a b s t r a c t

Objectives: Errors and adverse events associated with unexpected life-threatening events including unplanned transfer to the intensive care unit (ICU) and unexpected death after emergency department (ED) hospitalization are not well characterized. We performed this study to investigate the role of unexpected life-threatening events as a trigger to capture errors and adverse events for ED patient safety.

Methods: This prospective observational study enrolled adult non-trauma patients with unexpected life-threat- ening events within 24 h of general ward admission from the ED of a medical center in Taiwan. The period of study was one year (in 2013); the medical records of enrolled patients were reviewed to identify adverse events and errors. We measured the incidence rate of adverse events or errors. Preventability, type, and Physical injury severity of adverse events were investigated.

Results: Of 33,224 adult non-trauma Ward admissions from the ED, 100 admissions (0.3%) met the study criteria. Incidence rate was 2% and 15% for errors and adverse events, respectively. In admissions involving error, all were preventable and the error type was overlooked of severity. In admissions that involved adverse events, 93.3% were preventable. There were 20% of admissions that resulted in death and 60% developed with severe physical injury. The adverse event types were diagnosis issues (53.3%), Management issues (40%), and medication ad- verse events (6.7%).

Conclusions: Unexpected life-threatening events within 24 h of admission from the ED could be a useful trigger tool to identify preventable adverse events with serious physical injury in ED.

(C) 2016 The Authors. This is an open access article under the CC BY-NC-ND license

(http://creativecommons.org/licenses/by-nc-nd/4.0/).

Background and aims

Clinical deterioration leading to unexpected death or unplanned in- tensive care unit (ICU) admission can occur over the course of patient hospitalization on general wards. Among unplanned ICU transfer of general ward patients, higher in-hospital mortality and longer hospital length of stay was noted when compared to patients directly admitted to ICU [1]. Previous study using all hospital admissions as denominator revealed that 3.7% of all admissions had unplanned ICU transfer, these admissions accounted for 24.2% of all ICU admissions, 21.7% of all hospi- tal death [2,3]. Another research found that 5% of ED admissions experi- enced unplanned transfer to ICU from general ward, these patients accounted for about 25% of in-hospital deaths [1,3]. In some studies, a large portion of these unexpected life-threatening events were predict- able and preventable [4,5]. In about 70% of patients, clinical deteriora- tion was evident by Physiological parameters preceding the

* Corresponding author at: 123, Dabi Road, Niausung District, Kaohsiung, 83301, Taiwan.

E-mail address: [email protected] (W.-H. Lee).

unexpected events [6]. These are often not only explained by the natural course of the patient’s disease but can also result from dysfunctions or errors of the care system [7]. However, some previous studies revealed that the quality of care preceding unexpected events has been decreas- ing and is substandard due to subOptimal care and monitoring and neg- ligence [2,8-10]. According, improved recognition and response to changes in a patient’s condition was identified as a patient safety goal and Medical Emergency Team (MET) had been advocated to deal with these problems [7].

These unexpected life-threatening events of patients hospitalized via ED admission is a patient safety problem and may represent deficiencies in the care of these patients received in the ED. Previous studies found that about 1.6%-5% of those admitted to the general ward from the ED experienced unexpected death or Unplanned ICU admission [1,11]. Among unplanned ICU admissions, about 34% of transfers occurred within the first 24 h of admission and 19% of transfers were associated with errors in care [12]. The majority of these errors occurred in the ED.

Analysis of errors and adverse events can provide understanding of the medical care system in which the events occurred. Once we

http://dx.doi.org/10.1016/j.ajem.2016.11.062

0735-6757/(C) 2016 The Authors. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

can capture and measure the adverse events and errors, we can develop a system for improvement [5]. A reliable, feasible, and valid monitoring system to identify adverse events and errors can enhance patient safety in the ED. Commonly used patient safety measurement methods included reporting and trigger tool methods. A recent study showed trigger tool methods to be superior to reporting methods as active surveillance for ED patient safety [13]. As a trigger to capture adverse events or errors, we adopted the un- expected death and unplanned ICU transfer of ED admissions to the general medical ward within 24 h. We expected this trigger could help to identify adverse events or errors with severe patient safety problem or physical harm in the ED. The objective of this study was to investigate the incidence rate of this trigger to capture adverse events and errors of ED care. We also studied the preventability, type, and physical injury severity of the captured adverse events.

