Research in Social and Administrative Pharmacy 10 (2014) 494–507

Original Research

Quality of psychopharmacological medication use in nursing home residents Linda Simoni-Wastila, Ph.D.a, Yu-Jung Wei, Ph.D.a,*, Mario Luong, Pharm.D.a, Christine Franey, M.P.H.a, Ting-Ying Huang, B.S.a, Gail B. Rattinger, Pharm.D., Ph.D.a,b, Ilene H. Zuckerman, Pharm.D., Ph.D.a, Nicole Brandt, Pharm.D.c, Judith A. Lucas, Ed.D., R.N., G.C.N.S.-B.C.d,e a

Department of Pharmaceutical Health Services Research, University of Maryland School of Pharmacy, Baltimore, MD, USA b Fairleigh Dickinson University School of Pharmacy, Florham Park, NJ, USA c Department of Pharmacy Practice and Science, University of Maryland School of Pharmacy, Baltimore, MD, USA d Department of Behavioral and Community Health, Seton Hall University College of Nursing, South Orange, NJ, USA e Institute for Health, Health Care Policy, and Aging Research, Rutgers University, New Brunswick, NJ, USA

Abstract Background: Despite well-documented evidence regarding antipsychotic use in older adults residing in nursing homes (NHs), there is a lack of evidence-based use and quality benchmarks for other psychopharmacological medications (PPMs), including antidepressants, anxiolytics, and sedativehypnotics. Objective: To estimate the prevalence and patterns of use of PPMs and to measure the quality of PPM use. Methods: Using a 5% random sample of 2007 Medicare claims data linked to the Minimum Data Set 2.0, this cross-sectional study identified a nationally representative sample of 69,832 NH residents with R3 months of institutionalization. This study measured 1-year prevalence and quality of PPM use, as assessed by indication, dose, and duration of use defined and operationalized according to the current Centers for Medicare and Medicaid Services Unnecessary Medication Guidance for Surveyors and relevant practice guidelines. Results: Over two-thirds of residents (72.1%, n¼50,349) used R1 PPM in 2007, with the highest prevalence seen in antidepressants (59.4%), and the lowest in anxiolytics (8.9%). Almost two-thirds (61.0%) of PPM users used R2 PPM classes. Compared to other PPM therapeutic classes, antipsychotic users had greatest evidence of guideline adequate use by indication (95.8%) and dose (78.7%). In addition, longer duration of adequate treatment was observed among antipsychotic users (mean ¼ 208 days, standard deviation [SD] ¼ 118) as compared to anxiolytic (mean ¼ 159 days, SD ¼ 118) and sedative-hypnotic users (mean ¼ 183 days, SD ¼ 117). Conclusions: This study found that PPM use remains highly prevalent among long-stay Medicare NH residents. While antipsychotic use remained high (31.5%), little antipsychotic use was deemed inadequate by indication. However, the 1-year prevalence of use, dose, and duration of use of other PPMs remain high * Corresponding author. Tel.: þ1 410 706 1074; fax: þ1 410 706 5394. E-mail address: [email protected] (Y-J. Wei). 1551-7411/$ - see front matter Ó 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.sapharm.2013.10.003

Simoni-Wastila et al. / Research in Social and Administrative Pharmacy 10 (2014) 494–507

495

and potentially inadequate. Practitioners and policy-makers should heed both the high use and lower prescribing quality of antidepressants, anxiolytics, and sedative-hypnotics in NH residents. Ó 2014 Elsevier Inc. All rights reserved. Keywords: Quality of medication use; Psychopharmacological medications; Nursing homes

Introduction The Institute of Medicine defines the quality of health care as “the degree to which health services for individuals and populations increase the likelihood of desired health outcomes and are consistent with current professional knowledge.”1 According to this widely accepted definition, the quality of health care in the United States (U.S.) continues to fall short of the mark.2 A primary reason for suboptimal health care quality is inappropriate medication use, a problem that results in unnecessary and largely preventable excess morbidity, mortality, and costs.3 Suboptimal medication use causes significant morbidity and mortality among individuals residing in nursing homes (NHs).3–5 Of particular importance is the poor prescribing of psychopharmacological medications (PPMs), such as antipsychotics, due to their potential use as chemical restraints.6 This concern resulted in passage of the 1987 Nursing Home Reform Act to ensure the quality of care and medications received by NH residents.7 PPM use in NHs, however, continued to draw attention when antipsychotic use was linked to increased risks of death8–12 and disability due to stroke and other adverse effects among NH residents with dementia.13 These findings led the U.S. Food and Drug Administration (FDA) to issue black-box warnings regarding antipsychotic use in dementia patients.14 Other research had since documented the evidence of antipsychotic use with an increased mortality in the general NH population.5,8,15,16 Accordingly, in 2012 the Centers for Medicare & Medicaid (CMS) declared an initiative to reduce antipsychotic use among NH residents by 15% by the end of the year.17 Recognizing the importance of appropriate medication use in achieving optimal resident outcomes, in 2006, CMS updated the F-329 Interpretive Guidance (Unnecessary Medication Guidance [UMG]) to include a comprehensive medication review to assist facility surveyors looking at indication, dose, duration, and other aspects of medications.18 Additional tools, including the updated Beers Criteria, provide further attention to

potentially inappropriate medications and associated risks in older adults.8 Yet, even with prescribing criteria, the use of antipsychotics and other “high risk” medications in nursing home residents remains both controversial and common.4,19–22 Despite current national interest in antipsychotic prescribing quality, there remains a paucity of data establishing national rates of usedand quality of usedin nursing facilities. Even less is understood about the use of other potentially problematic PPMs, such as anxiolytics, antidepressants, and sedative-hypnotics. The few available studies focus on small, clinical or site-specific samples of nursing facilities and/or rely on data which lack dose, duration, and other information necessary to ascertain prescribing quality.19,22–25 Indeed, the only U.S. nursing home metric of PPM quality is collected by Nursing Home Compare, which reports state- and facility-level rates of antipsychotic prevalence in long- and shortstay residents in nursing facilities.26 These estimates, derived from the Minimum Data Set (MDS), lack detailed information, including specific agent administered, dose, duration, and indication, required to assess prescribing.17 Using nationally representative data of Medicare beneficiaries residing in long-term care (LTC) facilities, this study aimed to: 1) estimate the 1-year prevalence and patterns of use of PPMs, including antipsychotics, antidepressants, anxiolytics, and sedative-hypnotics; and 2) measure the quality of PPM use by operationalizing the CMS UMG for Surveyors criteria. Methods Study design and source This descriptive, cross-sectional study used a 5% random sample of U.S. Medicare beneficiaries from the 2006–2007 Chronic Condition Data Warehouse (CCDW) data linked to MDS 2.0 files to examine PPM use among beneficiaries residing in LTC settings. The latest available at the time of this study, the CCDW data contain detailed claims data for all Medicare Part A (inpatient),

496

Simoni-Wastila et al. / Research in Social and Administrative Pharmacy 10 (2014) 494–507

