011

Chin, M. H., Wang, L. C., Jin, L., Muliken, R., Walter, J., Hayley, D. C., ... Friedmann, P. D. (1999). Appropriateness of medication selection for older persons in an urban academic emergency department. Academic Emergency Medicine, 6(12), 1232-1242. PubMed http://dx.doi.org/10.1111/j.1553-2712.1999.tb00139.x

Fick, D. M., Waller, J. L., Maclean, J. R., Vanden Heuvel, R., Tadlock, J. G., Gottlieb, M., & Cangialose, C. B. (2001). Potentially inappropriate medication use in a Medicare managed care population: association with higher costs and utilization. Journal of Managed Care Pharmacy, 7(5), 407-413.

Fick, D. M., Cooper, J. W., Wade, W. E., Waller, J. L., Maclean, J. R., & Beers, M. H. (2003). Updating the Beers Criteria for Potentially Inappropriate Medication Use in Older Adults. Archives of Internal Medicine, 163(22), 2716-2724. PubMed http://dx.doi.org/10.1001 /archinte.163.22.2716

Gallagher, P. F., Barry, P. J., Ryan, C., Hartigan, I., & O'Mahony, D. (2008). Inappropriate prescribing in an acutely ill population of elderly patients as determined by Beers' criteria. Age and Ageing, 37(1), 96-101. PubMed http://dx.doi.org/10.1093/ageing/afm116

Hanlon, J. T., Schmader, K. E., Koronkowski, M. J., Weinberger, M., Landsman, P. B., Sams, G. P., & Lewis, I. K. (1997). Adverse drug events in high risk older outpatients. Journal of the American Geriatrics Society, 45(8), 945-948. PubMed

Li, P., Kim, M. M., & Doshi, J. A. (2010). Comparison of the performance of the CMS Hierarchical Condition Cateory (CMS-HCC) risk adjuster with the charlson and elixhauser comorbidity measures in predicting mortality. BioMed Health Services Research., 10, 245. http://dx.doi.org/10.1186/1472-6963-10-245

Meredith, S., Feldman, P. H., Frey, D., Hall, K., Arnold, K., Brown, N. J., & Ray, W. A. (2001). Possible medication errors in home healthcare patients. Journal of the American Geriatrics Society, 49(6), 719-724. PubMed http://dx.doi.org/10.1046/j.1532- 5415.2001.49147.x

Mort, J. R., & Aparasu, R. R. (2000). Prescribing potentially inappropriate psychotropic medications to the ambulatory elderly. Archives of Internal Medicine, 160(18), 2825-2831. PubMed http://dx.doi.org/10.1001/archinte.160.18.2825

Piecoro, L. T., Browning, S. R., Prince, T. S., Ranz, T. T., & Scutchfield, F. D. (2000). A database analysis of potentially inappropriate drug use in an elderly Medicaid population. Pharmacotherapy, 20(2), 221-228. PubMed http://dx.doi.org/10.1592/phco.20.3.221.34779

ResDAC. (2009). CMS Denominator File (CCW Version). Retrieved from http://www.resdac.umn.edu/ddvh/CMS_Part_D_Denominator_File_jul_2009.pdf

Rothberg, M. B., Pekow, P. S., Liu, F., Korc-Grodzicki, B., Brennan, M. J., Bellantonio, S., ... Lindenauer, P. K. (2008). Potentially inappropriate medication use in hospitalized elders. Journal of Hospital Medicine, 3(2), 91-102. PubMed http://dx.doi.org/10.1002/jhm.290 SAS Institute Inc. (2010). Retrieved from http://www.sas.com/

Shrank, W. H., Polinski, J. M., & Avorn, J. (2007). Quality indicators for medication use in vulnerable elders. [Supplemental material]. Journal of the American Geriatrics Society, 55, S373-S382. PubMed http://dx.doi.org/10.1111/j.1532-5415.2007.01345.x

The Henry J. Kaiser Family Foundation. (2004). Dual eligibles: Medicaid's role for low-income Medicare beneficiaries. Kaiser Commission on Medicaid and the Uninsured. Retrieved http://www.kff.org/medicaid/upload/Dual-Eligibles-Medicaid-s-Role-for-Low- IncomeMedicare-Beneficiaries-Fact-Sheet-2.pdf

The Henry J. Kaiser Family Foundation. (2011). The role of Medicare for the people dually eligible for Medicare and Medicaid. Retrieved from http://www.kff.org/medicare/upload/8138.pdf

U.S. Department of Economics and Statistics Administration, U.S. Census Bureau. (2010). Census regions and divisions of the United States [Map and Divisions/Regions with state FIPS codes]. Retrieved from http://www.census.gov/geo/www/us_regdiv.pdf

U.S. General Accounting Office, Health, Education, and Human Services Division. (July 1995). Prescription drugs and the elderly: Many still receive potentially harmful drugs despite recent improvements(GAO/HEHS-95-152).

