Uses of first line emergency services in Cuba

1. Introduction 

The historical function of hospital emergency departments was to care for patients who did not have the financial means to consult private physicians [1]. Their function shifted towards ensuring adequate response to accidents and ‘real’ emergencies when social security systems and financially accessible first line health services became available. However, many patients continue to visit emergency departments with problems that could have been solved at the level of a family doctor or health care centre. The literature extensively documents such inappropriate use or even ‘abuse’ and indicates that a substantial proportion of patients would have been better off being treated by a competent general practitioner [2], [3], [4], [5], [6], [7]. Even after introducing a broad array of initiatives to canalize ‘primary care patients’ to the corresponding level of care, ‘inappropriate use’ of the emergency services frequently continues to affect their adequate, or at least their efficient functioning [8], [9].

Cuba's national health system is also confronted with this problem, that was exacerbated at the beginning of the 1990s, when radical changes in the terms of trade with the former Soviet Union and the tightening of the USA blockade drastically reduced the national income [10], [11]. Health care spending dropped and scarce resources were concentrated at hospital level [12]. This resulted in patients, who could have been attended by family physicians, overcrowding hospital emergency units, perusing resources and possibly receiving sub-optimal care [13]. The health authorities developed plans for rationalizing service delivery and increasing efficiency without recurring to privatization or cost recovery. Despite the economic crisis, they ratified the principles of state responsibility and free universal coverage, effectiveness and participation [14], [15]. A programme of quality improvement was set up [11], [16], [17], [18], including an extra muros emergency care system in which first line policlinics and networks of community-based family doctors or general practitioners (GP's) played a central role [19], [20].

Cuban municipalities are divided into health areas, in which a policlinic organizes the first line care for about 30000 inhabitants [11], [19]. The policlinic covers the whole population with a family doctor program (one GP and one nurse per 700–800 inhabitants), and ensures the necessary diagnostic support and backstopping by specialists. There is a hospital per 150000 inhabitants. Until 1995, only hospital emergency services (HES) offered high standard emergency care. From 1996 onwards, each municipality developed a first line emergency service (FLES) consisting of one principal emergency policlinic (an existing policlinic), and – in function of population density – a variable number of (small) peripheral or auxiliary emergency centers (other policlinics or strengthened GP-cabinets) [21], [22]. In this new set up the patient can attend directly the principal first line emergency policlinic or, possibly, the nearest peripheral emergency centre where an ambulance can be called. The FLES is functioning 24h a day. The principal emergency policlinics are staffed with specialized personnel seconded by GP's and have a high technical plateau compared with the (poorly equipped) GP-cabinets. However, contrary to GP's the FLES is not ensuring continuous integrated care [23].

The first line emergency services have been contributing to diminishing pressure on the hospital emergency services, but at the same time the demand for non-emergency care at the family doctor level seemed to shift to the FLES [23]. We analyse here the use of the FLES and its determinants, in order to develop proposals to channel ‘non-emergency’ FLES-patients to their family doctors.

2. Methods 

3. Results 

Over the 2 years observation period, 24879 and 59795 patient contacts were registered with the principal emergency policlinic in Baracoa and Cerro, respectively. Overall, attendance was equally distributed over the days of the week. In both municipalities the under-five and the over 65 years age groups were over represented in comparison to the Cuban population pyramid (Table 1). Also, female attendance reached more than 60% in the adult groups. In Baracoa the patients were more frequently referred (p<0.01) and a higher proportion attended during day-time (p<0.01). The main motives for consultation, both in Baracoa and in Cerro, were ‘control of blood pressure’ and ‘shortness of breath’. The most frequent diagnosis in Baracoa was hypertension (16.0%), against asthma (20.2%) in Cerro. In Baracoa and Cerro respectively, trauma contributed only 3.7% and 4.3% of the consultations; 9.8% and 14.1% of patients were asymptomatic; 18% and 37% of the cases were discharged, without further technical intervention or treatment, after anamnesis and clinical examination only; 34.2% and 62.1% of the patients were referred to their family doctor; and hospitalization at home was suggested in 15.9% and 1.9%.