Part B (physician/supplier), and Part D Prescription Drug Event (PDE) services. Each Part A and B claim includes up to 10 and 8 diagnoses, respectively, coded using the International Classification of Disease, Ninth Revision, Clinical Modification (ICD-9-CM); Part D PDEs include information on drug name, fill date, days’ supply, quantity, and strength. The MDS, a federally-mandated assessment tool for patients residing in Medicare and/or Medicaid certified facilities, collects clinical data including cognitive and behavioral symptoms, physical functioning, and diagnoses. Each resident receives an MDS assessment at admission, quarterly intervals, and upon major changes in status.27 Study data were obtained through a data use agreement with CMS. The University of Maryland, Baltimore Institutional Review Board approved the conduct of this study (HP-00046873) and waived the requirement for written informed consent. Study sample The study cohort included a sample of Medicare nursing home residents, defined as beneficiaries with: 1) R1 MDS assessment record in 2007; 2) enrollment in Medicare Parts A, B, and D in any month during 2007; and 3) a long stay of 3 consecutive months or longer in nursing facilities. Residents with short stays (i.e., !3 consecutive months) were excluded from the sample because payments for drugs prescribed over the short stay was bundled into the Medicare Part A payment, which prevents discernment of short-stay drug use through the CCDW administrative data. To distinguish long stays, an algorithm was developed by combining evidence from Part A Skilled Nursing Facility (SNF) claims and MDS data for 2006 and 2007. Using service dates reported in both datasets, this MDS/SNF algorithm began by constructing NH episodes that started with admission (SNF or MDS admission date, whichever came first) and ended with discharge (MDS discharge date) or death. Information on mortality was obtained from the CCDW Beneficiary Summary files. For residents who entered facilities before 2006, the first NH episode was calculated from January 1, 2006. The length of NH stay for each episode was then calculated and used to determine long-stay residents who had at least one NH episode lasting at least 101 consecutive days based on the CMS definition of long stay.28 The MDS/SNF algorithm was validated against the Part B carrier claims to compare the performance in identifying NH residents. Although the

Part B data have been commonly used to detect whether individuals resided in NHs,29–32 they are inferior to MDS/SNF data combined due to the Medicare reimbursement policy that limits the quantities of Part B services covered during NH stays. This limitation also restricts the ability of Part B data to track the length of NH stay for residents. The validation analysis showed the MDS/ SNF algorithm captured 100% of residents with confirmatory NH evidence from the Part B data. Additionally, the algorithm captured 9606 residents (13.2% of 72,774 eligible samples) that were failed to detect with the Part B evidence alone (data not shown). From 72,774 eligible beneficiaries, patients were excluded if they had: 1) Medicare Advantage/Health Maintenance Organization insurance due to lack of medical claims for these beneficiaries (n ¼ 2232); and 2) no Part A and B claims data in the 6 months (July–December 2006) prior to the study period (n ¼ 710). A six month pre-index period was used to detect clinical conditions and/ or symptoms that may be adequate indications for PPMs administered in 2007.33 After applying exclusion criteria, a final cohort of 69,832 Medicare long-stay nursing home residents was derived. Psychopharmacological medications PPMs assessed in this study included four therapeutic drug classes: antipsychotics, antidepressants, anxiolytics, and sedative-hypnotics (Supplemental Table). Scheduled and pro re nata (prn) PPMs filled during 2007 were captured through Part D Prescription Drug Event file using National Drug Codes. A total of 98 distinct PPMs were assessed; 15 PPMs were classified in two of four therapeutic classes. The 1-year prevalence of PPM use was defined as the proportion of residents in the sample with R1 prescription fills for PPMs of interest during 2007. Among PPM users, this study also examined patterns of polyuse across therapeutic class, defined as use of R2 different therapeutic classes. Due to therapeutic interchangeability of many individual medications within pharmacologic classes, concurrent use of multiple individual drugs (i.e., polypharmacy) was not examined. Psychopharmacological medication quality indicators Psychopharmacological medication quality indicators (PMQIs) were assessed on three domainsd use by indication, dose, and durationdas defined

Simoni-Wastila et al. / Research in Social and Administrative Pharmacy 10 (2014) 494–507

in the UMG,18 and operationalized using information from FDA labeling, practice guidelines and clinical literature relevant to the geriatric population.18,22,34–38 Quality of indication among users of each therapeutic class was assessed by whether they had evidence of at least one indication deemed adequate per the UMG (Supplemental Table). Indications were captured during an 18-month period (July 2006–December 2007) using ICD-9-CM codes in Parts A and B claims, as well as MDS 2.0 fields (Supplemental Table). In particular, MDS fields were used to capture syndromes and/or symptoms less often recorded in claims data, such as delirium or behavioral symptoms in dementia. For example, followed by prior studies,39,40 depressed patients were defined if they had an administrative claim diagnosis of depression (ICD-9-CM codes 296.2x, 296.3x, 300.4, 311.xx) or MDS 2.0 diagnosis plus a Depression Rating Scale score of R3. Finally, to assess quality of use among residents utilizing R2 PPMs (n ¼ 50,349), this study assessed indication quality as: 1) fully adequate (all PPM classes used had evidence of at least one adequate indication); 2) partially adequate (at least one PPM class used had evidence of an adequate indication); and 3) fully inadequate (no PPM classes used had evidence of any adequate indications). These three categories were determined based on the nature distribution of indication quality measure among the study sample. To the authors’ knowledge, there is no standard way to categorize indication quality among users with multiple drug classes. Dose and duration of use were estimated only among users with evidence of adequate indications for use (n ¼ 42,655). Dose exposure was operationalized with a modified standardized daily dose (mSDD)23,24,25 which compared the average daily dose for each medication against the maximum recommended geriatric dose derived from the UMG and clinical literature.18,34,35,41–48 This study used maximum, rather than minimum, effective dose used in prior work23–25 to provide a ceiling threshold for safe and effective dose. Total dose exposure was measured over 12 months by summing individual medication mSDDs at the patient level. Patients’ dose exposures were categorized into 3 groups: PPM dose within acceptable range (mSDD % 1.0), PPM dose potentially out-ofrange (mSDD ¼ 1.01–3.00), and PPM dose likely out-of-range (mSDD O 3.00).23–25 Duration of therapy (DOT) was calculated by summing days with use in 2007. Days during hospitalizations and Part A covered SNF stays were excluded because medication use over this period was unobservable.

497

Covariates Key covariates used to describe the nursing home sample included patient sociodemographics (age, sex, race/ethnicity, marital status, geographic region, original reason for Medicare entitlement, low-income subsidy status [LIS], length of LTC stay) and clinical characteristics. Clinical characteristics included use of physical restraints, key comorbidities of interest (e.g., Alzheimer’s disease and related disorders [ADRD], acute myocardial infarction [AMI], arthritis, cancer, diabetes, hip/ pelvic fracture, ischemic heart disease [IHD], osteoporosis, stroke), and psychiatric and behavioral conditions (e.g., anxiety, bipolar disorder, delirium, delusion, depression, hallucination, schizophrenia). These conditions were measured using claims-based ICD-9-CM codes and/or MDS 2.0 elements (Supplemental Table). Additionally, this study also assessed patient behaviors, physical function, cognitive function, and relevant signs and symptoms during 2007 using five MDS 2.0 scales: Aggressive Behavior Scale (ABS),49 Activities of Daily Living (ADL) status,50 Cognitive Performance Scale (CPS),51 MDS Cognition Scale (MDS-COGS),52 and Changes in Health, Endstage disease and Symptoms and Signs (MDSCHESS) scale.53 Data analysis This study estimated annual prevalence and patterns of PPM use, overall and by therapeutic (e.g., antidepressants) and pharmacological subclasses (e.g., selective serotonin reuptake inhibitors, SSRIs). Descriptive analyses were utilized to describe and compare associations between medication measures and covariates. Descriptive statistics were summarized for three PMQI measures: evidence of adequate indications among PPM users and, among these users, dose and duration of use. Additionally, a hierarchy of appropriateness of medication use, taking into account the three PMQI indications was displayed for each of the four major therapeutic classes. All analyses were performed using SASÒ 9.2 (SAS Institute, Cary, NC).

Results Prevalence and patterns of PPM use More than two-thirds (72.1%) of nursing home residents (n ¼ 69,832) used at least 1 PPM class (Table 1). The most commonly used PPM

498

Simoni-Wastila et al. / Research in Social and Administrative Pharmacy 10 (2014) 494–507

Table 1 One-year prevalence of psychopharmacological medications (PPMs) use, overall and by therapeutic and pharmacological class among Medicare long-staya nursing home residents in 2007 Total sample

N

%

69,832 Number of classes used overall 0 1 2 3 4

19,483 19,623 19,643 9517 1566

Use by therapeutic and pharmacological classb Any antipsychotics 22,006 Atypical 20,519 Typical 3658 Any antidepressant use 41,503 Alpha-adrenoreceptor 10,909 antagonists Dopamine-reuptake 1740 blocking compounds MAOIs 170 Serotonin antagonists 5682 SNRIs 5561 SSRIs 28,083 Tetracyclic 2 TCAs and related 2477 compounds Any sedative-hypnotic use 23,996 Benzodiazepine hypnotics 302 Non-benzodiazepine 6109 hypnotics Melatonin receptor 313 agonists Other hypnotics 26 Sedating antidepressants 17,313 Sedating antihistamines 3997 Barbiturates 5 Any anxiolytic use 6129 Benzodiazepines, 419 short-acting Benzodiazepines, 91 long-acting Non-benzodiazepines 1971 anxiolytics Sedating antihistamines 3997

27.9 28.1 28.1 13.6 2.2 31.5 93.2 16.6 59.4 26.3 4.2 0.4 13.7 13.4 67.7 0.0 11.3 34.4 1.3 25.5 1.3 0.1 72.1 16.7 0.0 8.9 6.7 1.5 31.7 64.3

MAOIs ¼ monoamine oxidase inhibitors; SNRIs ¼ serotonin–norepinephrine reuptake inhibitors; SSRIs ¼ selective serotonin reuptake inhibitors; TCAs ¼ tricyclic antidepressants. a Long-term stays (R3 consecutive months) were identified based on evidence from Part A Skilled Nursing Facility (SNF) claims and Minimum Data Set data. b Residents might have used R1 drugs across PPM classes or within a therapeutic class.

class was antidepressants, followed by sedativehypnotics, antipsychotics, and anxiolytics. The majority (93.2%) of AP users used atypical agents. SSRIs constituted the most prevalent antidepressant subclass; the most commonly prescribed sub-classes for anxiolytics and sedative-hypnotics were sedating antihistamines and sedating antidepressants, respectively. Among PPM users, 61% utilized two or more therapeutic classes; only 3.1% received drugs from all PPM classes (Table 2). The most commonly observed inter-class combination was antidepressant and sedative-hypnotic use, followed by antipsychotics and antidepressants use (data not shown). More than one-quarter (26.9%) of antidepressant users exhibited evidence of intra-class multi-drug use, as compared to 15.5% of sedativehypnotic users, 9.9% of antipsychotic users, and 4.1% of anxiolytic users. Patient and clinical characteristics of PPM users The long-stay nursing home sample was predominately female, white, aged 75 years or older, and had an average LTC stay of 9 months (Table 3). Most users were entitled to Medicare due to age, or were LIS recipients; over half were widowed. Approximately the same ordering of comorbid conditions was observed among the total sample, PPM users, and therapeutic class users (Table 4). The most prevalent comorbidities were ADRD, IHD, and diabetes; the top three psychiatric/behavioral conditions were depression, anxiety, and delirium. There was marked variation in characteristics across PPM classes. For example, a higher proportion of antipsychotic users was diagnosed with ADRD (85.3% vs. 69.9%–74.6% for other PPMs) and received physical restraints (7.7% vs. 5.0–5.2% for other PPMs). Antipsychotic users also exhibited more aggressive behavior (ABS R 3), functional impairment (ADL R 24), and cognitive impairment (CPS R 4 or MDSCOGs R 3) (Table 5). However, results showed a lower proportion (24.1%) of frail antipsychotic users (MDS-CHESS R 2), compared to other PPM users (26.3%, 26.5%, and 27.2% for antidepressant, anxiolytic, and sedative-hypnotic users, respectively). Quality of PPM use Among PPM users, over two-thirds (70.4%, n ¼ 35,434) were identified as having “fully

Simoni-Wastila et al. / Research in Social and Administrative Pharmacy 10 (2014) 494–507

499

Table 2 Patterns of psychopharmacological medication (PPM) use across and within therapeutic classes PPM users

Number of therapeutic classes One

Across 4 classes

50,349

Two

Three

Four

n

%

n

%

N

%

n

%

19,623

39.0

19,643

39.0

9517

18.9

1566

3.1

– 3.7 1.4 0.0

– 195 9 1

Number of pharmacological subclassesa Within each class AP AD SH AX

Single 22,006 41,503 23,996 6219

19,835 30,342 20,277 5964

Two 90.1 73.1 84.5 95.9

2171 9419 3379 252

Three 9.9 22.7 14.1 4.1

– 1547 331 2

Four or more – 0.5 0.0 0.0

AP ¼ antipsychotic; AD ¼ antidepressant; AX ¼ anxiolytic; SH ¼ sedative-hypnotic. % ¼ row percentage. a Number of pharmacological sub-classes in each therapeutic drug: AP (2 sub-classes), AD (8 sub-classes), AX (4 sub-classes), SH (7 sub-classes).

