Impact of the Change to Inactivated Poliovirus Vaccine on the Immunization Status of Young Children in the United States: A Study From Pediatric Research in Office Settings and the National Medical Association
Objective. To determine whether the change from an all oral poliovirus vaccine (OPV) schedule to an inactivated poliovirus vaccine (IPV)-containing schedule has adversely affected the immunization status of young children in the United States.
Methods. Immunization data were abstracted from the medical records of children 8 to 35 months old seen consecutively for any reason in the offices of practicing pediatricians who are members of the Pediatric Research in Office Settings network of the American Academy of Pediatrics or the National Medical Association. Data on up to 120 eligible children were collected in each practice between March 1998 and January 2000. Patients were classified as fully immunized at 8 months old if they had received 3 diphtheria-tetanus-pertussis, 2Haemophilus influenzae type b, 2 hepatitis B, and 2 poliovirus vaccines. Study children who were ≥12 months of age at the time that data were collected were categorized as being fully immunized at 12 months if they had received the same vaccines before their first birthday. To assess the effect of type of poliovirus vaccines on these outcomes, study patients were classified as being in an IPV or OPV group based on the initial type of vaccine received. Logistic regression was used to calculate the odds ratios (ORs) and 95% confidence intervals (CIs) for IPV as a predictor of being fully immunized at 8 and 12 months of age, after adjusting for race/ethnicity of the patient, maternal education level, year of birth, and method of payment for vaccines. In addition, the effect of clustering of children within practices was accounted for by the use of generalized estimation equation techniques.
Results. Data were analyzed on 13 520 children from 177 practices in 42 states; 79.4% of patients were fully immunized at 8 months of age, and 88.7% of those eligible were fully immunized at 12 months of age. A total of 6910 patients (51.1%) were classified as OPV recipients, wheras 5282 (39.1%) received IPV. In addition, 1328 children (9.8%) were documented as having received poliovirus vaccine, but the particular type could not be determined. Compared with OPV recipients and after controlling for the confounding variables and the effect of clustering within practices, children in the IPV group were as likely as were OPV recipients to be fully immunized at 8 months of age (OR: 1.04; 95% CI: 0.88,1.23). At 12 months of age, the OR for IPV as a predictor of being fully immunized was 1.08 (95% CI: 0.90,1.30). When compared with OPV recipients, adjusted ORs for children in the undetermined poliovirus vaccine type group being fully immunized at 8 and 12 months of age were 0.84 (95% CI: 0.68,1.04) and 0.84 (95% CI: 0.67,1.07), respectively.
Conclusions. The results of this national study indicate that the implementation of an IPV-containing poliovirus vaccine schedule has not had an adverse effect on the immunization status of young children who were vaccinated in the offices of practicing pediatricians.
The eradication of poliomyelitis in the United States is one of the most gratifying public health achievements of the final half of the 20th century. The last reported indigenous cases of poliomyelitis caused by wild virus in the United States occurred in 19791; no indigenous cases have been reported in the Western Hemisphere since 1991.2 This remarkable decline in the incidence of poliomyelitis is directly attributable to the introduction of inactivated poliovirus vaccine (IPV) in 1955 and oral poliovirus vaccine (OPV) in 1963.3
Although OPV has been very effective in eliminating wild virus infection, ∼8 to 9 cases per year of vaccine-associated paralytic polio (VAPP) were reported annually between 1980 and 1994.4,,5 Thus, in 1997 the Advisory Committee on Immunization Practices and the American Academy of Pediatrics (AAP) recommended that poliovirus vaccine in the United States be delivered via a sequential IPV/OPV schedule, a change from the all-OPV schedule previously endorsed.3,,6 It was estimated that this change would decrease the number of cases of VAPP by 50% or more.3 When the sequential schedule was recommended, an additional change to an all-IPV schedule was contemplated for initiation once efforts to globally eradicate poliovirus progressed.
Although many experts endorsed this change in the polio immunization policy,7 the recommendation proved to be controversial. One major objection was that the additional injections required with the IPV/OPV sequential schedule might have a negative impact on compliance.8,,9 Of special concern were populations that have been difficult to vaccinate, such as poor and minority children.
