Optimizing time to IV methotrexate initiation for planned admissions to the inpatient hematology/oncology service Free

Background Parkland Hospital serves as a safety net for Dallas County and provides comprehensive cancer services. Efficient utilization of limited physical and financial resources is essential to providing high quality care in a safety net system.

Delays in intravenous methotrexate (IV MTX) initiation on the inpatient hematology/oncology service impact hospitalization length, cause inefficient utilization of hospital beds, and decrease patient satisfaction. Furthermore, late time-of-day for IV MTX initiation places critical medical decisions (made at specific timepoints) into the overnight window when there is decreased staffing, raising potential safety concerns.

As part of the American Society of Hematology Quality Improvement Training Institute, we assembled a multidisciplinary quality improvement team to evaluate the IV MTX workflow. Co-primary aims included optimizing the median time from admission to initiation and time-of-day of initiation of IV MTX. Secondary aims included improving median time to urine alkalinization, median patient arrival time (excluding patients seen same-day in clinic), and mean hospitalization length.

Methods A consensus workflow for IV MTX admissions was defined through multidisciplinary collaboration. Baseline data collection of key timepoints was performed on all planned admissions for IV MTX over an 8 month period; encounters with delays unrelated to the IV MTX workflow were excluded. Three subsequent Plan-Do-Study-Act (PDSA) cycles lasting 2 months each were conducted utilizing insights from analysis of prior periods.

During PDSA cycle 1, a standardized recommended patient arrival time of 7:30am was introduced along with a structured system for pre-admission reminder calls by unit staff. A novel checklist for methotrexate admissions was created to enhance recognition of key timepoints across nursing handoffs, and in-service education was conducted with the nursing staff.

During PDSA cycle 2, initial IV sodium bicarbonate orders were de-coupled from treatment plans and placed into a separate order panel specific for IV MTX admissions, thereby allowing admitting residents to begin the alkalinization process immediately upon arrival (without requiring an “okay to treat” order to be placed by a fellow/attending). Additionally, the frequency of urine pH measurements was standardized through this panel.

During PDSA cycle 3, the order panel was refined to include orders pertaining to peripheral IV placement and utilization of pre-existing central venous access devices. Additionally, a one-time oral sodium bicarbonate load (1950mg) was added to mitigate inefficiencies related to delays in IV access preventing timely initiation of IV bicarbonate infusions.

Results During the baseline data period (11/1/23-6/30/24), 29 eligible admissions for IV MTX were identified. During PDSA cycle 1 (8/1/24-9/30/24), 7 admissions were identified. During PDSA cycle 2 (11/1/24-12/31/24), 11 admissions were identified. During PDSA cycle 3 (3/1/25-4/30/25), 12 admissions were identified. Processes stability was confirmed (via statistical process control charts) in each PDSA cycle for all primary and secondary aims (with the exception of time to urine alkalinization in PDSA cycle 3).

The median time from admission to IV MTX initiation during the baseline period was 12.28 hours; this decreased to 10.83, 8.60, and 9.10 hours after PDSA cycles 1, 2 and 3, respectively. Median time-of-day to start IV MTX (10:26pm, 7:13pm, 8:06pm, 6:11pm) was similarly improved with a substantial increase in the percentage of patients initiated before the 7pm shift change (7%, 43%, 45%, 50%).

Median time to urine alkalinization (8.23, 8.47, 5.75, 6.34 hours) was improved following PDSA cycle 2. Median time of admission (9:46am, 8:23am, 8:19am, 9:55am) was improved in PDSA cycles 1 and 2. Mean hospitalization length was not significantly decreased (4.38, 4.50, 4.14, 4.22 days) in any of the PDSA cycles.

Conclusions Substantial improvements in time from admission to IV MTX initiation and time-of-day of IV MTX initiation were seen following three PDSA cycles targeting key areas identified via a multidisciplinary quality improvement initiative. These improvements decrease the number of critical decisions made overnight and facilitate more efficient bed flow. An extended analysis is planned to confirm sustainability of all improvements and address limitations related to variability from patient-specific factors and small sample sizes.