adequate use,” while the remaining were categorized as having evidence of “partially adequate use” or “fully inadequate use” (Table 6). Antipsychotic users had the highest proportion of evidence of adequate indications for its use, followed by antidepressant, sedative-hypnotic, and anxiolytic users. For each PPM class, the majority of users with evidence of adequate indications for use received within-range doses (mSDD % 1), with the highest proportion of within-range doses noted among antipsychotic users (Table 4). However, 5.8% of antipsychotic users with evidence of adequate indications for use exceeded mSDD O3.00, as did 4.4% of sedative-hypnotic users and 3.7% of anxiolytic users. Antipsychotic and antidepressant users had longer mean DOTs (208 and 227 days, respectively) compared to anxiolytic and sedative-hypnotic users (159 and 183 days, respectively). Further analysis of a hierarchy of appropriateness by three PMQI indicators revealed a different pattern of duration by dose across therapeutic classes. Longest mean duration of use occurred with the highest mSDD for antipsychotic and antidepressant users with evidence of adequate indications for use, while the opposite pattern was observed for anxiolytic and sedativehypnotic peers (Fig. 1). Discussion This descriptive study provides the first comprehensive, national view of prevalence and quality of PPM use in long-stay Medicare nursing home

residents in 2007. Over two-thirds (72.1%) of residents used PPMs; the majority (70.4%) had evidence of adequate indications for all PPMs used. This study observed substantial differences in 1-year prevalence and quality of use across PPM classes. For example, while the prevalence of antipsychotic use in nursing facilities remained high (31.5%), nearly all use (95.8%) had evidence of at least one indication for use deemed adequate by the UMG. This high estimate is consistent with data from a 2011 report conducted by the Office of Inspector General, indicating that 92.0% of atypical antipsychotic claims for nursing home residents were prescribed with evidence of an appropriate indication.54 Furthermore, this study also found the majority (78.7%) of antipsychotic users with evidence of at least one adequate indication for use received doses within acceptable clinical range (i.e., mSDD % 1), with a treatment duration of approximately 7 months. Findings suggest prescribers may have increased awareness of indications warranting antipsychotic use, if not dose and duration of use, among their long-term care patients following the FDA black-box warning linking death and antipsychotic use in elderly individuals.55 Lack of evidence regarding adequate indications for use of other PPM classes remains a concern. While a high proportion (59.4%) of residents received R1 antidepressant in 2007, almost 20% of antidepressant users had no evidence of adequate indications for use. Among sedative-hypnotic and anxiolytic users, high proportions also lacked evidence of adequate indications for use (34% and 76%, respectively). As well,

500

Table 3 Patient characteristics of total sample, any psychopharmacological medication (PPM) users, and users of four different therapeutic classes Characteristics

Total sample

Users by therapeutic classa

PPM users

Sample size Age !65 65–74 75–84 85þ Sex Female Male Race White Black Other Marital status Never married Married Widowed Separated/divorced Missing Region Northeast North central South West Reason for Medicare Age Disability ESRD only/with disability LIS status (yes/no) LTC stay (mean, SD)

AD users

AX users

SH users

%

n

%

n

%

n

%

n

%

n

%

69,832

100

50,349

100

22,006

100

41,503

100

6219

100

23,996

100

5720 8180 21,167 34,765

8.2 11.7 30.3 49.8

4471 6335 15,659 23,884

8.9 12.6 31.1 47.4

2563 3069 6949 9425

11.6 13.9 31.6 42.8

3462 5140 13,019 19,882

8.3 12.4 31.4 47.9

755 853 1937 2674

12.1 13.7 31.1 43.0

2130 3040 7512 11,314

8.9 12.7 31.3 47.1

53,041 16,791

76.0 24.0

38,413 11,936

76.3 23.7

16,035 5971

72.9 27.1

32,202 9301

77.6 22.4

4740 1479

76.2 23.8

18,520 5476

77.2 22.8

58,716 8341 2775

84.1 11.9 4.0

43,183 5268 1898

85.8 10.5 3.8

18,371 2731 904

83.5 12.4 4.1

36,333 3728 1442

87.5 9.0 3.5

5349 611 259

86.0 9.8 4.2

20,706 2338 952

86.3 9.7 4.1

8635 10,423 39,287 7300 4187

12.4 14.9 56.3 10.5 6.0

6166 7613 28,082 5625 2863

12.2 15.1 55.8 11.1 5.7

3644 3194 11,205 2752 1211

16.6 14.5 50.9 12.5 5.5

4493 6325 23,691 4624 2370

10.8 15.2 57.1 11.2 5.7

824 985 3401 733 276

13.2 15.8 54.7 11.8 4.4

2688 3708 13,566 2721 1313

11.2 15.5 56.5 11.3 5.5

15,895 19,971 25,951 8015

22.8 28.6 37.2 11.5

11,328 14,307 19,143 5571

22.5 28.4 38.0 11.1

5089 5948 8641 2328

23.1 27.0 39.3 10.6

9305 11,994 15,746 4458

22.4 28.9 37.9 10.7

1175 1668 2655 721

18.9 26.8 42.7 11.6

5352 6255 9614 2775

22.3 26.1 40.1 11.6

54,164 15,255 413 58,149 9

77.6 21.8 0.6 83.3 3.1

38,256 11,785 308 42,615 9

76.0 23.4 0.6 84.6 3.0

15,609 6311 86 19,450 9

70.9 28.7 0.4 88.4 2.8

31,903 9362 238 35,054 9

76.9 22.6 0.6 84.5 3.0

4431 1714 74 5261 9

71.2 27.6 1.2 84.6 3.0

18,245 5559 192 20,297 9

76.0 23.2 0.8 84.6 3.0

AP ¼ antipsychotic; AD ¼ antidepressant; AX ¼ anxiolytic; SH ¼ sedative-hypnotic; ESRD ¼ end-stage renal disease; LIS ¼ low-income subsidy; LTC ¼ long-term care. Drug class groups were not mutually exclusive.

a

Simoni-Wastila et al. / Research in Social and Administrative Pharmacy 10 (2014) 494–507

AP users N

Table 4 Clinical conditions of total sample, any psychopharmacological medication (PPM) users, and users of four different therapeutic classes Clinical conditions

Total sample

Users by therapeutic classa

PPM users

Sample size

AD users

AX users

SH users

%

n

%

n

%

n

%

n

%

n

%

69,832

100

50,349

100

22,006

100

41,503

100

6219

100

23,996

100

Comorbidities HCC counts (mean, SD) 6 ADRD 50,373 IHD 34,871 Diabetes 28,208 Arthritis 25,669 COPD 17,431 Osteoporosis 14,247 Stroke 13,007 Glaucoma 5802 Cancer 3595 Hip/pelvic fracture 3428 Acute MI 1712 Psychiatric/behavioral conditions Depression 38,863 Anxiety 23,974 Delirium 24,589 Insomnia 16,188 Seizure 7712 Parkinson’s disease 7136 Delusion 6612 Hallucination 3916 Schizophrenia 5058 Bipolar/mood 3102 Cerebral palsy 862 Physical restraints 3373

