Emergency calls concerning older patients: Are the appropriate questions asked?

Abstract

Aim

In the present study, we evaluated the triage process particularly for older patients after calls to Emergency Medical Call Centers (ECC), according to the geriatric assessment tool.

Methods

In this observational population‐based cross‐sectional study in the Rhône (France), we analyzed the audiotapes of all calls received by ECC concerning patients aged ≥75 years, during seven randomly selected days, over a period of 1 year. We analyzed whether information about seven key items, predefined by a panel of experts as essential for quality telephone triage of seniors, was actually collected.

Results

Among 4168 calls, 712 (17.1%) concerned patients >75 years (mean ± SD, age 84.6 ± 5.6 years). The mean duration of calls was 3 min 28 s. Information about living arrangements (alone or not), dependency, multiple pathologies, polymedication, ability to walk independently or with help, and hospitalization in the previous 3 months was not collected in 20%, 42%, 40%, 45%, 58% and 61% of calls, respectively. All seven geriatric items were collected for only 54 (7.8%) calls, and only three criteria collected for 277 (40%) calls. Nurse‐managed calls were significantly associated with the collection of less geriatric items compared with physician‐managed calls.

Conclusion

Key information is particularly important to guide the orientation, and further management of older patients may be lacking during the telephone triage of patients in ECCs. This may represent an important level of improvement of the triage process, to address the needs of older patients better and avoid inappropriate emergency department visits. Geriatr Gerontol Int 2023; 23: 54–59.

Emergency telephone calls have to manage complex needs quickly, particularly of the elderly. We analyzed the audiotapes of 712 calls for emergency services for the elderly according to seven geriatric items. Only three criteria were obtained for 40% of calls and seven geriatric items were obtained for 7.8% of calls.

Introduction

As life expectancy increases, there is a rise in chronic health problems and their related complications and, consequently, a growing demand on emergency medical services particularly for older patients. ¹ In France, patients call for medical emergencies a unique phone number (#15), and their calls are managed by nurses and/or physicians of the corresponding regional emergency call center (ECC). These ECCs are faced with the challenge of managing, in a rapid and adapted manner, the complex needs of an aging population. In 2014 in Sweden, more than half the calls made to ECC were related to patients >65 years old. ² Telephone triage of older patients is nonetheless difficult due to the diversity of clinical profiles and presentations in this population, as well as to the frequent issues regarding communication with older patients. ³ Faced with these problems, and without an adequate geriatric assessment tool, the ECC responders often have to refer patients to the emergency department (ED) for assessment. ⁴

This results in many inappropriate ED visits by older patients (12–21% of admissions ⁵ ), which can sometimes contribute to the saturation of the entire care chain. In addition, the management of frail older people in the ED is likely to be more complex and time‐consuming than for younger patients, ⁶ and EDs are often not well equipped ⁷ , ⁸ to handle older patients when they do not require acute care. ⁸ Finally, it has been shown that being admitted in ED may even become dangerous for the welfare of these vulnerable patients. ⁹

To improve the care and quality of life of older patients and to optimize the ED workload, it is essential to consider how calls to ECCs involving seniors are managed and to integrate elements of the underlying medical and social context of patients.

The main objective of this study was to assess the type and frequency of the information collected according to seven predefined key items of geriatric risk assessment during the process of triaging calls to ECC concerning older patients, and to identify the factors associated with not collecting this information.

Methods

We conducted a retrospective study in the centralized ECC of the Rhône county, comprising a population of 1.9 million inhabitants of whom 8% are aged ≥75 years. ¹⁰ This ECC manages about 700 000 emergency calls/year. Calls regarding individuals aged ≥75 years were selected. In total, seven non‐consecutive study days were randomly selected over a period of 1 year to consider weekly and seasonal variations in the number and the nature of the calls. Each day of the week from Monday to Sunday was represented in the sample, and each of the seven study days was approximately separated from the others by 1.7 months to represent all the seasons of the year.

