Publication
Article
Peer Reviewed
Pharmacy Practice in Focus: Oncology
This study evaluated factors affecting the duration of the infusion appointment times for 392 clinical trial patients.
Précis
This study evaluated factors affecting the duration of the infusion appointment times for 392 clinical trial patients. We found no significant correlations, although further research is needed.
Abstract
Objectives
This study identified variables significantly affecting the duration of appointment times for oncology clinical trial patients. Additionally, researchers sought to differentiate between intrinsic and extrinsic delays to determine areas needing optimization.
Study Design
Conducted as a single-center, retrospective quality improvement initiative, the study looked at data from adult patients enrolled in oncology clinical trials at UC Davis Comprehensive Cancer Center in Sacramento, California, from July 1, 2019, to July 1, 2022.
Methods
The analysis involved descriptive statistics and multiple linear regression to evaluate factors affecting infusion appointment duration and arrival-to-chair time. Variables included patient age, number of investigational products administered, appointment time (morning vs afternoon), type and phase of the study, pharmacokinetics at chair time, and the use of interactive response technology kit assignment.
Results
The study examined 392 individuals and found no significant correlations between the analyzed variables and infusion appointment duration (P = .085-.53) or arrival-to-chair time (P = .085-.93).
Conclusion
Although this analysis did not yield definitive results, it highlights the complexity of delays in oncology infusion chair times. The lack of significant correlations suggests intricate interactions among various factors, emphasizing the need for further research to uncover these complexities and optimize appointment efficiency for patients in clinical trials.
Introduction
Image credit: Seventyfour | stock.adobe.com
The UC Davis Comprehensive Cancer Center (UCDCCC) in Sacramento, California, is a National Cancer Institute–designated comprehensive cancer center that conducts early-phase clinical trial research, operating hundreds of oncology clinical trials annually. UCDCCC is the main hub for cancer treatment in the Central Valley region. High patient volume has affected the infusion center schedule, leading to a long wait list. Patients who travel long distances, such as from Southern California or out of state, need to complete multiple care-related activities in a single day, including obtaining laboratory tests and seeing their provider, all on top of receiving their cancer treatment. This is especially the case for clinical trial patients seeking care at UCDCCC because their local institution may not conduct clinical trials or be permitted to conduct study-related activities for the trial on which they are enrolled.
Typical reasons for extended infusion appointments, from a patient scheduling perspective, include satisfying trial requirements. Examples include accommodating predose and up to 10 hours postdose pharmacokinetic (PK) blood samples, infusions with durations of several hours, extensive postdose observation and monitoring periods, and numerous infusions. Examples of extenuating circumstances faced by many sites, which are not specific to clinical trials themselves, include patients arriving late, clinic providers experiencing delays in starting at the scheduled time, and other infusion patients utilizing a chair longer than expected. Although some of these reasons are consistent with challenges faced by standard of care, several are unique to clinical trials. Irrespective of the reason for prolonged treatment days, it is important to reduce patient chair time and improve care experiences at UCDCCC.
Many US and international institutions have experienced similar challenges with infusion center efficiency and attempted to create tools to shorten infusion chair wait times and improve patient satisfaction. Examples include developing simulation models, process mapping, and Plan-Do-Study-Act cycles.1-5
In 2023, a quality improvement (QI) study was performed at UC Davis Health to identify areas of intervention for increased workflow efficiency and decreased infusion chair time for clinical trial patients. This QI project was completed through EPIC (electronic health record software)-generated reports and examined the current workflow by calculating the time intervals for the following: comprehensive metabolic panel and complete blood count laboratory processing, clinical trial medication orders signed by physician to medication verification by pharmacist, patient arrival to study medication dispense, patient arrival to administration of study medication, and the time of appointment (morning [am] or afternoon [pm]). No significant findings were identified, aside from the fact that patients with am appointments waited, on average, 10 minutes longer than patients with pm appointments. It is unclear whether am appointments happened to be generally longer, or because of the complexity of treatment (eg, the patient had multiple study medications administered at the same visit, multiple PK blood draws), these patients were scheduled in the am to ensure treatment could be completed during operating infusion hours.
These follow-up study results extend the insights gained from the aforementioned 2023 UC Davis Health study that noted a significant difference between am and pm appointment delays, confirming trends identified in a 2019 study by the Infusion Efficiency Workgroup under the National Comprehensive Cancer Network.6
The study further investigated the variances in infusion chair time optimization for oncology clinical trial patients by comparing the duration of the appointment vs examined variables and arrival-to-chair time vs examined variables to ascertain their impact on the chair time duration.
Methods
This study was a single-center, retrospective, continuous quality improvement study. An EPIC report was generated for all adult patients with a scheduled infusion clinic appointment between July 1, 2019, and July 1, 2022, who were enrolled in a clinical trial that involved infusion treatment. From the report, the following variables were collected and examined per patient and infusion appointment encounter, with up to the first 3 encounters included for each patient: age, am or pm appointment, type of study, phase of the study, requiring PK blood draw at chair time, requiring investigational product (IP) supply allocated by interactive response technology (IRT) kit assignment, and number of IPs administered.