4.1 72.1 49.9 40.4 36.8 25.0 20.4 18.6 8.3 5.1 4.9 2.5

6 37,733 25,404 20,746 18,931 13,030 10,331 9381 4123 2468 2427 1164

4.1 74.9 50.5 41.2 37.6 25.9 20.5 18.6 8.2 4.9 4.8 2.3

5 18,772 10,277 8880 7545 5386 3956 3628 1707 920 889 394

3.9 85.3 46.7 40.4 34.3 24.5 18.0 16.5 7.8 4.2 4.0 1.8

6 30,968 21,178 17,227 15,988 10,851 8786 7865 3396 2020 2049 972

4.0 74.6 51.0 41.5 38.5 26.1 21.2 19.0 8.2 4.9 4.9 2.3

6 4329 3349 2759 2490 1880 1283 1152 459 300 311 159

4.3 69.6 53.9 44.4 40.0 30.2 20.6 18.5 7.4 4.8 5.0 2.6

6 17,532 12,727 9951 9632 6785 5268 4468 1952 1257 1289 615

4.2 73.1 53.0 41.5 40.1 28.3 22.0 18.6 8.1 5.2 5.4 2.6

55.7 34.3 35.2 23.2 11.0 10.2 9.5 5.6 7.2 4.4 1.2 4.8

34,266 20,437 19,070 13,490 5578 5572 5904 3486 4679 2897 527 2650

68.1 40.6 37.9 26.8 11.1 11.1 11.7 6.9 9.3 5.8 1.0 5.3

14,571 9543 9517 6136 2899 2892 4613 2572 4268 2321 259 1696

66.2 43.4 43.2 27.9 13.2 13.1 21.0 11.7 19.4 10.5 1.2 7.7

31,134 17,679 15,828 11,336 4463 4604 4479 2629 2969 2236 418 2058

75.0 42.6 38.1 27.3 10.8 11.1 10.8 6.3 7.2 5.4 1.0 5.0

4146 3308 2430 1928 700 609 724 430 536 374 96 343

66.7 53.2 39.1 31.0 11.3 9.8 11.6 6.9 8.6 6.0 1.5 5.5

16,821 10,489 9402 8303 2626 2502 2631 1583 1846 1408 240 1252

70.1 43.7 39.2 34.6 10.9 10.4 11.0 6.6 7.7 5.9 1.0 5.2

AP ¼ antipsychotic; AD ¼ antidepressant; AX ¼ anxiolytic; SH ¼ sedative-hypnotic; HCC ¼ hierarchical condition categories comorbidity index; ADRD ¼ Alzheimer’s disease and related disorders; IHD ¼ ischemic heart disease; COPD ¼ chronic obstructive pulmonary disease; AMI ¼ acute myocardial infarction. a Drug class groups were not mutually exclusive.

Simoni-Wastila et al. / Research in Social and Administrative Pharmacy 10 (2014) 494–507

AP users N

501

502

MDS scales

Total sample

Users by therapeutic classb

PPM users

AP users

Sample size

AD users

AX users

SH users

N

%

n

%

n

%

n

%

n

%

n

%

69,832

100

50,349

100

22,006

100

41,503

100

6219

100

23,996

100

Aggressive behavior scale (ABS) 0 42,171 60.4 1–2 20,160 28.9 3–12 7501 10.7 Activities of daily living (ADL) 0–10 16,176 23.2 11–18 19,983 28.6 19–23 16,638 23.8 24–28 17,035 24.4 Cognitive performance scale (CPS) 0 7947 11.4 1–2 19,684 28.2 3 21,859 31.3 4–6 20,342 29.1 MDS cognition scale (MDS-COGS) 0–1 5967 8.5 2 30,428 43.6 3–10 33,437 47.9 MDS-CHESSa Scale 0 16,813 24.1 1 35,388 50.7 2–5 17,631 25.2

28,109 16,038 6202

55.8 31.9 12.3

9165 8624 4217

41.6 39.2 19.2

23,742 13,002 4759

57.2 31.3 11.5

3495 1933 791

56.2 31.1 12.7

13,495 7661 2840

56.2 31.9 11.8

11,942 14,678 12,589 11,140

23.7 29.2 25.0 22.1

5537 5736 5453 5280

25.2 26.1 24.8 24.0

9581 12,385 10,605 8932

23.1 29.8 25.6 21.5

1602 1852 1531 1234

25.8 29.8 24.6 19.8

5787 7156 6004 5049

24.1 29.8 25.0 21.0

5096 14,202 17,207 13,844

10.1 28.2 34.2 27.5

920 4693 8702 7691

4.2 21.3 39.5 34.9

4397 12,110 14,267 10,729

10.6 29.2 34.4 25.9

802 1934 1991 1492

12.9 31.1 32.0 24.0

2910 7257 7724 6105

12.1 30.2 32.2 25.4

4145 20,907 25,297

8.2 41.5 50.2

1165 7669 13,172

5.3 34.8 59.9

3568 17,402 20,533

8.6 41.9 49.5

589 2713 2917

9.5 43.6 46.9

2138 10,612 11,246

8.9 44.2 46.9

11,742 25,646 12,961

23.3 50.9 25.7

5640 11,054 5312

25.6 50.2 24.1

9285 21,289 10,929

22.4 51.3 26.3

1337 3233 1649

21.5 52.0 26.5

5001 12,461 6534

20.8 51.9 27.2

AP ¼ antipsychotic; AD ¼ antidepressant; AX ¼ anxiolytic; SH ¼ sedative-hypnotic. a Changes in health, end-stage disease and symptoms and signs. b Drug class groups were not mutually exclusive.

Simoni-Wastila et al. / Research in Social and Administrative Pharmacy 10 (2014) 494–507

Table 5 MDS scales of total sample, any psychopharmacological medication (PPM) users, and users of four different therapeutic classes

503

Simoni-Wastila et al. / Research in Social and Administrative Pharmacy 10 (2014) 494–507

Table 6 Quality of psychopharmacological medication (PPM) use by indications (among users), and by dose and duration of use (among users with adequate indications), overall and by therapeutic class Quality indicators

Users (n)

Overall PPM

50,349

Quality by indication [n, (%)] Use deemed adequate 35,434 (70.4)a Use deemed inadequate 14,915 (29.6)b Among users (n) with evidence of adequate indications of use Quality by dose [n, (%)] 0.00 % mSDD % 1.00 1.01 % mSDD % 3.00 mSDD O 3.00 Quality by duration of use (mean  SD)

Users by therapeutic classc AP

AD

SH

AX

22,006

41,503

23,996

6219

21,082 (95.8) 924 (4.2)

33,642 (81.1) 7861 (18.9)

15,887 (66.2) 8109 (33.8)

1500 (24.1) 4719 (75.9)

