key: cord-216974-0al3vdh1 authors: Ravaud, Philippe; Ouay, Franck le; Depaulis, Etienne; Huckert, Alexandre; Vegreville, Bruno; Tran, Viet-Thi title: Reconfiguring health services to reduce the workload of caregivers during the COVID-19 outbreak using an open-source scalable platform for remote digital monitoring and coordination of care in hospital Command Centres date: 2020-03-12 journal: nan DOI: nan sha: doc_id: 216974 cord_uid: 0al3vdh1 The Covid-19 outbreak threatens to saturate healthcare systems in most Western countries. We describe how digital technologies may be used to automatically and remotely monitor patients at home. Patients answer simple self-reported questionnaires and their data is transmitted, in real time, to a Command Centre in the nearest reference hospital. Patient reported data are automatically filtered by algorithms to identify those with early warning signs. Open-source code of all software components required to deploy the remote digital monitoring platform and Command Centres is available. The coronavirus disease 2019 (Covid-19) has been declared a public health emergency of international concern 1 . On the 11 th of March 2020, 118 162 cases were confirmed in 113 countries, with 4 290 deaths 2 . The Covid-19 outbreak threatens to saturate healthcare systems in Western countries which are already at breaking point dealing with routine demands 3 . Indeed, the virus transmits easily (R0 between 1.4 and 3.28) and its current management involves mainly hospital care, despite the fact that most patients who will contract the virus will have few symptoms and long term effects [3] [4] [5] . Naïve calculations show that this strategy will not be sustainable as the epidemic progresses 6 . To anticipate for the escalation of the epidemic, since the 7 th of March 2020, patients with confirmed Covid-19 without signs of severity in France can be managed at home by their general practitioner 7 . Similarly, in the United Kingdom, a 24 hour, seven day a week, service has been set-up to manage patients who "do not require immediate admission" to the hospital", at home 5 . Yet, these solutions rely heavily on in-person management of patients and will still place an important workload on general practitioners and other care professionals involved. Our proposal is to exploit digital technologies to remotely monitor patients at home. Data from remote monitoring of patients are transmitted to a Command Centre, in the nearest reference hospital, and automated algorithms triage patients with early warning signs. This will spare human time and let physicians focus on patients who may need specific care actions ( Figure, Supplementary Material 1). In addition to reducing contacts with care professionals and risks of contamination, remote monitoring of patients and automated decisions may alleviate the workload of caregivers and delay the expected disorganization of care structures. In a few days, public health researchers and clinicians from the Assistance Publique Hôpitaux de Paris and University of Paris, software architects and developers co-constructed a prototype for the remote monitoring platform (Supplementary Material 2) . Open-source code of all software components required to deploy the remote monitoring platform and Command Centres was developed by Lifen and is available at: https://github.com/lifen-labs/covid. Patients with confirmed Covid-19 are assessed by clinicians (in hospital or in community) for: 1) absence of initial signs of severity (based on their age, comorbidities, initial presentation of the disease); 2) their ability to be quarantined at home (e.g., absence of a psychiatric disorder or of a loss of autonomy); and 3) their ability to perform the remote monitoring at home (e.g., basic computer literacy, smartphone availability). If all criteria are fulfilled and if the patient consents for remote digital monitoring, he is sent home with instructions for quarantine 8 . Information is automatically sent to the patient's general practitioner, informing them that one of their patients has been confirmed with Covid-19 and is now being monitored at home. Remote follow-up of patients at home was designed to be minimally disruptive. It consists of a self-reported questionnaire, once or twice a day. Patients receive a text-message with a direct secure link to an online questionnaire. No login is required. Questionnaires involve <10 items and collect self-reported symptoms with validated tools (e.g., temperature, dyspnoea, pain) and quarantine information (e.g., psychological state regarding the quarantine and the disease, change of the people who are at home with them). In case of emergency, patients can contact the Command Centre or the National emergency number. Command Centres are located in hospitals and involve human