Article Text

PDF

Original article
Tracking surgical day care patients using RFID technology
  1. L S G L Wauben1,2,3,
  2. A C P Guédon1,
  3. D F de Korne4,5,6,7,
  4. J J van den Dobbelsteen1
  1. 1Department of BioMechanical Engineering, Faculty of Mechanical, Delft University of Technology, Maritime and Materials Engineering, Delft, The Netherlands
  2. 2Faculty of Industrial Design Engineering, Delft University of Technology, Delft, The Netherlands
  3. 3Rotterdam University of Applied Sciences, Research Centre Innovations in Care, Rotterdam, The Netherlands
  4. 4Rotterdam Eye Hospital, Rotterdam Ophthalmic Institute, Rotterdam, The Netherlands
  5. 5Erasmus University Rotterdam, Institute of Health Policy & Management, Rotterdam, The Netherlands
  6. 6Singapore National Eye Centre, SingHealth, Singapore, Singapore
  7. 7Duke-NUS Graduate Medical School, Health Services & Systems Research, Singapore, Singapore
  1. Correspondence to Research Professor LSGL Wauben, Department of BioMechanical Engineering, Faculty of Mechanical, Maritime and Materials Engineering, Delft University of Technology, Mekelweg 2, Delft 2628 CD, The Netherlands; L.S.G.L.Wauben{at}tudelft.nl

Abstract

Objective Measure wait times, characterise current information flow and define requirements for a technological information system that supports the patient's journey.

Design First, patients were observed during eight random weekdays and the durations of actions performed at each phase of the surgical trajectory were measured. Patients were grouped into patients receiving general anaesthesia or local (or topical) anaesthesia. Second (active) Radio Frequency IDentification (RFID) technology was installed and patients were tracked during 52 weekdays. Length of hospital stay, length of stay and wait times per phase, and differences in wait times between the two types of administered anaesthesia were analysed. Third, interviews were conducted to characterise the current information flow between staff, and between staff and escorts (patients’ family/friends escorting them throughout their journey).

Results Observations (198 patients) showed that the average duration of actions for general anaesthesia patients took longer than for local anaesthesia patients, especially at the recovery phase (general anaesthesia: 0h16, local anaesthesia: 0h01).

RFID tracking (622 patients): Significant differences were seen for wait times between general and local anaesthesia patients at: preoperative ward (p=0.014), recovery (p<0.001) and postoperative ward (p<0.001). The average percentage of wait time during the entire hospital stay ranged from 64% to 68% (with variation in groups being substantial).

Interviews (30 escorts, 9 ward nurses and 8 holding/recovery nurses): Escorts did not use the current information system and ward nurses indicated problems with exchanging information concerning bringing/picking up patients to/from the holding/recovery that resulted in unnecessary wait times for some patients (mainly local anaesthesia patients).

Conclusions Most time spent in hospital is wait time. A Patient Tracking System was designed to automatically display the phase in which a patient is in. It provides transparency for patients and staff in the surgical trajectory and is expected to reduce intermittent communication, improve patient flow, reduce wait times and improve patient and staff satisfaction.

  • Assistive Technology
  • Inventions
  • Ophthalmology
  • Remote monitoring
View Full Text

Statistics from Altmetric.com

Introduction

In the next 10–15 years, the number of patients requiring eye care and eye surgery will increase.1–3 Meanwhile, besides focusing on clinical issues, hospitals are also urged to focus on social and organisational issues, such as service and patient satisfaction.1 ,4 ,5 Previous research has shown that patient dissatisfaction is mostly related to non-clinical aspects, such as long (recurrent) wait times, lack of information about the clinical process and its predictability and wait times, inattentiveness/unresponsiveness of staff and the physical environment.4 ,6–8

