PLEASE FOLLOW DIRECTIONS CAREFULLY AND PLEASE FOLLOW RUBRIC . I have attached the required articles

Evaluation of the Implementation of a

Multicomponent Intervention to Improve Health Care

Provider Communication About Human

Papillomavirus Vaccination

D1X XJenna E. Reno, D2X XPhD; D3X XSean T. O’Leary, D4X XMD, MPH; D5X XJennifer Pyrzanowski, D6X XMSPH; D7X XSteven Lockhart, D8X XMPH; D9X XJacob Thomas, D10X XMPH; D11X XAmanda F. Dempsey,D12X XMD, PhD, MPH

From the Adult and Child Consortium for Health Outcomes Research and Delivery Science (ACCORDS), School of Medicine, University of Colorado Denver, Aurora The authors have no conflicts of interest to disclose.> Address correspondence to Jenna E. Reno, PhD, ACCORDS, Adult and Child Consortium for Health Outcomes Research and Delivery Science, University of Colorado Denver, Mail Stop F443, 13199 E Montview Blvd, Suite 300, Aurora, CO 80045 (e-mail: [email protected]). Received for publication May 7, 2018; accepted August 7, 2018.

TAGGEDPABSTRACT

OBJECTIVE: To evaluate the relative use, usefulness, and facil- itators and barriers to use as perceived by providers of 5 differ-

ent components in a human papillomavirus vaccine

communication intervention—which was found to be effective at improving human papillomavirus vaccination rates.

METHODS: Four serial surveys of 108 providers (doctor of medicine, nurse practitioner, or doctor of osteopathic medi-

cine) from intervention clinics involved in the study assessed

the use and usefulness of the 5 communication intervention

components during a 12-month period.

RESULTS: Survey response rates were 79% to 86%. The fact sheet (64%−77%) and motivational interviewing techniques (MI; 86%) were the most used components—use was sus- tained during the 12-month period. These components also

were perceived as somewhat or very useful by most providers,

and this perceived usefulness increased over time (very or

somewhat useful at end of study, 97% fact sheet, 98% MI,

respectively). Although fewer providers reported using the

Web site

(15%−42%), or disease images (6%−17%), when these were used, most providers (67%−87%) felt they were somewhat or very useful. The decision aid was not used frequently (17%

−41% of providers), and 43% of providers felt it was not very or not at all useful. Facilitators and barriers were identified for

each component. The fact sheet and MI were perceived as the

easiest to integrate into the clinic workflow.

CONCLUSIONS: The fact sheet and MI were the most used and most useful intervention components. Both were easy to inte-

grate into clinic workflow, and their use was sustained over

time. Dissemination of similar interventions in the future

should focus on these 2 specific components.

TAGGEDPKEYWORDS: communication intervention; HPV vaccination; patient-provider communication; program evaluation

ACADEMIC PEDIATRICS 2018;18:882−888

TAGGEDPWHAT’S NEW

Our previous multicomponent intervention increased

human papillomavirus vaccine-series initiation and

completion. This study examines the 5 intervention

components and identifies which items appear to be

most impactful for driving these results.

TAGGEDPVACCINATION RATES FOR human papillomavirus (HPV) remain suboptimal in the United States.

1 As of 2016,

only 49.5% and 37.5% of female and male adolescents,

respectively, were up-to-date on their HPV vaccination

series, 2 well below the Healthy People 2020 target level

of 80%. 3

Despite the benefits of HPV vaccination, a substantial

proportion of parents are still hesitant to vaccinate their

children. 4 High levels of parental HPV vaccine hesitancy

are a significant contributor to low adolescent HPV vacci-

nation rates in the United States. 5,6

Interventions are

needed to effectively address this hesitation, especially

those that improve provider communication about

vaccines, 7 a key influential factor in parents’ vaccine

decision making. 8,9

Receiving a provider’s recommendation for the vac-

cine is one of the strongest predictors of HPV vaccine

utilization. 10,11

Unfortunately, research demonstrates

significant reluctance across medical specialties to rec-

ommend the vaccine routinely to all eligible adoles-

cents. 12,13

To address this, our research group

ACADEMIC PEDIATRICS

Copyright © 2018 by Academic Pediatric Association 882 Volume 18, Number 8

November−December 2018

developed a 5-component, provider-focused, HPV vac-

cine communication intervention (described previ-

ously 14 ). This intervention was found to be effective for

improving adolescent HPV vaccine initiation and com-

pletion rates compared with usual care. 14

Although previous research has shown the efficacy of

multicomponent interventions, 15

few have evaluated

which of the components seemed most impactful within

the trial. This study examines the use of our various inter-

vention components in further detail to understand which

were most valued, perceived as effective, and used. This

information is critical for understanding the mechanisms

most likely to have contributed to the success of the inter-

vention in increasing HPV vaccination rates and for clari-

fying which may be most necessary to disseminate in the

future.

