Abstract
Background:
The population that visits emergency departments frequently is heterogeneous and at high risk for mortality. This study aimed to characterize these patients in Ontario and Alberta, compare them with controls who do not visit emergency departments frequently, and identify subgroups.
Methods:
This was a retrospective cohort study that captured patients in Ontario or Alberta from fiscal years 2011/12 to 2015/16 in the Dynamic Cohort from the Canadian Institute for Health Information, which defined people with frequent visits to the emergency department in the top 10% of annual visits and randomly selected controls from the bottom 90%. We included patients 18 years of age or older and linked to emergency department, hospitalization, continuing care, home care and mental health–related hospitalization data. We characterized people who made frequent visits to the emergency department over time, compared them with controls and identified subgroups using cluster analysis. We examined emergency department visit acuity using the Canadian Triage and Acuity Scale.
Results:
The number of patients who made frequent visits to the emergency department ranged from 435 334 to 477 647 each year in Ontario (≥ 4 visits per year), and from 98 840 to 105 047 in Alberta (≥ 5 visits per year). The acuity of these visits increased over time. Those who made frequent visits to the emergency department were older and used more health care services than controls. We identified 4 subgroups of those who made frequent visits: “short duration” (frequent, regularly spaced visits), “older patients” (median ages 69 and 64 years in Ontario and Alberta, respectively; more comorbidities; and more admissions), “young mental health” (median ages 45 and 40 years in Ontario and Alberta, respectively; and common mental health–related and alcohol-related visits) and “injury” (increased prevalence of injury-related visits).
Interpretation:
From 2011/12 to 2015/16, people who visited emergency departments frequently had increasing visit acuity, had higher health care use than controls, and comprised distinct subgroups. Emergency departments should codevelop interventions with the identified subgroups to address patient needs.
People who present frequently to emergency departments are a minority that account for disproportionate health care spending:1 the highest 3% of this group comprise 30% of charges.2,3 They are also high users of other health care3–6 and are hospitalized and die more often than nonfrequent visitors to the emergency department,7,8 suggesting a need for interventions that optimize patient outcomes and service allocation.9 Effective interventions must recognize these patients’ clinical and demographic heterogeneity. Our previous work identified 4 subgroups among patients who presented frequently to emergency departments in British Columbia, including an older subgroup with prevalent cardiac-related conditions and a younger subgroup with mental health comorbidities, 10 corroborating other studies.11
There is an urgent need across Canada to identify subgroups among those who use emergency departments frequently, so that we can inform patient-focused, regionally specific interventions that could be nationally scalable where commonalities exist. We sought to test the generalizability of our BC-based findings and hypothesized that similar subgroups exist in other provinces. We aimed to characterize people who made frequent visits to the emergency department, compared to those who visited nonfrequently, and to identify subgroups in Ontario and Alberta.
Methods
Study design and setting
This was a retrospective administrative database study that captured patients who visited an emergency department in Ontario or Alberta between Apr. 1, 2011, and Mar. 31, 2016. Data were split into 5 fiscal years. For this study, we analyzed a combined data set from Ontario and Alberta, and we disaggregated data by province before analysis to facilitate interprovincial comparisons. We report study findings in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guideline.12
Participants
We analyzed annual subsets of patients aged 18 years or older who visited emergency departments most frequently (top 10%) and compared them to nonfrequent controls (remaining 90%) in Ontario and Alberta. We used the Canadian Institute for Health Information (CIHI) Dynamic Cohort of Complex, High System Users.
Data sources
CIHI created the Dynamic Cohort using in-house data sets to identify patients in the top 10% of acute care costs, lengths of stay, number of hospitalizations and number of emergency department visits each year.13
CIHI identified patients in the top 10% of emergency department visits using records submitted by Ontario and Alberta in the National Ambulatory Care Reporting System (NACRS).14 CIHI first stratified emergency department visit data by fiscal year, province and age (< 18 and ≥ 18 years). Within each stratum, CIHI identified the top 10% based on annual visit counts. CIHI also generated a control group by randomly selecting patients from the remaining 90%, using a 4:1 ratio.13 CIHI repeated the cohort selection process each fiscal year, adding new patients and updating information from previously included patients.
We used the “ED Visit Indicator” variable to differentiate emergency department visits from scheduled revisits. All emergency departments in Ontario and Alberta comply with level 3 NACRS reporting, which mandates that diagnoses are completed and reported using the International Statistical Classification of Diseases and Related Health Problems, 10th Revision, Canadian version (ICD-10-CA).15
CIHI performed all data linkages using personal health number. We linked NACRS records to the Discharge Abstract Database (DAD) for hospitalizations, and to the Continuing Care Reporting System, the Home Care Reporting System and the Hospital Mental Health Database (HMHDB).16–19 The HMHDB collates information on mental health–related admissions from 4 sources, depending on their availability in each jurisdiction: DAD, the Hospital Morbidity Database, the Hospital Mental Health Survey and the Ontario Mental Health Reporting System.18,20
Study variables and definitions
All study variables and their data sources are described in Appendix 1, Table S1, available at www.cmajopen.ca/content/10/1/E232/suppl/DC1.
Demographic characteristics
We examined sex, age, province and rural or urban residence using NACRS. A “0” in the second character of a postal code denoted a rural address.21,22
Emergency department visits
We summarized the characteristics of emergency department visits (ambulance arrival, triage level, diagnoses and disposition) in NACRS. Triage level was classified using the Canadian Triage and Acuity Scale (CTAS), a validated national tool with good inter-rater reliability; it specifies 5 acuity levels to assist emergency departments in prioritizing patient care.23–26
Diagnostic categories
Diagnoses in NACRS and DAD were classified using ICD-10-CA, which groups diagnoses into 22 chapters.27
The HMHDB reports diagnoses within mental health categories based on the ICD-10-CA for DAD, and the Diagnostic and Statistical Manual of Mental Disorders (DSM) for the Ontario Mental Health Reporting System (DSM-5) and the Hospital Mental Health Survey (DSM-III or DSM-IV-TR).19
We used a CIHI standard to quantify harms related to substance use in Canada,28 which we cross-referenced against published expert analyses29 to generate a list of ICD-10-CA codes that defined alcohol-related emergency department presentations pertaining to intoxication, withdrawal and complications (Appendix 1, Table S2).
Charlson Comorbidity Index
We calculated patients’ Charlson Comorbidity Index using primary emergency department diagnoses from NACRS. This index assesses 17 medical comorbidities and has predictive validity for mortality.30 Although it is usually based on hospitalization diagnoses, it has been validated using emergency department data.31–34
Statistical analysis
Index year and index visit for cluster analysis
We defined Apr. 1, 2013, to Mar. 31, 2014, as the index year for our cluster analysis, and the index visit as each patient’s last visit in that year. We used a 365-day period before the index visit to examine baseline characteristics. We chose our index year for consistency with our previous cluster analysis using BC data, and to facilitate comparison.10
Regularity index
We calculated regularity index for emergency department visits, to characterize the spacing between patients’ visits over the 365-day period before the index visit using the following equation: 1 ÷ (1 + variance of visits). Variance was based on the number of days between visits. This index ranged from 0 to 1 (closer to 1 indicated greater regularity).
To illustrate, a person who made 12 annual visits, 1 per month, would have an index close to 1; their index would be closer to 0 if they visited 12 times at more random intervals. The regularity index has been used in large cohort studies that examined temporal visit dispersion.10,35–38
Cluster analysis
We used cluster analysis to identify subgroups among people who visited emergency departments frequently.39 This well-described method organizes data into clusters by optimizing within-subgroup similarities and between-subgroup differences.10,40
For our clustering algorithm, we included variables pertaining to emergency department visit patterns and characteristics in NACRS. As is commonly done,41 we used our previous analyses5,10 and clinical experience to inform the inclusion of variables that would be clinically useful for emergency physicians.10,40 We excluded patients with missing information. We included 10 variables: (1) number of emergency department visits; (2) number of months that the patient visited an emergency department; ICD-10-CA emergency department diagnoses pertaining to (3) mental health, (4) circulatory, (5) respiratory or (6) digestive issues; (7) number of ICD-10-CA diagnostic chapters; (8) regularity index; (9) Charlson Comorbidity Index; and (10) age.
We employed a k-means algorithm and used the elbow plot and pseudo-F test as a guide to the appropriate cluster number.42,43 As is accepted in cluster analysis, we applied clinical experience to determine meaningful groupings.42,44 Four clusters optimally described our data with respect to statistical optimization and generating clinically meaningful subgroups (Appendix 1, Tables S3 and S4 and Figures S1 to S6).
We named each subgroup for ease of reference, based on observed patterns in demographics and emergency department use. We defined “short duration” as making emergency department visits over a median of 2 months or less, informed by previous related analyses.10
We described demographic characteristics and health care utilization using all available data sources from Apr. 1, 2011, to Mar. 31, 2016. We compared people who used emergency departments frequently to controls for the fiscal year from Apr. 1, 2015, to Mar. 31, 2016. We chose this year because it had the most recent data available, as well as for consistency (and to facilitate comparison) with our characterization of data in BC using the same fiscal year.5 As described above, we carried out cluster analysis to identify subgroups using the index year Apr. 1, 2013, to Mar. 31, 2014.
We performed all analyses using R (R Development Core Team, 2011).
Ethics approval
The University of British Columbia Clinical Research Ethics Board approved this study.
