Paramedics assessing elders at risk for independence loss
While the majority of the elderly are independent, frail older people are a sub-group at increased risk for preventable events such as falls, delirium or functional decline: 20% of people over 65 and 44% of those over 85 lack the support they need to function daily. A minor change can trigger a chain of events leading to adverse outcomes. For example, being trapped on the floor after a fall, developing skin breakdown and becoming delirious may ultimately require admission to a nursing home. Interventions that prevent adverse outcomes exist, but they are effective only when they are appropriately targeted to high-risk groups. Older adults are the highest users of ambulance or emergency medical services (EMS), and are five times more likely than younger adults to need an ambulance. Because of their social isolation and lack of support and timely access to primary care, frail older people frequently depend on EMS to function as a safety net and the proportion of non-urgent calls paradoxically increases with age. Consequently, paramedics are uniquely positioned to observe older persons in their homes and identify those at risk for adverse outcomes.
This proposal builds on the IAP # 73370 pilot study in which we developed and field-tested a standardized paramedic risk assessment tool. This structured checklist, known as a clinical prediction tool, uses predictive variables chosen from systematic literature reviews, an expert panel and paramedic focus groups. We developed an educational package including a professionally produced video of a simulated patient-encounter and trained 855 paramedics in study procedures. We tested the inter-rater agreement of items on the data collection tool in both simulated and actual patient encounters and found “substantial” to “almost perfect” agreement (kappa = 0.63 to 0.98). Pilot data on 210 patients demonstrate the feasibility of paramedic data collection, establish the patient recruitment rate and allow simplification of the data collection tool. Univariate analysis also established that paramedic observations do predict subsequent adverse outcomes among their older clients.
The goal of the current study is to complete the empiric derivation of the PERIL clinical prediction tool, using paramedic’s observations in the homes of older people to identify those at high risk for adverse outcomes after an EMS encounter.
We will conduct a population-based prospective observational study over three years in three urban EMS systems (Edmonton, Ottawa and Toronto). Persons aged 65 years or older who live independently and are not critically ill will be included. Paramedics have agreed to complete the PERIL prediction tool when attending to eligible patients at their residence. Our primary outcomes include subsequent death, hospitalization, two or more recurrent EMS encounters or loss of functional autonomy within one year of the index EMS encounter. Paramedic records will be linked to administrative databases to determine their use of health services. The secondary outcome of functional status will be assessed by telephone follow-up, blinded to paramedic records. We will use principal components analysis and logistic regression to test the statistical association between predictive variables and adverse outcomes. A sample size of 600 will have 80% power to detect odds ratios for predictive variables of 1.93-2.20, given a rate of adverse events of 15% or more. Our pilot data demonstrate we can complete the derivation within three years using the EMS systems in three cities.
The proportion of the population at risk for adverse outcomes due to frailty and the length of time they live at risk is increasing dramatically. Helping older people to remain healthy and independent is important. The PERIL prediction tool will serve as a simple flag that can alert providers at multiple levels within the health care system that a patient is at high risk for adverse and potentially preventable outcomes. Use of the PERIL tool may lead to interventions that will promote functional autonomy and healthy and successful aging among older persons at greatest risk. The current study will take advantage of 855 Toronto EMS paramedics already trained in the pilot study and the cohort of 210 patients already enrolled. This will be the largest ever study of the use of EMS systems to identify vulnerable older people, and will be the first conducted in multiple EMS systems.
Once the PERIL prediction tool is successfully derived, we will validate its predictive ability in an independent patient sample, assess the use of the tool to target high-risk patients for interventions and ultimately assess the impact this has on preventing adverse outcomes.