Abstract
Strategies to increase adolescent immunization rates have been suggested, but little is documented about which strategies clinicians actually use or would consider. In spring 2010, we surveyed primary care physicians from 2 practice-based research networks (PBRNs): Greater Rochester PBRN (GR-PBRN) and national pediatric COntinuity Research NETwork (CORNET). Network clinicians received mailed or online surveys (response rate 76%, n = 148). The GR-PBRN patient population (51% suburban, 33% rural, and 16% urban) differed from that served by CORNET (85% urban). For nonseasonal vaccines recommended for adolescents, many GR-PBRN and CORNET practices reported using nurse prompts to providers at preventive visits (61% and 52%, respectively), physician education (53% and 53%), and scheduled vaccine-only visits (91% and 82%). Strategies not used that clinicians frequently indicated they would consider included patient reminder/recall and prompts to providers via nurses or electronic health records. As preventive visits and immunization recommendations grow more complex, using technology to support immunization delivery to adolescents might be effective.
Introduction
Recommendations for adolescent immunization have evolved substantially since 2005, creating challenges and opportunities for the effective delivery of vaccines to adolescents. The American Academy of Pediatrics, the American Academy of Family Physicians, and the federally appointed Advisory Committee on Immunization Practices jointly recommend that adolescents routinely receive tetanus toxoid, reduced diphtheria toxoid, and acellular pertussis (Tdap) vaccine; quadrivalent meningococcal conjugate (MCV4) vaccine; and human papillomavirus (HPV) vaccine as well as annual seasonal influenza vaccine. 1 The National Immunization Survey–Teen conducted in 2011 describes the most recent overall US immunization coverage among adolescents aged 13 through 17 years: ≥1 dose Tdap since age 10 years, 78.2%; ≥1 dose meningococcal vaccine, 70.5%; and, among girls, ≥1 dose of HPV vaccine, 53.0%; and ≥3 doses of HPV vaccine, 34.8%. 2 Although better than in previous years, adolescent immunization coverage has not reached the levels of early childhood vaccines and US goals. 3 Seasonal influenza vaccination rates are also lower for adolescents than for younger children (ie, 33.7% among teens aged 13-17 years, compared with 74.6% among those aged 6-23 months). 4 Thus, despite recommendations, adolescent immunization coverage is suboptimal.
In 2008, the Working Group on Adolescent Vaccination in the Medical Home, convened by the Centers for Disease Control and Prevention (CDC), published a summary of barriers to adolescent immunization. Barriers were noted at 3 levels: the adolescent or family, health care providers, and the health care system. 5 Adolescent or family-level barriers include the observation that some older adolescents come to preventive care visits without a parent or guardian who has authority to give consent for immunization. Also, adolescents and their parents may lack knowledge about needed vaccinations or may not realize that it is generally safe to receive vaccinations during visits for acute problems as well as visits for preventive health care. Barriers at the provider level include insufficient time during preventive visits to discuss immunizations adequately, given the number of other issues that also need to be addressed. 6 At the system level, the preventive health care schedule includes fewer visits for adolescents than for younger children, providing fewer opportunities for immunization “catch up” if a visit or an immunization is missed. 7 This is particularly important because many adolescents do not receive regular recommended preventive care visits.8,9
The Task Force on Community Preventive Services outlined strategies to enhance practice-based immunizations, such as patient reminder/recall and provider point-of-care prompts.10,11 However, most of these strategies reflect immunization delivery research targeting young children and adults; findings might not be applicable to adolescents. Also, the research was done before recent, rapid expansions of electronic health record (EHR) functionality. 12 The strategies recommended in the guide should be evaluated not only in the context of clinical care for adolescents but also with respect to the degree of buy-in from practitioners who grapple routinely with the complexities of office-based immunization.
To influence physician behavior in ways that will increase adolescent immunization rates within real-world practice settings, it is first essential to understand clinicians’ intentions and skills. 13 The goal of this study was to assess clinicians’ experience with and perceptions of the feasibility, sustainability, and generalizability of potential strategies to increase immunization rates among adolescents affiliated with medical homes. We studied the perceptions of primary care clinicians in a broad array of practices, with diversity of patient populations and health care coverage, to determine whether their assessments of potential immunization strategies would differ. The specific objectives of this study were to survey primary care physicians serving a wide range of adolescents to (1) identify interventions to increase practice-based immunization coverage that were likely to be effective, feasible, sustainable, and generalizable in primary care settings nationally and (2) compare the types of options suggested by health care providers in 2 practice-based research networks (PBRNs) with differing characteristics, including EHR use.