Results
From 2011/12 to 2015/16, the annual cohort of people who made frequent visits to the emergency department ranged from 435 334 to 477 647 in Ontario (median ≥ 4 visits per year), and 98 840 to 105 047 in Alberta (median ≥ 5 visits per year; Tables 1 and 2; Appendix 1, Tables S5 and S6). We observed increases from 2011/12 to 2015/16 in the proportion of visits that were triaged as resuscitation, emergent or urgent (CTAS 1–3; Ontario: 59.7% v. 67.4%; Alberta: 33.7% v. 43.9%); visits that involved arrival by ambulance (Ontario: 17.6% v. 20.3%; Alberta: 11.8% v. 14.3%); and visits that involved admission to hospital (Ontario: 14.5% v. 15.2%; Alberta: 10.4% v. 11.9%). Mental health–related hospitalizations related to substance use (including alcohol use) also increased from 2011/12 to 2015/16 (Ontario: 19.1% v. 20.5%; Alberta: 28.9% v. 36.4%).
Table 1:
Demographic and health care utilization characteristics of people who made frequent visits to the emergency department in Ontario
| Characteristic | Apr. 1, 2011–Mar. 31, 2012 | Apr. 1, 2012–Mar. 31, 2013 | Apr. 1, 2013–Mar. 31, 2014 | Apr. 1, 2014–Mar. 31, 2015 | Apr. 1, 2015–Mar. 31, 2016 |
|---|---|---|---|---|---|
| Emergency department visit characteristics (NACRS metadata) | |||||
| No. of patients | 435334 | 446954 | 451568 | 465949 | 477647 |
| No. of visits per patient, median (IQR) | 4 (3–5) | 4 (3–5) | 4 (3–5) | 4 (3–5) | 4 (3–5) |
| No. of emergency department visits | 2004975 | 2053609 | 2079281 | 2149965 | 2213161 |
| Arrival by ambulance, n (%) | |||||
| Air ambulance | 486 (0) | 422 (0) | 369 (0) | 459 (0) | 411 (0) |
| Air and ground ambulance | 978 (0) | 1119 (0.1) | 1054 (0.1) | 1283 (0.1) | 1315 (0.1) |
| Ground ambulance | 352171 (17.6) | 376214 (18.3) | 397811 (19.1) | 428368 (19.9) | 446762 (20.2) |
| No ambulance | 1651340 (82.4) | 1675854 (81.6) | 1680047 (80.8) | 1719855 (80.0) | 1764673 (79.7) |
| Triage level (CTAS), n (%) | |||||
| 1 (resuscitation) | 13123 (0.7) | 14229 (0.7) | 16599 (0.8) | 18759 (0.9) | 20796 (0.9) |
| 2 (emergent) | 332219 (16.6) | 361122 (17.6) | 402949 (19.4) | 435122 (20.2) | 458014 (20.7) |
| 3 (urgent) | 849844 (42.4) | 896673 (43.7) | 931616 (44.8) | 975875 (45.4) | 1012638 (45.8) |
| 4 (less urgent) | 648833 (32.4) | 634937 (30.9) | 597374 (28.7) | 588964 (27.4) | 593436 (26.8) |
| 5 (nonurgent) | 151140 (7.5) | 137432 (6.7) | 121391 (5.8) | 114988 (5.3) | 113565 (5.1) |
| Unknown | 8040 (0.4) | 7607 (0.4) | 7615 (0.4) | 14344 (0.7) | 12344 (0.6) |
| Not available | 1776 (0.1) | 1609 (0.1) | 1737 (0.1) | 1913 (0.1) | 2368 (0.1) |
| Top 5 ICD-10-CA primary diagnosis chapters, n (%) | |||||
| 1 | Respiratory 162000 (8.1) |
Respiratory 170897 (8.3) |
Respiratory 160762 (7.7) |
Respiratory 180688 (8.4) |
Respiratory 178433 (8.1) |
| 2 | Musculoskeletal 136297 (6.8) |
Musculoskeletal 137722 (6.7) |
Musculoskeletal 140076 (6.7) |
Musculoskeletal 146490 (6.8) |
Musculoskeletal 154688 (7.0) |
| 3 | Abnormal clinical findings 412 928 (20.6) |
Abnormal clinical findings 427536 (20.8) |
Abnormal clinical findings 442744 (21.3) |
Abnormal clinical findings 458415 (21.3) |
Abnormal clinical findings 473430 (21.4) |
| 4 | Injury and poisoning 285123 (14.2) |
Injury and poisoning 294076 (14.3) |
Injury and poisoning 300560 (14.5) |
Injury and poisoning 305651 (14.2) |
Injury and poisoning 320186 (14.5) |
| 5 | General health status 215204 (10.7) |
General health status 212930 (10.4) |
General health status 210348 (10.1) |
General health status 205890 (9.6) |
General health status 202791 (9.2) |
| Visit disposition, n (%) | |||||
| Discharged | 1620491 (80.8) | 1658247 (80.7) | 1679353 (80.8) | 1732802 (80.6) | 1778194 (80.4) |
| Transferred or admitted | 290321 (14.5) | 306710 (14.8) | 315618 (15.2) | 326049 (15.2) | 337594 (15.2) |
| Left against medical advice | 93003 (4.6) | 87501 (4.3) | 83169 (4) | 89909 (4.2) | 96255 (4.3) |
| Died | 1160 (0.1) | 1151 (0.1) | 1141 (0.1) | 1205 (0.1) | 1118 (0) |
| Hospitalization characteristics (DAD metadata) | |||||
| No. of admissions | 255202 | 268573 | 276307 | 284787 | 292411 |
| No. of admissions per patient, median (IQR) | 1 (1–2) | 1 (1–2) | 1 (1–2) | 1 (1–2) | 1 (1–2) |
| Top 5 ICD-10 primary diagnosis chapters simplified, n (%) | |||||
| 1 | Circulatory 45573 (17.9) |
Circulatory 46974 (17.5) |
Circulatory 48260 (17.5) |
Circulatory 49267 (17.3) |
Circulatory 50029 (17.1) |
| 2 | Respiratory 9491 (11.6) |
Respiratory 32452 (12.1) |
Respiratory 32177 (11.6) |
Respiratory 36479 (12.8) |
Respiratory 35749 (12.2) |
| 3 | Digestive 36263 (14.2) |
Digestive 37876 (14.1) |
Digestive 39235 (14.2) |
Digestive 38983 (13.7) |
Digestive 41150 (14.1) |
| 4 | Abnormal clinical findings 26679 (10.5) |
Abnormal clinical findings 27441 (10.2) |
Abnormal clinical findings 27767 (10.0) |
Abnormal clinical findings 27771 (9.8) |
Abnormal clinical findings 27636 (9.5) |
| 5 | Injury and poisoning 24054 (9.4) |
Injury and poisoning 25134 (9.4) |
Injury and poisoning 26302 (9.5) |
Injury and poisoning 26175 (9.2) |
Injury and poisoning 27506 (9.4) |
| Mental health hospitalization–related characteristics (HMHDB metadata) | |||||
| No. of mental health–related admissions | 37100 | 38282 | 39030 | 39913 | 42708 |
| No. of mental health–related admissions per patient, median (IQR) | 1 (1–2) | 1 (1–2) | 1 (1–2) | 1 (1–2) | 1 (1–2) |
| Diagnosis category, n (%) | |||||
| Mood disorder | 11513 (31.0) | 11563 (30.2) | 11148 (28.6) | 11565 (29.0) | 12038 (28.2) |
| Schizophrenic and psychotic disorder | 9355 (25.2) | 9362 (24.5) | 9797 (25.1) | 9008 (22.6) | 9192 (21.5) |
| Substance-related disorder | 7097 (19.1) | 7200 (18.8) | 7484 (19.2) | 7849 (19.7) | 8758 (20.5) |
| Organic disorder | 3628 (9.8) | 4275 (11.2) | 4615 (11.8) | 4974 (12.5) | 5721 (13.4) |
| Other mental health disorder | 2679 (7.2) | 2863 (7.5) | 2898 (7.4) | 2988 (7.5) | 3187 (7.5) |
| Anxiety disorder | 1412 (3.8) | 1480 (3.9) | 1467 (3.8) | 1647 (4.1) | 1661 (3.9) |
| Personality disorder | 1245 (3.4) | 1353 (3.5) | 1388 (3.6) | 1672 (4.2) | 1918 (4.5) |
| Non–mental health disorder | 125 (0.3) | 133 (0.3) | 175 (0.4) | 175 (0.4) | 183 (0.4) |
| Unknown disorder (not available) | 46 (0.1) | 53 (0.1) | 58 (0.1) | 35 (0.1) | 50 (0.1) |
Note: CTAS = Canadian Triage and Acuity Scale, DAD = Discharge Abstract Database, HMHDB = Hospital Mental Health Database, ICD-10-CA = International Statistical Classification of Diseases and Related Health Problems, 10th revision, Canadian version, IQR = interquartile range, NACRS = National Ambulatory Care Reporting System.
