- Email: [email protected]
Anticholinergic Drug Burden in Noncancer Versus Cancer Patients Near the End of Life
Accepted Manuscript Anticholinergic drug burden in non-cancer vs. cancer patients near the end of life Michael J. Hochman, B.S, Arif H. Kamal, M.D., M.H.S, Steven Wolf, M.S, Greg P. Samsa, Ph.D, David C. Currow, M.P.H., FRACP, Amy P. Abernethy, M.D., Ph.D, Thomas W. LeBlanc, M.D., M.A PII:
S0885-3924(16)30298-6
DOI:
10.1016/j.jpainsymman.2016.03.020
Reference:
JPS 9189
To appear in:
Journal of Pain and Symptom Management
Received Date: 2 December 2015 Revised Date:
18 March 2016
Accepted Date: 30 March 2016
Please cite this article as: Hochman MJ, Kamal AH, Wolf S, Samsa GP, Currow DC, Abernethy AP, LeBlanc TW, Anticholinergic drug burden in non-cancer vs. cancer patients near the end of life, Journal of Pain and Symptom Management (2016), doi: 10.1016/j.jpainsymman.2016.03.020. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
RI PT
Anticholinergic drug burden in non-cancer vs. cancer patients near the end of life Michael J. Hochman, B.S.1 Arif H. Kamal, M.D., M.H.S.2,3 Steven Wolf, M.S.4 Greg P. Samsa, Ph.D.4 David C. Currow, M.P.H., FRACP5 Amy P. Abernethy, M.D., Ph.D.3 Thomas W. LeBlanc, M.D., M.A.2,6 1
Duke University School of Medicine Durham, NC 27710 2
SC
Duke Cancer Institute Duke University School of Medicine Durham, NC 27710 3
M AN U
Flatiron Health New York, NY 10012 4
Duke Biostatistics Core Department of Biostatistics and Bioinformatics Duke University School of Medicine Durham, NC 27710 5
TE D
Discipline, Palliative and Supportive Services and Department of Medicine Flinders University Adelaide, South Australia
6
EP
Division of Hematologic Malignancies and Cellular Therapy Department of Medicine Duke University School of Medicine Durham, NC 27710
AC C
Corresponding Author: Thomas LeBlanc Box 2715 Duke University Medical Center Durham, NC 27710 Email: [email protected] Tables: 2 Figures: 1 References: 46 Word count: 2,474
ACCEPTED MANUSCRIPT
M AN U
SC
RI PT
Abstract Context. Anticholinergic drugs can cause several side effects, impairing cognition and quality of life (QOL). Cancer patients are often exposed to increasing cumulative anticholinergic load (ACL) as they approach death, but this burden has not been examined in patients with non-malignant diseases. Objectives. To determine ACL and its impact in non-cancer versus cancer palliative care patients. Methods. We performed a secondary analysis of 244 subjects enrolled in a randomized controlled trial. ACL was quantified with the Anticholinergic Drug Scale. We used multivariable regression to calculate the effect of ACL on key outcomes, including drowsiness, fatigue, and QOL. Patients were stratified by diagnosis, and drugs were grouped as symptom management (SM) or disease management (DM). Results. Overall ACL in cancer and non-cancer patients was not significantly different (2.6 vs. 2.4; P=0.23). SM drugs caused greater anticholinergic exposure than DM drugs in both cancer and non-cancer patients (2.3 vs. 0.5, and 1.5 vs. 1.3, respectively; both P<0.05), however DM drugs exposed non-cancer patients to relatively more ACL than cancer patients (1.2 vs. 0.6, P<0.0001). ACL was associated with worse fatigue (OR, 1.08; CI, 1.002-1.17) and worse QOL (OR, 0.89; CI, 0.80-0.98). Conclusions. ACL is associated with worse fatigue and QOL, and may not differ significantly between cancer and non-cancer patients nearing end of life. SM drugs are more responsible for ACL in cancer and non-cancer patients, though DM drugs contribute significantly to ACL in the latter group. We recommend more attention to reducing anticholinergic use in all patients with life-limiting illness.
TE D
Keywords Cholinergic antagonists, functional status, comorbidity, palliative care, quality of life
AC C
EP
Running Title: “Anticholinergic burden in non-cancer patients near EOL”
ACCEPTED MANUSCRIPT
RI PT
Introduction Medications with anticholinergic activity may cause several side effects including dry mouth, blurry vision, tachycardia, drowsiness, and delirium (1). Long-term cumulative anticholinergic exposure may even be a risk factor for the development of dementia (2, 3). In palliative and end-of-life settings, anticholinergic medications (e.g. atropine, opioids) are used frequently and are often necessary for managing pain, dyspnea, secretions, and anxiety, among other symptoms.
M AN U
SC
Terminally ill patients often suffer new or worsening symptoms requiring additional pharmacotherapy (4). To date, analyses of anticholinergic drug burden at end of life have focused on patients with cancer. For example, one analysis of mostly cancer patients in Australia found increasing cumulative anticholinergic exposure with proximity to death, largely from opioids (5). As patients with non-malignant life-limiting illnesses live longer and palliative care expands to include more clinically heterogeneous populations with longer life expectancies, it is also important to consider anticholinergic effects in patients with non-cancer diagnoses.
TE D
Our aim is therefore to investigate the relationship between drugs with anticholinergic activity, cancer versus non-cancer diagnosis, and key patient outcomes (drowsiness, fatigue, wellbeing, performance status, and quality of life) in a cohort of older patients with serious, life-limiting illness. We hypothesize that as patients approach death, anticholinergic load (ACL) increases and is associated with worse outcomes, such as drowsiness and fatigue. We also expect cancer patients to be exposed to greater ACL near end of life due to greater opioid use.
AC C
EP
Methods Study Design The original study was a randomized, pragmatic trial evaluating the impact of statin discontinuation near end of life on 381 patients from 15 U.S. sites (6). Eligible patients were adults with a life-limiting diagnosis and life expectancy >1 month with recent functional deterioration. The Duke University IRB approved the study, and the original paper describes the complete methodology (6). This secondary analysis is possible because statins have no anticholinergic properties; their presence or absence in a patient’s drug regimen is insignificant to this study. We thus combined both study arms and grouped patients by primary diagnosis (cancer vs. non-cancer). We included data from patients alive for at least 30 days after enrollment, with more than one follow-up with medications recorded and complete medication lists (n=244). We abstracted medication lists for up to 8 follow-ups. Assessments Baseline assessments included demographics, primary diagnosis, hospice status, cognition, performance status, quality of life (QOL), symptom severity, and medications. Outcomes included survival, resource utilization, and performance status (as measured by Australia-modified Karnofsky Performance Status [AKPS] scale (7)), collected weekly during the first month, then monthly until death (or one year). Medication inventories and
ACCEPTED MANUSCRIPT
patient-reported outcomes were collected in person or by telephone at baseline and scheduled follow-ups (0, 2, 4, 8, 12, 16, 20, and 24 weeks). Medication reports included medication name and frequency (scheduled or as needed). Symptoms and QOL were measured by the Edmonton Symptom Assessment Scale (ESAS)(8) and the McGill Quality of Life Questionnaire (MQOLQ)(9), respectively.
M AN U
SC
RI PT
Data Collection The Anticholinergic Drug Scale (ADS) is used in several studies to quantify anticholinergic burden in geriatric populations, including nursing home residents and hospital inpatients; it is correlated with serum anticholinergic activity in long-term care settings (10, 11). Individual medications are assigned a score on a 4-point scale, where 0 indicates no known anticholinergic activity, 1 indicates potential anticholinergic activity per receptor binding assays, 2 indicates clinically significant effects at excess doses, and 3 refers to marked anticholinergic effects at usual doses. ADS was originally based on reported anticholinergic effects in the literature, available laboratory data, and the independent ratings of 3 geriatric psychiatrists (12); it is a modification of the Clinicianrated Anticholinergic Scale (CrAS)(10, 11). We classified medications by their intended use, either for chronic disease management (DM) or symptom management (SM) as per prior studies (4, 5). Drugs potentially used for both purposes (anti-epileptics, antidepressants, acid suppressants) were assigned SM or DM by one author (MJH), based on likelihood of being used for maintenance of chronic illness versus treatment of acute symptoms, with subsequent review by another author (TWL).
EP
TE D
Statistical Analysis We added the ACL scores for all drugs at each available time point; an example is shown in Table S1. We treated combination medications as multiple individual medications taken at a single time point, except for those that had an ACL only in combination (e.g., fluticasone-salmeterol). We included all medications prescribed “as needed” and excluded non-specific drugs (“inhaler,” “leg med”) and non-drugs (“oxygen”). One hundred nineteen of 413 unique medications and supplements were not listed in ADS and were excluded from the analysis. A list of those medications is included in the supplementary content (Table S2).
AC C
We calculated descriptive statistics for patients at baseline based on their primary diagnosis, and then compared cancer and non-cancer patients using the Chi-Square and Fisher’s Exact test for discrete variables and the Kruskal-Wallis test for continuous variables. To trace how ACL varied longitudinally as patients approached death, we used performance status to follow disease progression and looked at encounters as the unit of analysis instead of patients. We took this approach because follow-ups were not spaced evenly in the original trial design, and data were highly likely to be missing not-atrandom due to patient decline and death. Functional status has been used to explore ACL (5) and track symptom burden (13) in palliative care patients; it has also been found to be a significant predictor of survival in these populations (14-16). By evaluating ACL based on performance status (AKPS), we maintained our sample size (n=244; 1,210 encounters).