Methods

Study design and setting

We conducted a prospective observational study lasting for one year at an academic medical center in Southern Taiwan. All adult non-trauma patients older than 18 years of age who were admitted to the general medical ward from the ED between January 2013 and December 2013 were eligible for the study. All data was collected from the electronic health information system (HIS). This study was approved by the institutional review board of the hospital.

Selection of participants

We enrolled patients with life-threatening events (unplanned trans- fer to the ICU or unexpected death within 24 h of admission to the ward) from the HIS for the study. Patients with a terminal condition (terminal cancer, vital organ failure with Advance directive not to resus- citate) or transfer to ICU for post-surgical operational care were exclud- ed from this study.

Data collection and processing

The study nurses summarized the enrolled patient demographics, presenting illnesses, medical histories, descriptions, and investiga- tion results via medical records. Three certified Attending emergency physicians reviewed the summarized results and medical charts to identify adverse events and errors. An adverse event was defined as a physical injury or potential harm arising from medical services or interventions. An error was defined as the failure of a planned ac- tion to be completed as intended or the use of an incorrect plan for a specific aim. Some adverse events were attributable to errors. In this study, errors resulting in physical injury or potential harm were clas- sified as adverse events. Therefore, events classified as errors did not overlap with adverse events.

We adopted a widely used 6-point Likert scale to determine the reliability of investigators’ ability to identify errors and adverse events, with 1 and 6 indicating no evidence and strong evidence of an adverse event or error, respectively; if both reviewers scored the level of certainty N 4, the study incident was classified as an adverse event or error. If there was disagreement between the investigators, a third investigator reviewed the study incident; final determination of an adverse event or error was based on agreement between 2 out of the 3 investigators. inter-rater agreement of reviewer’s scoring was good for identification of adverse events. (Kappa statistic, 0.86) The investigators also classified the physical impact, type, and preventability of the identified adverse events and errors. If the two investigators disagreed on the classification of an adverse event, the final decision was made after discussion (Appendix 1).

Physical injury included definite physical injury or potential harm. If the adverse event would most likely cause physical injury–irrespective of whether definite harm or detectable harm was identified–we considered that adverse event to result in poten- tial harm. Physical injury severity classification comprised four clas- sifications by the degree of harm ranging from minimal to death. Minimal physical injury was defined as harm requiring little or no in- tervention. Potential harm or undetectable harm was classified as minimal physical injury. Moderate physical injury was defined as in- juries requiring hospitalization. Severe physical injury was defined as required intervention necessary to sustain life.

Types of adverse events or errors were classified into five categories including diagnostic issues, management issues, overlooked severity, procedural complications, and medication adverse effect. Diagnostic issues were defined by the fact that a new diagnosis could be established after more extensive evaluation that could have been performed before the patient left the ED. Management issues were defined as suboptimal management plans despite accurate diagnosis, or based on an inaccurate diagnosis. Overlooked severity was defined as the patient having been placed at an unnecessary risk of experiencing death or major disability. Medication adverse effect was defined as occurring when a patient experienced a symptom related to a medication regardless whether the medication was appropriately prescribed or taken. Procedural complication was defined as a patient experiencing adverse consequences of a procedure [7,14].

Analysis

For demographic and descriptive data, inter-group comparisons were made using Student’s t-test. Categorical variables were reported as numbers and percentages; inter-group comparisons were made using the ?² test or Fisher’s exact test, where appropriate. The level of significance was set at P b 0.05 (two-tailed). All analyses were per- formed using SAS(R) software (version 9.3, SAS Institute Inc., Cary, NC, USA).

