Support Care Cancer DOI 10.1007/s00520-015-3007-z

ORIGINAL ARTICLE

Polypharmacy in the terminal stage of cancer

Katharina A. Kierner1 & Dietmar Weixler2 & Eva K. Masel1 & Verena Gartner1 & Herbert H. Watzke1

Received: 29 July 2015 /Accepted: 27 October 2015 # Springer-Verlag Berlin Heidelberg 2015

Abstracts administration. The fulfilled the requirements of Purpose This study aims to determine the numbers of patients in the majority of patients; 90 % received treat- with advanced cancer receiving polypharmacy at the end of ment for pain and anxiety. Patients treated at the palliative their lives and analyze differences in drug prescription at a ward received more and psychoactive drugs while general oncology ward and a dedicated palliative care ward. those at the oncology ward received more anti-cancer drugs Methods A retrospective single-center cohort study at a uni- and fluids. versity hospital with a large cancer center was conducted. The Conclusions Polypharmacy still is a problem in the large ma- charts of 100 patients who had died because of advanced jority of patients with terminal cancer. Further studies should cancer were reviewed; data concerning sociodemographic be focused on the patients’ quality of life, drug interactions, variables and medications were collected at four predefined and adverse events. time points (9, 6, 3, 0 days before death). Results Nine days before death, polypharmacy was registered Keywords Polypharmacy . Palliative care . Advanced in 95 % of patients; they had prescriptions for 11 (9–13) dif- cancer . End of life ferent medications per day (median, IQR). Although this num- ber dropped significantly, on the last day as many as 61 % of the patients were still taking more than 4 drugs (median 6.5, Introduction IQR 4–9). No significant difference was noted between the oncology ward and the palliative care ward. Polypharmacy Polypharmacy is an increasing problem. In 1999, 25 % of the was largely dependent on the patients’ ECOG performance US-American population was receiving five or more medica- status as well as the type of ward, the number of days before tions. In the high-risk group of the elderly, 44 % of men and death, and age. It was not influenced by gender, the duration 57 % of women were taking five or more drugs [1]. This of hospital stays, and the devices facilitating drug implies a variety of potential drug interactions, adverse events, and economic problems. An increasing number of reports on polypharmacy have been published since the 1960s, mainly in Electronic supplementary material The online version of this article journals of and anesthesia [2–4]. At the time, (doi:10.1007/s00520-015-3007-z) contains supplementary material, which is available to authorized users. polypharmacy was defined as a combination of different drugs [5]; the exact number of drugs was not specified. Even today, * Katharina A. Kierner it is difficult to establish the precise definition of [email protected] polypharmacy. Veehof et al. define polypharmacy as the long-term concomitant intake of two or more medications 1 Unit of Palliative Care, Department of Internal Medicine I, for more than 240 days in a year [6]. In other publications, University Hospital of Vienna, Waehringer Guertel 18-20, polypharmacy is defined as the concomitant use of more than A-1090 Vienna, Austria four [7, 8]orfive[9, 10]drugs.Furthermore,theterm 2 Department of Anesthesiology, Head of Palliative Care Team, Horn polypharmacy should rightly include over-the-counter drugs Hospital, Horn, Austria and herbal or supplementary substances in addition to Support Care Cancer prescribed medications. Many patients depend on their con- diseases differs from that of solid cancers. Hematological can- comitant medications because of their . cers are associated with the use of different medications be- Therefore, polypharmacy could well be established by cause the patients tend to progress faster from curative to counting the number of potentially inappropriate medications palliative treatment. taken by an individual [11]. In recent studies, polypharmacy Patients were stratified by the medical team (either a spe- has also been defined as underuse (the indicated cialized team of oncologists at a dedicated oncology ward or a drug is not used) or medication duplication (similar medica- specialized palliative care team at a dedicated palliative care tions are used) [12]. ward). The same quality of documentation was used for the Polypharmacy is a widespread phenomenon in palliative