Table 2:
Demographic and health care utilization characteristics of people who made frequent visits to the emergency department in Alberta
| Characteristic | Apr. 1, 2011–Mar. 31, 2012 | Apr. 1, 2012–Mar. 31, 2013 | Apr. 1, 2013–Mar. 31, 2014 | Apr. 1, 2014–Mar 31, 2015 | Apr. 1, 2015–Mar. 31, 2016 |
|---|---|---|---|---|---|
| Emergency department visit characteristics (NACRS metadata) | |||||
| No. of patients | 98840 | 102781 | 103711 | 105047 | 102027 |
| No. of visits per patient, median (IQR) | 5 (4–7) | 5 (4–7) | 5 (4–7) | 5 (4–7) | 5 (4–7) |
| No. of emergency department visits | 680740 | 699962 | 700347 | 704256 | 690330 |
| Arrival by ambulance, n (%) | |||||
| Air ambulance | 433 (0.1) | 398 (0.1) | 403 (0.1) | 501 (0.1) | 443 (0.1) |
| Air and ground ambulance | 767 (0.1) | 766 (0.1) | 649 (0.1) | 650 (0.1) | 470 (0.1) |
| Ground ambulance | 79141 (11.6) | 85515 (12.2) | 89108 (12.7) | 94726 (13.5) | 97925 (14.2) |
| No ambulance | 600399 (88.2) | 613283 (87.6) | 610187 (87.1) | 608379 (86.4) | 591492 (85.7) |
| Triage level (CTAS), n (%) | |||||
| 1 (resuscitation) | 1649 (0.2) | 2195 (0.3) | 2570 (0.4) | 3153 (0.4) | 3109 (0.5) |
| 2 (emergent) | 50532 (7.4) | 57895 (8.3) | 63947 (9.1) | 70898 (10.1) | 77985 (11.3) |
| 3 (urgent) | 177861 (26.1) | 190429 (27.2) | 193700 (27.7) | 209718 (29.8) | 221905 (32.1) |
| 4 (less urgent) | 240461 (35.3) | 242305 (34.6) | 248299 (35.5) | 249633 (35.4) | 239741 (34.7) |
| 5 (nonurgent) | 159539 (23.4) | 159704 (22.8) | 153945 (22.0) | 134544 (19.1) | 116670 (16.9) |
| Unknown | 48740 (7.2) | 45577 (6.5) | 35774 (5.1) | 34405 (4.9) | 29364 (4.3) |
| Not available | 1958 (0.3) | 1857 (0.3) | 2112 (0.3) | 1905 (0.3) | 1556 (0.2) |
| Top 5 ICD-10-CA primary diagnosis chapters, n (%) | |||||
| 1 | General health status 176635 (25.9) |
General health status 174778 (25.0) |
General health status 170621 (24.4) |
General health status 162792 (23.1) |
General health status 155595 (22.5) |
| 2 | Abnormal clinical findings 97120 (14.3) |
Abnormal clinical findings 100856 (14.4) |
Abnormal clinical findings 104234 (14.9) |
Abnormal clinical findings 108528 (15.4) |
Abnormal clinical findings 109645 (15.9) |
| 3 | Injury and poisoning 83836 (12.3) |
Injury and poisoning 88125 (12.6) |
Injury and poisoning 89318 (12.8) |
Injury and poisoning 90696 (12.9) |
Injury and poisoning 88772 (12.9) |
| 4 | Respiratory 52425 (7.7) |
Respiratory 55897 (8.0) |
Respiratory 53191 (7.6) |
Respiratory 54796 (7.8) |
Respiratory 50195 (7.3) |
| 5 | Musculoskeletal 39632 (5.8) |
Musculoskeletal 40971 (5.9) |
Musculoskeletal 41019 (5.9) |
Digestive 40888 (5.8) |
Mental and behavioural 40297 (5.8) |
| Visit disposition, n (%) | |||||
| Discharged | 584541 (85.9) | 597069 (85.3) | 593365 (84.7) | 590099 (83.8) | 577751 (83.7) |
| Transferred or admitted | 71150 (10.4) | 75501 (10.8) | 77497 (11.0) | 80386 (11.5) | 82193 (11.9) |
| Left against medical advice | 24857 (3.7) | 27208 (3.9) | 29274 (4.2) | 33524 (4.8) | 30174 (4.4) |
| Died | 192 (0.0) | 184 (0.0) | 211 (0.0) | 247 (0.0) | 212 (0.0) |
| Hospitalization characteristics (DAD metadata) | |||||
| No. of admissions | 66843 | 70069 | 72127 | 73466 | 75014 |
| No. of admissions per patient, median (IQR) | 1 (1–2) | 1 (1–2) | 1 (1–2) | 1 (1–2) | 1 (1–2) |
| Top 5 ICD-10 primary diagnosis chapters simplified, n (%) | |||||
| 1 | Mental and behavioural 6736 (10.1) |
Mental and behavioural 7244 (10.3) |
Mental and behavioural 7824 (10.8) |
Mental and behavioural 8143 (11.1) |
Mental and behavioural 8608 (11.5) |
| 2 | Circulatory 8328 (12.5) |
Circulatory 8734 (12.5) |
Circulatory 8865 (12.3) |
Circulatory 9319 (12.7) |
Circulatory 9415 (12.6) |
| 3 | Respiratory 7196 (10.8) |
Respiratory 7906 (11.3) |
Respiratory 7957 (11.0) |
Respiratory 8778 (11.9) |
Respiratory 8679 (11.6) |
| 4 | Digestive 8945 (13.4) |
Digestive 9186 (13.1) |
Digestive 9534 (13.2) |
Digestive 9828 (13.4) |
Digestive 9910 (13.2) |
| 5 | Injury and poisoning 6966 (10.4) |
Injury and poisoning 7310 (10.4) |
Injury and poisoning 7556 (10.5) |
Injury and poisoning 7823 (10.6) |
Injury and poisoning 7904 (10.5) |
| Mental health hospitalization–related characteristics (HMHDB metadata) | |||||
| No. of mental health–related admissions | 7835 | 8393 | 9103 | 9292 | 9798 |
| No. of mental health–related admissions per patient, median (IQR) | 1 (1–2) | 1 (1–2) | 1 (1–2) | 1 (1–2) | 1 (1–2) |
| Diagnosis category, n (%) | |||||
| Mood disorder | 1806 (23.1) | 1751 (20.9) | 1823 (20) | 1789 (19.3) | 1854 (18.9) |
| Schizophrenic and psychotic disorder | 1335 (17.0) | 1434 (17.1) | 1512 (16.6) | 1294 (13.9) | 1297 (13.2) |
| Substance-related disorder | 2268 (28.9) | 2691 (32.1) | 3095 (34) | 3351 (36.1) | 3570 (36.4) |
| Organic disorder | 860 (11.0) | 875 (10.4) | 906 (10.0) | 919 (9.9) | 986 (10.1) |
| Other mental health disorder | 855 (10.9) | 909 (10.8) | 971 (10.7) | 1 122 (12.1) | 1142 (11.7) |
| Anxiety disorder | 353 (4.5) | 403 (4.8) | 427 (4.7) | 415 (4.5) | 459 (4.7) |
| Personality disorder | 337 (4.3) | 310 (3.7) | 342 (3.8) | 358 (3.9) | 446 (4.6) |
| Non–mental health disorder | 21 (0.3) | 20 (0.2) | 27 (0.3) | 44 (0.5) | 44 (0.4) |
Note: CTAS = Canadian Triage and Acuity Scale, DAD = Discharge Abstract Database, HMHDB = Hospital Mental Health Database, ICD-10-CA = International Statistical Classification of Diseases and Related Health Problems, 10th revision, Canadian version, IQR = interquartile range, NACRS = National Ambulatory Care Reporting System.
Frequent emergency department visitors versus controls
The group that made frequent emergency department visits (compared to nonfrequent controls) was older (Ontario: median age 52 yr v. 49 yr; Alberta: median age 46 yr v. 43 yr); had a higher proportion of females (Ontario: 55.4% v. 52.4%; Alberta: 55.3% v. 50.7%); more commonly lived in a rural location (Ontario: 20.5% v. 16.0%; Alberta: 34.5% v. 17.7%); arrived more commonly by ambulance (Ontario: 20.3% v. 14.8%; Alberta: 14.3% v. 11.9%); and were admitted to hospital more often (Ontario: 15.3% v. 10.5%; Alberta: 11.9% v. 11.0%; Table 3 and Appendix 1, Table S7).
