Background. The Vaccines for Children (VFC) program is designed to reduce the cost of vaccines for vulnerable children, including Medicaid-eligible children, American Indian/Alaska Native children, uninsured children, and underinsured children whose health insurance does not cover the cost of vaccinations. A desired consequence of the program is to promote comprehensive continuous medical care within a medical home for these children.
Objectives. To explore how having a medical home is associated with vaccination coverage among children eligible for the program.
Participants. A total of 24514 children 19 to 35 months of age sampled by the National Immunization Survey.
Design. VFC eligibility was evaluated for 24514 children 19 to 35 months of age who were sampled by the National Immunization Survey. Children were considered to have a medical home if they had a doctor, nurse, or physician's assistant who provided them with ongoing routine care, including well-child care, preventive care, and sick care, according to their parents. Sampled children were determined to be 4:3:1:3:3 up-to-date (UTD) if their vaccination providers reported administering ≥4 doses of diphtheria-tetanus toxoids-acellular pertussis vaccine, ≥3 doses of polio vaccine, ≥1 dose of measles-mumps-rubella vaccine, ≥3 doses of Haemophilus influenzae type b vaccine, and ≥3 doses of hepatitis B vaccine.
Results. Nationally, 44.9% of all children were VFC eligible and 93.0% of the VFC-eligible children received all vaccine doses at a provider enrolled in the VFC program. Compared with children who were not VFC eligible, VFC-eligible children were less likely to be UTD (70.8% vs 77.7%) and less likely to have a medical home (82.1% vs 95.0%). However, among VFC-eligible children, children who had a medical home were significantly more likely to be UTD, compared with children who did not have a medical home (72.3% vs 63.5%). Also, among VFC-eligible children who had a medical home, children who used their medical home consistently to receive all of their vaccination doses were significantly more likely to be UTD, compared with children who did not receive all of their doses from their medical home (75.3% vs 65.7%). Finally, the 4:3:1:3:3 vaccination coverage rate among VFC-eligible children who received all of their vaccination doses from their medical home was not significantly different from that among non-VFC-eligible children, after controlling for significant differences in sociodemographic factors between these groups (adjusted difference: 2.8%; 95% confidence interval: −0.1% to 5.7%).
Conclusions. Although the vaccination coverage rate among VFC-eligible children who had a medical home and received all vaccine doses from their medical home was essentially equivalent to that of non-VFC-eligible children, substantial percentages of VFC-eligible children either did not have a medical home or did not use their medical home to receive all of their recommended vaccinations. The vaccination coverage rate among these children was significantly lower. This suggests that there may be opportunities to increase vaccination coverage by removing barriers that prevent the adoption and consistent use of a medical home among these children.
In 1993, the Childhood Immunization Initiative was proposed to address significant gaps in vaccination coverage among young children in the United States.1 One goal of the Initiative was to ensure that, by 2000, ≥90% of all 2-year-old children received the recommended series of vaccinations. Among the strategies for achieving this goal was reduction of the costs of vaccines for parents. In October 1994, the Vaccines for Children (VFC) program was established to achieve this goal by giving publicly purchased vaccines at no cost to enrolled vaccination providers, for use for children who were eligible for the program.2
Before the establishment of the VFC program, private vaccination providers were likely to refer children whose parents could not afford vaccinations to public health department clinics for vaccinations.3–7 More recently, vaccine delivery in the United States has shifted largely from the public sector to the private sector,8–10 with an emphasis on vaccination in the context of primary care and the medical home.11 In 1992, and again in 2002, the American Academy of Pediatrics published policy statements11,12 asserting that
the medical care of infants, children, and adolescents ideally should be accessible, continuous, comprehensive, and family centered, coordinated, compassionate, and culturally effective. It should be delivered or directed by well-trained physicians who provide primary care and help to manage and facilitate essentially all aspects of pediatric care. The physician should be known to the child and family and should be able to develop a partnership of mutual responsibility and trust with them.
These characteristics define the “medical home.”