medical charts of both wards. care. Due to their advanced age, many patients in a palliative Sociodemographic, disease-related, and medical variables setting suffer from comorbidities. Multiple medications are were retrieved for each patient. We determined the number of prescribed to address these conditions. Furthermore, symp- each prescribed medication at four predefined time points: 9 toms such as dyspnea or nausea impair quality of life in pal- (=day −9), 6 (=day −6), 3 (=day −3) days before death and the liative care patients. Further drugs are needed to treat these day of death (=day 0). Drugs were categorized into the fol- symptoms. Some patients in the palliative care setting contin- lowing groups: opioids, non- analgesics, psychoactive ue to receive anti-cancer therapy in order to arrest further drugs (, anti-psychotic drugs, benzodiaze- tumor proliferation. Recently, LeBlanc et al. [13]presenteda pines, ), anti-thrombotic medications, anti-microbial concise review of the existing literature on polypharmacy in therapy (antibiotics, anti-fungal, anti-viral drugs), cardiovas- patients with advanced cancer and the role of drug discontin- cular medication (including anti-hypertensive drugs, anti- uation in this setting. arrhythmic agents, and diuretics), anti-tumor agents, fluids Patients receiving polypharmacy frequently experience (including crystalloids, colloids, parenteral nutrition, and drug-drug interactions and adverse drug reactions [14–17]. blood products), and others (including anti-epileptic drugs, The effects of pharmacological interactions include more fre- antacids, laxatives, corticosteroids, anti-emetic agents, and a quent hospital admissions as well as higher morbidity and variety of other compounds) (Appendix A). Any drug pre- mortality rates [18, 19]. scribed for the day of assessment alone (+/− 1 day) for a Palliative care outpatients are reported to receive a median new short-term condition was classified as acute medication number of six medications [20]. However, nearly all of these while any drug prescribed earlier than 1 day before or after the publications focus on patients with a fairly acceptable perfor- day of evaluation, as part of an intermittent treatment protocol, mance status. Just a few studies deal with polypharmacy at the was termed chronic medication. Polypharmacy was defined as end of life. Frechen et al. investigated the potential of drug- the prescription of five or more different drugs per day. drug interactions 2 weeks prior to death [21]. To our knowl- The study was conducted in accordance with the edge, the use of medications close to the time of death and the Declaration of Helsinki. use of medications in patients undergoing treatment by a ded- icated palliative care team have not been analyzed so far. Statistical analysis In the present study, we evaluated the number of medica- tions being taken by patients with advanced cancer at the end Continuous variables were expressed as means and standard of their lives. Patients treated by specialized oncologists were errors of means (SEM) or medians and interquartile ranges compared to those treated by a dedicated palliative care team. (IQR). Groups were compared with a paired or unpaired Student’s t test, Mann–Whitney U test, Kruskal-Wallis test, or analysis of variance (ANOVA), as appropriate. Material and methods Categorical variables were expressed as numbers and percent- ages of group size, and groups compared with the Chi-square The investigation was designed as a retrospective, longitudi- test. Two-sided P values <0.05 were considered statistically nal, single-center cohort study. Medical charts of 100 patients significant. Logistic regression modeling and multivariate over 18 years of age who had died from advanced cancer analysis were used to identify factors associated with (metastatic solid tumors) between January 2011 and polypharmacy. The IBM SPSS v20.0 software was used for March 2013 were reviewed. Only patients who had had a calculation. minimum stay of 9 days at one of two specialized wards (on- cology, palliative care) at the comprehensive cancer center of a tertiary care university hospital (Department of Internal Results Medicine I, University Hospital of Vienna) were included. Patients with hematological cancer (leukemia, multiple mye- The data of 100 patients were evaluated. Patient characteris- loma, lymphoma) were excluded because the course of these tics on day −9 (9 days before the patient’s death) are shown in Support Care Cancer