35,434

21,082

33,642

15,887

1500

18,408 (52.0) 13,848 (39.1) 3178 (9.0) 231  111

16,585 (78.7) 3275 (15.5) 1222 (5.8) 208  118

23,507 (69.9) 9759 (29.0) 376 (1.1) 227  111

11,604 (73.1) 3579 (22.5) 697 (4.4) 183  117

1126 (75.1) 318 (21.2) 56 (3.7) 159  118

AP ¼ antipsychotic; AD ¼ antidepressant; AX ¼ anxiolytic; SH ¼ sedative-hypnotic; mSDD ¼ modified standardized daily dose; SD ¼ standard deviation. % ¼ column percentage. a Users had all drug classes used for adequate indications. b Of 14,915 PPM users, 7221 had partial evidence of adequate indications for use (i.e.,R1 of drug classes had no evidence of adequate indications) and 7694 had evidence of no adequate indication for use (i.e., all drug classes had evidence of inadequate indications for use). c Drug class groups were not mutually exclusive.

dose and duration of these classes remained high and potentially worrisome. This study reported the demographics and behavioral conditions of PPM users and users of four therapeutic classes. The majority of PPM users were female, older adults, white, and widowed. A similar demographic distribution was observed in each drug group, results consistent with previously reported data.20,22,37,56 In particular, the study observed a markedly high proportion of widowed residents used PPMs, due in part to a high number of widows living in NHs,57 and poor physical and mental health among the widowed.58,59 In terms of psychiatric and behavioral conditions, over two-thirds of PPM users had depression, and one-third had symptoms of anxiety and/or delirium. These findings are anticipated as PPM drugs are considered effective treatment for these symptoms. Additionally, findings showed the majority (74.9%) of residents who used PPMs had ADRD diagnoses, with the highest percentage (85.3%) seen among residents with antipsychotic use. The use of antipsychotics has been a concern among residents with ADRD since evidence suggested its use increased risk of death among this population.8–12 These findings, along with the FDA warning on antipsychotic

use, may have affected clinical practice by changing therapy options clinicians use to treat ADRD patients,60 and consequently led to reductions in antipsychotic prescribing in this population.61 However, the decline in antipsychotic use raises a concern about the quality of care for symptomatic patients, given the lack of superior evidencebased alternative options. Further studies are warranted to elucidate the clinical impact of decrease in antipsychotics in ADRD patients. This study was framed around the broad, safety-oriented guidelines set forth in the UMG; as such, it remains debatable how findings may translate into clinical prescribing practices. There remains considerable controversy over the safety and necessity of PPM prescribing for those indications listed as adequate within the current UMG, such that CMS is reviewing the guidance around these medications (Personal communication, Dan Andersen. 21 February, 2013).17 Clinicians may be aware of recommendations regarding PPM indications, dosing, and duration, but still exercise sound clinical judgment in their decision to prescribe differently. Furthermore, considering the scrutiny surrounding antipsychotic prescribing in nursing facilities, this study may highlight a shift in prescribing patterns

504

Simoni-Wastila et al. / Research in Social and Administrative Pharmacy 10 (2014) 494–507

Prevalence

AP use 31.5% (n=22,006 )

AD use 59.4% (n=41,503)

Adequate indication

78.7%

95.8% (n=21,082)

81.1% (n=33,642)

SH use 34.4% (n=23,996)

>3.00: 5.8%

255± 110 220± 113

1.01-3.00: 29.0%

242± 105

>3.00: 1.1%

267± 86 167± 119

1.01-3.00: 21.2%

118± 111

>3.00: 3.7%

143± 95

73.1%

66.2% (n=15,887)

200± 118 234± 116

75.1%

24.1% (n=1,500)

Duration (mean ± SD)

1.01-3.00: 15.5%

69.9%

Long-stay residents (n=69,832) AX use 8.9% (n=6,219)

Dose (mSDD)

154 ± 120

1.01-3.00: 22.5%

130 ± 110

>3.00: 4.4%

149 ±109

Fig. 1. A hierarchy of appropriateness of use by adequate indications, dose, and duration for each of the four therapeutic classes. AP ¼ antipsychotic; AD ¼ antidepressant; AX ¼ anxiolytic; SH ¼ sedative-hypnotic; mSDD ¼ modified standardized daily dose; SD ¼ standard deviation.

toward other PPM classes with similar safety risks, including the potential for use as chemical restraints. Future research and clinical interventions may be needed to broadly address suboptimal psychopharmacological prescribing practices beyond focusing on a single therapeutic class. Policy implications This study provides important benchmarks for the annual prevalence and quality of PPM prescribing in long-stay nursing facility residents. These benchmarks are important as CMS and other stakeholders strive to reduce antipsychotic prescribing among nursing facility residents with dementia by 15% by the end of 2012.17 While practitioners and policy-makers have focused much attention on improving quality of antipsychotic use, this study highlights the need to consider the quality of prescribing of other PPM classes. For example, although the highly prevalent use (60%) of antidepressants is largely due to a high proportion (56%) of the NH residents diagnosed with depression-related symptoms,

considerable use (18.9%) is without evidence of adequate indications. This finding demonstrates a pressing need for examination of the quality of antidepressant prescribing in NHs. When prescribing PPMs to older adults in nursing facilities, prescribers should consider adequate medical justification for its use. Polices focusing on AP use17 may be expanded to other PPM classes to reduce the amount of inappropriate use of these drugs in long-term facilities. The broad scope of this study provides information useful to prescribers, nursing staff, policymakers and other stakeholders, including residents. This study demonstrates the need for further development of data-driven clinical benchmarks to assess, monitor, and guide the safe and effective use of PPMs while protecting clinical flexibility. Strengths and limitations This study improves on prior research focusing on single drug classes (mostly antipsychotics) and/or broad quality measures4,22,56,62 by examining four major psychopharmacological classes

Simoni-Wastila et al. / Research in Social and Administrative Pharmacy 10 (2014) 494–507

in tandem. Furthermore, this study measures PPM quality around three domains recommended by the UMG: evidence of adequate indication for use, dose, and duration of therapy.18 Defining and operationalizing PPM use and quality provides a necessary foundation to examine other important medication management domains, including safety and efficacy monitoring, and identification of and response to medication-related adverse consequences. Most significantly, this study provides nationally-representative measures of 1-year PPM prevalence and quality in the Medicare Part D nursing facility population. Several limitations merit discussion. First, findings concerning anxiolytics and sedative-hypnotics should be interpreted with caution due to incomplete data on the use of some important pharmacologic classes (i.e., benzodiazepines, barbiturates, and antihistamines) due to lack of Part D coverage. Second, using both claims and MDS data to capture indications may overestimate the presence of conditions deemed adequate for medication. Future work to validate the utilization of administrative and MDS data with medical records is needed. Indeed, given the importance of medication use and quality for CMS, nursing facilities, and residents and their caregivers, it remains vital that MDS and Medicare administrative claims data be validated against chart review of progress notes, nursing notes, behavior sheets and other information in order to optimize, benchmark, and monitor such use and quality. Finally, this one-year crosssectional study did not explore all aspects of medication use as described in the UMG, such as concomitant therapy, discontinuations, switching, augmentation, and titration or gradual dose reduction. These additional measures remain important targets for future research. Next steps Assessing patterns and quality of PPM prescribing over a longer period using subsequent years of CCDW data will be vital to providing a better understanding of PPM use in the Medicare longterm care population. Research to identify and quantify the influence of key patient, clinical, and provider factors on PMQI adherence and subsequent clinical and economic outcomes using longitudinal files and appropriate multivariable analyses to assess temporality and causality is needed. Such research will help to identify residents at high risk of suboptimal pharmacological care and target subsequent interventions to enhance medication