The objectives of this study were to evaluate the extent to which VFC-eligible children receive immunizations from VFC-enrolled providers, to determine how likely VFC-eligible children are to have a medical home, and to evaluate whether VFC-eligible children who have a medical home are better vaccinated than those without a medical home. In addition, among VFC-eligible children with a medical home, we evaluated whether children who receive all of their vaccine doses at their medical home are better vaccinated than those who receive some of their doses at their medical home and others at a different provider.
In our analyses, we used data collected between the third quarter of 2001 and the fourth quarter of 2002 by the National Immunization Survey (NIS). The NIS is a survey of US children 19 to 35 months of age conducted by the Centers for Disease Control and Prevention, for the purposes of monitoring vaccination coverage rates in the 50 states, the District of Columbia, and 27 other large metropolitan areas.
The NIS includes 2 phases of data collection. In the first phase, households with ≥1 age-eligible child are identified with list-assisted random-digit dialing (RDD). Cellular telephone numbers are excluded from the sampling frame. When a household with an age-eligible child is identified, the RDD interview is conducted and collects demographic information about each age-eligible child in the household, demographic information about the age-eligible child's mother, and sociodemographic information about the household. The income/poverty ratio for the household of each sampled child is determined with reported information on annual household income, the number of persons living in the household, and the number of children ≤18 years of age living in the household. The response rate of the NIS RDD interview is measured with the Council of American Survey Research Organizations (CASRO)13 response rate. The CASRO response rate is the product of the resolution rate, the screening completion rate, and the interview completion rate of households with age-eligible children. The resolution rate is the percentage of the total telephone numbers called that are classifiable as being nonworking, nonresidential, or residential. The screening completion rate is the percentage of known households that are screened successfully for the presence of age-eligible children. The interview completion rate is the percentage of households with ≥1 age-eligible child that complete the NIS RDD interview.
At the end of the NIS RDD interview, consent is sought to contact the age-eligible children's medical care providers. If consent is given, then the second data collection phase of the NIS is conducted. In the second phase, all of the vaccination providers named by the NIS RDD respondent are contacted by mail, to obtain the household's age-eligible children's provider-reported vaccination histories. In the mail survey, all providers are asked to indicate whether they are a “VFC provider,” ie, whether they are registered to receive and administer vaccines purchased through the VFC program. Data from the provider-reported vaccination histories are used to determine the number of visits paid to providers for vaccinations and to evaluate sampled children's vaccination status.
Because all of the children in our study were born in 1998 or thereafter, we evaluated the vaccination status of children according to the 1998 recommended vaccination schedule,14 which recommended the diphtheria-tetanus toxoids-acellular pertussis vaccine, the polio vaccine, the measles-mumps-rubella vaccine, the Haemophilus influenzae type b vaccine, and the hepatitis B vaccine. Sampled children were determined to be 4:3:1:3:3 up-to-date (UTD) if their vaccination providers reported administering ≥4 doses of diphtheria-tetanus toxoids-acellular pertussis vaccine, ≥3 doses of polio vaccine, ≥1 dose of measles-mumps-rubella vaccine, ≥3 doses of H influenzae type b vaccine, and ≥3 doses of hepatitis B vaccine.
Between the third quarter of 2001 and the fourth quarter of 2002, 74.5% of the sampled children whose parents completed the NIS RDD interview and gave consent for contacting vaccination providers were assigned randomly to receive the NIS Health Insurance Module (HIM). Information collected by the NIS HIM included whether sampled children had a medical home, whether they were covered by health insurance at the time that the NIS HIM was administered, and whether their health insurance paid for vaccinations. Sampled children were determined to have a medical home if the NIS respondent said that the child had a doctor, nurse, or physician's assistant who provided their child with ongoing routine care, including well-child care, preventive care, and sick care. A sampled child was determined to be uninsured if the NIS respondent reported that the child was not covered by private health insurance, Medicaid, the State Children's Health Insurance Program (SCHIP), military health care, the Title 5 program, the Indian Health Service, or any other health insurance. Sampled children were determined to be underinsured if the child was covered by any of these insurance types but the insurance did not cover all of the cost of the vaccines. Children were determined to be VFC eligible if they were (1) enrolled in their state's Medicaid program, their state's Medicaid expansion SCHIP program, or the Medicaid expansion component of their state's combination SCHIP program, (2) uninsured, (3) American Indian or Alaska Native, or (4) underinsured and determined to have been vaccinated at a facility that was reported by a vaccination provider to be a Federally Qualified Health Center (FQHC). Non-VFC-eligible children were children who were not VFC eligible. When a sampled child was reported to be covered by SCHIP, the income/poverty ratio of the household was used to verify eligibility in a SCHIP program that was part of the child's state Medicaid program.