Table 1. Sociodemographic and disease-related parameters re- each 4.8 % pulmonary and 4.8 % gastrointestinal, 4.2 % he- vealed no significant difference between patients at the oncol- matological, and 13.8 % other problems. The majority of the ogy ward and those at the palliative care ward. In general, the conditions required medical therapy. No significant difference distribution of cancer types did not differ significantly be- was noted between the wards in this regard. tween the wards, although some entities (such as breast can- Table 2 shows the numbers of patients treated with a drug cer) were more common at the palliative care ward while of a specific subgroup. Data are shown separately for the others (such as gastrointestinal cancer) were more common investigated time points and locations. Approximately 80 % at the oncology ward. The patients’ ECOG performance status of terminally ill cancer patients received opioids 1 week be- rose significantly from day −9 to day 0, indicating the deteri- fore their death; the number rose to 92 % on the day of death. oration of their condition on the day of death (P<0.0001, Significantly more patients treated by a palliative care team Kruskal-Wallis test). The patients had a variety of comorbid- had a prescription of opioids. Non-opioid analgesics were ities: 168 different conditions were recorded. Of these, 41.6 % prescribed to two thirds of the patients, but were replaced by were cardiovascular (including , atrial fibrilla- opioids closer to the day of death. More than 70 % of the tion), 15.7 % metabolic (including and hormonal dis- patients had prescribed psychoactive drugs, significantly more orders), 8.4 % neurologic or psychiatric, 7.8 % infections, patients at the palliative care ward compared to those at the

Table 1 Patient characteristics Total Oncology ward Palliative care ward Statistical significance between wards

Number of patients 100 50 50 Age (years) 63.4 64.7 61.8 n.s.a (52.7–71.2) (54.8–71.8) (50.0–71.0) Gender (female/male) (%) 56 %/44 % 50 %/50 % 62 %/38 % n.s.b Days from admission to death 18.0 16.5 21.0 n.s.a (12.3–27.8) (12.0–26.3) (13.0–30.0) ECOG day −93.03.0 3.3 n.s.b (3.0–4.0) (2.75–4.0) (3.0–4.0) ECOG day −64.04.0 4.0 n.s.b (3.0–4.0) (3.0–4.0) (3.0–4.0) ECOG day −34.04.0 4.0 n.s.b (4.0–4.0) (3.0–4.0) (4.0–4.0) ECOG day 0 5.0 5.0 5.0 n.s.b (5.0–5.0) (5.0–5.0) (5.0–5.0) PEG (yes/no) 6 %/94 %c 6%/94% 4%/96%c n.s.b Central venous line (yes/no) 58 %/37 %d 64 %/26 %d 52 %/48 % 0.01b Type of cancer: Gastrointestinale 30 % 36 % 24 % n.s.b Lung 20 % 14 % 26 % n.s.b Urogenitalf 15 % 14 % 16 % n.s.b Breast 14 % 12 % 16 % n.s.b Head and neck 5 % 6 % 4 % n.s.b Sarcoma 9 % 8 % 10 % n.s.b Skin 3 % 2 % 4 % n.s.b Unknown origin 4 % 8 % 0 0.04b

Data are presented as medians and interquartile ranges PEG percutaneous gastrostomy, ECOG Eastern Cooperative Oncology Group Performance Score a Mann–Whitney U test b Chi-square test c Missing data for PEG in one patient d Missing data for central venous line in five patients e The following were included: esophageal and gastric cancer, colon and rectal cancer, pancreatic neoplasms, and cholangiocarcinoma f The following were included: prostatic cancer, testicular cancer, ovarian cancer, uterine cancer, and renal cancer Support Care Cancer

Table 2 Proportion of patients receiving drugs of various Total Oncology ward Palliative care ward Statistical significance categories (n=100) (n=50) (n=50) between wards (Chi-square)

