Gastrointestinal Cancer, Gynecologic Cancer, Other Cancer, Thoracic Cancer
Conditions
Keywords
Other Cancer, Digital health, Patient reported outcomes, Symptom management
Brief summary
Deficits in the management of common symptoms cause substantial morbidity for cancer patients.Because the health care delivery system is structured to be reactive and not proactive, there are missed opportunities to optimize symptom control. Growth in Internet access and proliferation of smartphones has created an opportunity to re-engineer cancer care delivery. Electronic symptom tracking and feedback is a promising strategy to improve symptom control. Electronic patient reported outcome (ePRO) monitoring of cancer symptoms has been shown to decrease symptom burden, improve quality of life, reduce acute care and even extend survival. SIMPRO will use functioning ePRO prototypes to create and refine the electronic symptom management system eSyM
Detailed description
A multi-disciplinary team of investigators from 6 health systems have formed the Symptom Management IMplementation of Patient Reported Outcomes in Oncology (SIMPRO) Research Center. SIMPRO will use functioning ePRO prototypes to create and refine the electronic symptom management system eSyM. eSyM is the name of the platform the team will refine, integrate, implement and evaluate. eSyM addresses each of the 4 evidence gaps by: * Implementing eSyM in cancer centers in small, rural or community-based systems. * Integrating eSyM into the EHR (electronic health record) of the predominant vendor used nationwide. * Leveraging evidence-based tools, patient engagement, and population management. * Executing this work using the Consolidated Framework for Implementation Research across all phases to maximize the chances that eSyM and similar systems achieve their intended goals and decrease the morbidity of cancer treatment at a population level. This project contains 5 activities: 1. Obtain stakeholder feedback 2. Build and deploy eSyM 3. Pilot test eSyM 4. Pragmatic stepped-wedge cluster randomized trial 5. Integration of eSyM data to develop algorithms to estimate the risk of experiencing an outcome, including, but not limited to, ED usage and hospitalization among cancer patients
Interventions
OTHERStakeholder Survey (Control Period)
Before eSyM go-live, study team members from each site will solicit input via emailed survey, remote meetings and/or in-person meeting on the use of ePROs in oncology from stakeholders to obtain input regarding adaptation, anticipated challenges, and implementation.
OTHERStakeholder Survey (Intervention Period)
After eSyM go-live and on an ongoing basis, we will evaluate the implementation process at each of the sites with a focus on adoption, appropriateness, acceptability, sustainability, penetration, and scalability. We will do so through emailed surveys and/or discussions with health system leadership, clinicians, clinic support staff, and informatics/IT staff.
OTHERQualitative Interview
A small subset of patients and stakeholders were invited to take part in qualitative interviews after the eSyM trial rollout.
OTHERSASS Questionnaire
A subset of control and intervention patients will be asked to complete a research questionnaire called the SASS Questionnaire asking about their Self-efficacy, Attainment of information needs, Symptom burden, and Satisfaction with care.
OTHEReSyM
The electronic symptom management (eSyM) program is the EHR-integrated ePRO program being evaluated through this trial.
Sponsors
National Cancer Institute (NCI)
RTI International
Baptist Memorial Health Care Corporation
Dartmouth-Hitchcock Medical Center
MaineHealth
West Virginia University
Lifespan
Dana-Farber Cancer Institute
Study design
Allocation
RANDOMIZED
Intervention model
PARALLEL
Primary purpose
HEALTH_SERVICES_RESEARCH
Masking
NONE
Intervention model description
The primary study design is a hybrid effectiveness-implementation stepped-wedge cluster randomized trial involving several different research activities to assess outcomes.
Eligibility
Sex/Gender
ALL
Age
18 Years to No maximum
Healthy volunteers
No
Inclusion criteria
Stakeholder Feedback and Stakeholder Qualitative Interviews Population: * Age ≥ 18 years * The potential stakeholders are: patient advisory council members, health system leaders, clinicians, clinic support staff/administration, IT/Informatics staff Cluster Randomized Trial, Patient QualitativeInterviews, Pilot Testing & SASS Questionnaire Population: * Age ≥ 18 years * Priority population will be patients who meet one of the following: * Suspected thoracic cancer \[lung or bronchus\] AND is inpatient following thoracic surgery. * Suspected gastrointestinal cancer \[colorectal, pancreas, liver/biliary, esophagus,or gastric\] AND is inpatient following gastrointestinal surgery. * Suspected gynecologic cancer \[ovary, uterus, or cervix\] AND is inpatient following gynecologic surgery. * Diagnosis of thoracic cancer \[lung or bronchus\] AND scheduled to start a new treatment plan for thoracic cancer. * Diagnosis of gastrointestinal cancer \[colorectal, pancreas, liver/biliary, esophagus,or gastric\] AND scheduled to start a new treatment plan for gastrointestinal cancer. * Diagnosis of gynecologic cancer \[ovary, uterus, or cervix\] AND scheduled to start a new treatment plan for gynecologic cancer. * Total population allowed to use eSyM: * Any patient at any participating site.
Exclusion criteria
\- Participants not meeting the inclusion critera above.
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Emergency Department Treat/Release [EDTR] Event Occurrence Status at Day 30 | 30 days | The primary study outcome of the stepped wedge cluster RCT (randomized controlled trial) is the EDTR rate. This outcome will be defined in relation to the date of discharge from hospital (surgical cohort) or the initiation date of a new intravenous chemo regimen (medical oncology cohort). This is a binary outcome and we will analyze the absolute difference and odds ratio for the medical oncology and surgical cohorts combined and independently. |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Admissions Event Occurrence Status at Day 30 | 30 Days | This outcome will be defined in relation to the date of discharge from hospital (surgical cohort) or the initiation date of a new intravenous chemo regimen (medical oncology cohort). This is a binary outcome and we will analyze the absolute difference and odds ratio for the medical oncology and surgical cohorts combined and independently. |
| Admissions Event Occurrence Status at Day 90 | 90 Days | This outcome will be defined in relation to the date of discharge from hospital (surgical cohort) or the initiation date of a new intravenous chemo regimen (medical oncology cohort). This is a binary outcome and we will analyze the absolute difference and odds ratio for the medical oncology and surgical cohorts combined and independently. |
| Difference in Fatigue PROMIS Scores Reported by Participants Before and After eSyM Exposure | 30-90 days before and after eSyM go-live | Difference in PROMIS fatigue scores between the pre-live and post-live eSyM cohorts. Since the two cohorts were comprised of different patients, comparisons were performed at the population level using mean T-scores. Fatigue was assessed using the PROMIS Fatigue 4-item short form with responses converted to T-scores (score range: 50 indicates the population mean with a standard deviation of 10; decrease in score indicates a positive change). All PROMIS items were scored using the standardized scoring tool kit. The differences in mean T-scores and the corresponding 95% confidence intervals (CI) were calculated for each PROMIS item, comparing the post-live versus pre-live cohort. When comparing group-level change in PROMIS scores, even trivial differences can be statistically significant if the sample size is large enough. The threshold to evaluate within-group change or to make a between-group comparison generally ranges between 2 and 6 T-score points (Terwee et al., 2021). |
| Difference in Depression PROMIS Scores Reported by Participants Before and After eSyM Exposure | 30-90 days before and after eSyM go-live | Difference in PROMIS depression scores between the pre-live and post-live eSyM cohorts. Since the two cohorts were comprised of different patients, comparisons were performed at the population level using mean T-scores. Depression was assessed using the PROMIS Emotional Distress-Depression 4-item short form with responses converted to T-scores (score range: 50 indicates the population mean with a standard deviation of 10; decrease in score indicates a positive change). All PROMIS items were scored using the standardized scoring tool kit. The differences in mean T-scores and the corresponding 95% confidence intervals (CI) were calculated for each PROMIS item, comparing the post-live versus pre-live cohort. When comparing group-level change in PROMIS scores, even trivial differences can be statistically significant if the sample size is large enough. The threshold to evaluate within-group change or to make a between-group comparison generally ranges between 2 and 6 T-score points. |
| Difference in Anxiety PROMIS Scores Reported by Participants Before and After eSyM Exposure | 30-90 days before and after eSyM go-live | Difference in PROMIS anxiety scores between the pre-live and post-live eSyM cohorts. Since the two cohorts were comprised of different patients, comparisons were performed at the population level using mean T-scores. Anxiety was assessed using the PROMIS Emotional Distress-Anxiety 4-item short form with responses converted to T-scores (score range: 50 indicates the population mean with a standard deviation of 10; decrease in score indicates a positive change). All PROMIS items were scored using the standardized scoring tool kit. The differences in mean T-scores and the corresponding 95% confidence intervals (CI) were calculated for each PROMIS item, comparing the post-live versus pre-live cohort. When comparing group-level change in PROMIS scores, even trivial differences can be statistically significant if the sample size is large enough. The threshold to evaluate within-group change or to make a between-group comparison generally ranges between 2 and 6 T-score points. |
| Difference in Pain Interference PROMIS Scores Reported by Participants Before and After eSyM Exposure | 30-90 days before and after eSyM go-live | Difference in PROMIS pain interference scores between the pre-live and post-live eSyM cohorts. Since the two cohorts were comprised of different patients, comparisons were performed at the population level using mean T-scores. Pain interference was assessed using the PROMIS Pain Interference 4-item short form with responses converted to T-scores (score range: 50 indicates the population mean with a standard deviation of 10; decrease in score indicates a positive change). All PROMIS items were scored using the standardized scoring tool kit. The differences in mean T-scores and the corresponding 95% confidence intervals (CI) were calculated for each PROMIS item, comparing the post-live versus pre-live cohort. When comparing group-level change in PROMIS scores, even trivial differences can be statistically significant if the sample size is large enough. The threshold to evaluate within-group change or to make a between-group comparison generally ranges between 2 and 6 T-score points. |
| Emergency Department Treat/Release [EDTR] Event Occurrence Status at Day 90 | 90 days | This outcome will be defined in relation to the date of discharge from hospital (surgical cohort) or the initiation date of a new intravenous chemo regimen (medical oncology cohort). This is a binary outcome and we will analyze the absolute difference and odds ratio for the medical oncology and surgical cohorts combined and independently. |
| Difference in Physical Function PROMIS Scores Reported by Participants Before and After eSyM Exposure | 30-90 days before and after eSyM go-live | Difference in PROMIS physical function scores between the pre-live and post-live eSyM cohorts. Since the two cohorts were comprised of different patients, comparisons were performed at the population level using mean T-scores. Physical function was assessed using the PROMIS Physical Function 4-item short form with responses converted to T-scores (score range: 50 indicates the population mean with a standard deviation of 10; decrease in score indicates a positive change). All PROMIS items were scored using the standardized scoring tool kit. The differences in mean T-scores and the corresponding 95% confidence intervals (CI) were calculated for each PROMIS item, comparing the post-live versus pre-live cohort. When comparing group-level change in PROMIS scores, even trivial differences can be statistically significant if the sample size is large enough. The threshold to evaluate within-group change or to make a between-group comparison generally ranges between 2 and 6 T-score points). |
| Emergency Department Treat/Release [EDTR] Event Occurrence Status at Day 30 (Comparing eSyM Intervention Responders Versus Non-Responders) | 30 days | The secondary study outcome of the stepped wedge cluster RCT (randomized controlled trial) is the difference in the EDTR rate between eSyM responders and non-responders. This outcome will be defined in relation to the date of discharge from hospital (surgical cohort) or the initiation date of a new intravenous chemo regimen (medical oncology cohort). This is a binary outcome and we will analyze the absolute difference and odds ratio for the medical oncology and surgical cohorts combined and independently. NOTE: The total number of intervention participants in this analysis is greater than the number of intervention participants included in the primary analyses (outcomes 1-4). For outcomes outcomes 1-4, we used the patient's first event during the entire study period. To conduct this analysis of eSyM responders, we used the patient's first event during the intervention period. Approximately 10% of patients in this analysis were counted as a control event in the prior analyses. |
| Emergency Department Treat/Release [EDTR] Event Occurrence Status at Day 90 (Comparing eSyM Intervention Responders Versus Non-Responders) | 90 days | This outcome will be defined in relation to the date of discharge from hospital (surgical cohort) or the initiation date of a new intravenous chemo regimen (medical oncology cohort). This is a binary outcome and we will analyze the absolute difference and odds ratio for the medical oncology and surgical cohorts combined and independently. NOTE: The total number of intervention participants in this analysis is greater than the number of intervention participants included in the primary analyses (outcomes 1-4). For outcomes outcomes 1-4, we used the patient's first event during the entire study period. To conduct this analysis of eSyM responders, we used the patient's first event during the intervention period. Approximately 10% of patients in this analysis were counted as a control event in the prior analyses. |
| Admissions Event Occurrence Status at Day 30 (Comparing eSyM Intervention Responders Versus Non-Responders) | 30 Days | This outcome will be defined in relation to the date of discharge from hospital (surgical cohort) or the initiation date of a new intravenous chemo regimen (medical oncology cohort). This is a binary outcome and we will analyze the absolute difference and odds ratio for the medical oncology and surgical cohorts combined and independently. NOTE: The total number of intervention participants in this analysis is greater than the number of intervention participants included in the primary analyses (outcomes 1-4). For outcomes outcomes 1-4, we used the patient's first event during the entire study period. To conduct this analysis of eSyM responders, we used the patient's first event during the intervention period. Approximately 10% of patients in this analysis were counted as a control event in the prior analyses. |
| Admissions Event Occurrence Status at Day 90 (Comparing eSyM Intervention Responders Versus Non-Responders) | 90 Days | This outcome will be defined in relation to the date of discharge from hospital (surgical cohort) or the initiation date of a new intravenous chemo regimen (medical oncology cohort). This is a binary outcome and we will analyze the absolute difference and odds ratio for the medical oncology and surgical cohorts combined and independently. NOTE: The total number of intervention participants in this analysis is greater than the number of intervention participants included in the primary analyses (outcomes 1-4). For outcomes outcomes 1-4, we used the patient's first event during the entire study period. To conduct this analysis of eSyM responders, we used the patient's first event during the intervention period. Approximately 10% of patients in this analysis were counted as a control event in the prior analyses. |
| Difference in Self-Efficacy PROMIS Scores Reported by Participants Before and After eSyM Exposure | 30-90 days before and after eSyM go-live | Difference in PROMIS self-efficacy scores between pre-live and post-live eSyM cohorts. Since the two cohorts were comprised of different patients, comparisons were performed at the population level using mean T-scores. Self-efficacy was assessed using the PROMIS Self-Efficacy for Managing Symptoms 8-item short form with responses converted to T-scores (score range: 50 indicates the population mean with a standard deviation of 10; decrease in score indicates a positive change). All PROMIS items were scored using the standardized scoring tool kit. The differences in mean T-scores and the corresponding 95% confidence intervals were calculated for each PROMIS item, comparing the post-live versus pre-live cohort. When comparing group-level change in PROMIS scores, even trivial differences can be statistically significant if the sample size is large enough. The threshold to evaluate within-group change or to make a between-group comparison generally ranges between 2 and 6 T-score points. |
Countries
United States
Participant flow
Recruitment details
Eligible patients and stakeholders were recruited at the six participating sites between Jan. 2018 and Feb. 2023. As this was a stepped wedge trial, each period has distinct enrollment. For the entire trial, grand total enrollment was 39,942 patients (21,112 control patients and 18,830 intervention patients).
Participants by arm
| Arm | Count |
|---|---|
| eSyM Control Participants (Medical Oncology and Surgery Participants Combined) These patients (and/or proxy) were seen at a participating site prior to the trial rollout of eSyM and were not exposed to the eSyM intervention.
This group includes the following trial subsets:
* patients NOT exposed to the eSyM intervention through the stepped wedge cluster randomized trial (control cohort)
* patients participating in the SASS control questionnaire
Please note: demographics for both medical oncology and surgical patients are combined because the primary analysis evaluated the entire cohort combined together. | 21 |
| eSyM Intervention Participants (Medical Oncology and Surgery Participants Combined) These patients (and/or proxy) were seen at a participating site after the trial rollout of eSyM and were exposed to the eSyM intervention.
This group includes the following trial subsets:
* patients exposed to the eSyM intervention through the stepped wedge cluster randomized trial (intervention cohort)
* patients participating in the SASS intervention questionnaire
* patients participating in qualitative interviews
* patients participating in pilot testing
Please note: demographics for both medical oncology and surgical patients are combined because the primary analysis evaluated the entire cohort combined together. | 18 |
| Total | 39 |
Baseline characteristics
| Characteristic | eSyM Control Participants (Medical Oncology and Surgery Participants Combined) | eSyM Intervention Participants (Medical Oncology and Surgery Participants Combined) | Total |
|---|---|---|---|
| Age, Continuous | 62 years | 65 years | 63 years |
| Race/Ethnicity, Customized Hispanic | 808 Participants | 516 Participants | 1324 Participants |
| Race/Ethnicity, Customized Non-Hispanic Black | 2024 Participants | 1378 Participants | 3402 Participants |
| Race/Ethnicity, Customized Non-Hispanic White | 16903 Participants | 15848 Participants | 32751 Participants |
| Race/Ethnicity, Customized Other | 1377 Participants | 1088 Participants | 2465 Participants |
| Sex/Gender, Customized Female | 14273 Participants | 11929 Participants | 26202 Participants |
| Sex/Gender, Customized Male | 6837 Participants | 6900 Participants | 13737 Participants |
| Sex/Gender, Customized Other | 2 Participants | 1 Participants | 3 Participants |
Adverse events
| Event type | EG000 affected / at risk | EG001 affected / at risk |
|---|---|---|
| deaths Total, all-cause mortality | 4,898 / 21,112 | 4,161 / 18,830 |
| other Total, other adverse events | 0 / 0 | 0 / 0 |
| serious Total, serious adverse events | 0 / 0 | 0 / 0 |
Outcome results
Primary
Emergency Department Treat/Release [EDTR] Event Occurrence Status at Day 30
The primary study outcome of the stepped wedge cluster RCT (randomized controlled trial) is the EDTR rate. This outcome will be defined in relation to the date of discharge from hospital (surgical cohort) or the initiation date of a new intravenous chemo regimen (medical oncology cohort). This is a binary outcome and we will analyze the absolute difference and odds ratio for the medical oncology and surgical cohorts combined and independently.
Time frame: 30 days
Population: The primary analysis included data from both medical oncology and surgery. As a secondary analysis, we analyzed the data from MO and Surg separately and estimated the intervention effect for each outcome. Results for both combined and separate analyses are reported below.
| Arm | Measure | Group | Value (NUMBER) |
|---|---|---|---|
| Cluster Randomized Trial (Control Patients) | Emergency Department Treat/Release [EDTR] Event Occurrence Status at Day 30 | Medical Oncology Cohort | 5.8 percentage of patients with event |
| Cluster Randomized Trial (Control Patients) | Emergency Department Treat/Release [EDTR] Event Occurrence Status at Day 30 | Surgery Cohort | 5.3 percentage of patients with event |
| Cluster Randomized Trial (Control Patients) | Emergency Department Treat/Release [EDTR] Event Occurrence Status at Day 30 | Medical Oncology + Surgery Combined Cohort | 5.4 percentage of patients with event |
| Cluster Randomized Trial (Intervention Patients) | Emergency Department Treat/Release [EDTR] Event Occurrence Status at Day 30 | Medical Oncology Cohort | 6.9 percentage of patients with event |
| Cluster Randomized Trial (Intervention Patients) | Emergency Department Treat/Release [EDTR] Event Occurrence Status at Day 30 | Surgery Cohort | 5.8 percentage of patients with event |
| Cluster Randomized Trial (Intervention Patients) | Emergency Department Treat/Release [EDTR] Event Occurrence Status at Day 30 | Medical Oncology + Surgery Combined Cohort | 6.2 percentage of patients with event |
Comparison: Adjusted analysis comparing control versus intervention patients (chemo and surgical patients combined)p-value: 0.21495% CI: [0.944, 1.29]Regression, Logistic
Comparison: Adjusted analysis comparing control versus intervention patients (medical oncology patients only)p-value: 0.56595% CI: [0.715, 1.201]Regression, Logistic
Comparison: Adjusted analysis comparing control versus intervention patients (surgical patients only)p-value: 0.04295% CI: [1.008, 1.49]Regression, Logistic
Secondary
Admissions Event Occurrence Status at Day 30
This outcome will be defined in relation to the date of discharge from hospital (surgical cohort) or the initiation date of a new intravenous chemo regimen (medical oncology cohort). This is a binary outcome and we will analyze the absolute difference and odds ratio for the medical oncology and surgical cohorts combined and independently.
