Serious conditions among conveyed and non-conveyed patients presenting with nonspecific chief complaints to the ambulance service

Abstract

Background

It is a challenge for the ambulance service to identify which patients to convey to the emergency department (ED). Ideally this would be the patients with serious conditions requiring further care in the ED. However, patients presenting with non-specific chief complaints are difficult to prioritize and typically have normal or near normal vital signs despite up to one third having underlying serious conditions. The proportion of patients with serious conditions among non-conveyed patients with non-specific chief complaints (NSC) as assessed by ambulance clinicians is not known. Therefore, the primary aim was to study the proportion of serious conditions among conveyed and non-conveyed patients presenting to the ambulance service with NSCs. The secondary aim was to study the mortality rate among patients with NSCs in relation to conveyance and serious and nonserious conditions.

Method

A retrospective cohort study of patients ≥ 18 years of age presenting with NSCs to the ambulance service in Stockholm Region between January 1st, 2013 and December 31st, 2013. Patients were identified via the ambulance service electronic health record and followed via records from the the National Patient Register and Causes of Death Registry at Sweden’s National Board for Health and Welfare. Descriptive statistics as well as regression analyses were used.

Results

A total of 4744 patients were included with a median age of 76 years. A serious condition was present in 1398 (29.5%) of the patients. After index assessment by the ambulance service, 3780 (79.7%) were conveyed of which 1334 (35.3%) had serious conditions, compared to 964 (20.3%) who were non-conveyed of which 64 (6.6%) had serious conditions. 30-day mortality was 372 (9.8%) in the conveyance group compared to 32 (3.3%) in the non-conveyance group. If serious conditions were present, the mortality rates were 269 (20.2%) in the conveyance group compared to 11 (17.2%) in the non-conveyance group.

Conclusion

The results show that serious conditions were more than five times more common among conveyed as compared to non-conveyed patients. Mortality rates were three times higher in the conveyance group as compared to the non-conveyance group. Our results suggest that the ambulance personnel play an important role in identifying patients with serious conditions, hence, likely to be in need of treatment in the ED.

Trial registration

N/A

Background

The ambulance service plays a pivotal role in assessing, initiating treatment, and conveying patients to the emergency department (ED) as one of the first steps in the chain of emergency care. Patients present to the ambulance with symptoms, i.e., chief complaints, rather than diagnoses. All chief complaints that are not part of a set of specific complaints or signs or where the primary assessment by ambulance clinicians cannot be established are considered as non-specific chief complaints (NSC) [1]. Assessments in the ambulance are conducted based on objective findings, such as vital signs, in combination with recent medical history, clinical judgement, and guided by the service’s medical guidelines and triage protocols. It is a challenge for the ambulance service to identify which patients to convey to the emergency department (ED). Ideally this would be the patients with serious conditions, requiring further care in the ED.

However, patients presenting with NSCs, are difficult to prioritize and typically have normal or near normal vital signs despite up to one third having underlying serious conditions [2, 3]. This is concerning as these patients are often old people and conveyance to the ED may lead to long waiting times with inherent risks of increased morbidity and mortality [2, 4]. It is therefore important to distinguish which patients would benefit from ED care. Among the patients with NSC our hypothesis is that it is the patients with serious conditions who are more likely benefited by ED care.

The proportion of patients with serious conditions among non-conveyed patients with NSC is not known. Therefore, the primary aim was to study the proportion of serious conditions among conveyed and non-conveyed patients presenting to the ambulance service with NSCs. The secondary aim was to study the mortality rates among patients with NSCs in relation to conveyance and serious and nonserious conditions.

Method

Study design

This is a retrospective cohort of adult patients (18 years and older) who presented to the ambulance service in the Stockholm Region, Sweden, with a chief complaint compatible with NSC between January 1st, 2013, and December 31st, 2013. This study forms part of a comprehensive study of patients presenting to the ambulance with NSC and builds upon previous research, which identified that one in three conveyed patients had serious conditions and experienced higher mortality rates compared to those without serious conditions [2]. The current study extends these findings by incorporating an analysis of non-conveyed patients.

Settings and population

Stockholm Region had a population of approximately 2.1 million (2013) and was responsible for operating the ambulance service. This service was provided by AISAB (Ambulanssjukvården I Storstockholm AB [Ambulance care in Greater Stockholm Ltd]) [5], along with two private companies [6, 7]. AISAB handled approximately 80 000 (42%) of all annual ambulance assignments in the region. All ambulances were staffed with a team consisting of a specialist nurse and an emergency medical technician (EMT). The entire Stockholm Region’s ambulance assignments were performed by 71 ambulances, 31 of which were operated by AISAB.

After the initial assessment by ambulance clinicians, patients are conveyed to the ED, directly admitted to a geriatric department following consultation with a geriatric consultant, bypassing the ED entirely or non-conveyed. In the Stockholm Region, non-conveyance guidelines for the ambulance service stipulate that patients are required to have normal vital signs to be eligible for a non-conveyance decision [8].

Study participants/ population

All patients who presented with NSCs to the ambulance service operated by AISAB throughout the study period. The inclusion criteria encompassed all adult patients aged 18 years and above with an ambulance service record containing a NSC in accordance with the ambulance record system (CAK-net criteria), within the Stockholm region. Exclusion criteria included specific chief complaints, duplicated records, referrals, or patients deceased during the ambulance mission.

Definitions

NSCs were defined according to Nemec et al. “all complaints that are not part of the set of specific complaints or signs or where an initial working diagnosis cannot be definitively established” [1], and adapted to the ambulance service in the studied setting. The criteria for NSC in the current study were one or more of the categories “decreased general health condition,” “general, unspecific,” or “undefined symptoms” according to the ambulance record system CAK-net. The ambulance service is mandated to select at least one of the predefined presenting complaints from a drop-down menu.

Nemec et al.‘s [1] definition of serious conditions for the ED was adapted for EMS by converting it to ICD-10 codes, excluding chronic and non-acute conditions. Neoplasms were considered serious only if the patient was admitted or died within 30 days, while infectious diseases were deemed serious if hospital admission was required., see Ivic et al. for details [2].

Non-conveyance is defined as the patient not being conveyed to the ED after the initial assessment by ambulance personnel. Ambulance service priority was documented as outbound or inbound. Outbound priority was decided by the emergency medical communication centre (EMCC), which in Sweden is separate from the ambulance service. Inbound priority was defined as the priority set by the ambulance service after index assessment.

Data collection

NSCs were identified with the CAK-net electronic health care record, used by the ambulance service [9]. CAK-net allows the registration of up to three primary assessment categories, and the first is mandatory. The non-specific categories were: “decreased general health condition,” “general, unspecific,” and “undefined symptoms”. The exclusion criteria were applied manually after the first acquisition of patients assessed as NSC (Fig. 1). Patient data collected from CAK-net were: age, sex, vital signs, Glasgow Coma Scale [GCS] score, National Advisory Committee for Aeronautics [NACA] score, and ambulance service disposition (conveyed to the emergency department [ED] or not). The following data were obtained from the National Patient Register at Sweden’s National Board of Health and Welfare, including ICD-10 codes at ED discharge, ED disposition (whether patients were discharged home or admitted to the hospital), in-hospital length of stay (LOS), ICD-10 codes from in-hospital discharge and ED contact after index ambulance service assessment for those non-conveyed., Mortality rates were collected from the Causes of Death register at the National Board of Health and Welfare. Both the National Patient Register and the Causes of Death register held by the Sweden’s National Board of Health and Welfare are mandated by Swedish law, ensuring that all caregivers submit comprehensive patient data.

Statistical analysis

Descriptive statistics was applied. For comparing groups, a Chi²⁻test was applied to categorical variables, while a Mann-Whitney U test was used for numerical variables. Logistic regression analysis was conducted to evaluate the association between conveyance/ non-coveyance and the primary outcome of a serious condition. Secondary outcomes, (24-hour and 30-day mortality rates), were assessed following the index assessment by the ambulance service, distinguishing between conveyance and non-conveyance groups. The results are expressed as odds ratios (OR) with 95% confidence intervals (CI). All statistical analyses were performed using IBM SPSS Statistics for Windows, version 25 (IBM Corp. Armonk, New York, USA).

Ethics approval and consent to participate

The study was approved by the Regional Ethics Committee in Stockholm, Sweden (reg.nr 2014/1999-31/4; 2016/1724-32.). The requirement for consent to participate was waived by the Regional Ethics Committee in Stockholm.

Results

A total of 4744 patients presenting to the ambulance service with NSCs were included (Fig. 1). The median age was 76 years of age. A serious condition was present in 1398/4744 (29.5%) of these patients. 1334/3780 (35.3%) (OR 7.02, CI 95% 5.37–9.18)of the patients had a serious condition in the conveyance group, compared to 64/964 (6.6%)(OR 0.14, CI 95% 0.11–0.19) in the non-conveyance group. (Tables 1 and 2). Non-conveyance had a Negative Predictive Value (NPV) of 93.4% (p < 0.001), accurately predicting the absence of a serious condition. The Positive Predictive Value (PPV) for conveyance identifying a serious condition was 35.3% (p < 0.001). All patients who were non-conveyed at the index ambulance service contact and had a serious condition were assessed in an ED within seven days.

Fig. 1

Flowchart over inclusion and exclusion in the study

MICU: Mobile Intensive Care Unit; Geriatric fast track: Region Stockholm regional guideline for direct admittance to geriatric departments bypassing the emergency department.

Table 1 Baseline characteristics for patients presenting with NSCs to the EMS grouped as conveyed or non-conveyed

The overall 24-hour mortality rate was 47/4744 (1.0%). 30-day mortality rate was 404/4744 (8.4%).

The 24-hour mortality rate in the conveyance group was 42/3780 (1.1%) compared to 5/964 (0.5%) in the non-conveyance group. The 30-day mortality rate in the conveyance group was 372/3780 (9.8%) compared to 32/964 (3.3%) in the non-conveyance group (Table 1). Among patients with serious conditions, the 30-day mortality rate was 269/1,334 (20.2%, OR 4.09, 95% CI 3.50–5.22) in the conveyed group, compared to 11/64 (17.2%, OR 6.22, 95% CI 2.81–13.78) in the non-conveyance group (Table 3). There were no statistically significant differences in sex, age or time of day when comparing serious conditions and mortality rates.

Table 2 Association between serious condition and conveyance

Table 3 Association between mortality rates in conveyed and non-conveyed patients grouped by the presence of serious conditions

Discussion

The results show that serious conditions were five times more common among conveyed as compared to non-conveyed patients. Mortality rates were three times higher in the conveyance group as compared to those non-conveyed. Non-conveyance was associated with lower presence of serious conditions and mortality rates.

The proportion of serious conditions was higher in the conveyance group, indicating that ambulance clinicians select patients with serious conditions after assessment, thus prioritizing those likely to need treatment in the ED. The results support that ambulance clinicians identify and convey most patients with a serious condition. However, although patients with serious conditions are conveyed, the majority of conveyed patients do not have serious conditions. It is however important to remember that reasons other than serious conditions will motivate conveyance to the ED. Therefore, it is important to always have a patient-centered approach.

While this study did not address care needs and care dependency, these factors are already known to influence hospitalization [10, 11]. The proportion of non-conveyed patients aligns with findings from previous studies in the Swedish context [12, 13]. Ambulance personnel’s ability to identify serious conditions is supported by previous research, which shows that experiential knowledge and clinical judgment are used alongside measured vital signs during clinical assessments [14,15,16]. However, it is notable that not all patients with serious conditions were conveyed after the initial assessment; nevertheless, all these patients were subsequently assessed in the ED within seven days. Similar findings were reported by Magnusson et al. [17], where most non-conveyed patients who were hospitalized within 72 h had presented with NSCs at the initial assessment. Overall, these results support that ambulance clinicians select patients in need of ED assessment and care, including of the majority of patients that present with serious conditions among patients presenting with NSCs.