All calls made by or concerning patients aged ≥75 years during the 24 h of each of the seven randomly selected days were analyzed extensively during the complete triage process. In the normal ECC procedure, calls are initially handled by a Nurse Officer who then directs the patient to either an emergency physician or a general practitioner depending on the degree of emergency and severity of the health problem. We excluded calls that did not lead to a contact with a physician, such as inter‐hospital transfers or callers hanging up before the medical decision was made.

AV and NJ, two physicians that specialized in geriatric medicine, analyzed the anonymized audiotapes of all calls meeting the inclusion criteria. The collected data were identity of the person making the phone call to ECC and his/her relationship with the patient (relative, informal or professional caregiver), demographic data of the patient (age, sex), hour and day of the call, duration of symptoms (more or less than 24 h), and reason for the call, i.e., trauma (falls and wounds) or medical reasons (other than falls or wounds). The reason for the call was categorized into two main groups, trauma or medical, rather than being classified by medical specialty, as the causative factor in older patients is often multifactorial. ¹¹

An expert panel of 10 professionals was created, composed of geriatricians, ED and ECC physicians, and epidemiologists. Based on clinical experience and a literature review of the existing geriatric assessment tools, ¹² the expert panel agreed on a set of items that seemed essential for managing emergency calls from older patients, and could be collected by phone within the timeframe of an ECC interview (i.e., limited to 4.5 min). A consensus was reached on a list of seven key geriatric items: (i) place of living (home/nursing home), (ii) living with of another person (Y/N), (iii) dependence (Y/N), (iv) walking (unaided/with help), (v) hospitalization in the previous 3 months (Y/N), (vi) presence of multiple pathologies (Y/N), and (vii) presence of polymedication (Y/N). The patient's place of residence was considered as the first evaluation criteria as it influences the decision‐making process for older patients. ¹¹ , ¹³ Living alone or with another person is an item present in the Identification of Senior at Risk score (ISAR), a recognized standardized tool that helps in identifying patients losing or at risk of losing their autonomy. ¹⁴ , ¹⁵ The functional assessment of patients, such as walking with or without aid, was chosen as an essential item assessment according to Fried et al.'s criteria for defining frailty ¹⁶ that can be made over the telephone. Whether or not information about each of the seven geriatric items was asked by the healthcare professional managing the call was recorded. In addition, the caller (patient/relative/health professional), level of dependence (need for human aid), history of hospitalization within the previous 3 months, presence of multiple pathologies (at least three comorbidities, as described in the Comprehensive Geriatric Assessment ¹⁷ ) and polymedication (three medications or more) were collected. The report form was adapted from the ISAR tool. ¹⁸

Each item defined above that was not present in the audiotaped calls was considered as not asked by the ECC. The amount of geriatric data not recorded was used as the outcome criterion to assess the frequency and completeness of the geriatric evaluation.

Data were analyzed anonymously. Qualitative variables were expressed as count and compared using the chi‐squared test or Fisher's test where appropriate. Quantitative variables were expressed as mean ± SD and compared using the Student t‐test or the Wilcoxon test in case of non‐normal distribution.

We used univariate and multivariate linear regression model to analyze the association between the number of geriatric criteria recorded and several factors, such as the duration of the call, age and place of living of patients, and reason for the call. Finally, we compared the management of calls related to a fall to those related to another reason, as falls are particularly frequent and may be handled differently.

All tests were two‐tailed and a P < 0.05 was considered significant. All analyses were performed using the SAS 9.2 software (SAS Institute Inc., Cary, NC, USA).

Results

In total, 4168 calls were received over the 7‐day study period; among them, 712 (17%) concerned patients aged >75 years, and 20 of these were excluded for the following reasons: patient dead at the time of the call (n = 9), secondary transfer (n = 4), regulation by another French department ECC (n = 3), patient already hospitalized (n = 1), age error (n = 2), patient hanging up before the medical decision was made (n = 1). Overall, 692 calls were included (Fig. 1).

Figure 1.

Flow chart of calls received during the study period and eventually included in the study.