Results

From January 1 to December 31, 2013, there were 134,484 adult non-trauma patients who visited the ED, and among them, 33,224 were admitted to the ward. There were 180 patients who died or were transferred to the ICU within 24 h of admission from the ED. Excluded from this study were 24 patients who were transferred to the ICU due to post-surgery care and 50 patients with terminal disease with advance directive. Therefore, 100 patients (0.30%) were enrolled in this study (Fig. 1). Among these 100 patients, 84 patients who were admitted to the general ward from the ED were transferred to the ICU within 24 h. The other 16 patients died within 24 h admission to the general ward from the ED. There were 15 adverse events and 2 errors found from the results of this study.

Table 1 compares the demographic characteristics of patients who experienced unexpected life-threatening events with and without adverse events or errors. There was no statistically significant difference between the two groups of patients except for the cause of unexpected life-threatening events. Overall, the patients were elderly and men predominated. In terms of triage category, most of them were of urgent or emergent need. However, patients without adverse events and error seemed to be older and have higher triage acuity than those patients with adverse events and errors. Among 21 patients who died within 7 days, 18 patients experienced no adverse events or errors, even with a greater mortality rate. The most common causes of unexpected life-threat- ening events were sepsis (35%) and respiratory failure (41.2%) for patients with adverse events or errors, which differed significantly from those patients without adverse events or errors.

Fig. 1. Patient flow for 134,484 emergency department visits, January 1-December 31, 2013.

The incidence rate for adverse events and errors were 15% and 2%, respectively (Table 2), of which over 93% were preventable. The most

Table 1 common type of adverse events was diagnostic issues and about 50% Comparison of demographic characteristics of 100 patients who experienced unexpected were related to missed or delayed diagnosis of sepsis. Management life-threatening events with and without adverse events or errors. issue was the second most common event type and all were related to

	Patients without adverse	Patient with adverse
Patient characteristics	events or errors (%) (n = 83)	events or errors (%) (n = 17)	P value

inadequate assessment. Two patients experienced errors and all had overlooked severity. We found one event with medication adverse ef- fect and no procedural complications. In terms of physical injury

Age	70.1 +- 15.5	67.8 +- 12.8	0.935
Female	36 (43.3)	6 (35.2)	0.539
Length of stay	26.2	29.33	0.978
Triage category
(resuscitation) (emergency) (urgent) Diagnostic categories	6 (7.2) 34 (40.9) 43 (51.8)	1 (5.9) 5 (29.4) 11 (64.7)	0.539
				Table 2 Characteristics of adverse events and errors of patient population for 100 patients with un- expected life-threatening events.
Neurology Chest Cardiology	33 (39.8) 19 (22.9) 9 (10.9)	5 (35.3) 6 (17.6) 3 (5.9)	0.731			Adverse events n = 15	Errors n = 2
Gastroenterology	10 (12.0)	1 (5.9)			Incidence rate (%)	15 (15)	2 (2)
Infection	10 (12.0)	1 (5.9)			Preventability (%)	14 (93.3)	2 (100)
Other	2 (2.4)	1 (5.9)			Type (%)
Cause of unexpected					Diagnostic issue	8 (53.3)	0
life-threatening					Management issue	6 (40)	0
events					Overlooked severity	0	2 (100)
Respiratory failure	41 (49.4)	7 (41.2)	0.027		Medication adverse effect	1 (6.7%)	0
AMI	11 (13.3)	2 (11.8)			Procedural complication	0	0
Sepsis	4 (4.8)	6 (35.3)			Physical injury severity (%)
GI bleeding	4 (4.8)	0			Death	3 (20)	0
Conscious change	23 (27.7)	2 (11.8)			Severe	9 (60%)	0
Mortality within 7 days	18 (21.7)	3 (17.6)	0.4992		Moderate	3 (20%)	0
Abbreviations: AMI, acute myocardial infarction; GI, gastroenterology.					Minimal	0	0

severity, three sentinel events with fatal outcome were noted. Most of the events resulted in severe physical injury (60%).

