Patient safety is an essential component of health care quality, and reducing adverse events (AEs) a priority in health care delivery. The World Health Organization (WHO), the European council and the Ministry of Health of Spain urge adherence to the published recommendations pertaining to patient safety.1–3

In Spain, the ENEAS study4,5 found an incidence of AEs of 9.3%, of which 70% required additional treatment and 42.8% were considered preventable. When it comes to the emergency care setting, the evidence is scarce, although this is a setting in which the risk is potentially high.6 The EVADUR study,7 conducted in emergency departments in Spanish hospitals, found that AEs occurred in 12% of the patients, and the ERIDA study8 also found similar figures. Few studies have been conducted in paediatric emergency departments (PEDs).9,10 The main objective of our study was to characterise safety incidents in the PED setting.

We conducted a cross-sectional, observational and descriptive study in the PED of the Hospital Clínico Universitario Virgen de la Arrixaca, which serves a catchment population of 125000 children and has a large number of paediatric inpatient beds (150 in total) and managing a mean of 190 paediatric inpatients a day. The data collection method was based on the methodology applied in the ERIDA study8 and the local adverse events reporting system (SINASP).11

The sample included 207 patients; 3 were lost to follow-up at the time of the telephone call, leaving a valid sample of 204 patients (Table 1). The main reasons for visiting the PED are presented in Table 2 by category. When it came to the assigned triage level, 90 cases (44.1%) were classified as level 4, 100 (49.0%) as level 3 and 14 (6.9%) as level 2. A majority of children were managed by residents (69.6%), followed in frequency by adjunct paediatricians (22%), and the rest received care from other specialists.

In 76 cases (37.2%), the evaluation of the patient included at least one diagnostic test or procedure; in 30 cases (81.6% of this subset) a single diagnostic test or procedure. Plain radiographs were the tests ordered most frequently in the sample, in 12 cases (40%), followed by blood tests in 6 cases (20%). On the other hand, in 69 cases (33.8%) the patient received medication during the emergency care episode, with oral delivery in 65% of cases, inhaled in 20.3%, nebulised in 7.2% and via the intravenous (IV), muscular or rectal route in 7.1%.

No procedures were done in 83.3% of children, and the rest underwent a mean of 1 to 2 procedures, while 2 patients underwent 5 different procedures. The median length of stay in the PED was 21minutes (Q1,10; Q3, 41.8; IQR, 31.8). Data on timeliness of triage and care delivery in the sample can be found in Table 1. The discharge destination was the home in 179 cases (87.7%) and hospital admission in the remaining 25 patients (12.2%).

A better understanding of the reach and harm produced by safety incidents is essential for the purpose of developing safety interventions aimed at reducing preventable harm12 and funnelling resources toward necessary system improvements.10 The follow-up of patients during care delivery in the PED allowed the identification of safety incidents and/or AEs; this observation was reinforced by the review of health records and the interview with the legal guardian of the child.

More than half of the incidents were detected in the follow-up at 1 week, in agreement with previous studies that found that up to 50% of safety incidents are detected within 7 days from the end of the care episode.7,13,14 However, the possibility that other incidents occurred and were not detected or occurred past that 7-day window cannot be ruled out. Thus, a study conducted in a hospital in Ottawa (Canada), in which a follow-up phone call was made 3 weeks after the care episode,15 detected safety incidents within this time frame in 2.5% of the patients (95% CI, 1.8%–3.5%). The authors determined that, overall, adverse events were most likely to occur in the first week after the emergency department visit, and none occurred in the third week. The incidence and/or prevalence of AEs in the PED is not well established, as most studies in the literature have a retrospective design16 and were conducted through the review of medical records and reports or data from reporting systems,17–19 omitting any events or incidents that have not been thus documented. This aspect needs to be taken into account in future research, for, while many authors advocate for assessment at 7 days, there is no methodological consensus about the optimal time frames for follow-up.

In our study, we found that at least 12% of children managed in the PED experienced a care-related safety incident. This percentage was higher compared previous studies in which the reported incidence of harm was of approximately 9%.10,15,20 These discrepancies could be due to the methodological differences between our study and previous works, such as the direct observation of patients during the care episode in the PED and the follow-up at 7 days from PED discharge.

In our study, we found an overall incidence of safety incidents of 12.3%, which was very similar to the incidence found in the EVADUR and ERIDA studies, of 12% and 12.04%, respectively. The profile of children who experienced PED care-related safety incidents was characterised by a more severe reason for the visit (triage levels 2 and 3), requiring more diagnostic tests and procedures and prescription of oral treatment (in which there could be dose errors) or IV treatment (in which handing of peripheral catheters could lead to complications). We also ought to highlight the occurrence of safety incidents in children with musculoskeletal complaints who required immobilization and/or bandaging, who developed haematomas or haemorrhage secondary to procedures or who had moderate to severe earache and did not receive adequate pain relief at home.

In our study, the identified risk factors for safety incidents were previous history of a safety incident and performance of diagnostic tests, with the risk increasing with the number of tests, which was also the case of procedures: the more procedures the patient underwent, the more likely a safety incident. Discharge to inpatient care was also associated with an increased frequency of safety incidents. When it came to medication-related risk factors, the oral route was associated with the highest risk, although all such incidents were detected in time and did not harm the patient. Inhalation was the sole route of administration for which no incidents were detected.

We ought to underscore that nearly 80% of incidents were deemed preventable, of which half resulted from a clear error, based on the judgment of the evaluators. Our study corroborated the low frequency of incident reporting and documentation in health records and notification forms, as half the incidents were not recorded.

Although we did not find significant differences in the occurrence of safety incidents between children with and without chronic diseases, different studies have found an increased incidence in chronically ill children associated to the greater number of drugs and interventions that they require.21–23 In our study, only 2 children with chronic conditions experienced safety incidents, both during the care episode. The first one needed a CT scan that entailed insertion of a peripheral catheter for sedation, which turned out to be complicated, as 4 attempts were made, producing cutaneous lesions and haematomas. The other patient required admission for close monitoring and administration of IV antibiotherapy, and developed phlebitis during the stay due to inadequate care. A study about medical errors in paediatric inpatients with chronic conditions in the United States published in Pediatrics in 2012,21 which analysed data for 3739 hospitals and 3.1 discharge summaries, found an overall of medical error rate per 100 discharges of 3.0 (95% CI, 2.8–3.3), which increased to 5.3 (95% CI, 4.9–5.7) in children with chronic conditions compared to 1.3 (95% CI, 1.2–1.3) in children without chronic conditions.

The main limitation of our study is that we did not include patients with triage level 1 due to ethical/legal constraints, as these patients require urgent care and the parents/legal guardians would not be able to provide informed consent prior to inclusion. There may also have been some form of bias in data collection associated with the emergency care setting (high volume of patients with multiple rooms and care pathways).

As regards future lines of research, we want to emphasise the lack of studies regarding paediatric care in the emergency medical system (EMS), a care setting on which evidence is required and that may offer opportunities for improvement, as quality is an issue that spans the entire care delivery process, and not only hospital-based care.

Our study evinced that AEs are a problem in the PED setting the awareness of which needs to be increased to enable the implementation of improvement measures and prevent the occurrence of new incidents, including: improvement of communication between health care providers, reviewing analgesia protocols (implementing a stepwise approach to analgesia in children with moderate to severe pain), electronic prescription and double-checking prior to administration of medication.

This research did not receive any external funding.

The authors have no conflicts of interest to declare.