No abstract

Pulmonary embolism (PE) is associated with mortality. There are many clinical prediction tools to predict early mortality in acute PE but little consensus on which is best. Our study aims to validate existing prediction tools and derive a predictive model that can be applied to all patients with acute PE in both inpatient and outpatient settings. This is a retrospective cohort study of patients with acute PE. For each patient, the Pulmonary Embolism Severity Index (PESI), simplified PESI (sPESI), European Society of Cardiology (ESC), and Angriman scores were calculated. Scores were assessed by the area under the receive-operating curve (AUC) for 30-day, all-cause mortality. To develop a new prognostic model, elastic logistic regression was used on the derivation cohort to estimate β-coefficients of 8 different variables; these were normalized to weigh them. A total of 321 patients (mean age 60±17 years) were included. Overall 30-day mortality was 10.3%. None of the scores performed well; the AUCs for the PESI, sPESI, ESC, and Angriman scores were 0.67 (95% confidence interval [CI], 0.57-0.77), 0.58 (0.48-0.69), 0.65 (0.55-0.75), and 0.67 (0.57-0.76), respectively. Our new prediction model outperformed PESI, with an AUC of 0.82 (95% CI, 0.76-0.88). At a cutoff score of 100, 195 (60.1%) patients were classified as low risk. Thirty-day mortality was 2.1% (95% CI, 0.8%-5.2%) and 23.0% (16.5%-31.1%) for low- and high-risk groups, respectively ( P < .001). In conclusion, we have developed a new model that outperforms existing prediction tools in all comers with PE. However, further validation on external cohorts is required before application.

Introduction: The number of treatment options for myeloma and indolent lymphoma are expanding at an exponential rate, with few direct head-to-head comparisons on which to base efficacy measures. We sought to understand how patients, their caregivers and physicians weigh treatment characteristics in order to come to a decision on which treatment option to pursue. Methods: Patients, their caregivers and physicians were recruited and interviewed until data saturation was reached. A qualitative, thematic analysis was done to identify themes important to each stakeholder. Results: We found that, while all three groups valued efficacy the most, the consideration of other secondary characteristics of the treatment, such as cost, toxicity and logistical issues all differed subtly between the different groups. Patients valued minimising cost and toxicity, even at small trade-offs in efficacy. Caregivers and physicians valued efficacy foremost. Conclusion: Acknowledging and managing these differences is paramount because they influence shared decision-making and may affect patient outcomes in the short term, as well as their more general well-being in the long term.

Introduction Blood transfusion is an integral part of routine outpatient Haematology-Oncology care. Blood product administration requires the concerted effort of nursing and laboratory staff in accredited institutions. One of the challenges with scheduling transfusions is the unpredictability surrounding transfusion requirements and the amount of time required to administer blood products. This mismatch between capacity availability for ad hoc transfusions and clinical need has resulted in physicians pre-booking transfusion slots so patients can be transfused if needed. However, when patients do not require transfusions, their cancelled slots represent capacity which could have otherwise been used to administer chemotherapy. This problem is exacerbated in a pandemic, where demand for inpatient beds necessitates the transition of elective chemotherapy to the outpatient setting insofar as is possible. Aim We hypothesized that reducing the number of transfusion slots booked could help to save healthcare-related costs and improve capacity utilisation. We also sought to right-site blood transfusions away from the chemotherapy infusion unit and to an acute cancer care unit (ACCU). Methods On 1 May 2020, two simple workflow changes were made. First, we introduced a policy where transfusions could not be pre-booked. Physicians were reassured that their patients would be transfused before their patient's crossmatch sample expired and that urgent transfusions would be done on the same day. The only exceptions to this policy were regularly transfused patients (e.g. thalassaemia major patients on chronic transfusions) and infirm patients. Secondly, ad hoc blood transfusions were moved from the chemotherapy unit to the acute cancer care unit. Ad hoc transfusion timing was prioritised according to clinical need. Consecutive patients treated at the National University Cancer Institute, Singapore, from 1 July 2019 to 31 July 2020 were included. Scheduled appointments were extracted from the hospital's scheduling system and analysed. Patients who had appointments booked for blood product transfusions were included. Data was extracted from drug ordering systems to determine the number of blood products administered. Patients were divided into a historical control group (before 1 May) and a study group (after 1 May). The primary outcome measures were cancellation rate (defined as the number of cancellations over total number of slots booked for transfusion) and number of chair hours wasted. Secondary outcome measures included the number of patients who had to be admitted for blood transfusion due to lack of slot availability and cost savings reflected in unit chair hours made available. Categorical data were analysed by the chi-square test. Analysis was done with SPSS v22 (IBM, USA). Results Between 1 July 2019 and 31 July 2020, a total of 3144 slots were booked for transfusion. Each slot was booked for four hours. 1548 blood products were administered. In the control group, there were 1630 cancellations. This equated to 6520 hours of chemotherapy chair time (average of 652 hours/month). There were no nett cancellations in the study group, as total number transfused exceeded the number booked. Assuming the booking rate would have been similar without our intervention, the study resulted in 1956 unutilised chair hours saved. This reflects capacity created for administration of chemotherapy, and cost savings of 1956x, where x is the unit cost of one chair hour. The cancellation rate was 58.3% (1630 cancelled, 2800 booked) in the control group. This decreased to -9.9% in the study group (378 administered (i.e., no nett cancellations), 344 booked, p&lt;0.001, Figure 1). No patients had to be admitted for elective blood transfusion after 1 May 2020. One patient had to be admitted emergently for blood transfusion because of concurrent cardiac failure. The primary reason for admission was intravenous diuresis. All ad hoc transfusions were administered in ACCU. Conclusion Efficient utilisation brought about by two simple workflow changes can help to create capacity and save costs. Such strategies are especially critical in a pandemic, where healthcare resources are under major strain and existing capacity must be maximised. Disclosures No relevant conflicts of interest to declare.