Each state determines its own supply policy that specifies which vaccines are purchased with VFC funding and which vaccine providers receive the publicly purchased vaccines. In some states, the policy has been to distribute VFC vaccines to all health care providers. VFC-eligible children who lived in these states might have had greater access to VFC vaccines and thus might have had higher vaccination coverage rates. In our study, states were classified as having (1) a “universal policy” if all recommended vaccines were distributed to all health care providers to serve all children (Alaska, Connecticut, Idaho, Maine, Massachusetts, Nevada, New Hampshire, New Mexico, North Carolina, North Dakota, Rhode Island, South Dakota, Vermont, Washington, and Wyoming), (2) a “VFC and underinsured policy” if all publicly purchased vaccines were distributed to all health care providers to serve only VFC-eligible and underinsured children (Arizona, Florida, Georgia, Hawaii, Illinois, Maryland, Michigan, Minnesota, Montana, New York, South Carolina, and Texas), and (3) a “VFC-only policy” if all publicly purchased vaccines were distributed to private health care providers for only VFC-eligible children (Alabama, Arkansas, California, Colorado, Delaware, District of Columbia, Indiana, Iowa, Kansas, Kentucky, Louisiana, Mississippi, Missouri, Nebraska, New Jersey, Ohio, Oklahoma, Oregon, Pennsylvania, Tennessee, Utah, Virginia, West Virginia, and Wisconsin).
To evaluate the statistical significance of the estimated difference in 4:3:1:3:3 coverage rates between VFC-eligible children who had a medical home and VFC-eligible children who did not have a medical home, we first evaluated the 95% confidence interval (CI) for the estimated difference. If the 95% CI did not include 0, then we declared the difference to be statistically significant. As a second step in evaluating the estimated difference in coveragerates between these 2 groups, the distributions of selected sociodemographic factors were compared for the 2 groups. Factors whose distributions were significantly different between the 2 groups were noted. These factors were included as main effect predictors in a multivariate regression analysis of UTD status with respect to whether a child had a medical home, with data for VFC-eligible children only. The coefficient associated with having a medical home from the multivariate regression analysis summarized the estimated difference in vaccination coverage between the 2 groups, after accounting for the significant differences in the distributions of sociodemographic factors between the groups. We referred to this coefficient as the “adjusted difference” and declared this difference to be statistically significant when the 95% CI of the adjusted difference did not contain 0. A similar 2-step approach was used in evaluating whether estimated vaccination coverage rates differed significantly from other comparisons of 2 groups that are discussed in our report. All analyses were conducted with the statistical software package SUDAAN,15 a statistical software package that enables the sampling weights and complex survey design of the NIS to be taken into account. Zell et al16 and Smith et al17–19 provided overviews of the sampling design and methods used by the NIS, and Smith et al19,20 provided additional details about the purpose and design of the NIS HIM.
Sampling and Response Rates
Between the third quarter of 2001 and the fourth quarter of 2002, the CASRO response rate for the RDD portion of the NIS interview was 75.2% and 48529 NIS RDD interviews were completed. Among these, 31250 children were randomized to receive the NIS HIM and 30741 (98.4%) completed the module. Among children for whom NIS HIM interviews were completed, provider-reported vaccination histories were obtained for 24596 children (80.0%) in the second data collection phase of the NIS.