Table 3:
Patient and health care utilization characteristics of people who make frequent emergency department visits and controls (Apr. 1, 2015, to Mar. 31, 2016), by province
| Characteristic | Ontario | Alberta | ||
|---|---|---|---|---|
|
|
|
|||
| Frequent emergency department visits | Controls | Frequent emergency department visits | Controls | |
| Patient characteristics (NACRS metadata) | ||||
|
| ||||
| No. of patients | 477647 | 1711848 | 102027 | 404111 |
|
| ||||
| Gender, n (%) | ||||
|
| ||||
| Female | 264731 (55.4) | 896281 (52.4) | 56417 (55.3) | 204986 (50.7) |
|
| ||||
| Male | 212908 (44.6) | 815550 (47.6) | 45610 (44.7) | 199125 (49.3) |
|
| ||||
| Other | 6 (0.0) | 9 (0.0) | 0 (0.0) | 0 (0.0) |
|
| ||||
| Undifferentiated | 2 (0.0) | 8 (0.0) | 0 (0.0) | 0 (0.0) |
|
| ||||
| Age, yr, median (IQR) | 52 (33–70) | 49 (33–64) | 46 (30–64) | 43 (30–59) |
|
| ||||
| Rural or urban, n (%) | ||||
|
| ||||
| Rural | 98098 (20.5) | 273212 (16.0) | 35178 (34.5) | 71432 (17.7) |
|
| ||||
| Urban | 377318 (79.0) | 1435081 (83.8) | 65429 (64.1) | 330834 (81.9) |
|
| ||||
| Not available | 2231 (0.5) | 3555 (0.2) | 1420 (1.4) | 1845 (0.4) |
|
| ||||
| No. of visits per patient, median (IQR) | 4 (3–5) | 1 (1–1) | 5 (4–7) | 1 (1–2) |
|
| ||||
| Emergency department visit characteristics (NACRS metadata) | ||||
|
| ||||
| Total number of emergency department visits | 2213161 | 2106899 | 690330 | 562922 |
|
| ||||
| Arrival by ambulance, n (%) | ||||
|
| ||||
| Air ambulance | 411 (0.0) | 236 (0.0) | 443 (0.1) | 349 (0.1) |
|
| ||||
| Air and ground ambulance | 1315 (0.1) | 591 (0.0) | 470 (0.1) | 165 (0.0) |
|
| ||||
| Ground ambulance | 446762 (20.2) | 310071 (14.7) | 97925 (14.2) | 66255 (11.8) |
|
| ||||
| No ambulance | 1764673 (79.7) | 1796001 (85.2) | 591492 (85.7) | 496153 (88.1) |
|
| ||||
| Triage level (CTAS), n (%) | ||||
|
| ||||
| 1 (resuscitation) | 20796 (0.9) | 19316 (0.9) | 3109 (0.5) | 3192 (0.6) |
|
| ||||
| 2 (emergent) | 458014 (20.7) | 418003 (19.8) | 77985 (11.3) | 81244 (14.4) |
|
| ||||
| 3 (urgent) | 1012638 (45.8) | 967356 (45.9) | 221905 (32.1) | 215251 (38.2) |
|
| ||||
| 4 (less urgent) | 593436 (26.8) | 638023 (30.3) | 239741 (34.7) | 206498 (36.7) |
|
| ||||
| 5 (nonurgent) | 113565 (5.1) | 60197 (2.9) | 116670 (16.9) | 46307 (8.2) |
|
| ||||
| Unknown | 12344 (0.6) | 2695 (0.1) | 29364 (4.3) | 9957 (1.8) |
|
| ||||
| Not available | 2368 (0.1) | 1309 (0.1) | 1556 (0.2) | 473 (0.1) |
|
| ||||
| Visit disposition, n (%) | ||||
|
| ||||
| Discharged | 1778194 (80.3) | 1818664 (86.3) | 577751 (83.7) | 479990 (85.3) |
|
| ||||
| Transferred or admitted | 337594 (15.3) | 221797 (10.5) | 82193 (11.9) | 61864 (11.0) |
|
| ||||
| Left against medical advice | 96255 (4.3) | 63833 (3.0) | 30174 (4.4) | 20513 (3.6) |
|
| ||||
| Died | 1118 (0.1) | 2605 (0.1) | 212 (0.0) | 555 (0.1) |
|
| ||||
| Hospitalization characteristics (DAD metadata) | ||||
|
| ||||
| No. of patients with at least 1 admission, (%) | 157965 (33.1) | 186498 (10.9) | 38464 (37.7) | 50348 (12.5) |
|
| ||||
| No. of admissions | 292411 | 211916 | 75014 | 59728 |
|
| ||||
| ICD-10 primary problem chapter among admissions, n (%) | ||||
|
| ||||
| Infectious diseases (I) | 13966 (4.8) | 8194 (3.9) | 2624 (3.5) | 1738 (2.9) |
|
| ||||
| Neoplasms (II and III) | 18315 (6.3) | 11092 (5.2) | 3439 (4.6) | 2778 (4.7) |
|
| ||||
| Endocrine (IV) | 12236 (4.2) | 5688 (2.7) | 3028 (4.0) | 1543 (2.6) |
|
| ||||
| Mental and behavioural (V) | 10187 (3.5) | 4202 (2.0) | 8608 (11.5) | 4754 (8.0) |
|
| ||||
| Neurologic (VI) | 6983 (2.4) | 4583 (2.2) | 1655 (2.2) | 1264 (2.1) |
|
| ||||
| Eye and ear (VII and VIII) | 836 (0.3) | 1114 (0.5) | 242 (0.3) | 385 (0.6) |
|
| ||||
| Circulatory (IX) | 50029 (17.1) | 45727 (21.6) | 9415 (12.6) | 9660 (16.2) |
|
| ||||
| Respiratory (X) | 35749 (12.2) | 20607 (9.7) | 8679 (11.6) | 5353 (9.0) |
|
| ||||
| Digestive (XI) | 41150 (14.1) | 34488 (16.3) | 9910 (13.2) | 8722 (14.6) |
|
| ||||
| Skin (XII) | 5120 (1.8) | 2015 (1.0) | 1653 (2.2) | 645 (1.1) |
|
| ||||
| Musculoskeletal (XIII) | 8113 (2.8) | 4757 (2.2) | 1975 (2.6) | 1352 (2.3) |
|
| ||||
| Genitourinary (XIV) | 19100 (6.5) | 10740 (5.1) | 4695 (6.3) | 3563 (6.0) |
|
| ||||
| Pregnancy (XV) | 5417 (1.9) | 6688 (3.2) | 2848 (3.8) | 2853 (4.8) |
|
| ||||
| Perinatal (XVI) | 0 (0.0) | 1 (0.0) | 2 (0.0) | 0 (0.0) |
|
| ||||
| Congenital (XVII) | 124 (0.0) | 110 (0.1) | 44 (0.1) | 47 (0.1) |
|
| ||||
| Abnormal clinical findings (XVIII) | 27636 (9.5) | 15969 (7.5) | 5523 (7.4) | 3374 (5.6) |
|
| ||||
| Injury and poisoning (XIX) | 27506 (9.4) | 27818 (13.1) | 7904 (10.5) | 9135 (15.3) |
|
| ||||
| General health status (XXI) | 9944 (3.4) | 8122 (3.8) | 2770 (3.7) | 2560 (4.3) |
|
| ||||
| Not available | 0 (0.0) | 1 (0.0) | 0 (0.0) | 2 (0.0) |
|
| ||||
| Alcohol-related diagnoses among admissions, n (%) | ||||
|
| ||||
| Yes | 4000 (1.4) | 1402 (0.7) | 2482 (3.3) | 771 (1.3) |
|
| ||||
| No | 288411 (98.6) | 210514 (99.3) | 72532 (96.7) | 58957 (98.7) |
|
| ||||
| Continuing care (CCRS metadata) | ||||
|
| ||||
| Number of patients with continuing care use, n (%) | 19512 (4.1) | 20307 (1.2) | 1140 (1.1) | 1730 (0.4) |
|
| ||||
| Home care (HCRS metadata) | ||||
|
| ||||
| Number of patients with home care use, n (%) | 91582 (19.2) | 111368 (6.5) | 15901 (15.6) | 20439 (5.1) |
|
| ||||
| Mental health hospitalization–related characteristics (HMHDB metadata) | ||||
|
| ||||
| No. of patients with at least 1 mental health–related admission (%) | 25555 (5.4) | 16048 (0.9) | 5913 (5.8) | 4677 (1.2) |
|
| ||||
| Homelessness status among patients with at least 1 mental health–related admission, n (%) | ||||
|
| ||||
| Home | 1246 (4.9) | 247 (1.5) | 338 (5.7) | 80 (1.7) |
|
| ||||
| No. of mental health–related admissions | 42708 | 18243 | 9798 | 5653 |
|
| ||||
| Diagnosis category among mental health–related admissions n (%) | ||||
|
| ||||
| Mood disorder | 12038 (28.2) | 5654 (31.0) | 1854 (18.9) | 1379 (24.4) |
|
| ||||
| Substance-related disorder | 8758 (20.5) | 2600 (14.3) | 3570 (36.4) | 1086 (19.2) |
|
| ||||
| Schizophrenic and psychotic disorder | 9192 (21.5) | 3973 (21.8) | 1297 (13.2) | 1118 (19.8) |
|
| ||||
| Organic disorder | 5721 (13.4) | 3518 (19.3) | 986 (10.1) | 1063 (18.8) |
|
| ||||
| Other mental health disorder | 3187 (7.5) | 1533 (8.4) | 1142 (11.7) | 658 (11.6) |
|
| ||||
| Anxiety disorder | 1661 (3.9) | 618 (3.4) | 459 (4.7) | 217 (3.8) |
|
| ||||
| Personality disorder | 1918 (4.5) | 234 (1.3) | 446 (4.6) | 88 (1.6) |
|
| ||||
| Non–mental health disorder | 183 (0.4) | 86 (0.5) | 44 (0.4) | 44 (0.8) |
|
| ||||
| Unknown disorder or not available | 50 (0.1) | 27 (0.1) | 0 (0.0) | 0 (0.0) |
Note: CCRS = Continuing Care Reporting System, CTAS = Canadian Triage and Acuity Scale, DAD = Discharge Abstract Database, HCRS = Home Care Reporting System, HMHDB = Hospital Mental Health Database, IQR = interquartile range, ICD-10-CA = International Statistical Classification of Diseases and Related Health Problems, 10th revision, Canadian version, NACRS = National Ambulatory Care Reporting System.
