feat(proposal): work on background and motivation

bibliography/bibliography.bib

@@ -1009,3 +1009,64 @@ Publisher: {BioMed} Central},
 	urldate = {2023-04-27},
 	file = {BioButton®* Multi-parameter Wearable | Medtronic:/home/ulinja/Zotero/storage/Z5TF3VAL/healthcast-biobutton-multi-parameter-wearable.html:text/html},
 }
+
+@article{wu_predicting_2021,
+	title = {Predicting in-hospital mortality in adult non-traumatic emergency department patients: a retrospective comparison of the Modified Early Warning Score ({MEWS}) and machine learning approach},
+	volume = {9},
+	issn = {2167-8359},
+	url = {https://peerj.com/articles/11988},
+	doi = {10.7717/peerj.11988},
+	shorttitle = {Predicting in-hospital mortality in adult non-traumatic emergency department patients},
+	abstract = {Background A feasible and accurate risk prediction systems for emergency department ({ED}) patients is urgently required. The Modified Early Warning Score ({MEWS}) is a wide-used tool to predict clinical outcomes in {ED}. Literatures showed that machine learning ({ML}) had better predictability in specific patient population than traditional scoring system. By analyzing a large multicenter dataset, we aim to develop a {ML} model to predict in-hospital morality of the adult non traumatic {ED} patients for different time stages, and comparing performance with other {ML} models and {MEWS}. Methods A retrospective observational cohort study was conducted in five Taiwan {EDs} including two tertiary medical centers and three regional hospitals. All consecutively adult ({\textgreater}17 years old) non-traumatic patients admit to {ED} during a 9-year period (January first, 2008 to December 31th, 2016) were included. Exclusion criteria including patients with (1) out-of-hospital cardiac arrest and (2) discharge against medical advice and transferred to other hospital (3) missing collect variables. The primary outcome was in-hospital mortality and were categorized into 6, 24, 72, 168 hours mortality. {MEWS} was calculated by systolic blood pressure, pulse rate, respiratory rate, body temperature, and level of consciousness. An ensemble supervised stacking {ML} model was developed and compared to sensitive and unsensitive Xgboost, Random Forest, and Adaboost. We conducted a performance test and examine both the area under the receiver operating characteristic ({AUROC}) and the area under the precision and recall curve ({AUPRC}) as the comparative measures. Result After excluding 182,001 visits (7.46\%), study group was consisted of 24,37,326 {ED} visits. The dataset was split into 67\% training data and 33\% test data for {ML} model development. There was no statistically difference found in the characteristics between two groups. For the prediction of 6, 24, 72, 168 hours in-hospital mortality, the {AUROC} of {MEW} and {ML} mode was 0.897, 0.865, 0.841, 0.816 and 0.939, 0.928, 0.913, 0.902 respectively. The stacking {ML} model outperform other {ML} model as well. For the prediction of in-hospital mortality over 48-hours, {AUPRC} performance of {MEWS} drop below 0.1, while the {AUPRC} of {ML} mode was 0.317 in 6 hours and 0.2150 in 168 hours. For each time frame, {ML} model achieved statistically significant higher {AUROC} and {AUPRC} than {MEWS} (all P {\textless} 0.001). Both models showed decreasing prediction ability as time elapse, but there was a trend that the gap of {AUROC} values between two model increases gradually (P {\textless} 0.001). Three {MEWS} thresholds (score {\textgreater}3, {\textgreater}4, and {\textgreater}5) were determined as baselines for comparison, {ML} mode consistently showed improved or equally performance in sensitivity, {PPV}, {NPV}, but not in specific. Conclusion Stacking {ML} methods improve predicted in-hospital mortality than {MEWS} in adult non-traumatic {ED} patients, especially in the prediction of delayed mortality.},
+	pages = {e11988},
+	journaltitle = {{PeerJ}},
+	author = {Wu, Kuan-Han and Cheng, Fu-Jen and Tai, Hsiang-Ling and Wang, Jui-Cheng and Huang, Yii-Ting and Su, Chih-Min and Chang, Yun-Nan},
+	urldate = {2023-04-28},
+	date = {2021-08-24},
+	langid = {english},
+	note = {Publisher: {PeerJ} Inc.},
+	file = {Full Text PDF:/home/ulinja/Zotero/storage/H2MPDP9A/Wu et al. - 2021 - Predicting in-hospital mortality in adult non-trau.pdf:application/pdf},
+}
+
+@article{martin-rodriguez_analysis_2019,
+	title = {Analysis of the early warning score to detect critical or high-risk patients in the prehospital setting},
+	volume = {14},
+	issn = {1970-9366},
+	url = {https://doi.org/10.1007/s11739-019-02026-2},
+	doi = {10.1007/s11739-019-02026-2},
+	abstract = {The early warning score can help to prevent, recognize and act at the first signs of clinical and physiological deterioration. The objective of this study is to evaluate different scales for use in the prehospital setting and to select the most relevant one by applicability and capacity to predict mortality in the first 48 h. A prospective longitudinal observational study was conducted in patients over 18 years of age who were treated by the advanced life support unit and transferred to the emergency department between April and July 2018. We analyzed demographic variables as well as the physiological parameters and clinical observations necessary to complement the {EWS}. Subsequently, each patient was followed up, considering their final diagnosis and mortality data. A total of 349 patients were included in our study. Early mortality before the first 48 h affected 27 patients (7.7\%). The scale with the best capacity to predict early mortality was the National Early Warning Score 2, with an area under the curve of 0.896 (95\% {CI} 0.82–0.97). The score with the lowest global classification error was 10 points with sensitivity of 81.5\% (95\% {CI} 62.7–92.1) and specificity of 88.5\% (95\% {CI} 84.5–91.6). The early warning score studied (except modified early warning score) shows no statistically significant differences between them; however, the National Early Warning Score 2 is the most used score internationally, validated at the prehospital scope and with a wide scientific literature that supports its use. The Prehospital Emergency Medical Services should include this scale among their operative elements to complement the structured and objective evaluation of the critical patient.},
+	pages = {581--589},
+	number = {4},
+	journaltitle = {Intern Emerg Med},
+	author = {Martín-Rodríguez, Francisco and Castro-Villamor, Miguel Ángel and del Pozo Vegas, Carlos and Martín-Conty, José Luis and Mayo-Iscar, Agustín and Delgado Benito, Juan Francisco and del Brio Ibañez, Pablo and Arnillas-Gómez, Pedro and Escudero-Cuadrillero, Carlos and López-Izquierdo, Raúl},
+	urldate = {2023-04-28},
+	date = {2019-06-01},
+	langid = {english},
+	keywords = {Early warning score, Clinical research, Early mortality, Prehospital care, Prognosis},
+	file = {Full Text PDF:/home/ulinja/Zotero/storage/2LVIYDZR/Martín-Rodríguez et al. - 2019 - Analysis of the early warning score to detect crit.pdf:application/pdf},
+}
+
+@article{abbott_pre-hospital_2018,
+	title = {Pre-hospital National Early Warning Score ({NEWS}) is associated with in-hospital mortality and critical care unit admission: A cohort study},
+	volume = {27},
+	issn = {2049-0801},
+	url = {https://www.sciencedirect.com/science/article/pii/S2049080118300116},
+	doi = {10.1016/j.amsu.2018.01.006},
+	shorttitle = {Pre-hospital National Early Warning Score ({NEWS}) is associated with in-hospital mortality and critical care unit admission},
+	abstract = {Background
+National Early Warning Score ({NEWS}) is increasingly used in {UK} hospitals. However, there is only limited evidence to support the use of pre-hospital early warning scores. We hypothesised that pre-hospital {NEWS} was associated with death or critical care escalation within the first 48 h of hospital stay.
+Methods
+Planned secondary analysis of a prospective cohort study at a single {UK} teaching hospital. Consecutive medical ward admissions over a 20-day period were included in the study. Data were collected from ambulance report forms, medical notes and electronic patient records. Pre-hospital {NEWS} was calculated retrospectively. The primary outcome was a composite of death or critical care unit escalation within 48 h of hospital admission. The secondary outcome was length of hospital stay.
+Results
+189 patients were included in the analysis. The median pre-hospital {NEWS} was 3 ({IQR} 1–5). 13 patients (6.9\%) died or were escalated to the critical care unit within 48 h of hospital admission. Pre-hospital {NEWS} was associated with death or critical care unit escalation ({OR}, 1.25; 95\% {CI}, 1.04–1.51; p = 0.02), but {NEWS} on admission to hospital was more strongly associated with this outcome ({OR}, 1.52; 95\% {CI}, 1.18–1.97, p {\textless} 0.01). Neither was associated with hospital length of stay.