The primary objective was to determine any variable(s) that significantly affects the duration of the appointment time for oncology clinical trial patients.
The secondary objective was to determine whether the intrinsic delay from the variable(s) differs from the extrinsic delay that needs optimization. Examples of intrinsic variables include type of study (industry, national, investigator-initiated), phase of study (eg, phase 1, 2, 3, 4), requiring PK blood draw at chair time, requiring IP IRT kit assignment, and number of IPs administered; these are inherent to the clinical trial process and not directly controllable by the clinical team and are not changeable (as is age). Extrinsic variables are am/pm appointment times and arrival-to-chair time, which reflect elements that may require external optimization. To compare the 2 types of delays, we examined arrival-to-chair time (extrinsic variable) to all mentioned intrinsic variables.
Descriptive statistics were used to summarize the data. This included calculating continuous variables’ mean, median, SD, and range. Frequency and percentage were used for categorical variables. These descriptive analyses helped to understand the general characteristics of the data set and the distribution of appointment durations.
A multiple linear regression analysis was used to examine the relationship between the dependent variable (duration of appointment time) and the independent variables (am/pm appointment, type of study, phase of study, requiring PK blood draw at chair time, requiring IP IRT kit assignment, number of IP administered, and age). This model helped to determine which variables were statistically significant predictors of appointment duration and the extent of their impact. A separate regression analysis focused on intrinsic vs extrinsic variables to help identify factors needing optimization through policy changes or additional resources.
Results
For this study, data from the 392 participants included in the final 2023 study analysis were examined. Most patients (59.4%) received 1 IP during their encounter, did not get PK blood draw done at chair time (62.7%), did not get IRT assignment for IP (82.7%), had equal occurrence of industry study (41.4%) vs federal study (41.3%), and participated in a phase 1 or phase 1/2 study (41.2%). Additionally, 77.9% of appointments were in am vs pm (22.1%) (Table 1).
IP, investigational product; IIT, investigator-initiated trial; IRT, interactive response technology, PK, pharmacokinetics.
For the primary objective, For the primary objective, the appointment time (am/pm) was not significantly associated with the appointment duration (P = .2776). The model showed that pm appointments were, on average, 41 minutes shorter, although this difference is not statistically significant. The type of study (ie, industry, investigator-initiated trials [IITs], national) was also not significantly associated with appointment duration (overall model P = .4475). Each type of study showed no significant impact on the duration compared with the intercept. The phase of the study showed some indication trending toward association with the duration of the appointment (overall model P = .0555), with phase 3 studies being associated with a significant increase in appointment duration by 98.4 minutes (P = .0433) (Figure). Whether requiring PK at chair time was performed (Yes [Y]/No [N]) was not significantly associated with the duration of the appointment (P = .5344), and whether requiring IP IRT kit was assigned (Y/N) also did not significantly affect the duration of the appointment (P = .254). Age was significantly associated with the duration of the appointment (P = .0443), with each year increase in age reducing the appointment duration by 2.79 minutes. The number of IPs administered showed some indication of association with appointment duration (overall model P = .085), particularly for category 3, which significantly increased the duration by 151.21 minutes (P = .0127) (Table 2).
For the secondary objective, the time of the scheduled appointment (am/pm) showed no significant association with arrival-to-chair time (P = .2961). The pm appointments were associated with increased arrival-to-chair time by 5.41 minutes, which was not statistically significant. The type of study (ie, industry, IIT, national) did not significantly affect arrival-to-chair time (overall model P = .2747), with no significant differences observed across the types of study mentioned. The phase of the study had a suggestive, yet not universally significant, association with arrival-to-chair time (overall model P = .0851). Specifically, phase 2 and phase 3 studies showed a significant increase in arrival-to-chair time by 20.15 minutes (P = .0322), indicating some variability based on the study phase. Whether requiring PK at chair time did not significantly influence arrival-to-chair time (P = .267), indicating no significant impact of this variable. Requiring IP IRT kit assignment also showed no significant effect on arrival-to-chair time (P = .6808). Age was not significantly associated with arrival-to-chair time (P = .8127), with a coefficient indicating a negligible change in arrival-to-chair time per year of age. The number of IPs administered did not show a significant association with arrival-to-chair time (overall model P = .9305), indicating no significant impact on arrival-to-chair time delay (Table 3).
IP, investigational product; IRT, interactive response technology; PK, pharmacokinetics.
IP, investigational product; IRT, interactive response technology; PK, pharmacokinetic.
Discussion
The findings from our comprehensive analysis underscore the complexity of optimizing infusion chair time for patients with cancer enrolled in clinical trials. Despite initial assumptions about the potential impact of various intrinsic and extrinsic variables, our results suggest that few factors significantly influence the duration of patient appointments and arrival to chair time.
One noteworthy result involved the phase of the clinical trial. Phase 3 studies were associated with significantly longer appointment durations, with an increase of approximately 98.4 minutes. This finding was unexpected because phase 1 studies usually have many more checks and more extensive data collection with PKs, etc. Furthermore, phase 2 and phase 3 studies showed a significant increase in arrival-to-chair time, particularly reflecting the complexity and resource allocation necessary at these stages. These results highlight the need for targeted strategies to manage the higher demands of advanced trial phases effectively.