Nationally, 44.9 ± 1.1% of all children were VFC eligible (Table 1). VFC-eligible children were predominantly Medicaid-eligible children (81.3 ± 1.5%). VFC-eligible children had a significantly lower 4:3:1:3:3 vaccination coverage rate than did non-VFC-eligible children (70.8 ± 1.6% vs 77.7 ± 1.2%; difference: −6.9%; 95% CI: −9.0% to −4.9%). Figure 1 shows that the estimated state-level VFC eligibility rates varied from 25.8% (Minnesota) to 67.8% (Alaska).
Among all VFC-eligible children nationally, an estimated 93.0 ± 1.0% received all vaccine doses at a provider enrolled in the VFC program. The estimated rates ranged from 79.7% (Vermont) to 98.9% (New Mexico) (Fig 2). Table 2 lists the distribution of VFC-eligible children who did not receive all vaccine doses from VFC-enrolled providers according to selected child, maternal, and household characteristics. Table 2 shows that the VFC-eligible children who were significantly less likely to receive all vaccines from a VFC-enrolled provider were more likely to be non-Hispanic white than non-Hispanic black, more likely to have a mother who had more than a high school education, and more likely to live in a household with an annual income above the poverty level vs below the poverty level or in a household that had only 1 child vs ≥4 children.
Among VFC-eligible children, the estimated 4:3:1:3:3 vaccination coverage rate for children who had received all vaccine doses from VFC-enrolled providers was 71.3 ± 1.8% and did not differ significantly from the estimated rates for those who had received some doses (76.3 ± 10.2%; difference: −5.0%; 95% CI: −15.4% to 5.3%) or no doses (70.8 ± 10.0%; difference: 0.5%; 95% CI: −10.1% to 10.2%) from VFC-enrolled providers.
The Medical Home
Figure 3 shows that the estimated percentage of VFC-eligible children who had a medical home ranged from 53.9% (Nevada) to 96.7% (Vermont). Nationally, VFC-eligible children were significantly less likely than non-VFC-eligible children to have a medical home (82.1 ± 1.5% vs 95.0 ± 0.6%; difference: −12.8%; 95% CI: −14.5% to −11.1%). However, VFC-eligible children who had a medical home had significantly more visits to a vaccination provider for vaccines than did VFC-eligible children who did not have a medical home (6.5 ± 0.04 visits vs 5.8 ± 0.14 visits; difference: 0.7 visits; 95% CI: 0.5 to 0.8 visits). Also, VFC-eligible children who had a medical home had a significantly higher vaccination coverage rate than did VFC-eligible children who did not have a medical home (72.3 ± 1.7% vs 63.5 ± 4.7%; difference: 8.8%; 95% CI: 3.7% to 13.8%). Table 2 lists the distributions of VFC-eligible children who did not have a medical home according to selected characteristics. Among VFC-eligible children, children with no medical home were significantly more likely to be Hispanic or non-Hispanic black than non-Hispanic white, to receive all of their vaccinations at all private clinics instead of all public clinics, to be foreign-born, or to have a mother who had no more than a high school education, spoke Spanish, or was <30 years of age. After controlling for variability in the estimated 4:3:1:3:3 vaccination coverage rate attributable to these differences, the vaccination coverage rate for VFC-eligible children who had a medical home remained significantly higher than that for VFC-eligible children who did not have a medical home (adjusted difference: 10.3%; 95% CI: 5.3% to 15.3%).