To assess the impact of the change to the IPV/OPV sequential schedule, we conducted a nationwide study in practice settings to compare the immunization status of infants receiving poliovirus vaccine via an all-OPV schedule with that of infants begun on IPV. Although 2 published studies have so far found no evidence of a negative impact of an IPV/OPV sequential schedule, these studies were conducted regionally in specialized settings (a closed panel health maintenance organization [HMO] and public health clinics).10,,11 Because the majority of children in the United States receive their vaccines from office-based practitioners,12,,13 the inclusion of infants immunized in this setting is fundamental to assessing the impact of this major change in immunization policy. Before beginning the study, we postulated that the immunization status of children receiving IPV would be at least equivalent to that of children vaccinated with OPV. This was the main study hypothesis on which sample size calculations were based.
The study was conducted by the Pediatric Research in Office Settings (PROS) network of the AAP and the Pediatric Section of the National Medical Association (NMA). PROS is a national practice-based research network; 1465 practitioners in 491 practices from 48 states, Quebec and Puerto Rico are members. Membership in the NMA is primarily comprised of black physicians; NMA pediatricians have traditionally provided health care to a high proportion of underserved children. From both organizations, only practitioners providing primary care, including immunizations, to a defined population of patients were recruited for the study.
Participation in the project by practicing pediatricians was entirely voluntary. However, practitioners caring for a high proportion of minority children were specifically recruited for the study. In addition, an effort was made to include a sample of pediatricians who reflected current practice patterns with respect to poliovirus vaccines. To achieve this, in the fall of 1997 PROS practitioners and NMA pediatricians were surveyed regarding their current practices regarding use of IPV and/or OPV in their offices. A total of 1330 surveys were completed. Among respondents, 32% reported using an all-OPV schedule in late 1997, 5% followed an all-IPV schedule, 57% the IPV/OPV sequential schedule, and 6% had other poliovirus vaccine policies. These data were used during recruitment for the study in an attempt to assemble a group of participating practitioners with a similar distribution of IPV and OPV usage.
Ultimately, 177 practitioners from 42 states contributed immunization data for the study. Of these, 161 (92.5%) had previously completed the survey on poliovirus vaccine usage; 26% responded that they were using an all-OPV schedule, 2% an all-IPV schedule, 63% the IPV/OPV sequential schedule, and 9% used another schedule, such as offering both IPV and OPV and letting parents chose the form of vaccine for their children. Only 1 practitioner per primary care office collected immunization data.
The main outcome measure of the study was immunization data collected on children 8 to 35 months old seen consecutively, for any reason, by 1 of the participating practitioners. At the time of an office visit, the project was explained to parents of eligible children. If a parent consented to participation, he/she completed a parental study form that included information on the study child's date of birth, race/ethnicity of the child, maternal education level, and method of payment for immunizations for the patient. Study forms were available in English and Spanish. Only 1 child from a family was enrolled in the study; if >1 age-eligible patient was being seen simultaneously for an office visit, the youngest child was selected for participation.
Each participating practitioner was asked to enroll up to 120 children in the study. During the data collection period in each office, a log was kept of all visits by age-eligible children to the practitioner. There were 14 522 office visits by age-eligible children; 168 of these patients were subsequently determined to be ineligible for the study (eg, younger sibling had been enrolled, parents did not speak English or Spanish). A total of 306 parents declined participation in the project, and parents of 339 potentially eligible children were inadvertently not approached for enrollment. Thus, 13 709 children (95.5% of those eligible) were enrolled in the study.
After the office visit, the medical record of each study child was reviewed; the practice Vaccine Administration Record (VAR), VARs from other practices, and newborn nursery records (for hepatitis B vaccine) were photocopied. This information was sent to the research staff and an immunization history for each child constructed.
Before beginning the study, we conducted an assessment on the completeness of data in the VAR. The office medical records of 8 to 10 children, 8 to 35 months of age, seen consecutively in 31 PROS/NMA practices were reviewed. Data on 283 children were analyzed. For 237 patients (83.7%), all available immunization information was recorded in the office VAR. For the remaining 46 children, 70% of all supplemental immunization data were recorded in either VARs supplied by other providers or the newborn nursery record. The immunization rate of these 283 patients at 8 months old was estimated at 79.9% based on all data in the office medical record versus 77.0% when only information from the office VAR, other VARs, and the nursery record were included.
The primary outcome of the study was immunization status of infants at 8 months of age. Children were considered to be fully immunized if they had received 3 diphtheria-tetanus-pertussis/diphtheria-tetanus-acellular pertussis/diphtheria-tetanus, 2 IPV/OPV, 2 Haemophilus influenzae type b, and 2 hepatitis B vaccines before the time they were 8 months of age. Secondary outcomes included immunization status at 12 months old. Patients were considered fully immunized at 12 months of age if they had received 3 diphtheria-tetanus-pertussis/diphtheria-tetanus-acellular pertussis/diphtheria-tetanus, 2 IPV/OPV, 2 H influenzae type b, and 2 hepatitis B vaccines before their first birthday. This outcome was only determined on patients who were at least 12 months of age at the time that immunization data from their medical records were abstracted. In addition to these outcomes, the timeliness of receipt of the initial 2 poliovirus immunizations was assessed. Study children were defined as receiving their first poliovirus vaccine late if they were >90 days of age when receiving their first poliovirus immunization, and very late if they were >120 days old. Similarly, patients were considered late or very late for their second poliovirus immunization if the interval between the first and second vaccine was >90 or >120 days, respectively. Children who never received a first and/or second vaccine were classified as very late.
Immunization outcomes were determined only on study children for whom a valid date of birth was recorded on the parental study form or office log. In addition, outcomes were not calculated on enrolled patients if no VAR or other immunization information was sent. However, if the medical record of an enrolled child was reviewed but no vaccine information was found, the patient was considered unimmunized and included in the analyses. Among the 13 709 enrolled children, study forms had discrepant dates of birth for 70 patients, 55 had missing parental study forms, and 64 had mismatched parental study forms and VARs. Thus, the final analysis was based on data on 13 520 children (98.6% of those enrolled).
Study children were categorized as receiving poliovirus immunization via an IPV-containing schedule or an all-OPV schedule based on the initial dose of vaccine administered. Thus, a patient was classified as being in the IPV group if the first dose of poliovirus vaccine received was IPV, and in the OPV group if the first dose received was OPV. Classification to the IPV or OPV group was only made if a child's immunization data specifically indicated which form of the vaccine was administered or if manufacturer or lot number information was recorded that clearly revealed the type of vaccine used. In situations in which it was evident that a study child received a poliovirus immunization but the particular type was not specified, a categorization to either the IPV or OPV group was not made. Finally, it is possible that some parents' decision not to have poliovirus vaccine administered to their child was attributable to the type of immunization offered by their child's pediatrician. Study patients who did not receive any poliovirus vaccines were classified as being in the IPV or OPV group based on the type of vaccine received by the next youngest child in that pediatric practice on whom immunization data were collected. There was no record of any poliovirus immunization for 288 study children (2.1%). Based on this classification schema, 96 of these patients were placed in the IPV group and 192 in the OPV group.
To assess the effect of receiving IPV on the immunization status of children, logistic regression was performed with each of the outcome measures as the dependent variable, and poliovirus vaccine group (IPV, OPV, or undetermined type) as the main independent variable of interest. Potentially confounding variables such as the child's race/ethnicity (white, black, Hispanic, or other), year of birth, maternal education level (high school graduate or less, some college, or college graduate or more), and method of payment for immunizations (out-of-pocket, free to parents, partially paid by parents) were controlled for in the analyses. To assess the impact of the change to IPV among minority populations, logistic analyses including race–IPV interaction terms were also performed. Finally, to assess the impact of the change to IPV in families with the lowest education level, we repeated the logistic regression analysis including only the 1491 study children whose mothers did not graduate from high school.
The 13 520 children whose immunization data were analyzed for the study were from 177 different practices. To account for the possibility that children within a particular practice were more similar than were patients from different practices or clusters, generalized estimating equation (GEE) techniques were used in the logistic regression analyses. GEE is a form of regression analysis that adjusts for the effects of clustered sampling.14
Preliminary univariate analyses of categorical data were conducted using χ2 tests. For all comparisons, 2-sidedP values of <.05 or odds ratios (ORs) with a 95% confidence interval (CI) that did not include 1.0 were considered statistically significant.
Before collecting immunization data, it was not known what proportion of infants would receive IPV. Sample size calculations were based on a conservative assumption that 20% of study children would be placed in the IPV group. (Any greater use of IPV by the study population than 20% would increase our power to detect differences in immunization status related to the type of poliovirus vaccine received.) Sample size calculations were further complicated by the possibility that children within a particular practice (or cluster) would be more similar than would children from different practices. Because of this potential correlation among patients in the same cluster, the effective sample size was somewhat reduced. The effective sample size was calculated as: number of children/(1 + [cluster size − 1] × intraclass correlation coefficient [ICC]), where cluster size was estimated as 100. In a previous study on the immunization status of children followed by PROS practitioners,15 the ICC of immunization status by practice was 0.11. Based on this value for ICC and an assumption that 80% of those in the OPV group would be fully immunized at 8 months of age, a sample size of 15 000 would yield a >80% chance of detecting an OR of 0.65 or less for IPV as a predictor of immunization status (ie, that IPV children were significantly less likely to be fully immunized than those in the OPV group), corresponding to a difference of 8% between the IPV and OPV groups (at a 1-sided α level of 0.05).
The study was approved by the institutional review board of the AAP. Immunization data were collected from March 1998 through January 2000.
Characteristics of the 13 520 children on whom immunization data were analyzed are shown in Table 1. Although the majority of recruited patients were white, 32.2% of study children were from minority populations. Mothers of study patients were relatively well educated; 67.4% received some education beyond high school. Conversely, only 1491 (11.1%) had less than a 12th-grade education. The vast majority of parents indicated that their child received immunizations for little or no out-of-pocket expenditure; only 5.9% of parents responded that they paid for their child's vaccines totally out-of-pocket. Overall, 79.4% of the 13 520 children included in the study were fully immunized at 8 months of age, and 88.7% at 12 months of age (immunization rate at 12 months of age based on the 10 396 children who were at least 12 months of age at the time of the survey).
Also presented in Table 1 are the immunization rates at 8 months of age of children with different characteristics and P values assessing the differences in rates between different groups in each category. Among children of different race/ethnicity, the differences in immunization rates were statistically significant. This was attributable to the lower vaccination rate among black children; compared with all other ethnic groups black patients were significantly less likely to be fully immunized (68.3% vs 81.4%; P< .00 001). Mothers with higher levels of education were more likely to have fully immunized infants than were those with less education. Method of payment for vaccines was statistically associated with immunization rates; however, the magnitude of the effect was only moderate. Finally, during the period that study children received their first vaccines, there was a significant secular trend in immunization rates; however, the overall change was modest, ranging from 75.7% for infants born in 1995 to 80.2% for those born in 1997.
Classification of study children into the IPV or OPV group is also shown in Table 1. As can be seen, 39.1% of study children were placed in the IPV group and 51.1% in the OPV group. These classifications include 288 patients who had no record of ever receiving a poliovirus vaccine; their categorization into the IPV or OPV group was based on the type of immunization received by the next youngest child in the same practice. In addition, 1328 infants (9.8%) received some form of poliovirus vaccine, but the exact formulation could not be determined from the information available. As expected, utilization of IPV vaccine increased during the time that study children were born. Among children who received either IPV or OPV, for those born in 1995 and 1996 only 11.1% received an initial dose of IPV, whereas 64.2% of those born in 1997–1999 received this form of the vaccine.
The percentage of children who were fully immunized at 8 months of age is shown by poliovirus vaccine type in Table 1. In this univariate analysis, children receiving IPV were more likely to be fully immunized than were those in the OPV and undetermined poliovirus vaccine groups. In comparing vaccination status among patients known to receive an initial dose of IPV (n = 5186) with a similar group of OPV recipients (n = 6718), the OR for IPV as a predictor of being fully immunized at 8 months of age was 1.33 (95% CI: 1.21,1.47).
Because of the potential for significant confounding, logistic GEE regression was performed. In these analyses, immunization statuses at 8 and 12 months of age (fully immunized or underimmunized) were the dependent variables, with type of poliovirus vaccine (IPV, OPV, undetermined) the main independent variable. Also included in the models were maternal education level, method of payment for immunization, year of birth, and race/ethnicity. The results are presented in Table 2. As can be seen in the table, compared with those receiving OPV, there were no indications that children who received IPV were less likely to be fully immunized. At 8 months of age, patients in the IPV group were as likely as those in the OPV group to be fully immunized (OR: 1.04; 95% CI: 0.88,1.23); immunization status at 12 months of age was also similar among children who received IPV and OPV. These analyses were repeated after excluding the 288 children who had no record of ever receiving a polio immunization; the ORs for receipt of IPV as a predictor of being fully immunized at 8 and 12 months of age were 1.03 and 1.07, respectively (95% CI at 8 months: 0.86,1.23; and at 12 months: 0.87,1.32).
Overall, 91.4% of children in the IPV group and 90.1% of those in the OPV group received their initial dose of poliovirus vaccine before the age of 90 days; 94.9% and 93.3%, respectively, received their first dose before they were 120 days old. The interval between the first and second doses of vaccine was <90 days in 87.6% of IPV and 88.7% of OPV recipients, and <120 days in 94.2% and 92.1%, respectively. After adjusting for confounding variables, there were no differences in the timeliness of receipt of vaccines among OPV and IPV recipients (Table 3).
To evaluate whether the use of IPV had a different effect among minority populations than among white children, race–poliovirus vaccine type interaction terms were included in logistic regression models. For black children, the impact of the use of IPV was not different from white patients in terms of immunization status at 8 months of age (P = .93) or timeliness of the initial dose of poliovirus vaccine (P = .64). Analyses of all minority children versus white children had similar results, with no difference in either of these outcomes (P = .94 andP = .67, respectively). The results were also similar when data from the 1491 children whose mothers had less than a high school education were analyzed separately; the adjusted OR for IPV use as a predictor for being fully immunized at 8 months of age was 1.05 (95% CI: 0.75,1.47).
A separate analysis was performed including only the study children born 1997- 1999, when the use of the IPV/OPV sequential schedule was widely implemented; 58.9% of the enrolled patients were born during this period. In the analysis, maternal education level, race/ethnicity, and payment type were also included; GEE was used to control for the effect of within practice correlation. The adjusted OR for IPV as a predictor of immunization status at 8 months of age among children born in 1997–1999 was 1.21 (95% CI: 1.02,1.43).
When the IPV-containing poliovirus vaccine schedule was recommended, it was reasonable for health care providers and policy makers to question whether such a change was prudent. It was feared that the additional injections required to adequately vaccinate children might result in decreased and delayed immunization against poliovirus specifically, and, perhaps, lower immunization rates in general. Concerns about increasing number of injections and changes in immunization practices might be heightened in parents of underserved children in whom health beliefs regarding vaccines are less positive than among less disadvantaged families.16 However, despite multiple analyses specifically designed to elucidate any negative effects, the results of our study strongly indicate that the use of IPV has had no significant impact on the immunization status of young children in the United States. Patients who received an initial dose of IPV were at least as likely to be fully immunized at 8 and 12 months of age as those immunized with OPV; similarly, the timeliness of vaccination against polio was equivalent in IPV and OPV recipients. Furthermore, the results suggest that this lack of negative effect on immunization status was the same among black children, other minority populations, and white children.
The results of our study are similar to findings of investigators evaluating the impact of the use of IPV in different populations. The Centers for Disease Control and Prevention evaluated immunization status of >17 000 children who were enrolled in closed panel HMOs in California and Washington State.10 At 12 months of age, those who received IPV were as likely to be fully immunized as those vaccinated with OPV (adjusted OR: 1.1; 95% CI: 1.0,1.3). Similarly, Kolasa et al11 conducted a cohort study in urban public health clinics comparing the immunization status of children before and after the change in poliovirus vaccine. They reported comparable compliance with IPV and OPV for the first and second doses of vaccine. Thus, IPV does not seem to have an adverse effect on immunization status in infants vaccinated in public health clinics, HMOs, and in the offices of a national sample of primary care pediatricians.
The ease and rapidity of the transition to an IPV-containing schedule in the United States seems surprising. However, previous research indicates that parents are accepting of additional injections, especially if the risk of an adverse effect is reduced. Woodin et al17 reported that physicians were more concerned about multiple vaccine injections at 1 visit than parents of young children. Among parents surveyed, 58% indicated that they would rather have their child receive 4 injections at 1 visit than fewer injections and more visits. In another survey of parental choice, Thoms et al18 found that that 61.3% of parents would prefer IPV to OPV, even with the additional injections required. The major reason for this preference was the risk of VAPP associated with OPV. These investigators also reported that 75% of parents would consult their child's physician before making a choice between IPV and OPV. Because 75% of pediatricians in the United States are currently recommending IPV,19 if parents follow the advice of their child's health care provider, a high proportion of infants will receive IPV.
Although we believe that the validity of the findings in this study in assessing the impact of the change to IPV are strengthened by analyzing data obtained on a sample of children followed by practicing pediatricians from throughout the United States, there are inherent limitations in this methodology. Assessment of a child's immunization status was based solely on information abstracted from the medical record of 1 group of providers from a practice-based research network. These providers were not selected randomly, which could limit the generalizability of our findings. However, in a survey of a random sample of AAP pediatricians conducted during the time that data were collected for this project, 70% of respondents indicated that they recommended an IPV/OPV sequential schedule, 5% an all-IPV schedule, 16% an all-OPV schedule, and 9% had no policy.19 These results are consistent with the pattern of poliovirus vaccine use reported by participants in our study.
An additional limitation in our study is that the immunization data in the records reviewed were not always complete and, to facilitate the rapid collection of the vaccine information, only certain portions of the medical records of study children (containing ∼90% of all immunization data) were reviewed. Fortunately, there was no systematic bias in the degree of incompleteness of immunization information between IPV and OPV recipients.
A final problem in analyzing data abstracted from practice medical records was the difficulty in determining the precise type of poliovirus vaccine administered; for 9.8% of the sample it was impossible. By univariate analysis, children in the undetermined poliovirus vaccine group were less likely to be fully immunized at 8 months of age than either IPV or OPV recipients. This may indicate that many of these patients received vaccines at sites other than at the practice in which the immunization data were collected making the information in the practice medical record less complete than for children in the IPV or OPV groups. Alternatively, the lack of detail in the immunization record may reflect less attention given to the overall vaccination process for these children leading to more underimmunization.
Because the majority of study children who received OPV were born before 1997, whereas IPV recipients were predominately born in 1997 and 1998, significant secular trends in immunization rates would tend to bias the analyses in favor of IPV. There was a small but statistically significant change in the percentage of children who were fully immunized at 8 months of age between 1995 and 1998. We attempted to account for this by controlling for year of birth in the logistic models. Furthermore, analysis of data on patients born in 1997–1999, when there was widespread use of both types of poliovirus vaccine, indicated that children who received IPV were more likely to be fully immunized than were those who received OPV (OR: 1.21; 95% CI: 1.02,1.43). Thus, taken together, none of these limitations seems to undermine the major findings of the study.
Vaccination against poliovirus using an all-OPV schedule was remarkably successful for over 30 years. VAPP, although devastating, is an extraordinarily rare side effect of OPV. However, our results, combined with those of other investigators,10,,11 indicate that use of an IPV-containing schedule was implemented without adverse impact on the immunization status of infants in the United States, and has apparently been accompanied by a decrease in the incidence of VAPP.20 Our findings further suggest that implementation of the recently recommended all-IPV poliovirus vaccine schedule should be accomplished without delay, thereby eliminating the risk of VAPP entirely.21
This study was funded by the Centers for Disease Control and Prevention, National Immunization Program, and the Health Resources and Services Administration, Maternal and Child Health Bureau.
The following pediatric practices or individual practitioners completed this study: PROS (listed by AAP Chapter): Alabama, Drs Heilpern and Reynolds, PC (Birmingham), Stephen Brandt, MD (Ozark); Alaska, Anchorage Pediatric Group (Anchorage); Arizona, Orange Grove Pediatrics (Tucson), Cigna HealthCare (Tucson); California-1, Castro Valley Pediatrics (Castro Valley), Drs Anita Tolentino-Macaraeg and Arminda Tolentino (Hollister), Shasta Community Health Center (Redding), Cantor, Giedt, and Kamachi (Salinas); California-4, Edinger Medical Group, Inc (Fountain Valley); Colorado, Cherry Creek Pediatrics (Denver), Rocky Mountain Youth (Denver), Julie Brockway, MD (Ft Collins), Gino Figlio, MD (Lamar); Connecticut, Jeff Cersonsky, MD (Southbury), Carol Rizzolo, RPA-C (Southington); Delaware, A.I. Dupont Institute Children's (Middletown); Florida, Rose Joseph, MD (Fort Lauderdale); Georgia, Victor Lui, MD (Chamblee), The Pediatric Center (Stone Mountain); Hawaii, Jeffrey Lim, MD (Honolulu); Illinois, Southwest Pediatrics (Palos Park); Indiana, Northpoint Pediatrics (Indianapolis), Georgetown Medical Care (Indianapolis), Jeffersonville Pediatrics (Jeffersonville); Iowa, Integra Health Pediatrics (Cedar Rapids), Asha Madia, MD (Des Moines); Kansas, Ashley Clinic (Chanute); Kentucky, Michael Simon, MD, and Jennifer Crane MD (Lexington); Maryland, O'Donovan and Ahluwalia, MD, PA (Baltimore), Children's Medical Group (Cumberland), Shore Pediatrics (Denton), Shore Pediatrics (Easton), Clinical Associates PA (Towson); Massachusetts, Garden City Pediatrics (Beverly), Pediatric Associates of Norwood (Norwood), Worcester Pediatric Associates (Worcester); Michigan, Susan Hendrickson, MD (Bay City), Children's Hospital of Michigan (Detroit), MiddleBelt Pediatrics HCA (Farmington Hills), Downriver Pediatric Assoc, PC (Lincoln Park), Pediatric and Family Care (Rochester Hills), Mackinac Straits Primary Care (St Ignace), Orchard Pediatrics (West Bloomfield); Minnesota, Brainerd Medical Center (Brainerd), South Lake Clinic (Minnetonka); Missouri, Edith Colbert, MD (Chesterfield), Children's Mercy Hospital (Kansas City), Northland Pediatric Associates (North Kansas City); Montana, Stevensville Pediatrics (Stevensville); New Hampshire, Exeter Pediatric Associates (Exeter), Pediatric and Adolescent Medicine (Kingston), Laconia Clinic (Laconia); New Jersey, Kids Care Pediatrics (Egg Harbor Township), Delaware Valley Pediatric Associates, PA (Lawrenceville); New Mexico, Presbyterian Family Healthcare-Rio Bravo (Albuquerque); New York-1, Pediatric Associates (Camillus), Panorama Pediatric Group (Rochester), Genesee Health Service (Rochester), Elmwood Pediatric Group (Rochester), Edward Lewis, MD (Rochester), Brighton Hill Pediatrics (Syracuse); New York-2, Andrea J. Leeds, MD, PC (Bellmore), Sonia Vinas, MD (Brooklyn), Huntington Medical Group, PC (Huntington Station); North Carolina, Eastover Pediatrics (Charlotte), Hendersonville Pediatrics (Fletcher); North Dakota, Altru Clinic (Grand Forks), Medical Arts Clinic (Minot); Ohio, Bryan Medical Group (Bryan), Raj Gupta, MD (Dayton), North Central Ohio Family Care (Galion), Drs Harris and Rhodes (Lancaster), Children's Hospital Physicians Assoc (Twinsburg), Comprehensive Pediatrics (Westlake), Wooster Clinic (Wooster), St Elizabeth Health Center (Youngstown); Oklahoma, Shawnee Medical Center Clinic (Shawnee), Pediatric and Adolescent Care, LLP (Tulsa); Oregon, NBMC (Coos Bay); Pennsylvania, Ches Penn Health Services (Chester), Erdenheim Pediatrics, PC (Flourtown), Praful Bhatt, MD (Lock Haven), Plum Pediatrics (Pittsburgh), Pennridge Pediatric Associates (Sellersville), Han Pediatrics, Crozer-Chester Medical Center (Upland), Reading Pediatrics, Inc (Wyomissing); Rhode Island, Marvin Wasser, MD (Cranston); South Carolina, Cheraw Pediatrics (Cheraw), Palmetto Pediatrics (Columbia), Grand Strand Pediatrics (Surfside Beach); Texas, White Rock Pediatrics, PA (Dallas), Sarah Helfand, MD (Dallas), Winnsboro Pediatrics (Winnsboro); Utah, Gordon Glade, MD (American Fork), John Weipert, MD (American Fork), Alpine Pediatrics (American Fork), University of Utah Health Sciences Center (Salt Lake City), Mountain View Pediatrics (Sandy); Vermont, Judy Orton, MD (Bennington), CHP Brattleboro Pediatrics (Brattleboro), University Pediatrics (Burlington), Rebecca Collman, MD (Colchester), Mousetrap Pediatrics (Milton), Practitioners of Ped Medicine (South Burlington), CHP Timber Lane Pediatrics (South Burlington), Drs Joseph Hagan, Jr George Brown, and Jill Rinehart (South Burlington), St Johnsbury Pediatrics (St Johnsbury), University Pediatrics (Williston); Virginia, Drs Casey, Goldman, Lischwe, Garrett, and Kim (Arlington), Fishing Bay Family Practice (Deltaville), Pediatric Assoc of Richmond, Inc (Richmond), South Richmond Health Center (Richmond); Washington, Jemima Tso, MD (Chehalis), Columbia Basin Health Association (Othello), Redmond Pediatrics (Redmond), Rockwood Clinic (Spokane), Yakima Neighborhood Health Services, Inc (Yakima); West Virginia, Grant Memorial Pediatrics (Petersburg); and Wisconsin, Beloit Clinic SC (Beloit), Gundersen Clinic (La Crosse), Gundersen Clinic-Whitehall (Whitehall).
NMA (listed by state): California, Mayo De Lilly, MD (Los Angeles); Delaware, A.I. Dupont Pediatric Childrens Hospital (Wilmington); Florida, Manatee County Rural Health Services (Bradenton), Arlene Haywood, MD (Plantation); Georgia, Beverly Walker, MD (Albany), Marsha Glover, MD (Bainbridge), Evelyn Johnson, MD (Brunswick); Lousiana, Charmine Venters, MD (Baton Rouge); Maryland, Diana Abney, MD (Waldorf), Johns Hopkins Community Physicians (West Friendship); Missouri, William Pankey, MD (Kansas City), Jackie Wayne-Tenney, MD (Lees Summit), Homer E. Nash, Jr, MD (St Louis); and New York, Winston Price, MD (Brooklyn).
- Received October 18, 2000.
- Accepted January 17, 2001.
Reprint requests to (J.A.T.) 146 N Canal St, Suite 300, Seattle, WA 98103. E-mail:
- IPV =
- inactivated poliovirus vaccine •
- OPV =
- oral poliovirus vaccine •
- VAPP =
- vaccine-associated paralytic polio •
- AAP =
- American Academy of Pediatrics •
- HMO =
- health maintenance organization •
- PROS =
- Pediatric Research in Office Settings •
- NMA =
- National Medical Association •
- VAR =
- Vaccine Administration Record •
- GEE =
- generalized estimating equation •
- OR =
- odds ratio •
- CI =
- confidence interval •
- ICC =
- intraclass correlation coefficient
- Strebel PM,
- Sutter RW,
- Cochi SL,
- et al.
- Centers for Disease Control and Prevention
- American Academy of Pediatrics, Committee on Infectious Diseases
- Katz SL
- Judelsohn R
- Hinman AR
- Rodewald L,
- Peak R,
- Ezzati-Rice T,
- Zell E,
- Thompson K
- ↵Diggle PJ, Kiang KY, Zeger SL. Analysis of Longitudinal Data. Oxford, England: Clarendon Press; 1994:146–168
- Taylor JA,
- Darden PM,
- Slora E,
- Hasemeier CM,
- Asmussen L,
- Wasserman R
- Keane V,
- Stanton B,
- Horton, et al
- ↵Darden PM, Taylor JA, Brooks DA, Baker AE, O'Connor KG, Bocian AB. Polio immunization practices of pediatricians. Paper presented at Pediatric Academic Societies' 1999 Annual Meeting; May 4, 1999; San Francisco, CA
- Centers for Disease Control and Prevention
- American Academy of Pediatrics, Committee on Infectious Diseases
- Copyright © 2001 American Academy of Pediatrics