Abstract
Objectives
Fatigue is commonly reported in patients with Crohn’s disease (CD) and ulcerative colitis (UC), including patients with inactive disease. We explored the impact of fatigue on healthcare utilization (HCU) and work productivity and activity impairment (WPAI).
Methods
Data collected between 2017 and 2022 were analyzed from the CorEvitas IBD Registry. We compared HCU and WPAI among subjects with high fatigue (PROMIS ≥55) versus low fatigue at enrollment and subjects whose fatigue score worsened or persisted versus low fatigue at 6 months. HCU was defined as an inflammatory bowel disease-related hospitalization or emergency room visit. WPAI included presenteeism, absenteeism, and lost WPAI. Logistic regression analysis was performed.
Results
Study patients (640 CD, 569 UC) reported high rates of fatigue, 47% in CD and 38% in UC, that persisted at least 6 months in 88%-89% of patients. Patients with UC with high fatigue had 3-fold higher rates of HCU and 2-3-fold more absenteeism and activity impairment than patients with low fatigue. Patients with CD with high fatigue had no difference in HCU but did experience 2-4-fold more absenteeism, presenteeism, work productivity loss, and activity impairment. On subgroup analysis of patients in remission, those with high fatigue did not have higher rates of HCU but continued to have higher rates of WPAI.
Conclusions
Fatigue is associated with an increase in HCU only in the setting of concomitantly active disease. On the other hand, fatigue is associated with a negative impact on WPAI in the setting of both active and inactive disease.
Keywords: fatigue, work productivity, inflammatory bowel disease, disease activity, CorEvitas IBD Registry
Graphical Abstract
Graphical Abstract.
Introduction
Inflammatory bowel disease (IBD), commonly diagnosed in young adulthood, often affects people throughout their most productive years in the workforce. Chronic diseases can not only affect employment status (ability to work) but may affect work productivity by increasing absenteeism (days missed at work) and presenteeism (impaired productivity at work).1 Fatigue is reported as a significant predictor of work and activity impairment in people with chronic inflammatory diseases2 and is a commonly reported symptom among people with IBD, even during quiescent disease. Studies show fatigue is present in almost 50% of patients with IBD without active disease and in up to 80% of people with active disease.3–5 One prospective study showed that fatigue can persist up to 1 year after starting biologic therapy, 70% with fatigue at enrollment and 61% with persistent fatigue at 1 year.6 Fatigue has been associated with mental health conditions such as anxiety and depression, and other patient-reported outcomes including sleep disturbances.7,8
Fatigue has also been associated with increased healthcare resource utilization (HCU) in people with IBD,9,10 however, these studies have been limited. Understanding the impact of IBD, and its component driving symptoms, on healthcare costs is paramount given that healthcare expenditures in the United States for people with IBD have been shown to be very high with annual costs now exceeding $25 billion and continuing to increase annually.11 Moreover, over half of these costs were associated with inpatient and emergency room [ER] visits.
Despite the high prevalence of fatigue in IBD, there remains an incomplete understanding of its etiology and impact on individual lives and society. Further, while strategies have been proposed to screen for and treat fatigue,12–14 the effects of these interventions remain uncertain.6,15 With a better understanding of the impact, we hope this will drive urgency to find more effective interventions for fatigue. Therefore, the present study sought to measure the impact of fatigue in people with IBD on their work status, productivity, and HCU using a large US multicenter prospective cohort of patients.
Methods
Study Setting
The CorEvitas IBD Registry is a prospective observational cohort of adults diagnosed with Crohn’s disease (CD) or ulcerative colitis (UC) by a gastroenterologist. Registry patients are recruited from 100 private and academic sites with 260 gastroenterologists throughout 30 US states. Patients enrolled have initiated or switched to a biologic or Janus kinase (JAK) inhibitor at or within 12 months of enrollment, are not currently in a clinical trial, and have provided written informed consent. Patients have follow-up visits approximately every 6 months. Data are collected from both physicians and patients using Registry questionnaires completed at enrollment and follow-up. Data collected included demographics, medical history including history of comorbid diseases, disease characteristics and activity, treatment history, and patient-reported outcomes. At the time of this study, the registry included 1618 patients with CD and 1312 patients with UC; mean follow-up was 1.75 years and 1.86 years, respectively.
Study Population
The patients included in our study were enrolled in the CorEvitas IBD Registry between May 3, 2017, and April 1, 2022, and had both an enrollment and 6-month follow-up visit, with the PROMIS fatigue score recorded at both visits. Patients without a 6-month follow-up visit or with missing PROMIS fatigue score data were excluded.
Outcomes
The primary outcomes of this study are (1) to compare rates of HCU between CD or UC patients with and without fatigue and (2) to determine the impact of fatigue on employment rates and, among employed patients, work productivity and activity impairment (WPAI).
Measures
Fatigue was measured using the 6-item Patient-Reported Outcomes Measurement Information System (PROMIS) score (PROMIS Fatigue—Short Form 6a).16 The raw PROMIS score was converted to a standardized T-score with a mean of 50 and a standard deviation [SD] of 10. Based on previous studies, the PROMIS scores were dichotomized into two categories with clinically significant fatigue (called “high fatigue”) defined as patients with a PROMIS fatigue score ≥55 while low or no fatigue as <55.17 At the 6-month timepoint, worsening fatigue was defined as patients with high fatigue at enrollment and an increase in fatigue score ≥5 points at 6 months or patients with low fatigue at enrollment and fatigue score ≥55 at 6 months. Persistently high fatigue was defined as patients with high fatigue at enrollment whose fatigue did not worsen and did not have a decrease in their score to <55 at 6 months. The 5-point change in PROMIS score was determined to be the minimal clinically important change based on prior studies18 and correlation with our data set that a 0.5 SD correlated with a change in PROMIS of 4.7.19
Disease activity was determined using the Harvey Bradshaw Index (HBI) for CD and the partial Mayo score for UC, both measured at enrollment.20,21 Advanced therapy is defined as a biological or small-molecule medication. HCU was determined from any history of IBD-related hospitalizations and IBD-related emergency room visits at enrollment and, at the 6-month follow-up, about any IBD-related hospitalizations and emergency room visits since the enrollment visit. To determine work productivity among employed patients, the WPAI questionnaire, which evaluates absenteeism, presenteeism, work productivity loss, activity impairment, and employment status was collected at both the enrollment and 6-month follow-up visits.22,23
Additional covariates included a provider-reported history of anxiety and depression as well as the PROMIS measures for sleep disturbance, pain interference, anxiety, and depression.
Statistical Analysis
Descriptive statistics were used to summarize patient demographics, clinical and disease characteristics, and patient-reported outcomes by high fatigue versus low fatigue for CD and UC patients, separately. Categorical variables were summarized using frequency counts and percentages. Continuous variables were summarized using the number of observations, mean, and SD; medians and interquartile ranges were provided for variables with highly skewed distributions. Similar descriptive statistics were also calculated for the subgroup of patients who were in remission at enrollment by high fatigue versus low fatigue.
Logistic regression models were used to explore the associations between fatigue (high vs low) in CD and in UC and HCU, employment, and WPAI at enrollment and 6 months. Odds ratios [ORs] and adjusted ORs [aOR] along with each of their corresponding confidence intervals [CIs] were also reported. Additionally, a subgroup analysis was conducted for patients who were in remission at enrollment using similar methods. Potential imbalances between the groups were explored in the baseline tables using univariate testing (t-tests for continuous variables and chi-squared test for categorical variables). P values <.20 were considered as potential covariates to include in the adjusted models. All analyses were performed in R (Version 4.2.1).
Results
A total of 640 patients with CD and 569 with UC met inclusion and exclusion criteria and were analyzed in this study (Figure 1).
Figure 1.
Patient flow chart with exclusions. The patients who were included in the study were analyzed at enrollment (timepoint 1) and 6 months (timepoint 2).
