Facilitators and Barriers to Providing Vaccinations During Hospital Visits

We conducted semistructured interviews with parents and physicians to explore their perspectives on vaccinating during hospitalizations for acute illnesses. We conducted an inductive content analysis to identify facilitators and barriers. This study was completed at a freestanding children’s hospital in Washington. In Washington, 69.9% of children aged 19 to 35 months are up to date (UTD) on vaccines compared to 70.4% nationally.^27–29  Of note, Washington has the second highest vaccine exemption rate nationally.³⁰  This study was approved by the institutional review board.

Parents were eligible to participate if they had a child aged ≤16 years who was not UTD on ≥1 vaccine dose and was admitted to the inpatient medical unit between December 2018 and April 2019. Because 50% of hospitalized children are admitted after an ED visit,^31,32  and vaccine delivery can also occur in the ED, we only recruited hospitalized children who were admitted through the ED, excluding direct admissions and admissions for planned procedures. We enrolled parents of undervaccinated children to explore parent experiences surrounding vaccine discussions during their admission process and hospitalization.

To identify eligible parent participants, 2 team members reviewed the electronic medical records of children admitted to the medical unit. Sampling occurred on random weekdays when both team members were available. Parents were excluded if their child had a condition that would alter the routine vaccination schedule per the Advisory Committee on Immunization Practices recommendations.³³  A child’s UTD status was confirmed by using the statewide immunization registry. Children were considered not UTD if they were >30 days overdue for ≥1 required vaccine doses, including influenza.^6,34  State registry data were collected on the number of missing vaccines. Eligible parents were approached by a research assistant; written consent was obtained. Parents with limited English proficiency were excluded. We used quota sampling to increase the diversity of participants by degree of undervaccination, including parents of children who had received no vaccines to those missing only the current influenza vaccine. Parent interviews were completed in-person during the child’s hospitalization. Participants received a $25 gift card.

Physicians were identified by using purposive sampling to include PCPs, ED physicians, and pediatric hospitalists from diverse practices in the same geographic region as the hospital in which parents were recruited. We included physicians from each of these practice settings because all are key stakeholders in the vaccination status of hospitalized children admitted through the ED. Physicians were recruited by e-mail; interviews were completed by phone.

Interview guides were developed after literature review^3–5,18  and revised for clarity and topic relevance. Parent questions focused on their experiences vaccinating their child, vaccine discussions with their child's clinical team during the admission process or hospitalization, and their willingness to vaccinate their child during hospitalization. We focused on all recommended vaccines to obtain a broad range of facilitators and barriers.

Physician questions focused on (1) their experience discussing vaccines with parents during ED visits, hospitalization, or hospital follow-up visits (ED, hospitalist, and PCP participants, respectively); (2) willingness to discuss vaccines with parents during the hospital visit (ED and hospitalist participants); (3) barriers to providing vaccines; and (4) acceptability of vaccinating children during hospital visits. After each of the first 4 interviews with parents and physicians, the interview guide was iteratively refined (Supplemental Information).

A pediatric hospitalist with qualitative research experience conducted in-depth interviews with all participants. A research assistant documented field notes during the interviews. Interviews were audiorecorded, transcribed, and deidentified. We collected parent-reported demographic characteristics (age, sex, race, ethnicity, education, and child age) and physician demographic and practice characteristics (sex, race, setting, and years in practice).

Descriptive statistics were used to summarize demographic and practice characteristics. We counted how many vaccine types a child was missing from the state vaccine registry.

