An Opportunity in Cancer Prevention: Human Papillomavirus Vaccine Delivery in the Hospital Free

This quality improvement (QI) initiative was implemented in a 300-bed Midwestern tertiary academic children’s hospital. The pediatric hospital medicine (PHM) division consists of 32 attending physicians, 10 advanced practice providers, and 100 pediatric residents who care for patients on the PHM service. Our EHR is Epic (Verona, WI).

Our providers use discharge order sets to prepare the after-visit summary (AVS) distributed to families at discharge. The AVS includes information on the patient’s current immunization status. Therefore, discharge preparation is an ideal time to discuss giving immunizations in our hospital. The Wisconsin Immunization Registry (WIR) is a record of immunization status available in an independent tab in Epic. In July 2019, discharge order sets were modified in the EHR to include a snapshot of WIR information (Supplemental Fig 3). This change followed clinical decision support guidelines by prompting providers to review WIR within the AVS and consider vaccinations at discharge.⁸ 

Two provider training seminars were conducted in July 2019, 1 during a pediatric resident meeting and another at a PHM meeting. Training involved a presentation about HPV vaccine safety and efficacy, guidance for vaccine counseling, and a comparison of national, statewide, and institutional vaccination rates. Follow-up materials were emailed in August.

Adolescents ≥13 years who were discharged from the PHM service from May 2019 through February 2020 were included. This age was selected because the Centers for Disease Control recommends vaccination by age 12, and therefore patients age ≥13 are past due. Patients were admitted to PHM or critical care service and discharged by PHM service. These data were collected to interpret if continuity of care impacted discharge planning and vaccination at discharge. Patients who were not living in Wisconsin were excluded because of the inability to verify vaccination status through the WIR. Patients with a contraindication to vaccination were excluded. Those with severe cognitive disabilities were excluded because of unique reproductive needs.

Data collected via chart review included: demographics, length of stay (LOS), admitting service, HPV vaccination status on hospitalization, and if the patient received the HPV vaccine before discharge. Vaccination status information was collected from the WIR.

Our primary outcome measure was the HPV vaccination rate, defined as the proportion of patients who received the HPV vaccine before discharge among those patients due for the HPV vaccine. Patients were defined as due for the vaccine if they had never received a dose of the HPV vaccine or had not completed the vaccine series and had lapsed >6 months after a previous dose of the vaccine. Meningococcal immunization status on admission and at discharge was collected to use as a comparison and a secondary outcome. The process measure, documentation accuracy of immunization status in the history and physical note, was collected. Accurate documentation was defined as immunization status recorded in the history and physical being the same status reported in the WIR. The balancing measure was LOS defined as the number of days the patient was hospitalized.

Statistical process control p-charts were used to analyze HPV and meningococcal vaccination rates, documentation accuracy, and LOS.⁹  Special cause variation was defined as 8 points above the initial median. χ² tests were used to assess associations between HPV vaccination rate and patient factors including sex and whether the patient was due for the first or subsequent dose of the vaccine and admitting service.

This project was determined exempt by the institutional review board.

Patient and encounter characteristics are outlined in Table 1. After exclusions, 440 patients ages ≥13 were included. Of these patients, 36.4% were male and 63.6% were female. The average age was 15.3, and this was a primarily English-speaking population.

HPV vaccination rates increased from a baseline median of 4.6% to 21.2% showing special cause variation (Fig 1A). Meningococcal vaccination rates increased from a baseline median of 8.3% to 26.6% showing special cause variation (Fig 1B). The median documentation accuracy was 78.0% and there was no median shift throughout the study (Fig 2). Of the 136 adolescents due for HPV vaccination with accurate documentation, 25 (18.4%) received it before discharge compared to 11.8% (4 of 34) of adolescents due for HPV vaccination with inaccurate documentation.

Analysis of patient factor relationships with HPV vaccination rate demonstrated no significant associations between sex, the dose of HPV vaccine the patient was due for, or the admitting service with HPV vaccination rate before discharge. Of male patients due for the HPV vaccine, 20.7% (16 of 77) received the vaccine before discharge compared to 14.1% (14 of 99) female patients (P = .245). Of patients due for their first dose of the HPV vaccine series, 13.9% (16 of 115) received the HPV vaccine before discharge compared to 23.0% (14 of 61) of patients due for a booster HPV vaccine (P = .129). Of patients admitted to the critical care service who were due for the HPV vaccine, 11.8% (2/17) received the vaccine before discharge compared with 17.2% (27 of 157) of patients hospitalized under PHM for the duration of their hospitalization (P = .392). Of the 54 adolescents due for both the HPV and meningococcal vaccines, 17 (31.4%) received both before discharge, 37 (68.5%) received neither, and no patients received only the HPV or meningococcal vaccine.

The mean LOS remained at 2.8 days throughout the study.

The coronavirus 2019 pandemic impacted the ability to implement further interventions or plan-do-study-act cycles.

HPV vaccination rates increased from 4.6% to 21.2% after modifying discharge order sets in the EHR and implementing HPV vaccine provider education. This increase surpassed the aim of this study and is higher than rates of delivery of the HPV vaccine reported in a recent multicenter study.¹⁰  The reason for success is likely a synergistic combination of interventions. Implementation of modified discharge order sets may have streamlined workflow by placing valuable WIR information in the provider’s view while preparing discharge orders, aligning with clinical decision support guidelines.⁸  Other studies have also seen success in using provider training and EHR changes to improve vaccination rates.¹¹ 

Meningococcal vaccination rates increased from 8.3% to 26.6%. Provider education regarding the HPV vaccine may have unintentionally influenced rates of the meningococcal vaccine. Future work could explore this relationship by emphasizing the importance of administering all needed vaccines for those not up-to-date during hospitalization.

Previous literature has found that provider recommendation of the HPV vaccine is twice as high for females compared to males.¹²  Our analysis did not demonstrate this difference, which may be related to provider education reinforcing that both should receive the vaccine. Although not statistically significant, we observed higher HPV vaccination rates among patients who received the HPV vaccine previously, possibly related to hesitancy regarding the HPV vaccine among those who have not started the series. Future interventions could optimize patient education and measure HPV vaccine refusal to further elucidate this potential association. HPV vaccination rates were not significantly different based on admitting service, possibly because vaccinations are considered to be the responsibility of the discharging service.

One in 5 encounters had inaccurate documentation of immunization status in the admission note. Recognition of the need for vaccination earlier in the hospital stay may allow for more opportunities to discuss vaccinations, thereby increasing vaccination rates before discharge.

This study was conducted at a single hospital, with an almost entirely English-speaking patient population. Further research on this topic may be indicated in other populations. Because our baseline data are limited, we may not have accurately captured baseline vaccination rates. The coronavirus 2019 pandemic limited our ability to initiate additional interventions.

HPV and meningococcal vaccination rates among hospitalized adolescents increased after a combination of efforts including provider education and enhancements of EHR discharge order sets. Further improvement could be obtained through patient education combined with provider training, or training for health care workers in other roles. Future work could explore vaccine administration failures and optimization of identifying correct immunization status on admission.