Predicting Hospital Admission among High Acuity Triaged Patients Transported to the Emergency Department in Ontario, Canada: A Population-Based Cohort Study using Machine Learning

doi:10.21203/rs.3.rs-1726114/v1

Download PDF

Research Article

Predicting Hospital Admission among High Acuity Triaged Patients Transported to the Emergency Department in Ontario, Canada: A Population-Based Cohort Study using Machine Learning

https://doi.org/10.21203/rs.3.rs-1726114/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background

Paramedics are mandated to transport emergently triaged patients to the closest emergency department (ED). The closest ED may not be the optimal transport destination if further distanced ED’s can provide specialized care or are less crowded. Machine learning may support paramedic decision-making to transport a specific subgroup of emergently triaged patients that are unlikely to require hospital admission or emergency care to a more appropriate ED. We examined whether prehospital patient characteristics known to paramedics were predictive of hospital admission.

Methods

We conducted a retrospective cohort study using machine learning algorithms to analyze ED visits of the National Ambulatory Care Reporting System from Jan 1, 2018 to Dec 31, 2019 in Ontario, Canada. We included all adult (≥ 18 years) paramedic transports to the ED who had an emergent Canadian Triage Acuity Scale score (CTAS 2). Eight prehospital characteristic classes known to paramedics were used. We applied four machine learning algorithms that were trained and assessed using 10-fold cross-validation to predict the ED visit disposition of admission to hospital or discharged from ED. Predictive model performance was determined using the area under the receiving operating characteristic curve (AUC) with 95% confidence intervals and probabilistic accuracy using the Brier Scaled score. Variable importance scores were computed to determine the top 10 predictors of hospital admission. We also reported sensitivity, specificity, and positive and negative predictive values to support performance interpretation.

Results

All machine learning algorithms performed similarly for the prediction of which ED patient visits would be admitted to hospital (AUC 0.77–0.78, Brier Scaled 0.22–0.24). The characteristics most predictive of admission included age 65 to 105 years, referral source from a residential care facility, presenting with a respiratory complaint, and receiving home care.

Conclusions

Machine learning algorithms performed well in predicting ED visit dispositions using a comprehensive list of prehospital patient characteristics. To the best of our knowledge, this study is the first to utilize machine learning to predict ED visit outcomes from patient characteristics known prior to paramedic transport. This study has potential to inform paramedic regulations regarding the distribution of emergently triaged patients.

Machine learning

paramedic

prehospital

health services research

emergency department

prediction model

electronic health record

The Canadian Triage and Acuity Scale (CTAS) is used by Ontario paramedics to delineate medical acuity and patient care needs in the prehospital setting.(1, 2) In Ontario, patients assigned a prehospital ‘emergent’ CTAS score require time-sensitive emergency medicine, for which paramedics are mandated by legislation to transport to the most proximal emergency department (ED).(1) Empirically, the majority of emergent triaged paramedic transports do not receive emergency medicine in the ED, do not clinically deteriorate, are discharged home directly from the ED (56.6%) and a sizeable proportion receive physician interventions considered primary care-like (39.3%).(3, 4) Paramedics transporting high-acuity triaged patients exclusively to the nearest ED constitutes an operational inefficiency; all emergently triaged patients are handled the same in the prehospital setting, independent of patient repatriation for continuity of care, patient preference, ED workload, staffing, or bed availability.(1) In circumstances where emergently triaged conditions do not require emergency or time-sensitive care, transport to an alternative ED could benefit both the patient and overwhelmed EDs.(5)

Foreknowledge of which patient characteristics are predictive of an ED visit resulting in hospital admission, or discharge, can inform policy modification, permitting paramedics to transport emergently triaged patients with non-emergency care needs to the most appropriate hospital, rather than closest. Utilization of machine learning (ML) classifications algorithms could identify a subgroup of patient or visit characteristics in emergently triaged paramedic transported patients that are predictive of visits unlikely to require emergency intervention or hospital admission. ML has the potential to improve care efficiency of patient distribution away from crowded ED’s where appropriate, and support paramedic decision-making when limited hospital communication or clinical information is known of patient conditions, diagnostics, or interventions required in the ED.(6, 7)

The objective of this study was to utilize machine learning algorithms to predict hospital admission from the ED for adult patients arriving by paramedic transport with an emergent triage score. Additionally, we aimed to assess the predictive value and utility of prehospital patient characteristics as predictors of ED outcomes in machine learning. Our study could contribute evidence to the fields of paramedicine, emergency medicine, and medical informatics by reporting on the utility of prehospital data in machine learning to support future research.

Study Design

We conducted a population-level retrospective cohort study to analyze model performance of four machine learning algorithms using Canadian administrative ED patient records. The STROBE statement was followed for reporting of results.(8)

Population

All Ontario adult patients (≥ 18 years) transported to the ED by ground paramedics after utilization of the emergency alert system (911) and had an assigned ‘emergent’ triage acuity score were eligible for inclusion. We excluded ED visits if they arrived by air transport or did not have a recorded triage score. ED visits between January 1, 2018 to December 31, 2019 were eligible for inclusion, as this timeframe represents the most recently available two-year period prior to the Covid-19 pandemic when paramedic and ED utilization trends changed.(9)

Data Source

We used population-level administrative ED patient data from the National Ambulatory Care Reporting System (NACRS) database. NACRS is a hospital and community-based ambulatory care administrative database that collects patient and visit-related data at the time care is provided.(10, 11) All Ontario hospitals are required to submit electronic patient abstracts from the ED to comply with reporting and quality measures and these data are recorded in NACRS. NACRS data was accessed through the ICES (formerly known as the Institute for Clinical Evaluative Sciences); ICES is a non-profit, independent corporation that supports the study of health service and population-wide outcomes in Ontario.

Predictor and Outcome Variables

To be pragmatic, we extracted all patient characteristics available to paramedics prior to the ED visit in the NACRS database. These characteristics constitute a broad representation of patients, their conditions, and rationale for visit prior to ED arrival. We used hospital admission based on ED referral as the outcome, defined as a binary variable where a patient is hospitalized to receive further clinical investigation.(12)

Data Preparation and Variable Selection

We selected variables based upon data availability and completeness in NACRS, previous literature and clinical importance determined a priori by emergency physicians at our local academic hospital center. Missing data for each variable included was scant (< 0.1%), and directly stated where applicable; complete case analysis was utilized. We dummy-coded categorical variables were dummy-coded for non-tree-based algorithm models.(13)

Statistical Analysis and Machine Learning Algorithms

We reported descriptive statistics using measures of central tendency and frequency. We trained four independent ML classification algorithms in this study: logistic regression (LR), lasso logistic regression (Lasso LR), random forest (RF) and gradient boosted trees (GBT). We selected algorithms based on previous literature to predict patient ED visit outcomes in consultation with health care machine learning experts.(14–16) We used 10-fold cross-validation to calculate predictive performance for the models using the same series of variables. Cross-validation is a reliable measure of model performance and an approach commonly utilized in machine learning healthcare prediction research.(17) We managed and analyzed the data in R software, V.3.6 and built the machine learning algorithms were built using ‘mlr’ and ‘mlr3’ packages.(18, 19)

Hyperparameter Tuning

We utilized hyperparameter optimization in RF and GBT algorithms to maximize the average 10-fold cross validation AUC through tuning prior to computation of the final models.(20) For the RF algorithm, we used an exhaustive grid search for eligible tuning parameters and further trained models by optimizing the minimum node size and number of splitting variables. For GBT, we employed a broad random search of parameter space, followed by a focused random search within the best performing areas of the broad search. The hyperparameters trained in the GBT algorithm consisted of maximum number of iterations, maximum tree depth, shrinkage rate, minimum sum of instance weights required to stop splitting and gamma. A comprehensive depiction of all hyperparameter tuning used for the RF and GBT algorithms are shown in Appendix 1.

Performance Measurement and Variable Importance

We measured the performance of all algorithms was measured using the area under the receiver operating curve (AUC) resulting from 10-fold cross-validation and the corresponding 95% confidence interval. We assed calibration accuracy using the Brier Scaled score, a measure of probabilistic predictions on a scale of 0 (perfect accuracy) to 1 (perfect inaccuracy).(21) Additionally, we reported the sensitivity, specificity, positive predictive value, and negative predicted value for each algorithm. To determine the importance of each predictor, we used permutation methods to calculate the average decrease in AUC in the absence of the predictor. The predictors with the highest variable importance corresponded to the largest declines in AUC. To construct a comprehensive list of variable importance across all algorithms, we computed the mean reduction in AUC for all predicters by averaging their AUC reduction from all models. We ranked the mean AUC reductions to produce a top 10 most important predictors list.

Descriptive Results

Figure 1 shows exclusions used to construct the study cohort. A total of 584,925 paramedic transports to the ED with an emergent triage acuity were included, yielding 261,072 (44.6%) patients admitted to hospital and 323,853 (55.4%) patients discharged directly from ED. Our cohort represented 90.4% of all adult emergently triaged paramedic transported ED visits in Ontario.

Table 1 shows a summary of the clinical profile of all patients in our study cohort. The majority of patients were older adults aged 65 years or greater (296,111; 50.6%), male (296,902; 50.8%), referred to ED by self of family (490,242; 83.3%), previously diagnosed with hypertension (336,174; 57.5%) and had their ED visit occur in an urban region of Ontario (535,270; 91.5%). The most common presenting complaints on ED arrival were cardiovascular (178,524; 30.5%), neurologic (99,119; 16.9%) and gastrointestinal (61,104; 10.4%). Missing data was negligible and non-informative, presenting only in the geographic location of the ED (0.5%).

Table 1: Prehospital characteristics of high acuity emergency department patients transported by paramedics in Ontario, Canada, January 1 2018 to December 31, 2019.

Variable (Patient Characteristic)

N (%)

Age Group, years

18-39

40-64

65-105

112,260 (19.2)

176,554 (30.2)

296,111 (50.6)

Sex

Male

Female

296,902 (50.8)

288,023 (49.2)

Comorbidities^a

Hypertension

Diabetes

Chronic Obstructive Pulmonary Disease

Asthma

Rheumatoid Arthritis

Congestive Heart Failure

Bowel Disease

Cancer

336,174 (57.5)

183,657 (31.4)

160,216 (27.4)

133,618 (22.8)

16,279 (2.8)

105,367 (18.0)

7,747 (1.3)

98,377 (16.8)

Access to Primary Healthcare^b

Yes

500,373 (85.5)

84,552 (14.5)

Presenting Complaint^c

Cardiac Arrest

Cardiovascular

ENT (ears, nose, throat)

Environmental

Gastrointestinal

Genitourinary

Mental Health

Neurologic

Obstetrician-Gynecological

Ophthalmology

Orthopedic

Respiratory

Skin

Substance Misuse

Trauma

General/Minor

Other

125 (0.0)

178,523 (30.5)

10,260 (1.8)

1,488 (0.3)

61,104 (10.4)

10,913 (1.9)

27,382 (4.7)

99,119 (16.9)

2,805 (0.5)

1,565 (0.3)

37,516 (6.4)

70,078 (12.0)

5,644 (1.0)

32,459 (5.5)

9,297 (1.6)

32,206 (5.5)

4,441 (0.8)

Geographic Location of ED^d

Urban

Rural

Missing

535,270 (91.5)

46,771 (8.0)

2,884 (0.5)

Referral Source

Self/Family Member

Ambulatory Care Service

Private Practice

Residential Care Facility

Other

490,242 (83.8)

7,169 (1.2)

4,762 (0.8)

44,706 (7.6)

38,046 (6.5)

Receiving Home Care Services

Yes

129,628 (22.2)

455,297 (77.8)

ED Visit Disposition Outcome

Discharged

Admitted

323,853 (55.4)

261,072 (44.6)

Note: ED = emergency department.

^a Conditions present on ED arrival.

^b Identifies if a patient has access to either a family physician, family health team, walk-in clinic or other primary care setting.

^c Categorized by the Canadian Emergency Department Information System (CEDIS) of the Canadian Institute for Health Information (CIHI).

^dDetermined using forward sortation area (FSA) of Ontario postal codes.