Prevalence and Severity of Fatigue
Among 640 patients with CD, 299 (46.7%) reported significant fatigue (PROMIS score ≥55, hereafter noted as high fatigue) at enrollment; 104 (34.8%) were mild and 195 (65.2%) were moderate-severe. From enrollment to the 6-month visit, 35 (11.7%) patients with high fatigue had improvement in their fatigue, 107 (35.8%) had worsening fatigue and 157 (52.2%) patients remained persistently high.
Of 569 patients with UC, 218 (38.3%) had high fatigue at enrollment; 86 (39.4%) were mild and 132 (60.6%) were moderate-severe. From enrollment to the 6-month visit, 24 (11.0%) had resolution of their fatigue, 76 (34.9%) had worsening fatigue, and 118 (54.1%) had persistently high fatigue.
Characteristics of Patients With CD With High Fatigue
Compared to patients with CD with low fatigue, patients with CD with high fatigue at enrollment were on average younger, female, not working for pay, more likely to have depression or anxiety, and more likely to have an ileocolonic disease (P range from <.001 to .039, Table 1). Those with high fatigue were also more likely to be advanced therapy and steroid experienced (P = .011 and P = .001, respectively) or to use steroids at enrollment (P = .003).
Table 1.
Patient demographics, socioeconomic status, and clinical characteristics at enrollment for patients with Crohn’s Disease by fatigue status; high fatigue (PROMIS fatigue score ≥ 55) versus low fatigue (PROMIS fatigue score <55).
| High fatigue n = 299 |
Low fatigue n = 341 |
P-value | |
|---|---|---|---|
| Age in y; mean (SD) | 43.8 (16.4) | 47.1 (17.8) | .015 |
| Sex (%Male) | 108 (36.1%) | 166 (48.7%) | .001 |
| Race | .473 | ||
| White | 267 (89.6%) | 293 (86.4%) | |
| Black | 21 (7.0%) | 31 (9.1%) | |
| Other | 10 (3.4%) | 15 (4.4%) | |
| Ethnicitya | |||
| Hispanic | 24 (8.0%) | 31 (9.1%) | .632 |
| Jewish | 31 (10.4%) | 42 (12.3%) | .439 |
| Socioeconomics | |||
| Insurance typeb | |||
| Private | 226 (75.6%) | 265 (77.7%) | .525 |
| Medicare | 50 (16.7%) | 66 (19.4%) | .388 |
| Medicaid | 23 (7.7%) | 15 (4.4%) | .079 |
| Education | .173 | ||
| High school graduate/GED or less | 78 (26.1%) | 74 (21.8%) | |
| Some college/Assoc. degree | 82 (27.4%) | 83 (24.4%) | |
| College graduate or higher | 139 (46.5%) | 183 (53.8%) | |
| Work status | .039 | ||
| Full time/part-time | 170 (56.9%) | 221 (64.8%) | |
| Not working for pay | 129 (43.1%) | 120 (35.2%) | |
| Marital status | .071 | ||
| Single | 104 (34.8%) | 116 (34.1%) | |
| Married/partnered | 172 (57.5%) | 179 (52.6%) | |
| Other (separated, divorced, widowed) | 23 (7.7%) | 45 (13.2%) | |
| Comorbid diseases | |||
| Modified Charlson Comorbidity Indexc; mean (SD) | 0.3 (0.7) | 0.3 (0.5) | .095 |
| Depression | 64 (21.4%) | 21 (6.2%) | <.001 |
| Anxiety | 64 (21.4%) | 26 (7.6%) | <.001 |
| Current smoker | 38 (12.8%) | 23 (6.8%) | .040 |
| IBD disease characteristics | |||
| Disease duration; y (SD) | 14.2 (12.7) | 14.0 (12.5) | .869 |
| Age at diagnosis | 31.5 (15.7) | 34.9 (17.2) | .010 |
| Disease location | |||
| Ileal | 83 (27.8%) | 123 (36.1%) | .025 |
| Colonic | 79 (26.4%) | 96 (28.2%) | .624 |
| Ileocolonic | 130 (43.5%) | 118 (34.6%) | .021 |
| Upper GI involvement | 18 (6.0%) | 9 (2.6%) | .034 |
| Disease behavior | .934 | ||
| Penetrating | 36 (12.5%) | 36 (10.9%) | |
| Stricturing | 80 (27.9%) | 94 (28.5%) | |
| Inflammatory | 154 (53.7%) | 179 (54.2%) | |
| Perianal disease | 17 (5.9%) | 21 (6.4%) | |
| History of IBD-related surgery | 102 (34.1%) | 122 (35.8%) | .660 |
| History of IBD-related hospitalization | 206 (68.9%) | 203 (59.7%) | .016 |
| History of ER use for IBD-related issue | 193 (64.5%) | 185 (54.4%) | .009 |
| IBD medication use | |||
| Advanced therapy-experienced | .011 | ||
| 1 drug | 125 (41.8%) | 162 (47.5%) | |
| ≥2 drugs | 118 (39.5%) | 97 (28.4%) | |
| Steroid experienced | 216 (72.2%) | 205 (60.1%) | .001 |
| Current IBD medication use at enrollment | |||
| Biologic—anti-TNF | 135 (45.2%) | 149 (43.7%) | .712 |
| Other Biologic—anti-integrin, IL23 ± 12 | 81 (27.1%) | 88 (25.8%) | .713 |
| Small molecule use | 0 (0%) | 0 (0%) | .097 |
| Immunomodulator | 46 (15.4%) | 63 (18.5%) | .299 |
| 5-ASA | 62 (20.7%) | 64 (18.8%) | .532 |
| Steroids | 79 (26.4%) | 57 (16.7%) | .003 |
Abbreviations: IBD, inflammatory bowel disease; PROMIS, Patient-Reported Outcomes Measurement Information System; SD, standard deviation.
aRace and ethnicity are not mutually exclusive variables.
bInsurance type is not mutually exclusive.
cModified Charlson Comorbidity Index: Sum of prior (history of) myocardial infarction, congestive heart failure, peripheral vascular disease, cerebrovascular disease, chronic obstructive pulmonary disease (COPD), peptic ulcer disease, diabetes mellitus, lymphoma, solid-tumor cancer (excluding nonmelanoma skin cancer), renal disease (creatinine >3), mild liver disease (hepatic events), and moderate/severe liver disease (PSC, autoimmune hepatitis).
Most patients with low fatigue at enrollment were in remission (78.6%) compared to only 56.2% of patients with high fatigue as measured by the HBI score (P < .001, Table 2). Patients with CD with high fatigue were more likely to report worse sleep disturbance, pain interference, depression, and anxiety at enrollment compared to patients with low fatigue (all P < .001).
Table 2.
Disease activity at enrollment for patients with Crohn’s was assessed using the Harvey Bradshaw Index (HBI).
| High fatigue n = 299 |
Low fatigue n = 341 |
P-value | |
|---|---|---|---|
| Disease activity | <.001 | ||
| Remission (HBI <5) | 155 (56.2%) | 253 (78.6%) | |
| Mild disease (HBI 5-7) | 64 (23.2%) | 49 (15.2%) | |
| Moderate disease (HBI 8-16) | 53 (19.2%) | 20 (6.2%) | |
| Severe disease (HBI >16) | 4 (1.4%) | 0 (0.0%) | |
| PROMIS score | |||
| Fatigue; mean (SD) | 62.4 (5.6) | 44.9 (7.1) | <.001 |
| Within normal limits (<55) | 0 (0.0%) | 341 (100.0%) | <.001 |
| Mild (55-60) | 104 (34.8%) | 0 (0.0%) | |
| Moderate/severe (≥60) | 195 (65.2%) | 0 (0.0%) | |
| Sleep disturbance | <.001 | ||
| Within normal limits (<55) | 166 (55.5%) | 299 (88.2%) | |
| Mild (55-60) | 65 (21.7%) | 28 (8.3%) | |
| Moderate/severe (≥60) | 68 (22.7%) | 12 (3.5%) | |
| Pain interference | <.001 | ||
| Within normal limits (<55) | 97 (32.4%) | 274 (80.6%) | |
| Mild (55-60) | 68 (22.7%) | 45 (13.2%) | |
| Moderate/severe (≥60) | 134 (44.8%) | 21 (6.2%) | |
| Depression | <.001 | ||
| Within normal limits (<55) | 167 (56.0%) | 316 (92.9%) | |
| Mild (55-60) | 64 (21.5%) | 18 (5.3%) | |
| Moderate/severe (≥60) | 67 (22.5%) | 6 (1.8%) | |
| Anxiety | <.001 | ||
| Within normal limits (<55) | 120 (40.1%) | 256 (75.1%) | |
| Mild (55-60) | 74 (24.7%) | 64 (18.8%) | |
| Moderate/severe (≥60) | 105 (35.1%) | 21 (6.2%) | |
Details of the PROMIS score subcategories, including sleep disturbance, pain interference, depression, and anxiety is compared between patients by fatigue status; high fatigue (PROMIS fatigue score ≥55) versus low fatigue (PROMIS fatigue score <55).
Abbreviation: PROMIS, Patient-Reported Outcomes Measurement Information System.
Characteristics of UC Patients With High Fatigue
Patients with UC with high fatigue at enrollment were on average younger, female, more likely to smoke, and more likely to have depression or anxiety compared to patients with UC with low fatigue (P range <.001 to .030, Table 3). Patients with UC reporting high fatigue were more likely to be advanced therapy-experienced (P = .041) and less likely to take a 5-ASA at enrollment (P = .003) compared to patients with UC with low fatigue.
Table 3.
Patient demographics, socioeconomic status, and clinical characteristics at enrollment for patients with ulcerative colitis by fatigue status; high fatigue (PROMIS fatigue score ≥55) versus low fatigue (PROMIS fatigue score <55).
| High fatigue n = 218 |
Low fatigue n = 351 |
P-value | |
|---|---|---|---|
| Age in y; mean (SD) | 44.8 (16.1) | 47.9 (17.0) | .030 |
| Sex (% male) | 90 (41.3%) | 185 (52.7%) | .008 |
| Race | .312 | ||
| White | 192 (88.1%) | 292 (83.4%) | |
| Black | 12 (5.5%) | 28 (8.0%) | |
| Other | 14 (6.4%) | 30 (8.6%) | |
| Ethnicitya | |||
| Hispanic | 31 (14.2%) | 60 (17.1%) | .363 |
| Jewish | 26 (11.9%) | 39 (11.1%) | .766 |
| Socioeconomics | |||
| Insurance typeb | |||
| Private | 166 (76.1%) | 272 (77.5%) | .711 |
| Medicare | 35 (16.1%) | 58 (16.5%) | .883 |
| Medicaid | 12 (5.5%) | 7 (2.0%) | .023 |
| Education | .710 | ||
| High school graduate/GED or less | 45 (20.6%) | 67 (19.1%) | |
| Some college/Assoc. degree | 64 (29.4%) | 96 (27.4%) | |
| College graduate or higher | 109 (50.0%) | 188 (53.6%) | |
| Work status | .516 | ||
| Full time/part-time | 137 (62.8%) | 230 (65.5%) | |
| Not working for pay | 81 (37.2%) | 121 (34.5%) | |
| Marital status | .484 | ||
| Single | 74 (33.9%) | 102 (29.1%) | |
| Married/partnered | 117 (53.7%) | 202 (57.7%) | |
| Other (separated, divorced, widowed) | 27 (12.4%) | 46 (13.1%) | |
| Comorbid diseases | |||
| Modified Charlson Comorbidity Indexc; mean (SD) | 0.3 (0.5) | 0.3 (0.6) | .588 |
| Depression | 34 (15.6%) | 17 (4.8%) | <.001 |
| Anxiety | 51 (23.4%) | 24 (6.8%) | <.001 |
| Current smoker | 14 (6.5%) | 5 (1.4%) | <.001 |
| IBD disease characteristics | |||
| Disease duration; y (SD) | 10.3 (10.0) | 10.8 (9.9) | .319 |
| Age at diagnosis | 35.4 (15.4) | 37.7 (16.8) | .105 |
| Disease location | |||
| Proctitis | 83 (27.8%) | 123 (36.1%) | .528 |
| Proctosigmoiditis | 79 (26.4%) | 96 (28.2%) | .429 |
| Left-sided | 130 (43.5%) | 118 (34.6%) | .650 |
| Pan-colitis | 18 (6.0%) | 9 (2.6%) | .210 |
| History of IBD-related surgery | 6 (2.8%) | 10 (2.8%) | .946 |
| History or IBD-related hospitalization | 109 (50.0%) | 144 (41.0%) | .036 |
| History of IBD-related ER visit | 111 (50.9%) | 132 (37.7%) | .002 |
| IBD medication use | |||
| Advanced therapy-experienced | .041 | ||
| 1 drug | 102 (46.8%) | 134 (38.2%) | |
| ≥2 drugs | 45 (20.6%) | 66 (18.8%) | |
| Steroid experienced | 161 (73.9%) | 247 (70.4%) | .370 |
| Current IBD medication use at enrolment | |||
| Biologic—anti-TNF | 71 (32.6%) | 94 (26.8%) | .139 |
| Biologic—anti-integrin, Il23 ± 12 | 64 (29.4%) | 75 (21.4%) | .031 |
| Small molecule | 8 (3.7%) | 13 (3.7%) | .983 |
| Immunomodulator | 34 (15.6%) | 48 (13.7%) | .526 |
| 5-ASA | 96 (44.0%) | 199 (56.7%) | .003 |
| Steroids | 61 (28.0%) | 86 (24.5%) | .357 |
Abbreviations: IBD, inflammatory bowel disease; PROMIS, Patient-Reported Outcomes Measurement Information System; SD, standard deviation.
aRace and ethnicity are not mutually exclusive variables.
bInsurance type is not mutually exclusive.
cModified Charlson Comorbidity Index: Sum of prior (history of) myocardial infarction, congestive heart failure, peripheral vascular disease, cerebrovascular disease, chronic obstructive pulmonary disease (COPD), peptic ulcer disease, diabetes mellitus, lymphoma, solid-tumor cancer (excluding nonmelanoma skin cancer), renal disease (creatinine >3), mild liver disease (hepatic events), and moderate/severe liver disease (PSC, autoimmune hepatitis).
A lower proportion of patients with UC with high fatigue were in remission at enrollment (38.8% vs 53.7%, P < .001) and a larger proportion had moderate to severe disease (25.8% vs 12.8%, P < .001) compared to patients with UC with lower fatigue at enrollment (Table 4). Compared to patients with low fatigue, patients with high fatigue at enrollment reported more sleep disturbance, pain interference, depression, and anxiety at enrollment (all P < .001).
Table 4.
Disease activity at enrollment for ulcerative colitis patients was assessed using the partial Mayo Score.
| High fatigue n = 218 |
Low fatigue n = 351 |
P-value | |
|---|---|---|---|
| Disease activity (partial Mayo) | <.001 | ||
| Remission (0-1) | 81 (38.8%) | 181 (53.7%) | |
| Mild Disease (2-4) | 74 (35.4%) | 113 (33.5%) | |
| Moderate Disease (5-6) | 35 (16.7%) | 33 (9.8%) | |
| Severe Disease (7-9) | 19 (9.1%) | 10 (3.0%) | |
| PROMIS score | |||
| Fatigue; mean (SD) | 61.8 (5.2) | 44.3 (6.9) | <.001 |
| Within normal limits (<55) | 0 (0.0%) | 351 (100.0%) | <.001 |
| Mild (55-60) | 86 (39.4%) | 0 (0.0%) | |
| Moderate/severe (≥60) | 132 (60.6%) | 0 (0.0%) | |
| Sleep disturbance | <.001 | ||
| Within normal limits (<55) | 115 (52.8%) | 302 (86.3%) | |
| Mild (55-60) | 50 (22.9%) | 32 (9.1%) | |
| Moderate/severe (≥60) | 53 (24.3%) | 16 (4.6%) | |
| Pain interference | <.001 | ||
| Within normal limits (<55) | 82 (37.6%) | 280 (80.0%) | |
| Mild (55-60) | 56 (25.7%) | 49 (14.0%) | |
| Moderate/severe (≥60) | 80 (36.7%) | 21 (6.0%) | |
| Depression | <.001 | ||
| Within normal limits (<55) | 118 (54.1%) | 309 (88.0%) | |
| Mild (55-60) | 55 (25.2%) | 33 (9.4%) | |
| Moderate/severe (≥60) | 45 (20.6%) | 9 (2.6%) | |
| Anxiety | <.001 | ||
| Within normal limits (<55) | 84 (38.5%) | 253 (72.1%) | |
| Mild (55-60) | 48 (22.0%) | 64 (18.2%) | |
| Moderate/severe (≥60) | 86 (39.4%) | 34 (9.7%) | |
Details of the PROMIS score subcategories including sleep disturbance, pain interference, depression, and anxiety compared between patients by fatigue status; high fatigue (PROMIS fatigue score ≥55) versus low fatigue (PROMIS fatigue score <55).