Time frame: 30 Days
Population: The primary analysis included data from both medical oncology and surgery. As a secondary analysis, we analyzed the data from MO and Surg separately and estimated the intervention effect for each outcome. Results for both combined and separate analyses are reported below.
| Arm | Measure | Group | Value (NUMBER) |
|---|---|---|---|
| Cluster Randomized Trial (Control Patients) | Admissions Event Occurrence Status at Day 30 | Medical Oncology | 10.8 percentage of patients with event |
| Cluster Randomized Trial (Control Patients) | Admissions Event Occurrence Status at Day 30 | Surgery | 7.7 percentage of patients with event |
| Cluster Randomized Trial (Control Patients) | Admissions Event Occurrence Status at Day 30 | Medical Oncology + Surgery | 8.5 percentage of patients with event |
| Cluster Randomized Trial (Intervention Patients) | Admissions Event Occurrence Status at Day 30 | Medical Oncology | 10.7 percentage of patients with event |
| Cluster Randomized Trial (Intervention Patients) | Admissions Event Occurrence Status at Day 30 | Surgery | 8.2 percentage of patients with event |
| Cluster Randomized Trial (Intervention Patients) | Admissions Event Occurrence Status at Day 30 | Medical Oncology + Surgery | 9.1 percentage of patients with event |
Comparison: Adjusted analysis comparing control versus intervention patients (medical oncology and surgical patients combined)p-value: 0.97895% CI: [0.881, 1.139]Regression, Logistic
Comparison: Adjusted analysis comparing control versus intervention patients (medical oncology patients only)p-value: 0.12695% CI: [0.697, 1.046]Regression, Logistic
Comparison: Adjusted analysis comparing control versus intervention patients (surgical patients only)p-value: 0.25495% CI: [0.933, 1.302]Regression, Logistic
Secondary
Admissions Event Occurrence Status at Day 30 (Comparing eSyM Intervention Responders Versus Non-Responders)
This outcome will be defined in relation to the date of discharge from hospital (surgical cohort) or the initiation date of a new intravenous chemo regimen (medical oncology cohort). This is a binary outcome and we will analyze the absolute difference and odds ratio for the medical oncology and surgical cohorts combined and independently. NOTE: The total number of intervention participants in this analysis is greater than the number of intervention participants included in the primary analyses (outcomes 1-4). For outcomes outcomes 1-4, we used the patient's first event during the entire study period. To conduct this analysis of eSyM responders, we used the patient's first event during the intervention period. Approximately 10% of patients in this analysis were counted as a control event in the prior analyses.
Time frame: 30 Days
Population: The primary analysis included data from both medical oncology and surgery. As a secondary analysis, we analyzed the data from MO and Surg separately and estimated the intervention effect for each outcome. Results for both combined and separate analyses are reported below.~NOTE: To conduct this analysis of eSyM responders, we used the patient's first event during the intervention period. Approximately 10% of patients in this analysis were counted as a control event in outcomes 1-4.
| Arm | Measure | Group | Value (NUMBER) |
|---|---|---|---|
| Cluster Randomized Trial (Control Patients) | Admissions Event Occurrence Status at Day 30 (Comparing eSyM Intervention Responders Versus Non-Responders) | Medical Oncology | 7.5 percentage of patients with event |
| Cluster Randomized Trial (Control Patients) | Admissions Event Occurrence Status at Day 30 (Comparing eSyM Intervention Responders Versus Non-Responders) | Surgery | 6.2 percentage of patients with event |
| Cluster Randomized Trial (Control Patients) | Admissions Event Occurrence Status at Day 30 (Comparing eSyM Intervention Responders Versus Non-Responders) | Medical Oncology + Surgery | 6.7 percentage of patients with event |
| Cluster Randomized Trial (Intervention Patients) | Admissions Event Occurrence Status at Day 30 (Comparing eSyM Intervention Responders Versus Non-Responders) | Medical Oncology | 12.9 percentage of patients with event |
| Cluster Randomized Trial (Intervention Patients) | Admissions Event Occurrence Status at Day 30 (Comparing eSyM Intervention Responders Versus Non-Responders) | Surgery | 10.5 percentage of patients with event |
| Cluster Randomized Trial (Intervention Patients) | Admissions Event Occurrence Status at Day 30 (Comparing eSyM Intervention Responders Versus Non-Responders) | Medical Oncology + Surgery | 11.5 percentage of patients with event |
Comparison: Adjusted analysis comparing intervention responders versus non-responders (medical oncology and surgical patients combined)p-value: <0.00195% CI: [0.593, 0.682]Regression, Logistic
Comparison: Adjusted analysis comparing intervention responders versus non-responders (medical oncology patients only)p-value: <0.00195% CI: [0.55, 0.679]Regression, Logistic
Comparison: Adjusted analysis comparing intervention responders versus non-responders (surgical patients only)p-value: <0.00195% CI: [0.597, 0.719]Regression, Logistic
Secondary
Admissions Event Occurrence Status at Day 90
This outcome will be defined in relation to the date of discharge from hospital (surgical cohort) or the initiation date of a new intravenous chemo regimen (medical oncology cohort). This is a binary outcome and we will analyze the absolute difference and odds ratio for the medical oncology and surgical cohorts combined and independently.
Time frame: 90 Days
Population: The primary analysis included data from both medical oncology and surgery. As a secondary analysis, we analyzed the data from MO and Surg separately and estimated the intervention effect for each outcome. Results for both combined and separate analyses are reported below.
| Arm | Measure | Group | Value (NUMBER) |
|---|---|---|---|
| Cluster Randomized Trial (Control Patients) | Admissions Event Occurrence Status at Day 90 | Medical Oncology | 20.8 percentage of patients with event |
| Cluster Randomized Trial (Control Patients) | Admissions Event Occurrence Status at Day 90 | Surgery | 12.8 percentage of patients with event |
| Cluster Randomized Trial (Control Patients) | Admissions Event Occurrence Status at Day 90 | Medical Oncology + Surgery | 14.7 percentage of patients with event |
| Cluster Randomized Trial (Intervention Patients) | Admissions Event Occurrence Status at Day 90 | Medical Oncology | 21.9 percentage of patients with event |
| Cluster Randomized Trial (Intervention Patients) | Admissions Event Occurrence Status at Day 90 | Surgery | 13.1 percentage of patients with event |
| Cluster Randomized Trial (Intervention Patients) | Admissions Event Occurrence Status at Day 90 | Medical Oncology + Surgery | 16.4 percentage of patients with event |
Comparison: Adjusted analysis comparing control versus intervention patients (medical oncology and surgical patients combined)p-value: 0.20995% CI: [0.964, 1.183]Regression, Logistic
Comparison: Adjusted analysis comparing control versus intervention patients (medical oncology patients only)p-value: 0.92495% CI: [0.85, 1.159]Regression, Logistic
Comparison: Adjusted analysis comparing control versus intervention patients (surgical patients only)p-value: 0.09995% CI: [0.979, 1.286]Regression, Logistic
Secondary
Admissions Event Occurrence Status at Day 90 (Comparing eSyM Intervention Responders Versus Non-Responders)
This outcome will be defined in relation to the date of discharge from hospital (surgical cohort) or the initiation date of a new intravenous chemo regimen (medical oncology cohort). This is a binary outcome and we will analyze the absolute difference and odds ratio for the medical oncology and surgical cohorts combined and independently. NOTE: The total number of intervention participants in this analysis is greater than the number of intervention participants included in the primary analyses (outcomes 1-4). For outcomes outcomes 1-4, we used the patient's first event during the entire study period. To conduct this analysis of eSyM responders, we used the patient's first event during the intervention period. Approximately 10% of patients in this analysis were counted as a control event in the prior analyses.
Time frame: 90 Days
Population: The primary analysis included data from both medical oncology and surgery. As a secondary analysis, we analyzed the data from MO and Surg separately and estimated the intervention effect for each outcome. Results for both combined and separate analyses are reported below.~NOTE: To conduct this analysis of eSyM responders, we used the patient's first event during the intervention period. Approximately 10% of patients in this analysis were counted as a control event in outcomes 1-4.
| Arm | Measure | Group | Value (NUMBER) |
|---|---|---|---|
| Cluster Randomized Trial (Control Patients) | Admissions Event Occurrence Status at Day 90 (Comparing eSyM Intervention Responders Versus Non-Responders) | Medical Oncology | 18.8 percentage of patients with event |
| Cluster Randomized Trial (Control Patients) | Admissions Event Occurrence Status at Day 90 (Comparing eSyM Intervention Responders Versus Non-Responders) | Surgery | 10.8 percentage of patients with event |
| Cluster Randomized Trial (Control Patients) | Admissions Event Occurrence Status at Day 90 (Comparing eSyM Intervention Responders Versus Non-Responders) | Medical Oncology + Surgery | 13.7 percentage of patients with event |
| Cluster Randomized Trial (Intervention Patients) | Admissions Event Occurrence Status at Day 90 (Comparing eSyM Intervention Responders Versus Non-Responders) | Medical Oncology | 23.7 percentage of patients with event |
| Cluster Randomized Trial (Intervention Patients) | Admissions Event Occurrence Status at Day 90 (Comparing eSyM Intervention Responders Versus Non-Responders) | Surgery | 16.1 percentage of patients with event |
| Cluster Randomized Trial (Intervention Patients) | Admissions Event Occurrence Status at Day 90 (Comparing eSyM Intervention Responders Versus Non-Responders) | Medical Oncology + Surgery | 19.4 percentage of patients with event |
Comparison: Adjusted analysis comparing intervention responders versus non-responders (medical oncology and surgical patients combined)p-value: <0.00195% CI: [0.729, 0.811]Regression, Logistic
Comparison: Adjusted analysis comparing intervention responders versus non-responders (medical oncology patients only)p-value: <0.00195% CI: [0.737, 0.858]Regression, Logistic
Comparison: Adjusted analysis comparing intervention responders versus non-responders (surgical patients only)p-value: <0.00195% CI: [0.686, 0.796]Regression, Logistic
Secondary
Difference in Anxiety PROMIS Scores Reported by Participants Before and After eSyM Exposure
Difference in PROMIS anxiety scores between the pre-live and post-live eSyM cohorts. Since the two cohorts were comprised of different patients, comparisons were performed at the population level using mean T-scores. Anxiety was assessed using the PROMIS Emotional Distress-Anxiety 4-item short form with responses converted to T-scores (score range: 50 indicates the population mean with a standard deviation of 10; decrease in score indicates a positive change). All PROMIS items were scored using the standardized scoring tool kit. The differences in mean T-scores and the corresponding 95% confidence intervals (CI) were calculated for each PROMIS item, comparing the post-live versus pre-live cohort. When comparing group-level change in PROMIS scores, even trivial differences can be statistically significant if the sample size is large enough. The threshold to evaluate within-group change or to make a between-group comparison generally ranges between 2 and 6 T-score points.
Time frame: 30-90 days before and after eSyM go-live
Population: Since only a subset of eSyM patients completed the SASS questionnaire which included PROMIS measures, the analyzed population is smaller than other analytic populations. Distinct surveys were given to medical oncology and surgical patients to account for differences in their care, so the analytic cohorts below are distinct. Results for both the medical oncology and surgical subsets are included below.
| Arm | Measure | Group | Value (GEOMETRIC_MEAN) | Dispersion |
|---|---|---|---|---|
| Cluster Randomized Trial (Control Patients) | Difference in Anxiety PROMIS Scores Reported by Participants Before and After eSyM Exposure | Medical Oncology | 52.4 T-Score | Standard Deviation 9.8 |
| Cluster Randomized Trial (Control Patients) | Difference in Anxiety PROMIS Scores Reported by Participants Before and After eSyM Exposure | Surgery | 49.9 T-Score | Standard Deviation 9.3 |
| Cluster Randomized Trial (Intervention Patients) | Difference in Anxiety PROMIS Scores Reported by Participants Before and After eSyM Exposure | Medical Oncology | 50.2 T-Score | Standard Deviation 9.5 |
| Cluster Randomized Trial (Intervention Patients) | Difference in Anxiety PROMIS Scores Reported by Participants Before and After eSyM Exposure | Surgery | 47.4 T-Score | Standard Deviation 8.8 |
Comparison: Multivariable regression analyses (medical oncology cohort only)p-value: 0.1795% CI: [-2.532, 0.433]Regression, Linear
Comparison: Multivariable regression analyses (surgical cohort only)p-value: <0.000195% CI: [-3.259, -1.137]Regression, Linear
Secondary
Difference in Depression PROMIS Scores Reported by Participants Before and After eSyM Exposure
Difference in PROMIS depression scores between the pre-live and post-live eSyM cohorts. Since the two cohorts were comprised of different patients, comparisons were performed at the population level using mean T-scores. Depression was assessed using the PROMIS Emotional Distress-Depression 4-item short form with responses converted to T-scores (score range: 50 indicates the population mean with a standard deviation of 10; decrease in score indicates a positive change). All PROMIS items were scored using the standardized scoring tool kit. The differences in mean T-scores and the corresponding 95% confidence intervals (CI) were calculated for each PROMIS item, comparing the post-live versus pre-live cohort. When comparing group-level change in PROMIS scores, even trivial differences can be statistically significant if the sample size is large enough. The threshold to evaluate within-group change or to make a between-group comparison generally ranges between 2 and 6 T-score points.
Time frame: 30-90 days before and after eSyM go-live
Population: Since only a subset of eSyM patients completed the SASS questionnaire which included PROMIS measures, the analyzed population is smaller than other analytic populations. Distinct surveys were given to medical oncology and surgical patients to account for differences in their care, so the analytic cohorts below are distinct. Results for both the medical oncology and surgical subsets are included below.
| Arm | Measure | Group | Value (GEOMETRIC_MEAN) | Dispersion |
|---|---|---|---|---|
| Cluster Randomized Trial (Control Patients) | Difference in Depression PROMIS Scores Reported by Participants Before and After eSyM Exposure | Medical Oncology | 50.0 T-Score | Standard Deviation 9 |
| Cluster Randomized Trial (Control Patients) | Difference in Depression PROMIS Scores Reported by Participants Before and After eSyM Exposure | Surgery | 47.9 T-Score | Standard Deviation 8.5 |
| Cluster Randomized Trial (Intervention Patients) | Difference in Depression PROMIS Scores Reported by Participants Before and After eSyM Exposure | Medical Oncology | 48.7 T-Score | Standard Deviation 8.2 |
| Cluster Randomized Trial (Intervention Patients) | Difference in Depression PROMIS Scores Reported by Participants Before and After eSyM Exposure | Surgery | 46.0 T-Score | Standard Deviation 7.6 |
Comparison: Multivariable regression analyses (medical oncology cohort only)p-value: 0.0995% CI: [-1.985, 0.136]Regression, Linear
Comparison: Multivariable regression analyses (surgical cohort only)p-value: 0.00295% CI: [-2.474, -0.557]Regression, Linear
Secondary
Difference in Fatigue PROMIS Scores Reported by Participants Before and After eSyM Exposure
Difference in PROMIS fatigue scores between the pre-live and post-live eSyM cohorts. Since the two cohorts were comprised of different patients, comparisons were performed at the population level using mean T-scores. Fatigue was assessed using the PROMIS Fatigue 4-item short form with responses converted to T-scores (score range: 50 indicates the population mean with a standard deviation of 10; decrease in score indicates a positive change). All PROMIS items were scored using the standardized scoring tool kit. The differences in mean T-scores and the corresponding 95% confidence intervals (CI) were calculated for each PROMIS item, comparing the post-live versus pre-live cohort. When comparing group-level change in PROMIS scores, even trivial differences can be statistically significant if the sample size is large enough. The threshold to evaluate within-group change or to make a between-group comparison generally ranges between 2 and 6 T-score points (Terwee et al., 2021).
Time frame: 30-90 days before and after eSyM go-live
Population: Since only a subset of eSyM patients completed the SASS questionnaire which included PROMIS measures, the analyzed population is smaller than other analytic populations. Distinct surveys were given to medical oncology and surgical patients to account for differences in their care, so the analytic cohorts below are distinct. Results for both the medical oncology and surgical subsets are included below.
| Arm | Measure | Group | Value (GEOMETRIC_MEAN) | Dispersion |
|---|---|---|---|---|
| Cluster Randomized Trial (Control Patients) | Difference in Fatigue PROMIS Scores Reported by Participants Before and After eSyM Exposure | Medical Oncology | 57.4 T-Score | Standard Deviation 10.5 |
| Cluster Randomized Trial (Control Patients) | Difference in Fatigue PROMIS Scores Reported by Participants Before and After eSyM Exposure | Surgery | 51.1 T-Score | Standard Deviation 10.9 |
| Cluster Randomized Trial (Intervention Patients) | Difference in Fatigue PROMIS Scores Reported by Participants Before and After eSyM Exposure | Medical Oncology | 55.6 T-Score | Standard Deviation 9.8 |
| Cluster Randomized Trial (Intervention Patients) | Difference in Fatigue PROMIS Scores Reported by Participants Before and After eSyM Exposure | Surgery | 49.0 T-Score | Standard Deviation 11.2 |
Comparison: Multivariable regression analyses (medical oncology cohort only)p-value: 0.0495% CI: [-2.593, -0.062]Regression, Linear
Comparison: Multivariable regression analyses (surgical cohort only)p-value: 0.00495% CI: [-3.269, -0.646]Regression, Logistic
Secondary
Difference in Pain Interference PROMIS Scores Reported by Participants Before and After eSyM Exposure
Difference in PROMIS pain interference scores between the pre-live and post-live eSyM cohorts. Since the two cohorts were comprised of different patients, comparisons were performed at the population level using mean T-scores. Pain interference was assessed using the PROMIS Pain Interference 4-item short form with responses converted to T-scores (score range: 50 indicates the population mean with a standard deviation of 10; decrease in score indicates a positive change). All PROMIS items were scored using the standardized scoring tool kit. The differences in mean T-scores and the corresponding 95% confidence intervals (CI) were calculated for each PROMIS item, comparing the post-live versus pre-live cohort. When comparing group-level change in PROMIS scores, even trivial differences can be statistically significant if the sample size is large enough. The threshold to evaluate within-group change or to make a between-group comparison generally ranges between 2 and 6 T-score points.
Time frame: 30-90 days before and after eSyM go-live
Population: Since only a subset of eSyM patients completed the SASS questionnaire which included PROMIS measures, the analyzed population is smaller than other analytic populations. Distinct surveys were given to medical oncology and surgical patients to account for differences in their care, so the analytic cohorts below are distinct. Results for both the medical oncology and surgical subsets are included below.
| Arm | Measure | Group | Value (GEOMETRIC_MEAN) | Dispersion |
|---|---|---|---|---|
| Cluster Randomized Trial (Control Patients) | Difference in Pain Interference PROMIS Scores Reported by Participants Before and After eSyM Exposure | Medical Oncology | 54.0 T-Score | Standard Deviation 10.3 |
| Cluster Randomized Trial (Control Patients) | Difference in Pain Interference PROMIS Scores Reported by Participants Before and After eSyM Exposure | Surgery | 52.7 T-Score | Standard Deviation 9.8 |
| Cluster Randomized Trial (Intervention Patients) | Difference in Pain Interference PROMIS Scores Reported by Participants Before and After eSyM Exposure | Medical Oncology | 52.7 T-Score | Standard Deviation 9.7 |
| Cluster Randomized Trial (Intervention Patients) | Difference in Pain Interference PROMIS Scores Reported by Participants Before and After eSyM Exposure | Surgery | 51.8 T-Score | Standard Deviation 9.5 |
Comparison: Multivariable regression analyses (medical oncology cohort only)p-value: 0.295% CI: [-2.048, 0.431]Regression, Linear
Comparison: Multivariable regression analyses (surgical cohort only)p-value: 0.2795% CI: [-1.8, 0.496]Regression, Linear
Secondary
Difference in Physical Function PROMIS Scores Reported by Participants Before and After eSyM Exposure
Difference in PROMIS physical function scores between the pre-live and post-live eSyM cohorts. Since the two cohorts were comprised of different patients, comparisons were performed at the population level using mean T-scores. Physical function was assessed using the PROMIS Physical Function 4-item short form with responses converted to T-scores (score range: 50 indicates the population mean with a standard deviation of 10; decrease in score indicates a positive change). All PROMIS items were scored using the standardized scoring tool kit. The differences in mean T-scores and the corresponding 95% confidence intervals (CI) were calculated for each PROMIS item, comparing the post-live versus pre-live cohort. When comparing group-level change in PROMIS scores, even trivial differences can be statistically significant if the sample size is large enough. The threshold to evaluate within-group change or to make a between-group comparison generally ranges between 2 and 6 T-score points).
Time frame: 30-90 days before and after eSyM go-live
Population: Since only a subset of eSyM patients completed the SASS questionnaire which included PROMIS measures, the analyzed population is smaller than other analytic populations. Distinct surveys were given to medical oncology and surgical patients to account for differences in their care, so the analytic cohorts below are distinct. Results for both the medical oncology and surgical subsets are included below.
| Arm | Measure | Group | Value (GEOMETRIC_MEAN) | Dispersion |
|---|---|---|---|---|
| Cluster Randomized Trial (Control Patients) | Difference in Physical Function PROMIS Scores Reported by Participants Before and After eSyM Exposure | Medical Oncology | 42.3 T-Score | Standard Deviation 9.5 |
| Cluster Randomized Trial (Control Patients) | Difference in Physical Function PROMIS Scores Reported by Participants Before and After eSyM Exposure | Surgery | 47.9 T-Score | Standard Deviation 8.9 |
| Cluster Randomized Trial (Intervention Patients) | Difference in Physical Function PROMIS Scores Reported by Participants Before and After eSyM Exposure | Medical Oncology | 42.3 T-Score | Standard Deviation 9 |
| Cluster Randomized Trial (Intervention Patients) | Difference in Physical Function PROMIS Scores Reported by Participants Before and After eSyM Exposure | Surgery | 47.5 T-Score | Standard Deviation 9.4 |
Comparison: Multivariable regression analyses (medical oncology cohort only)p-value: 0.2295% CI: [-1.829, 0.429]Regression, Linear
Comparison: Multivariable regression analyses (surgical cohort only)p-value: 0.2895% CI: [-1.632, 0.472]Regression, Linear
Secondary
Difference in Self-Efficacy PROMIS Scores Reported by Participants Before and After eSyM Exposure
Difference in PROMIS self-efficacy scores between pre-live and post-live eSyM cohorts. Since the two cohorts were comprised of different patients, comparisons were performed at the population level using mean T-scores. Self-efficacy was assessed using the PROMIS Self-Efficacy for Managing Symptoms 8-item short form with responses converted to T-scores (score range: 50 indicates the population mean with a standard deviation of 10; decrease in score indicates a positive change). All PROMIS items were scored using the standardized scoring tool kit. The differences in mean T-scores and the corresponding 95% confidence intervals were calculated for each PROMIS item, comparing the post-live versus pre-live cohort. When comparing group-level change in PROMIS scores, even trivial differences can be statistically significant if the sample size is large enough. The threshold to evaluate within-group change or to make a between-group comparison generally ranges between 2 and 6 T-score points.
Time frame: 30-90 days before and after eSyM go-live
Population: Since only a subset of eSyM patients completed the SASS questionnaire which included PROMIS measures, the analyzed population is smaller than other analytic populations. Distinct surveys were given to medical oncology and surgical patients to account for differences in their care, so the analytic cohorts below are distinct. Results for both the medical oncology and surgical subsets are included below.
| Arm | Measure | Group | Value (GEOMETRIC_MEAN) | Dispersion |
|---|---|---|---|---|
| Cluster Randomized Trial (Control Patients) | Difference in Self-Efficacy PROMIS Scores Reported by Participants Before and After eSyM Exposure | Medical Oncology | 47.6 T-Score | Standard Deviation 8.4 |
| Cluster Randomized Trial (Control Patients) | Difference in Self-Efficacy PROMIS Scores Reported by Participants Before and After eSyM Exposure | Surgery | 50.1 T-Score | Standard Deviation 8.8 |
| Cluster Randomized Trial (Intervention Patients) | Difference in Self-Efficacy PROMIS Scores Reported by Participants Before and After eSyM Exposure | Medical Oncology | 47.3 T-Score | Standard Deviation 7.6 |
| Cluster Randomized Trial (Intervention Patients) | Difference in Self-Efficacy PROMIS Scores Reported by Participants Before and After eSyM Exposure | Surgery | 50.8 T-Score | Standard Deviation 8 |
Comparison: Multivariable regression analyses (medical oncology cohort only)p-value: 0.2395% CI: [-1.62, 0.386]Regression, Linear
Comparison: Multivariable regression analyses (surgical cohort only)p-value: 0.3995% CI: [-0.563, 1.431]Regression, Linear
Secondary
Emergency Department Treat/Release [EDTR] Event Occurrence Status at Day 30 (Comparing eSyM Intervention Responders Versus Non-Responders)
The secondary study outcome of the stepped wedge cluster RCT (randomized controlled trial) is the difference in the EDTR rate between eSyM responders and non-responders. This outcome will be defined in relation to the date of discharge from hospital (surgical cohort) or the initiation date of a new intravenous chemo regimen (medical oncology cohort). This is a binary outcome and we will analyze the absolute difference and odds ratio for the medical oncology and surgical cohorts combined and independently. NOTE: The total number of intervention participants in this analysis is greater than the number of intervention participants included in the primary analyses (outcomes 1-4). For outcomes outcomes 1-4, we used the patient's first event during the entire study period. To conduct this analysis of eSyM responders, we used the patient's first event during the intervention period. Approximately 10% of patients in this analysis were counted as a control event in the prior analyses.
Time frame: 30 days
Population: The primary analysis included data from both medical oncology and surgery. As a secondary analysis, we analyzed the data from MO and Surg separately and estimated the intervention effect for each outcome. Results for both combined and separate analyses are reported below.~NOTE: To conduct this analysis of eSyM responders, we used the patient's first event during the intervention period. Approximately 10% of patients in this analysis were counted as a control event in outcomes 1-4.
