Presentation
Engaging Patients with Health Information Technology (IT) for Resilient Cancer Care
Event Type
Oral Presentations
TimeFriday, April 1612:30pm - 12:50pm EDT
LocationDigital Health
DescriptionCancer patients are at a high risk for adverse events such as medical errors and unplanned hospitalization due to the complexity and toxicity of their treatment. Additionally, cancer care is increasingly being treated in ambulatory settings where cancer patients are vulnerable to undetected clinical deterioration. A recent study found that 22% of re-admissions in cancer patients were preventable [1]. Identifying patients at risk for clinical deterioration remains a challenge as clinicians are often not aware of patient complications between visits, and are therefore unable to intervene in a timely manner. Consequently, the patient and their caregiver(s) are responsible for monitoring and detecting clinical deterioration and communicating the problem with their clinical team before an acute event occurs. To proactively prevent clinical deterioration and improve resilience in cancer care, patients and their caregivers need to be engaged in their care and empowered to speak up when problems occur.
With the growing use of health information technology (IT), including patient-facing health IT, there are opportunities to leverage these technologies to identify clinical deterioration before an adverse event occurs. However, there remain few reliable systems to support outpatient cancer patients. The objective of this study is to describe a multi-modal approach to engage patients in their outpatient cancer care utilizing health IT. This study was conducted as a part of a larger project aimed at preventing unexpected clinical deterioration in cancer patients using active (self-reported) and passive (wearable technology) data to generate a predictive model of clinical deterioration. A key aspect of the predictive model is based on non-routine events (NREs), defined as any aspect of clinical care deviating from optimal or expected care for that specific patient [2, 3]. This study describes how health IT applications can be used by patients to support and improve the resilience of their cancer care.
Methods
We identified eligible cancer outpatients via chart review and care team recommendations. We approached eligible patients at the start of their treatment and invited them to participate in the study. If a patient agreed to participate, we collected data from the patient using a multi-modal approach leveraging several types of health IT: MyCap, FitBit, and Geolocation data. Patients’ participation in this longitudinal study lasted between 6-13 weeks.
MyCap:
We developed a secure patient-facing mobile app for participant data collection through MyCap™, REDCap’s mobile app deployment toolkit. Participants used MyCap to complete weekly validated patient reported outcome measures (PROMs) including the Distress Thermometer [4], symptom list, and abbreviated versions of global health status [5] and care experience metrics [6]. Patients also used MyCap to log any NREs that occurred. These data were stored in REDCap™ and exported into Microsoft Excel for analysis.
FitBit:
Enrolled patients were given a FitBit to monitor their health. The FitBit captures real-time personal data including sleep, activity level (e.g., steps), and heart rate. Patients can view these data through the associated application on their personal (or study-provided) smartphone. In addition, these data were downloaded to the research database weekly for analysis.
Geolocation data:
We monitored patient geolocation data using a pre-determined list of common places each patient visits when they are feeling well (e.g., grocery store, church) or not so well (e.g., pharmacy, emergency room). The geolocation data were collected from patients’ smartphones using Google Maps.
Results
To date, we have enrolled 26 cancer patients – 22 with head and neck, 5 pancreatic, and 1 gastrointestinal cancer. These patients are 83% male and the majority have a high school diploma or GED. Patient comfort with technology varied: 13% stated they are the first among their peers to adopt new technology, 35% stated they wait to see how the new technology works for others before adopting, 30% stated they wait until the technology is well established before adopting, and 22% stated they reluctantly adopt new technology. Despite the varying levels of comfort with and adoption of technology, we successfully engaged every patient to use the multiple health IT platforms throughout their study duration.
MyCap:
Participants have completed a total of 186 MyCap reports, including 139 weekly PROMS reports and 47 reports of NREs. In the weekly PROMs reports, patients provided detailed insights into their health and response to cancer treatment. Out of 139 reports, the average distress level in patients was 4.1 (SD: 2.8; range: 0-10). Patients reported facing several practical problems including making treatment decisions (25/139) and dealing with financial issues (19/139). Patients described their emotional problems including depression (18%), fear (19%), nervousness (16%), sadness (13%), worry (28%), and a loss of interest in activities (13%). Patients reported physical problems encountered, the most common being fatigue (39%). We gathered data from patients about their experience with their care team. Overall, patients were positive when describing their care team, with 79% stating that their care team ‘always’ listened carefully to them. Patients also reported that their care team explained things in a way that was easy to understand.
Enrolled patients completed a total of 47 NRE reports. Out of these 47 NREs, 57% occurred at home, 17% at a hospital, 4% in the emergency department, 2% in a clinic, and the remaining in other, unspecified locations. Only half (52%) of the patients experiencing an adverse event reported the event to a clinical staff member. The NREs had varying effects on the patient’s health ranging from no impact (33%) to serious impact (17%). An example of the latter was a blockage in the patient’s feeding tube resulting in a visit to the emergency department. Serious impact NREs were commonly related to severe pain, substantial loss of time at work/school, and/or the inability to accomplish activities of daily living.
FitBit:
All 26 enrolled patients used the FitBits and reported that they liked using the device to track their weight, diet, and sleep to assess how well they were doing throughout their treatment. We encountered some challenges using the FitBits. For instance, we had missing data when patients forgot to charge or wear their devices. At least weekly check-ins by study personnel with each participant minimized data loss.
Geolocation:
We collected geolocation data from 22 of the 26 enrolled participants (4 declined due to privacy concerns). We faced a couple of challenges collecting the geolocation data from patients. When a patient’s phone had a Google security update, this would interrupt data collection until the next interaction with the research staff, who could re-configure their phone. The COVID-19 pandemic changed the nature and potential value of the geolocation data for this study. During project conception, we had hypothesized that patient visits to some specific locations (e.g., church, son’s house) would indicate maintenance of their pre-cancer treatment state of health while unscheduled visits to medical locations would suggest worsening health. However, with the onset of COVID-19, we have seen substantially less patient travel outside of their homes. These pandemic effects appear to have diminished the value/sensitivity of geolocation data to predict clinical deterioration in our patient population.
In this study, we described how we leveraged health IT to better understand patient experiences throughout their cancer treatment. We successfully engaged patients through various health IT platforms. We described some challenges to deploying patient-facing health IT, especially during a pandemic.
Conclusion
Cancer patients are at high risk of undetected clinical deterioration in the outpatient setting. This study described 3 digital health platforms used to successfully engage patients in their cancer care. The data gathered through the various digital health platforms will be used to develop a clinical risk algorithm to proactively monitor, detect, and respond to unexpected clinical deterioration in outpatient cancer patients.
With the growing use of health information technology (IT), including patient-facing health IT, there are opportunities to leverage these technologies to identify clinical deterioration before an adverse event occurs. However, there remain few reliable systems to support outpatient cancer patients. The objective of this study is to describe a multi-modal approach to engage patients in their outpatient cancer care utilizing health IT. This study was conducted as a part of a larger project aimed at preventing unexpected clinical deterioration in cancer patients using active (self-reported) and passive (wearable technology) data to generate a predictive model of clinical deterioration. A key aspect of the predictive model is based on non-routine events (NREs), defined as any aspect of clinical care deviating from optimal or expected care for that specific patient [2, 3]. This study describes how health IT applications can be used by patients to support and improve the resilience of their cancer care.
Methods
We identified eligible cancer outpatients via chart review and care team recommendations. We approached eligible patients at the start of their treatment and invited them to participate in the study. If a patient agreed to participate, we collected data from the patient using a multi-modal approach leveraging several types of health IT: MyCap, FitBit, and Geolocation data. Patients’ participation in this longitudinal study lasted between 6-13 weeks.
MyCap:
We developed a secure patient-facing mobile app for participant data collection through MyCap™, REDCap’s mobile app deployment toolkit. Participants used MyCap to complete weekly validated patient reported outcome measures (PROMs) including the Distress Thermometer [4], symptom list, and abbreviated versions of global health status [5] and care experience metrics [6]. Patients also used MyCap to log any NREs that occurred. These data were stored in REDCap™ and exported into Microsoft Excel for analysis.
