Search

2085 Background: The Memorial Sloan Kettering (MSK) Urgent Care Center (UCC) functions as the emergency room for MSK. With 23,000+ visits annually, increasing volume and acuity means more days over capacity. Patients experience increased wait times to see clinicians, complete evaluation, and transfer to an inpatient bed. The UCC TeleTriage Program is a remote triage program which aims to align patient volume and need with available resources, improve patient experience, and streamline flow through the UCC. By managing resources more efficiently and expediting initial evaluation, the program promotes timely patient access to care, while maintaining MSK's standard of care. Methods: UCC TeleTriage began July 2018 with the Gastrointestinal Medical Oncology service. The Service Nurse refers patients to TeleTriage on weekdays, from 9a.m.- 4:30p.m. The TeleTriage clinician contacts each patient within 30 minutes of referral, takes the history, and determines the initial plan. Patients are directed to a local ER, clinic, or UCC based on level of acuity, real-time GPS, and specific need. For stable patients coming to UCC, TeleTriage focuses on initiating testing prior to registration in UCC. Results: TeleTriage patients have (virtual) contact with a UCC clinician within 30 minutes of referral, whereas non-TeleTriage patients wait 110 minutes or more. TeleTriage patients are discharged from UCC up to 42 minutes more rapidly. TeleTriage patients who receive imaging prior to registration in UCC receive a final disposition up to 93 minutes sooner. About 4% of TeleTriage patients are managed at home. In a small number of TeleTriage patients with severe complications of cancer-treatment, significant morbidity was avoided due to early intervention and coordination of care. Conclusions: TeleTriage patients have contact with a UCC clinician measurably faster than non-TeleTriage patients. Their evaluation is also started earlier. By managing less acute patients at remote sites or at home, TeleTriage can help patients avoid unnecessary travel, (time) expenditure, and hospital contact. TeleTriage patients who come to UCC, spend less time in UCC than non-TeleTriage patients and they discharge faster. By utilizing cancer care expertise, TeleTriage can significantly impact patient outcomes and utilize resources more effectively.

2022 Background: Studies have suggested superior outcomes for patients with cancer treated at National Cancer Institute (NCI) and academic hospitals, leading some to advocate for complex cancer care to be delivered at regional referral centers. However, growing demand at such centers may exceed their capacity to provide timely treatment, which could be detrimental to patient outcomes. We evaluated the relationship between hospital type, the average annual growth rate in patient volume (PV), and time to treatment initiation (TTI) trends. Methods: We used the National Cancer Database to identify patients undergoing initial treatment for a new diagnosis of cancer (breast, lung, prostate, colorectal, melanoma, bladder, non-Hodgkin lymphoma, renal, uterine or pancreatic) in 2007-2016. The exposure was hospital type (NCI, academic, community or integrated network). The primary outcome was TTI over time. We estimated both the average annual growth rate for PV and adjusted TTI trends by hospital type using linear mixed effects models, including a hospital type-by-time interaction and, when modeling TTI, a patient volume-by-time interaction. Results: We identified 4,218,577 patients treated at 1351 hospitals (49% at 897 community, 23% at 177 academic, 14% at 50 NCI and 14% at 227 integrated network hospitals). Over the study period, PV grew by 40% at NCI and 25% at academic hospitals, compared to 8% at community hospitals (p-value for trend both < 0.001). Meanwhile, mean TTI increased by 3.2 days at community, remained stable at academic (+0.3 days) and decreased by 4.3 days at NCI hospitals (p-value for trend both < 0.001 vs community). A higher annual PV growth rate was associated with a statistically but not clinically significant TTI increase (0.05 days for each 100 patient/year increase in the growth rate, p = 0.001). Conclusions: Patients with newly diagnosed cancer are increasingly receiving treatment at NCI and academic hospitals. While TTI at NCI and academic hospitals is longer than in the community, PV growth has been possible without delaying cancer treatment. Further study is needed to determine whether continued growth at this rate is sustainable. [Table: see text]

e16039 Background: There is an association between higher hospital and urologist case volume and improved survival outcomes for patients with bladder cancer. Limited data on facility volume and outcomes exists following trimodality bladder preservation with TURBT followed definitive chemoradiation (CRT). Methods: The National Cancer Database was queried for patients from 2004-2013 with urothelial bladder cancer (cT2-4aN0M0) receiving definitive CRT after TURBT. High-volume bladder preservation case volume (BPCV) was defined as facilities in the > 70th percentile of trimodality bladder preservation cases per ten years examined. We compared OS between high- and low-BPCV using Cox proportional hazard models. CRT was defined as > = 5040cGy pelvic RT with chemotherapy given within 21 days. Propensity matching was performed to match high- and low-volume centers. Results: A total of 666 treatment facilities treating 1,635 patients with bladder preservation were identified with a median follow-up of 26 months (range 2-136 months). A 70th percentile cutoff identified 497 patients that received treatment at 64 high-volume facilities. Median OS of patients treated at high BPCV centers was 37.0 months (95% CI, 27.9-33.6) versus 32.3 months (95% CI, 27.3-46.7) for patients treated at low BPCV centers (P = 0.004). High BPCV facilities were independently associated with a decreased hazard of death (HR, 0.86, 95% CI, 0.75-0.98; P = 0.031). When BPCV was analyzed as a continuous variable, an increasing number of bladder preservation cases per institute was also independently associated with improved OS (HR: 0.98, 95% CI, 0.97-0.99; P = 0.044). In the propensity score matched cohort, median OS of patients treated at high BPCV was 36.1 months (95% CI, 26.5-45.8) versus 28.1 months (95% CI, 23.9-32.3) for patients treated at a low BPCV facility (P = 0.016). Conclusions: In this observational cohort, treatment at a high BPCV facility was associated with improved OS. Causal factors for this finding may include greater institutional expertise in multidisciplinary bladder cancer care and potential advantages in treatment technology.

39 Background: In Ontario’s publicly funded cancer system, patient access to Gynecological Oncology (GyneOnc) care is threatened by growing demand, waiting times and need for lengthy patient travel. Cancer Care Ontario has published organizational standards that lay out the minimum clinical elements and infrastructure needed for a high quality program. Capacity planning and readiness assessments were used to guide GyneOnc expansion into non-academic regional cancer centres to alleviate system pressure. We evaluated early expansion activities to learn about critical factors and generate recommendations for future growth of this specialized oncology service. Methods: Capacity planning for new centres factored health human resources (HHR), patient volume and travel distances, ensuring that existing programs maintain critical mass. Evaluation included review of documents and readiness assessments related to the expansion projects undertaken to date. Interviews were done with GyneOncologists and administrators working at local and provincial levels to discover areas of risk, key factors for success, and recommended strategies for future expansion. Results: Gyne Onc programs need adequate resourcing (minimum 3 GyneOncs) to meet patient demand, support on-call and offer resilience to parental/medical leaves. Programs need engagement from other surgical and medical subspecialties, and clinical services (e.g. pathology, diagnostic imaging) for integrated, multidisciplinary care. Academic appointments, residencies, fellowships, and support for research provide links with universities, important for credibility. Endorsement by senior leadership - both management and clinical - is crucial. Organizational standards need specificity around the infrastructure and multidisciplinary environment needed for program success. Conclusions: HHR planning and readiness assessment are critical enablers in the expansion of specialized services like GyneOnc beyond academic centres. New programs need to consider the adequacy and appropriateness of staffing models, broad engagement of the clinical team, and linkages to academia, in order to ensure program sustainability and acceptability.

259 Background: Increasing patient volume and treatment complexity in the field of radiation oncology has resulted in increased number of errors possibly affecting patient safety. Effective methods of mitigating these errors include automation, computerization, simplification, and standardization. To improve quality of care and patient safety, our institution established consensus standardized treatment guidelines for each cancer site. However, physician orders for computed tomography (CT) simulation for radiation treatment planning continued to have extreme variability, with error rates of 31%. Therefore, a team was assembled to devise standardized orders to reduce error, improve patient safety, and improve quality of care in the CT simulation order process. Methods: For this study, we investigated 3 commonly treated sites at our institution: breast (14%), prostate (7%), and brain metastases treated with radiosurgery (14%). A standardized template CT simulation order was defined for each disease site using the consensus treatment guidelines. These orders were integrated into the electronic medical record (EMR) on March 5, 2018. To evaluate the efficacy of the intervention, CT simulation order data were queried for the two-month period before and after implementation of standardized template orders. Orders with variation from the treatment guidelines were counted to calculate error rates with and without standardized simulation orders. Results: In the two-months prior to implementation of the standardized order templates, 48 of 151 (31%) CT simulation orders for the three selected sites had variation from the consensus standardized orders. After implementation of the EMR standardized template, 17 of 129 simulation orders (13%) in the three selected sites had variations from the standard during this two-month period. Standardization of CT simulation orders using an EMR template reduced error rates from 31% to 13% (18% absolute reduction; 42% relative reduction). Conclusions: Simplification and standardization of CT simulation orders decreased error rates by 42%, thereby improving clinic efficiency and appropriate patient treatment.