Age also emerged as a significant factor, albeit in a somewhat unexpected direction. With each incremental year in age, appointment durations decreased slightly by approximately 2.79 minutes. This counterintuitive finding might be explained by more streamlined or less intensive treatment regimens often prescribed for older patients or perhaps more experienced patient handling by clinical staff. Moreover, this reduction could be reflective of an adjustment in clinical trial protocols or the health care team’s more efficiently adapting to older adults.
Interestingly, most other intrinsic variables—including the type of study, requiring PK blood draw at chair time, and requiring IP IRT kit assignment—showed no significant association with either the duration of the appointments or the arrival-to-chair time; this was a surprise. Time is added to the appointment for IRT kit assignment because it requires the pharmacy to wait on the assignment. These findings suggest that such intrinsic factors might be well managed within current clinical workflows or that their impact is minimal compared with other elements.
As for extrinsic variables, the timing of appointments (am/pm) was anticipated to influence efficiency potentially, but the data did not support this hypothesis. Although pm appointments were, on average, shorter by 41 minutes, this difference was not statistically significant. It is unclear whether am appointments happened to be generally longer, or because of the complexity of treatment (eg, the patient had multiple study medications administered at the same visit, multiple PK blood draws, etc), these patients were scheduled in the am to ensure treatment could be completed during operating infusion hours. Therefore, scheduling adjustments alone might not yield substantial improvements in chair-time efficiency.
The discovery that the number of IPs administered during a visit also significantly extended appointment times by approximately 151.21 minutes when a patient receives 3 IPs underscores the logistical and procedural burdens associated with multi-drug protocols. This insight points to the need for specialized processes or additional support for handling multi-IP sessions to reduce inefficiencies.
Although the current study results provide valuable insights into the factors influencing infusion chair time for oncology clinical trial patients, several limitations must be acknowledged.
First, the study duration of approximately 3 years might not be sufficient to capture longer-term trends and variations in clinical trial workflows and infusion times. Seasonal variations, periodic protocol changes, and other temporal factors could play a role in influencing appointment durations, which might not be fully encapsulated within this relatively short time frame.
Second, as a single-center study conducted within the UC Davis Health System, the findings may have limited generalizability to other institutions. The workflows, resources, and patient demographics at our center may differ from those at other comprehensive cancer centers or community treatment facilities. Consequently, the results might not be entirely applicable to different clinical settings with varying operational protocols or patient populations.
Last, the study involved nurses manually inputting time from the patient’s arrival to the end of the appointment, which introduced a potential for human error and variability in data accuracy. Relying on manual recording can lead to inconsistencies in time documentation, which may have affected the precision of our findings. Automated data collection systems could yield more reliable and consistent data, minimizing the risk of inaccuracies in manual entry.
Considering these limitations, future research should aim for longer study durations to better capture temporal variability, multicenter collaborations to enhance generalizability, and the implementation of automated time-tracking systems to improve data accuracy. Addressing these areas will help refine our understanding of infusion chair time optimization and lead to more robust and widely applicable conclusions.
Conclusion
into the factors influencing infusion chair time among patients with cancer participating in clinical trials. Although the patient’s age and the phase of the study were significantly associated with changes in appointment duration, most other intrinsic variables examined, including the type of study, PK at chair time, and IRT kit assignment, did not have a statistically significant impact. Moreover, extrinsic variables such as the time of the appointment (am/pm) also showed no significant effect on either the overall appointment duration or arrival-to-chair time.
The significant increases in duration for phase 3 studies and appointments involving multiple IPs suggest specific areas where targeted interventions could be effective. For example, additional staffing or dedicated resources during phase 3 trials might help to manage the increased complexity and requirements associated with treatments for these types of trials. Likewise, optimizing processes for multi-drug administrations could reduce procedural delays.
Kaycey Pearce, PharmD, is a clinical oncology pharmacist in the Department of Pharmacy and Investigational Drugs Services at Samuel Oschin Cancer Center at Cedars-Sinai in Los Angeles, California, and a former PGY-2 investigational drugs and research pharmacy resident at UC Davis Health in Sacramento, California.
Jennifer Murphy, PharmD, BCOP, BCPS, is a senior pharmacist in the Oncology & Investigational Drug Service at UC Davis Health in Sacramento, California.
The findings emphasize that although certain intrinsic factors are inherent to treatments associated with clinical trial protocols and may require tailored management strategies, many current workflow components are already optimized to a degree that limits further improvements through changes in those areas alone. Future efforts might focus on refining a standardization for clinical trials to ensure consistent care and remove the risk of error caused by different workflows specific to more complex trial phases, and administration of multi-agent regimens to achieve more noticeable enhancements in efficiency and reduction in appointment duration.
This study serves as a foundation for further research and quality improvement initiatives aimed at enhancing the experience and care of oncology trial patients. Continuous assessment and incremental adjustments, particularly focused on the identified significant factors, will be essential in moving toward more efficient and patient-friendly clinical trial processes.
REFERENCES