The proportion of people who made frequent emergency department visits that were triaged as resuscitation, emergent or urgent (CTAS 1–3) was higher in Ontario (67.4% v. 66.6%), but lower in Alberta (43.9% v. 53.2%) compared to nonfrequent controls. Those who made frequent emergency department visits had more episodes of continuing care (Ontario: 4.1% v. 1.2%; Alberta: 1.1% v. 0.4%), home care (Ontario: 19.2% v. 6.5%; Alberta: 15.6% v. 5.1%) and mental health admission (Ontario: 5.4% v. 0.9%; Alberta: 5.8% v. 1.2%) compared to controls; a high proportion of these were related to substance use (Ontario: 20.5% v. 14.3%; Alberta: 36.4% v. 19.2%).
Subgroups of frequent emergency department visitors
Our cluster analysis identified 4 subgroups that were similar in Ontario and Alberta (Tables 4 and 5; Appendix 1, Tables S8 and S9).
Table 4:
Cluster analysis and subgroup characterization among people who made frequent emergency department visits from Apr. 1, 2013, to Mar. 31, 2014 — Ontario
| Characteristic | Subgroup 1 (“Short duration”) | Subgroup 2 (“Older patients”) | Subgroup 3 (“Young mental health”) | Subgroup 4 (“Injury”) |
|---|---|---|---|---|
| Subgroup characteristics (clustering variables) | ||||
| No. of patients | 34116 | 74995 | 49167 | 292704 |
| Age, yr, median (IQR) | 49 (32–64) | 69 (54–80) | 45 (29–63) | 47 (30–65) |
| No. of visits to the emergency department, median (IQR) | 2 (1–3) | 3 (2–4) | 6 (4–8) | 2 (1–3) |
| Charlson Comorbidity Index, median (IQR) | 0 (0–0) | 1 (1–1) | 0 (0–0) | 0 (0–0) |
| No. of different discharge diagnosis chapters, median (IQR) | 2 (2–2) | 3 (2–4) | 5 (4–6) | 3 (2–3) |
| No. of months in the year that patients visited the emergency department, median (IQR) | 1 (1–2) | 3 (3–4) | 6 (5–7) | 3 (2–3) |
| Regularity index, median (IQR) | 1 (0.8–1) | 0.9 × 10−7 (0.6 × 10−7 to 6.8 × 10−5) | 1.3 × 10−7 (0.9 × 10−7 to 2.2 × 10−7) | 0.9 × 10−7 (0.5 × 10−7 to 1.2 × 10−4) |
| Patient characteristics (NACRS metadata) | ||||
| Gender, n (%) | ||||
| Female | 16875 (49.5) | 39575 (52.8) | 29136 (59.3) | 166034 (56.7) |
| Male | 17241 (50.5) | 35420 (47.2) | 20030 (40.7) | 126667 (43.3) |
| Other | 0 (0.0) | 0 (0.0) | 0 (0.0) | 1 (0.0) |
| Undifferentiated | 0 (0.0) | 0 (0.0) | 1 (0.0) | 1 (0.0) |
| Not available | 0 (0.0) | 0 (0.0) | 0 (0.0) | 1 (0.0) |
| Rural or urban, n (%) | ||||
| Rural | 7361 (21.6) | 15945 (21.3) | 10957 (22.3) | 58949 (20.1) |
| Urban | 26669 (78.2) | 58923 (78.6) | 37485 (76.2) | 232647 (79.5) |
| Emergency department visit characteristics (NACRS metadata) | ||||
| No. of emergency department visits | 113990 | 386011 | 563397 | 1187131 |
| Top 5 ICD-10 emergency department primary diagnoses, n (%) | ||||
| 1 | Drug therapies 13082 (11.5) |
COPD 23782 (4.2) |
Abdominal pain 23782 (4.2) |
Abdominal pain 46055 (3.9) |
| 2 | Abdominal pain 3944 (3.5) |
Asthma 15192 (2.7) |
Drug therapies 15192 (2.7) |
UTI 37335 (3.1) |
| 3 | Dressings 3151 (2.8) |
Pneumonia 15025 (2.7) |
UTI 15025 (2.7) |
Chest pain 31801 (2.7) |
| 4 | Cellulitis of the lower limb 3107 (2.7) |
Bronchitis 12552 (2.2) |
Chest pain 12552 (2.2) |
Drug therapies 26497 (2.2) |
| 5 | Cellulitis of the upper limb 2021 (1.8) |
CHF 10268 (1.8) |
Alcohol intoxication 10268 (1.8) |
Cellulitis of the lower limb 16941 (1.4) |
| Visit disposition, n (%) | ||||
| Discharged | 92928 (81.5) | 266750 (69.1) | 455419 (80.8) | 1003316 (84.5) |
| Left against medical advice | 4525 (4.0) | 8323 (2.2) | 32546 (5.8) | 43506 (3.7) |
| Admitted | 9544 (8.4) | 101667 (26.3) | 64298 (11.4) | 121653 (10.2) |
| Transferred to another facility | 6952 (6.1) | 8706 (2.3) | 10989 (2.0) | 18268 (1.5) |
| Died | 41 (0.0) | 565 (0.1) | 145 (0.0) | 388 (0.0) |
| Hospitalization characteristics (DAD metadata) | ||||
| No. of patients with at least 1 admission (%) | 6825 (20.0) | 45430 (60.6) | 21745 (44.2) | 72084 (24.6) |
| No. of admissions | 8078 | 100924 | 57359 | 110157 |
| Top 5 ICD-10-CA primary diagnosis chapters, n (%) | ||||
| 1 | Digestive 1574 (19.5) |
Circulatory 30001 (29.7) |
Digestive 12229 (21.3) |
Digestive 19370 (17.6) |
| 2 | Injury and poisoning 1351 (16.7) |
Respiratory 20441 (20.3) |
Abnormal clinical findings 6901 (12.0) |
Injury and poisoning 14330 (13.0) |
| 3 | Circulatory 1286 (15.9) |
Abnormal clinical findings 7836 (7.8) |
Injury and poisoning 5604 (9.8) |
Abnormal clinical findings 12961 (11.8) |
| 4 | Abnormal clinical findings 599 (7.4) |
Neoplasms and blood 7459 (7.4) |
Circulatory 4707 (8.2) |
Circulatory 12934 (11.7) |
| 5 | Genitourinary 560 (6.9) |
Digestive 6693 (6.6) |
Genitourinary 4615 (8.0) |
Genitourinary 9034 (8.2) |
| All alcohol-related diagnoses, n (%) | ||||
| No | 8027 (99.4) | 100643 (99.7) | 55351 (96.5) | 108918 (98.9) |
| Yes | 51 (0.6) | 281 (0.3) | 2008 (3.5) | 1239 (1.1) |
| Mental health hospitalization–related characteristics (HMHDB metadata) | ||||
| No. of patients with at least 1 admission (%) | 1103 (3.2) | 4955 (6.6) | 10048 (20.4) | 20031 (6.8) |
| No. of mental health–related admissions | 1926 | 8467 | 40 406 | 45 035 |
| Diagnosis category among mental health–related admissions, n (%) | ||||
| Mood disorder | 629 (32.7) | 1967 (23.2) | 10473 (25.9) | 14071 (31.2) |
| Schizophrenic and psychotic disorder | 524 (27.2) | 1291 (15.2) | 9852 (24.4) | 11582 (25.7) |
| Substance-related disorder | 337 (17.5) | 1150 (13.6) | 10686 (26.4) | 8766 (19.5) |
| Organic disorder | 177 (9.2) | 3097 (36.6) | 1379 (3.4) | 3606 (8.0) |
| Other mental health disorder | 137 (7.1) | 423 (5.0) | 3248 (8.0) | 3478 (7.7) |
| Anxiety disorder | 57 (3.0) | 344 (4.1) | 1582 (3.9) | 1772 (3.9) |
| Personality disorder | 51 (2.6) | 162 (1.9) | 2964 (7.3) | 1517 (3.4) |
| Non–mental health disorder | 11 (0.6) | 30 (0.4) | 177 (0.4) | 183 (0.4) |
| Unknown disorder | 3 (0.2) | 3 (0.0) | 45 (0.1) | 60 (0.1) |
Note: CHF = congestive heart failure, COPD = chronic obstructive pulmonary disease, DAD = Discharge Abstract Database, HMHDB = Hospital Mental Health Database, ICD-10-CA = International Statistical Classification of Diseases and Related Health Problems, 10th revision, Canadian version, IQR = Interquartile range, NACRS = National Ambulatory Care Reporting System, UTI = urinary tract infection.