Hyperparameter Tuning and Comparison of Predictive Model Performance

Appendix 1 shows the details of hyperparameter tuning ranges and optimal values used to produce the final RF and GBT algorithm outputs. Table 2 displays the final predictive model performance of all machine learning algorithms used in this study after benchmarking. Overall, each algorithm performed very similarly, with an AUC ranging from 0.77-0.78. The GBT model computed the best overall performance in classifying paramedic transported emergently triaged visits that were admitted to hospital, achieving an AUC of 0.78 (95% CI 0.78-0.78). The LR and lasso LR models computed the most accurate predictions with a Brier Scaled score of 0.22. Calibration plots of predicted probabilities were assessed for quality of algorithm predictions. All algorithms received excellent calibration with slopes ranging from 0.997 to 1.086, shown in Appendix 2.

Table 2: Measures of predictive accuracy across all machine learning algorithms.

Measure	LR	Lasso LR	RF	GBT
AUC (95% CI)	0.77 (0.76-0.77)	0.77 (0.76-0.77)	0.78 (0.77-0.78)	0.78 (0.78-0.78)
Brier Scale	0.22	0.22	0.24	0.24
Sensitivity	0.80	0.81	0.76	0.78
Specificity	0.58	0.57	0.63	0.37
PPV	0.70	0.70	0.71	0.71
NPV	0.71	0.71	0.70	0.71
Note: LR = logistic regression, Lasso LR = lasso logistic regression, RF = random forest, GBT = gradient boosted trees, AUC = area under the receiver operating characteristic curve, CI = confidence intervals, PPV = positive predictive value, NPV = negative predictive value.

Variable Importance

Figure 2 shows the mean variable importance scores of the top 10 predictors of an admitted ED visit following paramedic transport of an emergently triaged patient. The top 10 important predictors were (1) age 65 to 105 years, (2) referred to ED from a residential care facility, (3) presenting complaint is respiratory, (4) receiving home care, (5) presenting complaint is general or minor, (6) aged 40 to 64 years, (7) presenting complaint is cardiovascular, (8) presenting complaint is gastrointestinal, (9) presenting complaint is mental health and (10) comorbidity of hypertension. Appendix 3 shows the relative importance of each predictor across all machine learning algorithms, presented as the reduction in AUC in absence of the predictor. Appendix 4 shows the variable-relevance ranking of the top 10 predictors for each ML algorithm. Aged 65 to 105 years was consistently the most important variable ranked amongst all ML algorithms. Referral from residential care was ranked as either the second or third most important predictor and presenting complaint of respiratory was ranked from second to fourth position. Receiving home care’s rank was spread from second to sixth in the algorithms. All remaining top 10 predictors were distributed variably amongst each algorithm, with less uniformity in rankings.

We were able to make accurate ED admission predictions using readily available clinical variables (age, gender, presenting complaint, comorbidities) and non-clinical variables (geographic location of ED, referral source, receiving home care services) known to paramedics prior to ED transport that highlight a plausible stratification of emergently triaged patients. The acceptable algorithm results demonstrated that prehospital characteristics have predictive validity to forecast ED visit dispositions even prior to ED arrival. In the absence of additional important clinical and non-clinical variables, we were able to compute acceptable AUC outputs of 0.77-0.78 and probabilistic accuracy scores of 0.22-0.24 across all algorithms. All machine learning algorithms performed similarly in all measures of statistical precision (sensitivity, specificity).

We found that that older age, living in a residential facility, presenting to ED with a respiratory complaint, and receiving home care were the most informative predictors of hospital admission. These findings are consistent with previous literature which analyzed admission predictors of unscheduled ED visits.(22–25) A plethora of research has been dedicated to examining variables predictive of admission in the older adult cohorts, yielding results consistent with our predictors.(26–29) Receiving home care was included less as an admission predictor in the literature, representing a plausible underreporting of this variables influence on ED visit dispositions. Our results highlight that later-stages of life and requiring support care (residential or institutional) are the driving forces of ED admission from prehospital patient characteristics.

Several characteristics such as the ED located in an urban region, presenting complaints related to environmental reasons or the skin, comorbidities of bowel disease or asthma, and referral from ambulatory care services or private practices were not informative in predicting hospital admissions. Our non-informative predictors were somewhat consistent with prior work.(4,30) Specifically, comorbidities of asthma and diabetes are reported as risk markers to hospital admission, although not consistently in the literature.(31,32) Our study determined these two predictors were not informative of forecasting admission, but this could be a result of analyzing only paramedic transported visits in our study, where paramedics could have treated asthma or diabetes conditions prior to ED arrival in the prehospital field, thereby decreasing the likelihood of need for admission.

To our knowledge, our study is the first to employ machine learning algorithms to predict ED outcomes using data known prehospital to paramedics. Thus, our study contributes novel research to the scientific literature, and showcases a framework for future machine learning studies in the prehospital and paramedic fields. Given the easy accessibility of these characteristics to paramedics, stakeholders of paramedic practices and policymakers should consider incorporating the most important predictors into any future modifications of patient distribution legislation or clinical decision rules for paramedics.

With advanced knowledge of which patient characteristics are most predictive of admission, a patient distribution protocol could be constructed to stratify a two-tier emergently triaged patient cohort. The first subgroup could be defined as emergently triaged patients with a high predicted probability of admission who must be transported to the closest ED. The second subgroup would be patients emergently triaged but with low predicted probability of admission, who could be transported to the most appropriate ED, not necessarily the closest. Future research is needed to delineate the inclusion criteria of subgroups within emergently triaged patients, but our study is hypothesis generating and supports research to further explore this concept. Integration of threshold tuning to identify an ideal cut-point in the predicted probability with an acceptable sensitivity could be beneficial in defining the subgroups. Stratification of patients assigned an emergent triage score into two subgroups could optimize patient distribution, improve ED congestion and plausibly improve patient-important outcomes

Bearing in mind that the discriminative and calibration accuracy of the LR model was similar to more advanced ML algorithms, it is worth nothing that this model may be best for pragmatic implementation and decision-making at the clinical and policy level. Relative odds and confidence interval interpretation are standard in medical education, whereas ML interpretation is not. Additionally, the simplicity of LR supports ease of implementation of study findings towards the use of clinical decision rules and uptake of study findings by medical and health policy professionals, who are mostly likely to influence direct practice change.

An overarching goal of this study was to determine if patient characteristics known prehospital could be predictive of ED visit dispositions prior to arrival. Our study demonstrated that prehospital patient characteristics have predictive utility of a forthcoming ED visit. We establish that ML has ample utility for integration into prehospital research, and can be an important methodology to optimize paramedic clinical practices and logistical effectiveness. Inclusion of paramedic medical reports, in addition to ED administrative data resources, would undoubtedly increase the probabilistic accuracy of predictive algorithms, and contribute variables that encompass a more complete prehospital patient presentation than ED resources alone.

Future research should leverage ML to analyze the utility of patient characteristics to predict ED visit dispositions in low acuity transports. Ontario paramedics may be granted authorization to transport low acuity patients to non-ED community-based care alternatives, though identification and classification of suitable patients for redirection from ED is challenging prehospital.(33) Future research on low acuity transports should utilize ML to investigate which patient characteristics are most predictive of an ED visit that is both discharged and did not require any the emergency medicine intervention.(33) Knowledge of these predictors would be beneficial to support paramedic clinical guideline development to identify patients potentially suitable for redirection, given clinical information is limited prehospital to paramedics.

Limitations

Additional administrative clinical and non-clinical variables known to paramedics prehospital could not be included in this study, as this data is not recorded in NACRS. Accessibility of paramedic administrative data resources remains a constant challenge in Ontario, all data are held within each of 52 paramedic service independently and do not regularly permit access for research.(7) Additionally, potentially influential characteristics known prehospital that are accessible through other ICES administrative databases (i.e. prescription medications, time of visit, caregiver attendance) could have been included in our study, but were excluded due to their large incompleteness. However, we do not believe these clinical factors would have made sizable impact to our results.

ED visit disposition was predicted from prehospital administrative patient characteristics of paramedic transported emergently triaged patients. Each ML algorithm performed relatively similar with an acceptable AUC and probabilistic accuracy score using readily available clinical data known pre-hospital to paramedics. Our results could support policy modification regarding patient distribution to ease burdens of ED workloads when patients do not require emergency medical attention. Our study highlights that prehospital patient characteristics have abundant predictive utility in machine learning, and should be utilized in future paramedic and prehospital research.