The overall mortality rate was three times higher in the conveyance group as compared with the non-conveyance group. Mortality rates observed in the current study for conveyed patients were higher than those reported by Nemec et al. [1] and Karakoumis et al. [18], who both studied patients presenting with NSCs to the ED, however, without considering mode of arrival. Moreover, the prior studies [1, 18] were based on ED populations and as patients arriving to the ED by ambulance have in general higher mortality rates as compared with walk-in [19], hence, supporting the current results. The overall 30 day mortality rate in the non-conveyance group was more than twice that when compared to those reported by Heinonen et al. [20] who studied non-conveyance in a general population. When comparing mortality rates between the conveyance and non-conveyance groups for patients with serious conditions, the higher odds ratios in the non-conveyance group may be attributed to delays in receiving definitive care and diagnosis, potentially increasing mortality and compromising patient safety. However, it is also possible that some patient deaths were anticipated, a factor that was not addressed in the current study. Moreover, it is noteworthy that none of the non-conveyed patients with serious conditions died between the initial assessment by ambulance clinicians and the subsequent assessment in the ED within seven days.

Knowledge of non-conveyance decisions for patients presenting with NSCs is limited and uniform criteria are lacking [21]. The non-conveyance guidelines in Stockholm Region were limited to a few specific chief complaints and groups of patients, such as hypoglycemia and post-ictal patients [8]. For other chief complaints, such as NSCs the guideline stipulates that “prior to the decision of self-care, an evaluation must be made of the patient’s or significant other’s conditions to manage the selfcare safely” in addition to assessing normal vital signs [8]. Given that most patients in the current study were elderly and were required to have normal vital signs, the assessment may be further complicated by the fact that vital sign-based triage alone might be inadequate for identifying serious conditions in patients presenting with NSCs [2]. Furthermore, there is no universal consensus on acceptable under- or over-triage rates. Richards et al. [22] reported that an under-triage rate of 27% is reasonable, while Pointer et al. [23] considered 10% unacceptably high. Over-triage, which does not misclassify outcomes, is generally more accepted, though it may still impact patient safety [24, 25]. The findings in the current study are an indication that the assessment and decision for non-conveyance are based on variables beyond those measured in the current study, such as the environment the patient resides in when assessed by ambulance service, the patients’ wishes regarding their own care, information from next-of-kin or other caregivers, which is in concordance with previous studies [14, 16].

Limitations

The current study was limited by its retrospective study design. The electronic ambulance health records were not created for research purposes, but rather as local electronic health care records. The data retrieved from these records consisted of predefined variables that were imported into the pooled database used for analysis. The documentation of patients with NSCs by the ambulance service may have resulted in the exclusion of some patients from the data, as well as the inclusion of some patients with specific chief complaints. However, the current cohort is relatively large, and the results are therefore considered to be generalizable.

The validity of the results could be questioned as the data is from 2013. However, the current non-conveyance guidelines are the same as at the time of the study. Hence, there has been no major change in the management of patients with NSC and the decision to convey or not to the ED. Another limitation could be the overall increase in population, particularly among the elderly, leading to more patients presenting with NSCs. However, since the fundamental clinical practices and conveyance criteria have remained consistent, significant changes in the results are unlikely.

Conclusion

The results show that serious conditions were five times more common among conveyed as compared to non-conveyed patients. Mortality rates were three times higher in the conveyance group as compared to those non-conveyed. Non-conveyance was associated with lower presence of serious conditions and mortality rates. This implies that ambulance personnel demonstrate an ability to recognize patients with serious conditions, and assumed to be in need of treatment in the ED. Further research is needed on the clinical decision-making process regarding conveyance vs. non-conveyance when encountering patients presenting with NSCs.