Modeling and estimating the spatial distribution of healthcare workers

Abstract

This paper describes a spatial model for healthcare workers' location in a large hospital facility. Such models have many applications in healthcare, such as supporting timeand-motion efficiency studies to improve healthcare delivery, or modeling the spread of hospital-acquired infections. We use our model to estimate spatial distributions for healthcare workers in The University of Iowa Hospitals and Clinics (UIHC), a 700-bed comprehensive academic medical center spanning a total of 3.2 million square feet and employing about 8,000 healthcare workers. We model the UIHC as a metric space induced by walking distance between pairs of rooms, and with each room having a level of attractiveness representing the activity level in that room. We combine this with a model in which each healthcare worker has a center of activity and a probability density function that decays polynomially as we move away from the center. Using 12 million Electronic Medical Record (EMR) logins collected over 22 months, we solve for the model parameters for each room and each healthcare worker using heuristic techniques to make the problem computationally tractable. We then validate the model parameters obtained by comparing realworld expectations of healthcare worker behavior for several job categories to our model predictions (e.g., we verify that Unit Clerks are much more stationary than Respiratory Therapists). Finally we present solutions to two important applications. First, we use healthcare worker spatial distributions to generate random walks representing their movement through the hospital. We use these random walks to construct healthcare worker contact networks in order to study the spread of hospital-acquired infections. Second, using the healthcare worker spatial distributions, we find a near-optimal placement of hospital resources which minimizes the average distance a healthcare worker has to travel to access that resource.

Computational applications to hospital epidemiology

“…One of the questions we wanted to answer was whether knowing when workers visit rooms can be enough to predict their interactions (i.e., close proximity contacts). In a sense, this work was meant to validate previous work done in the compepi group in which contacts between healthcare workers were estimated by the spatial and temporal proximity of their respective computer logins, creating login networks [18,17,16]. We confirmed this hypothesis by being effectively able to predict contacts from visits to rooms.…”

“…Due to the unavailability of data, early studies used synthetic network models to study the spread of infection [79,56,78]. In spite that contact network epidemiology was proposed more than 30 years ago [45], it has been only in the recent years (mostly since 2011) that research groups started measuring detailed contact network data to study the transmission of infection [118,50,5,74,59,4,75,120,36,19,60], an effort joined by the compepi group [18,17,102,81,48,82,16,83].…”

“…In the worst cases, CDI may turn into toxic megacolon (the colon enlarges dramatically due to inflaction), which may require surgery, and even death. For this project, I extended an existing database which contained clinical, architectural, and computer usage data, that was initiated by previous members of the compepi group [18,17,102,16].…”

Tracking Nosocomial Diseases at Individual Level with a Real-Time Indoor Positioning System

“…They showed that healthcare workers are main transmitters of diseases and shall be vaccinated with a higher priority. Curtis et al [34] modeled dynamic contact networks by deriving spatial distributions of healthcare workers and generating random walks to predict human movements in a hospital. Prakash et al [15] constructed a time-varying network which follows an alternating connectivity behavior to model the day-night pattern of nurse shifts and derived a closed-form equation for the epidemic threshold with their network.…”

Nosocomial infection modeling and simulation using fine-grained healthcare data

“…Recent work suggests that healthcare worker contact graphs may indeed exhibit small world properties [61,29,28], so in this thesis we tend, in the limit, towards agentbased infectious disease simulations (e.g., one compartment for each agent in the simulation rather than one compartment for the entire population). We find that healthcare worker contact networks based on our agent-based simulator do indeed exhibit classic "small world" properties [117,87], and thus agent-based simulation is critical for understanding and controlling the spread of hospital-acquired infections such as "Clostridium difficile" (C. diff), methicillin-resistant "Staphylococcus aureus" (MRSA), or vancomycin-Resistant "Enterococcus" (VRE).…”