Current healthcare systems do not effectively promote weight reduction in patients with obesity and gastroesophageal reflux disease (GERD). The Reflux Improvement and Monitoring (TRIM) program provides personalized, multidisciplinary, health education and monitoring over 6 months. In this study we aimed to (i) measure the effectiveness of TRIM on GERD symptoms, quality of life, and weight, and (ii) examine patient health beliefs related to TRIM.
This prospective mixed methods feasibility study was performed at a single center between September 2015 and February 2017, and included adult patients with GERD and a body mass index ≥30 kg/m2. Quantitative analysis consisted of a pre- to post-intervention analysis of TRIM participants (+TRIM Cohort) and a multivariable longitudinal mixed model analysis of +TRIM vs. patients who declined TRIM (−TRIM Cohort). Primary outcomes were change in patient-reported GERD symptom severity (GerdQ) and quality of life (GerdQ-DI), and change in percent excess body weight (%EBW). Qualitative analysis was based on two focus groups of TRIM participants.
Among the +TRIM cohort (n=52), mean baseline GerdQ scores (8.7±2.9) decreased at 3 months (7.5±2.2; P<0.01) and 6 months (7.4±1.9; P=0.02). Mean GerdQ-DI scores decreased, but did not reach statistical significance. Compared with the −TRIM cohort (n=89), reduction in %EBW was significantly greater at 3, 6, and 12 months among the +TRIM cohort (n=52). In qualitative analysis, patients unanimously appreciated the multidisciplinary approach and utilized weight loss effectively to improve GERD symptoms.
In this mixed methods feasibility study, participation in TRIM was associated with symptom improvement, weight reduction, and patient engagement.
Obesity is an important independent risk factor for gastroesophageal reflux disease (GERD) (1, 2). In observational studies and systematic reviews, weight loss is reported as one of the few effective lifestyle modifications for GERD (1, 3, 4). The prevalence of obesity and GERD is rising, and these conditions contribute significant morbidity, resource utilization, and healthcare costs (5). Consequently, appropriate recognition and management of obesity is a core Centers for Medicare and Medicaid Services quality measure and US Preventive Services Task Force recommendation (6, 7). In addition, documentation of adequate weight management counseling for overweight patients is a valid quality measure in the initial evaluation of patients with GERD (8).
Although effective weight management in the care of GERD is a national priority, implementing clinically feasible and effective interventions is challenging (9). A multitude of factors influence the effectiveness of weight loss interventions. Health system barriers include provider time constraints, insufficient reimbursement, and inadequate provider training (9, 10). In addition, patient engagement is influenced by choices, attitudes, and beliefs; as such, effective interventions need to focus on changeable behaviors and objectives (9, 10, 11). Recognizing these factors, we designed a weight management program tailored to patients with obesity and GERD, known as The Reflux Improvement and Monitoring (TRIM) program. TRIM integrates multidisciplinary care teams and information systems to provide patient-centered weight management education over 6 months (12).
In this mixed methods feasibility study, we aimed to (i) measure the effectiveness of TRIM on symptoms, health-related quality of life (QOL), and weight, and (ii) examine patient health beliefs related to TRIM. We hypothesized that participation in TRIM is associated with reduced symptom severity, improved QOL, and weight loss.
Study design and setting
We performed a prospective mixed methods feasibility study at a single academic medical center over 18 months (1 September 2015 to 30 February 2017). The sequential explanatory mixed methods approach consisted of a quantitative pre- to post-intervention analysis and cohort comparison, and a qualitative analysis of patient focus groups to assess patient perspectives related to the intervention. The Northwestern University institutional review board approved the quantitative study and waived informed consent (STU no. :60007543), and deemed the focus group component as institutional review board exempt.
We enrolled eligible patients from 1 September 2015 to 30 November 2016. Electronic health record (EHR) decision support was used to identify adult patients (≥18 years old) with a diagnosis of GERD (ICD9 530.81, ICD10 K21.9) and a body mass index (BMI) ≥30 kg/m2. During general internal medicine or gastroenterology clinic visits, the EHR prompted providers to accept an autogenerated referral to TRIM. Referred patients received standardized instructions to initiate a consultation with a certified health educator. If a referred patient did not contact the educators within 1 week, the educators attempted to contact the referred patient by phone on three separate occasions. Eligible patients were separated into two cohorts: intervention cohort (+TRIM)—those who agreed to participate at the initial educator consultation—and comparator cohort (−TRIM)—those who declined enrollment. We followed all patients for a minimum of 3 months and up to 12 months from the time of referral.
