This is an open-access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Research Protocols, is properly cited. The complete bibliographic information, a link to the original publication on http://www.researchprotocols.org, as well as this copyright and license information must be included.
Inflammatory bowel disease (IBD) is a chronic condition of the bowel that affects over 1 million people in the United States. The recurring nature of disease makes IBD patients ideal candidates for patient-engaged care that is centered on enhanced self-management and improved doctor-patient communication. In IBD, optimal approaches to management vary for patients with different phenotypes and extent of disease and past surgical history. Hence, a single quality metric cannot define a heterogeneous disease such as IBD, unlike hypertension and diabetes. A more comprehensive assessment may be provided by complementing traditional quality metrics with measures of the patient’s quality of life (QOL) through an application like HealthPROMISE.
The objective of this pragmatic randomized controlled trial is to determine the impact of the HealthPROMISE app in improving outcomes (quality of care [QOC], QOL, patient adherence, disease control, and resource utilization) as compared to a patient education app. Our hypothesis is that a patient-centric self-monitoring and collaborative decision support platform will lead to sustainable improvement in overall QOL for IBD patients.
Participants will be recruited during face-to-face visits and randomized to either an interventional (ie, HealthPROMISE) or control (ie, education app). Patients in the HealthPROMISE arm will be able to update their information and receive disease summary, quality metrics, and a graph showing the trend of QOL (SIBDQ) scores and resource utilization over time. Providers will use the data for collaborative decision making and quality improvement interventions at the point of care. Patients in the control arm will enter data at baseline, during office visits, and at the end of the study but will not receive any decision support (trend of QOL, alert, or dashboard views).
Enrollment in the trial will be starting in first quarter of 2015. It is intended that up to 300 patients with IBD will be recruited into the study (with 1:1 allocation ratio). The primary endpoint is number of quality indicators met in HealthPROMISE versus control arm. Secondary endpoints include decrease in number of emergency visits due to IBD, decrease in number of hospitalization due to IBD, change in generic QOL score from baseline, proportion of patients in each group who meet all eligible outpatient quality metrics, and proportion of patients in disease control in each group. In addition, we plan to conduct protocol analysis of intervention patients with adequate HealthPROMISE utilization (more than 6 log-ins with data entry from week 0 through week 52) achieving above mentioned primary and secondary endpoints.
HealthPROMISE is a unique cloud-based patient-reported outcome (PRO) and decision support tool that empowers both patients and providers. Patients track their QOL and symptoms, and providers can use the visual data in real time (integrated with electronic health records [EHRs]) to provide better care to their entire patient population. Using pragmatic trial design, we hope to show that IBD patients who participate in their own care and share in decision making have appreciably improved outcomes when compared to patients who do not.
ClinicalTrials.gov NCT02322307; https://clinicaltrials.gov/ct2/show/NCT02322307 (Archived by WebCite at http://www.webcitation.org/6W8PoYThr).
Inflammatory bowel disease (IBD) is a chronic condition of the bowel that affects over 1 million people in the United States [
While diseases such as hypertension and diabetes render themselves well to quality improvement efforts because of standardized indicators such as blood pressure and hemoglobin A1C respectively, a single quality-of-care (QOC) metric cannot define a heterogeneous disease such as IBD, where optimal approaches to manage patients differ between different phenotypes. Furthermore, IBD profoundly affects patients not only physically but also in social, professional, and emotional activities [
Chronic diseases affect almost 1 out of every 2 Americans and produce a significant burden on US health care [
HealthPROMISE [
Quality of life measure.
Short Inflammatory Bowel Disease Questionnaire.
A more comprehensive assessment may be provided by complementing the QOL with quality of care metrics (
Quality of care.
Quality improvement efforts so far have shown that measuring even limited QOC metrics carries a prohibitively high administrative and cost burden. The estimated costs from the Institute for Healthcare Improvement quality improvement initiative for either congestive heart failure or diabetes ranged from $81,000 to $148,000 per organization [
Currently, there is no well-accepted national model for quality improvement. Most of the quality improvement projects to date involve some kind of data abstraction from the clinical encounters that is fed into a registry to allow benchmarking, risk adjustment, and quality reporting. This cycle takes anywhere from a few weeks to a few months and happens long after the patient has left the health care facility. Patients are not involved in measuring or improving quality. Thus, an important patient-physician “productive interaction” opportunity to improve outcomes at the point of care is missed [
The aim of this research protocol is to evaluate the patient-centric Web- and mobile-based application, HealthPROMISE, where IBD patients longitudinally measure their QOC and QOL metrics and physicians use this information for collaborative decision making and improving patient outcomes. Our hypothesis is that a patient-centric self-monitoring and collaborative decision support platform will lead to sustainable improvement in overall QOL for IBD patients.
This is a phase III, single-center, pragmatic randomized controlled trial (RCT) to evaluate if a patient-centric self-monitoring and collaborative decision support platform will lead to sustainable improvement in overall QOL for IBD patients. It is intended that 300 patients with IBD will be recruited into the study (allocation ratio 1:1;
Point of care recruitment and randomization.
Shows initial mock-up of app.
Patients will be recruited at outpatient and inpatient facilities in an academic center through informational paper and electronic flyers. Once enrolled, patients will receive a walkthrough of the app, which includes access to a training video. The provider dashboard will also have access to the training video. Eligible patients will be 18 years or older, have a mobile phone or access to the Internet at home, and be able to complete a Web-based questionnaire in English. Exclusion criteria include the inability to communicate with the investigators and comply with the study requirements, presence of short bowel syndrome or stoma, and presence of a condition or disease that, in the opinion of the investigators, may make it difficult for the patient to use the HealthPROMISE app, including, but not limited to, advanced dementia.
A combination of different questionnaires (eg, SIBDQ), symptom updates, and quality indicators relevant for evaluating patient status will be the data collected during this study through the HealthPROMISE app (
Shows completed app.
The Short Inflammatory Bowel Disease Questionnaire (SIBDQ) [
EQ-5D is a standardized instrument for measuring generic QOL [
eHEALS is an 8-item measure of eHealth literacy developed to measure consumers’ combined knowledge, comfort, and perceived skills at finding, evaluating, and applying electronic health information to health problems [
Quality indicators are included from a list of indicators published by national societies and finalized through a Delphi panel of IBD providers [
Utilization will be assessed through log-in, page views, health information updates, and response to alerts and reminders.
Provider workflow on app dashboard.
All outcome data (
Primary endpoint
Number of quality indicators met in HealthPROMISE versus control arm
Secondary endpoints
Decrease in number of emergency visits due to IBD
Decrease in number of hospitalization due to IBD
Change in generic QOL score (EQ-5D) from baseline
Proportion of patients in each group who meet all eligible outpatient quality metrics
Proportion of patients in disease control in each group
Emergency visits in each group
Hospitalizations in each group
General QOL scores in each group
Per protocol analysis of intervention patients with adequate HealthPROMISE use (more than 6 log-ins with data entry from week 0 through week 52) achieving above mentioned primary and secondary endpoints
Evaluation metrics for HealthPROMISE progress
Metric | Target Group | Goal / Timeline | |
|
|||
Number of providers trained | Investigators | 10 in 2 months | |
|
Number of patients enrolled | Participants | 300 in 6 months |
|
Recruitment and training of key personnel | Coordinator | 1 in 3 months |
|
Patient utilization of HealthPROMISE | Participants | Ongoing |
|
Provider utilization of HealthPROMISE | Providers | Ongoing |
|
Response to alert within 2 business days | Providers | >90% |
|
|||
|
Improvement in quality of care metrics | Provider, Center | Quarterly reports |
|
Improvement in quality of life | Provider, Center | Quarterly reports |
|
Readmission rate in two arms | Provider, Center | Quarterly reports |
We will use SAS 9.2 (SAS Institute, Inc) to calculate frequencies and percentages for categorical factors and means with standard deviations and/or percentiles for continuous factors. Pearson’s chi-square tests will be used for primary outcome (number of quality indicators met in HealthPROMISE vs control arm) and secondary outcomes. We will calculate percentage score for each patient at baseline and at week 104 ([number of quality metrics met/quality metrics eligible]*100). Change in the percentage score from baseline to week 104 will be aggregated for each arm to calculate percentage-point improvement in quality metric, similar to the strategy by Cebul et al [
Analysis of covariance (ANCOVA) will be performed to assess differences in the area under the curve of QOL scores while adjusting for baseline QOL score. To assess the association between patient and practice characteristics and achievement of eligible quality metrics, we will use multivariable analyses. Since the data will be hierarchically structured, with patients clustered within physicians and metrics clustered within patients, we will construct multilevel, generalized, linear mixed models with random effects to determine predictors of quality care, similar to the strategy used by Kanwal et al [
Interim analysis will be performed once 150 patients are followed up for week 52. If primary outcome is met by that time, all patients will be offered HealthPROMISE app and followed for the additional 52 weeks.