TAGGEDH1METHODSTAGGEDEND

TAGGEDH2INTERVENTION OVERVIEW TAGGEDEND

A full description of the randomized controlled trial

(RCT) can be found elsewhere 14 ; here, we provide a

brief overview. Practices recruited for the trial came

from a practice-based research network in central

Colorado. Eligibility criteria included having 1) a min-

imum of 400 active adolescent patients, 2) baseline

HPV vaccination rates of <80%, and 3) HPV

vaccines available at the clinic. Eight intervention and

8 control practices participated in the 12-month

(February 2015−January 2016) intervention (clinical trials.gov #NCT02456077).

TAGGEDH2INTERVENTION COMPONENTS TAGGEDEND

The 5 components of the intervention included 1) an

HPV fact sheet (tailored to concerns of parents at that

practice), 2) a tailored Web site for parents, 3) disease

images, 4) an HPV vaccine decision aid, and 5) communi-

cation training workshops. These have been described in

detail previously (with examples) 14

but are summarized

in the following. Unless otherwise specified, all interven-

tion clinics used all intervention components.

Fact sheets were customized by each of the intervention

clinics to reflect information that would answer the ques-

tions and concerns most salient to patients at that particu-

lar practice. Content was chosen from a fact sheet library

developed by the study team that consisted primarily of

information on 1) HPV incidence, 2) HPV-related disease,

and 3) HPV vaccination safety and efficacy. Although the

fact sheets differed in their specific content, the general

content was the same (see example in the supplementary

data from the trial 14 ).

A tailored Web site for parents, called “iVac HPV,”

was HPV specific and was composed of educational infor-

mation about HPV vaccination and infection that was cus-

tomized for parents based on responses to a short

survey. 16

Practices that chose to use the Web site placed

iPads and/or kiosks in the clinic’s waiting rooms, made

the Web site available on computers in the examination

rooms, and/or posted the Web site address on their clinic’s

Web site and encouraged parents to look at it before their

appointment. All but 1 of the 8 intervention clinics incor-

porated the Web site into their practice (screen shots

available 14 ).

Disease images, including sex-specific photo represen-

tations of genital warts, penile cancer, vaginal cancer, vul-

var cancer, cervical cancer, and oropharyngeal cancer

were provided in binders to clinics to use at their discre-

tion. However, only 2 clinics elected to use this interven-

tion component. An HPV vaccine decision aid included

information about the risks of vaccinating versus not vac-

cinating and action steps for parents to take based on their

current thoughts about the vaccine (example available 14 ).

HPV vaccine communication training consisted of a

self-administered, 30-minute webinar and 2 in-person, 1-

hour training sessions that covered use of the

“presumptive communication approach” to open the HPV

vaccine conversation, 17,18

as well as training in motiva-

tional interviewing (MI) strategies 19,20

when a parent was

hesitant to vaccinate. As the presumptive approach is

increasingly becoming a standard in vaccine recommen-

dation practice, 18,21

it was not considered a novel compo-

nent of the intervention. Thus, the presumptive approach

was not consistently measured throughout the interven-

tion. A detailed description of the motivational interview-

ing training has been described elsewhere 14,20

(see also

the Alliance for HPV Free Colorado 22 ).

TAGGEDH2STUDY OVERVIEW TAGGEDEND

Seven serial surveys were administered via e-mail or in

person throughout the intervention time period to track

use of intervention components by providers. A final sur-

vey was conducted at the end of the intervention period

(ie, “postintervention survey”). Surveys were modified

and adapted throughout the study to meet the evaluation

needs of the project; thus, not all items were included in

each survey. The current study includes data from 4 of

the surveys administered quarterly, including the postin-

tervention survey. All study activities were approved by

the Colorado Multiple Institutional Review Board.

TAGGEDH2PARTICIPANTSTAGGEDEND

Participants were 108 health care providers from 8

intervention clinics participating in the trial.