Abbreviation: PROMIS, Patient-Reported Outcomes Measurement Information System.
Healthcare Utilization
Compared to patients with low fatigue, patients with CD with high fatigue at enrollment had a similar history of IBD-related hospitalization (aOR 1.21, 95% CI 0.81, 1.82, adj. P = .354) and IBD-related ER visits (aOR 1.06, 95% CI 0.72, 1.56, adj. P = .776) after adjustment for baseline differences. At 6 months of follow-up, IBD-related hospitalizations and ER visits were higher for patients with high fatigue, but after adjustment for baseline differences, this was not significant (hospitalization: aOR 1.57, 95% CI 0.76, 3.23, adj. P = .223; ER visits: aOR 1.16, 95% CI 0.59-2.28, adj. P = .675; Table 5).
Table 5.
Hospitalizations and ER utilization at baseline and after enrollment (out to 6 months of follow-up) compared between patients with Crohn’s and ulcerative colitis with high versus low fatigue at enrollment.
| High fatigue | Low fatigue | OR (95% CI) | P-value | Adjusted OR (95% CI) a,b | Adjusted P-valuea,b | |
|---|---|---|---|---|---|---|
| Crohn’s disease | n = 299 | n = 341 | ||||
| History of IBD-related hospitalization at enrollment | 206 (68.9%) | 203 (59.7%) | 1.49 (1.08, 2.07) | .016 | 1.21 (0.81, 1.82) | .354 |
| History of ER use for IBD-related issues at enrollment | 193 (64.5%) | 185 (54.4%) | 1.53 (1.11, 2.1) | .009 | 1.06 (0.72, 1.56) | .776 |
| Hospitalization after enrollment | 36 (12.0%) | 18 (5.3%) | 2.46 (1.36, 4.43) | .003 | 1.57 (0.76, 3.23) | .223 |
| ER visit after enrollment | 38 (12.8%) | 24 (7.0%) | 1.93 (1.13, 3.3) | .016 | 1.16 (0.59, 2.28) | .675 |
| Hospitalization or ER visit after enrollment | 44 (14.8%) | 29 (8.5%) | 1.86 (1.13, 3.06) | .014 | 1.17 (0.63, 2.15) | .623 |
| Ulcerative colitis | n = 218 | n = 351 | ||||
| History of IBD-related hospitalization at enrollment | 109 (50.0%) | 144 (41.0%) | 1.44 (1.02, 2.02) | .037 | 1.1 (0.73, 1.65) | .663 |
| History of ER use for IBD-related issue at enrollment | 111 (50.9%) | 132 (37.7%) | 1.71 (1.22, 2.41) | .002 | 1.29 (0.86, 1.95) | .222 |
| Hospitalization after enrollment | 19 (8.7%) | 5 (1.4%) | 6.61 (2.43, 17.97) | <.001 | 3.66 (1.24, 10.8) | .019 |
| ER visit after enrollment | 18 (8.3%) | 6 (1.7%) | 5.17 (2.02, 13.25) | <.001 | 2.77 (0.98, 7.82) | .055 |
| Hospitalization or ER visit after enrollment | 24 (11.0%) | 7 (2.0%) | 6.08 (2.57, 14.37) | <.001 | 3.25 (1.26, 8.35) | .014 |
Abbreviations: CI, confidence interval; IBD, inflammatory bowel disease; OR, odds ratio.
aCD models: odds ratios and P values were obtained using logistic regression models; adjusted odds ratios and P values were obtained using logistic regression models adjusted for age, sex, education, marital status, current alcohol use, history of CVD, history of respiratory, history of depression, history of anxiety, number of prior biologics/JAKi/S1P, concomitant steroids, duration of current IBD therapy, Harvey Bradshaw index score (continuous), and WPAI current employment.
bUC models: odds ratios and P values were obtained using logistic regression models; adjusted odds ratios and P values were obtained using logistic regression models adjusted for age, sex, smoking status, current alcohol use, history of CVD, history of depression, history of anxiety, history of other non-serious medical condition, pan-colitis, concomitant biologics/JAKi/S1P, concomitant 5-ASAs, duration of current IBD therapy (years), and Partial Mayo score (continuous).
Patients with UC with high fatigue at enrollment were more likely to have IBD-related hospitalization or ER visits compared to patients with UC with low fatigue, although these were not statistically significantly different after adjustment for baseline differences (hospitalization: aOR 1.10, 95% CI 0.73, 1.65, adj. P = .663; ER visits: aOR 1.29, 95% CI 0.86, 1.95, adj. P = .222). At 6 months of follow-up, IBD-related hospitalization was statistically significantly higher for those with high fatigue after adjustment for baseline differences (aOR 3.66, 95% CI 1.24, 10.80, adj. P = .019), but the larger proportion of ER visits was not statistically significant (aOR 2.77, 95% CI 0.98, 7.82, adj. P = .055).
Employment Rates and Work Productivity
For patients with CD and UC, 402 (62.8%) and 374 (65.7%) were employed at enrollment, respectively, (Table 6). Enrollment employment rates were similar between patients with CD with high fatigue versus low fatigue (aOR 0.93, 95% CI 0.6, 1.45, adj. P = .762) as well as between patients with UC with high fatigue versus low fatigue (aOR 0.86, 95% CI 0.55, 1.33, adj. P = .489). However, patients with CD with high fatigue reported significantly more absenteeism (aOR 1.94, 95% CI 1.01, 3.72, adj. P = .046), more presenteeism (aOR 4.31, 95% CI 2.41, 7.7, adj. P < .001), more work productivity loss (aOR 4.20, 95% CI 2.25, 7.84, adj. P < .001) and more daily activity impairment (aOR 3.30, 95% CI 2.06, 5.27, adj. P < .001) compared to patients with CD with low fatigue. Similarly, patients with UC with high fatigue reported significantly more absenteeism (aOR 3.41, 95% CI 1.71, 6.82, adj. P < .001) and activity impairment (aOR 2.20, 95% CI 1.33, 3.64, adj. P = .002), and non-significantly greater presenteeism (aOR 1.76, 95% CI 0.98, 3.13, adj. P = .057).
Table 6.
