Table 2.

Papers selected for data analysis and synthesis.

Authors/Year	Country	Objective(s)	Method	Focus of Data Collection	Setting and Sample	Demographic Data
Snowdon et al. 2011 [37]	Australia	To understand and compare the patterns of psychotropic medications use in nursing homes.	A cross-sectional survey	Comparison of the patterns of psychotropic medication, using data from surveys from 1993, 1998, 2003, and 2009. Details of current prescription of medications during the last 14 days or since initiation of regular or PRN medication were recorded. Surveys reported the rate of PRN prescription and administration in the last 28 days. The administration of PRN medications at least 25 out of 28 days was considered regular medication use.	Number of nursing homes and percentage of their participation in the study were as follows: 46 (98%), 38 (97%), 51 (100%), and 44 (92%), in years 1993, 1998, 2003, and 2009, respectively.	Mean number of residents: 52.5, 52.0, 60.6, and 56.0, respectively. Percentages of 36% (n = 895, mean 78.7 years, SD = 12.1) and 64% (n = 1570, mean 84.2 years, SD = 9.6) were male and female, respectively. No information of their diagnosis was available.
Carder, 2012 [38]	USA	To identify if staff who were not registered nurses administered PRN medications to residents with dementia.	The qualitative analysis of interviews and medication record reviews	How is a decision on administering PRN medications to residents with dementia made? PRN administration staff were observed for 6 consecutive days, 16 h per day, including 72 scheduled medication rounds. Sixteen interviews were completed.	Sixteen unlicensed staff members supervised by registered nurses working at three assisted living settings with all 47 residents in urban and suburban areas.	Residents were mostly female. The mean age of staff was 38 years (SD = 8.3 years) and were mainly female (75%). Their job experience ranged 4 months to 8 years.
Rønningen et al., 2013 [39]	Norway	To document prescribing and administration of PRN medication in one nursing home, and to investigate how often, and how, the positive and negative effects of prescribed PRN medication are documented.	A cross-sectional survey of documentation of PRN medications	PRN prescription and administration prevalence. Description of documentation of PRN medication effects (beneficial and harmful) were documented in terms of quantity and quality.	PRN medicines management was documented and described for 108 patients in one nursing home over 15 weeks.	Mean age of residents was 84.5 years (SD = 9.4 years). They were mostly female (60.2%). No data on staff were provided.
Voyer et al., 2015 [40]	Canada	To identify behavioural and psychological symptoms in dementia that were associated with PRN antipsychotic medicine prescription and administration.	A secondary analysis on a prospective, observational, multisite cohort	Association between behavioural and psychological symptoms of dementia with PRN antipsychotic medicine use. Medication records of regular and PRN use of medication were reviewed. Atypical antipsychotic agents (risperidone (Risperdal®), olanzapine (Zyprexa®), and quetiapine (Seroquel®), and typical antipsychotic agents (haloperidol (Haldol®) were used. Prescription and administration of PRN antipsychotic agents during 7 days before the monthly assessment of behavioural and psychological symptoms of dementia assessments were considered.	A total of 146 nursing home residents from 7 settings. Subjects aged ≥65 years were included. Those without dementia, cognitive impairment, or behavioural and psychological symptoms were excluded.	Participants had a mean age of 85.6 years (SD = 7 years) and were mostly female (58.9%). They mainly were diagnosed with dementia (89.7%). Of 129 nurses, 76.7% were registered nurses and 90.7% were female. Their experience in geriatric wards was mainly greater than 10 years (72%).
Dörks et al., 2016 [41]	Germany	To examine the characteristics and potential predictors of PRN prescription and administration in nursing homes.	A multicentre survey	Investigation of characteristics and potential predictors of PRN medicine prescription and administration in nursing homes.	A total of 852 residents in 21 nursing homes organised by different institutions.	Their mean age was 83.5 years (SD = 10.5) and 76.5% were female. Their mean length of stay was 3.2 years (SD = 3.4).
Allers et al., 2017 [42]	Germany	To compare the use of antipsychotic medications in residents with and without dementia.	A cross-sectional survey	Assessment of the prevalence of antipsychotic medicine use and exploration of factors affecting their prescription. While data on all prescribed medications were collected from the residents’ medication schedules, antipsychotics prescribed as scheduled medication and on a PRN basis were studied and prescriptions of scheduled and PRN medication were compared.	All residents (n = 852) from 21 nursing homes without any exclusion criteria.	A percentage of 57.7% of the residents were diagnosed with dementia and their mean age was higher than those without it (84.9 vs. 81.4 years, but no standard deviations were presented). About three quarters of the residents were female and those with dementia were more commonly severely care-dependent (32.8 vs. 16.4%).
Westbury et al., 2018 [43]	Australia	To study the impact of an interdisciplinary intervention on the prescription of antipsychotics and benzodiazepines in older people’s residential care facilities.	A longitudinal study with comparisons over time	Investigation of the impact of an educational consultation intervention on the use of antipsychotics and benzodiazepines over 6 months. It assessed the possibility of substitution of medications prescribed regularly. The multi-strategic programme comprised: auditing psychotropic medication, staff education, and case reviews by the physician, pharmacist, and nurse at the beginning of the programme and at 3 and 6 months.	A national-level sample consisting of 150 older people’s residential care facilities hosting 12,157 people.	The residents’ mean age was 85.8 years (SD = 8.6).
Westbury et al., 2019 [44]	Australia	To analyse the use of psychotropics in a national sample of residential aged care facilities.	A retrospective cohort	Analysis of psychotropic use in a large national sample of residential aged care facility residents, derived from a project to promote the appropriate use of antipsychotics and benzodiazepines. A multi-strategic interdisciplinary intervention was devised consisting of a 6-month programme with cycles of audits, education, and a review of sedatives. Prescribing data were extracted via a custom-made website.	A large national-level sample of 150 residential aged care facilities consisting of 11,368 residents	Data on psychotropic prescribing was collected from for 139 of 150 facilities with a response rate of 93%. Clinical, diagnostic, and demographic data were not reported.

PRN: pro re nata; SD: standard deviation.