505

management and prevent adverse sequelae, including falls, hospitalization, and mortality. Conclusion In 2007, over two-thirds of long-stay Medicare nursing home residents used at least one PPM. The highest prevalence was seen for antidepressants, followed by sedative-hypnotics and antipsychotics. Although this study observed high prevalence of antipsychotic use with evidence of guideline adequate indications, for other PPM classes the quality of use by indication, as well as dose and duration, remain potentially problematic. Practitioners should heed the prescribing patterns of all PPMs in nursing facility residents. The CMS existing polices focusing on AP use may be expanded to other PPM classes to reduce the amount of inappropriate use of these drugs. Acknowledgments This research conducted at the University of Maryland School of Pharmacy was funded by CMS through a subcontract with Abt Associates (Contract no. HHSM-500-2005-0001881). All coauthors have no disclosures to report. Dr. Rattinger was supported by an NIH Institutional Career Development Grant (K12 HD043489). Supplementary material Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.sapharm. 2013.10.003. References 1. IOM. Medicare: A Strategy for Quality Assurance. Washington, DC: National Academy Press; 1990. 2. McGlynn E, Asch S, Adams J. The quality of health care delivered to adults in the United States. N Engl J Med 2003;348:2635–2645. 3. IOM. To Err Is Human: Building a Safer Health System. Washington, DC: National Academy Press; 2000. 4. Simoni-Wastila L, Ryder PT, Qian J, Zuckerman IH, Shaffer T, Zhao L. Association of antipsychotic use with hospital events and mortality among medicare beneficiaries residing in long-term care facilities. Am J Geriatr Psychiatry 2009;17:417–427. 5. Huybrechts KF, Schneeweiss S, Gerhard T, et al. Comparative safety of antipsychotic medications in nursing home residents. J Am Geriatr Soc 2012; 60(3):420–429.

506

Simoni-Wastila et al. / Research in Social and Administrative Pharmacy 10 (2014) 494–507

6. Hughes CM, Lapane KL, Mor V. Impact of legislation on nursing home care in the United States: lessons for the United Kingdom. BMJ 1999;319(7216): 1060–1062. 7. OBRA-87. Omnibus Budget Reconciliation Act of 1987. Public Law 00–203. Subtitle C. Nursing Home Reform Act. 42 USC 1395i-3(a)-(h)(Medicare); 1396r (a)-(h) (Medicaid); 1987. 8. Wang PS, Schneeweiss S, Avorn J, et al. Risk of death in elderly users of conventional vs. atypical antipsychotic medications. N Engl J Med 2005; 353:2335–2341. 9. Schneider LS, Dagerman KS, Insel P. Risk of death with atypical antipsychotic drug treatment for dementia: meta-analysis of randomized placebo-controlled trials. J Am Med Assoc 2005;294:1934–1943. 10. Kales HC, Kim HM, Zivin K, et al. Risk of mortality among individual antipsychotics in patients with dementia. Am J Psychiatry 2012;169:71–79. 11. Liperoti R, Onder G, Landi F, et al. All-cause mortality associated with atypical and conventional antipsychotics among nursing home residents with dementia: a retrospective cohort study. J Clin Psychiatry 2009;70:1340–1347. 12. Gill SS, Bronskill SE, Normand SL, et al. Antipsychotic drug use and mortality in older adults with dementia. Ann Intern Med 2007;146:775–786. 13. Rochon PA, Normand SL, Gomes T, et al. Antipsychotic therapy and short-term serious events in older adults with dementia. Arch Intern Med May 26 2008;168:1090–1096. 14. FDA. Public Health Advisory: Deaths With Antipsychotics in Elderly Patients With Behavioral Disturbance; April 11 2005. 15. Aparasu RR, Chatterjee S, Chen H. Risk of pneumonia in elderly nursing home residents using typical versus atypical antipsychotics. Ann Pharmacother Apr 2013;47:464–474. 16. Aparasu RR, Chatterjee S, Mehta S, Chen H. Risk of death in dual-eligible nursing home residents using typical or atypical antipsychotic agents. Med Care Nov 2012;50:961–969. 17. Mitka M. CMS seeks to reduce antipsychotic use in nursing home residents with dementia. J Am Med Assoc 2012;308. 119, 121. 18. CMS. Manual System. Unnecessary Medications Guidance. Pub 100–107: State Operations Provider Certification: Transmittal 22: Tag F329 (Rev 22; Issued 12-115-06; Effective/Implementation:12-1806) 438.25: Unnecessary Drug. http://www.cms. hhs.gov/transmittals/downloads/R225OMA.pdf; Accessed 12.10.09. 19. Kales H, Zivin K, Kim H, et al. Trends in antipsychotic use in dementia 2000-2007. Arch Gen Psychiatry 2011;68:190–197. 20. Chen Y, Briesacher B, Field T, Tija J, Lau D, Gurwitz J. Unexplained variation across US nursing homes in antipsychotic prescribing rates. Arch Int Med 2010;170:89–95.

21. Huybrechts K, Schneeweiss S, Gerhard T, et al. Comparative safety of antipsychotic medications in nursing home resident. J Am Geriatr Soc 2012; 60:420–429. 22. Briesacher BA, Limcangco MR, Simoni-Wastila L, et al. The quality of antipsychotic drug prescribing in nursing homes. Arch Int Med 2005;165: 1280–1285. 23. Boudreau RM, Hanlon JT, Roumani YF, et al. Central nervous system medication use and incident mobility limitation in community elders: the health, aging, and body composition study. Pharmacoepidemiol Drug Saf 2009;18:916–922. 24. Hanlon JT, Boudreau RM, Roumani YF, et al. Number and dosage of central nervous system medications on recurrent falls in community elders: the health, aging and body composition study. J Gerontol A Biol Sci Med Sci 2009;64:492–498. 25. Wright RM, Roumani YF, Boudreau R, et al. Effect of central nervous system medication use on decline in cognition in community-dwelling older adults: findings from the health, aging and body composition study. J Am Geriatr Soc 2009;57:243–250. 26. CMS. Nursing Home Compare. http://www.medicare.gov/NursingHomeCompare/search.aspx?bhcp¼1; Accessed 06.11.12. 27. Warehouse CCCD. Long Term Care Minimum Data Set (MDS) Identifiable Data Files, 2012. https:// www.cms.gov/IdentifiableDataFiles/10_LongTermCareMinimumDataSetMDS.asp. Accessed 20.02.12. 28. The Center for Medicare and Medicaid Services. Nursing Home Quality Initiative, 2013. http://www. cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/NursingHomeQualityInits/NHQIQualityMeasures.html. 29. Zuckerman IH, Sato M, Hsu VD, Hernandez JJ. Validation of a method for identifying nursing home admissions using administrative claims. BMC Health Serv Res 2007;7:202. 30. Yun H, Kilgore ML, Curtis JR, et al. Identifying types of nursing facility stays using medicare claims data: an algorithm and validation. Health Serv Outcomes Res Methodol 2010;10:100–110. 31. Koroukian SM, Xu F, Murray P. Ability of Medicare claims data to identify nursing home patients: a validation study. Med Care 2008;46(11):1184–1187. 32. Iwashyna TJ. On the detection of nursing home use in Medicare claims. Health Serv Outcomes Res Methodol 2003;4:187–196. 33. Guthrie B, Clark S, McCowan C. The burden of psychotropic drug prescribing in people with dementia: a population database study. Age Ageing Sep 2010;39:637–642. 34. APA Work Group on Alzheimer’s Disease and other DementiasRabins PV, Blacker D, et al. American Psychiatric Association practice guideline for the treatment of patients with Alzheimer’s disease and other dementias: 2nd ed. Am J Psychiatry Dec 2007;164(12 suppl):5–56.