ACCEPTED MANUSCRIPT
SC
RI PT
We calculated average ACL by determining the standard and least squares mean (LSM) of all encounters’ ACLs over the course of the study by adjusting for AKPS, and then stratified by cancer diagnosis and drug type (DM vs. SM). We used ANCOVA to test for statistical significance between the groups and ACL types while adjusting for AKPS, disease status, race, gender, cognitive status, and hospice status at baseline. Using multivariable ordinal logistic regression, we evaluated associations between overall ACL, QOL, and symptoms, adjusting for age, race, gender, primary diagnosis, baseline cognitive impairment, baseline hospice status, and performance status (with AKPS grouped as 10-30 [poor], 40-60 [moderate], and 70-100 [good]). As a secondary analysis, we compared the standard mean and LSM of ACL from DM drugs with that from SM drugs in cancer patients and non-cancer patients separately. Additionally, to account for the effects of polypharmacy, we examined the number of anticholinergic drugs patients used over the course of the study and tested whether this predicted our outcomes of interest. Statistics were done using SAS, version 9.4 (SAS Institute Inc).
M AN U
Finally, we ranked the relative contribution of drugs to cumulative anticholinergic exposure. We abstracted medication lists for the first time patients dropped below an AKPS of 70 to any lower performance level (n=165). We chose this point in order to evaluate medications taken as functional status declined at a standardized point in the illness trajectory. We calculated the relative contribution of drugs to patient ACL by multiplying the number of times a drug appeared on a patient’s list (the “count”) with its ACL score from ADS.
EP
TE D
Results Baseline Characteristics About half (n=118, 48.4%) of patients had a non-malignant disease (Table 1). On average, non-cancer patients were older (79.1 vs. 69.7 years; P<0.0001) and had worse performance status (median AKPS 50.0 vs. 60.0; P<0.0001). Differences in QOL and average ACL (6.7 vs. 6.3 and 3.2 vs. 2.8, respectively) were not statistically significant.
AC C
Longitudinal Anticholinergic Exposure Figure 1A displays how ACL varied in cancer and non-cancer patients as they approached death. Cancer patient ACL peaked at an AKPS of 30 (n=19), indicating a state of being almost completely bedbound; non-cancer patients’ ACL peaked at an AKPS of 40 (bedbound more than half of the time, n=134) and again at 10 (comatose status, n=3). Primary diagnosis and AKPS were both significantly associated with ACL (Table S3). Cancer patients received a larger ACL from SM drugs than DM drugs (2.3 vs. 0.5 ACL, least squares mean [LSM]; P<0.0001; see Figure 1B), as did non-cancer patients (1.5 SM vs. 1.3 DM ACL, LSM; P=0.04; see Figure 1C and Table S4). Overall, however, symptom management drugs exposed cancer patients to a higher ACL than non-cancer patients (2.3 vs. 1.3 ACL, LSM; P<0.0001), and chronic disease management drugs burdened non-cancer patients more than cancer patients (1.2 vs. 0.6 ACL, LSM;
ACCEPTED MANUSCRIPT
P<0.0001). Cumulative anticholinergic exposure between the two groups was not significantly different (2.6 vs. 2.4 ACL, LSM; P=0.23; see Table S5).
SC
RI PT
Multivariable Analysis In multivariable analysis, we explored associations between ACL and several other measures (Table 2). Increased ACL was significantly associated with worse patient fatigue (OR, 1.08; CI, 1.002-1.17; P=0.045) and worse QOL (OR, 0.89; CI, 0.80-0.98; P=0.02) after adjusting for performance status, baseline hospice status, baseline cognitive status, and demographics. There was no association between anticholinergic exposure and drowsiness or wellbeing. In a sensitivity analysis examining the relationship between number of anticholinergic drugs and our outcomes of interest, we found that increased anticholinergic drug usage was significantly associated with drowsiness (OR, 1.12; CI, 1.01-1.24; P=0.04) and worse QOL (OR 0.85; CI, 0.74-0.99; Table S6).
M AN U
Anticholinergic Contribution In cancer and non-cancer patients, the SM drugs that contributed most to anticholinergic burden were opioids and benzodiazepines. Furosemide and warfarin were common DM drugs to burden both patient groups with anticholinergic effects. Rank lists for drug anticholinergic contribution can be found in the supplementary content (Table S7). Cancer patients used more drugs with high ACL scores (ACL 3) than non-cancer patients (0.4 vs. 0.2, P<0.0002); the opposite held true for non-cancer patients consuming more low ACL (1-2) drugs (2.3 vs. 2.0, P<0.0011).
EP
TE D
Discussion Using data from a large, prospective, randomized controlled trial, our longitudinal analysis of ACL among patients near the end of life yielded three important findings: (1) patients with non-cancer diagnoses are exposed to similar cumulative anticholinergic burden as cancer patients near the end of life; (2) much of the anticholinergic drug burden near end of life is due to SM drugs, regardless of cancer versus non-cancer diagnosis (although DM drugs more significantly burden patients without cancer); and (3) drug anticholinergic activity is associated with worsening fatigue and quality of life.
AC C
We were surprised to find that non-cancer patients are exposed to similar ACL compared to cancer patients near end-of-life, despite less frequent use of opioids. This suggests significant use of medications with anticholinergic effects among all patients near the end of life, regardless of primary diagnosis. Although expert palliative care panels have argued that non-complex patients only require a limited number of drugs for quality symptom control near end of life, many of these have anticholinergic activity (17, 18). Our results highlight the challenge of tailoring medication use in patients with lifelimiting illnesses to their needs, and the importance of targeted attention to polypharmacy and anticholinergic use among all patients nearing end of life (19). In both cancer and non-cancer patients, ACL was largely due to SM drugs like opioids and benzodiazepines. This result emphasizes the significance and utility of these medications in palliative and hospice care. Pain and subsequent opioid use has been
ACCEPTED MANUSCRIPT
linked to terminal restlessness (20) and the need for palliative sedation (21) in end-of-life care. Although patients find both mental alertness and minimal pain important near the end of life, palliative care clinicians tend to value terminal lucidity less than their patients (22). Clearly, we must recognize the benefit of adequate pain management but balance it with reasonable cognitive and performance statuses.
SC
RI PT
We also found that anticholinergic drug burden due to DM drugs is relatively higher in non-cancer patients than cancer patients. This is a critical finding. As terminal illness is considered an indication for deprescribing (23), patients with chronic, life-threatening, non-malignant diseases may warrant targeted attention to decrease cumulative anticholinergic exposure via the thoughtful discontinuation of DM agents that are no longer indicated. Furthermore, our results suggest that clinicians should mind patient diagnosis when prescribing medications at end of life, as a more thorough evaluation of DM drugs may facilitate optimal care for terminal non-cancer patients.
TE D
M AN U
Finally, we found that drug ACL is independently associated with worse fatigue and quality of life. The magnitude of these effects was small, yet collectively suggests that cumulative anticholinergic effects are burdensome. These results are concordant with Agar et al., who report adjusted odds ratios and confidence intervals of 0.85 (0.81-0.90) for AKPS and 0.90 (0.85-0.95) for QOL per unit of ACL (5). Because that study included few non-cancer patients and no patients with baseline cognitive impairment, our findings add to the generalizability of this literature. In our sensitivity analysis, we found that the average number of anticholinergic drugs consumed had a detrimental relationship with patient QOL and drowsiness, indicating that anticholinergic polypharmacy and drug potency (as measured by ACL) similarly correspond with this pharmacologic burden. We recommend that future studies focus on whether interventions to reduce anticholinergic exposure can be safely and successfully applied to patients with significant, complex illness near end of life.
AC C
EP
Our approach has a few limitations. First, using performance status to trace disease progression is limited, as it does not necessarily decline linearly with time; some patients remained at the same performance level for the study duration and others varied between multiple levels. Second, a linear additive model for cumulative anticholinergic effects based on ADS does not reflect the nuances of drug-patient interactions or the dosedependence of anticholinergic effects (11, 24). Additionally, different scales assign anticholinergic potency based on varied methods. A different scale would potentially result in different outcomes to our study (25). Third, not all medications could be assigned an ACL score on the scale—a limitation narrowed by the fact that the scale is comprehensive for most medications known to have anticholinergic activity. Fourth, we have a limited ability to assign drug purpose or measure true drug exposure through patient interviews or other means, including dispensing records or serum drug concentrations. Finally, it is difficult to ascertain whether anticholinergic drugs negatively impact performance status near the end of life. Sicker patients with worse QOL may be consuming more anticholinergic drugs for symptom palliation simply because SM drugs are typically added as patients become sicker (4, 5), reflecting a natural tendency in the terminal care of patients. Still, as a robust secondary analysis of a
ACCEPTED MANUSCRIPT
rigorous randomized trial, our work represents a comprehensive and generalizable approach to ACL near end of life.
SC
RI PT
In conclusion, overall anticholinergic exposure may not differ significantly between cancer and non-cancer patients near the end of life, but it appears to be detrimental to patient fatigue and quality of life. SM drugs are most responsible for anticholinergic burden in cancer and non-cancer palliative care patients, and they burden cancer patients relatively more than DM drugs, while chronic disease-specific drugs carry a more significant burden in non-cancer patients near the end of life. Further studies to assess clinical benefit in decreasing anticholinergic burden in palliative care patients are warranted.
TE D
M AN U
Disclosures and Acknowledgments Funding for the statin discontinuation study that generated the data for this analysis was provided by the National Institute of Nursing Research (grant numbers UC4-NR012584 and U24-NR014637). This work also included the support of facilities and resources of the Veterans Affairs Health Care System (eg, Phoenix, Arizona, and Birmingham, Alabama). Dr. LeBlanc received funding support from a Junior Career Development Grant from the National Palliative Care Research Center during the time that this analysis was done. He currently receives support from a Sojourns Scholar Award Grant from the Cambia Health Foundation. The design, conduct, analysis, and write-up of the study were performed independently from any sponsoring agency or funding. No other authors have relevant conflicts of interest to report. The authors would like to acknowledge Michael McNeil for his original ideas during early portions of this work.