FitBit:
Enrolled patients were given a FitBit to monitor their health. The FitBit captures real-time personal data including sleep, activity level (e.g., steps), and heart rate. Patients can view these data through the associated application on their personal (or study-provided) smartphone. In addition, these data were downloaded to the research database weekly for analysis.
Geolocation data:
We monitored patient geolocation data using a pre-determined list of common places each patient visits when they are feeling well (e.g., grocery store, church) or not so well (e.g., pharmacy, emergency room). The geolocation data were collected from patients’ smartphones using Google Maps.
Results
To date, we have enrolled 26 cancer patients – 22 with head and neck, 5 pancreatic, and 1 gastrointestinal cancer. These patients are 83% male and the majority have a high school diploma or GED. Patient comfort with technology varied: 13% stated they are the first among their peers to adopt new technology, 35% stated they wait to see how the new technology works for others before adopting, 30% stated they wait until the technology is well established before adopting, and 22% stated they reluctantly adopt new technology. Despite the varying levels of comfort with and adoption of technology, we successfully engaged every patient to use the multiple health IT platforms throughout their study duration.
MyCap:
Participants have completed a total of 186 MyCap reports, including 139 weekly PROMS reports and 47 reports of NREs. In the weekly PROMs reports, patients provided detailed insights into their health and response to cancer treatment. Out of 139 reports, the average distress level in patients was 4.1 (SD: 2.8; range: 0-10). Patients reported facing several practical problems including making treatment decisions (25/139) and dealing with financial issues (19/139). Patients described their emotional problems including depression (18%), fear (19%), nervousness (16%), sadness (13%), worry (28%), and a loss of interest in activities (13%). Patients reported physical problems encountered, the most common being fatigue (39%). We gathered data from patients about their experience with their care team. Overall, patients were positive when describing their care team, with 79% stating that their care team ‘always’ listened carefully to them. Patients also reported that their care team explained things in a way that was easy to understand.
Enrolled patients completed a total of 47 NRE reports. Out of these 47 NREs, 57% occurred at home, 17% at a hospital, 4% in the emergency department, 2% in a clinic, and the remaining in other, unspecified locations. Only half (52%) of the patients experiencing an adverse event reported the event to a clinical staff member. The NREs had varying effects on the patient’s health ranging from no impact (33%) to serious impact (17%). An example of the latter was a blockage in the patient’s feeding tube resulting in a visit to the emergency department. Serious impact NREs were commonly related to severe pain, substantial loss of time at work/school, and/or the inability to accomplish activities of daily living.
FitBit:
All 26 enrolled patients used the FitBits and reported that they liked using the device to track their weight, diet, and sleep to assess how well they were doing throughout their treatment. We encountered some challenges using the FitBits. For instance, we had missing data when patients forgot to charge or wear their devices. At least weekly check-ins by study personnel with each participant minimized data loss.
Geolocation:
We collected geolocation data from 22 of the 26 enrolled participants (4 declined due to privacy concerns). We faced a couple of challenges collecting the geolocation data from patients. When a patient’s phone had a Google security update, this would interrupt data collection until the next interaction with the research staff, who could re-configure their phone. The COVID-19 pandemic changed the nature and potential value of the geolocation data for this study. During project conception, we had hypothesized that patient visits to some specific locations (e.g., church, son’s house) would indicate maintenance of their pre-cancer treatment state of health while unscheduled visits to medical locations would suggest worsening health. However, with the onset of COVID-19, we have seen substantially less patient travel outside of their homes. These pandemic effects appear to have diminished the value/sensitivity of geolocation data to predict clinical deterioration in our patient population.
In this study, we described how we leveraged health IT to better understand patient experiences throughout their cancer treatment. We successfully engaged patients through various health IT platforms. We described some challenges to deploying patient-facing health IT, especially during a pandemic.
Conclusion
Cancer patients are at high risk of undetected clinical deterioration in the outpatient setting. This study described 3 digital health platforms used to successfully engage patients in their cancer care. The data gathered through the various digital health platforms will be used to develop a clinical risk algorithm to proactively monitor, detect, and respond to unexpected clinical deterioration in outpatient cancer patients.
Authors