Simoni-Wastila et al. / Research in Social and Administrative Pharmacy 10 (2014) 494–507 35. APA Work Group on Delirium. American Psychiatric Association Practice Guideline for the Treatment of Patients With Delirium. http://psychiatryonline. org/content.aspx?bookid¼28§ionid¼1663978; Accessed 23.01.12. 36. Kamble P, Chen H, Sherer JT, Aparasu RR. Use of antipsychotics among elderly nursing home residents with dementia in the US: an analysis of National Survey Data. Drugs Aging 2009;26:483–492. 37. Karkare SU, Bhattacharjee S, Kamble P, Aparasu R. Prevalence and predictors of antidepressant prescribing in nursing home residents in the United States. Am J Geriatr Pharmacother 2011;9:109–119. 38. Wu CS, Ting TT, Wang SC, Chang IS, Lin KM. Effect of benzodiazepine discontinuation on dementia risk. Am J Geriatr Psychiatry 2011;19:151–159. 39. Hanlon J, Wang X, Castle N, et al. Potential underuse, overuse, and inappropriate use of antidepressants in older veteran nursing home residents. J Am Geriatr Soc 2011;59:1412–1420. 40. Burrows AB, Morris JN, Simon SE, Hirdes JP, Phillips C. Development of a minimum data setbased depression rating scale for use in nursing homes. Age Ageing 2000;29:165–172. 41. APA Work Group on Bipolar Disorder. American Psychiatric Association Practice Guideline for the Treatment of Patients With Bipolar Disorder: 2nd ed. http://psychiatryonline.org/content.aspx?bookid¼28§ionid¼1669577; Accessed 23.10.12. 42. APA Work Group on Major Depressive Disorder. American Psychiatric Association Practice Guideline for the Treatment of Major Depressive Disorder: 3rd ed. http://psychiatryonline.org/content.aspx? bookid¼28§ionid¼1667485; Accessed 23.01.12. 43. APA Work Group on Schizophrenia. American Psychiatric Association Practice Guideline for the Treatment of Patients With Schizophrenia: 2nd ed. http://psychiatryonline.org/content.aspx?bookid¼28§ionid¼1665359; Accessed 23.01.12. 44. Kastenschmidt EK, Kennedy GJ. Depression and anxiety in late life: diagnostic insights and therapeutic options. Mount Sinai J Med New York Jul–Aug 2011;78:527–545. 45. Geriatric Lexi-Drugs Online. Lexi-Comp, Inc. Updated Last Updated Date. Accessed 18.01.12. 46. Drugdex. Thomson Healthcare. Updated Last Updated Date. Accessed 06.12.10. 47. Rojas-Fernandez CH, Miller LJ, Sadowski CA. Considerations in the treatment of geriatric depression: overview of pharmacotherapeutic and psychotherapeutic treatment interventions. Res Gerontol Nurs 2010;3:176–186. 48. Tampi R, Williamson D, Muralee S, et al. Behavioral and psychological symptoms of dementia: part II: treatment. Clin Geriatr 2011;19:31–40.

507

49. Perlman C, Hirdes J. The aggressive behavior scale: a new scale to measure aggression based on the minimum data set. J Am Geriatr Soc 2008;56:2298–2303. 50. Morris J, Fries B, Morris S. Scaling ADLs within the MDS. J Gerontol A Biol Sci Med Sci Nov 1999;54:M546–M553. 51. Morris JN, Fries BE, Mehr DR, et al. MDS cognitive performance scale. J Gerontol Jul 1994;49: M174–M182. 52. Hartmaier SL, Sloane PD, Guess HA, Koch GG. The MDS cognition scale: a valid instrument for identifying and staging nursing home residents with dementia using the minimum data set. J Am Geriatr Soc 1994;42:1173–1179. 53. Hirdes JP, Frijters DH, Teare GF. The MDSCHESS scale: a new measure to predict mortality in institutionalized older people. J Am Geriatr Soc Jan 2003;51(1):96–100. 54. OIG. Medicare Atypical Antipsychotic Drug Claims for Elderly Nursing Home Residents. Washington, DC: The Office of the Inspector General for the Department of Health and Human Services; May 2011. OEI-07-08-00150. 55. FDA. Public Health Advisory: Deaths With Antipsychotics in Elderly Patients With Behavioral Disturbances; 2005. 56. Stevenson DG, Decker SL, Dwyer LL, et al. Antipsychotic and benzodiazepine use among nursing home residents: findings from the 2004 National Nursing Home Survey. Am J Geriatr Psychiatry 2010;18:1078–1092. 57. Kaye HS, Harrington C, LaPlante MP. Long-term care: who gets it, who provides it, who pays, and how much? Health Aff (Millwood) 2010;29:11–21. 58. Avis NE, Brambilla DJ, Vass K, McKinlay JB. The effect of widowhood on health: a prospective analysis from the Massachusetts Women’s Health Study. Soc Sci Med 1991;33(9):1063–1070. 59. Cafferata GL, Meyers SM. Pathways to psychotropic drugs. Understanding the basis of gender differences. Med Care 1990;28:285–300. 60. Saad M, Cassagnol M, Ahmed E. The impact of FDA’s warning on the use of antipsychotics in clinical practice: a survey. Consult Pharm 2010;25: 739–744. 61. Desai VC, Heaton PC, Kelton CM. Impact of the Food and Drug Administration’s antipsychotic black box warning on psychotropic drug prescribing in elderly patients with dementia in outpatient and office-based settings. Alzheimers Dement 2012; 8:453–457. 62. Kamble P, Sherer J, Chen H, Aparasu R. Off-label use of second-generation antipsychotic agents among elderly nursing home residents. Psychiatr Serv (Washington, D.C.) 2010;61:130–136.