Table 2.

Summary of eight studies included in the review which use a system to identify potentially inappropriate prescribing in advanced dementia or dementia in palliative care

First Author, Year	Country	Aim(s)	Design	N at baseline	Setting	Approach to identify inappropriate medications	Medication variables	Results
Studies which use number of medications as indication of potentially inappropriate prescribing i.e. polypharmacy
Blass et al. 2008 [15]	USA (Baltimore)	Identify how medication usage changed over time as resident with advanced dementia moves toward death, and identify correlates of increased medication usage.	Prospective cohort study (longitudinal)	125 residents	3 nursing homes	Number of medications prescribed i.e. polypharmacy.	Number of medications prescribed (regular + prn) at baseline, and factors associated with total number of medications.	Residents prescribed 14.6 medications each. Increase in palliative medicines i.e. opiates and a decrease in antibiotics, anti-dementia agents, cardiovascular agents and psychotropics as death approaches. No change in the number of medications given over time.
Studies using explicit criteria to identify potentially inappropriate prescribing
Holmes et al. 2008 [16]	USA	Evaluate the feasibility of developing consensus recommendations for appropriate prescribing for patients with advanced dementia.	Modified Delphi consensus panel (and medication record audit) (cross-sectional)	34 patients	3 long term care facilities	Using modified Delphi process (12 geriatricians), medications categorised for use in palliative care patients with advanced dementia; never, rarely, sometimes and always appropriate.	Determine frequency of inappropriate medication prescribing, using in-house developed explicit criteria.	Patients taking 6.5 medications each. Six patients taking ten or more medications daily. 29 % of patients taking a medication considered never appropriate.
Tjia et al. 2010 [17]	USA (Chicago)	Describe the pattern and factors associated with daily medication use in nursing home residents with advanced dementia.	Prospective cohort study (longitudinal)	323 residents	22 nursing homes	Drugs of questionable benefit i.e. ‘never appropriate’ according to medications classified by Holmes et al. 2008.	Resident characteristics associated with the use of daily medications and drugs deemed inappropriate.	Male, shorter length of stay at facility (<1 year), higher functional ability and diabetes independently associated with inappropriate drug use. Having a DNR order independently associated with a lower likelihood of inappropriate drug use.
Colloca et al. 2012 [20]	7 EU countries (Czech Republic, England, Finland, France, Germany, Italy, The Netherlands) and Israel	Identify prevalence and factors associated with use of inappropriate drugs in older adult patients with severe cognitive impairment.	Medication chart audit (cross-sectional)	1449 residents	57 nursing homes	The use of drugs classified as rarely or never appropriate by criteria developed by Holmes et al. 2008.	Inappropriate drug use defined as rarely or never appropriate in patients with severe cognitive impairment based on the Holmes criteria published in 2008.	Inappropriate drug use in 643 (44.9 %) of residents. Most commonly prescribed inappropriate drugs were lipid-lowering agents (9.9 %), antiplatelet agents (9.9 %), Ach inhibitors (7.2 %) and antispasmodics (6.9 %). Inappropriate drug use associated with diabetes, HF, stroke, recent hospitalization. An inverse relationship between inappropriate drug use and geriatrician at facility.
Toscani et al. 2013 [18]	Italy	Assess and compare treatments and prescriptions of patients with advanced dementia cared for in nursing homes and in home care and assess their appropriateness from a palliative care perspective.	Baseline data from multicentre prospective observational cohort study	245 residents	Nursing homes	Used criteria developed by Holmes et al. 2008.	The appropriateness of each prescription assessed according to the Holmes et al. 2008 classification.	Patients received 4.1 medications on average (range 0–13). Laxatives, antipsychotics, and anxiolytics were the most frequently prescribed in the nursing homes. 8.1 % of residents receiving at least one analgesic.
Tjia et al. 2014 [22]	USA	Estimate the prevalence of medications with questionable benefit used by nursing home residents with advanced dementia.	Medication record audit (cross-sectional)	5406 residents	Nursing homes	Medications deemed never appropriate for use in advanced dementia according to criteria developed by Holmes et al. 2008.	Use of medication of questionable benefit in advanced dementia based on previously published criteria and mean 90-day expenditures due to these medications per resident.	53.9 % of residents receiving at least one medication with questionable benefit. Anticholinesterase inhibitors (36.4 %), memantine (25.2 %) and lipid-lowering agents (22.4 %) most commonly prescribed medications with questionable benefit.
Other approaches to identify inappropriate prescribing
Shega et al. 2009 [19]	USA	Describe hospice medical directors practice patterns and experiences in the use and discontinuation of anticholinesterase inhibitors and memantine in hospice patients with dementia.	Mail survey (cross-sectional)	152 hospital medical directors	Hospice care	N/A	Associations between the likelihood of survey response and participant characteristics. Comparisons analysing whether or not a physician would recommend medication discontinuation based upon reported clinical benefit of anticholinesterase inhibitors and memantine use.	Of the respondents, 75 % and 33 % reported that at least 20 % of patients were taking anticholinesterase inhibitor or memantine at hospice admission. 80 % of respondents would recommend discontinuation of these agents, however, a subset believe they stabilize cognition (22 %), decrease challenging behaviours (28 %), maintain patient function (22 %,) reduce caregiver burden (20 %) and improve caregiver quality of life (20 %).
Parsons et al. 2014 [21]	NI (Northern Ireland), RoI (Republic of Ireland)	Evaluate the extent to which patient-related factors and physicians’ country of practice influenced decision making regarding medication use in patients with end-stage dementia.	Factorial survey design	662 health professionals	Community, nursing home, hospital	Medications selected due to contradictory evidence available to guide practice or because they have been identified in the limited literature as potentially inappropriate for individuals with advanced dementia: antibiotics, anticholinesterase inhibitors, memantine, lipid-lowering agents and antipsychotics.	Assess physician decision making regarding withholding or continuation/discontinuation of key medications in patients with end-stage dementia.	Considerable variability found regarding initiating/withholding antibiotics and continuing/discontinuing anticholinesterase inhibitors and memantine hydrochloride. Less variability found in decision making regarding lipid-lowering agents and antipsychotics. Patient place of residence and physician country of practice had the strongest and most consistent effects on decision making.