Table 5:
Cluster analysis and subgroup characterization among people who made frequent emergency department visits from Apr. 1, 2013, to Mar. 31, 2014 — Alberta
| Characteristic | Subgroup 1 (“Short duration”) | Subgroup 2 (“Older patients”) | Subgroup 3 (“Young mental health”) | Subgroup 4 (“Injury”) |
|---|---|---|---|---|
| Subgroup characteristics (clustering variables) | ||||
| No. of patients | 4301 | 18776 | 12827 | 67722 |
| Age, yr, median (IQR) | 44 (31–58) | 64 (48–78) | 40 (28–55) | 40 (27–58) |
| No. of visits to the emergency department, median (IQR) | 3 (2–4) | 3 (2–5) | 7 (4–11) | 3 (2–4) |
| Charlson Comorbidity Index, median (IQR) | 0 (0–0) | 1 (1–1) | 0 (0–0) | 0 (0–0) |
| No. of different discharge diagnosis chapters, median (IQR) | 2 (2–2) | 4 (3–5) | 5 (5–6) | 3 (2–4) |
| No. of months in the year that patients visited the emergency department, median (IQR) | 1 (1–1) | 4 (3–5) | 7 (6–8) | 3 (2–4) |
| Regularity index, median (IQR) | 1 (0.8–1) | 1 × 10−7 (0.7 × 10−7 to 3.8 × 10−5) | 1.6 × 10−7 (1.2 × 10−7 to 2.6 × 10−7) | 0.9 × 10−7 (0.6 × 10−7 to −8.6 × 10−5) |
| Patient characteristics (NACRS metadata) | ||||
| Gender, n (%) | ||||
| Female | 1868 (43.4) | 9744 (51.9) | 7766 (60.5) | 37888 (55.9) |
| Male | 17241 (50.5) | 35420 (47.2) | 20030 (40.7) | 126667 (43.3) |
| Other | 0 (0.0) | 0 (0.0) | 0 (0.0) | 1 (0.0) |
| Undifferentiated | 0 (0.0) | 0 (0.0) | 1 (0.0) | 1 (0.0) |
| Not available | 0 (0.0) | 0 (0.0) | 0 (0.0) | 1 (0.0) |
| Rural or urban, n (%) | ||||
| Rural | 1428 (33.2) | 6720 (35.8) | 5526 (43.1) | 22378 (33.0) |
| Urban | 2850 (66.3) | 11921 (63.5) | 6846 (53.4) | 44610 (65.9) |
| Emergency department visit characteristics (NACRS metadata) | ||||
| No. of emergency department visits | 499395 | 1587307 | 139770 | 780967 |
| Top 5 ICD-10 emergency department primary diagnoses, n (%) | ||||
| 1 | Drug therapies 12356 (48.0) |
Drug therapies 9492 (7.0) |
Drug therapies 31466 (15.3) |
Drug therapies 38202 (9.9) |
| 2 | Dressings 1574 (6.1) |
COPD 4599 (3.4) |
Dressings 6016 (2.9) |
Dressings 12771 (3.3) |
| 3 | Other medical care 823 (3.2) |
Bronchitis 4241 (3.1) |
Abdominal pain 5650 (2.7) |
UTI 9159 (2.4) |
| 4 | Cellulitis of lower limb 623 (2.4) |
CHF 3919 (2.9) |
UTI 4044 (2.0) |
Abdominal pain 9075 (2.4) |
| 5 | Periapical abscess 372 (1.4) |
Asthma 3025 (2.2) |
Alcohol intoxication 4329 (2.1) |
Orthopedic care 8380 (2.2) |
| Visit disposition, n (%) | ||||
| Discharged | 23892 (92.9) | 102510 (76.0) | 177626 (86.3) | 334800 (87.0) |
| Left against medical advice | 409 (1.6) | 3306 (2.5) | 10731 (5.2) | 14663 (3.8) |
| Admitted | 762 (3.0) | 24873 (18.4) | 14082 (6.8) | 27476 (7.1) |
| Transferred to another facility | 654 (2.5) | 4090 (3.0) | 3376 (1.6) | 7861 (2.0) |
| Died | 2 (0.0) | 119 (0.1) | 37 (0.0) | 53 (0.0) |
| Hospitalization characteristics (DAD metadata) | ||||
| No. of patients with at least 1 admission (%) | 603 (14.0) | 10926 (58.2) | 6209 (48.4) | 18857 (27.8) |
| No. of admissions | 728 | 26281 | 15937 | 29826 |
| Top 5 ICD-10-CA primary diagnosis chapters, n (%) | ||||
| 1 | Digestive 125 (17.2) |
Circulatory 6145 (23.4) |
Mental and behavioural 3522 (22.1) |
Digestive 4581 (15.4) |
| 2 | Injury and poisoning 120 (16.5) |
Respiratory 5693 (21.7) |
Digestive 2933 (18.4) |
Injury and poisoning 4404 (14.8) |
| 3 | Genitourinary 73 (10) |
Abnormal clinical findings 2175 (8.3) |
Injury and poisoning 1655 (10.4) |
Mental and behavioural 3526 (11.8) |
| 4 | Pregnancy 71 (9.8) |
Digestive 2033 (7.7) |
Abnormal clinical findings 1347 (8.5) |
Abnormal clinical findings 2647 (8.9) |
| 5 | Circulatory 57 (7.8) |
Endocrine 1558 (5.9) |
Respiratory 906 (5.7) |
Genitourinary 2405 (8.1) |
| All alcohol-related diagnoses, n (%) | ||||
| No | 726 (99.7) | 26117 (99.4) | 14479 (90.9) | 28939 (97.0) |
| Yes | 2 (0.3) | 164 (0.6) | 1458 (9.1) | 887 (3.0) |
| Mental health hospitalization–related characteristics (HMHDB metadata) | ||||
| No. of patients with at least 1 admission (%) | 58 (1.3) | 1399 (7.5) | 2842 (22.2) | 4522 (6.7) |
| No. of mental health–related admissions | 85 | 2400 | 10 065 | 9439 |
| Diagnosis category among mental health–related admissions, n (%) | ||||
| Mood disorder | 18 (21.2) | 433 (18) | 1629 (16.2) | 2137 (22.6) |
| Schizophrenic and psychotic disorder | 19 (22.4) | 288 (12) | 1350 (13.4) | 1555 (16.5) |
| Substance-related disorder | 20 (23.5) | 536 (22.3) | 4583 (45.5) | 3236 (34.3) |
| Organic disorder | 9 (10.6) | 739 (30.8) | 252 (2.5) | 587 (6.2) |
| Other mental health disorder | 12 (14.1) | 210 (8.8) | 1110 (11.0) | 1144 (12.1) |
| Anxiety disorder | 4 (4.7) | 146 (6.1) | 407 (4.0) | 421 (4.5) |
| Personality disorder | 1 (1.2) | 43 (1.8) | 708 (7.0) | 322 (3.4) |
| Non–mental health disorder | 2 (2.4) | 5 (0.2) | 26 (0.3) | 37 (0.4) |
Note: CHF = congestive heart failure, COPD = chronic obstructive pulmonary disease, DAD = Discharge Abstract Database, HMHDB = Hospital Mental Health Database, ICD-10-CA = International Statistical Classification of Diseases and Related Health Problems, 10th revision, Canadian version, IQR = Interquartile range, NACRS = National Ambulatory Care Reporting System, UTI = urinary tract infection.
The “short duration” subgroup (Ontario: n = 34 116 [7.6%]; Alberta: n = 4301 [4.2%]) had median ages of 49 and 44 years, respectively; made a median number of 2 and 3 visits per year; and had regularly spaced visits. They commonly visited emergency departments for intravenous therapy (which could include antibiotics), dressings and cellulitis. Fewer patients were hospitalized in the index year for general hospitalizations (Ontario: 20.0%; Alberta: 14.0%) and mental health–related (Ontario: 3.2%; Alberta: 1.3%) than other subgroups.
The “older patients” subgroup (Ontario: n = 74 995 [16.6%]; Alberta: n = 8776 [18.1%]) had median ages of 69 and 64 years, respectively; made a median number of 3 visits per year; and had higher Charlson Comorbidity Index scores than the other subgroups. More were hospitalized at least once in the index year (Ontario: 60.6%; Alberta: 58.2%) than in the other subgroups, commonly for circulatory (Ontario: 29.7%; Alberta: 23.4%) and respiratory issues (Ontario: 20.3%; Alberta: 21.7%).
The “young mental health” subgroup (Ontario: n = 49 167 [10.9%]; Alberta: n = 12 827 [12.4%]) had median ages of 45 and 40 years, respectively; made a median number of 6 and 7 visits per year; were more commonly female (Ontario: 59.3%; Alberta: 60.5%); made more mental health–related visits (Ontario: 11.6%; Alberta: 10.0%); made more alcohol-related visits (Ontario: 3.5%; Alberta: 9.1%); and more commonly left the emergency department against medical advice (Ontario: 5.8%; Alberta: 5.2%) compared to other subgroups. This group had more mental health–related hospitalizations (Ontario: 20.4%; Alberta: 22.2%), among which diagnoses related to substance use were prevalent (Ontario: 26.4%; Alberta: 45.5%).
The “injury” subgroup (Ontario: n = 292 704 [64.9%]; Alberta: n = 67 722 [65.4%]) had median ages of 47 and 40 years, respectively; made a median number of 2 and 3 visits per year; and made more injury-related visits than the other subgroups (Ontario: 17.7%; Alberta: 15.9%).
Interpretation
Our study characterized those who made frequent visits to the emergency department in Ontario and Alberta using linked population-level administrative data and cluster analysis to identify clinically important subgroups. Our results indicated that visit acuity among these patients increased over time, and that they made high use of health care services compared to nonfrequent controls. We identified 4 subgroups with distinct demographic, clinical and visit patterns.
Our results denote important patterns that require further exploration. Increasing visit acuity suggests that people who use the emergency department frequently may be at growing risk for poor outcomes. These patients were more commonly admitted to hospital; however, although emergency department visits were of higher acuity in Ontario compared to nonfrequent controls, they were of lower acuity in Alberta, similar to previous analyses.45 This finding may indicate that social complexities (e.g., unstable housing or older patients failing to thrive in unsupported home environments) or lack of community follow-up to enable safe discharge may influence admission decisions.
Increases in substance use are likely to be multifactorial and may suggest improved identification, growing prevalence or increasing illicit substance toxicity, particularly in the early years of the opioid epidemic, which were captured by our data. Our findings were in alignment with existing literature that shows an increasing burden of frequent emergency department use over time, including rising clinical severity, substance use and poor outcomes.5,46,47
Repeated presentations from the subgroups we identified suggest that system-level gaps led to a failure to meet patients’ needs. The “short duration” subgroup may represent patients with visits related to an acute event that required a period of medical care (e.g., infection, injury). Although we used the NACRS “ED Visit Indicator” to exclude scheduled visits, a portion of these visits could still have been scheduled — for intravenous antibiotics, anticoagulation or wound care, for instance.
Our “older patient” subgroup had prevalent medical comorbidity and admissions, suggesting that supports are needed to avoid hospitalization (e.g., specialist clinics, home visits, improved primary care, chronic disease management and end-of-life care).
Similarly, our “young mental health” subgroup had very high numbers of emergency department visits, prevalent substance use and mental health–related hospitalizations, suggesting a need for immediate access to low-barrier treatment for substance use disorders, as well as psychosocial supports (e.g., outreach teams, peer-based violence prevention programs, supportive housing and managed care plans).48 Finally, our “injury” subgroup pointed to a possible role for individual- and population-level public health injury-prevention messaging.
Our findings were in alignment with literature that demonstrated heterogeneity among people who made frequent visits to the emergency department,46,49 and with our previous BC characterization, which identified nearly identical subgroups: short-term, with regularly spaced visits over a short period; older patients with multiple comorbidities; middle-aged patients with visits for mental health issues and alcohol use; and younger patients visiting emergency departments for mental health concerns.10,46,49
The comparability of our results strongly suggests generalizability across Canada, indicating that effective interventions could be nationally scaled. However, we lacked the data to determine whether the racial or ethnic composition of subgroups differed regionally. Barriers, stigma and discrimination affect health equity, access and the quality of care for many racialized groups,50 and follow-up research and interventions must consider these factors critically.
The existing literature focuses mostly on case management and care plans, targeting people who make frequent visits to the emergency department in aggregate, and has shown moderate effectiveness at decreasing repeat visits and potentially saving costs.9,51 Researchers, clinicians, emergency departments and policy-makers should undertake qualitative examination and collaborative engagement of subgroups of people who use emergency departments frequently so that they can better understand people’s reasons for high use and unmet needs. They should also codesign and pilot patient-centred interventions and referral pathways.
Limitations
Our analysis was limited by data availability. Variables such as employment and race or ethnicity were unavailable, and we could link only to CIHI’s data holdings, which did not include provincial pharmacy records, physician billing records and vital statistics. This restricted our ability to assess health care utilization, family physician attachment and mortality fully. Nonetheless, CIHI’s Dynamic Cohort is comprehensive, and it provided access to a built-in control cohort.
Our study was also limited by data quality (e.g., diagnostic coding), although this was mitigated by CIHI’s routine quality assurance. Moreover, Ontario and Alberta submit level 3 NACRS data, increasing data completeness.
The accuracy and reliability of the NACRS “ED Visit Indicator” to differentiate emergency department visits from scheduled returns were uncertain. Our cohort likely included patients with scheduled visits, but we had no reliable way of verifying this hypothesis. Therefore, we could not confirm and exclude suspected scheduled visits based on the data available.
Finally, because of delays in acquisition and linkage, our data were not current; the most recent available year was 2015/16. Still, although interim change is possible, the trends we identified remain relevant; for instance, substance use visits have likely increased further in the ongoing opioid epidemic.
Conclusion
People who use emergency departments frequently are making increasingly higher acuity visits and comprise distinct subgroups (“short duration,” “older patients,” “young mental health” and “injury”). Clinicians and policy-makers must engage with patients to codevelop and advocate for effective interventions (both in the emergency department and outside of it) to address heterogeneous patient-specific needs.
Supplementary Material
Footnotes
Competing interests: Jessica Moe has received grant funding from the Canadian Institutes of Health Research, Health Canada Substance Use and Addictions Program, Canadian Association of Emergency Physicians, Vancouver Coastal Health Research Institute, Vancouver Foundation, Vancouver Physician Staff Association, UBC Department of Family Practice, Vancouver General Hospital Complex Pain and Addictions Service, BC Centre for Disease Control Foundation for Public Health, and the UBC Faculty of Medicine. Margaret McGregor is a board member of the Vancouver Coastal Health Authority. Kathryn Dong has received grant funding from the Canadian Research Initiative in Substance Misuse, committee honoraria from the College of Physicians and Surgeons of Alberta and the Edmonton Zone Medical Staff Association, financial support from the Royal College of Physicians and Surgeons of Canada and the Canadian Association of Emergency Physicians, and a medical leadership salary from Alberta Health Services. No other competing interests were declared.
This article has been peer reviewed.
Contributors: Jessica Moe conceived the study, designed the analysis, obtained research funding, analyzed the data, interpreted results, and provided overall study oversight. Elle (Yuequiao) Wang designed the analysis, analyzed the data, created tables, and interpreted results. Margaret McGregor, Michael Schull, Kathryn Dong, Brian Holroyd, Eric Grafstein, Corinne Hohl and Johanna Trimble provided feedback on study design, data analysis, and results interpretation. Fiona O’Sullivan assisted with data analysis and table creation. Kimberlyn McGrail served as a methodological expert, designed the analysis, analyzed the data, and provided feedback on results interpretation. Jessica Moe drafted the manuscript and all authors contributed substantially to its revision. All authors have reviewed the final version, have provided final approval for publication, and agree to be accountable for all aspects of the work.
Funding: This study received funding from the Canadian Institutes of Health Research and the Canadian Association of Emergency Physicians.
Data sharing: We accessed our data through a data request to the Canadian Institute for Health Information (CIHI). Additional investigators can access the data analyzed in this study through an independent data request to CIHI.
Supplemental information: For reviewer comments and the original submission of this manuscript, please see www.cmajopen.ca/content/10/1/E232/suppl/DC1.
References
- 1.Pan-Canadian Forum on High Users of Health Care — summary report. Toronto: Canadian Institute for Health Information; 2014. [accessed 2021 Apr. 28]. Available: secure.cihi.ca/free_products/highusers_summary_report_revised_EN_web.pdf. [Google Scholar]
- 2.Johnson TL, Rinehart DJ, Durfee J, et al. For many patients who use large amounts of health care services, the need is intense yet temporary. Health Aff (Millwood) 2015;34:1312–9. doi: 10.1377/hlthaff.2014.1186. [DOI] [PubMed] [Google Scholar]
- 3.Mitchell MS, León CLK, Byrne TH, et al. Cost of health care utilization among homeless frequent emergency department users. Psychol Serv. 2017;14:193–202. doi: 10.1037/ser0000113. [DOI] [PubMed] [Google Scholar]
- 4.Krieg C, Hudon C, Chouinard M-C, et al. Individual predictors of frequent emergency department use: a scoping review. BMC Health Serv Res. 2016;16:594. doi: 10.1186/s12913-016-1852-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Moe J, O’Sullivan F, McGregor MJ, et al. Characteristics of frequent emergency department users in British Columbia, Canada: a retrospective analysis. CMAJ Open. 2021;9:E134–41. doi: 10.9778/cmajo.20200168. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Hansagi H, Olsson M, Sjöberg S, et al. Frequent use of the hospital emergency department is indicative of high use of other health care services. Ann Emerg Med. 2001;37:561–7. doi: 10.1067/mem.2001.111762. [DOI] [PubMed] [Google Scholar]
- 7.Moe J, Kirkland S, Ospina MB, et al. Mortality, admission rates and outpatient use among frequent users of emergency departments: a systematic review. Emerg Med J. 2016;33:230–6. doi: 10.1136/emermed-2014-204496. [DOI] [PubMed] [Google Scholar]
- 8.Fuda KK, Immekus R. Frequent users of Massachusetts emergency departments: a statewide analysis. Ann Emerg Med. 2006;48:9–16. doi: 10.1016/j.annemergmed.2006.03.001. [DOI] [PubMed] [Google Scholar]