156 Background: The purpose of this study was to investigate the relationships between new patient volume per oncologist, the availability of a social worker, and the capacity to provide psychosocial care. Methods: Community practices within Michigan Oncology Quality Consortium (MOQC), a statewide quality collaborative, completed NCI’s 10-item Cancer Psychosocial Care Matrix (CPCM). The CPCM speaks to communication, assessment, coordination, training, and outcomes and quality, each on a separate Likert scale. Analyses describe the ability to meet psychosocial care and the availability of a social worker, new patients/full time oncologist, payer mix, and percent minorities. Results: To date,15 of 47 practices (32%) completed the CPCM. We report on 4 measures with sizeable variation between practices. Scores were the highest on the ability to communicate the importance of psychosocial care and implementing a plan (scale of 1 – 3; mean of 2.36, standard deviation, SD 0.72; mean of 1.79, SD 0.89). The lowest scores were related to linking patients with services and offering provider education (scale of 1 – 5: mean of 2.18, SD 1.08; mean of 2.20, SD 0.94). Linear regression showed an association with number of patients/oncologist and the ability to provide staff training of effective patient/provider communication and monitoring this activity’s impact. For every 100-person increase in the number of patients/oncologist, there was a -0.16 (p = 0.046) rating difference in staff training and a -0.25 (p = 0.006) rating difference to monitor the communication’s effect. A dedicated social worker was present in 3 practices. Regression analysis showed a relationship between a social worker and recognizing (p=0.02) and addressing psychosocial care (p

224 Background: Published guidelines for follow up after R0 resection of GEP-NETS are complex and emphasize closer surveillance in the first 3 years, as per other GI malignancies. NETS have a different pattern and timescale of recurrence, hence require more practical and tailored follow up. Aim: To examine real world follow up practices compared to published guidelines for follow up of patients with GEP-NETs. Methods: A detailed electronic cross-sectional survey was developed and distributed to members of the CommNETS Collaboration (Australia, NZ, Canada) and NANETs. Questions addressed demographics, knowledge and use of current guidelines and follow up practices relating to various prognostic factors. Descriptive statistics were obtained for all responses, stratified by country, patient volume and specialty. Results: There were 163 respondents: Australia: 59; NZ: 25; Canada: 46; US: 33. 50% were Medical Oncologists; 23% Surgeons; 13% Nuclear Medicine; 14% other. 38% of respondents were “very familiar” with NCCN NET guidelines; 33% with ENETS, 17% with ESMO, however only 15%, 27% and 10% found them “very useful”, respectively. 63% reported not using guidelines at their institution. Ranking of prognostic factors (top 5, decreasing order): grade, Ki67/mitotic count, T stage, N stage and site of origin. Follow up in 1st 2 years was most commonly every 6 months (62%); in years 3-5 every 12 months (59%) and > 5 years, 12 months (41%). Follow up patterns did not differ significantly by patient volume. The commonest investigations were CT scans (66%) and CgA (86%). When poor prognostic factors were introduced, increased visits and tests were recommended. Conclusions: This large international survey yields detailed information, highlighting variation in current follow up practices not well addressed by current guidelines. This forms a strong basis for the upcoming CommNETS/NANETS consensus meeting, which aims to produce user friendly, practical guidelines tailored to the expected pattern of recurrence in patients with NETs.

