• 
• 

Three Part Question

In [older, frail patients attended by an ambulance crew] which [screening tools for frailty] are most accurate at [identifying patients who can be safely left at home]?

Clinical Scenario

You are asked to attend to an 87 year old female patient who fell at home. She has no acute injuries, but is quite distressed at having been brought to hospital. The ambulance crew state that she did not fulfill their criteria for home management. You wonder if all pre-hospital screening tools for frailty are equally effective at deciding which patients can be safely left at home.

Search Strategy

EMBASE
[Aged/ or Elderly/ or Geriatrics/ or Geriatric.mp.] AND {[Emergency health service/ OR Emergency Medical Services.mp OR EMS.mp. OR Pre-hospital.mp. OR Prehospital.mp. or Ambulance.mp. or Ambulance transportation/ OR Paramedic.mp. OR Paramedic personnel/] AND [Medical decision making/ OR Decision making/ OR Clinical judgement .mp. OR Clinical decision making/ OR Clinical Decision Support Systems.mp. OR Clinical decision.mp. OR Clinical decision rule.mp. OR Screening/ screening.mp. OR Screening tool.mp. OR Triage.mp.]} AND [Transfer.mp.OR Transport OR Hospital admission/ or Non-transport.mp. OR Non-transport$.mp. OR Home care/mp. OR admission/ OR Re-attendance.mp. OR Return.mp. or Self care.mp. OR Non-conveyance.mp. OR Convey OR Conveyance]


MEDLINE
[Aged/ or Elderly/ or Geriatrics/ or Geriatric.mp.] AND [Emergency health service.mp. or Emergency Medical Services/ OR Pre-hospital.mp. OR Prehospital.mp. OR Ambulance.mp. OR Ambulances/ OR Paramedic.mp. OR Allied health personnel/] AND [Decision making/ OR Clinical judgement .mp. OR Clinical decision making/ OR Clinical Decision Support Systems/ OR Clinical decision.mp. OR Clinical decision rule.mp. OR Decision support techniques OR Screening/ screening.mp. OR Screening rule OR Triage] AND [Transfer.mp. OR Transport.mp. OR Transport$.mp. OR “transportation of patients”/ OR Patient admission/ OR Hospital admission .mp. OR Hospitalization/ OR Unscheduled admission.mp. OR Non-transport.mp. OR Non-transport$.mp. OR Self care/ OR Home care.mp. OR Home care services/ OR Re-attendance.mp. OR Return.mp. OR Self care/ OR Non-conveyance.mp. OR Convey.mp. OR conveyance.mp. OR convey$

Using EMBASE the search strategy found 854 articles of which found two relevant articles. Using MEDLINE the search strategy found 386 articles, with no further relevant articles.

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Lee JS, Richard Verbeek P, Shull MJ et al. 2016 Canada	Patients aged 65 and over attended by a paramedic crew from one of three EMS services in Ontario, Canada	Prospective observational study that tested a clinical decision rule that aimed to identify high-risk older adults using paramedics observations. The Paramedic assessing Elders at Risk of independence loss (PERIL) Rule is a checklist of 43 yes/no questions. Tested on 1065 subjects of which 764 had complete data.	Inter-observer Reliability; predictive accuracy of the best performing items on the PERIL scoresheet	Inter-observer reliability was good or excellent for 40/43 questions; 4 most significant predictors. 1)	Large amount of missing data. Study underpowered. Not externally validated. Based in Canada, therefore may not be generalisable to a UK population. Unable to collect data on patients not screened, possible
Snooks HA, Carter B, Dale J et al. September 2014 United Kingdom	Patients aged 65 and over; living in the catchment area of a participating falls service; and attended by a study paramedic following their first emergency call categorised by the call-taker as a fall during the study period.	Prospective cluster randomised trial study that evaluated effectiveness, safety and cost-effectiveness of Computerised Clinical Decision Support (CCDS) for paramedics attending older people who fall. The setting was 13 ambulance stations in two UK emergency ambulance services. 42 of 409 eligible paramedics, attended 779 older patients for a reported fall. 17 intervention paramedics used CCDS for 54 (12.4%) of 436 participants. 18 paramedics in control group who assessed 343 patients.	Referral to fall service; adverse events; job cycle time and pattern and effects of CCDS usage	Referrals: 42 (9.6%) patients were referred to falls services, compared with 17 (5.0%) of 343 participants seen by 19 control paramedics [Odds ratio (OR) 2.04, 95% CI 1.12 to 3.72]; Adverse events: No adverse events were related to the intervention. Non-significant differences between groups included: subsequent emergency contacts (34.6% versus 29.1%; OR 1.27, 95% CI 0.93 to 1.72); quality of life (mean SF12 differences: MCS 20.74, 95% CI 22.83 to +1.28; PCS 20.13, 95% CI 21.65 to +1.39) and non-conveyance (42.0% versus 36.7%; OR 1.13, 95% CI 0.84 to 1.52); Job cycle time: ambulance job cycle time was 8.9 minutes longer for intervention patients (95% CI 2.3 to 15.3); Pattern and effects of CCDS usage: 54/436 patients in the Intervention arm received the intervention of CCDS. Where used the CCDS increased the proportion of patients referred to fall service [(12/54 22.2% vs 30/382 (7.9%)] and those not conveyed to the ED [35/54 (64.8%)	Small number of eligible paramedics participated, may introduce selection bias. Large number of patients lost to follow-up or refusing to consent. Underpowered, samples size calculation of 865 however only 770 participants. Unequal groups due to illness of paramedics group in control group. In intervention group use of CCDS low. 22/54 cases where CCDS used only a single paramedic. Only looking specifically at falls rather than fraility. Not other presenting complaints.

Comment(s)

There is limited evidence regarding the use of clinical decision rules to decide which patients need to attend the ED and which can safely stay at home. Previous literature has explored the characteristics of patient not being transported, only two papers have formally assessed the accuracy of clinical decision rules in this setting. Neither paper explicitly assessed patients in terms of ‘frailty’, though the outcomes were relevant to this subset of patients. The first by Lee et al is based in Canada and would need validation in the UK. The second, while UK based suffered from low recruitment rates and potential selection bias. Several previous studies have analysed the epidemiology or characteristics of patients that paramedics do not transfer to hospital but only two have examined the use of a clinical decision aid.

Clinical Bottom Line

Further research is needed to identify the most effective screening tool for non-conveyance of older frail patients attended by pre-hospital practitioners.

References

1. Lee JS, Richard Verbeek P, Shull MJ et al. Paramedics assessing elders at risk for independence loss (PERIL): Derivation, reliability and comparative effectiveness of a clinical prediction rule. Canadian Journal of Emergency Medicine 2016 18 (2):121-132
2. Snooks HA, Carter B, Dale J et al. Support and assessment for fall emergency referrals (SAFER 1): Cluster randomised trial of computerised clinical decision support for paramedics. PLoS ONE 2014. 9 9 (9) (no pagination), 2014. Article Number: 0106436