Abstract
Background & Aims:
Patients with inflammatory bowel diseases (IBD) have a high risk of venous thromboembolism (VTE). We assess the timing and risk factors associated with readmission to the hospital for VTE among patients with IBD.
Methods:
We collected data from the Nationwide Readmissions Database on IBD index admissions resulting in readmission to the hospital for VTE within 60 days, from 2010 through 2014. We used univariable and multivariable regression to assess risk factors associated with VTE readmission with unadjusted (RR) and adjusted RR (aRR) as measures of effect. Time to VTE readmission was assessed in 10-day intervals up to 90 days.
Results:
We identified 872,122 index admissions of patients with IBD; 1160 resulted in readmission with VTE. More than 90% of readmissions occurred within 60 days of discharge from the index admission. Factors associated with hospital readmission with VTE included prior VTE, longer length of hospital stay, comorbidities, having a flexible sigmoidoscopy or colonoscopy at index admission, and age older than 18 years. Additional risk factors included Clostridium difficile infection at index admission (aRR, 1.47; 95% CI, 1.17–1.85) and discharge to a skilled nursing facility or intermediate care facility (aRR, 1.39; 95%CI 1.14–1.70) or discharge with home health services (aRR, 1.65; 95% CI, 1.41–1.94).
Conclusions:
Among patients admitted to the hospital with IBD, most readmissions with VTE occur within 60 days of discharge. Readmission with VTE is associated C difficile infection and discharge to a skilled nursing facility, intermediate care facility, or with home health services. Studies are needed to evaluate the potential benefit of extending VTE prophylaxis for patients admitted to the hospital with IBD for up to 2 months after discharge, to minimize risk.
Keywords: NRD, SNF, Ulcerative Colitis; Crohn’s Disease; Pulmonary Embolism
INTRODUCTION
Inflammatory bowel disease (IBD), comprising of ulcerative colitis (UC) and Crohn’s disease (CD), is an immune mediated disease of the bowel. The prevalence of IBD in North America exceeds 0.5% and is forecasted to rise over the next decade.1–3 Most patients with IBD require lifelong medication to treat disease activity, and when these drugs fail, hospitalization and surgery often follow. Consequently, the complications of IBD are among the most significant burden to the patient and their healthcare system.
Venous thromboembolism (VTE) is one of the most common and life threatening complications for patients with IBD.4 Due to ongoing inflammation, IBD has been shown to be an independent risk factor for the development of a VTE.5 Studies have suggested that patients with IBD have greater than two times the risk of VTE as compared to the general population, and an even higher risk when hospitalized or in the setting of a flare.6,7
Given the significant morbidity and mortality associated with VTE in IBD patients, emphasis has been placed on adherence to VTE prophylaxis guidelines during hospitalization.8–11 In addition to events during hospitalizations, surgical research supports that increased risk for VTE may extend after hospital discharge, and that continued prophylaxis may decrease risk of a VTE event including for the IBD population.12–14 In accordance with this, recent data on VTE rates during and after hospitalization in IBD patients showed that the absolute risk of VTE after major surgery remained elevated for up to 6-weeks post-discharge.15 Data examining patients hospitalized with medical illnesses demonstrated a similar increase in VTE risk post-discharge.16 Additionally, given the safety of pharmacologic prophylaxis in IBD patients,9 there may be a role for continued prophylaxis for those at highest risk for VTE post-discharge.
To date, there has been limited population data examining both the temporality and risk factors related to a post-discharge VTE event in IBD patients. From this study, we aimed to define which IBD patients would be most likely to benefit from post-discharge prophylaxis, as well as the optimal timeline for prophylaxis.
METHODS
Data Source:
We conducted a retrospective cohort study using 2010-2014 data from the Nationwide Readmissions Database (NRD). The NRD is an all-payer database compiled annually by the Healthcare Cost and Utilization Project (HCUP), containing state-level patient information with the ability to track patients across hospitalizations within a single state during an individual year. As of 2014, the NRD contains data relating to approximately 15 million discharges per year, from 22 states and 2,048 hospitals.17 The NRD contains a set of weights in the database that allow for national estimations, and when applied, are representative of approximately 35 million discharges annually. NRD contains patient information from a variety of hospital settings including public, community, and academic health centers (excluding the VA system), and has been utilized to evaluate inpatient readmissions.18–20 Given the de-identified and publically available nature of the NRD, the Institutional Review Boards of Columbia University Medical Center and the University of Southern California deemed this work exempt from review. Per HCUP Data User Agreement, cell sizes <11 were not reported and suppressed.
Study Population:
Patients in the database with International Classification of Diseases, Ninth Edition, Clinical Modification (ICD-9-CM) diagnosis codes for ulcerative colitis (UC: 556.x) or Crohn’s disease (CD: 555.x) during the index hospital admission were classified as having IBD.19 Exclusion criteria included acute VTE or inpatient mortality during index admission. NRD datasets are annually-derived and cannot be linked, thus only index admissions where the discharge occurred from Jan. 1st – Oct. 31st for each year were included (index hospitalizations during November and December were not included because readmissions for the subsequent 60 days could not be fully ascertained).
Outcomes Measured:
Our primary outcome of interest was 60-day risk for VTE readmission from index admission discharge. Readmissions complicated by acute VTE including both deep venous thrombosis (DVT) and pulmonary embolus (PE), were identified based on a new hospitalization with a primary diagnosis of VTE (Appendix).21,22 We only included readmissions that had a primary diagnosis of VTE so as to eliminate instances in which a VTE was acquired during the readmission.
For IBD patients that never experienced an acute VTE readmission, index admission was defined as their first hospitalization of the year. For those IBD patients who did have an acute VTE readmission, the first hospitalization within the preceding 60 days was utilized as the index hospitalization; for most cases, this ended up being the first hospitalization of the year as well. If one patient had several VTE readmissions within a year, only the first index admission leading to a VTE readmission was included in our analysis.
Patient and Hospital Characteristics:
The NRD contains information regarding patient and hospital factors that were analyzed in this study during the index admission. Demographic patient factors included patient age, gender, payer type, and median income quartile by ZIP code. Patient comorbidity status was calculated using the NRD included Agency for Healthcare Research and Quality comorbidity index, which includes comorbidities used in the Elixhauser Comorbidity Index such as heart failure and renal disease.17 Patients were then categorized based on number of underlying comorbid conditions (0, 1, >1 conditions). The presence of an operating room procedure on index admission was stratified by type of surgery. Patients who underwent a small or large bowel repair or resection, anorectal manipulation, or fistula surgery were included in the intestinal surgery group (Appendix),23 whereas all other patients who had an operation on index admission were characterized as having non-intestinal surgery. Additional clinical factors not already provided in the National Readmissions Database are represented with their corresponding ICD-9CM codes in the Appendix. Hospital-based characteristics that were captured included presence of a flexible sigmoidoscopy/colonoscopy during index admission, length of stay, and discharge disposition. Hospital factors included hospital bed size and hospital teaching status.
Data Analysis
Demographic, patient, and hospital factors were obtained from index admission data, and weighted to reflect national estimates based upon conversion factors provided by the NRD. We used univariable and log-linear multivariable regression models with unadjusted (RR) and adjusted (aRR) risk ratios to assess the relationship between the primary outcome and patient, hospital, and clinical factors. To evaluate temporality of VTE readmissions, we assessed the time from the discharge immediately preceding the VTE readmission as a categorical variable stratified in 10-day intervals for up to 90 days. In order to capture all VTE readmissions within 90 days of discharge, only discharges from Jan. 1st – Sept. 30th of 2010-2014 were included in this part of the analysis, as this allows for the full 90 days post-discharge to be observed. Mean index admission costs were calculated using the NRD charge to cost converter, and adjusted for inflation to represent costs in 2016 dollars. Stratified analysis based upon IBD subgroup (UC or CD) was also conducted. All statistical analyses were performed using SAS version 9.4 (Cary, NC).
Sensitivity Analysis
We conducted four separate sensitivity analyses. First, in order to account for the possible misclassification of chronic VTE, we repeated the analyses excluding those with a prior history of VTE. Second, in order to account for the small subset (12%) of patients who had a readmission in between an initial index admission and a subsequent VTE readmission, we performed a separate analysis eliminating all patients with an interval non-VTE related admission. Third, given that many patients with a secondary readmission diagnosis of VTE likely acquired the VTE prior to admission, we conducted an additional analysis including all patients with either a primary or secondary diagnosis of VTE on readmission. Lastly, in order to characterize risk among patients with high likelihood for active complications from IBD, we performed a sensitivity analysis including only patients whose primary ICD-9-CM code on index admission corresponded to an IBD diagnosis.
RESULTS
Baseline Characteristics
From Jan.-Oct. of 2010-2014, there were a total of 872,122 index admissions in IBD patients with a slightly higher proportion of women (57%) than men (43%), and a relatively uniform distribution of individuals in each age group from 18-80 years of age (Table 1). Approximately 4% of the entire cohort had C. difficile infection on index admission, 4% had a prior VTE, 7.3% had an intestinal surgery on index admission, and 3% had a flexible sigmoidoscopy or colonoscopy on index admission, which we included as a surrogate for a disease-related exacerbation.
Table 1:
Univariable analysis of index admission demographics, medical, and hospital related factors in patients with IBD.
| VTE Readmission N [%] | No VTE Readmission N [%] | Risk Ratio | 95% Confidence Interval | p-value | |
|---|---|---|---|---|---|
| Age (years) at time of index admission | |||||
| < 18 | 19 [0.05%] | 39,676 [99.95%] | 1.00 | (Reference) | |
| 18-30 | 95 [0.07%] | 129,299 [99.93%] | 1.56 | 0.95 – 2.55 | 0.08 |
| 31-40 | 122 [0.10%] | 123,361 [99.90%] | 2.10 | 1.29 – 3.42 | <0.01 |
| 41-50 | 125 [0.10%] | 127,449[99.90%] | 2.08 | 1.28 – 3.37 | <0.01 |
| 51-65 | 376 [0.18%] | 212,869 [99.82%] | 3.74 | 2.35 – 5.94 | <0.01 |
| 66-80 | 322 [0.19%] | 168,474 [99.81%] | 4.04 | 2.54 – 6.44 | <0.01 |
| >80 | 101 [0.14%] | 69,834 [99.86%] | 3.06 | 1.87 – 5.02 | <0.01 |
| Gender | |||||
| Male | 554 [0.15%] | 372,964 [99.85%] | 1.00 | (Reference) | |
| Female | 606 [0.12%] | 497,998 [99.88%] | 0.82 | 0.73 – 0.92 | <0.01 |
| Median household Zip code income quartile | |||||
| Quartile 4 (highest) | 283 [0.13%] | 224,927 [99.87%] | 1.00 | (Reference) | |
| Quartile 3 | 313 [0.14%] | 219,433 [99.86%] | 1.13 | 0.97 – 1.33 | 0.12 |
| Quartile 2 | 286 [0.14%] | 210,342 [99.86%] | 1.08 | 0.92 – 1.28 | 0.34 |
| Quartile 1 (lowest) | 268 [0.13%] | 203,406 [99.87%] | 1.05 | 0.89 – 1.24 | 0.57 |
| Missing | 9 [0.07%] | 12,854 [99.93%] | 0.56 | 0.29 – 1.09 | 0.09 |
| Insurance status | |||||
| Private | 442 [0.12%] | 369,302 [99.88%] | 1.00 | (Reference) | |
| Medicare | 532 [0.17%] | 319,632 [99.83%] | 1.39 | 1.23 – 1.58 | <0.01 |
| Medicaid | 121 [0.12%] | 100,981 [99.88%] | 1.00 | 0.82 – 1.23 | 0.98 |
| Self-pay | 26 [0.06%] | 41,044 [99.94%] | 0.53 | 0.35 – 0.78 | <0.01 |
| No charge | 4 [0.07%] | 5,645 [99.93%] | 0.58 | 0.21 – 1.57 | 0.29 |
| Other | 34 [0.10%] | 32,647 [99.9%] | 0.88 | 0.62 – 1.24 | 0.45 |
| Unknown | 0 [0%] | 1,711 [100%] | - | - | - |
| Tobacco use | |||||
| No | 1,018 [0.14%] | 743,762 [99.86%] | 1.00 | (Reference) | |
| Yes | 141 [0.11%] | 127,200 [99.89%] | 0.81 | 0.68 – 0.97 | 0.02 |
| Malignancy | |||||
| No | 1,110 [0.13%] | 860,259 [99.87%] | 1.00 | (Reference) | |
| Yes | 49 [0.45%] | 10,703 [99.55%] | 1.69 | 1.27 – 2.25 | <0.01 |
| Clostridium difficile on index admission | |||||
| No | 1,078 [0.13%] | 837,659 [98.87%] | 1.00 | (Reference) | |
| Yes | 81 [0.24%] | 33,303 [99.76%] | 1.90 | 1.51 – 2.38 | <0.01 |
| Comorbidities | |||||
| None | 113 [0.07%] | 173,626 [99.93%] | 1.00 | (Reference) | |
| One | 229 [0.11%] | 206,124 [99.89%] | 1.71 | 1.36 – 2.14 | <0.01 |
| ≥2 | 818 [0.17%] | 491,212 [99.83%] | 2.57 | 2.11 – 3.12 | <0.01 |
| History of VTE | |||||
| No | 1,035 [0.12%] | 836,352 [99.88%] | 1.00 | (Reference) | |
| Yes | 124 [0.36%] | 34,610 [99.64%] | 2.89 | 2.40 – 3.48 | <0.01 |
| Operating room procedure during index hospitalization | |||||
| None | 865 [0.13%] | 644,210 [99.87%] | 1.00 | (Reference) | |
| Intestinal surgery | 113 [0.18%] | 63,451 [99.82] | 1.32 | 1.09 – 1.61 | <0.01 |
| Non-intestinal surgery | 182 [0.11%] | 163,301 [99.89%] | 0.83 | 0.71 – 0.97 | 0.02 |
| Hospital teaching status | |||||
| Metropolitan nonteaching | 415 [0.14%] | 302,535 [99.86%] | 1.00 | (Reference) | |
| Metropolitan teaching | 676 [0.14%] | 489,814 [99.86%] | 1.01 | 0.89 – 1.14 | 0.90 |
| Nonmetropolitan | 69 [0.09%] | 78,613 [99.91%] | 0.64 | 0.50 – 0.83 | <0.01 |
| Hospital bed size | |||||
| Small | 121 [0.11%] | 109,704 [99.89%] | 1.00 | (Reference) | |
| Medium | 268 [0.13%] | 204,486 [99.87%] | 1.19 | 0.96 – 1.48 | 0.11 |
| Large | 771 [0.14%] | 556,772 [99.86%] | 1.26 | 1.04 – 1.52 | 0.02 |
| Flex Sig/Colonoscopy on index admission | |||||
| No | 940 [0.11%] | 848,368 [99.89%] | 1.00 | (Reference) | |
| Yes | 220 [0.96%] | 22,594 [99.04%] | 1.29 | 1.12 – 1.50 | <0.01 |
| Discharge status | |||||
| Routine Discharge | 747 [0.11%] | 673,098 [99.89%] | 1.00 | (Reference) | |
| Transfer to short-term hospital | 8 [0.11%] | 7,106 [99.89%] | 1.03 | 0.52 – 2.06 | 0.93 |
| Transfer to SNF, ICF, or other | 154 [0.21%] | 74,981 [99.79%] | 1.85 | 1.56 – 2.20 | <0.01 |
| Home health care | 240 [0.23%] | 105,229 [99.77%] | 2.05 | 1.78 – 2.38 | <0.01 |
| Against medical advice | 10 [0.10%] | 9,899 [99.90%] | 0.90 | 0.48 – 1.69 | 0.75 |
| Unknown | 0 [0%] | 648 [100%] | - | - | - |
| Mean length of index admission | 6.9 days | 5.4 days | 1.01 | 1.01 – 1.02 | <0.01 |
| Cost of index hospitalization* | $15,045.97 | $12,434.35 | - | - | - |
Adjusted for inflation to 2016 dollars
Of the 872,122 index admissions, 1,160 led to a VTE readmission (0.13%). Of those IBD patients who had a VTE readmission, there was a relatively equal proportion of UC (45%) to CD (55%) patients. When comparing mean length of stay, those who had a VTE readmission had a 1.5-day higher mean length of stay on index admission than those who did not.
Readmission Temporality
Risk for VTE readmission was highest the first 10 days after index hospitalization discharge and decreased for each subsequent 10-day period. 91% of 90-day VTE readmissions occurred within the first 60 days of discharge (Figure 1).
Figure 1:
Percentage of total VTE readmissions by number of days following discharge from index hospitalization
Factors Associated with VTE readmission:
When considering demographics using univariable analysis, we found all age groups for patients over 30 to have an increased risk of VTE readmission as compared to those less than 18 years of age, and found that women were at lower risk of VTE readmission (RR 0.82, 95% CI 0.73-0.92) compared to men (Table 1). We additionally found that IBD patients with Medicare were at higher risk for VTE readmission, and that those with self-pay were at lower risk for VTE readmission when compared to IBD patients with private insurance.
When considering hospital related factors, IBD patients seen at a non-metropolitan hospital on index admission had a lower risk of VTE readmission as compared to those cared for at a metropolitan nonteaching hospital. Additionally, being cared for at a large hospital carried a higher risk of VTE readmission in comparison to being cared for at a small hospital.
Patient clinical factors that were associated with the highest risk of VTE readmission on univariable analysis included a prior history of VTE (RR 2.89, 95% CI 2.40-3.48), the presence of two or more comorbidities as compared to none (RR 2.57, 95% CI 2.11-3.12), C. difficile infection on index admission (RR 1.90, 95% CI 1.51-2.38), and discharge to a skilled nursing facility (SNF)/intermediate care facility (ICF)/other (RR 1.85, 95% CI 1.56-2.20) or discharge home with health services (RR 2.05, 95% CI 1.78-2.38) as compared to routine discharge home. Additionally, malignancy, length of stay, and having a flexible sigmoidoscopy/colonoscopy on index admission were also significantly associated with risk of VTE readmission. Of note, though tobacco use was associated with a lower risk of VTE on univariable analysis (Table 1), this was likely due to confounding as it was no longer a significant predictor of VTE (RR 0.89, 95% CI 0.75 – 1.07) when adjusting for all other variables in the model (Table 2).
Table 2:
Multivariable analysis examining factors associated with a VTE readmission in those with an ICD diagnosis of IBD.
| Composite IBD Diagnosis (n=1160) | Ulcerative Colitis (n=522) | Crohn’s disease (n=638) | ||||
|---|---|---|---|---|---|---|
| Variable | Adjusted* Risk Ratio, (95% CI) | p-value | Adjusted* Risk Ratio, (95% CI) | p-value | Adjusted* Risk Ratio, (95% CI) | p-value |
| Age in years | ||||||
| < 18 | (Ref.) | (Ref.) | (Ref.) | |||
| 18-30 | 1.74 (1.06-2.86) | 0.03 | 10.80 (2.62-44.49) | <0.01 | 0.75 (0.42-1.33) | 0.33 |
| 31-40 | 2.32 (1.42-3.8) | <0.01 | 7.08 (1.69-29.64) | <0.01 | 1.62 (0.95-2.77) | 0.07 |
| 41-50 | 2.10 (1.28-3.43) | <0.01 | 7.21 (1.73-30.09) | <0.01 | 1.41 (0.82-2.41) | 0.21 |
| 51-65 | 3.40 (2.12-5.47) | <0.01 | 16.91 (4.15-68.88) | <0.01 | 1.82 (1.08-3.05) | 0.02 |
| 66-80 | 3.48 (2.12-5.73) | <0.01 | 15.51 (3.73-64.44) | <0.01 | 2.04 (1.18-3.55) | 0.01 |
| >80 | 2.39 (1.41-4.06) | <0.01 | 12.52 (2.96-53.01) | <0.01 | 1.14 (0.61-2.13) | 0.69 |
| Gender | ||||||
| Female | 0.79 (0.71-0.89) | <0.01 | 0.81 (0.68-0.96) | 0.02 | 0.79 (0.67-0.92) | <0.01 |
| Median household Zip code income quartile | ||||||
| Highest | (Ref.) | (Ref.) | (Ref.) | |||
| High | 1.15 (0.98-1.35) | 0.09 | 1.00 (0.79-1.27) | 0.99 | 1.31 (1.04-1.63) | 0.02 |
| Low | 1.14 (0.96-1.35) | 0.13 | 1.05 (0.83-1.34) | 0.67 | 1.22 (0.97-1.54) | 0.09 |
| Lowest | 1.13 (0.95-1.34) | 0.17 | 1.07 (0.83-1.38) | 0.61 | 1.21 (0.95-1.53) | 0.12 |
| Missing | 0.59 (0.30-1.14) | 0.12 | 0.60 (0.23-1.54) | 0.29 | 0.58 (0.23-1.47) | 0.25 |
| Insurance status | ||||||
| Private | (Ref.) | (Ref.) | (Ref.) | |||
| Medicare | 0.88 (0.74-1.04) | 0.14 | 0.88 (0.67-1.15) | 0.36 | 0.89 (0.71-1.11) | 0.30 |
| Medicaid | 1.06 (0.86-1.30) | 0.60 | 0.85 (0.60-1.21) | 0.35 | 1.25 (0.96-1.62) | 0.10 |
| Self-pay | 0.58 (0.39-0.86) | <0.01 | 0.24 (0.10-0.61) | <0.01 | 0.85 (0.54-1.34) | 0.49 |
| No charge | 0.59 (0.22-1.61) | 0.31 | - | - | 1.12 (0.41-3.04) | 0.83 |
| Other | 0.86 (0.61-1.22) | 0.40 | 0.49 (0.26-0.94) | 0.03 | 1.23 (0.81-1.87) | 0.33 |
| Tobacco use | ||||||
| Yes | 0.89 (0.75-1.07) | 0.23 | 0.58 (0.39-0.86) | <0.01 | 1.06 (0.86-1.31) | 0.59 |
| Malignancy | ||||||
| Yes | 1.17 (0.88-1.57) | 0.27 | 0.67 (0.41-1.11) | 0.12 | 1.78 (1.25-2.54) | <0.01 |
| Clostridium difficile | ||||||
| Yes | 1.47 (1.17-1.85) | <0.01 | 1.50 (1.12-2.00) | <0.01 | 1.49 (0.99-2.24) | 0.05 |
| Comorbidities | ||||||
| None | (Ref.) | (Ref.) | (Ref.) | |||
| One | 1.49 (1.18-1.87) | <0.01 | 0.72 (0.52-1.00) | 0.05 | 2.87 (2.03-4.06) | <0.01 |
| ≥2 | 1.78 (1.44-2.20) | <0.01 | 1.08 (0.82-1.43) | 0.58 | 2.93 (2.10-4.10) | <0.01 |
| History of VTE | ||||||
| Yes | 2.41 (1.99-2.90) | <0.01 | 2.01 (1.49-2.71) | <0.01 | 2.66 (2.09-3.39) | <0.01 |
| Operating room procedure during index admission | ||||||
| None | (Ref.) | (Ref.) | (Ref.) | |||
| Intestinal surgery | 1.13 (0.92-1.39) | 0.26 | 1.31 (0.97-1.76) | 0.08 | 0.99 (0.73-1.34) | 0.96 |
| Non-intestinal surgery | 0.74 (0.62-0.87) | <0.01 | 0.44 (0.32-0.59) | <0.01 | 0.99 (0.81-1.21) | 0.92 |
| Hospital teaching | ||||||
| Metro nonteaching | (Ref.) | (Ref.) | (Ref.) | |||
| Metro teaching | 1.05 (0.93-1.19) | 0.45 | 0.99 (0.82-1.19) | 0.90 | 1.08 (0.92-1.28) | 0.34 |
| Nonmetropolitan | 0.64 (0.50-0.84) | <0.01 | 0.64 (0.43-0.95) | 0.03 | 0.64 (0.46-0.91) | 0.01 |
| Flex Sig/Colonoscopy | ||||||
| Yes | 1.35 (1.16-1.57) | <0.01 | 1.24 (1.01, 1.53) | 0.04 | 1.40 (1.12-1.75) | <0.01 |
| Discharge status | ||||||
| Routine Discharge | (Ref.) | (Ref.) | (Ref.) | |||
| Short-term hospital | 0.91 (0.46-1.82) | 0.79 | 0.81 (0.30-2.16) | 0.67 | 0.97 (0.36-2.59) | 0.85 |
| SNF, ICF, or other | 1.39 (1.14-1.70) | <0.01 | 0.82 (0.60-1.12) | 0.21 | 2.13 (1.65-2.73) | <0.01 |
| Home health care | 1.65 (1.41-1.94) | <0.01 | 1.40 (1.11-1.78) | <0.01 | 1.85 (1.49-2.28) | <0.01 |
| Against medical advice | 1.02 (0.54-1.92) | 0.95 | 0.60 (0.14-2.69) | 0.51 | 1.47 (0.78-2.77) | 0.24 |
| Length of stay | 1.01 (1.0-1.01) | 0.03 | 1.00 (0.99-1.01) | 0.39 | 1.01 (1.0-1.01) | <0.01 |
Adjusted for all variables listed in the table
On multivariable analysis, age remained associated with increased risk for VTE readmission (Table 2). As in the univariable analysis, a history of VTE, length of stay, having a flexible sigmoidoscopy/colonoscopy on index admission, and the presence of comorbidities were also associated with VTE readmission among IBD patients. Additionally, a diagnosis of C. difficile infection during index admission (aRR 1.47, 95% CI 1.17-1.85, p<0.01) and discharge to a SNF/ICF/other (aRR 1.39, 95% CI 1.14-1.70) or home with health services (aRR 1.65, 95% CI 1.41-1.94) as compared to routine discharge home were significant predictors of VTE readmission among those with IBD. Factors associated with a lower risk of VTE readmission on multivariable analysis included female gender and self-pay insurance status as compared to private. Of note, having a non-intestinal surgery on index admission was also associated with a lower risk of a 60-day VTE readmission (aRR 0.74, 95% CI 0.62-2.87), though having intestinal surgery on index admission was not associated with a lower risk of VTE readmission (aRR 1.13, 95% CI 0.92-1.39; Table 2).
In our sensitivity analysis, we found similar results in our multivariable model when excluding all patients with a prior history of VTE (Supplementary Table 1), and when excluding the 142 patients whose first readmission was not for a VTE (Supplementary Table 2). Additionally, when examining patients with either a primary or secondary readmission diagnosis of VTE, the total number of VTE patients doubled to 2,387, and C. difficile infection as well as discharge to a SNF/ICF/other or discharge home with health services remained significant predictors of a VTE readmission (Supplementary Table 3). When only including patients who had a primary ICD code for IBD on index admission, all of these predictors remained significant except for discharge to SNF/ICF/other (Supplementary Table 4).
DISCUSSION
It is estimated that between 300,000 and 600,000 patients are diagnosed with a VTE each year in the U.S.24 In addition to increased healthcare related expenditures of over 11 billion dollars/year, there is also a significant increase in morbidity and mortality that is associated with VTE.25,26 Patients with IBD are known to be at higher risk for VTE than the general population, with this risk extending to the outpatient setting.7,26 Current guidelines focus upon VTE prophylaxis during hospitalization, but due to a paucity of data, there are no clear guidelines addressing post-discharge prophylaxis.27
In this nationwide study of hospitalized IBD patients from Jan.-Oct. of 2010-2014, we observed 1,160 (of 872,122) index admissions leading to a VTE readmission within 60 days. This likely under-represents the true burden of VTE among IBD patients, given that our analysis only captures those who required re-hospitalization, and only those who had a readmission with a primary diagnosis of VTE.
In order to help target future prophylaxis guidelines, we assessed time to VTE readmission in 10-day intervals up to 90 days. In accordance with prior data among hospitalized patients,28 we found that the majority of VTE readmissions occurred soon after discharge, with over 90% occurring within 60 days.
With the knowledge that VTE readmission among IBD patients most often occurred within 60 days of discharge, we then examined factors that were associated with a VTE readmission. From our data, we found that as compared to routine discharge home, discharge to an intermediate or skilled nursing facility was associated with a 39% increase risk in VTE readmission, and discharge home with services was associated with a 65% increase in VTE readmission. This is in accordance with prior data suggesting an increased risk of VTE among nursing home residents and those in long-term care facilities.29,30
Several factors known to increase VTE risk in the general population, such as advanced age and the presence of comorbidities, were also associated with an increased VTE readmission risk among hospitalized IBD patients.31,32 Elderly patients may be especially vulnerable, as they are more likely to have multiple comorbidities and require discharge to nursing facilities, both independent predictors of VTE readmission in this study. We expect that these results will become increasingly relevant, as the prevalence of IBD continues to rise in the elderly population.2
Additionally, in accordance with prior literature, having a disease related exacerbation, which we assessed using the presence of flexible sigmoidoscopy or colonoscopy on index admission given NRD’s lack of symptom and laboratory information, was also associated with a higher risk of VTE readmission.5–7 We found that IBD patients who underwent a non-intestinal surgery had a lower risk of VTE as compared to those who did not undergo any surgery. We suspect this is because IBD patients who did not undergo surgery likely include a group of patients who were hospitalized for a disease flare (a known risk factor for VTE),7,11,27 whereas those who were admitted and had non-intestinal surgeries likely include patients who were hospitalized for non-IBD related reasons (eg: orthopedic procedures).
C. difficile infection represented an additional risk factor that was significantly associated with VTE readmission risk among IBD patients. Specifically, we found that having C. difficile on index admission was associated with a 47% increase in the risk of VTE readmission when adjusting for all other factors. Analogous results were seen in a prior study comparing the odds of VTE among hospitalized IBD patients with C. difficile as compared to those without.33 Also supporting this association is a case-cross over study from Rogers et al., in which infection was found to be the most common risk factor for VTE related admissions among 16,781 patients studied.34
One of the major strengths of our study is the ability to review over 15 million discharges across the U.S. annually using the Nationwide Readmissions Database. However, this study also has limitations. Although our outcomes and covariates are based upon accepted ICD classifications within the literature, their accuracy is unable to be verified via chart review. Additionally, we are unable to address medication use such as oral contraceptives, use of VTE prophylaxis during hospitalization, disease severity, family history, and other IBD specific disease characteristics that may influence VTE risk, though a recent study by McCurdy et al. found no association between VTE prophylaxis during hospitalization and a post-discharge VTE event.31 Lastly, given that the NRD dataset tracks individual patient admissions for only one year, it is possible that a given patient, if admitted and readmitted multiple times in this study period, may be counted more than once. However, since we only considered a first VTE readmission per patient, this could occur a maximum of five times in the unlikely event that the same patient was admitted and then readmitted with a VTE in each of the five years of the study period.
In conclusion, based upon our retrospective nationwide study, we found the risk of VTE readmission among IBD patients to be highest within the first 60 days of hospital discharge, and associated with risk factors such as older age, presence of comorbidities, C. difficile infection and discharge to a SNF/ICF or home with health services. Given increased thrombotic risk postdischarge,31 as well as overall safety of VTE prophylaxis,9 extending prophylaxis for those at highest risk may have significant benefits. This, however, must be weighed against the risk of polypharmacy, insurance barriers to implementation, and overall cost, and should be rigorously evaluated through future prospective and cost-effectiveness studies.
Supplementary Material
Need to Know.
Background:
Inflammatory bowel diseases (IBD) are associated with an increased risk of venous thromboembolism (VTE), which carries significant morbidity and mortality. Little is known about the timing or risk factors for readmission to the hospital with a VTE.
Findings:
Among patients admitted to the hospital with IBD, most VTE readmissions occur within 60 days of discharge. Older age, comorbidities, Clostridium difficile infection, and discharge to a nursing or intermediate facility were associated with a higher risk of readmission with VTE.
Implications for Patient Care:
Studies are needed to evaluate the potential benefit of extending VTE prophylaxis for patients admitted to the hospital with IBD for up to 2 months after discharge, to minimize risk.
Acknowledgments
All Grants/Disclosures:
ASF: NIH: T32DK083256
TW: None
AA: Supported in part by grants from (NIH: R03 DK112909, Pfizer, the Chleck Family Foundation, and the Crohn’s and Colitis Foundation). Dr. Ananthakrishnan has also served on scientific advisory boards for Janssen, Takeda, Gilead, and Merck and has received research support from Pfizer.
SL: Supported in part by grants from Jansen, AbbVie, Eli Lilly, Pfizer, Celgene, Gilead. Dr. Lichtiger consults for Jansen, AbbVie, Takeda, and Gilead and receives honoraria as a speaker for AbbVie, Jansen, Takeda, and Pfizer.
GK: Honoraria for speaking or consultancy from Abbvie, Janssen, Pfizer, and Takeda. Dr. Kaplan has received research support from Janssen, Abbvie, GlaxoSmith Kline, Merck, and Shire. He shares ownership of a patent: TREATMENT OF INFLAMMATORY DISORDERS, AUTOIMMUNE DISEASE, AND PBC. UTI Limited Partnership, assignee. Patent 62/555,397. 7 Sept. 2017.
AF: NIH: K08HD082287
GL: Research support from AbbVie, honoraria for speaking from Merck and Pfizer
JDW: Consultant for Tesaro and Clovis oncology
FA: NIH: KL2TR001854
WJM: Consultant to Rebound Therapeutics, Viseon TSP, Medtronic, Penumbra, Stream Biomedical; Investor: Cerebrotech, Endostream, Viseon, Rebound.
BL: Consultant to Takeda Pharmaceuticals
Abbreviations:
- aRR
adjusted Risk Ratio
- CD
Crohn’s Disease
- DVT
Deep Venous Thrombosis
- HCUP
Healthcare Cost and Utilization Project
- IBD
Inflammatory Bowel Disease
- ICD-9-CM
International Classification of Diseases Ninth Edition Clinical Modification
- ICF
Intermediate Care Facility
- NRD
Nationwide Readmissions Database
- PE
Pulmonary Embolus
- SNF
Skilled Nursing Facility
- UC
Ulcerative Colitis
- VTE
Venous Thromboembolism
Appendix:
International Classification of Diseases, ninth revision, Clinical Modification (ICD-9-CM) Codes used in our study.
| Descrintion | ICD 9 CM Codes |
|---|---|
| VTE diagnosis (PE or DVT) | 415.1x, 451.1x, 453.4x, 453.8x, 453.9x, 673.2x, 673.8x, 671.3x, 671.4x, 671.5x |
| History of VTE | V12.51, V12.55 |
| Smoking status | 305.1 |
| Flexible Sigmoidoscopy/Colonoscopy | 45.22, 45.23, 45.24, 45.25, 45.42, and 45.43 |
| Clostridium difficile | 008.45 |
| Intestinal Surgery | 45.61-45.63, 17.3, 45.71–45.79, 45.81–45.83, 48.40–48.49, 48.50–48.59, 48.61–48.69, or 46.74, 46.76, 48.73, 48.93, 49.11, 49.12, 49.73, 57.83, 58.43, or 70.72–70.74 |
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Potential Competing Interests/Conflicts of Interest:
None
REFERENCES
- 1.Ng SC, Shi HY, Hamidi N, et al. Worldwide incidence and prevalence of inflammatory bowel disease in the 21st century: a systematic review of population-based studies. Lancet. 2018;390(10114):2769–2778. [DOI] [PubMed] [Google Scholar]
- 2.Molodecky NA, Soon IS, Rabi DM, et al. Increasing incidence and prevalence of the inflammatory bowel diseases with time, based on systematic review. Gastroenterology. 2012;142(1):46–54 e42,; quiz e30. [DOI] [PubMed] [Google Scholar]
- 3.Coward S, Clement F, Benchimol EI, et al. Past and Future Burden of Inflammatory Bowel Diseases Based on Modeling of Population-based Data. Gastroenterology. 2019. [DOI] [PubMed] [Google Scholar]
- 4.Nguyen GC, Sam J. Rising prevalence of venous thromboembolism and its impact on mortality among hospitalized inflammatory bowel disease patients. Am J Gastroenterol. 2008;103(9):2272–2280. [DOI] [PubMed] [Google Scholar]
- 5.Miehsler W, Reinisch W, Valic E, et al. Is inflammatory bowel disease an independent and disease specific risk factor for thromboembolism? Gut. 2004;53(4):542–548. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Kappelman MD, Horvath-Puho E, Sandler RS, et al. Thromboembolic risk among Danish children and adults with inflammatory bowel diseases: a population-based nationwide study. Gut. 2011;60(7):937–943. [DOI] [PubMed] [Google Scholar]
- 7.Grainge MJ, West J, Card TR. Venous thromboembolism during active disease and remission in inflammatory bowel disease: a cohort study. Lancet. 2010;375(9715):657–663. [DOI] [PubMed] [Google Scholar]
- 8.Benlice C, Holubar SD, Gorgun E, et al. Extended Venous Thromboembolism Prophylaxis After Elective Surgery for IBD Patients: Nomogram-Based Risk Assessment and Prediction from Nationwide Cohort. Diseases of the colon and rectum. 2018;61(10):1170–1179. [DOI] [PubMed] [Google Scholar]
- 9.Ra G, Thanabalan R, Ratneswaran S, Nguyen GC. Predictors and safety of venous thromboembolism prophylaxis among hospitalized inflammatory bowel disease patients. Journal of Crohn’s & colitis. 2013;7(10):e479–485. [DOI] [PubMed] [Google Scholar]
- 10.Tinsley A, Naymagon S, Enomoto LM, Hollenbeak CS, Sands BE, Ullman TA. Rates of pharmacologic venous thromboembolism prophylaxis in hospitalized patients with active ulcerative colitis: results from a tertiary care center. Journal of Crohn’s & colitis. 2013;7(12):e635–640. [DOI] [PubMed] [Google Scholar]
- 11.Papa A, Gerardi V, Marzo M, Felice C, Rapaccini GL, Gasbarrini A. Venous thromboembolism in patients with inflammatory bowel disease: focus on prevention and treatment. World J Gastroenterol. 2014;20(12):3173–3179. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Sweetland S, Green J, Liu B, et al. Duration and magnitude of the postoperative risk of venous thromboembolism in middle aged women: prospective cohort study. BMJ. 2009;339:b4583. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Najjar PA, Madenci AL, Zogg CK, et al. Implementation of a Comprehensive Post-Discharge Venous Thromboembolism Prophylaxis Program for Abdominal and Pelvic Surgery Patients. Journal of the American College of Surgeons. 2016;223(6):804–813. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Gross ME, Vogler SA, Mone MC, Sheng X, Sklow B. The importance of extended postoperative venous thromboembolism prophylaxis in IBD: a National Surgical Quality Improvement Program analysis. Diseases of the colon and rectum. 2014;57(4):482–489. [DOI] [PubMed] [Google Scholar]
- 15.Chu TPC, Grainge MJ, Card TR. The risk of venous thromboembolism during and after hospitalisation in patients with inflammatory bowel disease activity. Aliment Pharmacol Ther. 2018;48(10):1099–1108. [DOI] [PubMed] [Google Scholar]
- 16.Nemeth B, Lijfering WM, Cannegieter SC. Thromboprophylaxis after hospital discharge in acutely ill medical patients: need for trials in patients who are at high risk of venous thrombosis. Journal of thoracic disease. 2017;9(4):950–952. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.AHRQ. Introduction to the HCUP Nationwide Readmissions Database (NRD) from 2010-2014. 2017.
- 18.Nguyen NH, Ohno-Machado L, Sandborn WJ, Singh S. Obesity Is Independently Associated With Higher Annual Burden and Costs of Hospitalization in Patients With Inflammatory Bowel Diseases. Clin Gastroenterol Hepatol. 2018. [DOI] [PubMed] [Google Scholar]
- 19.Barnes EL, Kochar B, Long MD, et al. Modifiable Risk Factors for Hospital Readmission Among Patients with Inflammatory Bowel Disease in a Nationwide Database. Inflammatory bowel diseases. 2017;23(6):875–881. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Agrawal S, Garg L, Shah M, et al. Thirty-Day Readmissions After Left Ventricular Assist Device Implantation in the United States: Insights From the Nationwide Readmissions Database. Circulation Heart failure. 2018;11(3):e004628. [DOI] [PubMed] [Google Scholar]
- 21.Tamariz L, Harkins T, Nair V. A systematic review of validated methods for identifying venous thromboembolism using administrative and claims data. Pharmacoepidemiology and drug safety. 2012;21 Suppl 1:154–162. [DOI] [PubMed] [Google Scholar]
- 22.Wen T, Wright JD, Goffman D, et al. Postpartum venous thromboembolism readmissions in the United States. American journal of obstetrics and gynecology. 2018;219(4):401 e401–401 e414. [DOI] [PubMed] [Google Scholar]
- 23.Malarcher CA, Wheaton AG, Liu Y, et al. Hospitalizations for Crohn’s Disease - United States, 2003-2013. MMWR Morbidity and mortality weekly report. 2017;66(14):377–381. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Beckman MG, Hooper WC, Critchley SE, Ortel TL. Venous thromboembolism: a public health concern. American journal of preventive medicine. 2010;38(4 Suppl):S495–501. [DOI] [PubMed] [Google Scholar]
- 25.LaMori JC, Shoheiber O, Mody SH, Bookhart BK. Inpatient resource use and cost burden of deep vein thrombosis and pulmonary embolism in the United States. Clinical therapeutics. 2015;37(1):62–70. [DOI] [PubMed] [Google Scholar]
- 26.Andrade AR, Barros LL, Azevedo MFC, et al. Risk of thrombosis and mortality in inflammatory bowel disease. Clinical and translational gastroenterology. 2018;9(4):142. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Nguyen GC, Bernstein CN, Bitton A, et al. Consensus statements on the risk, prevention, and treatment of venous thromboembolism in inflammatory bowel disease: Canadian Association of Gastroenterology. Gastroenterology. 2014;146(3):835–848 e836. [DOI] [PubMed] [Google Scholar]
- 28.Amin AN, Varker H, Princic N, Lin J, Thompson S, Johnston S. Duration of venous thromboembolism risk across a continuum in medically ill hospitalized patients. Journal of hospital medicine. 2012;7(3):231–238. [DOI] [PubMed] [Google Scholar]
- 29.Heit JA, Silverstein MD, Mohr DN, Petterson TM, O’Fallon WM, Melton LJ, 3rd. Risk factors for deep vein thrombosis and pulmonary embolism: a population-based case-control study. Arch Intern Med. 2000;160(6):809–815. [DOI] [PubMed] [Google Scholar]
- 30.Wang TF, Wong CA, Milligan PE, Thoelke MS, Woeltje KF, Gage BF. Risk factors for inpatient venous thromboembolism despite thromboprophylaxis. Thrombosis research. 2014;133(1):25–29. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.McCurdy JD, Israel A, Hasan M, et al. A clinical predictive model for post-hospitalisation venous thromboembolism in patients with inflammatory bowel disease. Aliment Pharmacol Ther. 2019;49(12):1493–1501. [DOI] [PubMed] [Google Scholar]
- 32.Anderson FA Jr., Spencer FA. Risk factors for venous thromboembolism. Circulation. 2003;107(23 Suppl 1):I9–16. [DOI] [PubMed] [Google Scholar]
- 33.Bhandari S, Mohammed Abdul MK, Dhakal B, Kreuziger LB, Saeian K, Stein D. Increased Rate of Venous Thromboembolism in Hospitalized Inflammatory Bowel Disease Patients with Clostridium Difficile Infection. Inflammatory bowel diseases. 2017;23(10):1847–1852. [DOI] [PubMed] [Google Scholar]
- 34.Rogers MA, Levine DA, Blumberg N, Flanders SA, Chopra V, Langa KM. Triggers of hospitalization for venous thromboembolism. Circulation. 2012;125(17):2092–2099. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.