The patients concerned by the included calls were aged between 75 and 101 years, their mean ± SD age was 84.7 ± 5.6 years. In total, 215 (31%) calls were received over the weekend (Table 1), 479 (69.2%) were received between 08.00 and 18.00 h, and there were two peaks in the frequency of calls over the 24‐h period, i.e., between 08.00 and 00.00 h, and between 17.00 and 21.00 h (Fig. 2). Medical and traumatic problems accounted respectively for 517 (74.7%) and 175 (25.3%) ECC calls (Table 1).

Table 1.

Characteristics of the included Emergency Medical Call Centers calls and concerned patients

	Calls total, N = 692
		n (%)
Age (years)	75–84	355 (51.3)
	85–94	311 (44.9)
	>95	26 (3.8)
Sex	Male	256 (37.0)
	Female	436 (63.0)
Time of call	08.00–18.00 h	398 (57.5)
	18.00–08.00 h	294 (42.5)
Day of call	Monday–Friday	473 (68.4)
	Weekend	219 (31.6)
Duration of symptoms (h)	≤24	502 (73.0)
	>24	186 (27.0)
Reason for the call	Medical	517 (74.7)
	Trauma	175 (25.3)

Figure 2.

Distribution of the included calls over 24 h.

The main symptoms reported were shortness of breath, unconsciousness, fever, chest pain and behavioral disorder. Falls represented 165 (23.8%) of all calls.

The caller was the patient called in 104 (15%) cases, the remaining calls were made by different types of informal caregivers or health professionals, mostly nursing staff or healthcare assistants (n = 115), physicians (n = 27) and tele‐monitoring companies (n = 23). In total, 103 (14.9%) patients were living in nursing homes (Table 2).

Table 2.

Specific geriatric items recorded during the call

	Total number of calls: n = 692
		n (%)
Caller	Patient	105 (15.2)
	Entourage	358 (51.7)
	Caregiver	229 (33.1)
Place of living	Private home	582 (84.1)
	Secure‐care facility	103 (14.9)
	N/A †	7 (1.0)
Presence of another person in the home	Yes	357 (51.6)
	Alone	195 (28.2)
	N/A †	140 (20.2)
Dependence	Non‐dependent	120 (17.3)
	Dependent	283 (40.9)
	N/A †	289 (41.8)
Walking	Unaided	139 (20.1)
	With help	154 (22.2)
	N/A †	399 (57.7)
Hospitalization <3 months	No	46 (6.6)
	Yes	224 (32.4)
	N/A †	422 (61.0)
Multiple pathologies	≤3	179 (25.9)
	>3	237 (34.2)
	N/A †	276 (39.9)
Polymedication	No	117 (16.9)
	Yes	265 (38.3)
	N/A †	310 (44.8)

^† Not recorded during the call by the regulator officer. N/A, not available.

Information about the place of living was collected in 685 (99.0%) cases. Information about the presence of another person living with the patient was missing in 138 (20%) cases. Information about the level of dependency and ability to walk with or without technical assistance were missing in 291 (42%) and 401 (58%) cases, respectively. Information about the existence of multiple pathologies, polymedication and hospitalization during the previous 3 months were missing in 277 (40%), 311 (45%) and 422 (61%) calls, respectively. All seven geriatric evaluation criteria were obtained for only 54 (7.8%) calls, and 277 (40%) of calls had only three geriatric items recorded; six items were collected for 114 (16.5%) calls, five for 117 (16.9%), four for 144 (20.8%), three for 127 (18.4%), two for 75 (10.8%), one for 46 (6.6%) and none for 15 (2.2%). The mean ± SD duration of the telephone interview was 3 min 28 s± 2. In univariate regression analyses (Table 3), we found that greater duration of call, time of call 18.00–08.00 h, weekend calls and greater duration of symptoms were significantly associated with a higher number of geriatric criteria recorded during the call.

Table 3.

Univariate and multivariate linear regression analysis of the association of factors with the number of geriatric criteria assessed during the call

		Univariate analyses			Multivariate analyses
Variable		β	[95% CI]	Pr > |t|	β	[95% CI]	Pr > |t|
Duration of call, n = 674/692		0.004	[0.004; 0.005]	<0.0001	0.003	[0.002; 0.004]	<0.0001
Age		−0.015	[−0.038; 0.008]	0.1895	−0.017	[−0.038; 0.004]	0.1
Gender	Male	ref			ref
	Female	−0.100	[−0.365; 0.164]	0.4558	−0.076	[−0.316; 0.164]	0.53
Time of call	08.00–18.00 h	ref			ref
	18.00–08.00 h	0.301	[0.044; 0.559]	0.0216	0.127	[−0.112; 0.366]	0.29
Day of call	Week (Mon–Fri)	ref			ref
	Weekend	0.295	[0.021; 0.568]	0.0346	0.319	[0.068; 0.571]	0.01
Reason of call	Medical	ref			ref
	Trauma	−1.125	[−1.406; −0.843]	<0.0001	−0.589	[−0.905; −0.272]	0.0003
Place of living, n = 685/692	Home	ref			ref
	Nursing home	0.224	[−0.131; 0.579]	0.2150	0.135	[−0.241; 0.511]	0.48
Duration of symptoms (h), n = 688/692	<24	ref			ref
	>24	0.417	[0.132; 0.701]	0.0042	0.213	[−0.069; 0.494]	0.14
Caller	Patient	ref			ref
	Patient family	0.323	[−0.049; 0.695]	0.0887	0.456	[0.114; 0.798]	0.009
	Care pract.	0.134	[−0.260; 0.530]	0.5022	0.579	[0.170; 0.987]	0.006
Officer	Emerg. phys.	ref			ref
	General pract.	−0.350	[−0.633; −0.068]	0.0152	−0.064	[−0.347; 0.219]	0.66
	Nurse	−1.367	[−1.713; −1.020]	<0.0001	−0.502	[−0.896; −0.107]	0.01

CI, confidence interval; pract., practitioner; emerg., emergency; phys., physician.

Conversely, calls related to a trauma, and calls managed by a practitioner for old patients (POP) nurse (vs. an emergency physician) were associated with a lower number of geriatric criteria assessed during the call. In multivariate analyses, the greater duration of call and weekend calls, but also calls made by the family practitioner or the patient family (vs. the patient him/herself) were significantly associated with a higher number of geriatric criteria recorded during the call.

In total, 175 (25.3%) calls were made for traumatic problems, particularly falls. The two other main reasons for calling were faints (n = 57, 8.2%) and dyspnea (n = 45, 6.5%). Compared calls related to non‐traumatic problems, the duration of call tended to be shorter for fall‐related calls (mean ± SD, 2 min 32 s ± 2) and information about the number of treatments taken by the patient was collected less frequently (P < 0.001). These patients were more often referred to the ED (n = 121/N, %) than when calls were made for medical reasons (P = 0.009).

Discussion

This study analyzed the audiotapes of 692 telephone interviews concerning older patients, which were conducted over seven randomly selected days spread evenly between weekdays and weekends and over the four seasons of 1 year. Seventeen percent of the total calls concerned people aged ≥75 years (7.8% of the population in the Rhône region ¹⁰ ), and nearly about two of three patients were sent to an ED. We observed that all seven geriatric evaluation criteria that we selected were collected in <8% of the calls, and that in 40% of calls only three geriatric items were recorded. Notably, information about dependency was lacking in >40% of calls, and the information on recent hospitalization was not gathered in 60% of the calls. We also found that more geriatric criteria were collected in weekend calls, when the reason for the call was not related to a trauma, when the family practitioner or the family was calling (vs. the patient), and when the emergency practitioner managed the call (vs. a registered nurse).

At the ECC level, the distinction between real and perceived medical emergencies can sometimes be difficult. ¹⁹ In the aged population, telephone triage may be even harder due to more frequent communication problems, comorbidities, and/or complex psychosocial situations. ²⁰ A number of tools have been developed and tested for improving the triage of elderly people in EDs. ²¹ However, an effective triage by the ECC staff could save older patients from unnecessary or inappropriate visits to the EDs. Furthermore, the evaluation tools used in hospital settings are not adapted for use in the ECC, as they require a physical examination and are often time‐consuming.