AUC = area under the receiver operating curve

CI = confidence interval

CTAS = Canadian Triage and Acuity Scale

ED = emergency department

GBT = gradient boosted trees

ICES = Institute for Clinical Evaluative Sciences

LR = logistic regression

Lasso LR = lasso logistic regression

NACRS = National Ambulatory Care Reporting System

RF = random forest

Ethics Approval and Consent to Participate

ICES’ collection and use of NACRS secondary ambulatory data are authorized under Section 45 of Ontario’s Personal Health Information Protection Act (PHIPA) as a prescribed entity, which is exempt from review by a Research Ethics Board. (34,35) The use of the data in this study is authorized under section 45 and approved by ICES’ Privacy and Legal Office, which does not require participant consent according to national regulations in Ontario, Canada.(34)

Consent for Publication

Not applicable.

Availability of Data and Materials

All aggregate data herein are accessible to other interested parties by application to the corresponding author.

Competing Interests

The authors declare that they have no competing interests.

Funding

This work was substantially supported by Schlegel Chair in Clinical Epidemiology in Aging at McMaster University.

Authors’ Contributions

RPS and AJ led the conceptualization of the study objective and methodology. RPS designed the study, drafted and revised the manuscript. RPS and AJ conducted the analysis. All authors made critical revisions to the interpretation and manuscript. All authors agreed to be accountable for all aspects of this manuscript.

Acknowledgments

This study was supported by ICES, which is funded by an annual grant from the Ontario Ministry of Health (MOH) and the Ministry of Long-Term Care (MLTC).

Author Affiliations:

RPS, FIM, AJ: Department of Health Research Methods, Evidence and Impact, McMaster University, Hamilton, Canada

MZ: Department of Health Policy and Management, DeGroote School of Business, McMaster University, Hamilton, Canada

AJ: Institute for Clinical Evaluative Sciences, McMaster University, Hamilton, Canada