AC C
EP
References 1. Mintzer J, Burns A. Anticholinergic side-effects of drugs in elderly people. J R Soc Med 2000;93:457-62. 2. Gray SL, Anderson ML, Dublin S, et al. Cumulative Use of Strong Anticholinergics and Incident Dementia: A Prospective Cohort Study. JAMA Intern Med 2015. 3. Jessen F, Kaduszkiewicz H, Daerr M, et al. Anticholinergic drug use and risk for dementia: target for dementia prevention. Eur Arch Psychiatry Clin Neurosci 2010;260 Suppl 2:S111-5. 4. Currow DC, Stevenson JP, Abernethy AP, Plummer J, Shelby-James TM. Prescribing in palliative care as death approaches. J Am Geriatr Soc 2007;55:590-5. 5. Agar M, Currow D, Plummer J, et al. Changes in anticholinergic load from regular prescribed medications in palliative care as death approaches. Palliat Med 2009;23:25765. 6. Kutner JS, Blatchford PJ, Taylor DH, et al. Safety and Benefit of Discontinuing Statin Therapy in the Setting of Advanced, Life-Limiting Illness: A Randomized Clinical Trial. JAMA Intern Med 2015.
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
7. Abernethy AP, Shelby-James T, Fazekas BS, Woods D, Currow DC. The Australiamodified Karnofsky Performance Status (AKPS) scale: a revised scale for contemporary palliative care clinical practice [ISRCTN81117481]. BMC Palliat Care 2005;4:7. 8. Bruera E, Kuehn N, Miller MJ, Selmser P, Macmillan K. The Edmonton Symptom Assessment System (ESAS): a simple method for the assessment of palliative care patients. J Palliat Care 1991;7:6-9. 9. Cohen SR, Mount BM, Strobel MG, Bui F. The McGill Quality of Life Questionnaire: a measure of quality of life appropriate for people with advanced disease. A preliminary study of validity and acceptability. Palliat Med 1995;9:207-19. 10. Carnahan RM, Lund BC, Perry PJ, Pollock BG, Culp KR. The Anticholinergic Drug Scale as a measure of drug-related anticholinergic burden: associations with serum anticholinergic activity. J Clin Pharmacol 2006;46:1481-6. 11. Salahudeen MS, Duffull SB, Nishtala PS. Anticholinergic burden quantified by anticholinergic risk scales and adverse outcomes in older people: a systematic review. BMC Geriatr 2015;15:31. 12. Han L, McCusker J, Cole M, et al. Use of medications with anticholinergic effect predicts clinical severity of delirium symptoms in older medical inpatients. Arch Intern Med 2001;161:1099-105. 13. Kamal AH, Nipp RD, Bull J, Stinson CS, Abernethy AP. Symptom Burden and Performance Status among Community-Dwelling Patients with Serious Illness. J Palliat Med 2015;18:542-4. 14. Jang RW, Caraiscos VB, Swami N, et al. Simple prognostic model for patients with advanced cancer based on performance status. J Oncol Pract 2014;10:e335-41. 15. Lau F, Maida V, Downing M, et al. Use of the Palliative Performance Scale (PPS) for end-of-life prognostication in a palliative medicine consultation service. J Pain Symptom Manage 2009;37:965-72. 16. Myers J, Kim A, Flanagan J, Selby D. Palliative performance scale and survival among outpatients with advanced cancer. Support Care Cancer 2015;23:913-8. 17. Lindqvist O, Lundquist G, Dickman A, et al. Four essential drugs needed for quality care of the dying: a Delphi-study based international expert consensus opinion. J Palliat Med 2013;16:38-43. 18. Tait P, Morris B, To T. Core palliative medicines: meeting the needs of non-complex community patients. Aust Fam Physician 2014;43:29-32. 19. LeBlanc TW, McNeil MJ, Kamal AH, Currow DC, Abernethy AP. Polypharmacy in patients with advanced cancer and the role of medication discontinuation. Lancet Oncol 2015;16:e333-41. 20. White C, McCann MA, Jackson N. First do no harm... Terminal restlessness or druginduced delirium. J Palliat Med 2007;10:345-51. 21. Oosten AW, Oldenmenger WH, van Zuylen C, et al. Higher doses of opioids in patients who need palliative sedation prior to death: cause or consequence? Eur J Cancer 2011;47:2341-6. 22. Steinhauser KE, Christakis NA, Clipp EC, et al. Factors considered important at the end of life by patients, family, physicians, and other care providers. JAMA 2000;284:2476-2482. 23. Scott IA, Hilmer SN, Reeve E, et al. Reducing inappropriate polypharmacy: the process of deprescribing. JAMA Intern Med 2015;175:827-34.
ACCEPTED MANUSCRIPT
24. Kersten H, Wyller TB. Anticholinergic drug burden in older people's brain - how well is it measured? Basic Clin Pharmacol Toxicol 2014;114:151-9. 25. Duran CE, Azermai M, Vander Stichele RH. Systematic review of anticholinergic risk scales in older adults. Eur J Clin Pharmacol 2013;69:1485-96.
RI PT
Tables and Figures
SC
Table 1. Baseline demographic and clinical characteristics of statin trial patients in secondary analysis. SD—Standard deviation, AKPS—Australia-modified Karnofsky Performance Status, ACL—anticholinergic load
M AN U
Figure 1 ACL—Anticholinergic Load, AKPS—Australia-modified Karnofsky Performance Status, DM—disease management, SM—symptom management Distribution of mean ACL over performance status (AKPS) in a randomized controlled trial exploring statin discontinuation. Figure 1A shows the distribution among 244 patients stratified by primary diagnosis (cancer vs. non-cancer) with the number of patients in each column listed above. Figures 1B and 1C show the distribution for cancer patients (n=126) and non-cancer patients (n=118), respectively; medications are stratified by purpose (symptom or disease management).
TE D
Table 2. Multivariable ordinal logistic regression showing effect of ACL on several outcomes. OR—Odds Ratio, CI—Confidence Interval, ACL—anticholinergic load, QOL—quality of life
EP
Supplementary Content
AC C
Table S1. Sample of anticholinergic loads from the Anticholinergic Drug Scale (10) and example medication list with anticholinergic load calculation. ACL—anticholinergic load, DM—disease management, SM—symptom management, PRN—as needed, #—no assigned anticholinergic score in the ADS. Example patient medication list from week 8 of study; patient Australia-modified Karnofsky Performance Status is 60. Patient ACL calculation: (0*3) + (4*1) = 4. Table S2. Medications in statin trial not listed in ADS. Table S3. Associations between ACL, performance status, and diagnosis. ACL—anticholinergic load, DM—Disease Management, SM—Symptom Management, AKPS—Australia-modified Karnofsky Performance Status, PD—Primary Diagnosis 1 Additionally adjusted for race, gender, baseline cognitive ability, and baseline hospice status (ANCOVA)
ACCEPTED MANUSCRIPT
Table S4. Cumulative anticholinergic load in cancer and non-cancer patients grouped by drug purpose. ACL—anticholinergic load, LSM—least squares mean, DM—Disease Management, SM—Symptom Management
RI PT
Table S5. Cumulative anticholinergic load from disease and symptom management drugs as grouped by diagnosis. ACL—anticholinergic load, LSM—least squares mean, DM—Disease Management, SM—Symptom Management
SC
Table S6. Multivariable ordinal logistic regression showing effect of number of anticholinergic drugs (polypharmacy) on several outcomes. OR—Odds Ratio, CI—Confidence Interval, ACL—anticholinergic load, QOL—quality of life
AC C
EP
TE D
M AN U
Table S7. Relative drug anticholinergic contribution of symptom and disease-specific medications in non-cancer and cancer patients. Shown are the top 10 medications from each group.