Discussion

Patients who experienced an unplanned transfer to the ICU or died within 24 h admission from the ED were described as life-threatening events and represent a potential risk to patient safety. In our study, we used patient admitted from the ED with unexpected life-threatening events within 24 h as trigger tool for patient safety of the ED.

We identified that 17% of ED adult non-trauma patients who had life-threatening events within 24 h of ward admission had been ex- posed to adverse events or errors related to their ED care. The overall rate of adverse event or error (0.01%) was exceedingly low. It should not be interpreted as the environment having had unusually safe ED care. In our study, the criteria for record review was highly specific screening and involved only 0.3% of all non-trauma admissions and 0.07% of all adult non-trauma ED visits. However, reviewing this small proportion of the study population had high incidence rate to capture adverse events or error. The incidence rate is the proportion of selected patients of a trigger that was truly associated with an adverse events or errors [15]. The median rate of adverse events of in-hospital patients is about 9% and ranges from 9%-12% among ED visits [13,16-18]. There- fore, for life-threatening events within 24 h, ward admission is a very promising trigger for patient safety of ED care due to its low flag rate and high incidence rate. We adopted the electronic method to identify life-threatening events as a trigger from the HIS. Other than manual medical record review of trigger tool methodologies, it is more effective and objective [15]. The electronic trigger tool system had been demon- strated effective in the prior study [13].

Triggers are clinical events indicating the possibility of adverse events or errors. Compared to routine medical record screening, by reviewing the records of patients with trigger events capture more trig- ger events or errors. LimitED triggers are reported in the prior studies for a patient safety surveillance system of ED [15]. Although there are few studies related to emergency medicine for comparison, our findings are similar to other studies. Previous study adopting combination of 5 triggers in ED revealed that the incidence rate of errors and adverse events in the study cohort was 17.8%. The trigger with unplanned ICU transfer demonstrated 10.2% of errors and 8.5% of adverse events rate [19]. A prospective cohort study revealed that trigger with ED on a re- turn visit was about 5.7% positive predictive rate (PPV) for adverse events [13]. The incidence of adverse events with unplanned ICU admis- sion in other research is around 5.7%-20% overall for the general ward population [4,14,20]. When we adopted the trigger combining un- planned ICU transfer and unexpected death in this study, the proportion of captured adverse events was greater than errors among the study co- hort. We believed that the patient safety problem of ED care comprises a substantial proportion of adverse events of early unplanned ICU admission from the general ward.

Besides high incidence rate, our study captures adverse events with unique characteristics. These adverse events are highly preventable, resulting in serious physical injury and most of them are related to issues in clinical judgment. This highlights a need for improved clinical assessments to reduce preventable adverse events with fatal or severe physical injury. Our data revealed that the proportion of adverse events with serious physical injury is greater than other studies adopting different triggers. Prior study with return ED visits as a trigger demonstrated that 58% of adverse events resulted in transient minor disability and 1% resulted in death [13]. Another study pertaining to patient safety analysis of ED patients with heart failure and chronic obstructive pulmonary disease found that 75% adverse events were with symptoms only and 8.3% resulted in death [21]. We found three sentinel events with mortality. It is essential that a trigger can capture adverse events with serious physical injury which warrant systemic investigations such as Root

Cause Analysis. Through the investigation process, we can identify factors that underlie variation in performance or that predispose an event toward undesired outcomes.

About 50% of adverse events are reported preventable by prior studies [13]. However, the preventability of adverse events captured in our study is much higher (93%). It may be because all of our adverse event types involved clinical assessment problems. These include diag- nostic error, inadequate assessment, and overlooked severity. We found that there are no adverse events due to procedure complications. The provision of ED care requires a high Level of knowledge, sound judg- ment, fast and accurate response to clinical situations, and the capacity for an extended period of vigilance. The most common clinical assess- ment problem was diagnostic issue (53.3%). Sepsis was the most com- mon disease (50%) with missed or delayed diagnosis. In addition, patients suffered from adverse events and errors had greater statistical- ly significant proportion of sepsis disease than patients without adverse events and errors. Need for improvement of clinical management of septic patients in ED is crucial.