TAGGEDH2MEASURES TAGGEDEND

TAGGEDPUSE OF INTERVENTION COMPONENTS TAGGEDEND For 4 of the intervention components (the fact sheet,

decision aid, disease images, and Web site), use of the

item was measured in 3 ways. In serial surveys conducted

during the intervention, providers were asked to report the

estimated total number of HPV vaccine−eligible visits for the given month as well as the estimated total number

of visits in which each component was used. This resulted

in 2 measures: provider-level component use (ie, percent-

age of providers reporting the use of intervention

TAGGEDENDACADEMIC PEDIATRICS IMPROVING COMMUNICATION ABOUT HPV VACCINATION 883

components at 1 or more HPV vaccine-eligible visits by

intervention month) and patient-level component use (ie,

percentage of total vaccine-eligible visits in which an

intervention component was used by intervention month).

A third use measure, patient-type intervention component

use, was assessed for specific patient scenarios, including

parents of eligible patients: 1) whose HPV vaccine

hesitancy was unknown, 2) who were HPV vaccine hesi-

tant, 3) who refused the HPV vaccine during the sched-

uled visit, and 4) who had refused the HPV vaccine

during a previous visit. Providers reported in the postin-

tervention survey how often they used each intervention

component specifically for each of the 4 patient scenarios.

Responses included “most of the time (>75%),” “often

(51%−75%),” “sometimes (26%−50%),” and “rarely/ never (0%−25%).” Because of the self-report nature of the survey, what

was considered “use” of an intervention component var-

ied. For example, the use of the fact sheet or decision aid

could be interpreted as discussing the content of the item

with patients/parents or simply giving it to them to review

on their own. Use of the Web site was defined as “referred

patients to use the Web site outside of the medical office,

either before or after the scheduled visit AND/OR had

patients access the Web site at the medical office via iPad,

Kiosk, or Exam Room Desktop.” Follow-through on this

recommendation to use the Web site was not measured.

TAGGEDPUSE OF MI T AGGEDEND MI was measured differently than the other interven-

tion components. In serial surveys conducted during the

intervention, providers were asked a general question

regarding MI. “During the month of ______, how fre-

quently did you use Motivational Interviewing techni-

ques when discussing HPV vaccine recommendations?”

Responses included “most of the time (>75%),” “often

(51%−75%),” “sometimes (26%−50%),” and “rarely/ never (0%−25%).” The postintervention survey collected more detailed information on MI through a series of

questions that queried how often providers used specific

MI techniques (eg, open-ended questions, reflections,

affirmations, ruler-method, metaphors or hypothetical

scenarios, and asking permission to share information; a

full description of the MI methods taught in this study

with examples can be found elsewhere 20 ) with responses

mirroring those described previously.

TAGGEDPUSE OF THE PRESUMPTIVE APPROACHTAGGEDEND The presumptive approach was measured by using 1

item that stated: “When bringing up the HPV vaccine

for the first time with parents of 11- to 12-year-old

patients, please select the approach that best describes

your current practice.” Responses included the follow-

ing: 1) I first ask parents a question, such as: “Are you

interested in getting the HPV vaccine for your child

today” (participatory approach); 2) I first tell parents

what to expect, such as: “we’ve got three vaccines

today: Tdap, HPV, and Meningococcal vaccine” (pre-

sumptive approach); 3) I use both approaches

(presumptive and participatory) equally; 4) I don’t

generally bring up the vaccine because we have a

standing order for these vaccines. So, the only parents

I discuss the vaccines with are ones I already know

have questions; and 5) I don’t generally bring up the

vaccine at 11- to 12-year-old patient visits because

most parents don’t accept it, anyway. Only the second

response option was categorized as use of the pre-

sumptive approach. All other answers were categorized

as did not (consistently) use the presumptive approach

(as instructed in the communication training). This

item was only included in the postintervention survey,

and therefore, in this analysis use of the presumptive

approach was not assessed over time.

TAGGEDPUSEFULNESS OF INTERVENTION COMPONENTS TAGGEDEND Providers’ perceptions of the usefulness of each inter-

vention component (excluding the presumptive approach)

was measured at intervention midpoint and postinterven-

tion. Responses were based on a 4-point Likert-type scale

ranging from “not at all useful” to “very useful.”

TAGGEDPLIKELIHOOD TO CONTINUE USING TAGGEDEND In the postintervention survey, participants were asked

to report the likelihood of continuing to use each interven-

tion component (excluding the presumptive approach):

“how likely or unlikely are you to use each of the follow-

ing toolkit components in the future?” Responses were

measured on a 4-point scale and ranged from “very like-

ly” to “very unlikely.”

TAGGEDPFACILITATORS AND BARRIERS TO INTERVENTION COMPONENT USE TAGGEDEND

In the postintervention survey, providers were asked to

report the degree to which they agreed or disagreed with

barriers specific to using each intervention component

(excluding the presumptive approach). Responses were

measured on a 4-point scale ranging from “strongly

agree” to “strongly disagree.” In addition, each item

included a “not familiar enough with [intervention

component]” response.

TAGGEDH2DATA ANALYSIS TAGGEDEND

Basic frequencies and percentages were calculated for

variables analyzed from the survey. Chi-square tests, anal-

ysis of variance tests, and 2-sample t tests were used for

the majority of comparisons when appropriate. Spearman

correlations were used to test the association between Lik-

ert-style questions that were kept on a continuous scale.

The McNemar test was used to compare intervention

component usefulness at 2 different time points, as well

as for comparing fact sheet use between each of the

patient-specific scenarios. To make global comparisons

between patient-specific scenarios or between-interven-

tion components when the outcomes were dichotomous,

data were transposed, and generalized linear mixed mod-

els (PROC GLIMMIX; SAS Institute Inc, Cary, NC) were

used to determine P values. All analyses were performed

using SAS, Version 9.4 (SAS Institute Inc.).

TAGGEDEND884 RENO ET AL ACADEMIC PEDIATRICS

TAGGEDH1RESULTSTAGGEDEND Across the 4 study surveys, between 67% and 78% of

total providers participating in the intervention completed

surveys. Overall, 108 providers completed 1 or more of the

surveys. Table 1 describes the respondents’ characteristics.

TAGGEDH2USE OF INTERVENTION ITEMS TAGGEDEND

Over time, the majority of providers reported using at

least 1 of the intervention components in all 4 of the

months in which provider-level intervention component

use and patient-level intervention component use were

measured (Tables 2 and 3). Peak use was primarily seen

during July, 6 months into the intervention, with tapered

use in fall and winter months. When providers did not use

intervention components, 47.4% agreed that having none

or very few HPV vaccine−hesitant parents was the reason. Of all the various intervention components, the fact

sheet was the most often used (used with 30.5%−52.7% of patients) and the disease images were the least used

(0.7%−6.1% of patients [excluding clinics that opted out of using disease images]). The fact sheet was used most

often with vaccine-hesitant parents (most of the

time = 58.5%, often = 27.7%) and parents who refused the

vaccine during the current visit (most of the time = 63.1%,

often = 21.5%). The fact sheet was used less often with

parents of unknown hesitancy (most of the time = 15.4%,

often = 10.8%) and those who had refused the vaccine at a

previous visit (most of the time = 38.5%, often = 23.1%).

Although the decision aid was not used as often as the fact

sheet, it followed a similar pattern of use (Tables 2 and 3).

At postintervention, average reported use of interven-

tion components was greater in private versus public clin-

ics (significant for all but disease images; Table 4). The

majority of providers who reported using the fact sheet or

MI said they were either somewhat or very likely to con-

tinue using them (fact sheet, 98.5%; MI, 91.3%).

In the postintervention survey, 78.2% of providers

reported using the presumptive approach to open the vac-

cine conversation. In addition, 89.6% of participants

(n = 69/77 responses) reported having used MI with

vaccine-hesitant parents at least once during the interven-

tion. Of those who reported having used MI, the most

commonly used technique was using open-ended ques-

tions; 55.1% (n = 38) of providers reported using this

strategy most of the time (>75%) with parents who were

vaccine hesitant.

TAGGEDH2USEFULNESS OF INTERVENTION COMPONENTS TAGGEDEND

Across all of the intervention components, providers

reported that the fact sheet and MI were most useful. For

the fact sheet, 97% of using providers reported that it was

very or somewhat useful at the postintervention assess-

ment. Similarly, 98.5% of using providers said MI was

very or somewhat useful. Compared with intervention

midpoint, by the end of the intervention, among those

who reported using the fact sheet or MI, both were per-

ceived as very useful by a larger proportion of participants

(fact sheet: 57.4% mid, 63.1% post; MI: 35.8% mid,

49.3% post). The decision aid was reported as least

useful—43% responded it was “not that useful” or “not at all useful” at the postintervention assessment.

Although not as many providers used the disease

images, decision aid, or referred parents to the Web site

(n = 20, disease images; n = 37, decision aid; n = 27, Web

site), they were reported as somewhat useful or very use-

ful at postintervention assessment by the majority of pro-

viders who used them (80%, 56.2%, and 66.7%,

respectively).

TAGGEDH2FACILITATORS AND BARRIERS TO INTERVENTION COMPONENT USE TAGGEDEND

Overall, workflow integration was reported as a promi-

nent barrier to using some of the intervention components,

including the decision aid and Web site. Very few pro-

viders reported that it was easy to integrate these items

into clinic workflow (25.0% and 9.4%, respectively,

Table 5). In contrast, a majority of participants agreed the

fact sheet (90.8%, n = 59) and MI (81.2%, n = 56) were

easy to integrate into the patient visit workflow. There

was also a significant difference (P < .001) between how

participants responded to items measuring ease of work-

flow integration across all intervention components. Fur-

thermore, a majority of providers reported that “other

tools were adequate to help communicate HPV vaccine

recommendations to parents” as a reason underlying

the low utilization of the decision aid, disease images,

and Web site (70.0%, 88.9%, and 65.6%, respectively).

However, fewer providers reported this same sentiment

in regard to the fact sheet and MI. There was a significant

difference (P < .001) between how participants

responded to this item across all intervention components.

In addition, very few participants (13.2%) disagreed with

the statement “I did not have enough MI training/practice

to feel comfortable using the techniques.” However, as

many as 46.8% and 71.1% identified lack of familiarity as

a reason for not using the decision aid and Web site,

respectively.

Table 1. Demographics Across Surveys

% (n) range

Respondents (of 108) 67%−78% (72−83) Respondents with an HPV

vaccine−eligible visit* 84%−92% (65−72)

Type of practice

Private 50%−52% (53−53) Public 48%−50% (48−54)

Sex

Female 75%−77% (76−82) Role

MD 75%−76% (76−81) DO 9%−10% (10−10) Nurse practitioner 15%−15% (15−16)

HPV indicates human papillomavirus; MD, doctor of medicine;

and DO, doctor of osteopathic medicine.

*Denominator is the number of respondents who responded to

that survey. All other items in Table 1 use everyone who was invited

to the survey (n = 108) as the denominator.

TAGGEDENDACADEMIC PEDIATRICS IMPROVING COMMUNICATION ABOUT HPV VACCINATION 885

TAGGEDH1DISCUSSIONTAGGEDEND Overall, results demonstrate that the fact sheet and MI

were the most-used intervention components throughout

the 12-month intervention period and thus likely the most

impactful components of our intervention. This conclu-

sion was further strengthened by the finding that providers

also reported that the fact sheet and MI were the most use-

ful of the intervention components. Additionally, in com-

parison with the other components, few providers

reported that other intervention tools were sufficient sub-

stitutes to using the fact sheet and MI, suggesting that

most providers felt that the fact sheet and MI were more

effective tools for successfully communicating with

parents about HPV vaccine recommendations.

Our finding that the fact sheet and MI were the most

used among all intervention components may be because

providers had the greatest familiarity with these compo-

nents at the study outset. The fact sheet was created via a

collaborative process with providers from each clinic.

Thus, many of the providers in the intervention group

were not only highly familiar with the fact sheet’s content

but also had a level of ownership in its creation. The MI

training was conducted over 3 sessions lasting a total of

2.5 hours, and this training occurred at the beginning of

the study period. 20

Lack of familiarity was reported as a

significant barrier to using the Web site, decision aid, and

disease images. Taken together, these results suggest that

familiarity with the intervention components increased

their use and perceived usefulness. Had additional train-

ing or collaborative efforts to develop or modify the Web

site, decision aid, or disease images been conducted dur-

ing the study, it is possible that the use of these may have

increased due to greater familiarity with the tools.

In addition to the fact sheet and MI being perceived as

the most useful of the intervention components overall,

results demonstrate that perceptions of usefulness for

these items also increased over time. This finding suggests

that as providers had more opportunities to use these tools

(ie, their familiarity with tools increased) their perceptions

of usefulness also increased. Similarly, even though not

used as frequently, among the small number of providers

that did use the disease images and the Web site, they

were perceived as useful. Thus, although familiarity with

a communication tool at the beginning of the intervention

(eg, the fact sheet and MI) may have determined initial

use and perceptions of usefulness of the various interven-

tion components, continued use of items did not diminish

perceptions of usefulness. Future studies on multicompo-

nent interventions should consider ensuring provider

familiarity throughout the study with all intervention

items as a means for increasing overall use.

TAGGEDH2BARRIERS TO IMPLEMENTATION TAGGEDEND

Across the various intervention components, many of

the reported barriers to use significantly differed between

the various items. Although the fact sheet and MI were

perceived as relatively easy to integrate into the clinic

workflow, the decision aid and Web site appeared to be

perceived as more challenging, whereas the disease

images presented their own unique set of barriers (a more

detailed exploration of reasons why providers did or did

not use intervention components can be found else-

where 23 ). A significant number of providers reported con-

cerns about the disease images that they might be too

explicit or perceived as coercive (by comparison, very

few providers perceived other intervention components as

potentially coercive). Thus, although this “scare-tactic”

strategy of communication was perceived as effective by

those who used the disease images, our findings suggest

that providers were concerned about potential boomerang

effects that may have prevented their more prolific use.

Table 2. Provider-Level Intervention Component Use: Percentage of Providers Reporting Use of Intervention Components at 1 or More

HPV Vaccine−Eligible Visits by Intervention Month

Intervention Component April 2015 (N = 78) July 2015 (N = 65) October 2015 (N = 72) January 2016 (N = 79)

Fact sheet 77% (n = 53) 79% (n = 51) 71% (n = 51) 64% (n = 42)

Decision aid 41% (n = 28) 46% (n = 30) 25% (n = 18) 17% (n = 11)

Disease images 15% (n = 9) 15% (n = 8) 3% (n = 2) 6% (n = 4)

iVac Web site recommended to parents 29% (n = 20) 42% (n = 27) 21% (n = 15) 15% (n = 10)

HPV indicates human papillomavirus.

Motivational interviewing not included, as its use was measured differently than other intervention components.

Table 3. Patient-Level Intervention Component Use: Percentage of Total Vaccine-Eligible Visits in Which Intervention Component Used by

Intervention Month

Intervention Component April 2015 July 2015 October 2015 January 2016

Fact sheet 45% (n* = 1040) 53% (n* = 1316) 49% (n* = 1346) 31% (n* = 1316)

Decision aid 20% (n* = 1040) 18% (n* = 1316) 13% (n* = 1346) 4% (n* = 1316)

Disease images 3% (n† = 824) 6% (n† = 1032) 0.7% (n† = 1047) 3% (n† = 1316)

iVac Website recommended to parents 12% (n* = 1040) 11% (n* = 1316) 6% (n* = 1346) 2% (n* = 1316)

*n = total number of reported vaccine eligible patients across all providers/clinics that month.

†n = disease images were not used in all clinics; thus, n reflects the total number of vaccine-eligible patients only in clinics in which partici-

pants reported having access to the disease images.

T AGGEDEND886 RENO ET AL ACADEMIC PEDIATRICS

Thus, it remains unclear how effective scare tactics are in

motivating parents to vaccinate.

TAGGEDH2LIMITATIONS TAGGEDEND

A major limitation of this study is the relatively small

sample size. Although we sampled the majority of the

providers in the participating clinics, the overall number

was low (n = 108 total), meaning that differences in out-

comes were limited in their power to detect significance.

A second limitation is that intervention component use

was self-reported, rather than observed, and therefore

may have been impacted by recall bias, as were reports of

barriers and usefulness. Frequent surveying throughout

the study period was used in an attempt to mitigate this

potential limitation.

In addition, although the RCT trial design should have

evenly distributed any other factors that were not

Table 4. Average Use of Intervention Components. Postinterven-

tion Survey (N = 79)

n Mean SD P Value

Fact sheet .001

Private 36 0.41 0.34

Public 30 0.15 0.27

Decision aid .01

Private 36 0.10 0.23

Public 30 0.01 0.02

Disease images .27

Private 36 0.03 0.17

Public 30 0.01 0.01

Web site .03

Private 36 0.06 0.12

Public 30 0.01 0.04

SD indicates standard deviation.

Table 5. Facilitators and Barriers to Intervention Component Use

Fact Sheet Decision Aid Disease Images Web Site MI

Survey Items % Agreed by Intervention Component (n)

Facilitators

The amount of information on

the ____ was just right.*

98.5% (66) 78.0% (32) − 91.3% (21) −

It was easy to integrate the use

of the ____ into the patient

visit work-flow. †

90.8% (59) 25.0% (11) 53.3% (16) 9.4% (3) 81.2% (56)

There was enough time for the

parents to read the _____

during the scheduled visit.*

53% (35) 33.3% (15) 43.7% (14) 9.4% (3) 59.4% (41)

Barriers

The information on the ___ was

too complicated.*

6.1% (4) 31.0% (13) − 18.2% (2) −

I was concerned that the

______ would raise questions

among parents who were not

likely to be HPV vaccine

hesitant.

14.1% (9) 14.6% (6) − 0% (0) −

I was concerned that HPV

vaccine-hesitant parents

would be even more hesitant

after reading the ________.*

7.6% (5) 16.7% (7) 22.6% (7) 0% (0) 2.9% (2)

I saw little benefit in sending the

______ home with parents

who refused the vaccine.*

11.1% (7) 37.0% (17) − 20% (5) −

I was concerned that the

______ would feel coercive to

some parents.†

6.2% (4) 17.8% (8) 53.1% (17) 0% (0) 7.4% (5)

Other tools were adequate to

help communicate HPV vac-

cine recommendations to

parents (______ was not

needed). †

25.8% (17) 70.0% (35) 88.9% (32) 65.6% (19) 40.6% (28)

Disease images were too

explicit

− − 55.6% (20) − −

I did not have enough MI

training/practice to feel com-

fortable using the techniques.

− − − − 13.2% (9)

MI indicates motivational interviewing; HPV, human papillomavirus.

Survey items were adapted as appropriate for each intervention component; 42.1% reported that the disease images were not used in

their office. Dashes (−) indicate instances in which survey items were not applicable to the indicated intervention components. *P < .05 for differences across intervention components.

†P < .001 for differences across intervention components.

TAGGEDENDACADEMIC PEDIATRICS IMPROVING COMMUNICATION ABOUT HPV VACCINATION 887

measured but may have led to increases in vaccination

rates, it is possible with only 16 clinics in the trial that

some other influential but unmeasured factors could have

actually been unbalanced. Furthermore, provider-level

vaccination rate data were not available to compare with

individual provider intervention component use. Thus,

determinations that specific intervention component use

led to greater vaccination rates are speculative.

An additional limitation is that because the clinics

involved in the original RCT had to agree to participate in

the study, this could introduce biases that impact the

conclusions, for example, if clinics involved in the study

had greater levels of initial interest in improving their

vaccination rates than average. However, the main goal of

this study was not to provide generalizable information

about which intervention components would work in

every office but rather to begin to understand what spe-

cific components of the larger intervention were likely

most influential on the demonstrated improved HPV

vaccine rate results.

TAGGEDH1CONCLUSIONSTAGGEDEND Overall, the current study suggests that the fact sheet

and MI components of our multi-item provider HPV vac-

cine communication intervention were likely critical ele-

ments underlying the effectiveness of the intervention in

our larger randomized trial. However, although the other

intervention components were less used, when they were

used, they were perceived as useful. Intervention compo-

nent use overall was significantly lower in public clinics

compared with private ones, likely explaining differences

in the intervention effectiveness found in the larger trial. 14

Future research should examine the impact of a pared-

down intervention consisting of only the communication

training and fact sheet to determine whether these ele-

ments confer similar effectiveness to the larger, more

complex 5-component intervention originally tested.

Communication-based interventions such as these provide

a promising mechanism for addressing gaps in HPV vac-

cination among adolescents.

TAGGEDH1ACKNOWLEDGMENTS TAGGEDEND

Financial disclosure: The funding source for this study was the Cen-

ters for Disease Control and Prevention (5U01IP000801), who provided

input into the main study (randomized controlled trial) design and data

interpretation. The funding source had no direct involvement in the prep-

aration of the manuscript.

TAGGEDH1REFERENCESTAGGEDEND

1. Centers for Disease Control and Prevention. CDC Human Papillo-

mavirus (HPV) and Cancer. HPV and Cancer. 2017. Available at:

https://www.cdc.gov/cancer/hpv/. Accessed October 2, 2017 .

2. Walker TY, Elam-Evans LD, Singleton JA, et al. National, regional,

state, and selected local area vaccination coverage among adoles-

cents aged 13−17 years—United States, 2016. MMWR Morb Mortal Wkly Rep. 2017;66:874–882.

3. US Department of Health and Human Services. Healthy People

2020. Washington, DC: Office of Disease Prevention and Health

Promotion; 2010.

4. Edwards KM, Hackell JM. Committee on Infectious Diseases, the

Committee on Practice and Ambulatory Medicine. Countering vac-

cine hesitancy. Pediatrics. 2016;138:e20162146.

5. Brown B, Gabra MI, Pellman H. Reasons for acceptance or refusal

of human papillomavirus vaccine in a California pediatric practice.

Papillomavirus Res. 2017;3:42–45.

6. Darden PM, Thompson DM, Roberts JR, et al. Reasons for not vac-

cinating adolescents: national immunization survey of teens,

2008−2010. Pediatrics. 2013;131:645–651. 7. Dempsey AF, Lockhart S, Campagna EJ, et al. Providers’ time spent

and tools used when discussing the HPV vaccine with parents of

adolescents. Vaccine. 2016;34:6217–6222.

8. Finney Rutten LJ, St. Sauver JL, Beebe TJ, et al. Association of both

consistency and strength of self-reported clinician recommendation

for HPV vaccination and HPV vaccine uptake among 11- to

12-year-old children. Vaccine. 2017;35:6122–6128.

9. Clark SJ, Cowan AE, Filipp SL, et al. Parent perception of provider

interactions influences HPV vaccination status of adolescent

females. Clin Pediatr (Phila). 2016;55:701–706.

10. Dorell C, Yankey D, Kennedy A, et al. Factors that influence paren-

tal vaccination decisions for adolescents, 13 to 17 years old: national

immunization survey teen, 2010. Clin Pediatr (Phila). 2013;52:

162–170.

11. Kester LM, Zimet GD, Fortenberry JD, et al. A national study of

HPV vaccination of adolescent girls: rates, predictors, and reasons

for non-vaccination. Matern Child Health J. 2013;17:879–885.

12. Bratic JS, Seyferth ER, Bocchini JA. Update on barriers to human

papillomavirus vaccination and effective strategies to promote vac-

cine acceptance. Curr Opin Pediatr. 2016;28:407–412.

13. Vadaparampil ST, Malo TL, Sutton SK, et al. Missing the target for

routine human papillomavirus vaccination: consistent and strong

physician recommendations are lacking for 11-12 year old males.

Cancer Epidemiol Biomarkers Prev. 2016;25:1435–1446.

14. Dempsey AF, Pyrznawoski J, Lockhart S, et al. Effect of a health

care professional communication training intervention on adolescent

human papillomavirus vaccination: a cluster randomized clinical

trial. JAMA Pediatr. 2018;172. e180016-e180016.

15. Community Preventive Services Task Force. The Guide to Commu-

nity Preventive Services 2015. Atlanta, GA: The Community Guide;

2015.

16. Dempsey A, Maerten J, Beaty BL, et al. Utilization of an in-clinic,

iPad-based intervention to provide parents with personalized informa-

tion about adolescent vaccines. J Adolesc Health. 2015;56:S7–13.

17. Opel DJ, Heritage J, Taylor JA, et al. The architecture of provider-

parent vaccine discussions at health supervision visits. Pediatrics.

2013;132:1037–1046.

18. Brewer NT, Hall ME, Malo TL, et al. Announcements versus con-

versations to improve HPV vaccination coverage: a randomized

trial. Pediatrics. 2017;139:e20161764.

19. Rollnick S, Miller WR, Butler CC. Motivational Interviewing in

Health Care: Helping Patients Change Behavior. New York, NY:

Guilford Press; 2007.

20. Reno JE, O’Leary S, Garrett K, et al. Improving provider

communication about HPV vaccines for vaccine-hesitant parents

through the use of motivational interviewing. J Health Commun.

2018;23:313–320.

21. Malo TL, Hall ME, Brewer NT, et al. Why is announcement training

more effective than conversation training for introducing HPV vac-

cination? A theory-based investigation. Implement Sci. 2018;13:57.

22. Colorado Children’s Immunization Coalition. About the Alliance

for HPV Free Colorado. HPV Inf Provid. 2018. Available at:. https://

www.childrensimmunization.org/healthcare-professionals/hpv-

information. Accessed July 12, 2018 .

23. Lockhart S, Barnard J, O’Leary ST, et al. Provider and parent per-

spectives on enhanced communication tools for HPV vaccine-hesi-

tant parents. Acad Pediatr. 2018;18:776-782.

TAGGEDEND888 RENO ET AL ACADEMIC PEDIATRICS

©2018 Elsevier