Emergency Health Services Branch, Ministry of Health and Long-Term Care. Prehospital Canadian Triage & Acuity Scale: Prehospital CTAS Paramedic Guide [Internet]. Government of Ontario; 2016. Available from: https://www.lhsc.on.ca/media/2904/download
Bullard MJ, Unger B, Spence J, Grafstein E, the CTAS National Working Group. Revisions to the Canadian Emergency Department Triage and Acuity Scale (CTAS) adult guidelines. CJEM [Internet]. 2008 Mar [cited 2020 Dec 5];10(02):136–42. Available from: https://www.cambridge.org/core/product/identifier/S1481803500009854/type/journal_article
Strum RP, Tavares W, Worster A, Griffith LE, Costa AP. Identifying patient characteristics associated with potentially redirectable paramedic transported emergency department visits in Ontario, Canada: a population-based cohort study. BMJ Open [Internet]. 2021 Dec 1 [cited 2021 Dec 31];11(12):e054625. Available from: https://bmjopen.bmj.com/content/11/12/e054625
Strum RP, Mowbray FI, Worster A, Tavares W, Leyenaar MS, Correia RH, et al. Examining the association between paramedic transport to the emergency department and hospital admission: a population-based cohort study. BMC Emerg Med [Internet]. 2021 Dec [cited 2021 Nov 16];21(1):1–9. Available from: https://bmcemergmed.biomedcentral.com/articles/10.1186/s12873-021-00507-2
Shen Y, Lee LH. Improving the wait time to consultation at the emergency department. BMJ Open Qual [Internet]. 2018 Jan 3 [cited 2022 Mar 22];7(1):e000131. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5759711/
Hong WS, Haimovich AD, Taylor RA. Predicting hospital admission at emergency department triage using machine learning. PLOS ONE [Internet]. 2018 Jul 20 [cited 2022 Mar 13];13(7):e0201016. Available from: https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0201016
Strum RP, Costa AP. Implementing machine learning in paramedicine. CMAJ [Internet]. 2022 Jan 17 [cited 2022 Mar 13];194(2):E54–E54. Available from: https://www.cmaj.ca/content/194/2/E54
Elm E von, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP. Strengthening the reporting of observational studies in epidemiology (STROBE) statement: guidelines for reporting observational studies. BMJ [Internet]. 2007 Oct 18 [cited 2020 Dec 4];335(7624):806–8. Available from: https://www.bmj.com/content/335/7624/806
Ferron R, Agarwal G, Cooper R, Munkley D. The effect of COVID-19 on emergency medical service call volumes and patient acuity: a cross-sectional study in Niagara, Ontario. BMC Emerg Med [Internet]. 2021 Mar 29 [cited 2022 Jan 1];21(1):39. Available from: https://doi.org/10.1186/s12873-021-00431-5
Li G, Lau JT, McCarthy ML, Schull MJ, Vermeulen M, Kelen GD. Emergency Department Utilization in the United States and Ontario, Canada. Acad Emerg Med [Internet]. 2007 [cited 2020 Dec 4];14(6):582–4. Available from: https://onlinelibrary.wiley.com/doi/abs/10.1197/j.aem.2007.02.030
National Ambulatory Care Reporting System metadata (NACRS) | CIHI [Internet]. [cited 2020 Nov 14]. Available from: https://www.cihi.ca/en/national-ambulatory-care-reporting-system-metadata-nacrs
Calder LA, Arnason T, Vaillancourt C, Perry JJ, Stiell IG, Forster AJ. How do emergency physicians make discharge decisions? Emerg Med J [Internet]. 2015 Jan 1 [cited 2022 Jan 5];32(1):9–14. Available from: https://emj.bmj.com/content/32/1/9
Johnson MK and K. Feature Engineering and Selection: A Practical Approach for Predictive Models [Internet]. [cited 2022 Jan 6]. Available from: http://www.feat.engineering/
Jones A, Costa AP, Pesevski A, McNicholas PD. Predicting hospital and emergency department utilization among community-dwelling older adults: Statistical and machine learning approaches. PLOS ONE [Internet]. 2018 Nov 1 [cited 2022 Mar 15];13(11):e0206662. Available from: https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0206662
Graham B, Bond R, Quinn M, Mulvenna M. Using Data Mining to Predict Hospital Admissions From the Emergency Department. IEEE Access. 2018;6:10458–69.
Raita Y, Goto T, Faridi MK, Brown DFM, Camargo CA, Hasegawa K. Emergency department triage prediction of clinical outcomes using machine learning models. Crit Care [Internet]. 2019 Feb 22 [cited 2022 Mar 15];23(1):64. Available from: https://doi.org/10.1186/s13054-019-2351-7
Doupe P, Faghmous J, Basu S. Machine Learning for Health Services Researchers. Value Health [Internet]. 2019 Jul 1 [cited 2022 Mar 15];22(7):808–15. Available from: https://www.sciencedirect.com/science/article/pii/S1098301519301469
Binder M, Pﬁsterer F, Lang M, Schneider L, Kotthoﬀ L, Bischl B. mlr3pipelines – Flexible Machine Learning Pipelines in R. J Mach Learn Res. 2021 Jun;22:1–7.
Lang M, Binder M, Richter J, Schratz P, Pfisterer F, Coors S, et al. mlr3: A modern object-oriented machine learning framework in R. J Open Source Softw [Internet]. 2019 Dec 11 [cited 2022 Jan 11];4(44):1903. Available from: https://joss.theoj.org/papers/10.21105/joss.01903
Abbott D. Applied Predictive Analytics: Principles and Techniques for the Professional Data Analyst. John Wiley & Sons; 2014. 456 p.
Rufibach K. Use of Brier score to assess binary predictions. J Clin Epidemiol [Internet]. 2010 Aug 1 [cited 2022 Jan 6];63(8):938–9. Available from: https://www.jclinepi.com/article/S0895-4356(09)00363-1/fulltext
Hayward J, Hagtvedt R, Ma W, Gauri A, Vester M, Holroyd BR. Predictors of Admission in Adult Unscheduled Return Visits to the Emergency Department. West J Emerg Med. 2018 Nov;19(6):912–8.
Lin D, Worster A. Predictors of admission to hospital of patients triaged as nonurgent using the Canadian Triage and Acuity Scale. CJEM [Internet]. 2013 Nov [cited 2020 Dec 11];15(06):353–8. Available from: https://www.cambridge.org/core/product/identifier/S148180350000244X/type/journal_article
Mowbray FI, Aryal K, Mercier E, Heckman G, Costa AP. Older Emergency Department Patients: Does Baseline Care Status Matter? Can Geriatr J CGJ. 2020 Dec;23(4):289–96.
Mowbray F, Zargoush M, Jones A, de Wit K, Costa A. Predicting hospital admission for older emergency department patients: Insights from machine learning. Int J Med Inf [Internet]. 2020 Aug 1 [cited 2022 May 31];140:104163. Available from: https://www.sciencedirect.com/science/article/pii/S1386505619314510
Marcusson J, Nord M, Dong HJ, Lyth J. Clinically useful prediction of hospital admissions in an older population. BMC Geriatr [Internet]. 2020 Dec [cited 2020 Dec 11];20(1):95. Available from: https://bmcgeriatr.biomedcentral.com/articles/10.1186/s12877-020-1475-6
Gabayan GZ, Sarkisian CA, Liang LJ, Sun BC. Predictors of admission after emergency department discharge in older adults. J Am Geriatr Soc. 2015 Jan;63(1):39–45.
Vetrano DL, Landi F, De Buyser SL, Carfì A, Zuccalà G, Petrovic M, et al. Predictors of length of hospital stay among older adults admitted to acute care wards: a multicentre observational study. Eur J Intern Med [Internet]. 2014 Jan 1 [cited 2022 Mar 16];25(1):56–62. Available from: https://www.sciencedirect.com/science/article/pii/S0953620513009187
Aliyu MH, Adediran AS, Obisesan TO. Predictors of hospital admissions in the elderly: analysis of data from the Longitudinal Study on Aging. J Natl Med Assoc [Internet]. 2003 Dec [cited 2022 Mar 16];95(12):1158–67. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2594833/
Rahimian F, Salimi-Khorshidi G, Payberah AH, Tran J, Solares RA, Raimondi F, et al. Predicting the risk of emergency admission with machine learning: Development and validation using linked electronic health records. PLOS Med [Internet]. 2018 Nov 20 [cited 2022 Mar 16];15(11):e1002695. Available from: https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1002695
Donnan PT, Dorward DWT, Mutch B, Morris AD. Development and Validation of a Model for Predicting Emergency Admissions Over the Next Year (PEONY): A UK Historical Cohort Study. Arch Intern Med [Internet]. 2008 Jul 14 [cited 2022 Mar 16];168(13):1416–22. Available from: https://doi.org/10.1001/archinte.168.13.1416
Ownby DR, Abarzua J, Anderson JA. Attempting to predict hospital admission in acute asthma. Am J Dis Child 1960. 1984 Nov;138(11):1062–6.
Strum RP, Tavares W, Worster A, Griffith LE, Costa AP. Emergency department interventions that could be conducted in subacute care settings for patients with nonemergent conditions transported by paramedics: a modified Delphi study. Can Med Assoc Open Access J [Internet]. 2022 Jan 1 [cited 2022 Jan 11];10(1):E1–7. Available from: https://www.cmajopen.ca/content/10/1/E1
Government of Ontario. Personal Health Information Protection Act, 2004, S.O. 2004, c. 3, Sched. A [Internet]. Jul 24, 2014. Available from: https://www.ontario.ca/laws/statute/04p03#BK63
Government of Ontario. Ontario Regulation 329/04: General, Personal Health Information Protection Act [Internet]. Sect. 18. Prescribed entitites for the purposes of s. 45 (1) of the Act 02, 2021. Available from: https://www.ontario.ca/laws/regulation/040329#BK19

Download PDF

Version 1

posted

You are reading this latest preprint version

Predicting Hospital Admission among High Acuity Triaged Patients Transported to the Emergency Department in Ontario, Canada: A Population-Based Cohort Study using Machine Learning

Status:

Version 1

Abstract

Background

Methods

Results

Conclusions

Figures

Background

Methods

Study Design

Population

Data Source

Predictor and Outcome Variables

Data Preparation and Variable Selection

Statistical Analysis and Machine Learning Algorithms

Hyperparameter Tuning

Performance Measurement and Variable Importance

Results

DISCUSSION

CONCLUSION

Abbreviations

Declarations

References

Supplementary Files

Status:

Version 1