We used inductive content analysis to identify themes related to facilitators and barriers to vaccinating during the admission process and hospitalization.^35,36  The initial coding of transcripts was conducted concurrently with the interviews. After each set of 3 interviews, 2 authors independently coded the transcripts, reviewed the identified codes, and reached consensus on a preliminary set of codes and code definitions. This process was done iteratively until thematic saturation was reached, at which point recruitment was stopped. After the initial coding, both authors reviewed the codebook, refined the definitions, and discussed codes with a third investigator to clarify and finalize them. The codes were then used to inductively develop themes. Thematic evaluation focused on domains that participants identified as facilitators and barriers to vaccinating during hospital visits. Next, all authors reviewed and revised the relationships between the themes and subthemes to ensure clarity and comprehensiveness. Finally, we discussed the relationships between the themes to organize our results into a conceptual framework on the basis of an existing model for health behavior. Several established health behavior models and models of vaccine behavior were considered.³⁷  The adapted conceptual framework was revised iteratively until consensus was reached.

Twenty-one parents were interviewed. Parents’ ages ranged from 20 to 46 years; 86% were female; 76% were white. Twenty-four percent completed high school; 34% obtained a college or graduate degree. Mean child age was 3.8 years, ranging from 3 months to 16 years (Table 1). Vaccination status varied, with 30% having received no vaccines and 20% missing only influenza vaccine (Table 1).

Physician participants included 3 PCPs, 3 ED physicians, and 4 pediatric hospitalists. The physician sample was 70% female and 80% white. Practice settings included 3 community hospitals, 2 children’s hospitals, 2 private practices, and 1 managed care organization. Mean number of years in practice was 9, ranging from 1 to 24 years (Table 2).

Facilitators and barriers to hospital-based vaccination fell under 4 major themes: (1) systems-level factors, (2) physician-level factors, (3) parent-physician interactional factors, and (4) parent- and child-level factors (Table 3).

We organized the themes into a conceptual framework on the basis of the social ecological model (SEM). We used the SEM because the identified facilitators and barriers occurred at multiple levels: individual (parent and physician), interpersonal (parent-physician interaction), and organizational (system-level).^37–39  The levels correspond with previous vaccine intervention frameworks^38,39  and align with identified processes for vaccine delivery (identification of undervaccination, ordering vaccines, physician-parent communication, and parental acceptance of vaccines).^5,9  Our adaptation of the SEM reveals how the different themes at one level potentially influence other levels (Fig 1).

Systems-level factors referred to facilitators and barriers at the level of the hospital or health care system.

A major facilitator identified by parent and physician participants from all practice settings was the importance of having immunization data available at the point of care (Table 4). ED and hospitalist physicians identified that having vaccinations integrated into the workflow was a facilitator (Table 4). Many referenced successful influenza vaccine programs involving nurse screening and automatic vaccine ordering.

A major barrier identified by parents and ED and hospitalist physicians was a lack of awareness that the child was vaccine-eligible (Table 4). Parents were often unaware that their child was not UTD or did not know which vaccines their child was missing. Hospital-based physicians often relied on parental report of a child’s vaccination status. Being unaware of vaccine eligibility was described as a cause for underidentification of vaccination opportunities.

Another parent-identified barrier was that their child’s undervaccination went unaddressed; some parents did not recall having any hospital provider discuss their child’s vaccination status (Table 4). Hospital-based physicians, particularly ED physicians, identified logistic barriers to vaccinating (eg, the time required to correctly identify a child’s vaccination status, discuss vaccines with families, and deliver the vaccine) (Table 4).

A major facilitator among parent and all physician participants was the perception that it was acceptable for hospital-based physicians to address vaccines (Table 4). Parent participants discussed that hospital-based physicians have experience caring for children with vaccine-preventable diseases (VPDs), thus making them uniquely qualified to discuss vaccines with parents (Table 4). Similarly, ED and hospitalist participants reported drawing on their experience caring for children with VPDs when discussing vaccines with families. Lastly, all physician participants identified potential benefits to vaccinating children in the hospital as facilitators (Table 4). For example, parents with concerns about side effects may be willing to receive vaccines when their child was observed by medical personnel.

An important barrier discussed by ED and hospitalist participants was a perceived lack of skills to discuss vaccines with vaccine-hesitant families (Table 4). They reported many experiences in which they had been unable to persuade a family to vaccinate and reflected on a lack of skill at having these discussions.

This theme incorporates facilitators and barriers related to the parent-physician relationship. A major facilitator identified by parent and all physician participants was the development of a relationship or rapport before discussing vaccines (Table 4). Hospitalist participants identified that they were able to develop rapport with families over the child’s hospitalization and over multiple hospitalizations for children with medical complexity. Conversely, ED and hospitalist participants identified a lack of rapport as a barrier if they did not spend sufficient time with families before vaccine discussions. Parent participants discussed how they did not want to discuss vaccines with their child’s physician if they did not have rapport with them (Table 4).

Another facilitator identified by parent and all physician participants was having consistent information about vaccines across care settings (Table 4). Parents discussed that when the vaccine information from their hospital-based physicians was consistent with their PCP, they felt more confident in choosing to vaccinate. All physician participants identified hospital visits as an opportunity to reinforce PCP vaccine recommendations.

This theme encompassed facilitators and barriers related to the experiences and beliefs that parents have when considering vaccinations in the hospital. Many parents expressed willingness to have their child receive vaccines during hospital visits; ED and hospitalist participants provided further support for this facilitator on the basis of their experiences (Table 4).

Parent and all physician participants identified perceived susceptibility to VPDs as a facilitator (Table 4). They discussed that the prevalence of VPDs within the community (eg, measles) and having a sick child made parents concerned that their child may have a VPD when seeking hospital-based care. Both groups expressed that this made vaccine discussions during hospitalization salient. Lastly, parent and all physician participants identified that there are external triggers to vaccinating (eg, school requirements) that may be relevant at the time of hospitalization (Table 4).

Another parent-identified facilitator was their personal experience with vaccines (Table 4). If their child had received vaccines without negative side effects, they were willing to receive additional vaccines. All physician participants identified hospitalizations as an opportunity to highlight vaccines relevant to the child’s current illness (eg pneumococcal conjugate vaccine for a child with pneumonia) (Table 4).

Parent participants expressed that their child’s current illness was a barrier to vaccination; all physician participants had experienced parental resistance to vaccination because of the child’s illness (Table 4). Another barrier identified by both parents and physicians was that vaccination was not the primary focus of the child’s hospital visit and not a priority for parents if not directly relevant to the current illness (Table 4). Furthermore, some parent participants perceived a lack of susceptibility to VPDs; all physician participants reported conversations in which they experienced this perception as a barrier to vaccine acceptance (Table 4). Lastly, parent and all physician participants reported experiences in which the family preferred to receive vaccines at their PCP’s office (Table 4).

A major barrier was firmly held parental beliefs against vaccinating (Table 4). Additionally, some parents who had elected to follow a delayed vaccination schedule felt that a delayed schedule was typical (Table 4). In this scenario, they did not understand why the hospital-based physician was strongly recommending vaccines during the hospital visit.

Lastly, a physician-identified barrier was the perception of lack of trust in the medical system for parents whose children were not UTD (Table 4). Physicians discussed experiences in which, despite bringing their child to the hospital, parents did not trust the medical system around vaccines.

In this qualitative study, we identified facilitators and barriers to vaccinating hospitalized children. Important facilitators were the beliefs among parent and physician participants that hospital-based physicians should address vaccines with families and that many parents would be willing to accept vaccines in the hospital. Parent-identified barriers specific to hospital-based vaccinations included the child’s current illness, vaccines not being relevant to the current visit, and a preference to receive vaccines from their PCP. These barriers were in addition to other barriers that have also been identified in primary care, such as parental vaccine beliefs.^40–42 

When considering how to address parent-level barriers, an important barrier identified by ED and hospitalist physicians was a perceived lack of communication skills to address vaccinations with vaccine-hesitant families. This finding has not been well described in previous studies about inpatient vaccinations. Vaccine communication has been studied among PCPs, with many resources focused on developing long-term relationships.^43–47  ED and hospitalist physicians expressed interest in education for hospital-based vaccine communication. Adapting existing strategies, such as motivational interviewing,^44,47,48  for hospital settings is an important next step. Furthermore, there may be opportunities to capitalize on identified facilitators (eg, the ability to observe the child for side effects) that may resonate with parents and have yet to be fully explored in inpatient settings. Although some identified facilitators, like the presence of external triggers to vaccinate (eg, school requirements) are not unique to hospitalization, hospitalization may occur at a time when a family is primed to reconsider vaccinations.

Notably, there were instances in which aspects of hospitalization could serve as either a facilitator or barrier. For example, physicians identified that when a vaccine was relevant to the child’s illness, hospitalization provided an opportunity to address the need for vaccines. Conversely, parents may be less willing to accept vaccines during hospitalization because of the child’s illness. These findings highlight the complexity underlying parental vaccine decision-making⁴⁹  during hospitalization, similar to primary care.

Importantly, both parents and physicians identified that it was acceptable and appropriate for hospital-based physicians to address vaccines and were willing to discuss vaccines in hospital settings. These attitudes align with a survey about vaccine attitudes in parents of hospitalized children and hospital staff and with previous studies revealing vaccine uptake in hospitalized children when offered.^2,3,9,26  Physicians and parents highlighted the importance of rapport as a key component to vaccine communication during hospitalization, similar to the trust required for all vaccine communication.^50,51  It is promising that hospitalists identified the ability to build rapport during a child’s hospitalization.²⁶ 

Systems-based barriers were similar to those identified in previous studies: obtaining accurate vaccination histories is difficult, vaccinations are often not addressed during hospitalizations, and logistic barriers.^5,9,14,15  A major systems-level barrier was the lack of awareness that a child was vaccine-eligible. Systematically identifying undervaccinated children is the first step to comprehensively address and offer vaccines at all points of care.^1,52–54  Many physicians referenced successful influenza vaccination programs as a model for systematic solutions.¹²  Expanding inpatient influenza vaccine programs to include other vaccines can be explored as a method to improve vaccine delivery.

Our use of the SEM provides a conceptual framework to consider when designing and evaluating hospital-based vaccine interventions. The SEM has been used to examine factors influencing vaccine uptake at a population level^38,55  but has yet to be applied to hospital-based vaccine interventions. Our adaptation of the SEM highlights the importance of considering multilevel facilitators and barriers related to vaccination during hospitalization. Currently, inpatient vaccination rates are suboptimal.^1–4,9–17  Thus, developing interventions that address multilevel barriers within the SEM (ie, system, interpersonal, physician, and parent) may help to improve inpatient vaccination rates.

Our use of purposive sampling of physician participants may have introduced selection bias. As is inherent to qualitative research, there may be additional provider-level themes that were not identified. However, many physician-identified themes were consistent across care settings, and we obtained thematic saturation.

Parents were recruited from a single children’s hospital; physicians were recruited from a limited geographic region. Despite these limitations, this study provided an in-depth understanding of facilitators and barriers to vaccinating during hospitalization.

Because interviews require intensive communication, we enrolled only English-speaking participants. Additionally, 75% of parent participants and 80% of physicians self-identified their race as white. In past studies, researchers have identified sociodemographic characteristics related to vaccine uptake (eg, income, education, and country of birth).^56–58  However, little is known about how the demographic characteristics of parents and physicians influence vaccine communication. As this was a qualitative study, we are unable to quantify differences in the importance of the themes by sociodemographic characteristics. Future studies that include larger, more diverse populations of physicians and parents, including families with limited English proficiency, are needed.

Parents and physicians identified several key facilitators related to hospital-based vaccinations, most importantly that both groups were willing to discuss vaccines in hospital settings. Furthermore, we identified key barriers at multiple levels, such as identifying undervaccinated children, training hospitalists on vaccine communication, and parental concerns about vaccinating during the child’s current illness. Efforts to provide vaccines to children during hospitalization will likely need to address existing multilevel barriers.