Figure 4 shows that the estimated percentage of VFC-eligible children who had a medical home and received all of their vaccine doses from their medical home ranged from 31.7% (Arkansas) to 91.5% (Rhode Island). Among children with a medical home, VFC-eligible children were significantly less likely than non-VFC-eligible children to receive all of their vaccinations from their medical home (68.1 ± 1.8% vs 74.1 ± 1.3%; difference: −6.0%; 95% CI: −8.2% to −3.8%). However, among VFC-eligible children, children who received all of their vaccine doses at their medical home were significantly more likely to be 4:3:1:3:3 UTD than those who did not receive all of their vaccine doses at their medical home (75.3 ± 1.9% vs 65.7 ± 3.4%; difference: 9.6%; 95% CI: 5.7% to 13.5%). Table 2 lists the distribution of VFC-eligible children who had a medical home but did not receive all of their vaccine doses from their medical home according to selected child, maternal, and household characteristics. Among VFC-eligible children with a medical home, children who did not receive all of their vaccine doses at their medical home were more likely to be non-Hispanic white than Hispanic or non-Hispanic black; to have a mother who was widowed, divorced, or separated rather than married, spoke English, and was 20 to 29 years of age; lived in a non-metropolitan statistical area (MSA) rather than a non-central city MSA; received all of their vaccine doses from private providers; or lived in a state with a VFC and underinsured or a VFC-only supply policy. After controlling for variability in the 4:3:1:3:3 vaccination coverage rate attributable to these differences, the vaccination coverage rate for VFC-eligible children who received all of their vaccine doses from their medical home remained significantly higher than that for VFC-eligible children who did not receive all of their vaccine doses from their medical home (adjusted difference: 13.0%; 95% CI: 8.9% to 17.1%).
Among children who had a medical home and did not receive all vaccinations at their medical home, “convenience” was the reason given most frequently by parents for not seeking to have all vaccinations administered at the medical home (Table 3). Among children who had a medical home and did not receive all vaccinations at their medical home, “the high cost of vaccinations” was the second most frequently cited reason given by parents for not seeking to have all vaccinations administered at the medical home.
Although the 4:3:1:3:3 vaccination coverage rate for VFC-eligible children who received all of their vaccine doses from their medical home was nearly that of non-VFC-eligible children, the estimated 4:3:1:3:3 coverage rate for VFC-eligible children who received all of their vaccine doses from their medical home was lower (75.3 ± 1.9% vs 77.7 ± 1.2%; difference: −2.4%; 95% CI: −4.6% to −0.2%). However, these 2 groups differed significantly with respect to the distributions of the race/ethnicity of the child, maternal educational attainment, maternal preferred language, maternal age, household annual income, number of children in the household, MSA, and facility type. After controlling for variability in the estimated 4:3:1:3:3 vaccination coverage rate associated with these differences, the estimated coverage rate for VFC-eligible children who received all of their vaccine doses from their medical home was not significantly different from that for non-VFC-eligible children (adjusted difference: 2.8%; 95% CI: −0.1% to 5.7%).
A previous study demonstrated the high rate of penetration of the VFC program,21 and a number of studies have explored the use and benefits of medical homes among young children. Kempe et al22 and Szilagyi et al23 found that introduction of a state-funded insurance plan was associated with an increase in the use of primary care medical homes for preventive services, including immunizations. Santoli et al24 found that, of all children referred to the health department clinics, 12% used these clinics as a medical home. Shefer et al25 concluded that use of intensive immunization-promoting strategies in the Special Supplemental Nutrition Program for Women, Infants, and Children might improve utilization of well-child care visits and receipt of other clinical preventive services in the medical home. In a smaller study of 495 children vaccinated in Delaware in the early 1990s, Ortega et al26 found that the medical home was not significantly associated with immunization coverage. We note that the vaccination histories of these children were drawn from the Delaware immunization registry, which might have had incomplete reporting of children's vaccination histories at that time. Furthermore, many of these children received vaccinations at a time before the VFC-era and thus might have been more likely to be referred to public clinics to receive vaccinations.
Our study showed that, nationally, slightly less than one half of all children 19 to 35 months of age were eligible to receive publicly purchased vaccines at no cost from VFC-enrolled vaccination providers. The percentage of VFC-eligible children receiving vaccinations at a VFC provider was very high. This suggests that the VFC program has been largely successful in servicing the children it was mandated to serve. Furthermore, our study showed that a high percentage of VFC-eligible children had a medical home and, among those, a high percentage of children used their medical home consistently to receive all of their vaccinations. In the present study, both of these factors were found to have significant beneficial effects that were associated with higher vaccination coverage rates. In fact, VFC-eligible children who used their medical home consistently to receive all of their vaccine doses were found to have vaccination coverage rates that were higher than those of non-VFC-eligible children, after accounting for socioeconomic differences between these 2 groups. Other research5–10 demonstrated significant positive associations between reduction of financial barriers to being vaccinated, an increased likelihood of maintaining a consistent relationship with a child's primary health care provider, and receiving vaccinations from the primary health care provider. Results in the present study are congruent with these previous findings and suggest that the VFC program might have had a positive influence that has fostered the adoption and consistent use of a medical home, leading to high levels of vaccination coverage among children who, before the advent of the VFC program, might have had lower vaccination coverage.