Table 1. Characteristics of attendances at the first line emergency services (FLES) in Baracoa and Cerro, Cuba (July 1999–June 2001)

		Baracoa, n=2540	Cerro, n=2022
	1. Patient characteristics
	Mean age (years)	39.9 (S.D.: 24.7)	35.4 (S.D.: 22.9)
	Age distribution
	<1 year	2.7%	2.9%
	1–4 years	7.3%	8.7%
	5–14 years	9.8%	11.9%
	15–24 years	8.4%	11.2%
	25–44 years	27.9%	29.6%
	45–64 years	25.5%	23.5%
	≥65 years	18.4%	12.2%
	Female patients:	61.1%	58.6%
	Proportion female in child group (0–14 years)	46.6%	48.0%
	Proportion female in adult group (≥15 years)	64.7%	61.8%
	2. Consultation characteristics
	Spontaneous consultation	85.4%	99.2%
	Referred to FLES	14.6%	0.8%
	By general practitioner (GP)	2.2%	0.5%
	By peripheral emergency points	5.0%	0%
	By schools, factories, etc.	7.3%	0.3%
	Moment of consultation
	8 a.m.–4 p.m.	59.4%	44.3%
	5 p.m.–7 a.m.	40.6%	55.7%
	Main groups of motives for consultation
	Control blood pressure	14.1%	17.9%
	Short of breath	13.4%	21.4%
	Headache	11.7%	9.0%
	Fever	8.0%	6.7%
	Cough	2.5%	5.7%
	Other	50.3%	39.3%
	Main groups of diagnosis
	Hypertension	16.0%	a
	Asthma	9.6%	20.2%
	Upper respiratory infection	9.2%	17.3%
	Other	65.2%	62.5%
	Procedures
	Anamnesis and clinical examination only	18.2%	37.2%
	Technical exams (lab, RX, etc.)	8.3%	7.0%
	Specialist consultation	15.6%	1.8%
	Observation in emergency unit	16.5%	0.5%
	Herbal treatment	31.7%	5.7%
	Therapeutic measures (drugs, aerosol, etc.)	56.7%	53.0%
	Emergency	45.3%	38.7%
	Inappropriate use	57.6%	58.3%
	Outcome
	Discharged	38.4%	32.5%
	Referred to GP	34.2%	62.1%
	Referred to GP for hospitalization at home	15.9%	1.9%
	Hospitalization	11.5%	3.5%
	Deceased	0%	0.05%

a No precise data available.

In Cuba home hospitalization is defined as the intensive follow-up at home, by the first line health personnel, of patients that require bed rest and/or isolation, without having a compelling technical need for hospital admission [27], [28]. The percentage of patients hospitalized was three-times higher in Baracoa than in Cerro (11.5% versus 3.5%) (p<0.01), but the relative frequency of ‘inappropriate use’ was similar – just below 60% – in both municipalities. The agreement between the physicians’ classification of “emergency” based on the routine criteria used in Cuba and our constructed variable of ‘inappropriate use’ was poor (kappa of 0.31 for Baracoa and 0.34 for Cerro). In particular, at least 30% of the “emergency cases” received herbal treatment only.

Table 2 compares the characteristics of the patients attending during day- and night-time in the sample used for in depth analysis. In both municipalities night-time cases were less inappropriate users, even more prominently so in Cerro. This was also reflected in a higher frequency of child cases, of asthma crises, of therapeutic measures and of being hospitalized afterwards. During the night, less elderly were seen and less patients attending for control of blood pressure.

Table 2. Characteristics of day- and night-time attendances at the first line emergency services (FLES) in Baracoa and Cerro, Cuba (July 1999–June 2001)