Within the usual functioning, regardless of patient characteristics, the ECC physician organizes patient management plans based mainly on the reason for the call and the symptoms. However, recent studies have suggested that the paradigm for the management of older patients should shift from a pathology‐centered model to a more holistic assessment of patients. Therefore, ²² the triage of older people requires the use of other specific criteria involved in decision‐making in order to make an informed and personalized medical decision, as recommended in ECC guidelines. ³ , ⁴ , ²³

For older patients, it is crucial to adopt a holistic approach to patient care, not only based on pathological findings but also on other factors such as dependency, cognitive status, living environment, or social support, and hence to initiate an adapted management plan. ³ , ²³ , ²⁴ In this study, patient history or current medication were not elicited for one in two patients, although this information is essential for understanding the clinical presentation. Furthermore, older patient mobility, which reflects their autonomy, or previous hospitalizations, were not documented in 60% of calls. Finally, in 20% of calls, information regarding the presence of a family or formal caregiver at home was not obtained, although this is a crucial element for deciding whether monitoring at home can be considered.

There was no direct link between the length of the call and the amount of information gathered. This element can make one think that time is not the only barrier at stake: other determinants, such as the lack of knowledge of older patients and the specific aspects of their management, as well as the lack of tools to help in finding alternatives to sending patients to the ED, may also constitute barriers to the collection of relevant information. As two of three of the calls took place during working hours (08.00–20.00 h), it should be possible to mobilize more appropriate channels, such as geriatric hotlines or other care geriatric healthcare networks.

Calls related to falls accounted for more than a quarter of the calls concerning older people, and in such cases, triage was shorter and based on fewer geriatric data, some information was missing, particularly regarding patient current medication, despite recommendations. ²⁵ In an observational study about the pre‐hospital management of aged fallers, causes and/or complications of falls were responsible for the referral of 75% of older patients to the ED. ²⁶ Our study found similar results. Falls are a complex geriatric presentation that quadruple the risk of death within the year. ²⁷ The evaluation should determine whether a referral to the ED is necessary and beneficial for the patient. Finally, immediately directing the fallers towards specific geriatric healthcare pathways may represent a possible solution, and in case referral to the ED is necessary (excluding trauma assessment, suture necessary, etc.), a better anticipation of this access may help to improve patient management quality, safety and effectiveness.

The demographic statistics (DREES) in the Rhône department have revealed that almost 10% of people aged ≥75 years lived in a nursing home in 2012. ²⁸ In the present study, 15% of calls concerned this population. In this subgroup, information regarding the geriatric items was not better collected, although a specific emergency medical file exists and is accessible to the nursing‐home staff, highlighting the lack of coordination and communication between these two structures (ECC and nursing homes). ²⁹

There are several limitations of this study. First, the selected study days may not be representative of the whole year. We could not analyze a whole year of calls due to the large number of calls, and we aimed to have the best representation possible of all the days of the week and of the different seasons of the year. Secondly, the selected items for geriatric assessment may be disputable. Thirdly, the design of this study is retrospective. However, this allowed the gathering of data that was representative of the routine care process and not sensitive to the Hawthorne effect.

Conclusions

This study confirms that specific aspects of older patients' management are often not addressed by ECC healthcare providers during the telephone calls. Using a triage tool incorporating some of the items suggested could help in better directing the calls concerning older patients, as well as those not biologically old.

We underline that what we know from this study is these key elements are not often asked, and more study is needed to determine the impacts, positive or negative, of including additional geriatric elements in the phone triage of these patients.

Disclosure statement

The authors declare no conflict of interest.

Jomard N, Vincent A, Chammem R, et al. Emergency calls concerning older patients: Are the appropriate questions asked? Geriatr. Gerontol. Int. 2023;23:54–59. 10.1111/ggi.14509

Data availability statement

The datasets used and analyzed in this study are available from the corresponding author upon reasonable request.