Among VFC-eligible children with a medical home, many did not receive all of their vaccines from their medical home. These children were not as well vaccinated as VFC-eligible children who received all of their vaccinations at their medical home. Furthermore, VFC-eligible children with no medical home were less well vaccinated than VFC-eligible children with a medical home. These results are consistent with a growing body of literature that has documented the association of poor health outcomes among children with their dropping out of their medical home and failing to access primary care services.21,27,28 Failure to access preventive services that require multiple timely visits to the medical home is known to be correlated with suboptimal health outcomes.29–32
These findings suggest that there may be opportunities to increase vaccination coverage by being more deliberate in creating a medical home environment for every child. This environment is defined11 as a place where a bona fide concern for the child is expressed and demonstrated in a culturally effective manner, medical care is provided in an atmosphere of trust by a physician who is known to the child and the child's family, and the family is recognized by the physician as the child's principle caregivers and experts in the child's care. Furthermore, care at the medical home is provided and directed by the same well-trained pediatric physicians and health care professionals, who are able to manage all aspects of medical care from the child's infancy through adolescence and young adulthood; when multiple medical professionals provide the child with care, this care is coordinated through the medical home. Finally, the medical home is provided in the child's community; is accessible with public transportation, when available; accepts all insurance, including Medicaid; and accommodates changes in insurance. Because inconvenience was the most frequently cited barrier preventing the receipt of all vaccines at a medical home in our study, we recommend that efforts to increase the use and effectiveness of medical homes focus on reducing this barrier. Barriers that may be associated with inconvenience may include distance from home to the provider's location, lack of transportation to the location, and difficulties in scheduling appointments. Also, we note that 13.4% of VFC-eligible children who had a medical home were reported as not having received all vaccinations at their medical home because of the high cost of vaccinations. Because VFC-eligible children cannot be denied free VFC vaccines even if they cannot afford the cost of having these vaccines administered, this result suggests that other costs not covered by VFC, such as the cost of the office visit, may be a barrier to consistent use of a medical home and attaining the goal of being UTD with recommended vaccines.
The strengths of this study include the use of data from a survey with a very large sample from each state. One limitation of our study is that VFC eligibility was ascertained at the time of the NIS interview and not between children's birth and their 19th month of age, the period of time when many children receive their vaccinations. As a result, VFC eligibility during this important period is measured with error.33 The consequence of this error is that gaps in vaccination coverage are underestimated in the present study and the effects of having a medical home and using a medical home consistently may be more beneficial than reported in the present study.
Another limitation of the NIS is that it obtains information only from households that have a “land-line” telephone; the NIS does not sample telephone numbers that are known to be cellular telephone numbers and thus does not interview households that have a cellular telephone only. In survey research, these household are said to be “not covered” by the NIS sampling frame. Another limitation of the NIS is that nonresponse is incurred at various stage of data collection. Smith et al19 provided a thorough quantitative description of noncoverage of the target population by the NIS sampling frame and nonresponse among households covered by the sampling frame. Unfortunately, we do not know the UTD status of children living in households that were not covered by the NIS sampling frame or children who were NIS nonresponders. It is not possible to know precisely how our conclusions might change as a result of not knowing this information. However, if having missing information is correlated with not having a medical home, inconsistent use of a medical home, and not being UTD, then children with these characteristics would have been under-represented by the NIS HIM. If this was the case, then gaps in vaccination coverage in the present study would probably be underestimated and the effect of having a medical home and using a medical home consistently could be even more beneficial than we have concluded from our data.
- Accepted January 6, 2005.
- Address correspondence to Philip J. Smith, PhD, Centers for Disease Control and Prevention, National Immunization Program, MS E-32, 1600 Clifton Rd, NE, Atlanta, GA 30333. E-mail:
No conflict of interest declared.