+Conclusion
+Pre-hospital {NEWS} was associated with death or critical care unit escalation within 48 h of hospital admission. {NEWS} could be used by ambulance crews to assist in the early triage of patients requiring hospital treatment or rapid transport. Further cohort studies or trials in large samples are required before implementation.},
+	pages = {17--21},
+	journaltitle = {Annals of Medicine and Surgery},
+	author = {Abbott, Tom E. F. and Cron, Nicholas and Vaid, Nidhi and Ip, Dorothy and Torrance, Hew D. T. and Emmanuel, Julian},
+	urldate = {2023-04-28},
+	date = {2018-03-01},
+	langid = {english},
+	keywords = {Clinical research, Acute care emergency ambulance systems, Intensive care, Pre-hospital},
+	file = {ScienceDirect Full Text PDF:/home/ulinja/Zotero/storage/2HPZCFXG/Abbott et al. - 2018 - Pre-hospital National Early Warning Score (NEWS) i.pdf:application/pdf},
+}

proposal/README.md

@@ -36,5 +36,7 @@ AND TITLE-ABS-KEY ( "system*"  OR  "automat*"  OR  "smart*"  OR  "wearable*"  OR
 AND TITLE-ABS-KEY ( 'continuous' OR 'outpatient' OR 'home' OR 'remote' OR 'domestic')
 
 
-TITLE-ABS-KEY (("patient" OR "clinical" OR "medical") AND ("deterioration" OR "instability" OR "decompensation" OR "admission" OR "hospitalization" OR "escalation" OR "triage" OR "emergency")) AND TITLE-ABS-KEY ( "early warning"  OR  "early warning score"  OR  "warning"  OR  "score*"  OR  "EWS" ) AND TITLE-ABS-KEY ( "system*"  OR  "automat*"  OR  "smart*"  OR  "wearable*"  OR  "internet of thing*"  OR  "IOT"  OR  "digital"  OR  "sensor*"  OR  "signal"  OR  "intelligen*"  OR  "predict*"  OR  "monitor*"  OR  "sreen*"  OR  ( ( "vital*"  OR  "bio*" )  AND  ( "marker*"  OR  "sign*"  OR  "monitor*" ) )) AND TITLE-ABS-KEY ( 'continuous' OR 'outpatient' OR 'home' OR 'remote' OR 'domestic')
+TITLE-ABS-KEY (("patient" OR "clinical" OR "medical") AND ("deterioration" OR "instability" OR "decompensation" OR "admission" OR "hospitalization" OR "escalation" OR "triage" OR "emergency"))
+AND TITLE-ABS-KEY ( "early warning"  OR  "early warning score"  OR  "warning"  OR  "score*"  OR  "EWS" OR "early warning system" OR "rapid response system" )
+AND TITLE-ABS-KEY ( "Avoidance" OR "Prophylaxis" OR "Preclusion" OR "Anticipation" OR "Hindrance" OR "Obviation" OR "Deterrence" OR "Preemption" OR "Abstention" OR "Restraint" OR "Inhibition" OR "Exclusion" OR "Repression" OR "Suppression" )
 ```

proposal/proposal.tex

@@ -17,7 +17,7 @@
 \pagestyle{plain}
 
 % Citations
-\usepackage{csquotes}
+%\usepackage{cite}
 \usepackage[backend=biber, style=vancouver]{biblatex}
 \addbibresource{../bibliography/bibliography.bib}
 
@@ -32,31 +32,42 @@
 
 \section{Background}
 
-In hospital setting, calculation of EWSs has been shown to predict important clinical outcomes effectively, such as severe deterioration, likelyhood of ICU admission,
-and mortality\autocite{subbe_validation_2001, buist_association_2004, paterson_prediction_2006, alam_exploring_2015, bilben_national_2016, brekke_value_2019}.
-Traditionally, doctors and nursing staff collect and evaluate patient vitals data manually, which is limited due to lack of resources\cite{shaik_remote_2023}.
-Remote patient monitoring (RPM) can provide early pre-symptomatic detection of deterioration and hospital admission.
-The Covid pandemic in particular has sparked efforts to investigate remote patient monitoring solutions, and NEWS2 has proved beneficial in predicting critical outcomes\cite{gidari_predictive_2020, otoom_iot-based_2020, filho_iot-based_2021, carr_evaluation_2021}.
-Javanbakht et al. found that continuous vitals monitoring is more cost-effective than intermittent monitoring\cite{javanbakht_cost_2020}, however the findings of
-this study should be taken lightly due to potential bias reporting.
+Early warning scores (EWS) have been widely adopted internationally to identify deteriorating patients\cite{downey_strengths_2017}.
+A large body of scientific evidence validates the effectiveness of EWS in assessing severity of illness, and in predicting adverse clinical events,
+such as severe deterioration, likelihood of ICU admission, and mortality, both on hospital wards\cite{subbe_validation_2001, buist_association_2004, paterson_prediction_2006, alam_exploring_2015, bilben_national_2016, brekke_value_2019}
+and in ambulatory care \cite{ehara_effectiveness_2019, burgos-esteban_effectiveness_2022, paganelli_conceptual_2022}.