Opioids Day −9 78 % 62 % 94 % <0.0001 Day −6 80 % 66 % 94 % 0.001 Day −3 89 % 80 % 98 % 0.008 Day 0 92 % 88 % 96 % ns Non-opioid analgesics Day −966%62% 70% ns Day −668%68% 68% ns Day −361%60% 62% ns Day 0 39 % 30 % 48 % ns Psychoactive Day −9 74 % 60 % 88 % 0.003 Day −677%70% 84% ns Day −377%74% 80% ns Day 0 60 % 44 % 76 % 0.002 Anti-microbial Day −956%66% 46% ns Day −651%60% 42% ns Day −347%54% 40% ns Day 0 31 % 36 % 26 % ns Anti-thrombotic Day −982%76% 88% ns Day −683%78% 88% ns Day −376%72% 80% ns Day 0 55 % 52 % 58 % ns Anti-tumor Day −99%14% 4% ns Day −6 10 % 16 % 4 % 0.05 Day −38%14% 2% 0.03 Day0 2% 2% 2% ns Cardiovascular Day −954%54% 54% ns Day −653%54% 52% ns Day −349%54% 44% ns Day 0 23 % 26 % 20 % ns Fluids Day −9 57 % 68 % 46 % 0.05 Day −6 64 % 78 % 50 % 0.006 Day −3 75 % 86 % 64 % 0.02 Day 0 65 % 74 % 56 % ns Others Day −9 98 % 96 % 100 % ns Day −6 99 % 100 % 98 % ns Day −394%96% 92% ns Day 0 73 % 70 % 76 % ns

Psychoactive drugs include antidepressants, anti-psychotic drugs, and benzodiazepines; anti-microbial drugs include antibiotics, anti-fungal, and anti-viral therapy; cardiovascular medication includes anti-hypertensive and diuretic medication; fluids include crystalloids, colloids, parenteral nutrition, and blood products; others include anti-epileptics, antacids, laxatives, corticosteroids, and anti-emetic agents. A detailed description of the medication categories can be found in Appendix A Support Care Cancer oncological ward. The palliative care team administered fluids Wallis test) dropped significantly towards the day of death, less frequently. More patients at the dedicated oncology ward these figures clearly indicate an impressive drug burden short- were on anti-cancer therapy up to 3 days before death (16 vs. ly before death. On the day of death, one half of the patients 4%;P =0.05). A variety of other medications (see were still taking more than 6.5 medications per day, and a Appendix A) were prescribed to almost all patients up to quarter of them between 9 and 17 medications per day (IQR 3 days before death. 4–9). These numbers revealed no significant difference be- The majority of the prescribed drugs were classified as tween patients treated at the general oncology ward and those chronic medication. Patients received a median number of treated at the dedicated palliative care ward. nine (IQR 6–12; range 0–22) drugs as chronic medication Figure 2 shows the median number of medications in relation per day, but were given only one drug (median; IQR 1–2; to the day of assessment and the treatment ward (medians and range 0–7) for acute conditions; P<0.0001 (Mann–Whitney IQR). The median number of medications per patient remained U test). A reliable statistical evaluation could not be performed constant from day −9today−3, about 11 drugs per patient, and for acute conditions because of the small number of medica- dropped significantly to about a half of this value during 3 days tions used in this setting. Therefore, further analysis was fo- before death (P<0.0001, Kruskal-Wallis test). Patients at the pal- cused on the total numbers of prescribed drugs. liative care ward had significantly more prescribed medications on The distribution of medications at each time point is shown day −6 only (median 10 vs. 12 drugs; P=0.05, Mann–Whitney U in Fig. 1. Obviously, the criteria of polypharmacy were ful- test); the difference on the other remaining days was just one drug filled in a considerable number of patients. Nine days before and was not statistically significant. death, a median number of 11 medications (IQR 9–13) were While there was no difference in the total number of med- prescribed. Thus, one half of the patients were taking more ications between the palliative care and the oncology ward, than 11 medications per day, and a quarter of them between 13 but significant differences were noted between various drug and 18 medications. Although the number of patients on categories (Fig. 3). The large part of the patients’ daily med- polypharmacy (P=0.0001; Chi-square test) as well as the me- ication consisted of opioids, psychoactive drugs, and fluids. dian number of prescribed medications (P=0.0001; Kruskal- This proportion increased significantly, and the use of other