The other cause of diagnostic error was Neurological problems such as acute stroke or brain tumor. Usually, clinical key findings that could be determined in the ED to correct diagnosis were missed. Inadequate assessment was the second cause (35.1%) of clinical assessment prob- lem in our study. Patients with definite diagnosis but with delayed or missed treatment have a high risk of patient safety and poor prognosis, even leading to mortality. The delayed or missED treatments for airway problems frequently cause unplanned ICU admission or death [7,8,12]. Several key findings were missed in ED medical records led to diagnos- tic error. Complete History taking and physical examination as well as good oral or paper communication will help to decrease diagnostic error. However, it is a challenge to practice in the ED because heavy workload and limited time for clinical assessment. The most common cause for unexpected life threatening events was respiratory failure in this study. Previous studies with unplanned ICU transfer reported simi- lar results [3,7]. Proper evaluation of patient respiratory status during patients’ stay at the ED will help to decrease treatment error.

There were 83 patients experiencing life-threatening events without adverse events or errors during their first day of admission and were usually more critically ill. Their triage categories are higher and 7-day mortality is greater. Among 17 patients with adverse events or errors, three patients experienced fatal sentinel events, and no mortality was noted with another 14 patients. We believed that the life-threatening events could be prevented if the adverse events or errors did not occur. Although the amount and incidence rate of adverse events iden- tified in our study is small, it can capture an important proportion among the spectrum of adverse events and error in ED. This is because this proportion of events resulted with serious physical impact and is usually preventable.

Limitation

Our study has a number of potential limitations. The data came from a single teaching hospital. The Medical errors were detected by review of the electronic health record. This likely resulted in an undercounting of adverse events. When evaluating adverse events through examining the health record, we are subject to hindsight and outcome biases. Final- ly, preventability assessments can be subjective.

Conclusion

In our study, the overall incidence rate of adverse events and errors is low. We have shown that medical review for patients with unexpect- ed life-threatening events within 24 h of admission from the ED could be a useful trigger tool to identify preventable adverse events with seri- ous physical injury.

Financial support

Nil.

Conflicts of interest

Nil.

Appendix 1. Example of adverse events.

			Physical
Summary	Classification	Preventability	impact
Patient visited ED due to syncope and	Diagnostic	Preventable	Severe
was admitted to medical general	issue
ward under the diagnosis of
hyperglycemia. Antibiotic agent was
not prescribed despite fever and
leukopenia noted at ED. Respiratory
distress and hypotension developed
in the general ward and the patient
was transferred to ICU due to
unstable condition.
Patient visited ED due to consciousness	Management	Preventable	Severe
change and fever and was admitted	issue
to medical general ward under the
impression of sepsis. Nasogastric
intubation was not performed
despite poor consciousness. Choking
episode developed and emergent
intubation was done. Patient was
transferred to ICU due to acute
respiratory failure.
Patient visited ED due to dyspnea.	Overlooked	Preventable	Minimal
Chronic obstructive pulmonary	severity
disease was diagnosed and he was
admitted to medical general ward
with bi-level positive airway pressure
Ventilator use. For patients with
impending respiratory failure, the
ward could not provide appropriate
monitoring. Patient was transferred
to ICU for further care.
Patient visited ED due to confused	Medication	Unpreventable	Severe
consciousness for days. Delirium and	adverse
acute cerebral vascular infarction was	effect
diagnosed. He was admitted to the
general ward. Sedative medication
was prescribed due to irritable state.
Apnea developed after medication
injection. He was intubated and
transferred to the ICU.

Abbreviations: ED, emergency department; ICU, intensive care unit.