- 9.Soril LJ, Leggett LE, Lorenzetti DL, et al. Reducing frequent visits to the emergency department: a systematic review of interventions. PLoS One. 2015;10:e0123660. doi: 10.1371/journal.pone.0123660. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Moe J, O’Sullivan F, McGregor MJ, et al. Identifying subgroups and risk among frequent emergency department users in British Columbia. J Am Coll Emerg Physicians Open. 2021;2:e12346. doi: 10.1002/emp2.12346. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Soril LJ, Leggett LE, Lorenzetti DL, et al. Characteristics of frequent users of the emergency department in the general adult population: a systematic review of international healthcare systems. Health Policy. 2016;120:452–61. doi: 10.1016/j.healthpol.2016.02.006. [DOI] [PubMed] [Google Scholar]
- 12.Von Elm E, Altman DG, Egger M, et al. Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. BMJ. 2007;335:806–8. doi: 10.1136/bmj.39335.541782.AD. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Dynamic cohort of complex, high system us 2011–2015. Ottawa: Canadian Institutes of Health Research; 2017. [accessed 2021 Apr. 24]. Available: cihr-irsc.gc.ca/e/50129.html#section_1. [Google Scholar]
- 14.National Ambulatory Care Reporting System metadata (NACRS) Ottawa: Canadian Institute for Health Information; 2021. [accessed 2021 Apr. 24]. Available: www.cihi.ca/en/national-ambulatory-care-reporting-system-metadata-nacrs. [Google Scholar]
- 15.Data quality documentation, National Ambulatory Care Reporting System. Ottawa: Canadian Institute for Health Information; 2021. [accessed 2021 Apr. 28]. Available: www.cihi.ca/sites/default/files/document/nacrs-data-quality-current-year-information-2019-2020-en.pdf. [Google Scholar]
- 16.Discharge Abstract Database metadata (DAD) Ottawa: Canadian Institute for Health Information; 2021. [accessed 2021 Apr. 28]. Available: www.cihi.ca/en/discharge-abstract-database-metadata-dad. [Google Scholar]
- 17.Continuing Care metadata. Ottawa: Canadian Institute for Health Information; 2021. [accessed 2021 May 13]. Available: www.cihi.ca/en/continuing-care-metadata. [Google Scholar]
- 18.Home Care Reporting System metadata. Ottawa: Canadian Institute for Health Information; 2021. [accessed 2021 May 13]. Available: www.cihi.ca/en/home-care-reporting-system-metadata. [Google Scholar]
- 19.Hospital Mental Health Database metadata (HMHDB) Ottawa: Canadian Institute for Health Information; 2021. [accessed 2021 May 13]. Available: www.cihi.ca/en/hospital-mental-health-database-metadata-hmhdb. [Google Scholar]
- 20.Ontario Mental Health Reporting System metadata. Ottawa: Canadian Institute for Health Information; 2021. [accessed 2021 Apr. 28]. Available: www.cihi.ca/en/ontario-mental-health-reporting-system-metadata. [Google Scholar]
- 21.Postal code conversion file. Ottawa: Statistics Canada; 2021. [accessed 2021 Apr. 28]. Available: www150.statcan.gc.ca/n1/pub/92-154-g/92-154-g2017001-eng.htm. [Google Scholar]
- 22.Rural and small town Canada: analysis bulletin. Ottawa: Statistics Canada; 2001. [accessed 2021 Apr. 24]. Cat no 21-006-XIE. Available: www150.statcan.gc.ca/n1/pub/21-006-x/21-006-x2001003-eng.pdf. [Google Scholar]
- 23.Lee JY, Oh SH, Peck EH, et al. The validity of the Canadian Triage and Acuity Scale in predicting resource utilization and the need for immediate life-saving interventions in elderly emergency department patients. Scand J Trauma Resusc Emerg Med. 2011;19:68. doi: 10.1186/1757-7241-19-68. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Kuriyama A, Kaihara T, Ikegami T. Validity of the Japan Acuity and Triage Scale in elderly patients: a cohort study. Am J Emerg Med. 2019;37:2159–64. doi: 10.1016/j.ajem.2019.03.006. [DOI] [PubMed] [Google Scholar]
- 25.Mirhaghi A, Heydari A, Mazlom R, et al. The reliability of the Canadian Triage and Acuity Scale: meta-analysis. N Am J Med Sci. 2015;7:299–305. doi: 10.4103/1947-2714.161243. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.CTAS-ETG: Canadian Triage and Acuity Scale. Ottawa: CTAS National Working Group; 2016. [accessed 2021 Apr. 24]. Available: ctas-phctas.ca/?page_id=294. [Google Scholar]
- 27.International statistical classification of diseases and related health problems, 10th revision. Geneva: World Health Organization; 1992. [accessed 2021 Oct. 27]. Available: icd.who.int/browse10/2019/en. [Google Scholar]
- 28.Hospital stays for harm caused by substance use — appendices to indicator library. Ottawa: Canadian Institute for Health Information; 2020. [Google Scholar]
- 29.Saunders JB, Room R. Enhancing the ICD system in recording alcohol’s involvement in disease and injury. Alcohol Alcohol. 2012;47:216–8. doi: 10.1093/alcalc/ags024. [DOI] [PubMed] [Google Scholar]
- 30.Sundararajan V, Henderson T, Perry C, et al. New ICD-10 version of the Charlson Comorbidity Index predicted in-hospital mortality. J Clin Epidemiol. 2004;57:1288–94. doi: 10.1016/j.jclinepi.2004.03.012. [DOI] [PubMed] [Google Scholar]
- 31.Needham DM, Scales DC, Laupacis A, et al. A systematic review of the Charlson Comorbidity Index using Canadian administrative databases: a perspective on risk adjustment in critical care research. J Crit Care. 2005;20:12–9. doi: 10.1016/j.jcrc.2004.09.007. [DOI] [PubMed] [Google Scholar]
- 32.Murray SB, Bates DW, Ngo L, et al. Charlson Index is associated with one-year mortality in emergency department patients with suspected infection. Acad Emerg Med. 2006;13:530–6. doi: 10.1197/j.aem.2005.11.084. [DOI] [PubMed] [Google Scholar]
- 33.Olsson T, Terent A, Lind L. Charlson Comorbidity Index can add prognostic information to rapid emergency medicine score as a predictor of long-term mortality. Eur J Emerg Med. 2005;12:220–4. doi: 10.1097/00063110-200510000-00004. [DOI] [PubMed] [Google Scholar]
- 34.Charlson ME, Pompei P, Ales KL, et al. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40:373–83. doi: 10.1016/0021-9681(87)90171-8. [DOI] [PubMed] [Google Scholar]
- 35.Gibson DA, Moorin RE, Preen D, et al. Enhanced primary care improves GP service regularity in older patients without impacting on service frequency. Aust J Prim Health. 2012;18:295–303. doi: 10.1071/PY11050. [DOI] [PubMed] [Google Scholar]
- 36.Youens D, Harris M, Robinson S, et al. Regularity of contact with GPs: measurement approaches to improve valid associations with hospitalization. Fam Pract. 2019;36:650–6. doi: 10.1093/fampra/cmz002. [DOI] [PubMed] [Google Scholar]
- 37.Moorin RE, Youens D, Preen DB, et al. The association between general practitioner regularity of care and ‘high use’. BMC Health Serv Res. 2020;20:915. doi: 10.1186/s12913-020-05718-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Moorin RE, Youens D, Preen DB, et al. Association between continuity of provider-adjusted regularity of general practitioner contact and unplanned diabetes-related hospitalisation: a data linkage study in New South Wales, Australia, using the 45 and up study cohort. BMJ Open. 2019;9:e027158. doi: 10.1136/bmjopen-2018-027158. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Jain AK. Data clustering: 50 years beyond K-means. Pattern Recognit Lett. 2010;31:651–66. [Google Scholar]
- 40.Chapter 445: hierarchical clustering/dendrograms. Kaysville (UT): NCSS Statistical Software; 2020. [accessed 2021 May 13]. Available: ncss-wpengine.netdna-ssl.com/wp-content/themes/ncss/pdf/Procedures/NCSS/Hierarchical_Clustering-Dendrograms.pdf. [Google Scholar]
- 41.Haldar P, Pavord ID, Shaw DE, et al. Cluster analysis and clinical asthma phenotypes. Am J Respir Crit Care Med. 2008;178:218–24. doi: 10.1164/rccm.200711-1754OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Calinski T, Harabasz J. A dendrite method for cluster analysis. Commun Stat. 1974;3:1–27. [Google Scholar]
- 43.Sarstedt M, Mooi E. A concise guide to market research. 2nd ed. Heidelberg (Berlin): Springer-Verlag; 2014. [Google Scholar]
- 44.Androniceanu AGI, Georgescu I, Kinnunen J Digitalization clusters within the European Union. Proceedings from Conference: The International Business Information Management Conference (33rd IBIMA); 2019 Apr 10–11; Granada, Spain. King of Prussia (PA): IBIMA Publishing; 2019. [Google Scholar]
- 45.Chen A, Fielding S, Hu XJ, et al. Frequent users of emergency departments and patient flow in Alberta and Ontario, Canada: an administrative data study. BMC Health Serv Res. 2020;20:938. doi: 10.1186/s12913-020-05774-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.LaCalle E, Rabin E. Frequent users of emergency departments: the myths, the data, and the policy implications. Ann Emerg Med. 2010;56:42–8. doi: 10.1016/j.annemergmed.2010.01.032. [DOI] [PubMed] [Google Scholar]
- 47.Hulme J, Sheikh H, Xie E, et al. Mortality among patients with frequent emergency department use for alcohol-related reasons in Ontario: a population-based cohort study. CMAJ. 2020;192:E1522–E31. doi: 10.1503/cmaj.191730. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Snider CE, Jiang D, Logsetty S, et al. Feasibility and efficacy of a hospital-based violence intervention program on reducing repeat violent injury in youth: a randomized control trial. CJEM. 2020;22:313–20. doi: 10.1017/cem.2019.406. [DOI] [PubMed] [Google Scholar]
- 49.Bertoli-Avella AM, Haagsma JA, Van Tiel S, et al. Frequent users of the emergency department services in the largest academic hospital in the Netherlands: a 5-year report. Eur J Emerg Med. 2017;24:130–5. doi: 10.1097/MEJ.0000000000000314. [DOI] [PubMed] [Google Scholar]
- 50.Hausmann LRM, Jones AL, McInnes SE, et al. Identifying healthcare experiences associated with perceptions of racial/ethnic discrimination among veterans with pain: a cross-sectional mixed methods survey. PLoS One. 2020;15:e0237650. doi: 10.1371/journal.pone.0237650. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51.Moe J, Kirkland SW, Rawe E, et al. Effectiveness of interventions to decrease emergency department visits by adult frequent users: a systematic review. Acad Emerg Med. 2017;24:40–52. doi: 10.1111/acem.13060. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.