Tailored Interventions to Introduce Influenza Vaccination

November 1, 2005
Mary Patricia Nowalk, PhD, RD

,
Chyongchiou J. Lin, PhD

,
Richard Kent Zimmerman, MD, MPH

,
Judith A. Troy, MS

,
Alejandro Hoberman, MD

,
Diana H. Kearney, RN

,
Stephanie M. Cleary, BA

Volume 11, Issue 11

Background: Because of high rates of hospitalization forinfluenza infections among very young children (< 2 years), theAdvisory Committee on Immunization Practices initiated a newpolicy in 2002 that encouraged vaccination of healthy childrenaged 6 to 23 months against influenza.

Objective: To evaluate the effectiveness of implementing tailoredinterventions to introduce influenza vaccination of children6 to 23 months of age in inner-city practices.

Study Design: A before-after trial with historical and concurrentcontrols was conducted in 6 health centers in low-income urbanlocations.

Methods: Intervention sites were selected, and interventions(directed at 1534 patients who were 6 to 23 months old) wereimplemented from a menu of strategies. Vaccination rates weremeasured from medical record reviews. Focus groups of nursingstaffs provided evaluative information on strategies.

P

Results: Influenza vaccination rates improved significantly atthe intervention health centers compared with the control center.Preintervention (2001-2002) rates ranged from 0% to 7.6%,and intervention (2002-2003) rates ranged from 15.2% to 49.2%(< .001). The number of interventions ranged from 6 to 11. Sitesthat used more interventions (odds ratio, 1.24; 95% confidenceinterval, 1.15-1.34) and had staff support of the vaccination effort(odds ratio, 1.91; 95% confidence interval, 1.40-2.60) had highervaccination rates.

Conclusions: Tailored interventions resulted in successful introductionof influenza vaccination of 6-to 23-month-old children ininner-city health centers. More strategies and enthusiastic staff supportmay result in higher vaccination rates.

(Am J Manag Care. 2005;11:717-724)

Influenza disproportionately affects the very youngand the very old. Young children are hospitalized forinfluenza-related illness at rates similar to thoseamong older adults. Among children, the rates rangefrom 104 hospitalizations per 10 000 children per yearfor those younger than 6 months to 19 hospitalizationsfor those aged 1 to 3 years.1 Among adults 65 years andolder, a rate of 17.4 hospitalizations per 10 000 adultsper year has been reported.2 A more recent study3found that 70% of 182 children hospitalized forinfluenza infections were younger than 2 years old.Hospitalization rates among children aged 0 to 4 yearswith high-risk medical conditions are much higher(approximately 50 hospitalizations per 10 000 childrenper year) than those of children without high-risk conditions.2 Fortunately, case-fatality rates from influenzaamong very young children are low. However, influenzaattack rates are 10% to 40% among preschool-age children,4-6 putting older persons and persons with chronicmedical conditions at high risk of infection throughfamily transmission of the disease.

Among the estimated 8 million children aged 6months to 17 years with high-risk medical conditions,2influenza vaccination rates are disappointingly low. Forinstance, studies show influenza vaccination rates amongchildren with asthma of 9% to 10% in health maintenanceorganizations7 and 25% in an allergy clinic.8 More recently,in practices with immunization registries, implementationof reminder and recall systems for high-riskchildren resulted in a 42% influenza vaccination rate.9

At its February 2002 meeting, the AdvisoryCommittee on Immunization Practices (ACIP) extendedits influenza vaccination recommendations toencourage annual influenza vaccination for all childrenaged 6 to 23 months when feasible. Given the novelty ofimmunizing children among whom influenza vaccinationwas being encouraged for the first time, a creativeapproach to intervention was necessary. Without specifyingwhich combination is most effective in any givensetting, the Task Force on Community PreventiveServices10 suggests a combination of patient-, providerandsystem-oriented strategies to improve vaccinationrates. Miller et al11 and Crabtree et al12 recommend tailoringinterventions to account for differences in medicaloffice culture, philosophy, and structure, as well aspatient sociodemographic characteristics and medicalacuity, to improve provision of preventive services ingeneral. In the present study, vaccination programs usingtailored interventions were implemented at 5 inner-cityhealth centers in Pittsburgh during the 2002-2003influenza season, the first season in which influenzavaccination was encouraged for 6-to 23-month-old children.Inner-city health centers were selected as interventionsites because many of the children treatedthere are economically disadvantaged, are more vulnerableto the effects of influenza disease, and historicallyhave low vaccination rates.13 The objectives of the presentstudy were to use qualitative and quantitative methodsto describe tailored interventions implemented byinner-city health centers and to report influenza vaccinationrates among children 6 to 23 months old.

METHODS

Health Centers

This project was approved by the institutional reviewboards of the University of Pittsburgh and Children'sHospital of Pittsburgh. Five health centers participatedin the intervention. Three of the inner-city centers wereresidency training sites and faculty practices (1 pediatricand 2 family medicine) that were affiliated withthe University of Pittsburgh School of Medicine. Theother 2 were faith-based family health centers locatedin disadvantaged neighborhoods. All sites served childrenfrom low-income families and agreed to engage inactivities to increase influenza vaccination rates amongthe target group. A sixth health center (a family practiceresidency training and faculty practice site) was notincluded in the interventions but served as a concurrentcontrol site for influenza vaccination data only.

Interventions

There were 2 types of interventions, those initiatedby the investigators, which were comparable for allhealth centers, and those designed and executed by thestaff of each health center, which differed across sites.Members of the intervention team visited each healthcenter to meet with staff (eg, medical directors, headnurses, and administrators) to present a menu of interventionoptions and to brainstorm about the best set ofinterventions to use for their particular site. All interventionsites were encouraged to establish standingorders to allow nursing staff to vaccinate eligible childrenwithout requiring a visit with or a written orderfrom a physician. Other optional strategies were suggested,such as the following: (1) patient-orientedstrategies, including mailings to all eligible patients,telephone calls (to remind, schedule, or inform parents),and posters, fliers, or brochures in waiting andexamination rooms; (2) provider-oriented strategies,including electronic or paper chart reminders, e-mailsto medical and nursing staffs, and competitions amongproviders; and (3) system-oriented strategies, includingwalk-in influenza vaccinations during regular officehours or during special sessions ("flu shot clinics").

The investigators donated posters, fliers, and buttonsto promote influenza vaccination; offered design expertise(with assistance from the University of PittsburghCenter for Minority Health) for low literacy mailingsand reminder postcards when requested; and providedsupport staff to send mailings and make telephone calls.A financial payment was provided to each site for assistancewith the research study.

Children receiving medical assistance obtain theirvaccinations through the Vaccines for ChildrenProgram; however, it did not pay for vaccinating healthychildren 6 to 23 months old against influenza in 2002-2003. Therefore, all intervention sites received a supplyof influenza vaccine proportionate to their respectivepediatric populations to allow them to provide the vaccinationfree of charge to 6-to 23-month-old childrenwho were without alternative payment means.

During August and September 2002, educational sessionswere held at all intervention sites to educatephysicians and nursing staff about influenza in general,about the safety and efficacy of the vaccine, about thenew ACIP recommendation to encourage vaccination ofhealthy 6-to 23-month-old children, and about theobjectives and specifics of the study. All nursing staffreceived a token thank you gift (an influenza vaccinationpromotion button and chocolates in a gift bag) todemonstrate the investigators' appreciation for theextra work involved in establishing this program.

Each site implemented interventions and, as theinfluenza vaccination season progressed, added strategiesas necessary. For instance, one site made telephonecalls to schedule vaccinations when response to itsmailing was low.

Patients and Vaccination Data

Eligible children were identified using electronic medicalrecords (at 3 sites) and billing records (at 2 sites). Forthe intervention group, all children in a practice whowere born between December 1, 2000, and March 31,2002, and had had an office visit between April 1, 2002and September 30, 2002 (to ensure that they were stillactive patients) were included. Children from the samepractices who were born between December 1, 1999, andMarch 31, 2001, and had been seen in the office betweenApril 1, 2001, and September 30, 2001, comprised thepreintervention group. Date of birth, and to the extentthat they were available, race, sex, and insurance statusdata were provided by each health center.

Vaccination data were collected for children whowere the appropriate age for the preintervention andintervention seasons. The influenza season was definedas October 1 through February 28. Although influenzavaccine is not frequently administered in February, theCenters for Disease Control and Prevention recommendsthat vaccine continue to be given to unvaccinatedindividuals as long as the influenza virus iscirculating. Two sites provided electronic data sets.Data from the other 3 sites were abstracted manuallyusing an electronic database (CASA software, Centersfor Disease Control and Prevention). All data were combinedin an electronic database, deidentified, cleaned,and prepared for analysis.

Focus Groups

Following the 2002-2003 influenza vaccination season,focus groups with representatives from the nursingstaff at each intervention site were conducted by atrained facilitator, assisted by a scribe who took notes.Neither the facilitator nor the scribe was aware of thevaccination rates achieved by the sites. The facilitatorenumerated the various interventions each site hadundertaken to increase influenza vaccinations.Discussions centered on the efficacy of each strategy,any suggestions for improvement, and the nursing staff'sfeelings about immunizing very young healthy childrenagainst influenza. Participants were also asked why theythought that parents accepted or refused influenzavaccination for their children. All focus groups wereaudiotaped, and tapes were subsequently transcribedand summarized. These data were primarily used toevaluate the various strategies and served as a basis fordeveloping the next year's intervention. They also providedan invaluable opportunity to assess nursing staffsupport of the program, problems they encountered,and parents' reaction to the new recommendation toencourage vaccination.

Statistical Analysis

P

&#967:2 Test and Fisher exact test were used to compareinfluenza vaccination rates between the preinterventionand intervention seasons and between the interventionand control sites. The focus group results were summarizedusing standard techniques for qualitative data.Logistic regression analysis was used to determine theeffects of the number of interventions used and of staffsupport on the first-dose vaccination rate. The regressionmodel was performed with and without the largestsite, which accounted for 74% of patients. Analyses wereperformed using SPSS version 11.0 for Windows (SPSSInc, Chicago, Ill) and SAS version 8.2 (SAS Institute,Cary, NC). Statistical significance was set at < .05 forall tests.

RESULTS

Patient Demographics

Table 1 gives the demographic characteristics of thechildren at the 5 intervention health centers and thecontrol site. Approximately one half of the childrenwere female. Among those whose race was known, theproportion of African American children was higherthan the 12% found among the general population ofAllegheny County, Pennsylvania.14 High rates of medicalassistance among this population are an indicator oftheir low socioeconomic status.

Influenza Vaccination Rates

Table 2 summarizes the influenza vaccinationrates (first and second doses) for the interventionand control sites during the preintervention (2001-2002) and intervention (2002-2003) influenza seasons.Children younger than 9 years old receivinginfluenza vaccination for the first time should receive2 doses. Therefore, most patients in the presentstudy would have been expected to be eligible for 2doses. For all intervention sites combined and ateach individual intervention site, first-dose influenzavaccination rates were significantly improved,compared with their historical data and those ofthe control site. For the second dose, vaccinationrates increased significantly for all interventionsites combined and at 3 of the 5 individual interventionsites.

Interventions

Table 3 lists each intervention site, gives all the variousinterventions used, indicates whether each siteused the intervention, and summarizes the evaluationof the intervention from the focus group participants.All intervention sites used at least 1 patient-orientedstrategy, 1 provider-oriented strategy, and 1 system-orientedstrategy. The number of interventions variedamong sites, ranging from 6 to 11.

An unexpected revelation of the focus group processwas that the nursing staff members at only 3 of theintervention sites were supportive and considered therecommendation to encourage vaccination as a routinepart of care. The staffs at the other 2 sites expressedreluctance to add another vaccination to the scheduleand indicated that they did not vaccinate without aphysician's visit and order (ie, they did not use standingorders).

The Figure shows the number and type of interventionsused by each site and the corresponding first-doseinfluenza vaccination rates. In general, the 3 sites thatused the largest numbers of intervention strategies andwhose staffs embraced the idea had higher vaccinationrates than the 2 sites that implemented fewer strategiesand whose staffs were not supportive of influenza vaccinationfor 6-to 23-month-old children. The logisticregression analyses showed that influenza vaccinationrates were significantly associated with the number ofinterventions used (odds ratio, 1.24; 95% confidenceinterval, 1.15-1.34) and with staff support of the effort(odds ratio, 1.91; 95% confidence interval, 1.40-2.60).However, when the largest site was removed from theregression models, these associations becameinsignificant.

Table 4 lists the reasons that nurses cited in thefocus groups for parental acceptance or refusal ofinfluenza vaccination for their children. Notably, neitherrace nor income was mentioned.

DISCUSSION

In the 2002-2003 influenza season, the ACIP for thefirst time encouraged influenza vaccination for healthychildren aged 6 to 23 months. This policy differed fromprevious ACIP childhood vaccination recommendationsin that (1) it lacked the effect of a full recommendation,(2) the timing for administration was basedon age and season of the year (as opposed to age alonefor other childhood vaccinations), and (3) it immediatelyfollowed 2 years of significant shortages or delaysin production of influenza vaccine. The present studywas undertaken to introduce influenza vaccinationamong healthy children aged 6 to 23 months who werepatients at inner-city health centers. These childrenwere chosen as the target population because socioeconomicfactors such as income, race, parental educationallevel, family size, and mother's age, as well asparental attitudes and levels of trust, are related to vaccinationstatus.15-18

Influenza vaccination rates have been successfullyincreased by 17% to 20% by using reminder and recallsystems among children with high-risk medical conditions.9,19 However, various other interventions specificallytargeting low-income children and those treated atinner-city health centers have had varyingdegrees of success (5%-20% increases)in increasing general vaccination rates.20-22 Crabtree et al12 suggest that tailoredinterventions that account for differencesamong practices, as opposed to a one-size-fits-all approach, may be more successfulin increasing provision ofpreventive services such as vaccinations.Among adults, tailored interventions thatwere designed in concert with participatinghealth centers to improve influenzavaccination rates in inner-city health centershave been effective in eliminatingracial disparities.23 Therefore, in thisstudy, we worked with administratorsand designated staff members at eachhealth center to develop interventionsthat they believed were feasible and mostlikely to succeed in increasing influenzavaccination rates among their pediatricpopulations.

Vaccination rates increased significantlyamong 6-to 23-month-old childrenat all intervention sites, demonstratingthe feasibility of introducing a new ageandseason-based vaccination schedulefor children of this age. The National ImmunizationSurvey reported a 2002-2003 influenza vaccination rateof 7.4% among children 6 to 23 months old.24 TheBehavioral Risk Factor Surveillance System reported a2004-2005 vaccination rate of 49.4% after the adoptionof influenza vaccination for children 6 to 23 months ofage into the recommended childhood schedule.25Zimmerman et al26 report that adding influenza vaccinationto the childhood schedule does not delay thetiming of receipt of other vaccines.

Although all intervention sites had some strategies incommon (eg, free vaccinations, posters in waiting andexamination rooms, and staff education), there wasconsiderable variability in the numbers and types ofstrategies used by each health center. Moreover, healthcenters with higher vaccination rates reported full staffsupport of the effort. The number of strategies used wasnot related to the size of the practice or staffing levels.For instance, the site with the fewest patients used 8strategies, while the site with the second largest patientpopulation used 6 strategies. At sites with staff support,influenza vaccination was perceived as and communicatedto the parents as a routine part of well-child care.The 2 centers with the lowest influenza vaccinationrates engaged in fewer vaccination promotion activitiesand reported in focus groups that there had been lessthan enthusiastic support of the programby the nursing staff. The awareness-to-adherencemodel of physician vaccinationbehavior (a 4-stage model developed toexplain physician use of vaccination guidelines)suggests that an awareness of vaccinationrecommendations by physiciansalone is insufficient to ensure high vaccinationrates.27 According to the model, physiciansmust intellectually agree withrecommendations, decide to adopt them inthe care they provide, and regularly adhereto them at appropriate times. This modelmay apply to other clinical staff as well.

The second-dose vaccination rates weremuch lower than the first-dose rates. Thismay be due to parents' lack of familiaritywith the need for a first-time vaccinee toreceive a second dose for maximum protectionagainst influenza infection, or to theinability of office staff to devote time to promotingreceipt of the second dose. It may also simply reflect alow priority given to taking a healthy young child to ahealth center during the winter months if other vaccinationsare not due.

Importance of Communication

This study demonstrated the value of appropriateand effective communication for modifying healthbehavior at several levels. First, the investigatorsworked closely with the health centers to help themdevelop intervention strategies that were likely to befeasible, effective, and sustainable. Centers were providedwith well-documented successful strategies and wereallowed to choose or creatively modify them, therebyintegrating administration and clinical staff in theprocess from the outset. Second, communicating theobjectives of the study and the scientific justification forthe ACIP recommendation to encourage vaccinationthrough provider education was effective. However, welearned from the focus groups that more time shouldhave been devoted to clarifying the details of vaccination(eg, dosing, needle size, and 1 vs 2 doses). We alsofound that simply presenting factual information wasinsufficient to convince staff in some health centers ofthe value of vaccinating 6-to 23-month-old children andthat we should have allowed time for addressing theirconcerns. Carpiano et al28 show that indicators of officeteamwork such as having shared visions and staff roleexpectations, as well as indicators of tenacity such ashaving a champion of preventive services and a staffopen to innovation, were significantly related to immunizationrates. At the 2 intervention sites with the lowestvaccination rates, the lack of staff support suggests afailure to share the administrative interest in vaccination,while the decision not to use standing orders mightindicate a lack of openness to innovation.

Other studies29-32 report that physicians' recommendationsfor vaccination, as well as patients' beliefs thattheir physician recommends vaccination, are associatedwith increased influenza and pneumococcal vaccinationrates. From the focus groups in our study, welearned that provider recommendation and parentaltrust in the provider, as well as a desire to preventchildhood illness, were reasons why parents acceptedthe new recommendation to encourage vaccination. Incontrast, belief that flu shots cause influenza and fearsof other adverse effects were reasons why parentsdeclined vaccination in our study. This knowledge canhelp providers communicate the benefits of andaddress fears related to influenza vaccination.

Finally, the relationships among the number ofinterventions used, staff support, and vaccination ratessuggest that the introduction of a new recommendationmay require a multidimensional approach to producethe best results. A concerted effort by an enthusiasticand supportive office staff to educate, inform, andremind parents communicates that influenza vaccinationis important and is worth the extra effort requiredto protect children.

Strengths and Limitations

This multisite intervention targeted a group of low-incomechildren attending different types of inner-cityhealth centers. Clinically meaningful and statisticallysignificant increases in influenza vaccination ratesoccurred, despite the fact that this interventionoccurred during the first year of the recommendation toencourage vaccination (ie, 2 years before a full recommendationwill have been in place). Compared withcontrol data, tailored interventions were successful inraising vaccination rates.

This before-after multi-intervention study is limitedby the fact that multiple interventions did not allow isolationof the most powerful among them. However, theliterature on methods to increase vaccination rates issufficiently mature that meta-analyses and systematicreviews have been published on this subject.10,33Therefore, the need is to determine whether thesealready proven interventions can be translated toinfluenza vaccination among young children, particularlythe economically disadvantaged, not whether theyare successful in idealized settings. Another possiblelimitation is that most of the children were from 2 racialgroups. Therefore, the response of other racial and ethnicgroups is unknown. Finally, some sites had not fullyrecorded the race and ethnicity of their patients, leavinggaps in the demographic data.

CONCLUSIONS

Our findings suggest that introducing influenza vaccinationamong 6-to 23-month-old children in inner-cityhealth centers is feasible by means of different strategies.Health centers that used more strategies vaccinatedmore children than those using fewer strategies. The useof patient-oriented, provider-oriented, and system-orientedstrategies, as well as enthusiastic staff support,communicates to parents that influenza vaccination isimportant and can result in higher vaccination rates.

Acknowledgments

We thank Charles Bemm, MD, Bruce Block, MD, David Greenberg, MD,William H. Markle, MD, J. Todd Wahrenberger, MD, and their staffs at theparticipating health centers; Stuart Weinberg, MD, for his database managementefforts; and Feng Shou Ko, MS, for his assistance with the statisticalanalyses.

From the Departments of Family Medicine and Clinical Epidemiology (MPN, CJL, RKZ,JAT, SMC), Pediatrics (AH, DHK), and Radiation Oncology (CJL), School of Medicine, andthe Departments of Behavioral and Community Health Sciences (CJL, RKZ) and HealthPolicy and Management (CJL), Graduate School of Public Health, University of Pittsburgh,Pittsburgh, Pa.

This study was supported by award TS-894 from the Centers for Disease Control andPrevention, Atlanta, Ga, and the Association of Teachers of Preventive Medicine, Washington,DC; and by grant P60 MD-000-207 to the EXPORT (Excellence in Partnerships forCommunity Outreach, Research on Disparities in Health, and Training) Health Project atthe Center for Minority Health, Graduate School of Public Health, University of Pittsburgh,from the National Center on Minority Health and Health Disparities, National Institutes ofHealth, Bethesda, Md. The contents of this article are the responsibility of the authors anddo not necessarily reflect the official views of the Centers for Disease Control andPrevention, the Association of Teachers of Preventive Medicine, or the National Institutesof Health.

Address correspondence to: Mary Patricia Nowalk, PhD, RD, Department of FamilyMedicine and Clinical Epidemiology, School of Medicine, University of Pittsburgh, 3518Fifth Avenue, Pittsburgh, PA 15261. E-mail: tnowalk@pitt.edu.

N Engl J Med.

1. Neuzil KM, Mellen BG, Wright PF, Mitchel EF Jr, Griffin MR. The effect ofinfluenza on hospitalization, outpatient visits, and courses of antibiotics in children.2000;342:225-231.

J Infect Dis.

2. Simonsen L, Fukuda K, Schonberger LB, Cox NJ. The impact of influenza epidemicson hospitalizations. 2000;181:831-837.

Pediatrics.

3. Quach C, Piche-Walker L, Platt R, Moore D. Risk factors associated with severeinfluenza infections in childhood: implication for vaccine strategy. 2003;112(pt 1):e197-e201.

Am J Public Health.

4. Sullivan KM, Monto AS, Longini IM Jr. Estimates of the US health impact ofinfluenza. 1993;83:1712-1716.

Rev Infect Dis.

5. Glezen WP. Considerations of the risk of influenza in children and indicationsfor prophylaxis. 1980;2:408-420.

Pediatr Infect Dis J.

6. Glezen WP, Taber LH, Frank AL, Gruber WC, Piedra PA. Influenza virus infectionsin infants. 1997;16:1065-1068.

Vaccine.

7. Kramarz P, DeStefano F, Gargiullo PM, et al; Vaccine Safety Datalink Team.Influenza vaccination in children with asthma in health maintenance organizations.2000;18:2288-2294.

Ann Allergy Asthma Immunol.

8. Chung EK, Casey R, Pinto-Martin JA, Pawlowski NA, Bell LM. Routine andinfluenza vaccination rates in children with asthma. 1998;80:318-322.

Pediatrics.

9. Daley MF, Barrow J, Pearson K, et al. Identification and recall of children withchronic medical conditions for influenza vaccination. 2004;113(pt1):e26-e33.

Am J Prev Med.

10. Task Force on Community Preventive Services. Recommendations regardinginterventions to improve vaccination coverage in children, adolescents, and adults.2000;18:92-96.

J Fam Pract.

11. Miller WL, Crabtree BF, McDaniel R, Stange KC. Understanding change inprimary care practice using complexity theory. 1998;46:369-376.

J Fam Pract.

12. Crabtree BF, Miller WL, Aita VA, Flocke SA, Stange KC. Primary care practiceorganization and preventive services delivery: a qualitative analysis. 1998;46:403-409.

JAMA.

13. Glezen WP, Greenberg SB, Atmar RL, Piedra PA, Couch RB. Impact of respiratoryvirus infections on persons with chronic underlying conditions. 2000;283:499-505.

Table DP-1: Profile of General Demographic

Characteristics: 2000: Geographic Area: Allegheny County, Pennsylvania.

14. US Bureau of the Census. Washington, DC: US Bureau of the Census; November 6, 2001.

Annu Rev Public Health.

15. Cutts FT, Orenstein WA, Bernier RH. Causes of low preschool immunizationcoverage in the United States. 1992;13:385-398.

Pediatrics.

16. Strobino D, Keane V, Holt E, Hughart N, Guyer B. Parental attitudes do notexplain underimmunization. 1996;98(pt 1):1076-1083.

17. Centers for Disease Control and Prevention. Estimated vaccination coveragewith individual vaccines and selected vaccination series among children 19-35months of age by state. US National Immunization Survey Q3/2002-Q2/2003.Available at: www/cdc/gpv/nip/coverage/nis/02-03/toc-0203.htm. Accessed March14, 2004.

Am J Public Health.

18. Prislin R, Dyer JA, Blakely CH, Johnson CD. Immunization status and sociodemographiccharacteristics: the mediating role of beliefs, attitudes, and perceivedcontrol. 1998;88:1821-1826.

Pediatrics.

19. Lieu TA, Capra AM, Makol J, Black SB, Shinefield HR. Effectiveness and cost-effectivenessof letters, automated telephone messages, or both for underimmunizedchildren in a health maintenance organization. 1998;101:e3-e10.

Pediatrics.

20. Rodewald LE, Szilagyi PG, Humiston SG, Barth R, Kraus R, Raubertas RF. Arandomized study of tracking with outreach and provider prompting to improveimmunization coverage and primary care. 1999;103:31-38.

Arch Pediatr

Adolesc Med.

21. LeBaron CW, Starnes DM, Rask KJ. The impact of reminder-recall interventionson low vaccination coverage in an inner-city population. 2004;158:255-261.

Arch Pediatr Adolesc Med.

22. Hambidge SJ, Davidson AJ, Phibbs SL, et al. Strategies to improve immunizationrates and well-child care in a disadvantaged population: a cluster randomizedcontrolled trial. 2004;158:162-169.

Am J Public Health.

23. Zimmerman RK, Nowalk MP, Raymund M, et al. Tailored interventions toincrease influenza vaccination in neighborhood health centers serving the disadvantaged.2003;93:1699-1705.

MMWR Morb Mortal Wkly

Rep.

24. Centers for Disease Control and Prevention. Childhood influenza-vaccinationcoverage: United States, 2002-03 influenza season. 2004;53:863-866.

MMWR Morb Mortal Wkly Rep.

25. Centers for Disease Control and Prevention. Estimated influenza vaccinationcoverage among adults and children: United States, September 1, 2004-January31, 2005. 2005;54:304-307.

Am J Prev Med.

26. Zimmerman RK, Hoberman A, Nowalk MP, et al. Feasibility of influenzaimmunization for inner-city children aged 6 to 23 months. 2004;27:397-403.

Med Care.

27. Pathman DE, Konrad TR, Freed GL, Freeman VA, Koch GG. The awarenessto-adherence model of the steps to clinical guideline compliance: the case of pediatricvaccine recommendations. 1996;34:873-889.

Prev Med.

28. Carpiano RM, Flocke SA, Frank SH, Stange KC. Tools, teamwork, and tenacity:an examination of family practice office system influences on preventive servicedelivery. 2003;36:131-140.

Am J Med.

29. Zimmerman RK, Santibanez TA, Janosky JE, et al. What affects influenza vaccinationrates among older patients? an analysis from inner-city, suburban, rural,and Veterans Affairs practices. 2003;114:31-38.

Am J Prev Med.

30. Zimmerman RK, Nowalk MP, Bardella IJ, et al. Physician and practice factorsrelated to influenza vaccination among the elderly. 2004;26:1-10.

Ann Fam Med.

31. Santibanez TA, Zimmerman RK, Nowalk MP, Jewell IK, Bardella IJ. Physicianattitudes and beliefs associated with patient pneumococcal polysaccharide vaccinationstatus. 2004;2:41-48.

Vaccine

32. Zimmerman RK, Santibanez TA, Fine MJ, et al. Barriers and facilitators ofpneumococcal vaccination among the elderly. . 2003;21:1510-1517.

Can J Public Health.

33. Gyorkos TW, Tannenbaum TN, Abrahamowicz M, et al. Evaluation of theeffectiveness of immunization delivery methods. 1994;85(suppl 1):S14-S30.