Associations of fatigue (high vs low) with work productivity and activity impairment measured at enrollment and 6 months for patients with Crohn’s disease and ulcerative colitis.
| High fatigue | Low fatigue | OR (95% CI) | P-value | Adjusted OR (95% CI) | Adjusted P-valuea,b | |
|---|---|---|---|---|---|---|
| Crohn’s disease | n = 299 | n = 341 | ||||
| Enrollment | ||||||
| Employed | 177 (59.2%) | 225 (66.0%) | 0.75 (0.54, 1.03) | .077 | 0.93 (0.6, 1.45) | .762 |
| Any absenteeismc | 45 (26.6%) | 28 (13.4%) | 2.35 (1.39, 3.96) | .001 | 1.94 (1.01, 3.72) | .046 |
| Any presenteeismd | 147 (84.5%) | 113 (50.9%) | 5.25 (3.23, 8.55) | <.001 | 4.31 (2.41, 7.7) | <.001 |
| Any lost work productivitye | 143 (85.6%) | 112 (53.8%) | 5.11 (3.06, 8.51) | <.001 | 4.2 (2.25, 7.84) | <.001 |
| Any activity impairmentf | 255 (85.6%) | 193 (56.9%) | 4.49 (3.04, 6.61) | <.001 | 3.3 (2.06, 5.27) | <.001 |
| At 6 months | ||||||
| Employed | 174 (58.2%) | 223 (65.4%) | 0.74 (0.53, 1.01) | .061 | 1.06 (0.54, 2.09) | .857 |
| Any absenteeismc | 35 (21.0%) | 27 (12.7%) | 1.83 (1.05, 3.16) | .032 | 1.47 (0.77, 2.79) | .245 |
| Any presenteeismd | 127 (75.1%) | 116 (52.0%) | 2.79 (1.8, 4.32) | <.001 | 1.94 (1.15, 3.25) | .012 |
| Any lost work productivitye | 126 (76.8%) | 115 (54.0%) | 2.83 (1.8, 4.44) | <.001 | 1.91 (1.12, 3.28) | .018 |
| Any activity impairmentf | 251 (84.2%) | 190 (55.7%) | 4.24 (2.91, 6.19) | <.001 | 3.39 (2.15, 5.33) | <.001 |
| Ulcerative colitis | n = 218 | n = 351 | ||||
| Enrollment | ||||||
| Employed | 140 (64.2%) | 234 (66.7%) | 0.9 (0.63, 1.28) | .550 | 0.92 (0.59, 1.42) | .701 |
| Any absenteeismc | 48 (36.6%) | 23 (10.7%) | 4.83 (2.76, 8.45) | <.001 | 3.41 (1.71, 6.82) | <.001 |
| Any presenteeismd | 102 (73.4%) | 120 (51.7%) | 2.57 (1.63, 4.06) | <.001 | 1.76 (0.98, 3.13) | .057 |
| Any lost work productivitye | 99 (76.2%) | 117 (54.4%) | 2.67 (1.65, 4.34) | <.001 | 1.58 (0.84, 2.98) | .157 |
| Any activity impairmentf | 176 (81.1%) | 199 (56.7%) | 3.28 (2.2, 4.89) | <.001 | 2.2 (1.33, 3.64) | .002 |
| At 6 months | ||||||
| Employed | 134 (61.5%) | 231 (65.8%) | 0.83 (0.58, 1.18) | .294 | 0.86 (0.55, 1.33) | .489 |
| Any absenteeismc | 24 (20.0%) | 16 (7.3%) | 3.16 (1.6, 6.22) | <.001 | 2.26 (0.96, 5.33) | .062 |
| Any presenteeismd | 84 (63.2%) | 105 (45.7%) | 2.04 (1.32, 3.16) | .001 | 1.48 (0.85, 2.58) | .162 |
| Any lost work productivitye | 79 (66.4%) | 104 (47.7%) | 2.16 (1.36, 3.44) | .001 | 1.61 (0.9, 2.88) | .110 |
| Any activity impairmentf | 155 (71.1%) | 181 (51.9%) | 2.28 (1.59, 3.27) | <.001 | 1.66 (1.09, 2.54) | .019 |
Abbreviations: CI, confidence interval; OR, odds ratio.
aCD models: odds ratios and P values were be obtained using logistic regression models; adjusted odds ratios and P values were obtained using logistic regression models adjusted for age, sex, education, marital status, current alcohol use, history of CVD, history of respiratory, history of depression, history of anxiety, number of prior biologics/JAKi/S1P, concomitant steroids, duration of current IBD therapy, Harvey Bradshaw index score (continuous), and WPAI current employment.
bUC models: odds ratios and P values were be obtained using logistic regression models; adjusted odds ratios and P values were obtained using logistic regression models adjusted for age, sex, smoking status, current alcohol use, history of CVD, history of depression, history of anxiety, history of other non-serious medical condition, pan-colitis, concomitant biologics/JAKi/S1P, concomitant 5-ASAs, duration of current IBD therapy (years), and Partial Mayo score (continuous).
cAny absenteeism (>0% of work hours missed due to IBD reported at enrollment).
dAny presenteeism (>0% of impairment while working due to IBD reported at enrollment).
eAny work productivity loss (>0% of overall work hours affected by IBD reported at enrollment).
fAny activity impairment (>0% of daily activities affected by IBD reported at enrollment).
At 6 months of follow-up, both patients with CD and patients with UC with high fatigue did not have a significant difference in employment status (both P > .05). However, patients with CD with high fatigue had significantly greater presenteeism (aOR 1.94, 95% CI 1.15, 3.25, adj. P = .012), work productivity loss (aOR 1.94, 95% CI 1.15, 3.25, adj. P = .018), and any activity impairment (aOR 3.39, 95% CI 2.15, 5.33, adj. P < .001) compared to patients with low fatigue. Absenteeism was not significantly different after adjustment for covariates (aOR 1.47, 95% CI 0.77, 2.79, adj. P = .245). At 6 months, patients with UC with high fatigue had more activity impairment (aOR 1.66, 95% CI 1.09, 2.54, adj. P = .019) than patients with UC with low fatigue. Although not statistically significant, absenteeism was trending towards significance in adjusted models (aOR 2.26, 95% CI 0.96, 5.33, adj. P = .062) indicating patients with UC with high fatigue may have higher absenteeism than patients with low fatigue. Presenteeism and work productivity loss were not significantly different between patients with UC with high and low fatigue after adjustment for covariates (both adj. P > .05).
Subgroup Analysis of Patients in Remission With Fatigue
Within the group of patients without active disease, 38% of patients with CD and 31% of patients with UC reported high fatigue. At enrollment, patients with CD in remission with high fatigue did not have a higher rate of prior hospitalizations (aOR 1.44, 95% CI 0.87, 2.37, adj. P = .156) or ER visits (aOR 1.28, 95% CI 0.79, 2.07, adj. P = .323) than patients with low fatigue (Table 7). Similarly, for patients with UC in remission with high fatigue, there were no differences in rates of prior hospitalization (aOR 0.93, 95% CI 0.49, 1.78, adj. P = .837) or ER visits (aOR 1.28, 95% CI 0.67, 2.44, adj. P = .452) as compared to those with low fatigue. At 6 months of follow-up, patients with CD and UC in remission at enrollment continued to not have any differences in rates of ER visits or hospitalizations between those with high versus low fatigue at enrollment (aORs ranged from 1.68-4.10, all adj. P > .05).
Table 7.
Associations of fatigue (high vs low) with employment, work productivity, and healthcare utilization in patients with Crohn’s disease and ulcerative colitis in remission at enrollment.
| High fatigue | Low fatigue | OR (95% CI) | P-value | Adjusted OR (95% CI)a,b | Adjusted P-valuea,b | |
|---|---|---|---|---|---|---|
| Crohn’s disease | n = 155 | n = 253 | ||||
| WPAI at enrollment | ||||||
| Employed | 93 (60.0%) | 164 (64.8%) | 0.81 (0.54, 1.23) | .328 | 0.8 (0.45, 1.4) | .425 |
| Any absenteeismc | 16 (18.2%) | 16 (10.5%) | 1.89 (0.89, 4) | .096 | 1.67 (0.67, 4.17) | .269 |
| Any presenteeismd | 71 (78.0%) | 73 (45.3%) | 4.28 (2.38, 7.68) | <.001 | 4.61 (2.25, 9.43) | <.001 |
| Any lost work productivitye | 68 (78.2%) | 72 (47.7%) | 3.93 (2.15, 7.16) | <.001 | 4.4 (2.07, 9.36) | <.001 |
| Any activity impairmentf | 117 (76.0%) | 128 (50.8%) | 3.06 (1.96, 4.78) | <.001 | 2.82 (1.67, 4.75) | <.001 |
| History of hospitalization at enrollment | 104 (67.1%) | 147 (58.1%) | 1.47 (0.97, 2.23) | .071 | 1.44 (0.87, 2.37) | .156 |
| History of ER visit at enrollment | 94 (60.6%) | 135 (53.4%) | 1.35 (0.9, 2.02) | .151 | 1.28 (0.79, 2.07) | .323 |
| History of hospitalization or ER visit at enrollment | 112 (72.3%) | 166 (65.6%) | 1.37 (0.88, 2.11) | .163 | 1.33 (0.79, 2.24) | .290 |
| Hospitalization after enrollment | 13 (8.4%) | 11 (4.3%) | 2.01 (0.88, 4.62) | .098 | 2.36 (0.86, 6.47) | .094 |
| ER visit after enrollment | 14 (9.1%) | 15 (5.9%) | 1.59 (0.74, 3.38) | .232 | 1.68 (0.68, 4.14) | .262 |
| Hospitalization or ER visit after enrollment | 16 (10.4%) | 19 (7.5%) | 1.43 (0.71, 2.87) | .317 | 1.28 (0.56, 2.91) | .558 |
| Ulcerative colitis | n = 81 | n = 181 | ||||
| WPAI at enrollment | ||||||
| Current employment | 52 (64.2%) | 119 (65.7%) | 0.93 (0.54, 1.62) | .808 | 1.06 (0.53, 2.14) | .865 |
| Any absenteeismc | 9 (18.0%) | 6 (5.4%) | 3.88 (1.3, 11.58) | .015 | 2.52 (0.64, 9.92) | .187 |
| Any presenteeismd | 27 (51.9%) | 42 (35.6%) | 1.95 (1.01, 3.79) | .047 | 2.17 (0.98, 4.84) | .057 |
| Any lost work productivitye | 27 (54.0%) | 42 (37.5%) | 1.96 (1, 3.84) | .051 | 2.06 (0.91, 4.63) | .081 |
| Any activity impairmentf | 51 (63.0%) | 72 (39.8%) | 2.57 (1.5, 4.42) | <.001 | 2.45 (1.31, 4.58) | .005 |
| History of hospitalization at enrollment | 34 (42.0%) | 72 (39.8%) | 1.1 (0.64, 1.86) | .738 | 0.93 (0.49, 1.78) | .837 |
| History of ER visit at enrollment | 35 (43.2%) | 65 (35.9%) | 1.36 (0.8, 2.32) | .262 | 1.28 (0.67, 2.44) | .452 |
| History of hospitalization or ER visit at enrollment | 42 (51.9%) | 78 (43.1%) | 1.42 (0.84, 2.41) | .189 | 1.22 (0.64, 2.33) | .538 |
| Hospitalization after enrollment | 3 (3.7%) | 2 (1.1%) | 3.44 (0.56, 1.01) | .180 | 3.99 (0.55, 28.96) | .172 |
| ER visit after enrollment | 3 (3.7%) | 2 (1.1%) | 3.44 (0.56, 21.01) | .180 | 4.1 (0.54, 31.34) | .174 |
| Hospitalization or ER visit after enrollment | 4 (4.9%) | 3 (1.7%) | 3.08 (0.67, 14.1) | .147 | 3.26 (0.58, 18.32) | .180 |
Abbreviations: CI, confidence interval; OR, odds ratio; WPAI, work productivity and activity impairment.
aCD models: odds ratios and P values were obtained using logistic regression models; adjusted odds ratios and P values were obtained using logistic regression models adjusted for age, sex, white race, BMI categorical, current alcohol use, history of CVD, history of respiratory comorbidities, history of depression, history of anxiety, number of prior biologic/JAKi/S1Ps, 5-ASA experienced, concomitant ISTs, concomitant steroids, duration of current IBD therapy.
bUC models: Odds ratios and P values were obtained using logistic regression models; adjusted odds ratios and P values were obtained using logistic regression models adjusted for sex, white race, Jewish ethnicity, private insurance, smoking status, history of anxiety, time since IBD diagnosis, concomitant biologics/JAKi/S1Ps, concomitant 5-ASAs, duration of current IBD therapy, and NSAID use.
cAny absenteeism (>0% of work hours missed due to IBD reported at enrollment).
dAny presenteeism (>0% of impairment while working due to IBD reported at enrollment).
eAny work productivity loss (>0% of overall work hours affected by IBD reported at enrollment).
fAny activity impairment (>0% of daily activities affected by IBD reported at enrollment).
For patients with CD at enrollment, those with high fatigue had similar rates of employment and absenteeism but had significantly higher rates of presenteeism (aOR 4.61, 95% CI 2.25, 9.43, adj. P < .001), work productivity loss (aOR 4.4, 95% CI 2.07, 9.36, adj. P < .001), and activity impairment (aOR 2.82, 95% CI 1.67, 4.75) as compared to those with low fatigue (Table 7). Similarly, for patients with UC at enrollment, those with high fatigue had similar rates of employment and absenteeism as those with low fatigue. Those with high fatigue trended towards higher rates of presenteeism (aOR 2.17, 95% CI 0.98, 4.84, adj. P = .057) and work productivity loss (aOR 2.06, 95% CI 0.91, 4.63, adj. P = .081) and had significantly more activity impairment (aOR 2.45, 95% CI 1.31, 4.58, adj. P = .005). At 6 months, patients with CD with high fatigue at enrollment continued to have higher rates of presenteeism [aOR 2.06, 95% CI 1.07, 3.97, adj. P = .031] and activity impairment [aOR 3.16, 95% CI 1.86, 5.37, adj. P < .001] (data not shown). For patients with UC at 6 months, those with high fatigue at enrollment had significantly higher rates of presenteeism [aOR 3.32, 95% CI 1.3, 8.46, adj. P = .012], work productivity loss [aOR 3.25, 95% CI 1.23, 8.58, adj. P = .017], and activity impairment [aOR 1.91, 95% CI 1.02, 3.57, adj. P = .043].
Discussion
In this study, we found that patients with both CD and UC commonly report high rates of fatigue, 47% in CD and 38% in UC. Of these patients with high fatigue, 60%-65% reported moderate to severe fatigue. This is in line with previously published prevalence rates of fatigue in patients with IBD7,24 and contrasts with the 10%-20% rate of fatigue reported in the general population.3,4,25 After 6 months of follow-up, this fatigue persists or worsens in 88%-89% of patients who initially reported fatigue. High fatigue was associated with anxiety, depression, sleep disturbance, and pain interference as in other cohorts.7,8
Our first outcome of interest was the impact of fatigue on HCU. We found that during the 6 months of follow-up, patients with UC with high fatigue had 3-fold higher rates of hospitalization or ER visits than patients with low fatigue even after adjustment for potential confounders. Further, this association held true for hospitalizations when assessed separately. Interestingly in a subgroup analysis of patients with UC with no disease activity, we found that patients with high fatigue at enrollment did not have higher rates of hospitalization or ER visits suggesting the increase in utilization seen in the cohort as a whole may be driven more by disease activity with associated fatigue than fatigue itself. For patients with CD with high fatigue, there were higher rates of hospitalization and ER visits in those with high fatigue, but this was no longer significant after adjustment for confounders. When looking at the subgroup of patients with CD in remission, those with high fatigue did not have higher rates of ER visits or hospitalizations. The literature on the impact of fatigue in patients with IBD regarding HCU is sparse. In a single center, a retrospective study assessing symptoms that were associated with HCU, fatigue was found to be associated with a 2-fold increase in HCU on unadjusted analysis however this lost significance once the analysis was adjusted for confounders.9 This study was limited by its smaller population size and by the heterogeneity of combining UC and CD in their analysis. Moreover, fatigue was not the primary focus of the study.
Recently, a large administrative database was analyzed to assess the impact of fatigue versus no fatigue on HCU in newly diagnosed patients with IBD. In this analysis, the presence of fatigue was determined if an ICD-9/10 code was present after the diagnosis of IBD was made.10 They found that the presence of fatigue was associated with an almost 2-fold increase in HCU. A major strength of this study was its large size and matching, however, the dataset was limited by the definition of fatigue using ICD-9/10 codes may have underestimated the presence of fatigue, provided no data on fatigue severity, and did not have data on IBD disease activity. Our study, despite being smaller, used the PROMIS fatigue scores to objectively assess the presence and severity of fatigue as well as used disease activity scores to assess the activity of IBD. Our data similarly suggests that in both CD and UC, high fatigue is associated with more utilization of ER and hospitalization, but our subgroup analysis interestingly suggests this may be more due to the underlying disease activity rather than fatigue itself.
The impact of fatigue on work productivity and daily activities was our next outcome of interest. The presence of fatigue did not affect any difference in rates of employment at enrollment or after 6 months of follow-up in patients with UC or CD. However, in employed patients with CD with high fatigue, there was a 2-4-fold increase in absenteeism, presenteeism, work productivity loss, and activity impairment at enrollment and this impairment continued to be seen at 6 months with 2-3-fold increased presenteeism, and lost WPAI. At 6 months, there was no longer significant absenteeism. On the other hand, for employed patients with UC with high fatigue there was a 2-3-fold increase in absenteeism and activity impairment at enrollment compared to the patients with low fatigue, with these activity impairments continuing to be 2-fold increased at 6 months of follow-up. The other measures were numerically higher for those with fatigue at each of the time points, but not significant after adjustment. In the subgroup analysis of only patients in remission, there remained 2-4-fold increased rates of presenteeism, lost work productivity, and activity impairment in those with CD and a 2.5-fold increased rate of activity impairment in UC. There have been several studies published with only one in a US population that has looked at the impact of IBD on work productivity.26 This single-center study of 440 patients found that patients with IBD had more presenteeism but not absenteeism as compared to controls without IBD even when patients did not have active IBD. While fatigue was not a primary focus of the study, when queried about the symptoms limiting work productivity, patients commonly (42%) reported fatigue as a limiting symptom. A multicenter study from the Netherlands found that after correcting for disease activity, fatigue was a strong determinant for work productivity loss in patients with IBD.27
Strengths of this study include the use of the CorEvitas IBD Registry given its large size, rigorous design and implementation, prospective data collection, and inclusion of important granular details that are not available in administrative databases. Further, it has wide generalizability given the inclusion of patients from 100 unique sites in 30 states across the United States who are cared for by over 260 gastroenterologists. A potential limitation of the study is the inclusion criteria for the Registry of having recently started a new advanced therapy for IBD. It is possible this may be a group biased to higher rates of active disease, and therefore our findings cannot be generalized to patients who are not using these therapies. However, our subgroup analysis of patients in remission at enrollment helps bring more clarity to the impact of fatigue across the spectrum of active and inactive disease. Also of note, the visual analog scale (VAS) scales for work productivity and ability to do regular activities used in the WPAI questionnaire for our study ranged from 0 to 100 versus 0 to 10 in the licensed WPAI questionnaire that is validated by Reilly et al.22 This difference will not impact the results of our study quantitatively because we divided the VAS scores for these questions by 10. There is a possibility of measurement error introduced in our study due to the use of a version that is slightly different than the validated version. However, we do not expect that the VAS scale would have been interpreted differently by patients in the different comparator groups, solely due to the use of a 100-point versus a 10-point VAS scale in the questionnaire. The WPAI captures data only for employed patients, and thus the results cannot be generalized to patients who are not employed for pay. Last, there is the possible limitation of recall bias in collecting data on healthcare utilization, however, we would expect this bias to be similar across the groups with low versus high fatigue.
In conclusion, our study demonstrates that in a large, multicenter US population of patients with IBD, fatigue is highly prevalent and persists over time. Moreover, fatigue is associated with an increase in HCU, specifically ER visits and hospitalizations, only in the setting of concomitantly active disease. On the other hand, fatigue is associated with a negative impact on work productivity in the setting of both active and inactive diseases. These findings highlight the necessity for clinicians to screen for fatigue and comorbidities12 and offer promising psychological interventions13,14,28 alongside medications. Further studies into understanding the best treatment strategies for patients with IBD with fatigue, especially those without associated active disease, are needed.
Contributor Information
Linda A Feagins, Division of Gastroenterology and Hepatology, Dell Medical School at the University of Texas at Austin, Austin, TX, USA.
Page Moore, CorEvitas LLC, Waltham, MA, USA.
Margaux M Crabtree, CorEvitas LLC, Waltham, MA, USA.
Melissa Eliot, CorEvitas LLC, Waltham, MA, USA.
Celeste A Lemay, CorEvitas LLC, Waltham, MA, USA.
Anita M Loughlin, CorEvitas LLC, Waltham, MA, USA.
Jill K J Gaidos, Section of Digestive Diseases, Yale University School of Medicine, New Haven, CT, USA.
Author Contributions
LAF: study design, interpretation of the data, drafting, and revision of the manuscript, final approval of the manuscript; PM: acquisition of data, study design, interpretation of data, review, and revision of the manuscript, final approval of the manuscript; MMC: acquisition of data, study design, analysis of the data, interpretation of data, review, and revision of the manuscript, final approval of the manuscript; ME: analysis of the data, interpretation of data, review and revision of the manuscript, final approval of the manuscript; CAL: interpretation of data, final approval of the manuscript; AML: study design, interpretation of data, final approval of the manuscript; JKJG: study design, interpretation of the data, review and revision of the manuscript, final approval of the manuscript.
Funding
This study was sponsored by CorEvitas, LLC. Access to study data was limited to CorEvitas and CorEvitas statisticians completed all the analysis; all authors contributed to the interpretation of the results. CorEvitas has been supported through contracted subscriptions in the last two years by AbbVie, Amgen, Inc., Arena, Boehringer Ingelheim, Bristol Myers Squibb, Chugai, Eli Lilly and Company, Genentech, GSK, Janssen Pharmaceuticals, Inc., LEO Pharma, Novartis, Ortho Dermatologics, Pfizer, Inc., Sun Pharmaceutical Industries Ltd., and UCB S.A.
Conflicts of Interest
LAF: Research Support-Takeda, CorEvitas LLC; PM: Employee-CorEvitas LLC; MMC: Employee-CorEvitas LLC; ME: Employee-CorEvitas LLC; CAL: Employee-CorEvitas LLC; AML: Employee-CorEvitas LLC; JKJG: Consulting- MRM Health; Speaker Bureau—AbbVie.
Data Availability
Data are available from CorEvitas, LLC through a commercial subscription agreement and are not publicly available. No additional data are available from the authors.
References
- 1. Fouad AM, Waheed A, Gamal A, Amer SA, Abdellah RF, Shebl FM.. Effect of chronic diseases on work productivity: a propensity score analysis. J Occup Environ Med. 2017;59(5):480-485. doi: https://doi.org/ 10.1097/JOM.0000000000000981 [DOI] [PubMed] [Google Scholar]
- 2. Enns MW, Bernstein CN, Kroeker K, et al. ; CIHR Team in Defining the Burden and Managing the Effects of Psychiatric Comorbidity in Chronic Immunoinflammatory Disease. The association of fatigue, pain, depression and anxiety with work and activity impairment in immune mediated inflammatory diseases. PLoS One. 2018;13(6):e0198975. doi: https://doi.org/ 10.1371/journal.pone.0198975 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Romkens TE, van Vugt-van Pinxteren MW, Nagengast FM, van Oijen MGH, de Jong DJ.. High prevalence of fatigue in inflammatory bowel disease: a case control study. J Crohns Colitis. 2011;5(4):332-337. doi: https://doi.org/ 10.1016/j.crohns.2011.02.008 [DOI] [PubMed] [Google Scholar]
- 4. Schreiner P, Rossel JB, Biedermann L, et al. ; Swiss IBD Cohort Study Group. Fatigue in inflammatory bowel disease and its impact on daily activities. Aliment Pharmacol Ther. 2021;53(1):138-149. doi: https://doi.org/ 10.1111/apt.16145 [DOI] [PubMed] [Google Scholar]
- 5. Borren NZ, van der Woude CJ, Ananthakrishnan AN.. Fatigue in IBD: epidemiology, pathophysiology and management. Nat Rev Gastroenterol Hepatol. 2019;16(4):247-259. doi: https://doi.org/ 10.1038/s41575-018-0091-9 [DOI] [PubMed] [Google Scholar]
- 6. Borren NZ, Tan W, Colizzo FP, et al. Longitudinal trajectory of fatigue with initiation of biologic therapy in inflammatory bowel diseases: a prospective cohort study. J Crohns Colitis. 2020;14(3):309-315. doi: https://doi.org/ 10.1093/ecco-jcc/jjz148 [DOI] [PubMed] [Google Scholar]
- 7. Chavarria C, Casanova MJ, Chaparro M, et al. Prevalence and factors associated with fatigue in patients with inflammatory bowel disease: a multicentre study. J Crohns Colitis. 2019;13(8):996-1002. doi: https://doi.org/ 10.1093/ecco-jcc/jjz024 [DOI] [PubMed] [Google Scholar]
- 8. Holten KIA, Bernklev T, Opheim R, et al. Fatigue in patients with newly diagnosed inflammatory bowel disease: results from a prospective inception cohort, the IBSEN III study. J Crohns Colitis. 2023;17(11):1781-1790. doi: https://doi.org/ 10.1093/ecco-jcc/jjad094 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Bogale K, Maheshwari P, Kang M, et al. Symptoms associated with healthcare resource utilization in the setting of inflammatory bowel disease. Sci Rep. 2022;12(1):10577. doi: https://doi.org/ 10.1038/s41598-022-14838-y [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Ananthakrishnan AN, Desai R, Lee WJ, Griffith J, Chen N, Loftus EV.. Economic burden of fatigue in inflammatory bowel disease. Crohns Colitis 360. 2023;5(3):otad020. doi: https://doi.org/ 10.1093/crocol/otad020 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Singh S, Qian AS, Nguyen NH, et al. Trends in U.S. health care spending on inflammatory bowel diseases, 1996-2016. Inflamm Bowel Dis. 2022;28(3):364-372. doi: https://doi.org/ 10.1093/ibd/izab074 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Kreijne JE, Lie MR, Vogelaar L, van der Woude CJ.. Practical guideline for fatigue management in inflammatory bowel disease. J Crohns Colitis. 2016;10(1):105-111. doi: https://doi.org/ 10.1093/ecco-jcc/jjv168 [DOI] [PubMed] [Google Scholar]
- 13. Emerson C, Barhoun P, Olive L, et al. A systematic review of psychological treatments to manage fatigue in patients with inflammatory bowel disease. J Psychosom Res. 2021;147:110524. doi: https://doi.org/ 10.1016/j.jpsychores.2021.110524 [DOI] [PubMed] [Google Scholar]
- 14. Emerson C, Skvarc D, Mikocka-Walus A, Olive L, Gibson PR, Fuller-Tyszkiewicz M.. People with inflammatory bowel disease prefer cognitive behavioral therapy for fatigue management: a conjoint analysis. Dig Dis Sci. 2024;69(7):2345-2353. doi: https://doi.org/ 10.1007/s10620-024-08468-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Farrell D, Artom M, Czuber-Dochan W, Jelsness-Jørgensen LP, Norton C, Savage E.. Interventions for fatigue in inflammatory bowel disease. Cochrane Database Syst Rev. 2020;4(4):CD012005. doi: https://doi.org/ 10.1002/14651858.CD012005.pub2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Cella D, Choi SW, Condon DM, et al. PROMIS((R)) adult health profiles: efficient short-form measures of seven health domains. Value Health. 2019;22(5):537-544. doi: https://doi.org/ 10.1016/j.jval.2019.02.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Davis SP, Chen DG, Crane PB, Bolin LP, Johnson LA, Long MD.. Influencing factors of inflammatory bowel disease-fatigue: a path analysis model. Nurs Res. 2021;70(4):256-265. doi: https://doi.org/ 10.1097/NNR.0000000000000517 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Bartlett SJ, Gutierrez AK, Andersen KM, et al. Identifying minimal and meaningful change in a patient-reported outcomes measurement information system for rheumatoid arthritis: use of multiple methods and perspectives. Arthritis Care Res (Hoboken). 2022;74(4):588-597. doi: https://doi.org/ 10.1002/acr.24501 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Norman GR, Sloan JA, Wyrwich KW.. Interpretation of changes in health-related quality of life: the remarkable universality of half a standard deviation. Med Care. 2003;41(5):582-592. doi: https://doi.org/ 10.1097/01.MLR.0000062554.74615.4C [DOI] [PubMed] [Google Scholar]
- 20. Lewis JD, Chuai S, Nessel L, Lichtenstein GR, Aberra FN, Ellenberg JH.. Use of the noninvasive components of the Mayo score to assess clinical response in ulcerative colitis. Inflamm Bowel Dis. 2008;14(12):1660-1666. doi: https://doi.org/ 10.1002/ibd.20520 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Harvey RF, Bradshaw JM.. A simple index of Crohn’s-disease activity. Lancet. 1980;1(8167):514. doi: https://doi.org/ 10.1016/s0140-6736(80)92767-1 [DOI] [PubMed] [Google Scholar]
- 22. Reilly MC, Zbrozek AS, Dukes EM.. The validity and reproducibility of a work productivity and activity impairment instrument. PharmacoEcon. 1993;4(5):353-365. doi: https://doi.org/ 10.2165/00019053-199304050-00006 [DOI] [PubMed] [Google Scholar]
- 23. Reilly MC, Gerlier L, Brabant Y, Brown M.. Validity, reliability, and responsiveness of the work productivity and activity impairment questionnaire in Crohn’s disease. Clin Ther. 2008;30(2):393-404. doi: https://doi.org/ 10.1016/j.clinthera.2008.02.016 [DOI] [PubMed] [Google Scholar]
- 24. D’Silva A, Fox DE, Nasser Y, et al. Prevalence and risk factors for fatigue in adults with inflammatory bowel disease: a systematic review with meta-analysis. Clin Gastroenterol Hepatol. 2022;20(5):995-1009.e7. doi: https://doi.org/ 10.1016/j.cgh.2021.06.034 [DOI] [PubMed] [Google Scholar]
- 25. Walker EA, Katon WJ, Jemelka RP.. Psychiatric disorders and medical care utilization among people in the general population who report fatigue. J Gen Intern Med. 1993;8(8):436-440. doi: https://doi.org/ 10.1007/BF02599621 [DOI] [PubMed] [Google Scholar]
- 26. Zand A, van Deen WK, Inserra EK, et al. Presenteeism in inflammatory bowel diseases: a hidden problem with significant economic impact. Inflamm Bowel Dis. 2015;21(7):1623-1630. doi: https://doi.org/ 10.1097/MIB.0000000000000399 [DOI] [PubMed] [Google Scholar]
- 27. van Gennep S, Gielen ME, Rietdijk ST, et al. Work productivity loss is determined by fatigue and reduced quality of life in employed inflammatory bowel disease patients: a prospective multicentre cohort study. Eur J Gastroenterol Hepatol. 2021;33(1S Suppl 1):e594-e602. doi: https://doi.org/ 10.1097/MEG.0000000000002178 [DOI] [PubMed] [Google Scholar]
- 28. Regev S, Schwartz D, Sarid O, et al. Randomised clinical trial: psychological intervention improves work productivity and daily activity by reducing abdominal pain and fatigue in Crohn’s disease. Aliment Pharmacol Ther. 2023;57(8):861-871. doi: https://doi.org/ 10.1111/apt.17399 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
Data are available from CorEvitas, LLC through a commercial subscription agreement and are not publicly available. No additional data are available from the authors.