Vanderbilt University



Vanderbilt University

Department of Biomedical Engineering

Evaluation of Workflow and Accessibility of the Vanderbilt Children’s Hospital Neonatal Intensive Care Unit Data Entry System

John N. Fonge

Kendra S. Mills

Brandy C. Scott

Advisor: Dr. William Walsh, Chief of Nurseries Pediatrics Neonatology

Introduction

Vanderbilt University Children's Hospital Neonatal Intensive Care Unit (NICU) is a 60,000 square foot Level IIC Unit composed of sixty-one beds. The unit is divided into seven clusters, or pods, each containing six or seven 13 x 14 foot infant rooms arranged so that the nursing staff can see into all rooms from the corridor. On average, the NICU admits about 1200 infants per year.

Until recently, nurses in the unit have used what we will refer to as a "double charting system" due to a lack of immediate access to computers. For instance, a nurse would take a patient's vital signs, write the information on any scrap piece of paper they have accessible, and then transfer the values into the computer system in the main corridor whenever they had time. This could be anywhere from ten minutes to one hour after the initial "documentation." If the numbers had been smudged, or if they lose the paper, this could potentially lead to insufficient documentation. To reduce this error, clinical engineers at Vanderbilt Children's Hospital’s NICU have made several attempts to bring the computer stations into the patient room in an ergonomically friendly way where the systems are accessible to the nursing staff while being non-invasive to patients and their families. This way, vital information can be logged immediately, keeping data more current than before and reducing human error.

Research conducted on bedside charting at Minneapolis VA Medical Center helped to support this technological move. In a similar setting to that at the Vanderbilt Children's Hospital, conversion to a bedside computing system yielded many advantages. Prior to installation, nurses spent 7% of their time gathering data and 17% charting data. Use of the computer system reduced time spent gathering information to 4% and time spent charting to 10%.

When the rooms of Vanderbilt’s NICU were initially converted from double rooms to single rooms, each room was equipped with a mobile cart known to the nurses as a COW, or computer on wheels (Figure 1). These systems, however, had many shortcomings. Since they were mobile laptop computers, it was imperative that they remained charged. Therefore, it was the responsibility of the nurses to plug the computers into an outlet after use. This very rarely happened as the life of a nurse in an intensive care unit is very busy. Also, it was often times difficult to locate the COWs and the wheels of the cart were known to fail.

In addition to logging vital signs, the computing systems are also used for the administration of drugs through a program called AdminRX. This program requires the nurses to scan the identification wristbands of the patients and compare them to the barcodes on the medication to make sure that the patients are receiving the proper medication. If the COWs were dead or misplaced, this could be a very daunting task. The failure of this function of the computing systems helped the engineers to move to a system where computing stations could be permanently placed into each one of the rooms in the unit starting with Pod C.

At our entry to this project, a prototype computer charting station had been placed at the bedside of one of the patient rooms in Pod C. This station helped to alleviate the double charting and drug administration problems, but was not ergonomically friendly to the nurses. We refer to the dual screen system as Prototype I (Figure 2). It had two standard size flat screens, one with the vital sign reading and one with the data input screen. They were attached together side by side and then connected to a common keyboard and mouse. The entire unit was mounted to the wall by an arm that provided horizontal and vertical movement. The system was bulky and considered instable by the vast majority of the nurses. Initially it was our job to define that instability and then come up with a way to make the system more stable.

After all the negativity surrounding the dual screen system, the clinical engineers decided to abandon this system and use a more lightweight machine (Figure 3). The vital screen was mounted to the wall and a single screen was left on the station for inputting data, mounted to the wall by an Ergotron arm (Figure 4). A KVM switch was used to give the system the capability of toggling between the vital screen and the informatics screen. This setup is identified as Prototype II.

After the system was revamped, it became our new goal to determine a way to stabilize the new system, and compare the ergonomics of the two systems to quantitatively and qualitatively assess the improvements. We also researched better ways of bringing charting to the bedside that exist on the market today, potentially at costs cheaper than those used to create the prototypes.

Methods

First, we researched ways of making the initial dual screen prototype more stable. We decided that an adjustable leg attached to the bottom of the Prototype I nursing monitor system would probably reduce the keyboard shakiness. The clinical engineers then cut the bottom off of a walking cane and attached it to the keyboard desktop. However, instead of being centered on the bottom of the keyboard, the leg was attached towards the back of the desktop, which did very little to positively affect stability. Also, there was a metal screw that had to be twisted on the leg to adjust the height. This required a level of strength that had some of the nurses having to seek assistance to make adjustments.

When the Prototype II was created, we suggested finding a leg that used a push button pump-and-spring system similar to that used to vertically adjust a computer desk chair. The clinical engineer found a leg in a paintbrush extension arm manufactured by Shur-Line for approximately $20 (Figure 5). The brush part of the arm was cut off and a metal plate was welded to the metal part of the arm and then bolted to the base of the station. This arm was attached to center of the desktop for better support.

After the new prototype was completed, we developed an assessment to quantitatively measure improvement. The factors that were important to the nursing staff were a greater range of motion, a more stable workbench, and a decrease in the amount of force required to make lateral and vertical adjustments. Prototype I was tested without the arm, since that is the condition in which we found it. Prototype II was tested to include all of the improvements that had been made.

To determine the differences in these systems, we first measured the width of each. Then we lifted each system as high as it could go on their respective arms and measured the distance from the base of the computer to the floor. Afterwards, we lowered each system as far down it could go and repeated the measurement. From this we found the range of vertical motion.

It was not only important that the range of motion increased, but it was also important that the force required to adjust the height decreased. To gauge the difference in the amount of force required, we measured the amount of time it took for one group member to move the system from the highest point to the lowest point. Then, we repeated the measurement with the group member moving the system from the lowest point to the highest point. This process was repeated ten times for each movement. We then averaged the values to find the average time required for vertical movement, making the assumption that if the motion required less force, then the movement would take less time.

To quantify "keyboard shakiness," we measured the duration of oscillation and the number of oscillations per second. This was done by placing a two pound weight on the keyboard, quickly releasing it, and counting the number of oscillations visible to the human eye over a period of time. For better accuracy in these measurements, the same group member was commissioned to measure the oscillations. We also measured the keyboard deflection, or the amount of downward movement (in inches) when the keyboard is exposed to a force. To do this, we measured the difference between the front of the keyboard and the floor with and without the two pound weight.

To qualify our findings, we created surveys for the nurses to take. One was for the Prototype I, and the other was for the Prototype II. We asked the nurses to rate the ergonomics of each system independently so that we could gauge which best fit their needs. These surveys were only administered to nurses who had worked with both prototypes. In addition to the surveys, we interviewed members of the nursing staff to gather feedback regarding the two prototypes. To assess our effectiveness in pleasing our customer, the nurses, we created a Quality Function Deployment diagram (Figure 6).

Results and Discussion

In this project we faced many difficulties in trying to develop ways to quantify “shakiness” and stability. We did research on ergonomic tools that are on the market now and the techniques, practices, and principles that govern stability. This project was particularly difficult because we were asked to perform an engineer’s assessment on something that seemed to be common sense: if an object moves when disturbed, it is instable; if an object does not move it is stable.

Choosing a way to measure stability was also difficult in terms of finding a way to best simulate the force that the average human would exert on the keyboard when typing. After researching and brainstorming, we decided that disturbing the system with a two pound weight would be the best way to simulate the force because it was most similar to the force exerted when the nurses rested their hands on the keyboard; and would therefore be similar to the maximum amount of weight that a person would exert when typing. We also decided to remove the weight instantaneously to simulate the release that would take place when the nurse presses a key on the keyboard. After performing the tests, we noticed drastic differences between the two systems and decided that our assessment was suited for measuring and comparing stability.

In performing our ergonomic assessment, we found that the Prototype II was quantifiably a drastic improvement over the initial prototype. Data from the assessment can be found in Table 1. We found that the time it takes to raise and lower the arm decreased by 69%, which is a much needed improvement since the nurses need to devote the majority of their time to taking care of patients, not adjusting the data entry system.

Based on the deflection tests performed, we found that the stability of the keyboard improved by 75% and the oscillation duration decreased by 90%. The overall “shakiness” of the keyboard that existed in the initial prototype is basically non-existent in the second prototype.

The nurses shared our sentiments when qualitatively assessing the systems. They all agreed, for the most part, that the Prototype II was more user-friendly. The nurses found the pump on the support leg easy to use for the adjustments of altering the height. The new Ergotron arm is also a favorite compared to the previous arm used in Prototype I. The Ergotron arm used in the Prototype II was less in weight than the arm used Prototype II. Measurements were not taken to support this, but in one of the rooms in Pod C, the new lighter system was attached to the wall using the old arm. To our surprise, some of the instability returned. From this we learned that all of the improvements were necessary to make the machine more stable. Initially, we assumed that changing the arm would only decrease the amount of force necessary to move the system around. The new arm, the stability leg, and reducing the weight all helped to make the system more stable. A breakdown of the costs of all the new parts is provided in Table 2. The total for this project was about $2225 per unit minus the computer work station.

Although the engineers at Vanderbilt had to use the “build it from scratch” method to produce these prototypes with us as consultants, the future of bedside data entry is definitely promising. The company Ergotron from which the arm holding the unit was purchased has recently created the combo arm with medium CPU holder which would be great for the NICU. We believe that one of the downfalls of the system is that the parts were not designed by the same manufacturer which made it difficult for them to work together smoothly. This arm/holder combo, however, was designed to work together and hold the weight of the CPU, display, and keyboard. It makes the same vertical and horizontal adjustments and can adjust vertically 15 inches, which is comparable to the 15.1875 inches we measured in the second prototype. In addition to that, the arm also has the capability of tilting the display forward and backwards for ease of viewing; something a few of the nurses mentioned would be helpful that we were not able to offer due to the limitations of our stand. The price for this unit is $882.00, which is comparable to what was spent on mounting in the prototypes. Hopefully this system would provide even more stability without the use of a stabilizing leg, and other accommodations that the nurses would appreciate.

Another issue we researched was the effect the technology would have on the time nurses spent with their patients. This is particularly important in Vanderbilt Children’s NICU because now that the vital screens are mounted to the walls, they can be view from the interior room of the Pod, and be entered in the computer in the interior room without even entering the patient room. This was a fear of the researchers at the Minneapolis VA Medical Center as well. After completing their study, however, they found that nurses spent the same amount of time (43%) in patient rooms before and after the data entry systems were installed. This was very promising news. Even though the amount of time spent charting decreased, the nurses redirected that time to caring for their patients, increasing overall time spent providing care.

Conclusion

Changing the setup of the nursing monitoring stations from the first prototype to the second prototype provided a more stable system. It also allowed for easier vertical and horizontal motion. Based on the nurses’ feedback, the additions incorporated in the second prototype provided a better ergonomic setup, and eases their transition in going to a paperless system.

The medical personnel of the NICU department seem pleased with the progression of this project. Currently, the clinical engineering staff is working to install bedside nursing stations in all of the rooms in Pod C.

As biomedical engineers, it was interesting to go into a hospital setting and learn about the future of data entry. However, this project was almost purely mechanical and so disconnected from what we have learned in engineering school that we were unsure of the purpose we were supposed to serve for the first couple weeks of the project.

This project proved beneficial because it taught us the true purpose of our degrees: to solve problems. Even though we had very little knowledge or experience in stability or ergonomics, we definitely knew how to go about finding solutions to this problem after understanding scope of the problem.

We enjoyed working with nursing staff and the clinical engineering staff at Vanderbilt Children’s Hospital NICU, but we would not recommend having biomedical engineers complete a project similar to this one in the future.

Acknowledgements

We would like to thank the nursing staff of the Neonatal Intensive Care Unit of Vanderbilt Children’s Hospital, Allen Dyer of the Clinical Engineering Department, and Alex Mackowski of Vanderbilt University Department of Biomedical Engineering for all of their assistance.

Appendix

Figure 1: COW (Computer on Wheels)

[pic]

Figure 2: Prototype I

[pic]

Figure 3: Prototype II, Computer Station

[pic]

Figure 4: Prototype II, Wall-mounted Vital Signs Screen

[pic]

Figure 5: Prototype II, Support Leg

[pic]

Figure 6: Quality Function Deployment Diagram based on surveys taken by nurses which show that Prototype II is favored over Prototype I.

[pic]

Table 1: Results of Ergonomic Assessment of Prototypes I & II

|  |Prototype I |Prototype II |

|Screen Size (inches) |41.125 |18.25 |

|Range of Height (inches) |16.125 |15.1875 |

|Time to raise monitor (sec) |7.65 (N = 10) |2.37 (N = 10) |

|Time to lower monitor (sec) |5.55 (N = 10) |1.69 (N = 10) |

|duration of oscillation (sec) |14.24 |1.5 |

|oscillations/sec |3 |2 |

|keyboard deflection with 2kg weight (inches) |0.75 |0.1875 |

Table 2: NICU Equipment Proposal: Assessment of Costs

|Item Number |Description |Cost |

|1 |28" Flat Screen Monitor |$650 |

|2 |Mounting bracket for 28" Screen |$70 |

|3 |Ergotron arm and combo |$800 |

|4 |Extensions for keyboard |$55 |

|5 |Keyboard Stabilizer |$50 |

|6 |Headwall adapters |$100 |

|7 |KVM and misc. cables |$100 |

|8 |Wall Channel |$40 |

|9 |Scanner Bracket |$30 |

|10 |Mount for Philips CPU |$40 |

|11 |Plant Operations Work |$250 |

|12 |CWS CPU and mount |Can be reused |

TOTAL COST: $2225 per room

References

Gregory, K. D. (2003, January). Nursing Documentation Time During Implementation of an Electronic Medical Record. Journal of Nursing Administration, 33(1), 24-30.

Langowski, C. (1995, April/June). The Times They Are Changing: Effects of Online Nursing Documentation Systems. Quality Management in Health Care, 14(2), 121-125.

Moody, L. E., Slocumb, E., Berg, B., Jackson, D. (2004, November/December). Electronic Health Records Documentation in Nursing: Nurses’ Perceptions, Attitudes, and Preferences. CIN: Computers, Informatics, Nursing, 22(6), 337-344.

Pierpont, G. L., Thilgen, D.(1995, June). Effect of computerized charting on nursing activity in intensive care. Critical Care Medicine, 23(6), 1067-1073.

Splitzer, A. R. (2007, September). The electronic medical record and the data warehouse in neonatal practice - improving patient care through modern technology. Neonatology Today, 2(9), 1-8.

Walsh, W. F., McCullough, K. L., & White, R. D. (2006, October). Room for the Improvement: Nurses’ Perceptions of Providing Care in a Single Room Newborn Intensive Care Setting. Advances in Neonatal Care, 6(5), 261-270.

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download