personnel, including physicians and nurses who will analyse the constant influx of information from the remote monitoring. These people are equipped with real-time and decision-support tools, and assess whether patient care needs to be modified. They take necessary actions (e.g., intensifying monitoring, sending medical assistance, calling the patient for reassurance etc.) or provide feedback to patients and General Practitioners. Each time a patient sends new information by completing a self-reported questionnaire, their data are updated in real time. Automatic algorithms flag patients, using pre-defined decision rules, in four categories: -"Green" patients are asymptomatic and have no problems with the quarantine. No action from clinicians is required for these patients. Automatic messages are sent to reassure patients and to remind them to continue completing the regular questionnaires. -"Yellow" patients are stable with no signs of severity. No action from clinicians is required for these patients. Automatic messages are sent to reassure patients and to remind them to continue completing the regular questionnaires. -"Orange" patients are those with a recent change of symptoms and who may require closer monitoring. When a patient is flagged "Orange", the frequency of questionnaires is increased. In addition, an action is required from clinicians. -"Red" patients are those with rapid evolution of symptoms, signs of severity and/or those who have problems with the quarantine. Rapid action is required from clinicians. -"Patients who did not complete the questionnaire after 8h". Patients are highlighted and are called by the Command Centre. Whenever new data is received at the Command Centre, a summary of the patient status and of decisions taken is sent to their GP. The platform was designed to be compatible with nationwide deployment, in various size hospitals. The remote digital monitoring platform was also envisioned to evolve according to the evolution of the epidemic. First, the platform could be modified for other purposes, such as estimating the number of patients developing the disease by using simple self-reported questionnaires to identify the moment when asymptomatic contact subjects develop the disease. Secondly, the platform was thought to be compatible with the potential enrichment of patientreported information with data from biometric monitoring devices, such as pulse oximeters (these could be either provided to patients or via patients' own smartphones as in a Bring Your Own Device approach 9 ). This could help further pinpoint patients requiring hospital care and spare precious hospital beds. Thirdly, our platform could be integrated in a larger ecosystem of Covid-19 management integrating the remote follow-up of patients using different communication channels (e.g., telephone, smartphone application, etc.) according to the age, ability of the patient to use a smartphone, or severity of the disease etc. Finally, the remote digital monitoring of patients will provide continuous streams of data which will automatically update interactive dashboards about the local or regional state of the epidemic and about the capacity of the different Command Centres. In addition, data from all command centres can be pooled to provide real-time data visualizations on the state of the epidemic at national level or to constitute a research database. Figure. Key Obtained daily through a smartphone-app OR by phone (for people unable to fill questionnaires on a smartphone) Pulse oximeter (either by using specific sensor supplied or using a smartphone application) Data for surveillance of Covid-19 at local, regional or national level and research Clinical Characteristics of Coronavirus Disease 2019 in China Novel coronavirus (COVID-19) Situation 2020 Preventing a covid-19 pandemic Early Transmission Dynamics in Wuhan, China, of Novel Coronavirus-Infected Pneumonia The reproductive number of COVID-19 is higher compared to SARS coronavirus jusqu'à 850 000 personnes infectées et 50 000 morts Lignes directrices pour la prise en charge en ambulatoire des patients Covid-19 sans critère de gravité Paris: Ministère des solidarités et de la Santé Reliability of smartphone measurements of vital parameters: A prospective study using a reference method We thank greatly Prof Xavier Lescure (Infectious Diseases), Prof Enrique Casalino (Emergency Medicine) and Dr François Grolleau (Intensive Care Medicine) who helped us build the clinical questionnaires and decision rules. We also thank Dr Youri Yordanov (Emergency Medicine) for useful discussions about this paper. We thank Elise Diard for her help in drafting figures. Franck le Ouay, Etienne Depaulis and Alexandre Huckert are co-founders of Lifen. Bruno Vegreville is employed by Inato. Philippe Ravaud holds shares in Inato.