Methods
Study design and setting: RESORT (REStOring health of acutely unwell adulTs) is an observational longitudinal study cohort of patients admitted to the Royal Melbourne Hospital (Victoria, Australia) for geriatric rehabilitation (recruitment commenced October 16, 2017, and ended March 18, 2020), comprising a total of 1890 older adults. A description of the cohort, methods, and design in the RESORT study is available elsewhere.21,22 Briefly, patients on the geriatric rehabilitation wards of the Royal Melbourne Hospital, Melbourne, Australia, completed a standardized comprehensive geriatric assessment at both admission and discharge by a multidisciplinary team inclusive of physicians, nurses, physiotherapists, and occupational therapists, and followed up at 3 months post-discharge via phone interviews. Written informed consent was obtained by all patients or a nominated proxy. Patients were excluded if they were receiving palliative care at admission, were transferred to acute care prior to consenting to the study or lacked both the capacity to provide informed consent and a nominated proxy. The study was approved by the Melbourne Health Human Research Ethics Committee (HREC/17/MH/103).
Data sources and collection: Age, sex, and length of stay were obtained from medical records. Other demographic characteristics (e.g., current living situation, living status, primary language, walking ability) were obtained from a survey completed by the patient and/or carer. The primary reason for hospital admission was extracted from medical records and grouped into musculoskeletal, neurological, cardiorespiratory, infection and other (including gastrological, urology, hematology, ophthalmology, vascular, psychiatry related conditions, cancer, and metabolic disorders) by trained researchers. Physicians assessed morbidity using the Charlson Comorbidity Index (CCI)23 and the Cumulative Illness Rating Scale (CIRS).24 CCI and CIRS scores ranged from 0-37 and 0-56, respectively. Frailty was assessed by physicians using the Clinical Frailty Scale (CFS),25 ranging from 1-9, with greater scores indicating a higher level of frailty. Cognitive impairment assessment was completed by physicians and was defined as a dementia diagnosis captured by the CCI, CIRS, or medical records, or a score on the Standardized Mini-Mental State Examination of < 24 points,26 Montreal Cognitive Assessment score < 26 points,27 or the Rowland Universal Dementia Scale < 23 points.28 Functional independence status was examined by occupational therapists using the Katz Activities of Daily Living (ADL) 29 and Lawton and Brody Instrumental Activities of Daily Living (IADL)30 on admission and at discharge from geriatric rehabilitation. Scores of ADLs and IADL ranged between 0–6 and 0–8 points, respectively, with higher scores indicating higher level of independence. ADL and IADL scores at three months post-discharge were assessed by trained researchers via telephone interview with the patient and/or carer. All types of medications (chronic, acute, over the counter) that patients were taking across the care transitions (acute admission, admission to and discharge from geriatric rehabilitation wards) were extracted from medical records.
Potentially inappropriate medications: PIMs and PPOs were quantified according to version 2 of the Screening Tool of Older People’s Prescription (STOPP) and the Screening Tool to Alert Doctors to the Right Treatment (START) criteria, respectively.6These explicit screening tools, developed using expert opinion using available evidence from literature and clinical guidelines, consisted of 80 STOPP and 34 START items to identify unnecessary medications and omitted medications, respectively. PIMs/PPOs were identified across the three care transitions (acute admission, on admission to and at discharge from geriatric rehabilitation ward) during geriatric rehabilitation care. The data were extracted by a qualified pharmacist, with regular independent checks by another qualified pharmacist.
Outcome variables: The primary outcomes of interest were hospital readmissions, and mortality after hospital discharge. Readmission was defined as an unplanned acute readmission after discharge and was obtained from medical records and at 3-month phone interviews. Mortality was defined as death of patient at 3- or 12-months post-discharge and was obtained from the Registry of Births, Deaths and Marriages Victoria and medical records. The secondary outcome was functional status at 3-months post discharge from geriatric rehabilitation.
Data analysis: Descriptive statistics was undertaken to summarize the main characteristics of older patients. Continuous variables were reported as mean with corresponding standard deviation, or median with interquartile ranges if not normally distributed. Categorical variables were reported as absolute frequencies and percentages. The prevalence of PIPs (PIM/PPO) was calculated as a ratio between the number of patients exposed to at least 1 PIP (PIM/PPO) and the total number of patients included. To investigate the associations between PIPs identified at discharge and subsequent outcomes (e.g., hospital readmission, mortality, functional outcomes), PIPs were measured as a dichotomous variable (yes/no), as well as a continuous variable (number of PIPs). Only patients discharged alive were considered in the estimations related to analyses involving associations between PIPs and post-discharge health outcomes. For comparison between PIP and outcome variables, Chi-square test were employed for categorical/dichotomous variables, and the independent t-test was used for continuous variables. The McNemar and Cochran’s Q tests were used to compare the prevalence of PIM/PPO between two and three care transitions, respectively. Multivariate logistic regression analyses were employed to identify the independent associations between discharge PIMs/PPOs, and number of discharge medications with primary health outcomes. Multivariate regression analyses were further applied on PIM categories, and it was undertaken by merging similar PIMs together (e.g., Groups B and C, Groups D and K), Group A (indication) and fewer PIMs (E-J, L, M). Similarly, PPOs were categorised as Group A (cardiovascular medications), Group I (Vaccines) and other groups (B-D, F-H) and their associations were assessed against the primary outcomes. Linear regression analyses were used to determine the associations between discharge PIMs/PPOs and secondary outcomes (i.e., ADL and IADL). All multivariate analyses were adjusted for age, Charlson comorbidity index (CCI), cognitive impairment, length of stay during geriatric rehabilitation and number of discharge medications.