Fig. 1 Distributions of the number of different medications per patient. Legend: Histogram of the numbers of different medications. Shaded bars, oncology ward; solid bars, palliative care ward. The dashed line indicates the limit of polypharmacy (>4 medications per day). x-axis, number of medications; y-axis, number of patients. Inserts, number of patients with <5 (left)or>4(right) medications at the respective wards. Chi-square test Support Care Cancer

medications declined, as the patients approached their death. Psychoactive drugs, especially benzodiazepines, were used significantly more often at the palliative care ward (17 vs. 10 % on the day of death). Anti-cancer agents accounted for a maximum of 1.6 % of the patient’s daily medication at the oncology ward and only 0.5 % at the palliative care ward. Multivariate analysis was performed to identify factors influencing the number of prescribed drugs (Table 3). The ECOG performance status proved to be the strongest predictor of polypharmacy: the poorer the patient’s condition, the fewer medications were prescribed (Spearman’s rank correlation, R=−0.35; P<0.0001). Additionally, the time before the pa- tient’s death had a significant impact on the number of pre- scribed drugs. The patient’s age and the location of treatment were less, yet still statistically significant. Patients at the pal- liative care ward received slightly, but significantly more med- Fig. 2 Number of medications with reference to the wards. Legend: rectangles indicate the median, vertical lines the interquartile range. ications than did patients at the general oncology ward. No Open rectangles/dashed line, oncology ward; filled rectangles/solid line, significant association was registered between polypharmacy palliative care ward; x-axis, observation day; y-axis, number of and gender, the total duration of hospital stay, the insertion of a – medications. Difference between wards: *P=0.05 (Mann Whitney U central venous catheter, or a percutaneous gastrostomy tube to test); differences over time, P<0.0001 (Kruskal-Wallis test) facilitate the administration of medication.

Discussion

To our knowledge, this is the first detailed analysis of the quantity, composition, and course of medical therapy in ter- minally ill patients with advanced cancer. The major findings of the study are detailed below. First, the investigation clearly shows that nearly all patients with advanced cancer receive polypharmacy at the end of their lives. One week before death, 96 % of the patients were re- ceiving five or more drugs per day. Although this number fell as the patients approached their death, as many as 61 % had a prescription of more than five medications on the final day, and a quarter of them more than nine (9–17)drugs.Thisisin line with the published literature, as similar numbers have