Methods
We obtained approval for this project from the Research Subjects’ Review Board of the University of Rochester.
Recruitment of Survey Participants
Because we sought responses from a wide variety of primary care physicians, we recruited participants through 2 primary care PBRNs. The Greater Rochester PBRN (GR-PBRN) is composed of primary care physicians who serve suburban, rural, and urban patients in a multicounty area near Rochester, New York. For this study, this original cohort (63 pediatric and family medicine practices) was enhanced by the addition of 59 other pediatric and family medicine practices in the western New York area that expressed willingness to participate in past immunization surveys. All together the GR-PBRN denominator included 122 practices.
The COntinuity Research NETwork (CORNET) is a well-established, national PBRN of pediatric resident continuity practices that is supported by the Academic Pediatric Association (APA). 14 CORNET practices engage in research on diverse topics, including health care issues of underserved children, other health disparities, and resident education. CORNET clinicians include pediatric faculty, pediatric residents, and nurse practitioners who provide care to a large proportion of generally urban, low-income, publicly insured children mostly in academic medical center outpatient pediatric clinics. 15 CORNET study activities take place solely within academic sites where residents are trained in continuity clinics; in some institutions, faculty participate in study activities not only through their oversight of residents’ clinical care but also through caring for their own patients. CORNET site investigators are the faculty leadership for CORNET studies. At the time of this study, 73 active training programs were enrolled in CORNET.
Survey Tool
The 20-question survey focused on delivery of 2 categories of vaccines recommended for adolescents—nonseasonal vaccines (defined for this study as Tdap, MCV4, and HPV vaccines) and seasonal influenza vaccine. For each category, the survey presented a list of 9 groups of potential strategies (gleaned from a literature review) that might be used in practices to promote adolescent immunization. Clinicians were asked to indicate which strategies (1) were already being used by their practices and (2) if not being used, which strategies they would consider implementing in their practices. Survey questions included demographic information (eg, practice type and estimated race/ethnicity of patient population) as well as information on practices’ uses of EHRs.
Procedure
During May to August 2010, we surveyed physicians from the 2 primary care PBRNs. For the GR-PBRN, we mailed surveys to 1 physician network liaison from each practice along with a $5 gift card. We sent 3 follow-up mailings at 2-week intervals until a reply was received. If there was still no response, we placed calls to determine if the liaison would prefer to receive the survey by fax or email instead of ground mail; the preferred mode was then implemented. In total, we sent up to 5 rounds of the survey.
For the CORNET sites, we posted an invitation on the CORNET listserv asking site investigators to complete the survey online. Following the initial listserv postings, we sent up to 5 rounds of emails directly to site investigators reminding them to complete the survey. Once site investigators completed the surveys, they were able to opt in to receive a $5 gift card, which was mailed to them.
Data Analysis
We used Pearson χ2 tests and the Mann-Whitney U tests to determine differences between the 2 PBRNs, for categorical and nonnormal continuous variables, respectively. We compared the nonseasonal adolescent vaccinations and influenza vaccinations using McNemar’s test. All analyses were done using Stata 11.2. 16
Results
Participants
For the GR-PBRN, the response rate was 72.1% (88 of 122 practices); for CORNET, the response rate was 82.2% (60 of 73 active practices). Thus, the combined response rate was 75.9% (148 of 195 practices). Characteristics of the respondents’ practices are shown in Table 1. As expected by design of the PBRNs, GR-PBRN practices primarily served suburban (51.1%) or rural (33.0%) patients, and CORNET practices predominantly served urban (85.4%) patients. GR-PBRN practices had a higher proportion of patients who were white and had private insurance and a lower proportion of black and Hispanic/Latino patients compared with CORNET practices. Most practices in both networks participated in the Vaccines for Children program (P = .14); 60% of GR-PBRN and all CORNET practices were pediatric practices. GB-PBRN practices tended to be smaller; the mean number of patients was 5819 for GR-PRBRN practices and 10 446 for CORNET practices.
Demographic Characteristics of Respondents’ Practices by Practice-Based Research Network.
Abbreviations: GR-PBRN, Greater Rochester Practice-Based Research Network; CORNET, National Pediatric COntinuity Research NETwork; SD, standard deviation.
P values from Pearson χ2 test.
Not all respondents’ answers summed to 100%.
P values from 2-sample Wilcoxon rank-sum test.
Outcomes
Nonseasonal Adolescent Immunizations
For practices in each PBRN, Table 2 shows the 9 listed groups of strategies and the proportion of respondents’ practices that (1) already were using the strategies or (2) among those not using, the proportion that would consider using listed strategies to increase coverage with nonseasonal adolescent immunizations (Tdap, MCV4, and HPV vaccines). With respect to strategies that practices already were using, 3 findings are important. First, all the practices were using at least 1 strategy for nonseasonal adolescent vaccines. More than half of both GR-PBRN and CORNET practices reported currently using nurse prompts to providers at preventive visits (60.9% and 52.5%, respectively), physician education (53.4% and 53.3%), and scheduled vaccine-only visits (90.9% and 81.7%). Vaccine-only visits were the most commonly used strategy for both PBRNs. Second, only about one-third of GR-PBRN and less than one-tenth of CORNET practices were using patient reminder/recall specific to immunizations. Third, when compared with GR-PBRN practices, CORNET practices were more likely already to use audit and feedback (ie, review of a sample of patient charts with feedback to providers).
Abbreviations: GR-PBRN, Greater Rochester Practice-Based Research Network; CORNET, National Pediatric COntinuity Research NETwork; EHR, electronic health record; tech, technician; f/u, follow-up; CME, Continuing Medical Education; CNE, Continuing Nurse Education.
Denominators varied for questions because of small numbers of clinicians not completing every question. For each cell in this table, the number of respondents providing data is shown as the denominator. Overall, for the GR-PBRN, the survey response rate was 72% (88 of 122 practices); for CORNET, the response rate was 82% (60 of 73 active practices).
Nonseasonal vaccines were defined to include tetanus toxoid, reduced diphtheria toxoid, and acellular pertussis (Tdap) vaccine; quadrivalent meningococcal conjugate (MCV4) vaccine; and human papillomavirus (HPV) vaccine.
P values from Pearson χ2 test.
When asked which strategies practices were “not using but would consider using,” more than half of both GR-PBRN and CORNET physicians reported practices would consider many specific strategies, including patient reminder/recall and health care provider prompts. By network, the most commonly chosen strategies that clinicians would consider were patient reminder/recall about vaccines needed (72.7% GR-PBRN) and health care provider prompts by staff at preventive visits (92.9% CORNET). CORNET clinicians (when compared with GR-PBRN clinicians) were more likely to be willing to consider most strategies. Differences were statistically significant for many strategies, including provider prompts, vaccine clinics, and audit/feedback. Of note, although some differences between PBRNs appeared to be relatively large, the differences sometimes did not reach statistical significance because the numbers of practices not already using a particular strategy were small.
Table 3 shows a similar analysis, but instead of comparing practices in the 2 PBRNS, it compares practices that did (n = 88) or did not (n = 60) have an EHR (ie, for both PBRNs combined). Other than differences in use or consideration of EHR prompts, the practices differed only in that practices with EHR, compared with those without EHR, were less likely to give patients paper educational materials and were more likely to consider staff education as a way to increase adolescent nonseasonal immunization rates.
Abbreviations: GR-PBRN, Greater Rochester Practice-Based Research Network; CORNET, National Pediatric COntinuity Research NETwork; EHR, electronic health record; tech, technician; f/u, follow-up; CME, Continuing Medical Education; CNE, Continuing Nurse Education.
Denominators varied for questions because of small numbers of clinicians not completing every question. For each cell in this table, the number of respondents providing data is shown as the denominator. Overall, for the GR-PBRN, the survey response rate was 72% (88 of 122 practices); for CORNET, the response rate was 82% (60 of 73 active practices).
Nonseasonal vaccines were defined to include tetanus toxoid, reduced diphtheria toxoid, and acellular pertussis (Tdap) vaccine; quadrivalent meningococcal conjugate (MCV4) vaccine; and human papillomavirus (HPV) vaccine.
P values from Pearson χ2 test.
For the GR-PBRN, we compared the responses of pediatricians (PEDs; n = 53) and family physicians (FPs; n = 35). The proportion of respondents using specific strategies for Tdap, MCV4, and HPV vaccines varied by specialty in 2 cases: pediatric practices were more likely already to be using patient reminder/recall about preventive care visits needed (73.6% PEDS, 37.1% FPs; P < .001) but less likely to be using EHR-based prompts to health care providers to give specific vaccines at the time of the visit (13.5% PEDs, 34.3% FPs; P = .02). Among those not using a strategy, the proportion of respondents who would consider that strategy for Tdap, MCV4, and HPV vaccines varied by specialty in a single case: pediatric practices were more likely to consider using patient education via electronic means such as videos, emails, and Web sites (51.0% PEDs, 24.2% FPs; P = .02).
Influenza Immunization
By PBRN, Table 4 shows the proportion of respondents’ practices that (1) already were using or (2) (among those not using the strategy) would consider using each of 9 listed groups of strategies to increase coverage with influenza immunization. More than half of both GR-PBRN and CORNET practices reported currently using 2 strategies: influenza vaccine clinics (73.6% and 66.1%, respectively) and vaccine-only visits (87.5% and 85.0%) scheduled during regular office hours. The latter was the most commonly used strategy for both PBRNs. Although the strategies in use were similar between practices in the different PBRNs, again, when compared with CORNET, GR-PBRN practices were more likely to be already using patient reminder/recall about vaccines needed; CORNET practices were more likely to be already using audit and feedback as well as staff education.
Proportion of Respondents’ Practices a Using or Considering Strategies for Seasonal Influenza Vaccine by Practice-Based Research Network.
Abbreviations: GR-PBRN, Greater Rochester Practice-Based Research Network; CORNET, National Pediatric COntinuity Research NETwork; EHR, electronic health record; tech, technician; f/u, follow-up; CME, Continuing Medical Education; CNE, Continuing Nurse Education.
Denominators varied for questions because of small numbers of clinicians not completing every question. For each cell in this table, the number of respondents providing data is shown as the denominator. Overall, for the GR-PBRN, the survey response rate was 72% (88 of 122 practices); for CORNET, the response rate was 82% (60 of 73 active practices).
P values from Pearson’s χ2 test.
All but 2 practices (both GR-PBRN, representing 2% of the network’s practices) were using at least 1 strategy for influenza immunization. Only 1 of these 2 practices would consider future implementation of at least 1 strategy.
More than half of the respondents from both GR-PBRN and CORNET practices reported that although they were not currently using patient reminder/recall and EHR prompts to health care providers for influenza vaccinations, they would consider these strategies. The most commonly chosen strategies to implement—that were not already in use—were prompts to the health care provider by EHR (63.8% GR-PBRN) or staff (92.6% CORNET). Again, CORNET clinicians (when compared with GR-PBRN clinicians) were more likely to be willing to consider every strategy; as with nonseasonal vaccines, the difference reached statistical significance for many strategies.
Table 5 shows a similar analysis, combining PBRNs and comparing practices that did or did not have an EHR. Other than differences in use or consideration of EHR prompts, the groups were different only in that practices with EHR, compared with those without EHR, were far more likely to use standing orders for influenza vaccination (47.6% vs 29.3%, P = .03) and were more likely to consider staff education as a way to increase adolescent influenza immunization rates.
Proportion of Respondents’ Practices a Using or Considering Strategies for Seasonal Influenza Vaccine by Practice Use of EHR (88 Users, 60 Nonusers).
Abbreviations: GR-PBRN, Greater Rochester Practice-Based Research Network; CORNET, National Pediatric COntinuity Research NETwork; EHR, electronic health record; tech, technician; f/u, follow-up; CME, Continuing Medical Education; CNE, Continuing Nurse Education.
Denominators varied for questions due to small numbers of clinicians not completing every question. For each cell in this table, the number of respondents providing data is shown as the denominator. Overall, for the GR-PBRN the survey response rate was 72% (88 of 122 practices); for CORNET the response rate was 82% (60 of 73 active practices).
P values from Pearson χ2 test.
The proportion of respondents using specific strategies for influenza vaccine varied by specialty in a few cases among the GR-PBRN practices: PEDs were less likely to use EHR prompts to health care providers (to give influenza vaccine) both at preventive visits (15.4% PEDs, 37.1% FPs, P = .02) and at illness or follow-up visits (9.6% PEDs, 34.3% FPs, P = .004) and to use standing orders (30.2% PEDs, 51.4% FPs, P = .05). However, PEDs were more likely than FPs to use vaccine clinics on evenings and weekends during regular office hours (56.6% PEDs, 34.3% FPs, P = .04). Among those not using a strategy, the proportion of respondents who would consider that strategy for influenza vaccine did not vary significantly by specialty group.
Nonseasonal Adolescent Versus Seasonal Influenza Immunization (Results Not Shown)
Both GR-PBRN and CORNET practices were significantly more likely (P < .05) to be using certain strategies for influenza compared with nonseasonal adolescent vaccinations: patient reminders, vaccine clinics during regular office hours (daytime and other), and standing orders. GR-PBRN practices, but not CORNET practices, were significantly more likely (P < .05) to use several strategies for nonseasonal adolescent vaccinations, including EHR and staff prompts at preventive visits, patient educational materials, physician and staff education, and audit and feedback.
Discussion
In this study conducted among primary care physicians belonging to 2 PBRNs with differing characteristics, almost all clinician respondents reported that their practices used at least 1 strategy to increase coverage among adolescents for recommended nonseasonal vaccines and at least 1 strategy to promote seasonal influenza vaccine coverage. Vaccine-only visits comprised the most commonly used strategy for nonseasonal and seasonal immunization. Many practices would consider additional strategies to enhance immunization delivery. For nonseasonal adolescent vaccines, most practices were already using nurse prompts to providers at preventive visits, physician education, and scheduled vaccine-only visits. Strategies not currently used that clinicians frequently would consider included patient reminder/recall about vaccines or well visits and EHR or nurse prompts to providers. Respondents from both PBRNs were significantly more likely to be using certain strategies for influenza vaccination than for nonseasonal adolescent vaccinations; these included patient reminder/recall, vaccine clinics during regular office hours (daytime and other), and standing orders. Although the reason for this was not addressed in our study, it may be that offices used these additional measures because of differences in the health care system (eg, quality measures), office system (eg, the need to deliver all the vaccine purchased before it expires), or patients (eg, the need to put special effort forth to overcome low influenza immunization rates).
The Task Force on Community Preventive Services, an independent body of volunteer public health and prevention experts appointed by the Director of the CDC, developed The Community Guide to Preventive Services. The guide includes a systematic review of the available evidence supporting the use of interventions to increase the use of universally recommended vaccinations. 11 Interventions are grouped into categories, and each category includes at least 1 strategy applicable to primary care offices. For example, the guide suggests that primary care offices may enhance access to vaccination services by reducing client out-of-pocket costs (eg, by participating in the Vaccines For Children program 17 ), increasing community demand for vaccinations by sending client reminder/recall messages for vaccinations,18-20 and implementing provider-based interventions such as provider reminders 21 (eg, prompts via EHRs or nurse staff). Our study demonstrates that a high proportion of surveyed practices are currently using at least 1 strategy to promote immunization of adolescents.
A limitation of the Task Force recommendations is that the reviews are based on evidence that does not uniformly include studies of adolescent vaccination. The Working Group on Adolescent Vaccination in the Medical Home convened by CDC suggested that strategies recommended for childhood and adult vaccinations should be implemented for adolescents if evaluation shows them to be successful for this population and that medical homes should perform quality improvement projects to enhance delivery of adolescent vaccinations. The Working Group concluded, “By incorporating evidence-based strategies and coordinating effectively with other health care sites used by adolescents, medical homes will be the pivotal settings for the delivery of adolescent vaccinations.”5(pS22) We surveyed primary care practices that serve as the medical homes for adolescents. Our results are promising in that, with 1 exception, surveyed practices that were not currently using strategies were willing to consider new strategies to improve adolescent immunization delivery.
Recommendations for practice-based strategies are not easy to implement in busy primary care practices that have multiple, competing priorities. In this study, less than one-third of all practices used reminder/recall for routinely recommended vaccines, and half or fewer used it for influenza vaccination. Yet most respondents, particularly those affiliated with practices using EHRs, indicated a willingness to consider patient reminder/recall. Recently, studies have documented the challenges to patient reminder/recall for childhood vaccinations, including resources needed, cost, complex algorithms, and difficulties in reaching patients. 22 However, evidence has recently been published supporting the effectiveness of reminder/recall among adolescents in private practices and school-based health centers in Denver.23,24 Promising strategies include EHR-based reminder/recall systems, especially those using text messaging 25 or centralized reminder/recall 26 that would not require each practice to re-create a patient reminder/recall system.
Our study also highlights health care providers’ interest in point-of-care prompts. Use of these strategies has been shown to be effective in some settings, including influenza vaccination of children. 27 Although only a minority of respondents’ practices used provider prompts, the majority of those that did not use them would consider adopting them. Among the GR-PBRN, interest in using provider prompts was particularly high for practices that had EHRs. Theoretically, EHRs have the potential to mechanize provider prompts if the EHR contains accurate immunization information and algorithms to identify which patients are eligible for vaccinations. In fact, such “alerts” are part of meaningful use designations. Furthermore, state immunization information systems (previously known as “registries”) could be used either to link with EHRs or to deliver provider prompts. 28 Provider prompts represent another promising strategy for adolescent immunization delivery, but studies are needed to evaluate their effectiveness. 29
In this study, we compared responses from 2 different PBRNs: GR-PBRNs, which included PEDs and family medicine physicians primarily serving suburban and rural patients in a single region of New York, and CORNET, which was composed exclusively of PEDs practicing in urban academic centers throughout the United States. The proportion of respondents using specific strategies for Tdap, MCV4, and HPV vaccines varied by PBRN in 4 cases: CORNET practices were less likely to already use patient reminder/recall about vaccines needed or preventive care visits needed and lists of patients in need of vaccination but were more likely to use audit/feedback. Similar results were found for influenza vaccination. Among those not using a strategy, CORNET practices were more willing to consider many new strategies for adolescent immunization despite the complexities of their being in academic medical centers and including resident teaching clinics.
Limitations
This study’s limitations include the following: (1) The respondents were from only 2 PBRNs. The GR-PBRN is geographically based and, thus, might not be representative of practices in other parts of the country. CORNET practices, although located across the United States, include only academic pediatric centers with continuity clinics—a very different, but limited, setting for health care services. (2) The survey was sent in early 2010 when practices were making rapid changes in their uses of health information technology. Repeating the survey in the future might show many more practices using, or willing to consider using, EHR-based prompts. Similarly, advances in computer-generated patient reminder/recall through low-cost media (eg, text messages) might alter primary care use of patient reminders in the near future.30,31 (3) The summary presented here is based on self-reports from physicians; there was no in-office validation or direct observation of strategies reported as currently used.
Conclusions
A very high proportion of physician respondents reported using at least 1 strategy to increase coverage of seasonal and nonseasonal vaccines among adolescents receiving care in their practices. Furthermore, to enhance the delivery of these vaccines, many would consider additional strategies, especially health information technology–based approaches such as patient reminder/recall and prompts to health care providers. As primary care preventive visits and immunization recommendations grow more complex, using technological advances to support immunization delivery to adolescents might be effective and needed. Future evaluations of efforts to increase rates of adolescent immunizations would be important additions to the medical literature.
Footnotes
Acknowledgements
The authors express their thanks to the primary care physicians from the Greater Rochester Practice Based Network and the COntinuity Research NETwork (CORNET) who generously gave of their time to participate in this study.
Declaration of Conflicting Interests
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Dr. Humiston is a consultant to the Immunization Action Coalition, a not-for-profit organization that works to increase immunization rates and prevent disease by creating and distributing educational materials for health professionals and the public that enhance the delivery of safe and effective immunization services.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Centers for Disease Control and Prevention (CDC), National Center for Immunizations & Respiratory Diseases, Atlanta, GA (Grant 1U01P00012), and the CTSI (Grant UL RR024160).