Sarah Kadish

Conformance checking and performance improvement in scheduled processes

Service processes, for example in transportation, telecommunications or the health sector, are the backbone of todays economies. Conceptual models of service processes enable operational analysis that supports, e.g., resource provisioning or delay prediction. In the presence of event logs containing recorded traces of process execution, such operational models can be mined automatically.In this work, we target the analysis of resource-driven, scheduled processes based on event logs. We focus on processes for which there exists a pre-defined assignment of activity instances to resources that execute activities. Specifically, we approach the questions of conformance checking (how to assess the conformance of the schedule and the actual process execution) and performance improvement (how to improve the operational process performance). The first question is addressed based on a queueing network for both the schedule and the actual process execution. Based on these models, we detect operational deviations and then apply statistical inference and similarity measures to validate the scheduling assumptions, thereby identifying root-causes for these deviations. These results are the starting point for our technique to improve the operational performance. It suggests adaptations of the scheduling policy of the service process to decrease the tardiness (non-punctuality) and lower the flow time. We demonstrate the value of our approach based on a real-world dataset comprising clinical pathways of an outpatient clinic that have been recorded by a real-time location system (RTLS). Our results indicate that the presented technique enables localization of operational bottlenecks along with their root-causes, while our improvement technique yields a decrease in median tardiness and flow time by more than 20%. HighlightsWe provide a broad extension to the notions of conceptual, operational and continuous conformance checking techniques.We present novel theoretical results in terms of scheduling algorithms for Fork/Join networks that re-sequence arriving cases. We prove that these scheduling algorithms are guaranteed to never perform worse than the baseline approach.We test the new techniques for process improvement on real-world data, and show that the proposed algorithms yield a 20%-40% improvement in median flow time and tardiness.

Data-Driven Performance Analysis of Scheduled Processes

The performance of scheduled business processes is of central importance for services and manufacturing systems. However, current techniques for performance analysis do not take both queueing semantics and the process perspective into account. In this work, we address this gap by developing a novel method for utilizing rich process logs to analyze performance of scheduled processes. The proposed method combines simulation, queueing analytics, and statistical methods. At the heart of our approach is the discovery of an individual-case model from data, based on an extension of the Colored Petri Nets formalism. The resulting model can be simulated to answer performance queries, yet it is computational inefficient. To reduce the computational cost, the discovered model is projected into Queueing Networks, a formalism that enables efficient performance analytics. The projection is facilitated by a sequence of folding operations that alter the structure and dynamics of the Petri Net model. We evaluate the approach with a real-world dataset from Dana-Farber Cancer Institute, a large outpatient cancer hospital in the United States.

Implementation to Optimization: A Tailored, Data-Driven Approach to Improve Provider Efficiency and Confidence in Use of the Electronic Medical Record

PURPOSE Nine months after the implementation of a new electronic medical record (EMR) system at a single institution, physicians (MDs), nurse practitioners (NPs), and physician assistants (PAs) expressed frustration with its use. We aimed to test if an individually tailored training approach reduced time spent with the EMR and increased confidence. MATERIALS AND METHODS Two hours of training were conducted in a one-on-one manner with a trainer. Content was individualized according to the following: provider survey, EMR utilization profile, and shadowing in clinic. Surveys assessed confidence before training and immediately after training. Changes in time spent in various EMR activities before training and after training were compared. RESULTS Three trainers delivered one-on-one training to 133 MDs, 42 NPs, and 10 PAs who specialized in medical oncology. Participants reported an increase in confidence across all activities, and almost all providers (98%) who responded to our survey agreed that the training enhanced their efficiency. A non-statistically significant trend toward reduction in the overall time in the system was observed. Time in system was reduced primarily in activities such as documentation and ordering of laboratory tests, imaging, medications, and chemotherapy. CONCLUSION A personalized and data-driven training approach was highly regarded by providers. EMR usage reports provided extensive data to identify and prioritize training content and were valuable to measure the impact of training on provider time in system. With the growth of EMR implementation and the reported relationship of EMR use to burnout, continuous and personalized training after EMR implementation is effective to reduce the time in system and increase confidence.

Measuring Chemotherapy Appointment Duration and Variation Using Real-Time Location Systems.

Introduction:Clinical schedules drive resource utilization, cost, and patient wait time. Accurate appointment duration allocation ensures appropriate staffing ratios to daily caseloads and maximizes scarce resources. Dana-Farber Cancer Institute (DFCI) infusion appointment duration is adjusted by regimen using a consensus method of experts including pharmacists, nurses, and administrators. Using real-time location system (RTLS), we examined the accuracy of observed appointment duration compared with the scheduled duration. Methods:Appointment duration was calculated using RTLS at DFCI between August 1, 2013, and September 30, 2013. Duration was defined as the total time a patient occupied an infusion chair. The top 10 administered infusion regimens were investigated (n = 805). Results:Median observed appointment durations were statistically different than the scheduled durations. Appointment durations were shorter than scheduled 98% (C), 95% (I), and 75% (F) of the time and longer than scheduled 77% (A) and 76% (G) of the time. Fifty-six percent of the longer than scheduled (A) appointments were at least 30 minute longer. Conclusion:RTLS provides reliable and unbiased data to improve schedule accuracy. Replacing consensus with system-based data may improve clinic flow, relieve staff stress, and increase patient satisfaction. Further investigation is warranted to elucidate factors that impact variation in appointment duration.

Measuring chemotherapy appointment duration and variation utilizing real-time location systems (RTLS).

154 Background: Clinical schedules drive resource utilization, cost, and patient wait time. Accurate appointment durations ensure appropriate staffing ratios to the daily caseload and maximizes scarce resources. Dana-Farber Cancer Institute (DFCI) adjusts infusion appointment durations for each chemotherapy regimen using a consensus method of experts including pharmacists, nurses, and administrators. Utilizing RTLS, we examined the accuracy of appointment duration compared to suggested duration. METHODS Appointment duration was calculated using RTLS in three disease centers at DFCI between August 1st and September 30th, 2013. Duration was defined as the amount of time a patient occupied an infusion chair. The top 10 administered infusion regimens were statistically investigated (n=805). RESULTS All median observed appointment durations were statistically different than the suggested durations. Appointment duration was shorter than scheduled 98% [C], 95% [I], and 75% [F] of the time and longer than scheduled 77% [A] and 76% [G] of the time. Almost all C and I appointments were more than 30 minutes shorter than scheduled. Among A appointments longer than scheduled, 56% were more than 30 minutes longer than scheduled. CONCLUSIONS RTLS provides reliable and unbiased data to improve schedule accuracy. Replacing consensus with system-based data may improve clinic flow, relieve staff stress, and increase patient satisfaction. Further investigations should elucidate factors that cause variation in appointment duration. [Table: see text].

Standardizing coordination between surgical oncology and reconstructive surgery for breast cancer patients undergoing mastectomy with immediate reconstruction.

110 Background: Timely diagnosis and treatment of breast cancer, endorsed by organizations such as ASCO and NCCN, are essential to ensure optimal clinical outcomes and patient satisfaction. Inefficient care coordination may adversely affect care quality. At our cancer center, 75% of patients who undergo mastectomy seek a reconstructive surgery consult and over 60% elect mastectomy with immediate reconstruction. We sought to evaluate and reduce the time to reconstructive surgery consult and first definitive surgery (FDS) by streamlining coordination between services. METHODS We studied 330 patients who underwent mastectomy with immediate reconstruction between January 2011 and April 2013. Time intervals between initial surgical consult, reconstruction consult, and FDS were calculated. After examining existing best practices in patient referral and scheduling, we established targets of 7 days from initial consult to reconstruction consult and 28 days from initial consult to FDS. To achieve these targets, facilitated sessions were held with administrative and clinical experts to create a standard referral and scheduling process, including a referral template and establishing surgical teams based on clinic and operating room alignment. The interventions were implemented over a 6-month period. RESULTS Mean days from initial consult to reconstructive surgery consult decreased, with significant improvement in reaching the 7 day target. No significant changes from time of initial consult to FDS were observed. CONCLUSIONS Standardizing coordination has led to timelier reconstructive surgery consults for patients undergoing mastectomy with immediate reconstruction. Other factors, such as operating room availability, pre-operative testing and patient preference should be explored to reduce the time to FDS. [Table: see text].