ACCEPTED MANUSCRIPT Non-Cancer (N=118)
Cancer (N=126)
Total (N=244)
P value <0.000 1
Age Mean (SD)
79.1 (10.6)
69.7 (10.5)
74.3 (11.5) 0.9596
53 (44.9%)
57 (45.2%)
110 (45.1%)
RI PT
Sex Female
0.1740
104 (88.1%) 11 (9.3%) 3 (2.5%)
Ethnicity Hispanic
205 (84.0%) 34 (13.9%) 5 (2.0%)
4 (3.2%)
7 (5.9%) 99 (83.9%) 5 (4.2%) 7 (5.9%) 0 (0.0%)
Primary Diagnosis
D TE
Insurance Medicaid Medicare Other Insurance Private Uninsured
AC C
Statin Trial Group Continued Statins Discontinued Statins
0.0112
15 (6.1%) 182 (74.6%) 16 (6.6%) 30 (12.3%) 1 (0.4%) <0.000 1
29 (24.6%) 4 (3.4%) 20 (16.9%) 17 (14.4%) 0 (0.0%) 0 (0.0%) 0 (0.0%)
0 (0.0%) 0 (0.0%) 0 (0.0%) 0 (0.0%) 1 (0.8%) 9 (7.1%) 87 (69.0%)
29 (11.9%) 4 (1.6%) 20 (8.2%) 17 (7.0%) 1 (0.4%) 9 (3.7%) 87 (35.7%)
0 (0.0%)
29 (23.0%)
29 (11.9%)
37 (31.4%) 11 (9.3%)
0 (0.0%) 0 (0.0%)
37 (15.2%) 11 (4.5%)
EP
COPD Cerebrovascular Disease Congestive Heart Failure Dementia Leukemia Malignant Lymphoma Malignant Tumor (with metastases) Malignant Tumor (with no metastases) Other Primary Dx Renal Disease
8 (6.3%) 83 (65.9%) 11 (8.7%) 23 (18.3%) 1 (0.8%)
0.1929
11 (4.5%)
M AN U
7 (6.0%)
101 (80.2%) 23 (18.3%) 2 (1.6%)
SC
Race White Black Other Race
0.9933 60 (50.8%) 58 (49.2%)
64 (50.8%) 62 (49.2%)
124 (50.8%) 120 (49.2%)
ACCEPTED MANUSCRIPT <0.000 1
Baseline Cognitive Impairment? Yes
45 (38.1%)
7 (5.6%)
52 (21.3%) <0.000 1
Baseline Hospice Status? 60 (50.8%)
27 (21.6%)
<0.000 1
Charlson Comorbidity Index score Mean (SD)
3.7 (2.3)
5.8 (2.8)
50.0 (13.6)
6.3 (2.5)
Drowsiness Mean (SD)
3.1 (3.2)
Fatigue Mean (SD)
Baseline ACL Mean (SD)
D TE
3.8 (2.8)
6.7 (1.4)
3.7 (3.1)
4.8 (2.8)
<0.000 1
60.0 (13.5) 0.9625
6.5 (1.9) 0.1985 3.5 (3.2) 0.2333 4.6 (2.9) 0.4675
4.1 (2.9)
4.0 (2.9) 0.2892
2.8 (2.4)
OR per unit of ACL (95% CI) Drowsiness 1.07 (0.99 - 1.15) Fatigue 1.08 (1.002 – 1.17) QOL 0.89 (0.80 – 0.98) Wellbeing 1.07 (1.00 – 1.15)
AC C
Outcome
4.2 (2.9)
EP
Wellbeing Mean (SD)
60.0 (11.5)
M AN U
McGill Quality of Life Mean (SD)
4.8 (2.8)
SC
Performance Status (AKPS) Median (SD)
87 (35.8%)
RI PT
Enrolled in Hospice
P value 0.073 0.045 0.024 0.068
3.2 (2.6)
3.0 (2.5)
RI PT
ACCEPTED MANUSCRIPT
5
D
67
TE
36 3
2
1
6
1
124 166
134
3
86 192 6 45
EP
201 67
19
AC C
Mean ACL Scores
4
M AN U
Disease Cancer Non-cancer
SC
A. Anticholinergic loads according to primary diagnosis
11
4
18
1
0 100
90
80
70
60
50
AKPS
40
30
20
10
RI PT
ACCEPTED MANUSCRIPT
B. Anticholinergic loads in cancer patients according to drug purpose
M AN U
Medication Type Symptom Management Disease Management
4
EP
TE
D
3
2
AC C
Mean ACL Scores
SC
5
1
0
100
90
80
70
60 50 AKPS
40
30
20
10
RI PT
ACCEPTED MANUSCRIPT
C. Anticholinergic loads in non−cancer patients according to drug purpose
SC
3
M AN U
EP
TE
D
2
AC C
Mean ACL Scores
Medication Type Symptom Management Disease Management
1
0 100
90
80
70
60
50
AKPS
40
30
20
10
ACCEPTED MANUSCRIPT
AC C
Cyclobenzaprine Pimozide
Disopyramide Ranitidine
ACL 0 0 1 1 1 # 1 # 0 #
Codeine Warfarin
Gabapentin Trazodone
SC
Fluticasone Propanolol
M AN U
Type DM SM SM SM DM -DM DM DM DM
Chlorpromazine Scopolamine topical
Chlorthalidone Valproic acid
TE D
Frequency Regular Regular Regular PRN<50% Regular Regular Regular Regular Regular Regular
EP
Drug Name Ipratropium Albuterol Alprazolam Morphine Prednisone Oxygen Paroxetine Formoterol Levothyroxine Tiotropium
Benztropine Meclizine
RI PT
Anticholinergic load=3 Amitriptyline Atropine Diphenhydramine Hydroxyzine Anticholinergic load=2 Carbamazepine Cimetidine Meperidine Oxcarbazepine Anticholinergic load=1 Alprazolam Bromocriptine Diltiazem Hydrocortisone Anticholinergic load=0 Aspirin Bupropion Heparin Ipratropium
ACCEPTED MANUSCRIPT
EP
AC C
SC
RI PT
Piperazine Polyvinyl Prasugrel Pregabalin Probiotic Lysine Rasagiline Rivastigmine Roflumilast Sodium biphosphate Sevelamer Sildenafil Silodosin Sitagliptin Sodium polystyrene sulfate Sorbitol Sunitinib Tacrolimus Tazobactam Temsirolimus Tenofovir Thalidomide Thyrotropin Tiotropium Travaprost Travaprost ophthalmic Tretinoin topical Trolamine salicylate Ubiquinone Valacyclovir Vardenafil Vemurafenib Vinorelbine Vitamin A Voriconazole Zinc acetate Zinc oxide Zoledronic acid Zyflamend
M AN U
Eszopiclone Ezetimibe Fish oil/Omega 3 Fondaparinux Formoterol Gemcitabine Glucose Glutamine Glycopyrrolate Granisetron L-tryptophan Lacosamide Lactobaccilus Lactose Lapatinib Leucovorin Levetiracetam Levosalbutamol Lutein Melatonin Meloxicam Memantine Methadone Methocarbamol Methylnaltrexone Micafungin Minoxidil Modafinil Mycophenolate mofetil Nicotine Olmesartan Ondansetron Opium tincture Oseltamivir Oxaliplatin Paclitaxel Palonosetron Pancreatin Paricalcitol Pazopanib
TE D
Acetone Alfuzosin Aliskiren Alpha-galactosidase Alteplase Amikacin Aprepitant Azelaic acid Azelastine Aztreonam Baclofen Benzonatate Bevacizumab Bimatoprost Bupivacaine Buprenorphine Cabergoline Carboxymethylcellulose Cholecalciferol Cinacalcet Clobetasol Dabigatran Dalteparin Daptomycin Darbepoetin alfa Denosumab Depo-testosterone Desvenlafaxine Dextrose Dicloxacillin Docetaxel Doxercalciferol Dronabinol Duloxetine Dulasteride Efavirenz Emtricitabine Eplerenone Erlotinib Ertapenem
ACCEPTED MANUSCRIPT
P value for AKPS1 ACL (all drugs) <.0001 ACL (SM drugs) 0.0001 ACL (DM drugs) 0.0004
P value for PD1 0.6823 0.0030 <.0001
P value for AKPS*PD1 0.0279 0.0199 0.9154
AC C
EP
TE D
M AN U
SC
RI PT
Outcome
ACCEPTED MANUSCRIPT
ACL (DM drugs). Mean, LSM. 1.5, 1.3 0.8, 0.5
ACL (SM drugs). Mean, LSM. 1.7, 1.5 2.6, 2.3
P value 0.0461 <0.0001
AC C
EP
TE D
M AN U
SC
RI PT
Primary Diagnosis Non-Cancer Cancer
ACCEPTED MANUSCRIPT
P value
Cancer 3.0, 2.6 0.8, 0.6
1.5, 1.2
<0.0001
2.6, 2.3
1.7, 1.3
<0.0001
0.23
AC C
EP
TE D
M AN U
SC
ACL (all drugs). Mean, LSM. ACL (DM drugs). Mean, LSM. ACL (SM drugs). Mean, LSM.
Groups Non-Cancer 3.0, 2.4
RI PT
Outcome
ACCEPTED MANUSCRIPT
P value
0.04 0.07 0.033 0.20
RI PT
OR per Number of Anticholinergic Drugs (95% CI) Drowsiness 1.12 (1.01 - 1.24) Fatigue 1.12 (0.99 – 1.24) QOL 0.85 (0.74 – 0.99) Wellbeing 1.07 (0.97 – 1.18)
AC C
EP
TE D
M AN U
SC
Outcome
ACCEPTED MANUSCRIPT
Count ACL Contribution 36 1 36 21
1
21
18
1
18
5
3
15
15 15 4 3 4 2
1 1 3 3 2 3
15 15 12 9 8 6
Disease Management (DM) Drugs Non-Cancer Patients Rank Medication Count ACL Furosemide 37 1 1 Warfarin 19 1 2 Sertraline 11 1 3 Digoxin 10 1 4
10
8 7 6 5
1
1 1 1 1
8 7 6 5
Digoxin Famotidine Hydralazine Chlorthalidone
3 3 3 2
1 1 1 1
3 3 3 2
1
5
Diltiazem
2
1
2
TE D
10
Cancer Patients Contribution Medication Count ACL 37 Furosemide 16 1 19 Ranitidine 6 2 11 Warfarin 6 1 10 Isosorbide 4 1 (all formulations) 10 Sertraline 4 1
EP
6 7 8 9
Fluticasonesalmeterol Diltiazem Prednisone Hydralazine Isosorbide (all formulations) Paroxetine
5
AC C
5
M AN U
SC
RI PT
Symptom Management (SM) Drugs Non-Cancer Patients Cancer Patients Rank Medication Count ACL Contribution Medication Lorazepam 24 1 24 Oxycodone 1 (all formulations) Morphine 17 1 17 Dexamethasone 2 (all formulations) Oxybutynin 5 3 15 Morphine 3 (all formulations) Oxycodone 12 1 12 Diphenhydramine 4 (all formulations) Promethazine 3 3 9 Lorazepam 5 Tramadol 8 1 8 Prochlorperazine 6 Chlorpheniramine 2 3 6 Promethazine 7 Dicyclomine 2 3 6 Atropine 8 Diphenhydramine 2 3 6 Cyclobenzaprine 9 Fentanyl 5 1 5 Oxybutynin 10
Contribution 16 12 6 4 4
S0885-3924(16)30298-6
DOI:
10.1016/j.jpainsymman.2016.03.020
Reference:
JPS 9189
To appear in:
Journal of Pain and Symptom Management
Received Date: 2 December 2015 Revised Date:
18 March 2016
Accepted Date: 30 March 2016
Please cite this article as: Hochman MJ, Kamal AH, Wolf S, Samsa GP, Currow DC, Abernethy AP, LeBlanc TW, Anticholinergic drug burden in non-cancer vs. cancer patients near the end of life, Journal of Pain and Symptom Management (2016), doi: 10.1016/j.jpainsymman.2016.03.020. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
RI PT
Anticholinergic drug burden in non-cancer vs. cancer patients near the end of life Michael J. Hochman, B.S.1 Arif H. Kamal, M.D., M.H.S.2,3 Steven Wolf, M.S.4 Greg P. Samsa, Ph.D.4 David C. Currow, M.P.H., FRACP5 Amy P. Abernethy, M.D., Ph.D.3 Thomas W. LeBlanc, M.D., M.A.2,6 1
Duke University School of Medicine Durham, NC 27710 2
SC
Duke Cancer Institute Duke University School of Medicine Durham, NC 27710 3
M AN U
Flatiron Health New York, NY 10012 4
Duke Biostatistics Core Department of Biostatistics and Bioinformatics Duke University School of Medicine Durham, NC 27710 5
TE D
Discipline, Palliative and Supportive Services and Department of Medicine Flinders University Adelaide, South Australia
6
EP
Division of Hematologic Malignancies and Cellular Therapy Department of Medicine Duke University School of Medicine Durham, NC 27710
AC C
Corresponding Author: Thomas LeBlanc Box 2715 Duke University Medical Center Durham, NC 27710 Email: [email protected] Tables: 2 Figures: 1 References: 46 Word count: 2,474
ACCEPTED MANUSCRIPT
M AN U
SC
RI PT
Abstract Context. Anticholinergic drugs can cause several side effects, impairing cognition and quality of life (QOL). Cancer patients are often exposed to increasing cumulative anticholinergic load (ACL) as they approach death, but this burden has not been examined in patients with non-malignant diseases. Objectives. To determine ACL and its impact in non-cancer versus cancer palliative care patients. Methods. We performed a secondary analysis of 244 subjects enrolled in a randomized controlled trial. ACL was quantified with the Anticholinergic Drug Scale. We used multivariable regression to calculate the effect of ACL on key outcomes, including drowsiness, fatigue, and QOL. Patients were stratified by diagnosis, and drugs were grouped as symptom management (SM) or disease management (DM). Results. Overall ACL in cancer and non-cancer patients was not significantly different (2.6 vs. 2.4; P=0.23). SM drugs caused greater anticholinergic exposure than DM drugs in both cancer and non-cancer patients (2.3 vs. 0.5, and 1.5 vs. 1.3, respectively; both P<0.05), however DM drugs exposed non-cancer patients to relatively more ACL than cancer patients (1.2 vs. 0.6, P<0.0001). ACL was associated with worse fatigue (OR, 1.08; CI, 1.002-1.17) and worse QOL (OR, 0.89; CI, 0.80-0.98). Conclusions. ACL is associated with worse fatigue and QOL, and may not differ significantly between cancer and non-cancer patients nearing end of life. SM drugs are more responsible for ACL in cancer and non-cancer patients, though DM drugs contribute significantly to ACL in the latter group. We recommend more attention to reducing anticholinergic use in all patients with life-limiting illness.
TE D
Keywords Cholinergic antagonists, functional status, comorbidity, palliative care, quality of life
AC C
EP
Running Title: “Anticholinergic burden in non-cancer patients near EOL”
ACCEPTED MANUSCRIPT
RI PT
Introduction Medications with anticholinergic activity may cause several side effects including dry mouth, blurry vision, tachycardia, drowsiness, and delirium (1). Long-term cumulative anticholinergic exposure may even be a risk factor for the development of dementia (2, 3). In palliative and end-of-life settings, anticholinergic medications (e.g. atropine, opioids) are used frequently and are often necessary for managing pain, dyspnea, secretions, and anxiety, among other symptoms.
M AN U
SC
Terminally ill patients often suffer new or worsening symptoms requiring additional pharmacotherapy (4). To date, analyses of anticholinergic drug burden at end of life have focused on patients with cancer. For example, one analysis of mostly cancer patients in Australia found increasing cumulative anticholinergic exposure with proximity to death, largely from opioids (5). As patients with non-malignant life-limiting illnesses live longer and palliative care expands to include more clinically heterogeneous populations with longer life expectancies, it is also important to consider anticholinergic effects in patients with non-cancer diagnoses.
TE D
Our aim is therefore to investigate the relationship between drugs with anticholinergic activity, cancer versus non-cancer diagnosis, and key patient outcomes (drowsiness, fatigue, wellbeing, performance status, and quality of life) in a cohort of older patients with serious, life-limiting illness. We hypothesize that as patients approach death, anticholinergic load (ACL) increases and is associated with worse outcomes, such as drowsiness and fatigue. We also expect cancer patients to be exposed to greater ACL near end of life due to greater opioid use.
AC C
EP
Methods Study Design The original study was a randomized, pragmatic trial evaluating the impact of statin discontinuation near end of life on 381 patients from 15 U.S. sites (6). Eligible patients were adults with a life-limiting diagnosis and life expectancy >1 month with recent functional deterioration. The Duke University IRB approved the study, and the original paper describes the complete methodology (6). This secondary analysis is possible because statins have no anticholinergic properties; their presence or absence in a patient’s drug regimen is insignificant to this study. We thus combined both study arms and grouped patients by primary diagnosis (cancer vs. non-cancer). We included data from patients alive for at least 30 days after enrollment, with more than one follow-up with medications recorded and complete medication lists (n=244). We abstracted medication lists for up to 8 follow-ups. Assessments Baseline assessments included demographics, primary diagnosis, hospice status, cognition, performance status, quality of life (QOL), symptom severity, and medications. Outcomes included survival, resource utilization, and performance status (as measured by Australia-modified Karnofsky Performance Status [AKPS] scale (7)), collected weekly during the first month, then monthly until death (or one year). Medication inventories and
ACCEPTED MANUSCRIPT
patient-reported outcomes were collected in person or by telephone at baseline and scheduled follow-ups (0, 2, 4, 8, 12, 16, 20, and 24 weeks). Medication reports included medication name and frequency (scheduled or as needed). Symptoms and QOL were measured by the Edmonton Symptom Assessment Scale (ESAS)(8) and the McGill Quality of Life Questionnaire (MQOLQ)(9), respectively.
M AN U
SC
RI PT
Data Collection The Anticholinergic Drug Scale (ADS) is used in several studies to quantify anticholinergic burden in geriatric populations, including nursing home residents and hospital inpatients; it is correlated with serum anticholinergic activity in long-term care settings (10, 11). Individual medications are assigned a score on a 4-point scale, where 0 indicates no known anticholinergic activity, 1 indicates potential anticholinergic activity per receptor binding assays, 2 indicates clinically significant effects at excess doses, and 3 refers to marked anticholinergic effects at usual doses. ADS was originally based on reported anticholinergic effects in the literature, available laboratory data, and the independent ratings of 3 geriatric psychiatrists (12); it is a modification of the Clinicianrated Anticholinergic Scale (CrAS)(10, 11). We classified medications by their intended use, either for chronic disease management (DM) or symptom management (SM) as per prior studies (4, 5). Drugs potentially used for both purposes (anti-epileptics, antidepressants, acid suppressants) were assigned SM or DM by one author (MJH), based on likelihood of being used for maintenance of chronic illness versus treatment of acute symptoms, with subsequent review by another author (TWL).
EP
TE D
Statistical Analysis We added the ACL scores for all drugs at each available time point; an example is shown in Table S1. We treated combination medications as multiple individual medications taken at a single time point, except for those that had an ACL only in combination (e.g., fluticasone-salmeterol). We included all medications prescribed “as needed” and excluded non-specific drugs (“inhaler,” “leg med”) and non-drugs (“oxygen”). One hundred nineteen of 413 unique medications and supplements were not listed in ADS and were excluded from the analysis. A list of those medications is included in the supplementary content (Table S2).
AC C
We calculated descriptive statistics for patients at baseline based on their primary diagnosis, and then compared cancer and non-cancer patients using the Chi-Square and Fisher’s Exact test for discrete variables and the Kruskal-Wallis test for continuous variables. To trace how ACL varied longitudinally as patients approached death, we used performance status to follow disease progression and looked at encounters as the unit of analysis instead of patients. We took this approach because follow-ups were not spaced evenly in the original trial design, and data were highly likely to be missing not-atrandom due to patient decline and death. Functional status has been used to explore ACL (5) and track symptom burden (13) in palliative care patients; it has also been found to be a significant predictor of survival in these populations (14-16). By evaluating ACL based on performance status (AKPS), we maintained our sample size (n=244; 1,210 encounters).