| Arm | Measure | Group | Value (NUMBER) |
|---|---|---|---|
| Cluster Randomized Trial (Control Patients) | Emergency Department Treat/Release [EDTR] Event Occurrence Status at Day 30 (Comparing eSyM Intervention Responders Versus Non-Responders) | Medical Oncology Cohort | 5.7 percentage of patients with event |
| Cluster Randomized Trial (Control Patients) | Emergency Department Treat/Release [EDTR] Event Occurrence Status at Day 30 (Comparing eSyM Intervention Responders Versus Non-Responders) | Surgery Cohort | 5.1 percentage of patients with event |
| Cluster Randomized Trial (Control Patients) | Emergency Department Treat/Release [EDTR] Event Occurrence Status at Day 30 (Comparing eSyM Intervention Responders Versus Non-Responders) | Med Onc + Surgery Combined Cohort | 5.3 percentage of patients with event |
| Cluster Randomized Trial (Intervention Patients) | Emergency Department Treat/Release [EDTR] Event Occurrence Status at Day 30 (Comparing eSyM Intervention Responders Versus Non-Responders) | Medical Oncology Cohort | 7.7 percentage of patients with event |
| Cluster Randomized Trial (Intervention Patients) | Emergency Department Treat/Release [EDTR] Event Occurrence Status at Day 30 (Comparing eSyM Intervention Responders Versus Non-Responders) | Surgery Cohort | 6.7 percentage of patients with event |
| Cluster Randomized Trial (Intervention Patients) | Emergency Department Treat/Release [EDTR] Event Occurrence Status at Day 30 (Comparing eSyM Intervention Responders Versus Non-Responders) | Med Onc + Surgery Combined Cohort | 7.1 percentage of patients with event |
Comparison: Adjusted analysis comparing intervention responders versus non-responders (chemo and surgical patients combined)p-value: <0.00195% CI: [0.776, 0.913]Regression, Logistic
Comparison: Adjusted analysis comparing intervention responders versus non-responders (medical oncology patients only)p-value: <0.00195% CI: [0.688, 0.885]Regression, Logistic
Comparison: Adjusted analysis comparing intervention responders versus non-responders (surgical patients only)p-value: 0.03295% CI: [0.799, 0.99]Regression, Logistic
Secondary
Emergency Department Treat/Release [EDTR] Event Occurrence Status at Day 90
This outcome will be defined in relation to the date of discharge from hospital (surgical cohort) or the initiation date of a new intravenous chemo regimen (medical oncology cohort). This is a binary outcome and we will analyze the absolute difference and odds ratio for the medical oncology and surgical cohorts combined and independently.
Time frame: 90 days
Population: The primary analysis included data from both medical oncology and surgery. As a secondary analysis, we analyzed the data from MO and Surg separately and estimated the intervention effect for each outcome. Results for both combined and separate analyses are reported below.
| Arm | Measure | Group | Value (NUMBER) |
|---|---|---|---|
| Cluster Randomized Trial (Control Patients) | Emergency Department Treat/Release [EDTR] Event Occurrence Status at Day 90 | Medical Oncology | 11.6 percentage of patients with event |
| Cluster Randomized Trial (Control Patients) | Emergency Department Treat/Release [EDTR] Event Occurrence Status at Day 90 | Surgery | 8.8 percentage of patients with event |
| Cluster Randomized Trial (Control Patients) | Emergency Department Treat/Release [EDTR] Event Occurrence Status at Day 90 | Medical Oncology + Surgery | 9.5 percentage of patients with event |
| Cluster Randomized Trial (Intervention Patients) | Emergency Department Treat/Release [EDTR] Event Occurrence Status at Day 90 | Medical Oncology | 13.8 percentage of patients with event |
| Cluster Randomized Trial (Intervention Patients) | Emergency Department Treat/Release [EDTR] Event Occurrence Status at Day 90 | Surgery | 9.7 percentage of patients with event |
| Cluster Randomized Trial (Intervention Patients) | Emergency Department Treat/Release [EDTR] Event Occurrence Status at Day 90 | Medical Oncology + Surgery | 11.2 percentage of patients with event |
Comparison: Adjusted analysis comparing control versus intervention patients (medical oncology and surgical patients combined)p-value: 0.0595% CI: [1, 1.272]Regression, Logistic
Comparison: Adjusted analysis comparing control versus intervention patients (medical oncology patients only)p-value: 0.90395% CI: [0.836, 1.225]Regression, Logistic
Comparison: Adjusted analysis comparing control versus intervention patients (surgical patients only)p-value: 0.01495% CI: [1.04, 1.418]Regression, Logistic
Secondary
Emergency Department Treat/Release [EDTR] Event Occurrence Status at Day 90 (Comparing eSyM Intervention Responders Versus Non-Responders)
This outcome will be defined in relation to the date of discharge from hospital (surgical cohort) or the initiation date of a new intravenous chemo regimen (medical oncology cohort). This is a binary outcome and we will analyze the absolute difference and odds ratio for the medical oncology and surgical cohorts combined and independently. NOTE: The total number of intervention participants in this analysis is greater than the number of intervention participants included in the primary analyses (outcomes 1-4). For outcomes outcomes 1-4, we used the patient's first event during the entire study period. To conduct this analysis of eSyM responders, we used the patient's first event during the intervention period. Approximately 10% of patients in this analysis were counted as a control event in the prior analyses.
Time frame: 90 days
Population: The primary analysis included data from both medical oncology and surgery. As a secondary analysis, we analyzed the data from MO and Surg separately and estimated the intervention effect for each outcome. Results for both combined and separate analyses are reported below.~NOTE: To conduct this analysis of eSyM responders, we used the patient's first event during the intervention period. Approximately 10% of patients in this analysis were counted as a control event in outcomes 1-4.
| Arm | Measure | Group | Value (NUMBER) |
|---|---|---|---|
| Cluster Randomized Trial (Control Patients) | Emergency Department Treat/Release [EDTR] Event Occurrence Status at Day 90 (Comparing eSyM Intervention Responders Versus Non-Responders) | Medical Oncology | 12.1 percentage of patients with event |
| Cluster Randomized Trial (Control Patients) | Emergency Department Treat/Release [EDTR] Event Occurrence Status at Day 90 (Comparing eSyM Intervention Responders Versus Non-Responders) | Surgery | 8.6 percentage of patients with event |
| Cluster Randomized Trial (Control Patients) | Emergency Department Treat/Release [EDTR] Event Occurrence Status at Day 90 (Comparing eSyM Intervention Responders Versus Non-Responders) | Medical Oncology + Surgery | 9.9 percentage of patients with event |
| Cluster Randomized Trial (Intervention Patients) | Emergency Department Treat/Release [EDTR] Event Occurrence Status at Day 90 (Comparing eSyM Intervention Responders Versus Non-Responders) | Medical Oncology | 15.1 percentage of patients with event |
| Cluster Randomized Trial (Intervention Patients) | Emergency Department Treat/Release [EDTR] Event Occurrence Status at Day 90 (Comparing eSyM Intervention Responders Versus Non-Responders) | Surgery | 11.2 percentage of patients with event |
| Cluster Randomized Trial (Intervention Patients) | Emergency Department Treat/Release [EDTR] Event Occurrence Status at Day 90 (Comparing eSyM Intervention Responders Versus Non-Responders) | Medical Oncology + Surgery | 12.9 percentage of patients with event |
Comparison: Adjusted analysis comparing intervention responders versus non-responders (medical oncology and surgical patients combined)p-value: <0.00195% CI: [0.839, 0.949]Regression, Logistic
Comparison: Adjusted analysis comparing intervention responders versus non-responders (medical oncology patients only)p-value: 0.00395% CI: [0.796, 0.954]Regression, Logistic
Comparison: Adjusted analysis comparing intervention responders versus non-responders (surgical patients only)p-value: 0.02795% CI: [0.835, 0.989]Regression, Logistic