The goal of TRIM is to provide personalized health education and goal setting to patients with obesity and GERD. The structure and educational content of TRIM was designed by a multidisciplinary team of gastroenterologists, a nutritionist, certified health educators, and health informatics analysts. TRIM is a 6-month program consisting of an initial consultation, followed by three weekly follow-up sessions and then five monthly sessions with an educator. Initial and follow-up sessions are conducted in person or over the phone according to patient preference. Initial consultation involves baseline weigh-in, standardized education, personalized goal-setting, weekly meal planning, and instructions on use of electronic and/or written nutrition tracking tools. The educator emphasizes the causal relationship between obesity and GERD symptoms. During follow-up sessions, the educators review change in weight, dietary, and exercise goals, reinforce education, and provide resources as needed. At each session the educators conduct the GerdQ questionnaire and record patient-reported weights and GerdQ responses in templates embedded in the EHR (Figure 1).
GERD symptom severity was measured by the GerdQ, a validated six-item patient-reported GERD instrument that calculates symptom severity over the preceding week based on response to four positive predictors (heartburn, regurgitation, sleep disturbance, and antacid use) and two negative predictors (epigastric pain and nausea). Scores range from 0 to 18; a higher GerdQ indicates greater symptom severity and 8 is the proposed cutoff. The GerdQ-DI, a subscore of the GerdQ based on response to sleep disturbance and antacid medication, additionally measures QOL impairment (Supplementary Table S1 online) (13).
Our a priori goal was to analyze all data available 18 months after the beginning of the study; we did not prespecify a sample size or impute missing values.
The primary outcomes were change in GerdQ score, GerdQ-DI subscore, and percent excess body weight (%EBW).
Baseline clinical data and longitudinal weight data for both +TRIM and −TRIM cohorts and GerdQ data for +TRIM cohort were extracted from the institutional enterprise data warehouse. Weight data were derived from either patient-reported weights through the TRIM program or vital sign charted data.
Primary analysis: pre- to post-intervention comparison within intervention (+TRIM) cohort
The primary analysis used two-tailed paired t-test to examine differences in total GerdQ, GerdQ-DI, and %EBW among the TRIM group at baseline vs. 3- and 6-month intervals following enrollment.
Secondary analysis: comparison of intervention (+TRIM) and no intervention (−TRIM) cohorts
As this study was not designed as a controlled trial, we conducted a secondary analysis using a longitudinal multivariable mixed model to compare change in %EBW between the +TRIM and −TRIM cohorts up to 12 months from time of referral. First, bivariate analyses of baseline clinical data between the two cohorts were compared via two-tailed t-test and χ2 analysis as appropriate. We a priori determined covariates of age, sex, race, baseline %EBW, and referring clinician type. Variables significant at the P<0.05 level were included in the multivariable longitudinal mixed model. Longitudinal data were clustered by participant (1-X observations/participant), measured in units of patient-month, and analyzed at 3-, 6-, and 12-month intervals.
We post hoc performed a second multivariable longitudinal mixed model adjusting for all covariates regardless of significance in bivariate analysis. Longitudinal data were clustered by participant (1-X observations/participant), measured in units of patient-month, and analyzed at 3 and 6 months. Outcomes were only measured up to 6 months because of missing covariate data and loss of observations past 6 months.
Given the variability in data collection of weights for the +TRIM cohort, we a priori planned a sensitivity analysis using Spearman’s correlation to examine the correlation between patient-reported weights and vital sign documented weights for weights reported within a 24 h range. This analysis was limited to the setting where patient-reported weights were collected before vital sign documented weights to reduce reporting bias.
The P values of <0.05 were considered statistically significant. Statistical analysis for the quantitative analysis was conducted using STATA 14.2 (College Station, TX) and SAS 9.4 (Cary, NC).
