The COVID-19 pandemic is challenging the healthcare system capacity all around the world. Allocating ICU beds, ventilators and medical staff are crucial and important issues in this pandemic. Progressive and intensive care units had to be built in a matter of weeks. Quick deployment of materials, personnel, equipment, and also ICT infrastructures is and will be key in these situations.

What it does

Bluebeds is a flexible and customizable wireless remote monitoring solution, that aims to connect any kind of medical device.

With our medical device kits it is possible to monitor patients in real time that are not in the ICU. This can prevent ICU hospitalization that are not necessary.

The idea is also to create a predictive model that advise you on monitored patients that will need an ICU.

The platform will require on admission the compilation of an assessment based on Brescia model. The answers will give a complexity index of the patient. Our prediction model use this informations to predict the estimate of the length of stay of the patient. So, we can anticipate when the bed will be free.

On the long run we want to:

  • reduce healthcare costs by cost-effectively deploying remote patient monitors to non-ICU/non-PCU departments and integrating already available resources
  • increase patient safety by alerting on patient deterioration
  • reduce the workload of nursing staff and prevent "alarm fatigue"
  • contribute data to future studies

How we built it

We started to develop the web platform considering the implementation on cloud and on-premise. We used responsive design to support the use of mobiles and tablets. Medical devices are connected to an IoT infrastructure and through a mqtt broker all the data is transferred in real time to the web application. We also developed a solution to solve the issue of being able to associate medical devices with individual patients, so mixing up devices can be prevented.

Challenges we ran into

The biggest challenge was to explain the project and align everyone to a level of autonomy, especially for those who have no experience in the healthcare sector.

Accomplishments that we are proud of

During this hackathon we managed to implement the predictive model API on the length of stay for ICU beds with the integration into the platform. We have given a corporate brand to the start-up and we have improved some functional details of the project.

What we learned

To confront with different professional figures Regulations and certification processes for software and medical devices Studies about "avoidable hospitalizations"

What's next for Bluebeds

Currently acute care monitoring is often limited to spot checks of physiologic parameters and clinical assessment performed 4-6 times a day. As the bedside nursing ratio is often 4:1 to 6:1 or greater on the acute care ward or other intermediate care units (versus 2:1 in an ICU), there is a substantial negative impact of non-useful or non-actionable alarms. Due to the imbalance of bedside nursing ratio and the number of non-actionable alarms, nursing staff can get accustomed to ignoring alarms and in doing so may overlook alarms in a true emergency situation. This phenomenon is known as ‘alarm fatigue’.

We want to scale our project with providing patient monitor devices that conform the HL7 international standards and are able to wirelessly connect to our platform for non-ICU wards. This will allow such wards to continuously monitor patients cost effectively. With machine learning added, nurses with less experience in emergency care settings will be able to get better predictions on patient status, and initiate the early placement of patient to an intermediate or progressive care unit.

We want to continue the project after the hackathon, we have been accepted to the Innosuisse coaching program, we are applying to #VersusVirus hackathon funding program. Initial funding will allow us to buy medical devices and continue the implementation of the platform. We plan to work on interoperability, continue the development of the predictive models with realtime patient monitoring data, and reach potential customers to start a clinical trial.


  1. Drew BJ, Harris P, Zègre-Hemsey JK, et al. Insights into the problem of alarm fatigue with physiologic monitor devices: a comprehensive observational study of consecutive intensive care unit patients. PLoS ONE. 2014;9:e110274.
  2. Wung SF, Schatz MR. Critical care nurses’ cognitive ergonom-ics related to medical device alarms. Crit Care Nurs Clin N Am. 2018;30:191.
  3. Winters BD, Cvach MM, Bonafide CP, et al. Technological distractions (Part 2): a summary of approaches to manage clini-cal alarms with intent to reduce alarm fatigue. Crit Care Med. 2018;46:130–7.
  4. Curry JP, Jungquist CR. A critical assessment of monitoring prac-tices, patient deterioration, and alarm fatigue on inpatient wards: a review. Patient Saf Surg. 2014;8:1–20.
  5. Min A, Scott LD. Evaluating nursing hours per patient day as a nurse staffing measure. J Nurs Manage. 2016;24:439–48.
  6. Ding, X., Li, J., Liang, H. et al. Predictive model for acute respiratory distress syndrome events in ICU patients in China using machine learning algorithms: a secondary analysis of a cohort study. J Transl Med 17, 326 (2019).
  7. Fernandes M, Mendes R, Vieira SM, Leite F, Palos C, Johnson A, et al. (2020) Predicting Intensive Care Unit admission among patients presenting to the emergency department using machine learning and natural language processing. PLoS ONE 15(3): e0229331.
  8. Brekke IJ, Puntervoll LH, Pedersen PB, Kellett J, Brabrand M (2019) The value of vital sign trends in predicting and monitoring clinical deterioration: A systematic review. PLoS ONE 14(1): e0210875.
  9. Navpreet K. Dhillon, MD,1 Ara Ko, MD, MPH,1 Eric J. T. Smith, BA,1 Mayumi Kharabi, RN, MSN,2 Joseph Castongia, MHA,2 Michael Nurok, MD,1 Bruce L. Gewertz, MD,1 and Eric J. Ley, MDcorresponding author1 Potentially Avoidable Surgical Intensive Care Unit Admissions and Disposition Delays
  10. Goldhill DR1, Worthington L, Mulcahy A, Tarling M, Sumner A. The patient-at-risk team: identifying and managing seriously ill ward patients.
  11. Zografakis-Sfakianakis M1,2,3, De Bree E1,2, Linardakis M1, Messaritaki A2, Askitopoulou H1, Papaioannou A1,2, Aggouridakis P1,2. The value of the Modified Early Warning Score for unplanned Intensive Care Unit admissions of patients treated in hospital general wards.
  12. Kerlin, MD, MSCE1,2,3; Yihao Yuan, MS4; Rachel Kohn, MD, MSCE1,2,3; George L. Anesi, MD, MBE, MSCE1,2,3; Peter W. Groeneveld, MD, MS3,4,5; Rachel M. Werner, MD, PhD3,4,5; Scott D. Halpern, MD, PhD1,2,3 Potentially preventable intensive care unit admissions in the United States, 2006 - 2015

Built With

+ 4 more
Share this project: