Objective. To evaluate the impact of the 2002–2003 recommendation to “encourage when feasible” the influenza vaccine for healthy children 6 to 23 months of age.
Methods. A cross-sectional study of children who were 6 to 59 months of age and presented to a large, pediatric resident's continuity clinic or the affiliated acute care clinic in the summers of 2002 and 2003 was performed. The influenza vaccination status of children in the winter before enrollment and factors that influenced this status were determined by parental questionnaire.
Results. Of 245 and 329 children in the 2002 and 2003 study populations, influenza vaccinations increased from 7% to 18%, respectively. For healthy children 6 to 23 months of age, influenza vaccinations increased from 1% in 2002 to 17% in 2003 and accounted for most of the increase seen in the study population. Multivariate analysis revealed that the strongest predictor of pediatric influenza vaccination was parental recall of a physician recommendation (odds ratio: 39.3; 95% confidence interval: 17.3-89.4 in 2003). However, 65% of parents of high-risk children did not recall a physician recommendation despite the fact that the influenza vaccine was recommended for high-risk patients during both study years.
Conclusion. Concomitant with the 2002-2003 influenza vaccine recommendation for healthy children 6 to 23 months of age, influenza vaccinations for these children increased significantly. Because most children for whom the influenza vaccine was recommended were not vaccinated, physicians have the opportunity to increase the pediatric influenza vaccination rate by improving their recommendation strategies.
Influenza virus infects 30% to 40% of children each winter1–5 and causes significant morbidity such as hospitalizations, outpatient visits, and antibiotic courses.6–8 Children are also a major reservoir for influenza infection and frequently spread influenza disease to other family members and community contacts.9–11 Hence, vaccinating children can reduce the rate and severity of influenza infections for children and their families.12,13
For decades, the Advisory Committee on Immunization Practices (ACIP) and the American Academy of Pediatrics have recommended the influenza vaccine for children ≥6 months of age with high-risk conditions and their families. These high-risk conditions include 1) asthma and other chronic pulmonary diseases, 2) hemodynamically significant cardiac diseases, 3) immunosuppressive disorders or therapy, 4) human immunodeficiency virus infection, 5) sickle cell anemia and other hemoglobinopathies, 6) diseases that require long-term aspirin therapy, 7) chronic renal dysfunction, and 8) chronic metabolic diseases.13,14 Despite these long-standing recommendations, few high-risk children (10%–30%) receive the annual influenza vaccine.15–18
Influenza-attributable hospitalization rates of children <2 years of age are comparable to that of adults ≥65 years of age for whom the annual influenza vaccine is recommended.7 Hence, the ACIP modified the 2002–2003 recommendations by “encouraging when feasible” the influenza vaccine for healthy children 6 to 23 months of age.14 We evaluated the impact of this new recommendation on the influenza vaccination status of children and identified factors that influenced their vaccination status.
All children who were younger than 5 years and received primary care in a large, university-affiliated pediatric resident's continuity clinic and whose parent spoke English or Spanish (2003 only) were eligible. A trained interviewer approached parents who presented to the continuity clinic or the affiliated acute care clinic on weekdays from August 16 to August 29, 2002, and from May 21 to July 31, 2003. The first summer was before the new influenza vaccine recommendation for healthy, young children; the second summer was after the new recommendation. The Vanderbilt University Institutional Review Board reviewed and approved this study.
The pediatric continuity and acute care clinics provided care to an underserved, urban population. In 2002 and 2003, the continuity and acute care clinics had 7407 and 8551 visits for children <60 months of age, respectively. The age groups in the clinics were 16% 0 to 5 months, 39% 6 to 23 months, and 45% 24 to 59 months. The racial demographics were ∼45% black, 30% white, 20% Hispanic, and 5% other. Most children (∼80%) resided in Davidson County, the metropolitan area of Nashville, TN, and had Medicaid insurance. The Vaccines for Children Program provided free vaccines for >95% of these children. During the study period, strategies to identify and vaccinate children did not change.
After obtaining consent, a trained interviewer administered the questionnaire in English or Spanish (2003 only). The questionnaire, which had previously been designed and pretested,15 was modified and pilot-tested for the outpatient setting. It elicited the child's demographic information, medical history, influenza vaccination status, family influenza vaccination status, and parental reasons for the child's influenza vaccination status.
A child was considered vaccinated when a parent reported that his or her child received at least 1 dose of influenza vaccine after September 1 of the previous year. To confirm this report, all parents of vaccinated children were asked when the influenza vaccine was administered. For all children, the vaccination status was verified in the Tennessee State Immunization registry that documents all routine and some influenza vaccines received in the health department or Vanderbilt Pediatric Continuity Clinic but not any vaccines from most other clinics in the region. When the registry but not the parent indicated that a child was vaccinated, the child was considered vaccinated and the parental reasons for that vaccination status were excluded. A family member was considered vaccinated when a member of the household excluding the enrolled child was reported to have received the influenza vaccine in the previous winter.
A search of Epic (Madison, WI), a computerized outpatient clinic database used at Vanderbilt University, obtained the child's resident physician and the number of continuity clinic and acute care visits for each patient from October 1 to January 31 for each study year, the months when influenza vaccine is usually given. Personal identifiers were removed, and all data were coded so that the researchers were blinded to the identity of the patients, residents, and attending physicians.
Demographic information and medical history of the child were obtained by parental report. Race was characterized as white, black, Hispanic, or other. When a parent identified the child as biracial, the child was classified in the nonwhite group. The 6- to 23-month and 24- to 59-month age groups were determined by the age of the child on December 31 before enrollment. We identified high-risk children who were specifically recommended to receive the influenza vaccine in the 2000 Red Book13 by the following questions. Parents were asked whether their child had a history of asthma, cancer, cystic fibrosis, kidney disease, diabetes mellitus, heart disease, sickle cell anemia, immunodeficiency, or other chronic, high-risk conditions. We asked about premature birth and length of oxygen therapy, daily medications, and hospitalizations to detect any additional high-risk conditions.15
To identify parental factors that influenced the influenza vaccination status of children, all parents were asked whether a physician recommended the vaccine for their child. Using similar methods as previously described, parents of unvaccinated children were asked, “There are often many reasons for not receiving a flu shot. Which of the following reasons influenced your decision for your child last winter?” Potential reasons included lack of knowledge, concern about side effects, availability of vaccine, lack of a physician recommendation, inconvenience, and other. Parents of vaccinated children were asked, “Why did your child receive the flu shot last winter?” Potential reasons included protection of a family member, good experience with the vaccine, previous influenza infection, information from the media or a friend, physician recommendation, and other. All comments including “other” reasons were documented, and parents could cite multiple reasons for a child's vaccination status.15
All data were double-entered into Access 4.0 (Redmond, WA) and compared to verify the accuracy of data entry. The primary outcome—influenza vaccination status in the winter before enrollment—and all variables, except for race, were coded as dichotomous variables. Demographic information and the proportion of vaccinated children by high-risk condition, age group, having a vaccinated family member, having a continuity clinic visit between October and January, and physician characteristics were evaluated using 2-sided, χ2 or Fisher exact tests. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated using univariate and multivariate logistic regression to determine which factors identified a priori predicted a child's influenza vaccination status. All statistical analyses were performed using Stata 6.0 (College Station, TX).
Of 302 and 358 parents approached in 2002 and 2003, 281 (93%) and 340 (95%) children were enrolled, respectively. Of enrollees, 245 (87%) and 329 (97%) children were 6 to 59 months of age on December 31 before enrollment and thus comprised the study population (Table 1). In both years, the gender and age groups were comparable. The racial distribution changed significantly (P < .001), likely reflecting the enrollment of Spanish-speaking families in 2003. Sixteen percent of the population had high-risk conditions: asthma (12%), congenital heart disease (1%), cancer (1%), kidney disease (1%), immunodeficiency (1%), and other high-risk conditions (<1%) such as diabetes insipidus and bronchopulmonary disease.
Influenza vaccination rates increased for all racial groups over the 2 study years. In 2002, only 16 (7%) children were vaccinated as compared with 58 (18%) in 2003 (P < .001; Table 2). Of all children by age and high-risk groups, only healthy children 6 to 23 months of age had a significant increase in influenza vaccinations from 2002 (1%) to 2003 (17%; P < .001). Although more high-risk children received the influenza vaccine in 2003, the increase was significant only for children with asthma (13%–37%; P = .03).
Recalling a physician recommendation (OR: 39.3; 95% CI: 17.3–89.4 in 2003) was the strongest predictor of a child's influenza vaccination status by univariate and multivariate analyses (Table 3). Parental recall of a physician recommendation remained at 35% for high-risk children, whereas it increased from 7% in 2002 to 18% in 2003 (P < .001) for healthy children. For high-risk children, 50% in 2002 and 83% in 2003 were vaccinated when a parent recalled a physician recommendation; the respective proportions were 50% and 63% for healthy children. In contrast, when a parent did not recall a physician recommendation, few healthy or high-risk children (≤5%) were vaccinated. From 2002 to 2003, parental recall of a physician recommendation increased significantly for children 6 to 23 months of age (9%–22%; P = .004) but not for children 24 to 59 months of age (14%–19%; P = .2). Another significant predictor of the influenza vaccine in 2003 was a family member's being vaccinated (OR: 2.4; 95% CI: 1.05–5.7). The proportion of healthy children with a vaccinated family member remained ∼22%, whereas the proportion for high-risk children increased from 30% in 2002 to 63% in 2003 (P = .001). It is interesting that having a continuity clinic visit between October and January (when the influenza vaccine was administered) did not predict the influenza vaccination status of young children in either study year.
Because physician recommendation significantly influenced a child's influenza vaccination status, we evaluated the practice patterns of physicians by the proportion of their patients in our study population who were vaccinated. In 2003, 5 (9%) of 57 pediatric residents had, on average, 50% (range: 40%–75%) of their patients vaccinated (high vaccinators); the other 52 residents (low vaccinators) had, on average, 13% (range: 0%–30%) of their patients vaccinated (P < .001). The 5 high vaccinators accounted for 20 (34%) of the 58 vaccinated patients, had different attending physicians in continuity clinic, and represented all years of pediatric residency training. More parents recalled a physician recommendation from high vaccinators (21 of 40 [53%]) than from low vaccinators (46 of 296 [16%]; P < .001). More patients of high vaccinators (29 of 40 [73%]) had a continuity clinic visit from October through January than patients of low vaccinators (142 of 289 [49%]; P = .006). Of children with a continuity clinic visit at this time, more parents recalled a physician recommendation from high vaccinators (20 of 29 [69%]) than they did from low vaccinators (27 of 142 [19%]; P < .001). In contrast, few children without a continuity clinic visit at this time had parents who recalled a physician recommendation (1 of 11 [9%] for high vaccinators and 19 of 147 [13%] for low vaccinators; P = .7) or were vaccinated (2 of 11 [17%] and 15 of 147 [10%], respectively; P = .4). Between October 2002 and January 2003, 161 unvaccinated children made 363 continuity or acute care visits.
To address the concern that patients of high vaccinators systematically differed from those of low vaccinators, we compared patient characteristics. The children in each group were similar in regard to gender, age groups, and proportion with high-risk conditions. Approximately one third of parents in each group reported concern about the side effects of the influenza vaccine. The only identified difference was the racial composition of the clinic populations: 13% white, 35% black, 40% Hispanic, and 13% other for high vaccinators and 29% white, 43% black, 24% Hispanic, and 4% other for low vaccinators (P = .009).
To understand further parental reasons for the vaccination status of children, we asked parents which factors influenced their decision. Parents of unvaccinated children most commonly reported lack of knowledge about the influenza vaccine (50%) and lack of physician recommendation (54%), which together accounted for 62% of the cited reasons. Concern about vaccine side effects (31%) influenced some parents, but lack of time and low vaccine availability (7%) were infrequently cited. For vaccinated children, physician recommendation (>80%) was the most commonly cited reason for receiving the vaccine. Other reasons cited in 2003 but not 2002 were past experience with the influenza vaccine (45%), information from the media or a friend (29%), protection of a family member (21%), and previous experience with an influenza infection (21%). Among 141 children with a vaccinated family member, 65 (46%) reported that the family's experience with the influenza vaccine influenced their decision to vaccinate their child.
This study is the first, to our knowledge, to suggest that the 2002–2003 recommendation to “encourage when feasible” the influenza vaccine for healthy children 6 to 23 months of age14 affected behavior. Of all children 6 to 59 months of age, the proportion of vaccinated children increased significantly from 7% in 2002 to 18% in 2003 (P < .001) largely because vaccination of healthy children 6 to 23 months increased from 1% to 17% (P < .001). Hence, the new ACIP recommendation seemed to influence influenza vaccinations for the target population, healthy children 6 to 23 months of age, in its first year of implementation.
Similar to previous studies,15–18 we found that most high-risk children did not receive the annual influenza vaccine, despite the recommendation. The proportion of vaccinated high-risk children increased from 18% in 2002 to 31% in 2003 (P = .2), but the small sample size limited our power to detect a difference. Despite this limitation, the proportion of vaccinated children with asthma increased significantly from 13% to 37% (P = .03). These data suggest that the vaccination status of high-risk children can increase, but system improvements are needed to vaccinate all high-risk children.
This study has several potential limitations. Because the number of eligible children in continuity and acute care clinics at any given time varies, not all parents of eligible children were approached. To minimize selection bias, we approached the eligible child who most recently checked into the clinic and enrolled >93% of these children. Influenza vaccination status was determined by parental report, so misclassification bias might have occurred. However, misclassification seems less likely because self-reported influenza vaccination status has been validated in adults19 and the pediatric influenza vaccination rate of 2001–2002 was compatible with published rates.15–18 That all parents reported that the influenza vaccine was given between October and January and that 11 of 12 children with influenza vaccine documented in the Tennessee State Immunization Registry correctly identified their status were reassuring. Physician recommendation about the influenza vaccine was determined by parental report and not by chart review or physician interview. Although parents probably did not recall all physician recommendations, this method did reflect the message that parents received.
Another limitation is the generalizability of this study. Physicians in an academic setting, such as ours, may have more exposure to teachings that emphasize the importance of the new influenza vaccine recommendation than physicians in other clinical settings. Nonetheless, the finding that rates can be altered significantly is important as are the results detailing the reasons for the increase and for the acceptance or nonacceptance of vaccination.
Consistent with our previous study and adult studies, physician recommendation strongly predicted the influenza vaccination status of a child.15,20–24 However, most parents (93% in 2002 and 82% in 2003) did not recall a physician recommendation. Approximately two thirds of the reasons cited for not vaccinating children can be addressed by improving parental education about the influenza vaccine, explicitly recommending it, and reducing missed opportunities. In 2003, high vaccinators were more likely to see their patients in continuity clinic during the months when influenza vaccine was administered and to have parents recall a recommendation for the influenza vaccine than low vaccinators. Furthermore, few children without a continuity clinic visit between October and January received the influenza vaccine in 2002–2003 even when they had an acute care visit at this time. Reducing missed opportunities would significantly improve the pediatric influenza vaccination rate.
Because many children do not have clinic visits when the vaccine is available, physicians need to discuss the influenza vaccine throughout the year and develop systems (eg, tracking systems, standing order forms, influenza vaccination times) to implement effectively this recommendation.25 Public health initiatives could complement such an effort. Sixteen (5%) parents of children enrolled in 2003 heard about the pediatric influenza vaccine from the media or a friend. Expanding the public health initiatives for children may also improve the pediatric vaccination rate because such initiatives paralleled a 34% increase in the influenza vaccination rate for the elderly.26 Encouraging the influenza vaccine for both children and their families may reap additional benefits. Of children with a vaccinated family member, 46% of parents reported that a good experience with the influenza vaccine strongly influenced their decision to vaccinate.
These results need to be confirmed in other clinical settings and over subsequent influenza seasons. Although the temporal increase in vaccinating healthy children 6 to 23 months of age with the new influenza vaccine recommendation is reassuring, most children with and without high-risk conditions do not receive the influenza vaccine. To improve the health of children, we need to identify and eliminate system barriers that impede the implementation of the influenza vaccine recommendations.27
With the implementation of the new ACIP recommendation, more healthy children 6 to 23 months of age received the influenza vaccine. Parental recall of a physician recommendation was the strongest predictor of a child's influenza vaccination status. Hence, clinicians could improve the pediatric influenza vaccination rate by explicitly recommending the influenza vaccine, reducing missed opportunities, and evaluating the systems that they use to identify and vaccinate such children. Improving the pediatric influenza vaccination rate for young and high-risk children has the potential to significantly reduce pediatric morbidity from influenza virus.
Dr Poehling is supported by the Robert Wood Johnson Foundation's Generalist Physician Faculty Scholar program.
This study was made possible by the generous cooperation and assistance of many people: Robin McClendon, Gretchen Edwards, and Betty Goodman for data acquisition; Arnette Edwards and Gretchen Edwards for data entry; and the entire clinic staff under the leadership of Gregory Plemmons, MD, and Rebecca Swan, MD. We appreciate the contributions of Kathryn Edwards, MD, and Veronica Gunn, MD, MPH, who critically reviewed this manuscript. We also thank all of the parents of young children who so willingly shared their perspectives.
- Glezen WP. Emerging infections: pandemic influenza. Epidemiol Rev.1996;18 :64– 76
- Wright PF, Ross K, Thompson J, Karzon DT. Influenza A infections in young children: primary natural infection and protective efficacy of live vaccine-induced or naturally acquired immunity. N Engl J Med.1980;296 :829– 834
- ↵Fox JP, Hall CE, Cooney MK, Foy HM. Influenzavirus infections in Seattle families, 1975–1979. Study design, methods and the occurrence of infections by time and age. Am J Epidemiol.1982;116 :212– 227
- ↵Longini IM, Koopman JS, Monto AS, Fox JP. Estimating household and community transmission parameters for influenza. Am J Epidemiol.1982;115 :736– 751
- Monto AS, Koopman JS, Longini IM. Tecumseh study of illness. XIII. Influenza infection and disease, 1976–1981. Am J Epidemiol.1985;121 :811– 822
- ↵Fox JP, Cooney MK, Hall CE, Foy HM. Influenzavirus infections in Seattle families, 1975–1979. Patterns of infection in invaded households and relation of age and prior antibodies to occurrence of infection and related illness. Am J Epidemiol.1982;116 :228– 242
- ↵American Academy of Pediatrics. Influenza. In: Pickering LK, ed. 2000 Red Book: Report of the Committee on Infectious Diseases. 25th ed. Elk Grove Village, IL: American Academy of Pediatrics; 2000:351–359
- ↵Poehling KA, Speroff T, Dittus RS, Griffin MR, Hickson GB, Edwards KM. Predictors of influenza virus vaccination status in hospitalized children. Pediatrics.2001;108 (6). Available at: www.pediatrics.org/cgi/content/full/108/6/e99
- Szilagyi PG, Holl JL, Rodewald LE, et al. Evaluation of New York State's Child Health Plus: children who have asthma. Pediatrics.2000;105 :719– 727
- Copyright © 2004 by the American Academy of Pediatrics