An ontology-based nurse call management system (oNCS) with probabilistic priority assessment

Abstract

Background: The current, place-oriented nurse call systems are very static. A patient can only make calls with a button which is fixed to a wall of a room. Moreover, the system does not take into account various factors specific to a situation. In the future, there will be an evolution to a mobile button for each patient so that they can walk around freely and still make calls. The system would become person-oriented and the available context information should be taken into account to assign the correct nurse to a call.The aim of this research is (1) the design of a software platform that supports the transition to mobile and wireless nurse call buttons in hospitals and residential care and (2) the design of a sophisticated nurse call algorithm. This algorithm dynamically adapts to the situation at hand by taking the profile information of staff members and patients into account. Additionally, the priority of a call probabilistically depends on the risk factors, assigned to a patient.

Methods: The ontology-based Nurse Call System (oNCS) was developed as an extension of a Context-Aware Service Platform. An ontology is used to manage the profile information. Rules implement the novel nurse call algorithm that takes all this information into account. Probabilistic reasoning algorithms are designed to determine the priority of a call based on the risk factors of the patient.

Results: The oNCS system is evaluated through a prototype implementation and simulations, based on a detailed dataset obtained from Ghent University Hospital. The arrival times of nurses at the location of a call, the workload distribution of calls amongst nurses and the assignment of priorities to calls are compared for the oNCS system and the current, place-oriented nurse call system. Additionally, the performance of the system is discussed.

Conclusions: The execution time of the nurse call algorithm is on average 50.333 ms. Moreover, the oNCS system significantly improves the assignment of nurses to calls. Calls generally have a nurse present faster and the workload-distribution amongst the nurses improves.

Figure 1

The traditional place-oriented and static nurse call system vs. the future person-oriented and context-aware approach. The architecture of traditional nurse call systems can be viewed in the left part of the figure. Each room has at least one button which can be used by the patient to call a nurse. All the buttons in a room are connected to a Node. All the Nodes of a department are connected with each other and a Controller. The Controller has the intelligence to know what must happen when a call is made, for example which nurses must be called. A PC can be used to configure the controller. The nurses possess beepers or portable phone on which they can receive calls. Within a department, the Nodes can be further divided into different, possibly overlapping, nursing groups. A nurse will only receive calls of the nursing groups that this nurse is assigned to. The proposed architecture of the person-oriented and context-aware nurse call system can be viewed in the right part of the figure. Each patient has a mobile button so that they can walk around freely and still make calls. These calls are picked up by the sensor network and processed by the Controller. The Controller calls a nurse to handle the call. The nurse receives the call on his or her PDA.

Figure 2

General concept of the oNCS platform with probabilistic risk assessment and profile management

Figure 3

Fragment of the ontology that models the context information about the staff members and patients. Fragment of the ontology which models the patients and staff members of the hospital who can answer calls. The squares represent the classes. The arrows with the striped lines indicate subclass relationships. The other arrows and lines indicate relations between classes (object properties).

Figure 4

Fragment of the ontology that models the context information about the characteristics and risk factors. Fragment of the ontology which models (1) the characteristics of the helpers and (2) the risk factors of the patients. To highlight the possibilities of the system, a (not exhaustive) list of risk factors was assembled by experts from both the medical and nurse call domain. The squares represent the classes. The arrows with the striped lines indicate subclass relationships. The other arrows and lines indicate relations between classes (object properties).

Figure 5

Fragment of the ontology that models the context information about the calls and tasks. Fragment of the ontology which models the calls and tasks. It mainly indicates which calls can be made by patients and staff members and which staff members are allowed to handle these calls. Additionally, it models the possible priorities that a calls or tasks can have. The squares represent the classes. The arrows with the striped lines indicate subclass relationships. The other arrows and lines indicate relations between classes (object properties).

Figure 6

oNCS algorithm to find the correct staff member to handle a call. This figure shows the flow chart of the oNCS algorithm, which finds a correct staff member to handle a call. It first determines which kind of calls has been made. Normal, sanitary, service and (sanitary) assistance calls employ the same basic algorithm which is visualized in Figure 7 (Flow A). The difference is that for normal, sanitary and (sanitary) assistance calls only nurses can be called. For service calls caretakers can also be called. It is also made sure that the nurse that made the (sanitary) assistance call, cannot be called to answer this call. Urgency, medical and technical calls each have their own algorithm, which is visualized in this figure.