		Baracoa (n=2540)			Cerro (n=2022)
		Day-time (8 a.m.–4 p.m.) (n=1510)	Night-time (5 p.m.–7 a.m.) (n=1030)	p-valuea	Day-time (8 a.m.–4 p.m.) (n=896)	Night-time (5 p.m.–7 a.m.) (n=1126)	p-valuea
	1. Patient characteristics
	Mean age (years)	42.4 (S.D.: 24.3)	37.0 (S.D.: 24.9)	<0.01	39.0 (S.D.: 23.0)	32.6 (S.D.: 22.5)	<0.01
	Age distribution			<0.01			<0.01
	<1 year	2.6%	2.8%		1.5%	4%
	1–4 years	6.2%	8.9%		7.1%	9.8%
	5–14 years	7.8%	12.8%		10.7%	12.8%
	15–24 years	8.3%	8.4%		8.4%	13.5%
	25–44 years	27.0%	29.3%		30.1%	29.2%
	45–64 years	27.9%	21.9%		26.8%	21.0%
	≥65 years	20.2%	15.7%		15.4%	9.7%
	Female patients	61.5%	60.4%	0.57	59.3%	58.0%	0.56
	Proportion female in child group (0–14 years)	43.8%	49.2%	0.21	46.8%	48.7%	0.69
	Proportion female in adult group (≥15 years)	65.0%	64.1%	0.63	62.2%	61.4%	0.73
	2. Consultation characteristics
	Spontaneous consultation	86.1%	82.8%	0.02	99.2%	99.2%	0.96
	Referred to FLES	13.9%	17.2%	0.03	0.8%	0.8%	0.24
	By general practitioner (GP)	2.8%	2.1%		0.8%	0.5%
	By peripheral emergency points	4.5%	5.9%		0%	0%
	By schools, factories, etc.	6.6%	9.1%		0%	0.3%
	Main groups of motives for consultation			<0.01			<0.01
	Control blood pressure	18.0%	8.4%		24.2%	12.8%
	Short of breath	10.6%	17.7%		14.4%	25.6%
	Headache	10.7%	13.0%		7.1%	10.4%
	Fever	7.0%	9.2%		7.0%	6.5%
	Cough	2.9%	2.0%		5.8%	5.7%
	Other	50.8%	49.7%		41.5%	39.0%
	Main groups of diagnosis			<0.01			<0.01
	Hypertension	16.9%	14.7%		b	b
	Asthma	7.7%	12.3%		12.7%	26.2%
	Upper respiratory infection	9.8%	8.4%		17.0%	17.6%
	Other	65.6%	64.6%		70.3%	56.2%
	Procedures			<0.01			<0.01
	Anamnesis and clinical exam only	22.4%	12.1%		45.2%	30.8%
	Technical exams (lab, RX, etc.)	9.6%	6.4%		9.0%	5.4%
	Specialist consultation	13.2%	19.1%		1.0%	2.5%
	Observation in emergency unit	13.2%	21.3%		0.6%	0.5%
	Herbal treatment	29.2%	35.6%		4.5%	6.7%
	Therapeutic measures (drugs, aerosol, etc.)	49%	67.8%		43.4%	60.7%
	Emergency	40.4%	52.5%	<0.01	25.6%	46.3%	<0.01
	Inappropriate use	60.7%	53.1%	<0.01	65.4%	52.7%	<0.01
	Outcome			<0.01			<0.01
	Discharged	42.3%	32.7%		23.4%	39.7%
	Referred to GP	32.2%	37.1%		71.5%	54.5%
	Referred to GP for hospitalization at home	15.9%	16.0%		2%	1.9%
	Hospitalization	9.6%	14.2%		3%	3.8%
	Deceased	0%	0%		0%	0.1%

a For differences between day- and night-time. b No precise data available.

Determinants of inappropriate use were assessed in a logistic regression model (Table 3). There was no marked correlation between inappropriateness of care seeking and gender or age. Moment of consultation was a significant determinant in both settings: inappropriate use was some 50% more frequent during the day than at night. In Baracoa, but not in Cerro, referred patients were four times less likely to be classified as inappropriate users. In both settings these patients had markedly different characteristics (Table 4). More often they were adults and coming during night-time. They were also more frequently held in observation in the emergency unit, more in need of a consultation by a specialist, and up to 12 times more frequently hospitalized.

Table 3. Independent determinants of inappropriate use of FLES in Baracoa and Cerro, Cuba (July 1999–June 2001)

		Baracoa				Cerro
		Appropriate cases (n=1078)	Inappropriate cases (n=1462)	Crude OR (95% CI)	Adjusted OR (95% CI)	Appropriate cases (n=843)	Inappropriate cases (n=1179)	Crude OR (95% CI)	Adjusted OR (95% CI)
	Referred to FLES
	Yes	273	114	–	–	9	7	–
	No	805	1348	4.0 (3.2–5.1)	3.9 (3.1–5)	834	1172	1.8 (0.7–4.9)	a
	Moment of consultation
	5 p.m.–7 a.m.	484	546	–	–	533	593	–	–
	8 a.m.–4 p.m.	594	916	1.4 (1.2–1.6)	1.3 (1.1–1.6)	310	586	1.7 (1.4–2.0)	1.7 (1.4–2.1)
	Age group
	Adults	862	1174	–	a	658	891	–	–
	Children	216	288	1.0 (0.8–1.2)		185	288	1.1 (0.9–1.4)	1.2 (1–1.5)
	Sex
	Male	428	559	–	a	349	489	–	a
	Female	650	903	1.0 (0.9–1.2)		494	690	1 (0.8–1.2)

a Not significant in stepwise logistic regression.

Table 4. Characteristics of and procedures for referred vs. spontaneously consulting FLES patients in Baracoa and Cerro, Cuba (July 1999–June 2001)

		Baracoa				Cerroa
		Not referred cases (n=2153)	Referred cases (n=387)	Crude OR (95% CI)	Adjusted OR (95% CI)	Not referred cases (n=2006)	Referred cases (n=16)	Crude OR (95% CI)
	Age
	Children	451	53	–	–	467	6	2.0 (0.7–5.5)
	Adults	1702	334	1.7 (1.2–2.3)	2.3 (1.5–3.4)	1539	10	–
	Sex
	Female	1329	224	–	b	1176	8	–
	Male	824	163	1.2 (0.9–1.5)		830	8	1.4 (0.5–3.8)
	Moment of consult
	8 a.m.–5 p.m.	1300	210	–	b	889	7	–
	6 p.m.–7 a.m.	853	177	1.3 (1–1.6)		1117	9	1 (0.4–2.7)
	Inappropriate user
	Yes	1348	114	–	–	1172	7	–
	No	805	273	4.0 (3.2–5.1)	1.5 (1.1–1.9)	834	9	2.1 (1.5–2.8)
	Specialist consult
	No	1971	173	–	–	1974	11	–
	Yes	182	214	13.3 (10.4–17.2)	6.2 (4.6–8.5)	32	5	27.8 (9.2–83.3)
	Observation in emergency unit
	No	1867	254	–	–	1995	16
	Yes	286	133	3.4 (2.7–4.4)	1.9 (1.4–2.6)	11	0
	Hospitalized
	No	2051	198	–	–	1940	12	–
	Yes	102	189	19.2 (14.5–25.6)	11.7 (8.5–16.2)	66	4	9.8 (3.1–31.5)

a Multivariate logistic regression not possible due to limited number of referred cases. b Not significant in stepwise logistic regression.

As the appropriateness of care seeking was determined by the moment of consultation, we contrasted morning (8 a.m.–12 a.m.) and night (11 p.m.–3 a.m.) attendance (Table 5). In both localities night-time consultation was consistently associated with younger age. Also, interventions indicating a higher level of medical need, such as observation in the emergency unit, specialist consultation and hospitalization, were consistently associated with the time of attendance, albeit with some differences between the two study sites.

Table 5. Characteristics of and procedures for night-time (11 p.m.–3 a.m.) vs. morning (8 a.m.–12 a.m.) FLES patients in Baracoa and Cerro, July 1999–June 2001

		Baracoa (n=1200)				Cerro (n=867)
		Morning cases (n=1031)	Night cases (n=169)	Crude OR (95% CI)	Adjusted OR (95% CI)	Morning cases (n=665)	Night cases (n=202)	Crude OR (95% CI)	Adjusted OR (95% CI)
	Age groupa
	Adults	864	122	–	–	547	145	–	–
	Children	167	47	2.0 (1.4-2.9)	3.1 (1.8-5.1)	118	57	1.8 (1.3-2.6)	1.9 (1.4-2.8)
	Sexa
	Female	633	107	–	b	393	112	–	b
	Male	398	62	0.9 (0.6–1.3)		272	90	1.2 (0.8–1.6)
	Referred
	No	900	118	–	b	661	200	–	b
	Yes	131	51	2.9 (2.0–4.3)		4	2	1.6 (0.3–9.1)
	Inappropriate use
	Yes	62	658	–	–	443	99	–	–
	No	107	373	3.0 (2.2–4.3)	2.1 (1.4–3.0)	222	103	2.1 (1.5–2.8)	2.1 (1.5–2.9)
	Specialist consultation
	No	905	118	–	b	659	195	–	1
	Yes	126	51	3.1 (2.1–4.5)		6	7	3.9 (1.3–11.9)	3.8 (1.2–11.6)
	Observation in emergency unit
	No	908	118	–	–	663	201	–	b
	Yes	123	51	3.2 (2.2–4.6)	2.2 (1.5–3.4)	2	1	1.6 (0.1–18.2)
	Hospitalization
	No	943	118	–	–	647	193	–	b
	Yes	88	51	4.6 (3.1–6.8)	3.2 (2.1–4.8)	18	9	1.7 (0.7–3.8)

a Interaction between age and sex: OR 0.5 (0.2–0.9). b Not significant in stepwise logistic regression.

4. Discussion 

While a body of literature on appropriate use of emergency services claims that a variable proportion of patients would have been better off by consulting a competent general practitioner, explicit objective criteria for ‘appropriate’ use are frequently lacking [29], and definitions of ‘minor problems’ and of ‘inappropriate use’ continue to be controversial [30]. In the present study we define “inappropriate use” through a construct inspired on elements from earlier studies [24], [25], [26], and using the available information on the reason for encounter, the diagnosis and the medical conduct. While our definition has undeniably limitations, it assures a high specificity—be it at the expense of sensitivity. As such, our estimates constitute a lower boundary for the true frequency of inappropriate use. Our construct is also, certainly in the Cuban context, a proxy for the volume of health problems that could – and should – easily be handled at the level of a GP cabinet during regular office hours.

We had shown in the past [23] that the overall use of emergency services in Cuba is rather high, particularly in Cerro. At the time of our study, the utilization rates were 0.4 and 1.0cpy for the FLES and 0.6 and 1.4cpy for the HES in Baracoa and Cerro, respectively. We demonstrate here that Baracoa has a lower proportion of spontaneous use and a higher proportion of hospitalization.

To fully understand these differences, a contextual analysis is needed. First, in Baracoa many GP's live in the communities they serve and they can better ensure permanence and more easily develop a relationship of confidence with the population, which encourages patients to contact them rather than to recur directly to the emergency services. Distance and difficulties with transport also favor recourse to the family doctors. Secondly, the centrally located principal emergency policlinic in Baracoa diminishes the pressure on the HES more globally. In Cerro, the principal emergency policlinic is situated in the periphery: It complements the HES for part of the municipality but is not an attractive alternative for the population that resides closer to the hospitals [23]. Thirdly, the higher proportion of hospitalizations in FLES users in Baracoa is related to the fact that the FLES-ambulance is one of the few vehicles available in rural areas to transport critically ill patients and pregnant women from their homes to the hospital. Even if directly referred such patients are administratively included as “hospitalization through the FLES”, which increases the figures on hospital referrals.

Regardless of these differences, the main problem in both municipalities is the high volume of inappropriate use of first line emergency services, especially by non-referred patients and during daytime. Various studies use the proportion of hospitalized cases as a proxy-indicator for the appropriate use of emergency services [8], [31], [32], [33]. In hospital emergency services of the industrialized world these proportions vary from 18–19% in the UK and Ireland, to 34–71% in different studies in France [8], [32], [34]. But hospitalization figures from the Cuban FLES (3.5% in Cerro and 11.5% in Baracoa) cannot be easily compared to these international data, nor be used as a proxy for appropriate use: being a genuine first line service, it is frequently bypassed by cases for which the need for hospitalization is manifest.

Not withstanding, almost 60% of inappropriate use seems high. Still, the FLES ensure permanence and accessibility for emergency care at the first line and can best be compared with the ‘out-of-hours first line (emergency) care’ in European health systems. However, the Cuban FLES functions around the clock and specialists are at hand. Family doctors are comparatively less accessible (restricted opening hours, absent for home visits, etc.) and provide a lower technical platform. It has been observed before that this incites patients to use emergency services as first line services, bypassing GP's who are better placed to ensure continuous and integrated care [31], [33]. This is also compatible with a higher relative frequency of inappropriate use during daytime: When patients feel the subjective ‘urgent’ need to consult a doctor, the attractiveness – for the reasons set out above – of the emergency services leads them to directly consult FLES, also for non-emergencies. The lower frequency of inappropriate use amongst cases referred to the FLES by the family doctor only confirms this hypothesis.

In some countries, financial barriers have been put in place to dissuade non-appropriate health care seeking at emergency services [6], but this is not considered an option by the Cuban authorities. When financial constraints are of marginal importance, the decision to visit a family doctor depends on the patient's perception of his problem, his personality and his expectations and preferences [35], [36], [37], but also on the patient's confidence in the physician (the perception of) his technical capacity, his accessibility and the material conditions he works in [38]. Many authors consider that “inappropriate use” is the result of a rational process that leads the patient to identify the most appropriate way to solve his/her health problem [6], [31], [32], [34], [39], [40]. Therefore, more than at ‘inappropriate use’, this study points at weaknesses in and opportunities for the Cuban family doctor program.

To strengthen the central position of family doctors, an adequate technical plateau must be developed at their cabinets, permanence should be improved by strengthening the collaboration between family doctors working within the same neighborhood, and during day time access to the FLES ought to be limited to referred patients only. Documented counter-referral of “inappropriate” FLES-users to their family doctor has to become the rule and concurrently a clear information and communication campaign should be launched on the nature and purpose of the “family doctor – FLES” set up. Finally, a more systematic GP-involvement in the emergency network could ensure growing confidence of patients attending FLES in ‘their’ family doctor and his technical skills.