U.S. Department of Health and Human Services Centers for Medicare & Medicaid Services. (2011). Evaluation of the CMS-HCC Risk Adjustment Model: Final Report. Retrieved from https://www.cms.gov/MedicareAdvtgSpecRateStats/downloads/Evaluation_Risk_Adj_Mod el_2011.pdf

Wolters Kluwer Health. (2010). MDDB[R] [Database]. Medi-Span[R]. Retrieved from http://www.medispan.com/marketing/ContentPage.aspx?contentId=09e0f1ed-80a9-4a87 8bc9-ad778d7b6615

Zhan, C., Sangl, J., Bierman, A. S., Miller, M. R., Friedman, B., Wickizer, S. W., & Meyer, G. S. (2001). Potentially inappropriate medication use in the community-dwelling elderly: Findings from the 1996 Medical Expenditure Panel Survey. Journal of the American Medical Association, 286(22), 2823-2829. PubMed http://dx.doi.org/10.1001/jama.286.22.2823

Steven A. Blackwell, Melissa A. Montgomery, Dave K. Baugh, Gary M. Ciborowski, and Gerald F. Riley

Department of Health and Human Services--Centers for Medicare & Medicaid Services
Exhibit 1. Number (%) of Medicare Part D Enrollees (1),
Part D Medication Recipients (2), and Part D Beers
Medication Recipients (3) by Dual Eligibility Status,
Gender, Age Group, Race/Origin, and Recipient
Disease/Prescription Burden, 2007, Age 65 and Over

                                              Medicare Part D
Enrollment            Medicare Part              Medication
Status                 D Enrollees               Recipients

                  Dual (4)     Non-Dual     Dual (4)     Non-Dual
                                 (5)                       (5)

Total            4,087,132    13,640,634   3,842,987    12,711,537
  Enrollees
Dual +           17,727,766                16,554,524
  Non-Dual

Characteristic

Gender
  Female         2,896,156    8,231,834    2,773,385    7,782,019
                       71%          60%          72%          61%
  Male           1,190,976    5,408,800    1,069,602    4,929,518
                       29%          40%          28%          39%
Age Group
  65-74          1,712,519    6,689,743    1,589,583    6,135,394
                       42%          49%          41%          48%
  75-84          1,535,344    5,063,791    1,447,053    4,776,125
                       38%          37%          38%          38%
  85+              839,269    1,887,100      806,351    1,800,018
                       21%          14%          21%          14%
Race/Ethnic
Origin
  Black            739,692      838,651      692,531      748,908
                       18%           6%          18%           6%
  Hispanic         349,581      208,963      323,349      187,500
                        9%           2%           8%           1%
  Other            476,511      409,429      447,666      365,682
                       12%           3%          12%           3%
  White          2,521,348    12,183,591   2,379,441    11,409,447
                       62%          89%          62%          90%
Region
  Midwest          715,288    3,136,893      672,029    2,915,818
                       18%          23%          17%          23%
  Northeast        834,093    2,462,599      786,330    2,290,672
                       20%          18%          20%          18%
  South          1,586,864    4,571,109    1,501,861    4,306,873
                       39%          34%          39%          34%
  West             948,002    3,183,830      877,734    2,935,617
                       23%          23%          23%          23%
  Other              2,885      286,203        5,033      262,557
                        0%           2%           0%           2%
Disease
Burden (6)
  High                  --           --    1,581,353    2,535,709
                                                 41%          20%
  Medium-High           --           --    1,280,618    3,001,627
                                                 33%          24%
  Medium-Low            --           --      809,137    3,161,885
                                                 21%          25%
  Low                   --           --      171,836    4,008,857
                                                  5%          32%
  Missing               --           --           43         3459
                                                  0%           0%
Prescription
Burden (7)
  High                  --           --    1,920,713    2,136,493
                                                 50%          17%
  Medium-High           --           --      904,573    3,178,394
                                                 24%          25%
  Medium-Low            --           --      568,285    3,641,130
                                                 15%          29%
  Low                   --           --      449,416    3,755,520
                                                 12%          30%

                     Medicare Part
Enrollment         D Beers Medication
Status                Recipients

                 Dual (4)    Non-Dual
                                (5)

Total            1,641,024   4,237,056
  Enrollees
Dual +           5,878,080
  Non-Dual

Characteristic

Gender
  Female         1,233,222   2,867,030
                      75%         68%
  Male            407,802    1,370,026
                      25%         32%
Age Group
  65-74           722,766    2,069,865
                      44%         49%
  75-84           609,035    1,595,984
                      37%         38%
  85+             309,223     571,207
                      19%         13%
Race/Ethnic
Origin
  Black           297,600     256,796
                      18%          6%
  Hispanic        140,500      78,236
                       9%          2%
  Other           171,045      96,665
                      10%          2%
  White          1,031,879   3,805,359
                      63%         90%
Region
  Midwest         266,321     896,928
                      16%         21%
  Northeast       275,215     595,704
                      17%         14%
  South           723,836    1,643,208
                      44%         39%
  West            373,637     965,582
                      23%         23%
  Other             2,015     135,634
                       0%          3%
Disease
Burden (6)
  High            754,683    1,020,108
                      46%         24%
  Medium-High     528,711    1,046,342
                      32%         25%
  Medium-Low      296,085     980,086
                      18%         23%
  Low              61,527    1,188,915
                       4%         28%
  Missing              18       1,605
                       0%          0%
Prescription
Burden (7)
  High           1,011,792   1,094,577
                      62%         26%
  Medium-High     357,443    1,294,605
                      22%         31%
  Medium-Low      180,503    1,153,533
                      11%         27%
  Low              91,286     694,341
                       6%         16%

(1) A Medicare Part D enrollee is a beneficiary
enrolled for the entire 12 months in a Medicare
Part D program and who is eligible to receive
Medicare Part D prescription benefits, whether
or not the individual received a prescription in 2007.

(2) A Medicare Part D recipient is a beneficiary
enrolled for the entire 12 months in a Medicare
Part D program and received at least one
medication in 2007.

(3) A Medicare Part D Beers medication recipient
is a Part D enrollee who received at least one
Beers drug in 2007, based on the Fick update
(Fick et al., 2003). Legend drugs considered
nappropriate based on dose were excluded from
this analysis.

(4) Beneficiaries enrolled in Part D for the
entire twelve months as dual enrollees for
calendar year 2007.

(5) Beneficiaries enrolled in Part D for the
entire twelve months as non-dual enrollees for
calendar year 2007.

(6) Disease burden based on hierarchical
condition category risk scores acquired from
the Centers for Medicare & Medicaid Services.
Derivation based on quartiles.

(7) Prescription burden based on total number
of prescriptions filled per enrollee during
calendar year 2007. Derivation based on quartiles.

SOURCE: Medicare Denominator file.

Exhibit 2: Individual-Level Multivariate Logistic
Regression Model Predicting Beers (1) Drug Use for Medicare
Part D Medication Recipients (2) Age 65 and Over, 2007

                                 Odds Ratio               Wald
                                 (3,4) Point
Characteristic                     Estimate    Chi-Square    p value

Region
                 Midwest           1.2 17       12408.699    <0.0001
                 South              1.685      108646.680    <0.0001
                 West               1.582       69182.337    <0.0001
                 Other              2.857       58786.921    <0.0001
                 Northeast          1.0Dual
Enrollee
Status
                 Dual Enrollee     1.0 23       255. 526     <0.0001
                 Non-Dual
                 Enrollee           1.0Age
                 85+                0.647       66689.701    <0.0 001
                 75-84              0.834       21911.636    <0.0001
                 65-74              1.0Race/Ethnic
Origin
                 Black             0.9 60       447. 149     <0.0 001
                 Hispanic           1.100        923.360     <0.0001
                 Other              0.835       4639.998     <0.0001
                 White              1.0Gender
                 Female            1.3 49       65217.890    <0.0 001
                 Male               1.0Disease
Burden (5)
                 High              1.1 47       5872 .231    <0.0 001
                 MediumHigh
                 1.030        310.365     <0.0001
                 Medium-Low         0.954        809.500     <0.0001
                 Low                1.0Prescription
Burden (6)
                 High              4.5 31      69135 3.773   <0.0 001
                 MediumHigh
                 2.909      339165.765    <0.0001
                 Medium-Low         2.013      176669.525    <0.0001
                 Low                1.0
                                    95% Wald
                                   Confidence
Characteristic                       Limits

Region
                 Midwest         1.213    1.221
                 South           1.680    1.690
                 West            1.577    1.588
                 Other           2.833    2.882
                 Northeast
Dual Enrollee
Status
                 Dual Enrollee   1.020    1.026
                 Non-Dual
                 Enrollee
Age
                 85+             0.644    0.649
                 75-84           0.832    0.836
                 65-74
Race/Ethnic
Origin
                 Black           0.9 56   0.9 63
                 Hispanic        1.093    1.107
                 Other           0.831    0.840
                 White
Gender
                 Female          1.3 46   1.3 52
                 Male
Disease
Burden (5)
                 High            1.1 43   1.1 51
                 MediumHigh
                 1.027    1.034
                 Medium-Low      0.951    0.957
                 Low
Prescription
Burden (6)
                 High            4.5 15   4.5 47
                 MediumHigh
                 2.899    2.918
                 Medium-Low      2.007    2.020
                 Low

(1) Legend drugs considered inappropriate based
on dose were excluded from this analysis (Fick et
al., 2 003).

(2) A Medicare Part D medication recipient is a
Part D enrollee who received at least one
medication in 2007.

(3) Max-rescaled [R.sup.2] = 0.1125. c = 0.673.

(4) Higher ratio (>1) = greater odds for
receiving a Beers medication.

(5) Disease burden based on hierarchical
condition category risk scores acquired
from the Centers for Medicare & Medicaid
Services. Derivation based on quartiles.

(6) Prescription burden based on total number
of prescriptions filled per enrollee during
calendar year 2007. Derivation based on quartiles.

SOURCE: Medicare Part D Prescription Drug
Event and Medicare Denominator files.

Exhibit 3. Beers (1) Filled Prescriptions as a
Percentage of All Filled Prescriptions (2) by
Therapeutic Category (3), 2007.

Based on Dual Enrollee Status for Part
D Medication Recipients (4) Age 65 and Over

                                       Dual Enrollee
                                   Medication Recipients

                                               Percentage of
                               Total Number    Total Filled
                                    of           that were
                               Prescriptions       Beers
Therapeutic Category (5)          Filled       Prescriptions

Analgesics and Anesthetics       23,501,291         6.4%
Cardiovascular Agents           110,069,016         2.9%
Central Nervous Syst. Drugs      33,450,681        10.5%
Endocrine & Metabolic Drugs      41,369,345         1.8%
Gastrointestinal Agents          24,397,356         2.5%
Genitourinary Products            6,791,797        21.1%
Hematological Agents             12,047,463         1.0%
Neuromuscular Drugs              11,231,617        11.3%
Respiratory Agents               15,662,217         6.5%

All Categories                  278,520,783         4.8%

                                     Non-Dual Enrollee
                                   Medication Recipients

                                               Percentage of
                               Total Number    Total Filled
                                    of           that were
                               Prescriptions       Beers
Therapeutic Category (5)          Filled       Prescriptions

Analgesics and Anesthetics       32,354,985         6.7%
Cardiovascular Agents           209,670,439         2.6%
Central Nervous Syst. Drugs      37,154,453        15.2%
Endocrine & Metabolic Drugs      71,454,498         4.0%
Gastrointestinal Agents          25,551,378         3.6%
Genitourinary Products           12,115,859        19.9%
Hematological Agents             18,478,252         1.9%
Neuromuscular Drugs              12,483,831        16.2%
Respiratory Agents               19,504,294         4.8%

All Categories                  438,767,989         5.2%

(1) Legend drugs considered inappropriate based on
dose were excluded from this analysis (Fick et al., 2003).

(2) Filled prescriptions include both original
prescriptions and refills.

(3) Medi-Span[R] is a product of Wolters Kluwer
Health. See http://www.wkhealth.com for details.

(4) A Medicare Part D medication recipient is a
Part D enrollee who received at least one
medication in 2007.

(5) All Beers drugs included in this analysis were
classified into one of these nine categories
(Blackwell et al., 2008).

SOURCE: Medicare Part D Prescription Drug Event
and Medicare Denominator files.
COPYRIGHT 2012 Centers for Medicare & Medicaid Services
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2012 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Blackwell, Steven A.; Montgomery, Melissa A.; Baugh, Dave K.; Ciborowski, Gary M.; Riley, Gerald F.
Publication:Medicare & Medicaid Research Review
Article Type:Report
Date:Apr 1, 2012
Words:8370
Previous Article:Assessing the usability of max 2008 encounter data for comprehensive managed care.
Next Article:Wisconsin's experience with Medicaid auto-enrollment: lessons for other states.
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters