Fall 2023 Projects
Project Summary
The Infection Prevention (IP) team at UPMC monitors patients with Epidemiologically Significant Organisms (ESOs). These rare but potentially life threatening organisms pose significant risks to patients and the hospital system. The IP team has developed a protocol for the care of ESO patients, involving nursing staff, observers, and housekeeping. The current IP protocol is insufficient and breakdowns occur that cause unnecessary patient exposure. Each lapse represents unacceptable levels of transmission risk that impacts patient safety. This is costly to UPMC in time, staffing, equipment, and testing, even if a transmission event does not occur. We would like to understand and develop interventions that better document these lapses with the goal to reduce transmission and patient harm.
The team conducted a review of healthcare literature and UPMC’s existing protocol documentation, creating a task breakdown structure using the latter. The team also conducted interviews with various stakeholders, including observers, nurses, patient care technicians, and IP team members. A What-if analysis was conducted to assess the risk associated with each responsibility in the protocol, revealing areas with the highest risk of lapses. These methods highlighted a breakdown in communication among hospital staff roles, particularly during patient handoff and transport. The solution to this problem was the creation of a dashboard with two views, one for the observer to virtually inspect the protocol and one for the IP team to track patient progress and observer documentation within the ESO patient care protocol. The dashboard’s main benefit is the creation of a data collection tool to better understand where lapses occur. Our team recommends using the dashboard to identify where lapses occur and to develop new questions to improve adherence where needed. The use of this dashboard will help to foster a sense of shared decision making to promote a culture of ESO awareness.
Project Summary
UPMC Shadyside is a 520-bed tertiary care hospital in Pittsburgh, Pennsylvania. The hospital’s centralized telemetry unit (CTU) monitors and distributes telepacks. Telepacks are attached to patients to detect abnormal heart activity. Despite having more than enough of these devices to meet demand, the CTU routinely runs low on inventory, leading to subsequent delays in telepack delivery. This project focused on identifying why this issue was occurring and recommending corrective actions.
The team conducted a root cause analysis and corresponding data analysis to quantify the current condition of CTU inventory. The CTU’s physical inventory system was determined to be difficult for staff to maintain due to a reliance on paper slips. Additionally, the CTU’s virtual inventory system was developed with inaccurate assumptions and as a result, did not reflect real inventory levels. Finally, a lack of standardization regarding inventory replenishment was identified. These deficiencies corresponded to an annual cost exceeding $5000 in unproductive labor.
Recommendations for the CTU focused on inventory system modifications and staff process changes. For the physical system, the use of paper slips for inventory was eliminated and a change to how telepacks are stored was suggested. For the virtual system, additional telepack statuses were developed to promote accurate device tracking and automated messaging was recommended to reduce device return times. Finally, routine telepack collection times and an inventory reorder point were established. All of these recommendations were included in an implementation plan. Upon implementation, these recommendations are expected to decrease instances of low inventory, expedite telepack deliveries and returns, and allow CTU staff to dedicate more attention to their primary task of monitoring patient heart rhythms.
Project Summary
IDL Worldwide is a commercial print shop that serves brands such as Chick-fil-A, Sheetz, and Giant Eagle. They print, package, and ship signs off to stores. The team is focusing on the packaging department which is responsible for the assembly, packing, and kitting of signs. Projects are cyclical and IDL must outsource labor when they lack the capacity for fulfilling projects. Due to this nature of projects, there is an increased cost of labor and a dependency on temporary workers. The team is going to determine the size of an automated collation machine that has reasonable labor cost savings whilst satisfying customer demand. Our goals are to create a plan for implementing this machine and provide a recommendation for its configuration. The team analyzed historical data by using data visualization to identify trends of store frequencies, grouping similar frequencies to determine required bays, and a mathematical model comparing manual and machine throughput to calculate labor cost savings. One brand, EG, was a special case in which a clustering algorithm was used to group signs by dimensions to create multiple batches. The team calculated 25 bays to be optimal and received two machine quotes of $992,000 and $882,000, with the former having a bay size 29” by 24” and the latter 19.5” by 27.5”. The cost difference accounts for an increase in bay size and only accounts for a 2% increase in the volume of signs going onto the machine. The estimated labor cost savings using the machine is $97,692 per year, which is larger than IDL’s current spending on outsourcing labor. This calculation was found to be sensitive to picking time. We recommend that IDL completes labor standard validation to justify labor cost savings due to its sensitive nature, and they should meet with MKW to discuss further purchasing plans.
Project Summary
NurturePA is a nonprofit organization dedicated to promoting the social and emotional health of new mothers and their children. The organization runs a free mentorship program that pairs mothers with mentors according to the level of support they require, referred to as their “needs risk level.” Currently, NurturePA depends on Allegheny County to provide needs risk levels for each mother. However, the county is not transparent in its classification process and may stop providing classifications entirely; plus, NurturePA wishes to expand to other counties. A solution was required to identify low versus high needs risk mothers in the same way as Allegheny County but using only the data available to NurturePA.
Multiple classification algorithms were tested on data with three parameters: household income range, race, and single parent status. Algorithms’ performance was evaluated across accuracy, recall (out of only high needs risk mothers, how many are correctly classified), and a custom “NPA Metric” which combines the previous two while weighing recall more heavily. A logistic regression model performed the best and was incorporated into code scripts for prediction. When a new mother is enrolled in the program, their information can be sent to the model, and the predicted needs risk level is output to the NurturePA website. An updating script was also created–if more data is collected in the future, it can change the model based on that new data, aiming to improve performance. Additionally, a dashboard was constructed to provide visual insights into the model. It includes an interactive feature where users can select values for each parameter and the associated needs risk level prediction and probability of accuracy are displayed. As a result of this project, NurturePA can reach approximately 400 more mothers outside Allegheny County and ensure they receive the appropriate level of care and support.
Project Summary
C3 Controls current order picking process is negatively impacting the time it takes to pick items from inventory. The current process is very manual, requires employees to compare long part numbers between products and order sheets, and is not able to automatically track the inventory of these parts. Time studies, analysis on provided data, and process observations were conducted to identify the current state order picking times, inspection time, pick error rate, and process inefficiencies. Studies showed room for improvement through reduction of time to pick parts and elimination of the inspection step with implementation of a barcoding software. Literary research provided evidence that barcoding technology has the ability to reduce human errors. This barcoding software will essentially perform the inspection step for employees automatically, hence removing the need for that additional step. A MOST analysis was conducted to further prove the barcoding technologies ability to reduce order picking time through shortening the time to pick a part from the shelf. Utilizing these assumptions, a new order picking time was estimated. From here, a cost analysis was conducted using the new process time savings and employee hourly wage. These values and potential vendor costs were used in predicting a payback period for two different suppliers. C3 Controls has the potential to shorten their order picking process time and automate their inventory tracking by utilizing a barcoding software like that provided by Fishbowl Inventory.
Project Summary
This project aimed to improve the timesheet process at GE Vernova, located in Charleroi, PA, by exploring alternatives to the current paper-based system. Initially, the team conducted a comprehensive analysis of the existing process in all three departments.
The investigation revealed significant issues with error rates and missing timesheets, resulting in approximately 560 hours of lost time each week. To address these concerns, three potential solutions were developed: a streamlined paper timesheet, an iPad application, and the implementation of a manufacturing execution system (MES).
These alternatives were compared in an effort to evaluate their ability to mitigate errors and enhance visibility for both employees and management. The streamlined paper timesheet was designed to standardize the process in the three departments and enforce an error check. The iPad application and MES were identified as highly effective in reducing errors and ensuring comprehensive data entry without manual input. In addition, detailed implementation plans were created for each alternative.
A final recommendation was formulated based on a cost-benefit analysis, highlighting the integration of the iPad application with existing systems as the most viable solution for GE Vernova.
In conclusion, this project presents feasible and pragmatic alternatives to the current paper-based timesheet system, emphasizing significant reductions in errors and lost hours available through the adoption of modernized solutions.