Focus group design and participants
The qualitative analysis explored the health beliefs surrounding symptoms of GERD and weight loss among TRIM participants. The study team formulated key informant interview guides based on the Health Belief Model to explore perceptions, knowledge, self-efficacy, and attitudes related to TRIM, with a particular focus on understanding if multidisciplinary care, education, and monitoring engages and promotes weight loss for patients with GERD (14). Two focus groups were held in September 2016 in Chicago, IL. Each focus group was conducted on a weeknight and lasted for 2 h. The same participants were present in each focus group. One moderator with a background in gastrointestinal behavior and psychology led the focus groups and two observers took field notes during the focus groups. Despite institutional review board exemption, we obtained written consent to audio record the focus groups.
We recruited focus group participants by calling a panel of TRIM participants. The sampling strategy aimed to include an equal division of sex, age, and race. We then scheduled focus groups with 6 to 10 participants.
Verbatim transcriptions of the audio recordings were coded using MAXQDA software program (Verbi GmbH, Marburg, Germany). Analysis was conducted using Thematic Analysis and Grounded Theory. First, two researchers reviewed the transcripts, writing reflexive and analytic memos to identify themes and concepts. Subsequently, the researchers developed a codebook based on the research questions, theory constructs, and reoccurring themes and concepts. Both researchers independently coded the transcripts, after which they discussed any discrepancies until they reached a consensus. Finally, the researchers systematically reviewed each code and provided a detailed description of its meaning, context, and range of agreement and disagreement.
Primary quantitative analysis: pre- to post-intervention comparison
Subjects and baseline characteristics
Of the 123 patients enrolled in TRIM, 52 (42%) patients had at least 3 months of follow-up data available at the time of study analysis and were included in the +TRIM cohort. The mean age was 54.4±12.8 years and 40/52 (77%) were female. Baseline weight measurements were: mean weight 101.8±19.1 kg, median BMI 34.3 kg/m2 (interquartile range 32.5–40.5), and mean %EBW 41.8±10.1. At enrollment, all patients were taking single- or double-dose proton pump inhibitor (PPI).
Baseline mean GerdQ (8.7±2.9) significantly decreased at 3 months (7.5±2.2, P<0.01), and 6 months (7.4±1.9, P=0.02). Baseline mean GerdQ-DI (1.5±1.7) significantly decreased at 3 months (0.9±1.5; P=0.02) and was lower at 6 months (0.8±1.3; P=0.38) (Figure 2).
Compared with mean %EBW at baseline (41.8±10.1), mean %EBW decreased at 3 months (39.4±10.6, P=0.25) and 6 months (36.3±10.6, P=0.04).
Secondary quantitative analysis: cohort comparison
The cohort comparison included 52 +TRIM participants and 89 −TRIM patients. Baseline variables were balanced between the two groups (Table 1).
In the primary longitudinal mixed model, all covariates fell out of the model. Compared with the −TRIM cohort, the +TRIM cohort had a significantly greater reduction in %EBW at 3 months (−1.1 vs. −2.0 points; β −0.95, 95% confidence interval (CI) −1.9 to −0.1, P=0.04), 6 months (−1.0 vs. −3.3 points; β −2.4, 95% CI −3.5 to −1.2, P<0.001), and 12 months (+0.1 vs. −3.7 points, β −3.8, 95% CI −6.5 to −1.2, P<0.01) (Figure 3). With regard to BMI, among +TRIM participants, mean BMI decreased from 39.6±8.9 kg/m2 at 3 months to 39.2 kg/m2 ±10.6 kg/m2 at 6 months and to 38.6±10.3 kg/m2 at 12 months. Among −TRIM participants BMI was 37.5±6.4 kg/m2 at 3 months, 35.8±6.3 kg/m2 at 6 months, and 38.5±6.2 kg/m2 at 12 months. The prevalence of obesity (BMI >30 kg/m2) among +TRIM participants was 97% at baseline, 95% at 3 months, 89% at 6 months, and 86% at 12 months, whereas among No TRIM participants was 99% at baseline, 98% at 3 months, 94% at 6 months, and 100% at 12 months. The remainder of the groups were overweight (BMI 25–29.9 kg/m2).
Being in the +TRIM cohort, as opposed to −TRIM, was associated with an additional weight reduction of 4.0 kg at 6 months (P<0.001) and 5.7 kg at 12 months (P=0.02).
After adjusting for all prespecified covariates in the post hoc multivariable longitudinal mixed model, the +TRIM cohort had a greater reduction in %EBW compared with the −TRIM cohort at 3 months (P=0.05) and 6 months (P<0.01) (Table 2).