References

1. Liu V, Kipnis P, Rizk NW, Escobar GJ. Adverse outcomes associated with delayed in- tensive care unit transfers in an integrated healthcare system. J Hosp Med 2012;7: 224-30.
2. Escobar GJ, Greene JD, Gardner MN, et al. Intra-hospital transfers to a higher level of care: contribution to total hospital and intensive care unit mortality and length of stay (LOS). J Hosp Med 2011;6:74-80.
3. Delgado MK, Liu V, Pines JM, et al. Risk factors for unplanned transfer to intensive care within 24 hours of admission from the emergency department in an integrated healthcare system. J Hosp Med 2013;8:13-9.
4. Zegers M, de Bruijne MC, Wagner C, et al. Adverse events and potentially prevent- able deaths in Dutch hospitals: results of a retrospective patient record review study. Qual Saf Health Care 2009;18:297-302.
5. Vlayen A, Verelst S, Bekkering GE, et al. Incidence and preventability of adverse events requiring intensive care admission: a systematic review. J Eval Clin Pract 2012;18:485-97.
6. Boulain T, Runge I, Delorme N, et al. Patients hospitalized in general wards via the emergency department: early identification of Predisposing factors for death or un- expected intensive care unit admission-A historical prospective. Emerg Med Int 2014;2014:203747.
7. Bapoje SR, Gaudiani JL, Narayanan V, et al. Unplanned transfers to a medical inten- sive care unit: causes and relationship to Preventable errors in care. J Hosp Med 2011;6:68-72.
8. Frost SA, Alexandrou E, Bogdanovski T, et al. Unplanned admission to intensive care after emergency hospitalisation: risk factors and development of a nomogram for individualising risk. Resuscitation 2009;80(2):224-30.
9. Smith T, Den Hartog D, Moerman T, et al. Accuracy of an expanded early warning score for patients in general and trauma surgery wards. Brit J Surg 2012;99: 192-7.
10. Ludikhuize J, Smorenbueg SM, de Rooij SE, et al. Identification of deterioration patients on general wards; measurement of Vital parameters an potential effectiveness of the modified early warning score. J Crit Care 2012;27: 424.e7-424.e13.
11. Petersen JA, Mackel R, Antonsen K, et al. serious adverse events in a hospital using early warning score – what went wrong? Resuscitation 2014;85:1699-703.
12. Gold CA, Mayer SA, Lennihan L, et al. Unplanned transfers from hospital wards to the neurological intensive care unit. Neurocrit Care 2015;23:159-65.
13. Calder L, Pozgay A, Riff S, et al. Adverse events in patients with return emergency de- partment visits. BMJ Qual Saf 2015;24:142-8.
14. Dunn KL, Reddy P, Moulden A, et al. Medical record review of deaths, unexpected in- tensive care unit admissions, and clinician referrals: detection of adverse events and insight into the system. Arch Dis Child 2006;91:169-72.
15. Naessens JM, O’Byrne TJ, Johnson MG, et al. Measuring hospital adverse events: assessing inter-rater reliability and trigger performance of the Global Trigger Tool. Int J Qual Health Care 2010;22:266-74.
16. Soop MFU, Koster M, et al. The incidence of adverse events in Swedish hospitals: a retrospective medical record review study. Int J Qual Health Care 2009;21: 285-91.
17. Vincent CNG, Woloshynowych M. Adverse events in British hospitals: preliminary retrospective record review. BMJ 2001;322:517-9.
18. Chern CH, How CK, Wang LM, et al. Decreasing clinically significant adverse events using feedback to emergency physicians of telephone follow-up outcomes. Ann Emerg Med 2005;45:15-23.
19. Klasco RSWR, Wong M, et al. Assessing the rates of error and adverse events in the ED. Am J Emerg Med 2015;33:1786-9.
20. Reese J, Deakyne SJ, Blanchard A, et al. Rate of preventable early unplanned intensive care unit transfer for direct admissions and emergency department admissions. Hosp Pediatr 2015;5:27-34.
21. Calder L, Tierney S, Jiang Y, et al. Patient safety analysis of the ED care of patients with heart failure and COPD exacerbations: a multicenter prospective cohort study. Am J Emerg Med 2014;32:29-35.