ACCEPTED MANUSCRIPT
SC
RI PT
We calculated average ACL by determining the standard and least squares mean (LSM) of all encounters’ ACLs over the course of the study by adjusting for AKPS, and then stratified by cancer diagnosis and drug type (DM vs. SM). We used ANCOVA to test for statistical significance between the groups and ACL types while adjusting for AKPS, disease status, race, gender, cognitive status, and hospice status at baseline. Using multivariable ordinal logistic regression, we evaluated associations between overall ACL, QOL, and symptoms, adjusting for age, race, gender, primary diagnosis, baseline cognitive impairment, baseline hospice status, and performance status (with AKPS grouped as 10-30 [poor], 40-60 [moderate], and 70-100 [good]). As a secondary analysis, we compared the standard mean and LSM of ACL from DM drugs with that from SM drugs in cancer patients and non-cancer patients separately. Additionally, to account for the effects of polypharmacy, we examined the number of anticholinergic drugs patients used over the course of the study and tested whether this predicted our outcomes of interest. Statistics were done using SAS, version 9.4 (SAS Institute Inc).
M AN U
Finally, we ranked the relative contribution of drugs to cumulative anticholinergic exposure. We abstracted medication lists for the first time patients dropped below an AKPS of 70 to any lower performance level (n=165). We chose this point in order to evaluate medications taken as functional status declined at a standardized point in the illness trajectory. We calculated the relative contribution of drugs to patient ACL by multiplying the number of times a drug appeared on a patient’s list (the “count”) with its ACL score from ADS.
EP
TE D
Results Baseline Characteristics About half (n=118, 48.4%) of patients had a non-malignant disease (Table 1). On average, non-cancer patients were older (79.1 vs. 69.7 years; P<0.0001) and had worse performance status (median AKPS 50.0 vs. 60.0; P<0.0001). Differences in QOL and average ACL (6.7 vs. 6.3 and 3.2 vs. 2.8, respectively) were not statistically significant.
AC C
Longitudinal Anticholinergic Exposure Figure 1A displays how ACL varied in cancer and non-cancer patients as they approached death. Cancer patient ACL peaked at an AKPS of 30 (n=19), indicating a state of being almost completely bedbound; non-cancer patients’ ACL peaked at an AKPS of 40 (bedbound more than half of the time, n=134) and again at 10 (comatose status, n=3). Primary diagnosis and AKPS were both significantly associated with ACL (Table S3). Cancer patients received a larger ACL from SM drugs than DM drugs (2.3 vs. 0.5 ACL, least squares mean [LSM]; P<0.0001; see Figure 1B), as did non-cancer patients (1.5 SM vs. 1.3 DM ACL, LSM; P=0.04; see Figure 1C and Table S4). Overall, however, symptom management drugs exposed cancer patients to a higher ACL than non-cancer patients (2.3 vs. 1.3 ACL, LSM; P<0.0001), and chronic disease management drugs burdened non-cancer patients more than cancer patients (1.2 vs. 0.6 ACL, LSM;
ACCEPTED MANUSCRIPT
P<0.0001). Cumulative anticholinergic exposure between the two groups was not significantly different (2.6 vs. 2.4 ACL, LSM; P=0.23; see Table S5).
SC
RI PT
Multivariable Analysis In multivariable analysis, we explored associations between ACL and several other measures (Table 2). Increased ACL was significantly associated with worse patient fatigue (OR, 1.08; CI, 1.002-1.17; P=0.045) and worse QOL (OR, 0.89; CI, 0.80-0.98; P=0.02) after adjusting for performance status, baseline hospice status, baseline cognitive status, and demographics. There was no association between anticholinergic exposure and drowsiness or wellbeing. In a sensitivity analysis examining the relationship between number of anticholinergic drugs and our outcomes of interest, we found that increased anticholinergic drug usage was significantly associated with drowsiness (OR, 1.12; CI, 1.01-1.24; P=0.04) and worse QOL (OR 0.85; CI, 0.74-0.99; Table S6).
M AN U
Anticholinergic Contribution In cancer and non-cancer patients, the SM drugs that contributed most to anticholinergic burden were opioids and benzodiazepines. Furosemide and warfarin were common DM drugs to burden both patient groups with anticholinergic effects. Rank lists for drug anticholinergic contribution can be found in the supplementary content (Table S7). Cancer patients used more drugs with high ACL scores (ACL 3) than non-cancer patients (0.4 vs. 0.2, P<0.0002); the opposite held true for non-cancer patients consuming more low ACL (1-2) drugs (2.3 vs. 2.0, P<0.0011).
EP
TE D
Discussion Using data from a large, prospective, randomized controlled trial, our longitudinal analysis of ACL among patients near the end of life yielded three important findings: (1) patients with non-cancer diagnoses are exposed to similar cumulative anticholinergic burden as cancer patients near the end of life; (2) much of the anticholinergic drug burden near end of life is due to SM drugs, regardless of cancer versus non-cancer diagnosis (although DM drugs more significantly burden patients without cancer); and (3) drug anticholinergic activity is associated with worsening fatigue and quality of life.
AC C
We were surprised to find that non-cancer patients are exposed to similar ACL compared to cancer patients near end-of-life, despite less frequent use of opioids. This suggests significant use of medications with anticholinergic effects among all patients near the end of life, regardless of primary diagnosis. Although expert palliative care panels have argued that non-complex patients only require a limited number of drugs for quality symptom control near end of life, many of these have anticholinergic activity (17, 18). Our results highlight the challenge of tailoring medication use in patients with lifelimiting illnesses to their needs, and the importance of targeted attention to polypharmacy and anticholinergic use among all patients nearing end of life (19). In both cancer and non-cancer patients, ACL was largely due to SM drugs like opioids and benzodiazepines. This result emphasizes the significance and utility of these medications in palliative and hospice care. Pain and subsequent opioid use has been
ACCEPTED MANUSCRIPT
linked to terminal restlessness (20) and the need for palliative sedation (21) in end-of-life care. Although patients find both mental alertness and minimal pain important near the end of life, palliative care clinicians tend to value terminal lucidity less than their patients (22). Clearly, we must recognize the benefit of adequate pain management but balance it with reasonable cognitive and performance statuses.
SC
RI PT
We also found that anticholinergic drug burden due to DM drugs is relatively higher in non-cancer patients than cancer patients. This is a critical finding. As terminal illness is considered an indication for deprescribing (23), patients with chronic, life-threatening, non-malignant diseases may warrant targeted attention to decrease cumulative anticholinergic exposure via the thoughtful discontinuation of DM agents that are no longer indicated. Furthermore, our results suggest that clinicians should mind patient diagnosis when prescribing medications at end of life, as a more thorough evaluation of DM drugs may facilitate optimal care for terminal non-cancer patients.
TE D
M AN U
Finally, we found that drug ACL is independently associated with worse fatigue and quality of life. The magnitude of these effects was small, yet collectively suggests that cumulative anticholinergic effects are burdensome. These results are concordant with Agar et al., who report adjusted odds ratios and confidence intervals of 0.85 (0.81-0.90) for AKPS and 0.90 (0.85-0.95) for QOL per unit of ACL (5). Because that study included few non-cancer patients and no patients with baseline cognitive impairment, our findings add to the generalizability of this literature. In our sensitivity analysis, we found that the average number of anticholinergic drugs consumed had a detrimental relationship with patient QOL and drowsiness, indicating that anticholinergic polypharmacy and drug potency (as measured by ACL) similarly correspond with this pharmacologic burden. We recommend that future studies focus on whether interventions to reduce anticholinergic exposure can be safely and successfully applied to patients with significant, complex illness near end of life.
AC C
EP
Our approach has a few limitations. First, using performance status to trace disease progression is limited, as it does not necessarily decline linearly with time; some patients remained at the same performance level for the study duration and others varied between multiple levels. Second, a linear additive model for cumulative anticholinergic effects based on ADS does not reflect the nuances of drug-patient interactions or the dosedependence of anticholinergic effects (11, 24). Additionally, different scales assign anticholinergic potency based on varied methods. A different scale would potentially result in different outcomes to our study (25). Third, not all medications could be assigned an ACL score on the scale—a limitation narrowed by the fact that the scale is comprehensive for most medications known to have anticholinergic activity. Fourth, we have a limited ability to assign drug purpose or measure true drug exposure through patient interviews or other means, including dispensing records or serum drug concentrations. Finally, it is difficult to ascertain whether anticholinergic drugs negatively impact performance status near the end of life. Sicker patients with worse QOL may be consuming more anticholinergic drugs for symptom palliation simply because SM drugs are typically added as patients become sicker (4, 5), reflecting a natural tendency in the terminal care of patients. Still, as a robust secondary analysis of a
ACCEPTED MANUSCRIPT
rigorous randomized trial, our work represents a comprehensive and generalizable approach to ACL near end of life.
SC
RI PT
In conclusion, overall anticholinergic exposure may not differ significantly between cancer and non-cancer patients near the end of life, but it appears to be detrimental to patient fatigue and quality of life. SM drugs are most responsible for anticholinergic burden in cancer and non-cancer palliative care patients, and they burden cancer patients relatively more than DM drugs, while chronic disease-specific drugs carry a more significant burden in non-cancer patients near the end of life. Further studies to assess clinical benefit in decreasing anticholinergic burden in palliative care patients are warranted.
TE D
M AN U
Disclosures and Acknowledgments Funding for the statin discontinuation study that generated the data for this analysis was provided by the National Institute of Nursing Research (grant numbers UC4-NR012584 and U24-NR014637). This work also included the support of facilities and resources of the Veterans Affairs Health Care System (eg, Phoenix, Arizona, and Birmingham, Alabama). Dr. LeBlanc received funding support from a Junior Career Development Grant from the National Palliative Care Research Center during the time that this analysis was done. He currently receives support from a Sojourns Scholar Award Grant from the Cambia Health Foundation. The design, conduct, analysis, and write-up of the study were performed independently from any sponsoring agency or funding. No other authors have relevant conflicts of interest to report. The authors would like to acknowledge Michael McNeil for his original ideas during early portions of this work.