A total of 68 weights met inclusion criteria for the sensitivity analysis. The median difference between the weights was 0.19 kg (interquartile range −0.09 to 0.63). There was significantly high correlation between patient-reported weights and vital sign charted weights (Spearman’s rho 0.99, P<0.0001). In addition, the correlation between change in excess body weight and GerdQ score at 6 months was 0.41 (P=0.07).
Qualitative focus group analysis
Nine patients from the TRIM program attended each of the two focus groups (22% men, median age 60 years). Themes and sample quotes from the qualitative analysis are depicted in Supplementary Table S2.
In summary, GERD was perceived as a serious, frightening, and painful disease that negatively affected QOL at home and work. Many felt that because of a lack of empathy, clinicians often failed to treat their GERD seriously or provide helpful recommendations. Although all participants wished to reduce PPI use, barriers included the fear of pain returning and the difficulty in committing to exercise and diet as an alternative approach.
With regard to the TRIM program, the group unanimously expressed positive feelings for their multidisciplinary TRIM team, as the team provided cues to healthy behavioral actions. All participants identified TRIM as a beneficial combined approach to manage GERD symptoms and lose weight.
This mixed methods feasibility study consisted of a pre- to post-intervention analysis, a comparison of intervention vs. no intervention cohorts, and a qualitative analysis of two patient focus groups to understand the effect of TRIM, a multidisciplinary patient-centered weight management program for patients with obesity and with GERD. Overall, patients participating in TRIM had continuous improvements in symptom severity and QOL, as well as reduction in weight. Reductions in %EBW were significantly greater among TRIM participants compared with patients who declined TRIM participation. Finally, qualitative patient focus groups reinforced the benefits of multidisciplinary care and education as mechanisms to promote healthy behaviors. Thus, our feasibility study supports that TRIM is an acceptable and practical program associated with positive outcomes.
The epidemic of obesity continues to rise, and contributes a significant healthcare burden. Recently, several major societies developed joint practice guidelines and episode-of-care bundles on obesity and weight management, education, and resources (POWER) (11, 15). These papers identify obesity as a major modifiable cause of gastrointestinal diseases such as GERD, and exert that gastroenterologists are integral to the multidisciplinary care of obesity (11). According to POWER, obesity therapy must integrate multidisciplinary care and behavioral programs providing individualized dietary and physical activity plans.
Although developed before POWER, TRIM provides these precise cornerstones of obesity therapy. Moreover, TRIM by design overcomes common implementation challenges such as difficulty in patient identification, provider unwillingness to refer patients, and patient disengagement (9, 10, 11). TRIM uniquely leverages the EHR to identify patients via automated triggers, generates semiautomated referrals, and stores longitudinal patient-reported outcomes via embedded questionnaire templates. The EHR functionality minimizes workflow disruption and human error, provides ease of access to participants, and standardizes patient instruction and data collection. In addition, the functionality is generalizable across institutions and EHR portals. Thus, although in this study TRIM was personalized to patients with GERD, the multidisciplinary and electronic design and structure is such that it can be, and should be, incorporated into broader obesity programs, especially when obesity is specifically tied to a specific disease state (e.g., nonalcoholic fatty liver disease).
Although multidisciplinary care, education, and engagement are considered key elements to weight management programs for GERD, the supportive literature is limited by small sample sizes and lack of follow-up or real-world generalizability (16, 17). In this study we followed patients for 6 months during intervention and up to an additional 6 months after the end of intervention. The positive longitudinal results in this study emphasize the sustainability of TRIM with regard to self-management strategies and patient accountability. Of note, symptom improvement remained stable at 3 and 6 months despite an ongoing overall weight reduction, suggesting that mild persistent symptoms may be independent of weight change.
The qualitative layer in this study adds an additional strength and novelty that cannot be measured through quantitative analysis alone. Patients were extremely positive about the TRIM experience, and particularly valued the multidisciplinary approach and the relationship they developed with the health educators. Furthermore, patient feedback identified key areas for future assessment such as addressing barriers faced in the work setting and cultural differences related to healthy behaviors. Patients were enthusiastic about sharing experiences and learning from others, and requested more resources related to group intervention and support forums. Overall patients were accepting digital programs (e.g., a smartphone application) as a more accessible, long-term solution.
Through the focus groups we uncovered that many patients were unaware of and did not independently recognize the link between weight and GERD before TRIM participation. Yet, patients shared a strong desire to alleviate symptoms and reduce PPI use. Results from our mixed methods study suggest that weight loss becomes a more desirable and attainable goal once patients associate weight loss with symptom relief, further underscoring the value of health education. The idea that symptom relief is a powerful facilitator of healthy behaviors is intriguing and likely transcends symptom-based disorders across medicine.
There are important limitations in this study, many of which are inherent to feasibility studies. Despite several attempts to reach participants, approximately half of the group was lost to follow-up, introducing bias related to incomplete follow-up. We believe that the loss in follow-up relates to the work hour constraints of education sessions. In addition, the qualitative data suggest that the referring provider’s enthusiasm and encouragement may drive patient participation. In fact, it is possible that some patients simply needed to be told to “lose” weight and/or the association between obesity and GERD and then decided to proceed with weight loss on their own. Similarly, the 6-month program length may not be appropriate for all patients; some patients may require a longer duration of follow-up and some may feel they have the “tools” to succeed after 1–2 visits. Future iterations of the program should maintain the flexibility to adapt to individual patient needs. Although a loss in sample size, we limited the analysis to measured data and did not impute missing data from incomplete follow-up. Given the risk of variability and potentially reduced reliability of weights derived from two sources, we performed an a priori planned sensitivity analysis that demonstrated significantly high correlation between different weight collection types. We did not include waist measurement as an outcome in this feasibility study; however, we hope to provide participants with tape measures and instructions in order to incorporate anthropometric measurements in future studies. In addition, this was a pragmatic quality improvement study aimed to assess the effectiveness of implementing several best practice interventions simultaneously in order to minimize provider confusion and ease incorporation into the healthcare workflow (18). Although we recognize that the lack of a graded approach limits the ability to discern independent impacts, we inferred from our qualitative analysis that two aspects were of critical importance: (i) education that increased the awareness that weight loss can improve symptoms, particularly as reflux symptoms were described as life threatening, quality impairing, and troublesome at night and at work; and (ii) the frequent monitoring and patient-reporting required patients to be accountable. Recognizing that our interventions were higher quality than the current standard, our intention was to provide this best practice to all eligible patients. Finally, the comparator cohort (−TRIM) consisted of patients who self-declined participation in TRIM. There likely exists inherent differences in patients and their circumstances that affect selection to participate such as motivation, literacy, and access; this introduces a potential selection bias from unmeasurable or unaccounted confounders. The sample size limited our ability to perform propensity score matching as an alternate analysis. In addition, GerdQ scores were only collected for the +TRIM cohort, and thus we were unable to compare patient-reported outcomes between the +TRIM and −TRIM cohorts. Similarly, we appreciate that there is likely a bias toward TRIM participants positively reporting outcomes, inherent to all single-arm intervention studies focused on patient-reported outcomes However, the improvement in objective outcomes, especially excess weight loss, are consistent with the symptom improvement. Future studies should include a predefined control group and collect standardized data from both intervention and control groups.
The results from this work have generated further research ideas. Simultaneous to the TRIM program, we implemented a PPI tapering program for all patients with obesity and initiating PPI therapy. While 12-month follow-up of the PPI tapering initiative is ongoing, initial results suggest that among those who participated in TRIM and derived symptom improvement, a third are willing to taper their PPI therapy, and of these over 70% are able to withdraw PPIs completely (19). From a health services perspective, it is essential to understand why a significant portion of patients declined TRIM enrollment and why some clinicians declined patient referral to TRIM. Future intervention designs should explore mechanisms to further automate the referral process and examine the external generalizability and viability of TRIM.
In conclusion, this prospective mixed methods feasibility study highlights the positive effects of TRIM. Patients participating in TRIM had significant symptom reduction and when compared with patients not participating in TRIM, significant weight loss. Strengths of TRIM include optimized EHR functionality, minimal interference with clinician workflow, multidisciplinary care coordination, and a patient-centered approach. Lifestyle management programs tailored to patients with obesity and with GERD, such as TRIM, should be integrated into health promotion practices.
SUPPLEMENTARY MATERIAL is linked to the online version of the paper at http://www.nature.com/ajg