Table 3 Multivariate analysis of factors Parameter P value influencing the numbers of drugs ECOG score <0.0001 Location (ward) 0.01 Number of days before death 0.02 Age 0.05 Gender n.s. Duration of admission n.s. Central venous line n.s. Gastrotomy n.s. Fig. 3 Means of drug subclasses relative to total quantities of medication. Shaded bars, oncology ward; solid bars, palliative care ward. x-axis, n=400; dependent variable, total numbers number of days before death; y-axis, percentage of total medications of drugs Support Care Cancer been reported by several authors for cancer patients at the end palliative care centers, not closely associated with a hospital or of their lives [22–25]. an oncology team, may be different in this regard. The present The second notable feature of the present investigation was study was conducted in a retrospective setting. The actual that it revealed no significant difference in the total number of quantities of medication may have been overestimated be- prescribed drugs for cancer patients at a general oncology cause patients experiencing a variety unexpected complica- ward and a dedicated palliative care ward. At the time of tions and needing additional therapy were included. death, patient with advanced cancer had, on average, 6.5 med- However, we collected real-life data representing the actual ications at a specialized oncology ward and 7.3 medications at clinical situation. A study design such as ours would not be a dedicated palliative care ward. This was somewhat surpris- feasible in a prospective setting, as it would not be possible to ing because one generally assumes that the focus of treatment predict the day of a patient’s death and prospectively collect at a dedicated palliative care ward is to improve the patients’ data on predefined days before death. Moreover, a prospective quality of life rather than employ curative measures, and the design may have influenced the prescription behavior of the prescription strategy of a palliative care team would differ teams. Therefore, the retrospective design appears to be one of from that of oncologists. the greatest advantages of the study. Although there are some differences between the two teams, A further limitation is that the results provide data on the our study clearly shows that the principles of palliative care, total numbers of drugs, but the consequences of polypharmacy, such as relieving pain or anxiety, are fulfilled in the large ma- such as drug-drug interactions or the indications of drugs, were jority of patients during the last week of their lives: nearly 90 % not studied. The study was not powered for an analysis of this of the patients were receiving opioids, psychoactive drugs, an- nature, which would have required many more patients in view algesics, or sedation. Patients treated at a dedicated palliative of the heterogeneity of the medications. care unit received more sedation at the end of their lives, but the In conclusion, the present investigation showed that other medications did not differ significantly from those of polypharmacy is used in the large majority of patients with patients undergoing treatment at a specialized oncology ward. advanced cancer, even in dedicated palliative care wards. One reason for the similarity in the two settings may be the fact Future projects should be focused on drug-drug interactions that oncologists are well versed with the methods of palliative and the impact of polypharmacy on the patient’squalityof care, and the palliative care team was consulted for the treat- life. Another approach would be to analyze and score the ment of patients without curative options. indications for the different therapies and thus establish Similar observations were made by Currow et al. [26], who whether polypharmacy is actually necessary in palliative care reported that palliative care patients participating in a regional or is an outcome of the subjective estimation of treatment palliative care program in South Australia received, on average, teams. Finally, more controlled trials with structured programs 6.3 drugs on the day of death. In contrast to our data, Currow will be needed to reduce the quantity of medication and gen- et al. registered an increase in the number of medications to- erate scientific evidence on the discontinuation of medication wards death. In our study, the mean number of drugs declined in terminally ill cancer patients. from 10.5 to 6.5 (oncology ward) and from 11.5 to 7.3 (palli- ative care ward). Our results may reflect the beneficial effects of Acknowledgments We gratefully acknowledge the dedicated work and drug discontinuation in this population (for a review, see [13]). cooperation of the staff at the oncology and palliative care wards of the In general, the ECOG performance status had the greatest Department of Medicine I, Medical University of Vienna. impact on the number of drugs: patients in poor clinical con- dition or with terminal disease had significantly fewer medi- Compliance with ethical standards cations. Other minor differences were observed: anti-cancer Conflict of interests None therapy was more frequently used at the oncology ward and rarely by the palliative care team. This can be explained by the different treatment objectives. The purpose of admission to a dedicated oncology ward is to treat the patient’sunderlying disease. Surprisingly, medical devices enabling convenient References drug application, such as central venous lines (port systems or indwelling catheters) or PEG, had no significant impact on 1. Kaufman DW, Kelly JP, Rosenberg L et al (2002) Recent patterns the number of medications. of medication use in the ambulatory adult population of the United The limitations of the present study are worthy of note. It States: the Slone survey. JAMA 287:337–344 was a single-center investigation with experienced and closely 2. Merlis S, Sheppard C, Collins L, Fiorentino D (1970) Polypharmacy in psychiatry: patterns of differential treatment. interacting teams, active standard operating procedures, and Am J Psychiatry 126:1647–1651 quality control; this might explain the absence of major dif- 3. Cullen SC (1957) Polypharmacy in anesthesia. Surg Gynecol ferences between the wards. In contrast, stand-alone teams in Obstet 104:113–114 Support Care Cancer