- ↵US Department of Health and Human Services. The Childhood Immunization Initiative: HHS fact sheet, July 6, 2000. Available at: www.hhs.gov/news/press/2000pres/20000706a.html. Accessed May 12, 2005
- ↵Social Security Online. Compilation of the Social Security laws: Program for Distribution of Pediatric Vaccines: SEC 1928 [42 USC 1396s]. Available at: www.ssa.gov/OP_Home/ssact/title19/1928.htm. Accessed May 12, 2005
- ↵Schulte JM, Bown GR, Zetzman MR, et al. Changing immunization referral patterns among pediatricians and family practice physicians, Dallas County, Texas, 1988. Pediatrics.1991;87 :204– 207
- Ruch-Ross HS, O'Connor KG. Immunization referral practices of pediatricians in the United States. Pediatrics.1994;94 :508– 513
- Zimmerman RK, Mieczkowsksi T, Mainzer H, et al. Effect of the Vaccines for Children program on physician referral of children to public vaccine clinics: a pre-post comparison. Pediatrics.2001;108 :297– 304
- ↵American Academy of Pediatrics, Medical Home Initiative for Children with Special Needs Project Advisory Committee. The medical home. Pediatrics.2002;110 :184– 186
- ↵American Academy of Pediatrics. The medical home. Pediatrics.1992;90 :774
- ↵Frankel LR. The report of the CASRO task force on response rates. In: Wiseman F, ed. Improving Data Quality in Sample Surveys. Cambridge, MA: Marketing Science Institute; 1983: 1–11. Available at: www.casro.org/resprates.cfm. Accessed May 12, 2005
- ↵Research Triangle Institute. SUDAAN User's Manual, Release 8.0. Research Triangle Park, NC: Research Triangle Institute; 2002
- Smith PJ, Rao JNK, Battaglia MP, et al. Compensating for provider nonresponse using response propensities to form adjustment cells: the National Immunization Survey. Vital Health Stat 2.2001;(133):1–17. Available at: www.cdc.gov/nchs/data/series/sr_02/sr02_133.pdf. Accessed May 12, 2005
- ↵Smith PJ, Hoaglin DC, Battaglia MP, et al. Statistical methodology of the National Immunization Survey, 1994–2002. Vital Health Stat 2.2005;(138):1–55. Available at: www.cdc.gov/nchs/data/series/sr_02/sr02_138.pdf. Accessed May 12, 2005
- ↵Smith PJ, Simpson D, Battaglia MP, et al. Split sampling design for topical modules in the National Immunization Survey. In: 2000 Proceedings of the Section on Survey Research Methods. Alexandria, VA: American Statistical Association; 2000:653– 658
- ↵Santoli JM, Rodewald LE, Maes EF, Battaglia MP, Coronado VG. Vaccines for Children Program, United States, 1997. Pediatrics.1999;104 (2). Available at: www.pediatrics.org/cgi/content/full/104/2/e15
- ↵Kempe A, Beaty B, Englund BP, Roark RJ, Hester N, Steiner JF. Quality of care and use of the medical home in a state-funded capitated primary care plan for low-income children. Pediatrics.2000;105 :1020– 1028
- ↵Szilagyi PG, Holl JL, Rodewald LE, et al. Evaluation of children's health insurance: from New York State's Child Health Plus to SCHIP. Pediatrics. 200;105(3 suppl):687–691
- ↵Ortega AN, Stewart DC, Dowshen SA, Katz SH. The impact of a pediatric medical home on immunization coverage. Clin Pediatr (Phila).2000;39 :89– 96
- ↵Bordley WC, Margolis PA, Lannon CM. The delivery of immunizations and other preventive services in private practices. Pediatrics.1996;97 :467– 473
- ↵Biemer PP, Groves RM, Lyberg LE, Mathiowetz NA, Sudman S. Measurement Errors in Surveys. New York, NY: Wiley Series in Probability and Statistics; 2004
- Copyright © 2005 by the American Academy of Pediatrics