AC C
EP
References 1. Mintzer J, Burns A. Anticholinergic side-effects of drugs in elderly people. J R Soc Med 2000;93:457-62. 2. Gray SL, Anderson ML, Dublin S, et al. Cumulative Use of Strong Anticholinergics and Incident Dementia: A Prospective Cohort Study. JAMA Intern Med 2015. 3. Jessen F, Kaduszkiewicz H, Daerr M, et al. Anticholinergic drug use and risk for dementia: target for dementia prevention. Eur Arch Psychiatry Clin Neurosci 2010;260 Suppl 2:S111-5. 4. Currow DC, Stevenson JP, Abernethy AP, Plummer J, Shelby-James TM. Prescribing in palliative care as death approaches. J Am Geriatr Soc 2007;55:590-5. 5. Agar M, Currow D, Plummer J, et al. Changes in anticholinergic load from regular prescribed medications in palliative care as death approaches. Palliat Med 2009;23:25765. 6. Kutner JS, Blatchford PJ, Taylor DH, et al. Safety and Benefit of Discontinuing Statin Therapy in the Setting of Advanced, Life-Limiting Illness: A Randomized Clinical Trial. JAMA Intern Med 2015.
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
7. Abernethy AP, Shelby-James T, Fazekas BS, Woods D, Currow DC. The Australiamodified Karnofsky Performance Status (AKPS) scale: a revised scale for contemporary palliative care clinical practice [ISRCTN81117481]. BMC Palliat Care 2005;4:7. 8. Bruera E, Kuehn N, Miller MJ, Selmser P, Macmillan K. The Edmonton Symptom Assessment System (ESAS): a simple method for the assessment of palliative care patients. J Palliat Care 1991;7:6-9. 9. Cohen SR, Mount BM, Strobel MG, Bui F. The McGill Quality of Life Questionnaire: a measure of quality of life appropriate for people with advanced disease. A preliminary study of validity and acceptability. Palliat Med 1995;9:207-19. 10. Carnahan RM, Lund BC, Perry PJ, Pollock BG, Culp KR. The Anticholinergic Drug Scale as a measure of drug-related anticholinergic burden: associations with serum anticholinergic activity. J Clin Pharmacol 2006;46:1481-6. 11. Salahudeen MS, Duffull SB, Nishtala PS. Anticholinergic burden quantified by anticholinergic risk scales and adverse outcomes in older people: a systematic review. BMC Geriatr 2015;15:31. 12. Han L, McCusker J, Cole M, et al. Use of medications with anticholinergic effect predicts clinical severity of delirium symptoms in older medical inpatients. Arch Intern Med 2001;161:1099-105. 13. Kamal AH, Nipp RD, Bull J, Stinson CS, Abernethy AP. Symptom Burden and Performance Status among Community-Dwelling Patients with Serious Illness. J Palliat Med 2015;18:542-4. 14. Jang RW, Caraiscos VB, Swami N, et al. Simple prognostic model for patients with advanced cancer based on performance status. J Oncol Pract 2014;10:e335-41. 15. Lau F, Maida V, Downing M, et al. Use of the Palliative Performance Scale (PPS) for end-of-life prognostication in a palliative medicine consultation service. J Pain Symptom Manage 2009;37:965-72. 16. Myers J, Kim A, Flanagan J, Selby D. Palliative performance scale and survival among outpatients with advanced cancer. Support Care Cancer 2015;23:913-8. 17. Lindqvist O, Lundquist G, Dickman A, et al. Four essential drugs needed for quality care of the dying: a Delphi-study based international expert consensus opinion. J Palliat Med 2013;16:38-43. 18. Tait P, Morris B, To T. Core palliative medicines: meeting the needs of non-complex community patients. Aust Fam Physician 2014;43:29-32. 19. LeBlanc TW, McNeil MJ, Kamal AH, Currow DC, Abernethy AP. Polypharmacy in patients with advanced cancer and the role of medication discontinuation. Lancet Oncol 2015;16:e333-41. 20. White C, McCann MA, Jackson N. First do no harm... Terminal restlessness or druginduced delirium. J Palliat Med 2007;10:345-51. 21. Oosten AW, Oldenmenger WH, van Zuylen C, et al. Higher doses of opioids in patients who need palliative sedation prior to death: cause or consequence? Eur J Cancer 2011;47:2341-6. 22. Steinhauser KE, Christakis NA, Clipp EC, et al. Factors considered important at the end of life by patients, family, physicians, and other care providers. JAMA 2000;284:2476-2482. 23. Scott IA, Hilmer SN, Reeve E, et al. Reducing inappropriate polypharmacy: the process of deprescribing. JAMA Intern Med 2015;175:827-34.
ACCEPTED MANUSCRIPT
24. Kersten H, Wyller TB. Anticholinergic drug burden in older people's brain - how well is it measured? Basic Clin Pharmacol Toxicol 2014;114:151-9. 25. Duran CE, Azermai M, Vander Stichele RH. Systematic review of anticholinergic risk scales in older adults. Eur J Clin Pharmacol 2013;69:1485-96.
RI PT
Tables and Figures
SC
Table 1. Baseline demographic and clinical characteristics of statin trial patients in secondary analysis. SD—Standard deviation, AKPS—Australia-modified Karnofsky Performance Status, ACL—anticholinergic load
M AN U
Figure 1 ACL—Anticholinergic Load, AKPS—Australia-modified Karnofsky Performance Status, DM—disease management, SM—symptom management Distribution of mean ACL over performance status (AKPS) in a randomized controlled trial exploring statin discontinuation. Figure 1A shows the distribution among 244 patients stratified by primary diagnosis (cancer vs. non-cancer) with the number of patients in each column listed above. Figures 1B and 1C show the distribution for cancer patients (n=126) and non-cancer patients (n=118), respectively; medications are stratified by purpose (symptom or disease management).
TE D
Table 2. Multivariable ordinal logistic regression showing effect of ACL on several outcomes. OR—Odds Ratio, CI—Confidence Interval, ACL—anticholinergic load, QOL—quality of life
EP
Supplementary Content
AC C
Table S1. Sample of anticholinergic loads from the Anticholinergic Drug Scale (10) and example medication list with anticholinergic load calculation. ACL—anticholinergic load, DM—disease management, SM—symptom management, PRN—as needed, #—no assigned anticholinergic score in the ADS. Example patient medication list from week 8 of study; patient Australia-modified Karnofsky Performance Status is 60. Patient ACL calculation: (0*3) + (4*1) = 4. Table S2. Medications in statin trial not listed in ADS. Table S3. Associations between ACL, performance status, and diagnosis. ACL—anticholinergic load, DM—Disease Management, SM—Symptom Management, AKPS—Australia-modified Karnofsky Performance Status, PD—Primary Diagnosis 1 Additionally adjusted for race, gender, baseline cognitive ability, and baseline hospice status (ANCOVA)
ACCEPTED MANUSCRIPT
Table S4. Cumulative anticholinergic load in cancer and non-cancer patients grouped by drug purpose. ACL—anticholinergic load, LSM—least squares mean, DM—Disease Management, SM—Symptom Management
RI PT
Table S5. Cumulative anticholinergic load from disease and symptom management drugs as grouped by diagnosis. ACL—anticholinergic load, LSM—least squares mean, DM—Disease Management, SM—Symptom Management
SC
Table S6. Multivariable ordinal logistic regression showing effect of number of anticholinergic drugs (polypharmacy) on several outcomes. OR—Odds Ratio, CI—Confidence Interval, ACL—anticholinergic load, QOL—quality of life
AC C
EP
TE D
M AN U
Table S7. Relative drug anticholinergic contribution of symptom and disease-specific medications in non-cancer and cancer patients. Shown are the top 10 medications from each group.