Figure 7

Flow A: oNCS algorithm to find the correct staff member to handle a normal, sanitary, service or (sanitary) assistance call

Figure 8

The architecture of the oNCS platform. This figure represents the architecture of the oNCS platform. The Context Framework Layer is the most important layer. Within this layer the Context Interpreter controls all the context information. The ontology determines the structure of the Knowledge Base. The Knowledge Base contains all the data that conforms to the ontology. The Context Model provides access to the ontology by using Jena. Pellet is used to check the consistency of the model. The layer also holds all the Rules that work with the information in the Knowledge Base. The different Context Providers allow importing external information into the framework. This information is then added to the Knowledge Base. This new information can come from a database (Persistence Layer) or directly from a device (Device Layer and Context Gathering Layer). Currently three Context Providers are provided: the Person Provider, the Environment Provider and the Call Provider. The Query Services are used to extract information from the Knowledge Base. The Query Services can be used to visualize the knowledge or to use the information in another application (Application Layer). The methods in the Context Providers and Query Services were also made available as Web Services.

Figure 9

The floor plan of the studied department. This figure represents the floor plan of the studied department of the Ghent University Hospital. The department contains patients that are fairly mobile. The most important spaces to notice on the floor plan are the rooms and the sanitary areas. The department contains 26 beds. The floor plan indicates for each room how many beds it contains. Most rooms have their own sanitary space, but there are also some shared sanitary spaces. The nursing post is the place where nurses reside when they are not helping patients. This space is used to for example prepare medication or write reports. The head nurse has her own office. Patients do not have access to the storage and service spaces, the terrace, the rinse areas and the kitchen. The doors on the left and right of the floor plan are used to go to other departments. Generally patients use the elevator on the right of the floor plan to leave the department. The elevator in the middle of the floor plan is generally only used by staff members.

Figure 10

Distribution of time of the nurses across different kinds of tasks

Figure 11

Reasons for patients' call light use [41].

Figure 12

Number of calls as a function of the nurse arrival times. This figure shows the number of calls that have a nurse present (y-axis) as function of the arrival times of these nurses (x-axis in seconds) for both the oNCS system and current, place-oriented system. This means that the nurse has arrived at the place where the patient made the call. Note that the first part of the x-axis has a time-step of 5 seconds, while the second part has a time-step of 60 seconds. The two parts are separated by the striped vertical line.

Figure 13

Percentage of (sanitary) assistance calls as a function of the nurse arrival times. This figure shows the percentage of assistance and sanitary assistance calls (y-axis) as function of the arrival times of these nurses (x-axis in seconds) for both the oNCS system and the current, place-oriented system. This means that the nurse has arrived at the place where the patient made the (sanitary) assistance call. Note that the first part of the x-axis has a time-step of 5 seconds, while the second part has a time-step of 60 seconds. The two parts are separated by the striped vertical line.

Figure 14

Percentage of normal and sanitary calls as a function of the nurse arrival times. This figure shows the percentage of normal and sanitary calls (y-axis) as function of the arrival times of these nurses (x-axis in seconds) for both the oNCS system and the current, place-oriented system. This means that the nurse has arrived at the place where the patient made the normal or sanitary call. Note that the first part of the x-axis has a time-step of 5 seconds, while the second part has a time-step of 60 seconds. The two parts are separated by the striped vertical line.

Figure 15

oNCS system: number of calls as function of nurse arrival times for different call priorities. This figure visualizes the number of calls that have a nurse present (y-axis) as function of the arrival times of these nurses (x-axis in seconds) for different call priorities for the oNCS system. This means that the nurse has arrived at the place where the patient made the call. This allows evaluating (1) the influence of the priority of the call on the arrival time of the nurse (2) the distribution of the calls amongst the different priorities. Note that the first part of the x-axis has a time-step of 5 seconds, while the second part has a time-step of 60 seconds. The two parts are separated by the striped vertical line.

Figure 16

Place-oriented system: number of calls as function of nurse arrival times for different call priorities. This figure visualizes the number of calls that have a nurse present (y-axis) as function of the arrival times of these nurses (x-axis in seconds) for different call priorities for current, place-oriented system. This means that the nurse has arrived at the place where the patient made the call. This allows evaluating (1) the influence of the priority of the call on the arrival time of the nurse (2) the distribution of the calls amongst the different priorities. Note that the first part of the x-axis has a time-step of 5 seconds, while the second part has a time-step of 60 seconds. The two parts are separated by the striped vertical line.