Introduction Chemotherapy is a dynamic, complex process involving cross-functional healthcare teams and comprises dosing, scheduling, safety checks, compounding and administration. Coupled with team silos, legacy systems, escalating workload and cost, efficient chemotherapy delivery is increasingly challenging, resulting in negative staff and patient experience. A design thinking methodology focused on end-users is ideal for addressing complex problems with no clear best practices. Aim We hypothesized that a multidisciplinary team using a data-driven, design thinking approach to redesign chemotherapy workflows can reduce time to treatment, improve operational efficiency and staff and patient experience. Methods A process mapping exercise was undertaken to understand the chemotherapy process. Patients and staff from different job groups were shadowed. The problem statement was "60% of patients are waiting more than an hour from their appointment time to start treatment". The following examples of "how might we" questions were used for the ideation phase: 1. How might we increase advanced chemotherapy preparations (premakes) for patients? 2. How might we ensure only premakes are listed in the mornings? Separately, we also designed an anonymized database to track chemotherapy delivery and care provision outcomes by writing an algorithm to link data extracted from appointment, queue management and chemotherapy systems. New workflows were drafted, iterated, and implemented from 1 May 2020 with the following major changes: 1. No same day blood tests and chemotherapy, with physicians reminded to complete chemotherapy orders by 3pm the day before to allow advance compounding. 2. All chemotherapy regimens were consolidated into a directory containing properties like infusion duration, premake eligibility (based on drug stability and cost) and other scheduling characteristics. This was made searchable via an Excel (Microsoft, USA) algorithm, which also recommended ideal booking slots for the scheduling team. Premakes were prioritized for morning (0830 - 1030) slots. 3. Outcome targets were agreed on and tracked daily. These were made accessible to all staff via a dashboard. The workgroup met weekly to discuss targets, barriers and iterate workflows. Daily, intra-group communication was facilitated by TigerConnect (TigerConnect, USA). We included consecutive outpatients treated at our institution from 1 Jan - 27 Jul 2020. Patients were split into two groups: a historical control group (1 Jan - 30 Apr) and a post-intervention study group (1 May - 27 Jul). The primary outcome measure was the difference between appointment time and time treatment started. Secondary outcome measures included (a) proportion of premade chemotherapy; (b) number of patients starting treatment within an hour of appointment time; and (c) number of patients finishing treatment after 6pm. Continuous data are reported as median (25th-75th centile) and analysed with the Mann-Whitney U test, while categorical data were assessed with the chi-square test. Analysis was done with SPSS v22 (IBM, USA). Results Results are summarized in Table 1. From 1 Jan - 27 July 2020, 14314 treatments were completed. Of these, 5946 (41.5%) were in the 0830 - 1030 slots prioritized for premade chemotherapy. 18.8% of patients arrived after their appointment time. The proportion of premade chemotherapy increased to 70.8% from 30.6% (p&lt;0.001). The median time to start treatment decreased from 83 (51-128) minutes in the control to 49 (27-87) minutes in the study group (p&lt;0.001). This translated into an improvement for the day overall (Figure 1). The proportion of patients with morning appointments starting treatment within 1 hour of their appointment time increased to 58.4% from 31.7% (p&lt;0.001). For the whole day, this increased to 59.6% from 37.8% (p&lt;0.001), resulting in less patients finishing treatment after 6pm (20.5% to 10.6%, p&lt;0.001). Conclusion We have shown that a multidisciplinary group using a data-driven, design thinking approach to address team silos, reorganize and track work processes can improve the time taken to start treatment. Changes were made at no added cost to the healthcare system and using accessible software. Potential cost savings in terms of less overtime claims for staff have yet to be factored in. Addressing patient punctuality and registration and triage processes will help further decrease time to treatment. Disclosures No relevant conflicts of interest to declare.