ACCEPTED MANUSCRIPT Non-Cancer (N=118)
Cancer (N=126)
Total (N=244)
P value <0.000 1
Age Mean (SD)
79.1 (10.6)
69.7 (10.5)
74.3 (11.5) 0.9596
53 (44.9%)
57 (45.2%)
110 (45.1%)
RI PT
Sex Female
0.1740
104 (88.1%) 11 (9.3%) 3 (2.5%)
Ethnicity Hispanic
205 (84.0%) 34 (13.9%) 5 (2.0%)
4 (3.2%)
7 (5.9%) 99 (83.9%) 5 (4.2%) 7 (5.9%) 0 (0.0%)
Primary Diagnosis
D TE
Insurance Medicaid Medicare Other Insurance Private Uninsured
AC C
Statin Trial Group Continued Statins Discontinued Statins
0.0112
15 (6.1%) 182 (74.6%) 16 (6.6%) 30 (12.3%) 1 (0.4%) <0.000 1
29 (24.6%) 4 (3.4%) 20 (16.9%) 17 (14.4%) 0 (0.0%) 0 (0.0%) 0 (0.0%)
0 (0.0%) 0 (0.0%) 0 (0.0%) 0 (0.0%) 1 (0.8%) 9 (7.1%) 87 (69.0%)
29 (11.9%) 4 (1.6%) 20 (8.2%) 17 (7.0%) 1 (0.4%) 9 (3.7%) 87 (35.7%)
0 (0.0%)
29 (23.0%)
29 (11.9%)
37 (31.4%) 11 (9.3%)
0 (0.0%) 0 (0.0%)
37 (15.2%) 11 (4.5%)
EP
COPD Cerebrovascular Disease Congestive Heart Failure Dementia Leukemia Malignant Lymphoma Malignant Tumor (with metastases) Malignant Tumor (with no metastases) Other Primary Dx Renal Disease
8 (6.3%) 83 (65.9%) 11 (8.7%) 23 (18.3%) 1 (0.8%)
0.1929
11 (4.5%)
M AN U
7 (6.0%)
101 (80.2%) 23 (18.3%) 2 (1.6%)
SC
Race White Black Other Race
0.9933 60 (50.8%) 58 (49.2%)
64 (50.8%) 62 (49.2%)
124 (50.8%) 120 (49.2%)
ACCEPTED MANUSCRIPT <0.000 1
Baseline Cognitive Impairment? Yes
45 (38.1%)
7 (5.6%)
52 (21.3%) <0.000 1
Baseline Hospice Status? 60 (50.8%)
27 (21.6%)
<0.000 1
Charlson Comorbidity Index score Mean (SD)
3.7 (2.3)
5.8 (2.8)
50.0 (13.6)
6.3 (2.5)
Drowsiness Mean (SD)
3.1 (3.2)
Fatigue Mean (SD)
Baseline ACL Mean (SD)
D TE
3.8 (2.8)
6.7 (1.4)
3.7 (3.1)
4.8 (2.8)
<0.000 1
60.0 (13.5) 0.9625
6.5 (1.9) 0.1985 3.5 (3.2) 0.2333 4.6 (2.9) 0.4675
4.1 (2.9)
4.0 (2.9) 0.2892
2.8 (2.4)
OR per unit of ACL (95% CI) Drowsiness 1.07 (0.99 - 1.15) Fatigue 1.08 (1.002 – 1.17) QOL 0.89 (0.80 – 0.98) Wellbeing 1.07 (1.00 – 1.15)
AC C
Outcome
4.2 (2.9)
EP
Wellbeing Mean (SD)
60.0 (11.5)
M AN U
McGill Quality of Life Mean (SD)
4.8 (2.8)
SC
Performance Status (AKPS) Median (SD)
87 (35.8%)
RI PT
Enrolled in Hospice
P value 0.073 0.045 0.024 0.068
3.2 (2.6)
3.0 (2.5)
RI PT
ACCEPTED MANUSCRIPT
5
D
67
TE
36 3
2
1
6
1
124 166
134
3
86 192 6 45
EP
201 67
19
AC C
Mean ACL Scores
4
M AN U
Disease Cancer Non-cancer
SC
A. Anticholinergic loads according to primary diagnosis
11
4
18
1
0 100
90
80
70
60
50
AKPS
40
30
20
10
RI PT
ACCEPTED MANUSCRIPT
B. Anticholinergic loads in cancer patients according to drug purpose
M AN U
Medication Type Symptom Management Disease Management
4
EP
TE
D
3
2
AC C
Mean ACL Scores
SC
5
1
0
100
90
80
70
60 50 AKPS
40
30
20
10
RI PT
ACCEPTED MANUSCRIPT
C. Anticholinergic loads in non−cancer patients according to drug purpose
SC
3
M AN U
EP
TE
D
2
AC C
Mean ACL Scores
Medication Type Symptom Management Disease Management
1
0 100
90
80
70
60
50
AKPS
40
30
20
10
ACCEPTED MANUSCRIPT
AC C
Cyclobenzaprine Pimozide
Disopyramide Ranitidine
ACL 0 0 1 1 1 # 1 # 0 #
Codeine Warfarin
Gabapentin Trazodone
SC
Fluticasone Propanolol
M AN U
Type DM SM SM SM DM -DM DM DM DM
Chlorpromazine Scopolamine topical
Chlorthalidone Valproic acid
TE D
Frequency Regular Regular Regular PRN<50% Regular Regular Regular Regular Regular Regular
EP
Drug Name Ipratropium Albuterol Alprazolam Morphine Prednisone Oxygen Paroxetine Formoterol Levothyroxine Tiotropium
Benztropine Meclizine
RI PT
Anticholinergic load=3 Amitriptyline Atropine Diphenhydramine Hydroxyzine Anticholinergic load=2 Carbamazepine Cimetidine Meperidine Oxcarbazepine Anticholinergic load=1 Alprazolam Bromocriptine Diltiazem Hydrocortisone Anticholinergic load=0 Aspirin Bupropion Heparin Ipratropium
ACCEPTED MANUSCRIPT
EP
AC C
SC
RI PT
Piperazine Polyvinyl Prasugrel Pregabalin Probiotic Lysine Rasagiline Rivastigmine Roflumilast Sodium biphosphate Sevelamer Sildenafil Silodosin Sitagliptin Sodium polystyrene sulfate Sorbitol Sunitinib Tacrolimus Tazobactam Temsirolimus Tenofovir Thalidomide Thyrotropin Tiotropium Travaprost Travaprost ophthalmic Tretinoin topical Trolamine salicylate Ubiquinone Valacyclovir Vardenafil Vemurafenib Vinorelbine Vitamin A Voriconazole Zinc acetate Zinc oxide Zoledronic acid Zyflamend
M AN U
Eszopiclone Ezetimibe Fish oil/Omega 3 Fondaparinux Formoterol Gemcitabine Glucose Glutamine Glycopyrrolate Granisetron L-tryptophan Lacosamide Lactobaccilus Lactose Lapatinib Leucovorin Levetiracetam Levosalbutamol Lutein Melatonin Meloxicam Memantine Methadone Methocarbamol Methylnaltrexone Micafungin Minoxidil Modafinil Mycophenolate mofetil Nicotine Olmesartan Ondansetron Opium tincture Oseltamivir Oxaliplatin Paclitaxel Palonosetron Pancreatin Paricalcitol Pazopanib
TE D
Acetone Alfuzosin Aliskiren Alpha-galactosidase Alteplase Amikacin Aprepitant Azelaic acid Azelastine Aztreonam Baclofen Benzonatate Bevacizumab Bimatoprost Bupivacaine Buprenorphine Cabergoline Carboxymethylcellulose Cholecalciferol Cinacalcet Clobetasol Dabigatran Dalteparin Daptomycin Darbepoetin alfa Denosumab Depo-testosterone Desvenlafaxine Dextrose Dicloxacillin Docetaxel Doxercalciferol Dronabinol Duloxetine Dulasteride Efavirenz Emtricitabine Eplerenone Erlotinib Ertapenem
ACCEPTED MANUSCRIPT
P value for AKPS1 ACL (all drugs) <.0001 ACL (SM drugs) 0.0001 ACL (DM drugs) 0.0004
P value for PD1 0.6823 0.0030 <.0001
P value for AKPS*PD1 0.0279 0.0199 0.9154
AC C
EP
TE D
M AN U
SC
RI PT
Outcome
ACCEPTED MANUSCRIPT
ACL (DM drugs). Mean, LSM. 1.5, 1.3 0.8, 0.5
ACL (SM drugs). Mean, LSM. 1.7, 1.5 2.6, 2.3
P value 0.0461 <0.0001
AC C
EP
TE D
M AN U
SC
RI PT
Primary Diagnosis Non-Cancer Cancer
ACCEPTED MANUSCRIPT
P value
Cancer 3.0, 2.6 0.8, 0.6
1.5, 1.2
<0.0001
2.6, 2.3
1.7, 1.3
<0.0001
0.23
AC C
EP
TE D
M AN U
SC
ACL (all drugs). Mean, LSM. ACL (DM drugs). Mean, LSM. ACL (SM drugs). Mean, LSM.
Groups Non-Cancer 3.0, 2.4
RI PT
Outcome
ACCEPTED MANUSCRIPT
P value
0.04 0.07 0.033 0.20
RI PT
OR per Number of Anticholinergic Drugs (95% CI) Drowsiness 1.12 (1.01 - 1.24) Fatigue 1.12 (0.99 – 1.24) QOL 0.85 (0.74 – 0.99) Wellbeing 1.07 (0.97 – 1.18)
AC C
EP
TE D
M AN U
SC
Outcome
ACCEPTED MANUSCRIPT
Count ACL Contribution 36 1 36 21
1
21
18
1
18
5
3
15
15 15 4 3 4 2
1 1 3 3 2 3
15 15 12 9 8 6
Disease Management (DM) Drugs Non-Cancer Patients Rank Medication Count ACL Furosemide 37 1 1 Warfarin 19 1 2 Sertraline 11 1 3 Digoxin 10 1 4
10
8 7 6 5
1
1 1 1 1
8 7 6 5
Digoxin Famotidine Hydralazine Chlorthalidone
3 3 3 2
1 1 1 1
3 3 3 2
1
5
Diltiazem
2
1
2
TE D
10
Cancer Patients Contribution Medication Count ACL 37 Furosemide 16 1 19 Ranitidine 6 2 11 Warfarin 6 1 10 Isosorbide 4 1 (all formulations) 10 Sertraline 4 1
EP
6 7 8 9
Fluticasonesalmeterol Diltiazem Prednisone Hydralazine Isosorbide (all formulations) Paroxetine
5
AC C
5
M AN U
SC
RI PT
Symptom Management (SM) Drugs Non-Cancer Patients Cancer Patients Rank Medication Count ACL Contribution Medication Lorazepam 24 1 24 Oxycodone 1 (all formulations) Morphine 17 1 17 Dexamethasone 2 (all formulations) Oxybutynin 5 3 15 Morphine 3 (all formulations) Oxycodone 12 1 12 Diphenhydramine 4 (all formulations) Promethazine 3 3 9 Lorazepam 5 Tramadol 8 1 8 Prochlorperazine 6 Chlorpheniramine 2 3 6 Promethazine 7 Dicyclomine 2 3 6 Atropine 8 Diphenhydramine 2 3 6 Cyclobenzaprine 9 Fentanyl 5 1 5 Oxybutynin 10
Contribution 16 12 6 4 4