Study endpoints will be primarily assessed using intention-to-treat (ITT) analysis; however, per-protocol analysis will also be performed. The study is to be powered such that there is a>80% probability of demonstrating a difference with a
We assume that 128 out of 150 subjects (85%) in the intervention arm will meet all quality indicators (primary outcome) and expect that this percentage will be at least 15% lower in the control arm. A sample size of 95 patients will be needed in each arm to achieve at least 80% power to detect the difference with a 5% one-sided significance level.
Accounting for an estimated 30% attrition rate, we will require a total of 250 IBD patients to be enrolled in the study. Since some patients may agree to enroll but not download the app or use the PIN, we will recruit a total of 300 patients in the study.
For secondary outcomes related to QOL, the control arm is not a placebo arm and physicians are free to initiate any therapy based on patients' symptoms. Hence, we will assess the difference in proportion of patients achieving MCID in the HealthPROMISE arm versus control arm in the study. Using the distribution-based approach, an effect size of 0.5 SD is the closest estimate for determining MCID for SIBDQ and EQ-5D. Assuming that 20% more patients in HealthPROMISE arm will achieve MCID than in the control arm, the sample size of 250 patients will have 88% power to detect the difference with a 5% one-sided significance level and an estimated 30% attrition rate.
This pragmatic trial will help us study if a patient-centric self-monitoring and collaborative disease management app and dashboard can lead to improvement in care provided to IBD patients. Our hypothesis is that IBD patients using the HealthPROMISE platform will have significant improvement in QOC metrics, QOL, and resource utilization by the end of the 2-year study period when compared to IBD patients in the control arm (using a health education app alone).
HealthPROMISE can be a sustainable platform in the long run because it is patient-centric, device and disease agnostic, and not dependent on proprietary EHRs. As most of the data is entered by patients, the cost of running, supporting, and sustaining HealthPROMISE is very low compared to traditional disease registries. HealthPROMISE has a rapid form generator capability to allow it to be customized for other chronic diseases. Additionally, the decision support that generates alerts and dashboard reports is within the stand-alone HealthPROMISE app and not dependent on proprietary EHRs. We aim to integrate HealthPROMISE with personal health records, partner with national societies, and support through consortia so it can become a new standard of quality care for IBD and other chronic diseases.
HealthPROMISE is a unique cloud-based PRO and decision support tool that empowers both patients and providers. Patients track their QOL and symptoms, and providers can use the visual data in real time (integrated with EHRs) to provide better care to their entire patient population. Using pragmatic trial design, we hope to show that IBD patients who participate in their own care and share in decision making have appreciably improved outcomes when compared to patients who do not [
Initial critique of the proposal before it got funded in second attempt.
Informed consent form.
CONSORT-EHEALTH checklist V1.6.2 [
inflammatory bowel disease
quality of life
quality of care
patient-reported outcome
electronic health records
National Institutes of Health
randomized controlled trial
Short Inflammatory Bowel Disease Questionnaire
minimal clinically important difference
patient activation measure
The study is supported by the Crohn's & Colitis Foundation of America (grant #253624) and the National Institutes of Health (5 K23 DK97451-02) with Ashish Atreja as the principal investigator.
The app was developed in-house at Sinai AppLab.