23 Background: The objectives of the study were to determine whether a Palliative Care Unit (PCU) provides benefits not just from a cost perspective, but from a patient and caregiver quality of life (QOL) perspective. Methods: (1) Calculate the total costs of the PCU; (2) Leverage a threshold analysis to estimate the Quality-Adjusted Life Years (QALYs) required for the PCU to be cost effective; and (3) Determine whether it is feasible for the program to yield the required number of QALYs. Setting was the Johns Hopkins Health System Palliative Care Unit (PCU) in Baltimore, MD. Analysis was based on patient volume from March 2013-2014. Results: There were 209 palliative patients. The costs for the societal perspective was $2,044,364 and the required number of QALYs to deem it cost effective were 11.36. The net costs for the hospital perspective was $625,777 (gross cost was $993,528; however, the program generated $367,751 in savings for the hospital through treating patients in the PCU as opposed to other functional units), and the required number of QALYs to deem it cost effective were 3.48. To determine whether the program is able to achieve the number of QALYs required, the study team generated aggregated QALYs based on other studies’ evidence for palliative care’s improvement of quality of life for patients and their caregivers. Combining the QALYs generated from the aggregated calculations for patients (0.12) and caregivers (4.60), the program had the potential to yield a total of 4.73 QALYs. Conclusions: This analysis suggests that the PCU is cost effective from the hospital perspective in the sense that the benefits it provides to patients’ and caregivers’ quality of life outweighs the cost of care. Future studies should continue to evaluate palliative care from a cost effectiveness perspective that incorporates a consideration of the quality of life improvements, rather than just cost-reduction.

231 Background: The department of Radiation Oncology at the Dana-Farber/Brigham and Women’s Cancer Center operates seven linear accelerators treating approximately 200 patients per day. These machines are regularly shut-down for upgrades, requiring the redistribution of patients to available machines. Previous machine downtimes have caused disruptions to patient care and required extended operating hours to accommodate displaced patient volume. For the next machine downtime, we sought to create a model to optimize the reassignment of patients to available machines, maximizing treatment type-machine compatibility while minimizing overtime, costs, and patient care impact. Methods: This collaboration between Radiation Oncology and the Process Improvement team retooled the Assignment Problem Method, an Industrial Engineering technique used to assign tasks, in order to solve for case allocation to radiation machines. A spreadsheet model was developed to optimize the allocation of treatment cases for each machine, without violating working hours or treatment type-machine constraints. Input data included case duration, radiation technique, daily demand, and machine compatibility for each treatment type. In addition to finding an optimal case allocation, department administrators created “what-if” scenarios to identify how demand fluctuations would shift cases across machines. Proposed solutions were validated by comparing the model’s output to feasible treatment volumes identified by department administrators. Results: The model revealed that under idealized conditions, the current daily patient volume can be accommodated with two fewer machines and 1.5 fewer operating hours each day. The time it took department administrators to prepare for the next downtime was reduced from several weeks of manual planning, to a few days using the model. Conclusions: Simple industrial engineering tools can help solve complex problems. Developing optimization models can eliminate complex calculations, allowing end-users to refine and implement a suitable solution without fear of mathematical error. A robust model can provide end users the ability to assess the impact of future changes in capacity or demand.

e17703 Background: Antibiotic use in pediatric ALL patients varies significantly. Studies in other disciplines suggest increased patient volume is related to improved guideline adherence and fewer ...

6018 Background: To date, no quality of care indicators (QI) specific for NSCLC are widely accepted. We proposed a set of QI and are reporting on quality of care using data from FIQCC consortium which is comprised of 11 oncology practices. In addition, this study explored the impact of patient volume on quality of care. Methods: Major care guidelines (NCCN, ACCP, ESMO, Ontario CC, ELCWP) were systematically reviewed for potential QI. Survey was conducted among NCCN NSCLC panel in 2008 for QI with strong agreement ratings. Chart abstraction manual was created, structured chart abstraction conducted for all NSCLC patients first seen by oncologist in 2006 in each site, and independent audits performed to ensure accuracy and reliability. Results: 10 NSCLC-specific QI were established. 531 charts were sampled from 11 sites: 4 sites classified as higher volume (median 2006 NSCLC-patient volume = 314; range 244-862) and 7 as lower volume (median = 129; range 90-217). Patient median age was 68 years; 14% stage-I,...