To cope with the growing demand for surgical care and to manage the increasing healthcare costs, hospitals need to focus on patient satisfaction and on operational efficiency, particularly in the surgical trajectory where the costs are the highest.1 ,3 ,7 ,9–11 Recently, the potential of Radio Frequency IDentification (RFID) technology is being explored to facilitate healthcare processes, improve its efficiency and improve patient safety.12 RFID technology can be used to uniquely identify and localise objects, for example, medical asset/equipment tracking, patient/staff identification and workflow tracking, anticounterfeiting and medication safety.13–19 Data generated by RFID systems can also provide valuable information to improve the efficiency of processes in the surgical trajectory, reduce wait times, improve nurse allocation and improve patient flow.7 ,18 To improve patient satisfaction, patients’ expectations on wait time should be met by means of providing accurate information and realistic estimates of wait times and steps in the patient's journey.8 ,18 Data generated by RFID systems can be used to inform patients, as well as staff, about their progress in the surgical trajectory.

A first step to improve the efficiency of the surgical trajectory and to improve patient satisfaction is to acquire insight on the current patient and information flow. Therefore, the objective of this study is threefold: (1) measure wait times for families and patients undergoing eye surgery during surgical day care, (2) characterise current information flow between staff and patients, and between staff from different departments, and (3) define the requirements for a technological information system that manages the patient's journey.

Methods

This study was conducted at the main surgical centre (including four operating rooms—ORs) in the Rotterdam Eye Hospital and was divided into three parts: observations, RFID tracking and interviews.

Observations

Adult patients admitted for surgical day care were followed and observed during eight random weekdays. Children (<18 years), emergency patients and clinical patients (ie, patients who need to stay during the night) were excluded. First, five adults were shadowed to get an insight on the surgical day care trajectory. Each patient went through 12 phases after registering at the hospital, with each phase representing a specific location: (1) waiting room, (2) intake room, (3) waiting room, (4) dressing room, (5) (day) ward I, (6) holding, (7) OR, (8) recovery, (9) (day) ward I or II, (10) dressing room, (11) waiting room and (12) checkout room. At nine phases (excluding the waiting room), actions were performed by staff or patients (eg, changing clothes, handover, time-out, administering medication, performing surgery). Duration of actions performed at these nine phases was recorded by two researchers (one stationed at the ward and one at the surgical centre). No identifiable patient data or data on the surgical procedure were collected; information was collected only on the type of anaesthesia used, and the time of arrival and departure at the OR. The latter were obtained from the hospital information system (these times are manually recorded by the nurse anaesthetist at the OR).

Patients were grouped based on the type of anaesthesia administered: general anaesthesia (GA) versus local or topical anaesthesia (LTA) as the type of anaesthesia especially influences the recovery in the postoperative phases (recovery and postoperative ward). Durations of actions performed at the ward (preoperative and postoperative), holding, OR and recovery were calculated.

RFID tracking

RFID technology was used to automatically track the patient's location and measure the length of stay per phase. Adult patients admitted for surgical day care were tracked during 52 successive weekdays using active RFID technology. Again, children, emergency patients and clinical patients were excluded. The active RFID tag (pulse rate 0.8, frequency 433.92 MHz, power 1 mW, weight 24 g) was attached to the patient's wristband (see figure 1A), and was tracked by readers which were placed at eight locations shown in figure 1B. The readers (GW3D, RePoint, the Netherlands) and controllers (to store the data locally) were integrated in the ceiling and connected to the hospital's existing wired network. The location of the tags was determined by its signal strength as multiple nearby readers could detect the signal. Rough data were pushed and stored at the stand-alone server (Dell OptiPlex 790) that was placed at the nursing station. At the end of the research period, data were collected and analysed.

Figure 1

(A) Active RFID tag attached to the patient's identification wrist band. (B) Layout of the surgical trajectory and location of the RFID readers. RFID, Radio Frequency IDentification; OR, operating room.

Patients received the RFID tag at the registration desk and the nurse at day ward (=ward nurse) collected the tag during the checkout meeting. After use, the tags were cleaned with alcohol and could be used again (on the same day) to track other patients. Again, no patient data or data on the surgical procedure were collected; only data on the type of anaesthesia, and time of arrival and departure at the OR were collected. Patients were grouped based on the type of anaesthesia (GA vs LTA).

Standard descriptive statistical methods were used to generate length of hospital stay and length of stay per phase (ie, preoperative ward, holding, OR, recovery and postoperative ward) using IBM SPSS Statistics V.20 for Mac. Additionally, wait times per phase were calculated. For the recovery and postoperative ward ‘wait-recovery time’ was calculated as wait time and recovery time (recovering from surgery and anaesthesia) could not be separated. Mann-Whitney U tests were performed to calculate significant differences in wait times per phase between the two types of anaesthesia.

Interviews

Escorts accompanying patients, ward nurses and nurses from the holding/recovery (=holding/recovery nurses) were interviewed during the RFID tracking part. We have chosen to interview the escorts instead of the patients, as the latter have to wait during the entire trajectory and we did not want to disturb the patients. The escorts were interviewed 5–10 min after the patient left the ward. Nurses were interviewed during breaks or off-peak moments. One researcher used a semistructured approach and asked open questions concerning: their previous experience with the hospital, current information flow between staff and patients/escorts, current information flow between ward nurses and holding/recovery nurses (using the current information systems), and their desired future requirements. The interviews took a maximum of 15 min and notes were taken.

The current information system used at the ward is a magnetic whiteboard, which is placed across the registration desk. Coloured cards (male/female/child), including name and type of anaesthesia, are placed in different columns representing the different locations/phases. When a patient moves to a different phase the card is moved accordingly. The current information system used at the holding and recovery is a printed OR schedule on which the nurses mark the progress of an individual patient using highlighters.

Results

Observations

In total, 198 patients were included in the observations (137 GA, 61 LTA). Table 1 shows the duration of actions per phase. The average duration of actions performed at the OR, recovery and postoperative day ward took longer for GA than for LTA.

Table 1

Observations: duration of actions [h:mm] per phase

All patients followed the same trajectory along eight locations. Based on this set trajectory, these locations represented in figure 1B were selected for placement of the readers for the RFID tracking part of this study.

RFID tracking

In total, 829 patients admitted for surgical day care received a tag. However, 207 patients were excluded as the tag was not detected by the reader in the OR corridor (n=154), type of anaesthesia was unknown (n=29), the patient did not wear or had removed the tag (n=20), or the recorded OR time was less than a minute (n=4) indicating a technical flaw. In total, 622 patients (=75.0%; 405 GA, 217 LTA) were included in the analysis.

In line with hospital policy, patients were asked to arrive and register an average of 2 h before the planned surgery. Table 2 shows that in practice, 66.9% (n=271) of GA patients and 59.0% (n=128) of LTA patients arrived early compared with the planned arrival time, on average 00:18 and 00:24 early, respectively.

Table 2

Early and late: arrival of patients [h:mm:ss], time spent in the hospital, start of surgery and start of first surgery

Surgery performed under GA took between 00:17 and 3:59, an average of 1:05 (SD 00:33, median 00:57). Table 2 shows that surgery performed under GA started late in most cases (n=312). Twelve first case surgeries started more than 1 h late, which was caused by adding patients to the OR schedule (n=3), changing the order of patients (n=2), or for no specific reason (n=7). Surgery performed under LTA took between 00:06 and 2:16 with an average of 00:35 (SD 00:17, median 00:31) and started late in most cases (n=139). Six first case surgeries started more than 1 h late, which was caused by adding patients to the OR schedule (n=2) or for no given specific reason (n=4).

On average, GA patients spent 7:01 in hospital and LTA patients 4:17 (table 2). Figure 2 shows the wait times and wait-recovery times per phase. Mann-Whitney U tests showed significant differences in wait times between GA and LTA at the preoperative ward (p=0.014), at the recovery (p<0.001) and at the postoperative ward (p<0.001). No significant differences were found at the holding (p=0.0496). For GA patients the total percentage of wait and wait-recovery time during the entire hospital stay ranged from 0% to 87.0% with an average of 68.2%. For LTA patients, this ranged between 20.8% and 85.7% with an average of 64%.

Figure 2

Boxplot summaries for wait time and wait-recovery time per phase (median, IQR, minimum and maximum values, o=outlier, *=extreme case) and average, SD, median and number of patients not having to wait per phase.

Interviews

In total, 30 escorts, 9 ward nurses (out of 15) and 8 holding/recovery nurses (out of 10) were interviewed.

Escorts

Most patients’ escorts (n=23) had previous experience with the hospital and all escorts felt comfortable asking the nurse(s) questions. Although 13 escorts noticed the whiteboard, only one used it. Eighteen escorts received information on the duration of the surgical procedure and the arrival time at the postoperative ward, and eight escorts received information on what time to go home. In the future, most escorts would like to be informed about: progress at the surgical centre (n=19), arrival time at the postoperative ward (n=22) and general information about the surgical procedure (n=16). Twenty-one escorts would prefer a public screen in the waiting room to a personal device to portray the progress information. They did not have any privacy concerns related to this public screen and did not mind their names being visible to other patients and escorts.

Day ward nurses

Most problems experienced concerned the holding asking the ward to bring a patient (n=7) or the recovery asking to pick up a patient (n=6). Furthermore, nine nurses indicated that the whiteboard is not updated regularly. In the future, they would like to be informed (via a technological information system) about: registration of the patient (n=7), intake meeting conducted (n=9), patient ready for the holding (n=8), patient ready to be picked up from the recovery (n=8) and patient ready for checkout (n=8).

Holding/recovery nurses

Only few problems arose related to the patient flow and the paper OR schedule: only one nurse indicated that the schedule is not marked when a patient is requested from the ward or has arrived at the recovery. The OR schedule is always marked when the patient arrives at the holding. In the future, most nurses would like to be informed (via a technological information system) about: patient on their way to the holding (n=7), and ward nurses on their way to pick up the patient at the recovery (n=6). Six nurses also indicated that a digital information system could replace the phone calls between the holding/recovery and the ward.

Discussion

This study showed that both observations and RFID tracking are practical tools to measure wait times for patients undergoing eye surgery during surgical day care. However, using RFID has the advantage that tracking and time recording are performed automatically and in real time. The results showed that wait times were long; on average 66.7% of the entire hospital stay was wait and wait-recovery time, and most patients had to wait in each phase of their surgical journey.

Significant differences were found between GA and LTA for wait times at the preoperative and postoperative ward, and at the recovery. No specific reason could be found for the differences at the preoperative ward; these were not caused by the outliers or by the LTA patients arriving early. The difference in wait and recovery time in the postoperative phase is largely caused by the time to recover from surgery and anaesthesia (at the recovery and at the ward). Patients receiving LTA, basically do not need to recuperate at the recovery and can go straight back to the ward. The postoperative trajectory is expected to affect the patient satisfaction less as here the patients and escorts play an important role as indicated by themselves when they are ready to leave the hospital.20

For patients, wait time starts once they arrive at the hospital. From the moment the patients register until the start of the surgery, the patients expect to wait the indicated time (anticipated wait) or less: in this case an average of 2 h, including the total time spent at the preoperative ward and at the holding. In this study, 69.5% of patients (295 GA, 137 LTA) had to wait longer than anticipated, which reduces patient satisfaction.5 ,8 The longer wait was partly caused by patients arriving early, surgery starting late, sporadic communication between the ward nurses and the patients/escorts, and intermittent information exchange between the ward and the surgical centre. The latter was supported by the observations and discussions with hospital’s management, showing that currently most ward nurses experienced information problems related to bringing and picking up patients from the surgical centre. It also revealed that the phone calls between the ward and the surgical centre were redundant and disruptive for the ward nurses (although they are so used to these disruptions that they consider it as the normal way of working). Furthermore, the ward nurses also found questions by the escorts concerning the patient's progress disruptive. Overall, this leaves the ward nurses less time to perform their clinical tasks with concentration.

Real-time information about the patient flow can support communication between departments concerning transfer of patients and can help nurses to better anticipate and/or to automatically reschedule the surgical procedure to limit long wait times. Providing real-time realistic information about wait times and providing reasons for delays could also improve patient satisfaction with wait time.5 ,6 ,18 ,20 ,21 Shaikh et al22 have shown that most respondents would prefer a display with a time tracker to provide information about their wait time when visiting the emergency department. A display, automatically presenting the phase of the patient to the escorts and the estimated wait times, could also reduce the number of questions concerning the patient's progress, stimulate active involvement and actions of the patients/escorts (eg, go to the intake room themselves without a nurse assisting them) and reduce anxiety.13 ,20 ,22 Kim et al23 have shown that mean wait times were shortened when patients were automatically allocated to examination rooms; this also increased workflow efficiency by reducing staff effort and consequently, reducing costs. However, automatic presentation of the patient's phase first requires a technological system. RFID technology is already used in the healthcare domain12 ,13 ,15 ,16 ,18 ,19 ,23 and this study showed that RFID is able to record and show real-time data on the patient's location and time spent in the different phases. Although the technology can be designed in such a way that its influence on daily routine is limited, the organisation, its working routines and protocols have to change as well.13 In order for such a system to be used and adopted, the system should be designed by actively involving staff in designing an intuitive and simple system that relieves staff from redundant tasks, and fits in with the particular context and workflow.13 ,14 ,18 ,19 ,24 ,25

Patient Tracking System

Based on the results of this study, a ‘Patient Tracking System’ was designed in close cooperation with the ward nurses, patients/escorts and the hospital's management. Figure 3 shows the user interface of the Patient Tracking System that replaces the whiteboard at the ward and the extra display that will be placed in the waiting room. The Patient Tracking System automatically displays the phase in which a patient is in, and in the near future, this system will also empower patients by including predictions about, for example, when the patient can get dressed to go to the holding or when the patient can go home. The aim of the Patient Tracking System is to provide transparency for patients and staff into the surgical trajectory. The Patient Tracking System is expected to reduce intermittent communication between departments, improve the efficiency of the process between the ward and the holding/recovery, reduce wait times, and improve patient and staff satisfaction.

Figure 3

User interface of the Patient Tracking System. The patient cards include the patient's name, the time a patient arrived in a specific phase and a coloured dot representing the responsible ward nurse for that specific patient.

During this project, we also encountered some organisational and technical challenges. First, 84% of tags were not returned and were lost, which is high compared to Stahl et al,16 who only lost 5% of tags. Potential reasons for this loss were inattentiveness of the staff, unclear instructions to the staff, unawareness of the costs and reuse of the tags, and the collection process not being integrated into daily routines, protocols and checklists. Second, signals transmitted by the tags were read through the walls or were not seen by the reader in the OR corridor, which is a common problem in RFID tracking.14 ,15 For the newly developed system, using the tags as readers as well as transmitters solves these problems. This increases the number of readers and thereby, the accuracy and reliability of tracking.

This study was limited by excluding children, emergency patients and surgical procedural data. However, we deliberately excluded these data as we wanted to demonstrate the benefits of the RFID system first without changing the current working routine too much. Based on the results of this study, all patients’ routes have been standardised, enabling us to include all patients. The tracking data were not yet used to immediately improve the communication between the departments or reduce wait times; it only provided indirect information. However, the Patient Tracking System, which is now implemented in the hospital, directly informs the patients and nursing staff. The next step is to change routines, for example, by requesting patients to arrive earlier than the 2 h prior to surgery and ward nurses automatically collecting LTA patients from the recovery once the patient enters this phase in the Patient Tracking System.

Acknowledgments

The authors would like to thank Repoint BV for providing the RFID hardware and the design of the user interface of the Patient Tracking System, LMA Vankan, MSc, for her assistance during the observations, and JV Sluiman, MSc, for his assistance during the observations, interviews and the design of the user interface of the Patient Tracking System.

References

View Abstract

Footnotes

  • Funding This work was supported by the ‘Provincie Zuid Holland’, Project ID: CRZH101005.

  • Competing interests None.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.