 
-A variety of wearable medical sensors capable of continuously recording vital parameters have been developed recently\cite{noauthor_visi_nodate, noauthor_equivital_nodate, noauthor_vitls_nodate, noauthor_caretaker_nodate, noauthor_medtronic_nodate}.
+Two common implemetations are the \textit{National Early Warning Score 2} (NEWS2) and the
+\textit{Modified Early Warning Score} (MEWS)\cite{burgos-esteban_effectiveness_2022}.
+Both are calculated by capturing various vital parameters from the patient at a specific point in time, followed by numerical aggregation of the
+captured data according to the specifically used score.
+Traditionally, doctors and nursing staff perform collection and evaluation of the data manually, inputting data into an EWS-calculator by hand.
+Frequency of scoring, miscalculations and practical integration are known setbacks of NEWS2 and other scores\cite{eisenkraft_developing_2023}.
+% which is limited due to lack of resources\cite{shaik_remote_2023}.
+
+Remote patient monitoring (RPM) can improve detection of deterioration\cite{shaik_remote_2023} by greatly reducing the
+amount of human interaction required to take measurements and perform EWS calculations.
+A number of studies have explored RPM combined with automated EWS calculation in hospitals\cite{filho_iot-based_2021, un_observational_2021, karvounis_hospital_2021, eisenkraft_developing_2023}.
+With hospitals facing critical patient demand during the SARS-CoV-2 pandemic, interest in exploring remote patient monitoring options surged,
+and NEWS2 emerged as an effective tool for predicting severe infection outcomes\cite{gidari_predictive_2020, otoom_iot-based_2020, filho_iot-based_2021, carr_evaluation_2021},
+while reducing person-to-person contact during patient monitoring.
+%Javanbakht et al. found that continuous vitals monitoring is more cost-effective than intermittent monitoring\cite{javanbakht_cost_2020}, however the findings of
+%this study should be taken lightly due to potential bias reporting.
+Since then, a variety of wearable medical sensors capable of continuously recording vital parameters have been developed and are
+commercially available\cite{noauthor_visi_nodate, noauthor_equivital_nodate, noauthor_vitls_nodate, noauthor_caretaker_nodate, noauthor_medtronic_nodate}.
 
 \section{Motivation}
-EWSs may also be a viable tool for predicting deterioration outside of hospitals\cite{ehara_effectiveness_2019, burgos-esteban_effectiveness_2022, paganelli_conceptual_2022}, allowing for preemptive action to be taken.
-However, some known setbacks of the NEWS and other scales are the frequency of scoring and response, integration into practice, and miscalculation by healthcare providers\cite{eisenkraft_developing_2023}.
 
-Several studies have examined vitals monitoring using wearables for at-home-patients in a laboratory setting,
+While the application of EWS in ambulant care facilities and hospitals has been thoroughly investigated, very little research has been done to
+assess their practicability for remote monitoring of at-risk patients at home.
+Some studies have examined monitoring individual vital signs for abnormalities using wearables for at-home-patients in a laboratory setting,
 however in most of them, no automated EWS calculations were made\cite{archip_iot_2016, azimi_medical_2016, chowdary_efficient_2018, yeri_iot_2020, lee_all-day_2020, athira_design_2020, phaltankar_curaband_2021, thippeswamy_prototype_2021}.
-
-Sahu et al. used the PM6750\cite{sahu_internet--things-enabled_2022}, an experimental vitals data monitoring device capable of continuous measurements in a laboratory setting.
-However, the methodology of real-time EWS calculation using data gathered in the laboratory is unclear and was not demonstrated.
-Furthermore, the device used to take continuous measurements, requires a large number of sensors and cables to be continuously attached
-to the patient's body\cite{noauthor_pm6750_nodate}, restricting movement.
-
 Anzanpour et al. developed a monitoring system which collects vitals data and calculates EWSs in 2015, however due to limited or nonexistent
 availability of remotely operable sensors for all vital signs relevant to EWSs, the work was limited to using a laboratory prototype
 requiring some manual interaction in transferring vitals data\cite{anzanpour_internet_2015}.
+Sahu et al. documented their development of an EWS-supported digital early warning system using the PM6750\cite{sahu_internet--things-enabled_2022},
+an experimental vitals data monitoring device capable of taking continuous measurements in a laboratory setting\cite{noauthor_pm6750_nodate}.
+However, the methodology of real-time EWS calculation using data gathered in the laboratory is inconsistent and was not demonstrated.
 
 Patients appreciate the face-to-face aspect of early warning score monitoring as it allows for reassurance, social interaction, and gives them further opportunity to ask questions about their medical care\cite{downey_patient_2018}.