4. Grant-Whyte H (1961) Irrational polypharmacy in anaesthesia. S 16. Popa MA, Wallace KJ, Brunello A et al (2014) Potential drug in- Afr Med J 35:1016–1021 teractions and chemotoxicity in older patients with cancer receiving 5. (1966) Patients and polypharmacy. Can Med Assoc J 94:456. chemotherapy. J Geriatr Oncol 5:307–314. doi:10.1016/j.jgo.2014. 6. Veehof L, Stewart R, Haaijer-Ruskamp F, Jong BM (2000) The 04.002 development of polypharmacy. A longitudinal study. Fam Pract 17. VanLeeuwen RWF, Swart EL, Boom FA et al (2010) Potential drug 17:261–267 interactions and duplicate prescriptions among ambulatory cancer 7. Gnjidic D, Hilmer SN, Blyth FM et al (2012) Polypharmacy cutoff patients: a prevalence study using an advanced screening method. and outcomes: five or more medicines were used to identify BMC Cancer 10:679. doi:10.1186/1471-2407-10-679 community-dwelling older men at risk of different adverse out- 18. Garfinkel D, Zur-Gil S, Ben-Israel J (2007) The war against comes. J Clin Epidemiol 65:989–995. doi:10.1016/j.jclinepi.2012. polypharmacy: a new cost-effective geriatric-palliative approach 02.018 for improving drug therapy in disabled elderly people. Isr Med – 8. Jyrkkä J, Vartiainen L, Hartikainen S et al (2006) Increasing use of Assoc J 9:430 434 medicines in elderly persons: a five-year follow-up of the Kuopio 19. Lees J, Chan A (2011) Polypharmacy in elderly patients with can- – 75+Study. Eur J Clin Pharmacol 62:151–158. doi:10.1007/s00228- cer: clinical implications and management. Lancet Oncol 12:1249 005-0079-6 1257. doi:10.1016/S1470-2045(11)70040-7 20. Riechelmann RP, Krzyzanowska MK, O’Carroll A, Zimmermann 9. Bushardt RL, Massey EB, Simpson TWet al (2008) Polypharmacy: C (2007) Symptom and medication profiles among cancer patients misleading, but manageable. Clin Interv Aging 3:383–389 attending a palliative care clinic. Support Care Cancer 15:1407– 10. Jyrkkä J, Enlund H, Lavikainen P et al (2011) Association of 1412. doi:10.1007/s00520-007-0253-8 polypharmacy with nutritional status, functional ability and cogni- 21. Frechen S, Zoeller A, Ruberg K et al (2012) Drug interactions in tive capacity over a three-year period in an elderly population. dying patients: a retrospective analysis of hospice inpatients in Pharmacoepidemiol Drug Saf 20:514–522. doi:10.1002/pds.2116 Germany. Drug Saf 35:745–758. doi:10.2165/11631280- 11. Hajjar ER, Hanlon JT, Sloane RJ et al (2005) Unnecessary drug use 000000000-00000 in frail older people at hospital discharge. J Am Geriatr Soc 53: 22. Nauck F, Ostgathe C, Klaschik E et al (2004) Drugs in palliative – 1518 1523. doi:10.1111/j.1532-5415.2005.53523.x care: results from a representative survey in Germany. Palliat Med 12. Maggiore RJ, Gross CP, Hurria A (2010) Polypharmacy in older 18:1007–10 – adults with cancer. Oncologist 15:507 522. doi:10.1634/ 23. Lundy E, McMullan D, McShane P et al (2013) Polypharmacy and theoncologist.2009-0290 drug omissions across hospices in Northern Ireland. J Palliat Med 13. LeBlanc TW, McNeil MJ, Kamal AH et al (2015) Polypharmacy in 16:1446–1449. doi:10.1089/jpm.2013.0090 patients with advanced cancer and the role of medication discontin- 24. McLean S, Sheehy-Skeffington B, O’Leary N, O’Gorman A (2013) uation. Lancet Oncol 16:e333–e341. doi:10.1016/S1470-2045(15) Pharmacological management of co-morbid conditions at the end of 00080-7 life: is less more? Ir J Med Sci 182:107–112. doi:10.1007/s11845- 14. Goldberg RM, Mabee J, Chan L, Wong S (1996) Drug-drug and 012-0841-6 drug-disease interactions in the ED: analysis of a high-risk popula- 25. Sera L, McPherson ML, Holmes HM (2014) Commonly prescribed tion. Am J Emerg Med 14:447–450. doi:10.1016/S0735-6757(96) medications in a population of hospice patients. Am J Hosp Palliat 90147-3 Care 31:126–131. doi:10.1177/1049909113476132 15. Tamura BK, Bell CL, Inaba M, Masaki KH (2012) Outcomes of 26. Currow DC, Stevenson JP, Abernethy AP et al (2007) Prescribing polypharmacy in nursing home residents. Clin Geriatr Med 28:217– in palliative care as death approaches. J Am Geriatr Soc 55:590– 236. doi:10.1016/j.cger.2012.01.005 595. doi:10.1111/j.1532-5415.2007.01124.x