For many health care providers, the COVID-19 pandemic offered vital takeaways: the importance of determination in the face of adversity, the impact of patient-centered care, or even just a reminder to practice hand hygiene. One team, however, used one of their lessons to expand a research project. The Computational Epidemiology (CompEpi) Group at the University of Iowa has been studying how contact patterns affect the spread of infections within healthcare facilities for more than a dozen years. CompEpi was founded by Phil Polgreen, MD, MPH, professor in Infectious Diseases, and Computer Science professors Alberto Segre, PhD; Ted Herman, PhD; and Sriram Pemmaraju, PhD. They have received funding from the NIH, NSF, private industry, and the CDC to develop and apply computational approaches for studying the movement of infections within healthcare settings.

In their most recent publication for Nature Scientific Data, Contact Observations from an Intensive Care Unit, members of CompEpi—Roger Struble MD, MPH, recent faculty addition in Pulmonary, Critical Care, and Occupational Medicine, Computer Science graduate student Hieu Vu, and Computer Science professor Bijaya Adhikari, PhD, as well as Polgreen and Herman—used mobile sensors to measure how healthcare workers move and interact in a busy intensive care unit. The goal of this study was to develop an approach to measure the movement of and interactions between healthcare professionals (and the environment) in an intensive care unit. For Struble, during the pandemic, observing patients converting from negative to positive COVID status “underscored how little we truly understand about the everyday movement of clinicians, staff, and equipment—and how those patterns of movement might facilitate the spread of pathogens.”

The CompEpi group has a long history of using sensor networks to model the spread of infections in intensive care units, dialysis centers, and long-term care facilities. However, in this latest work, the team used ultra-wideband enabled sensors, which can accurately measure distances within centimeters. Previously, at best, prior approaches could only detect motion on one side of the room or another. In this study, the deployed sensor network could detect if someone was on a particular side of a patient bed, at the sink or at the computer keyboard.

For this study, the team strategically placed sensors around the MICU and also distributed sensor badges to healthcare professionals working in the unit at the beginning of each shift. These sensors collected a huge amount of data; approximately 15 million records were collected from the sensor network over the course of a week in the MICU. These records were used to measure the time and duration of staff interactions spent in different locations. “A single 24-hour period revealed up to hundreds of transitions between sinks, monitors, keyboards, supply carts, and the patient’s bedside,” Struble said, “more than most of us guessed.” A key feature of this publication is that the team released de-identified mobility data to the scientific community, so that other research groups can use these unique data to help parameterize models to understand how infections spread in healthcare settings.

Struble says the success of the project—and the key to its future application—is a result of the cross-departmental and cross-college collaborators being “united behind a shared vision.” CompEpi, he said, helps “foster an environment where curious questions turn into exciting results, empowering investigators at every career stage.”

In the future, the team plans to couple their movement maps with microbiologic culture data and simulations to test targeted interventions to mitigate the spread of infections. Struble has already collected a second round of data to map even more granular movements within MICU rooms, to help “estimate the probability of transferring multidrug-resistant organisms from environmental ‘hot spots’ to patients.”

Struble stressed that this work would not be possible without a large number of people who supported the study. The entire staff of the MICU participated including nurses, clinicians, therapists and support staff. Shelby Francis, and the research interns in the Mobile Technology Lab were critical throughout the deployment. Finally, the project was supported by the CDC-funded Modelling Infectious Diseases in Healthcare Networks (MInD), which was awarded to the CompEpi group (MPIs Segre and Polgreen).

Animations produced by CompEpi in previous studies tracking movement of health care workers: