Type of GI Provider Does Not Affect Rates of Vaccination Amongst Inammatory Bowel Disease Patients

Background: Patients with Inammatory Bowel Disease (IBD) are at heightened risk of infection, and are often under vaccinated. At our institution, IBD patients may follow up with one of three gastroenterology (GI) providers: IBD specialists, GI Fellows and General GI providers. Aims: Our primary objective was to identify whether the type of GI provider had an effect on rates of vaccinations for IBD patients. The seven vaccines studied were listed in the American College of Gastroenterology’s 2017 Guidelines on Preventative Care in IBD, and include Inuenza, Tetanus/Diphtheria/Pertussis, Hepatitis A (HAV), Hepatitis B (HBV), Pneumococcus (PCV13 and PPSV23), Meningococcus, and Human Papillomavirus (HPV). Methods: Retrospective case control study of IBD patients, looking at vaccination rates for each of the seven vaccinations listed above, and type of GI provider followed. Other data collected included patient demographics, IBD type, treatment regimen and insurance. Results: Of 338 patients, 65 (19.2%) 110 163 with HBV was only a difference in by provider. Bivariate analysis showed that patients who followed with IBD specialists and GI fellows were more likely to be for HBV than patients who GI = p = and OR = 2.73, p = .007 respectively). Conclusion


Introduction
In ammatory Bowel Disease (IBD) is an in ammatory condition with an increasing incidence and prevalence throughout the world [1]. The pathogenesis of IBD, while not completely understood, is thought to involve a dysregulated immune response to gut ora, as well as environmental factors that promote intestinal in ammation [2]. Due to this in ammation, IBD patients often face the prospect of long-term immunosuppressive therapy in order to control their symptoms. As a result, serious or opportunistic infections can pose a problem for this patient population [3]. Many of these infections can be preventable through regular vaccination, however the literature shows that IBD patients have suboptimal rates of vaccination, in most cases, lower than non-IBD patients [4].
The reason behind the lower vaccination rate is likely multifactorial, but one possibility may be due to whom the patient identi es as their primary provider. Patients with IBD are likely to follow up with both a gastroenterology (GI) provider and a primary care physician (PCP), but most, if not all will see their GI provider more often [5]. This close relationship may lead to patients treating their GI provider as their PCP, leading to confusion over whose responsibility it is to vaccinate IBD patients [6]. There is data that shows that many GI providers are less likely to vaccinate patients than PCPs and are less aware of the appropriate immunizations for IBD patients [7][8][9]. As a result, IBD patients often miss opportunities to receive necessary vaccines [10]. To address the under vaccination of IBD patients, the American College of Gastroenterology (ACG) published clinical guidelines in 2017 with the goal of increasing awareness of this issue among GI providers [11]. These guidelines represent the rst o cial set of recommendations for the vaccination of IBD patients, an important step towards educating GI providers about the importance of vaccinating this distinct patient population.
Many factors have been shown to contribute to differences in vaccine rates, including insurance status, income, education level and the presence of preexisting conditions such as IBD [12]. Studies have also examined vaccine-speci c differences among IBD patients, but no study has sought to examine the effect that the speci c type of GI provider has on vaccination rates of IBD patients.
The objective of this study is to identify whether there are differences in vaccination rates among IBD patients who follow up with these three different types of GI providers: IBD specialist, GI fellow, and general GI provider. We also aim to examine other factors that may in uence the vaccination rates of IBD patients, including age, race, sex, type of IBD diagnosis, immunosuppression status, and socioeconomic factors such as insurance status.

Methods
We performed a retrospective case-control study of vaccination rates in IBD patients who followed at the Monte ore Medical Center (Bronx, NY) with either an IBD specialist, a GI fellow, or a general GI provider.

Patient Selection
Our study utilized Clinical Looking Glass (CLG), a proprietary software used at Monte ore that assists in searching through patient records, to create a database of patients with a con rmed diagnosis of IBD who followed up with a GI provider at the institution. Patients were identi ed by searching for those with an ICD 9 diagnosis code of regional enteritis, idiopathic colitis or ulcerative colitis (555.1, 555.2, 555.9, 556.0-556.9). Inclusion criteria included (a) patients 18 years or older, (b) patients with an ICD 9 diagnosis listed above and (c) patients with a documented follow up visit with a GI provider at the institution between 1/1/18 and 12/31/19. Exclusion criteria included (a) patients younger than 18 years and (b) patients who have not followed up with a GI provider at the institution since 1/1/18.

Study Outcomes
The primary outcome assessed was whether patients were current on the following vaccines: In uenza, Tetanus/Diphtheria/Pertussis (TDaP), Hepatitis A (HAV), Hepatitis B (HBV), Pneumococcus (PCV13 and PPSV23), Meningococcus, and Human Papillomavirus (HPV). These vaccines represent the seven recommended inactivated vaccines in the ACG clinical guidelines for health maintenance in IBD patients. We also utilized guidelines published by the Advisory Committee on Immunization Practices (ACIP) to determine which patients were current on vaccinations [13]. Two vaccinations, Meningococcus and HPV, were recommended primarily for younger people, so for analysis, patients considered eligible for these vaccines had to be age 27 or younger. The remainder of the criteria that determined if a patient was current on a vaccine are listed in Table 1.

Pneumococcus
Patients were considered current for Pneumococcus if they received a dose of PCV13, followed by PPSV23 eight weeks to one year later. Patients were also considered current if they received a PPSV23 dose, followed by a PCV13 dose one year later. Patients who had received their rst vaccination less than one year ago and who were currently awaiting a second were also considered to be current.
Meningococcus Patients 27 years old and younger were considered current for Meningococcus if they had received a conjugate vaccine followed by a booster shot, or the serogroup B shot.

HPV
Patients 27 years old and younger were considered current for HPV if they had received three doses of the HPV vaccine within six months.

Demographic and Clinical Data
Chart review was conducted using the electronic medical record to gather demographic information about each patient who met criteria for the study (age, sex, race and type of insurance) as well as information about the disease and its management (type of IBD diagnosis, treatment regimen and type of GI provider followed). Finally, information regarding vaccination status was obtained (date and type of vaccination).

Statistics
Our initial database was exported from CLG to Excel, where data from chart review was added. Once collated, our spreadsheet was imported into IBM SPSS for Mac (Version 25.0. Armonk, NY). Categorical variables were coded numerically to streamline statistical analysis. The Shapiro-Wilk test was used to assess normal distribution in continuous descriptive variables (age). As age was not normally distributed, it was converted to a categorical variable using 60 as a cutoff in order to separate the cohort into a "young" and "elderly" age group.
Univariate chi-squared and regression analysis was rst done to assess for independent predictors that affected vaccination status for each of the seven vaccines that were studied.
Bivariate logistic regression analysis was then done to identify signi cant variables that affected vaccination rates. Each model was adjusted for age, sex, race, type of GI provider, IBD diagnosis, immunosuppression status, and insurance. Statistically signi cant values were determined by a two-sided p value < 0.05.

Results
A total of 338 patients were diagnosed with IBD and followed up with a Monte ore GI provider between the dates of 1/1/2018 and 12/31/2019 ( Table 2). The average age was 50.7 years. There were 180 (53.30%) male and 158 (46.70%) female patients.
The most predominant race was Hispanic (36.10%), followed by Black (31.70%), White (29.0%), Asian (2.40%), and Other (0.80%). There were 165 (48.80%) patients who had a diagnosis of Crohn disease, 169 (50%) who had Ulcerative Colitis (UC), and 4 (1.20%) who had indeterminate colitis. There were 163 (48.30%) patients in this cohort who followed up with an IBD specialist, 110 (32.50%) who followed up with a general GI provider, and 65 (19.20%) who followed up with a GI fellow. The most predominant type of insurance in this cohort was private (51.48%), followed by Medicaid (24.85%) and Medicare (23.67%). There were 155 (45.86%) patients who were on immunosuppressive therapy (either biologic, immunosuppressant or a combination of both with or without steroids), 131 (38.76%) patients who were on non-biologic therapy (aminosalicylates or anti-diarrheals), and 52 (15.39%) patients who were not on any therapy. Of note, no patient in this cohort was on steroid monotherapy and any patient on steroids was also on either a biologic or immunomodulator.  Univariate chi-squared analysis (Table 3) showed that patients with Crohn disease were more likely to be vaccinated for in uenza than those with UC (p = .029). For the TDaP vaccination, patients who were female, Black, Hispanic and immunocompetent were more likely to be vaccinated than male, White and immunosuppressed patients (p = .002, .009 and .019 respectively). For the HAV vaccination, patients who were Black, Hispanic and immunosuppressed were more likely to be vaccinated rather than White and immunocompetent patients (p = .011 and .033 respectively). For the HBV vaccination, patients were more likely to be vaccinated if they were younger than 60, Black, Hispanic, followed by IBD Specialists or GI Fellows, and if they were immunosuppressed when compared to patients older than 60, White, followed by a general GI and if they were immunocompetent (p = .0001 for all). For the pneumococcus vaccinations, patients older than 60, Black, Hispanic, and those with public insurance were more likely to be vaccinated than patients younger than 60, White and with private insurance (p = .0001, .003, .002 respectively). For the HPV vaccination, patients who were male and those with UC were more likely to be vaccinated than females and those with Crohn disease (p = .007 and .030 respectively) In bivariate logistical regression analysis (

Discussion
Patients with IBD are at higher risk for certain serious infections which may be partly due to the use of immunosuppressive medication but also due to their underlying disease. Therefore, vaccinations in these patients are a critical part of their treatment plan, however rates of vaccinations in this patient population are lagging. This study is the rst that examines the effect that type of GI provider plays on vaccination status in IBD patients. Focusing on seven inactivated vaccines recommended in the ACG clinical guidelines for preventive care in IBD patients, we performed an analysis of 338 IBD patients followed by GI providers at a single institution to see if type of GI provider affected vaccination status in IBD patients. The results showed that type of GI provider did not in uence an IBD patient's vaccination status for all but one of the seven inactivated vaccines. The HBV vaccination was the only vaccination where type of GI provider was signi cant for likelihood to be up to date with vaccination where IBD specialists and GI Fellows outperformed general GI providers. The signi cance of HBV speci cally may be explained by the fact that IBD specialists may be more likely to prescribe immunosuppressive therapies such as a biologic or immunomodulator where immunization to HBV is considered standard of care prior to giving the medication. Also, insurers tend to request vaccination history including HBV status prior to approving the use of biologic, possibly contributing to the signi cant number of patients up to date with the HBV vaccination. GI Fellows also outperformed general GI providers as well, likely due to GI fellow supervision by IBD specialists when seeing IBD patients at this institution.
Looking beyond type of provider, there were several other interesting observations in regards to vaccinations in this IBD patient population. Race played a signi cant role in vaccination rates in multivariate analysis for four of the vaccines studied (TDaP, HAV, HBV, and Pneumococcus). For each of these vaccines, Black and Hispanic patients were better vaccinated than White patients. These results were surprising, given that other studies on racial differences in vaccine administration typically found that Black and Hispanic patients had lower rates of vaccine coverage [14,15]. When comparing this cohort's vaccination rates to the US national averages for these four vaccines, Black patients with IBD in this cohort were better vaccinated than Black Americans nationally (TDaP, HAV, HBV, and Pneumococcus rates of 57.9%, 27.1%, 52.3% and 37.4% in our cohort versus 51.1%, 10.2%, 27.0% and 34.2%) while White patients with IBD in this cohort were less likely to be vaccinated than White Americans nationally (TDaP, HAV, HBV, and Pneumococcus rates of 36.7%, 10.2%, 24.5% and 16.3% in our cohort versus 63.4%, 14.0%, 36.2% and 47.5%) [16]. These differences could be attributed to the work of the institution itself since it is located in the Bronx, New York, one of the most racially diverse counties in the United States. The institution is sensitive to health inequities and disparities in their patients and as a result, has expertise in handling health issues in these minority patients. Also, since IBD patients utilize the health care system more than the general population, this may have allowed for higher vaccination rates at an institution that is focused on minority health issues as a whole [17].
Age was another factor that played a role in rates of vaccination for In uenza and the HBV vaccination. IBD patients 60 and older were better vaccinated against In uenza, while IBD patients under 60 were better vaccinated for HBV. This data is consistent with national trends in vaccination seen in both the general population and IBD patients speci cally, showing that that younger patients were more likely to be current with HBV vaccination status and less likely to be current on in uenza when compared to older patients [16,18]. One reason for this trend could be due to enhanced in uenza counselling by providers to older patients [19]. For the HBV vaccination, as it is typically indicated for high risk patients such as those with IBD, one reason to explain why younger patients are better vaccinated is that the incidence of IBD is greater in younger people, leading to rates of vaccination that skew towards the younger population [20,21].
Immunosuppressive therapy also played a role in vaccination status for TDaP and Pneumococcus vaccines. Immunosuppressed IBD patients had a lower rate of being vaccinated for TDaP and a higher rate of being vaccinated for pneumococcus. A reason for why this effect was seen may be because TDaP is more likely to be given to patients by their PCP, rather than their GI provider [7]. PCP's are also less likely to immunize patients on immunosuppression, possibly explaining the negative effect immunosuppressive therapy had on TDaP [22]. Pneumococcus is primarily indicated for high risk patients (when under 65) such as those on immunosuppressive therapy. Given that this therapy is as an indication for the vaccine, it is likely that high-risk patients were better vaccinated due to the indication [23].
Finally, sex was seen to play a role in TDaP vaccination status, with women being more likely than men to be vaccinated. While there is no consistent data that supports this trend on a national level, a possible explanation for this is that ACIP guidelines since 2013 have recommended a TDaP shot for women during every pregnancy [24]. Pregnant women therefore may have led to better TDaP vaccination rates because of this additional opportunity to be vaccinated.
This study had several limitations that are important to note. The rst is related to the reliance on accuracy of data from one institution's medical records. While efforts are made to update patient records to re ect care that was provided at outside institutions, it is possible that there were missed vaccinations if they were given elsewhere and not documented in the institution's medical records. Another limitation is a lack of information as to what provider speci cally ordered and administered vaccines. A GI provider may have requested a certain vaccination for their patient, however, if they were unable to administer it in their own practice it is out of their control whether or not a patient actually gets the requested vaccination. Finally, this study was unable to gather information on any reasons for why patients were not vaccinated. Previous studies have shown that patient anxiety plays a role in low vaccination rates among IBD patients, and it is possible that it played a role among this cohort as well [9].
Despite limitations, there were several strengths of this study including a large cohort of racially and socioeconomically diverse IBD patients following up at a single institution. Also, this study was the rst to examine if differences in type of GI provider follow up affected vaccination rates. While type of GI provider may not in uence the likelihood of vaccination, this study did show other factors that in uenced vaccination rates speci cally the effect of race. Black and Hispanic IBD patients were more likely to be vaccinated compared to White IBD patients for Tdap, the HAV vaccination, the HBV vaccination, and Pneumococcus vaccination, a nding which is not consistent with national vaccination rates amongst races. This may re ect more the work of the single institution that is located in a racially diverse county of New York and is focused on minority health issues. Future research should include efforts to better understand the effects of the institution itself on this nding or if there are other factors that shaped these results. Vaccination rates for the institution should be analyzed to compare with the IBD patients' vaccination rates to see if this nding is consistent with the institution as a whole or just to the IBD patients at the institution. Despite the interesting ndings regarding race, this study shows IBD patients at this institution still lag in receiving appropriate vaccinations. Future studies should be done to better understand IBD patients and all GI providers attitudes and knowledge of vaccinations, as well as assessing for any barriers to vaccinations. With this information, areas to explain gaps in vaccination can be identi ed and prospective studies can be designed to address these issues speci cally in hopes of improving vaccination rates for these patients.

Declarations Ethics Approval
This study was approved by the Institutional Review Board at Albert Einstein College of Medicine and was deemed to be ethically conducted and carried out.

Consent for Publication
This study does not contain individual personal data, therefor consent for publication is not applicable.

Availability of Data and Materials
The datasets generated during and/or analyzed during the current study are not publicly available as they contain identi able patient information and are stored on a secure hard drive. However, data can be de-identi ed and made available from the corresponding author on reasonable request.

Funding
No outside sources of funding were acquired for this study.
Author's contribution MT collected, analyzed and interpreted patient data, as well as was a major contributor to writing the manuscript. Both NM and JK helped devise the study and played major roles in writing and re ning the manuscript.