Background. There is concern that commercial health insurance reimbursement levels for immunizations and well-child visits may not be meeting the delivery and practice overhead costs within some areas of the country. There is also concern that insufficient physician reimbursement levels may negatively affect the quality of children’s health care.
Objective. We examined the relationships between commercial health insurance reimbursement levels to physicians for pediatric services and rates of immunization and well visits for children and adolescents.
Design. Quality of care was measured by examining state-level immunization and well-visit rates for 2002, which were obtained from the National Committee for Quality Assurance, Health Plan Employer Data and Information Set (HEDIS). Reimbursement data were obtained from the American Academy of Pediatrics Medical Cost Model. Variations in the child and adolescent HEDIS measures were examined as a function of physician reimbursement levels for pediatric services across states. HEDIS data were available for a total of 32 states. Partial correlations controlled for pediatrician concentration, as collected from the US Bureau of the Census and the American Medical Association Physician Masterfile data.
Results. Compliance with HEDIS immunization rates for all recommended vaccines was 60% for children and 24% for adolescents. By excluding the varicella vaccine, these rates increased to 70% for children and 44% for adolescents. Adherence rates for well visits were also higher for infants (60%) and children (59%) than for adolescents (34%). Physician reimbursement levels for pediatric services varied from $16.88 per member per month to $32.06 per member per month across states. Statistically significant positive correlations for reimbursement levels were found for 8 of the 16 HEDIS measures examined. Correlations with reimbursement levels were found for childhood immunizations (r = 0.42), infant well visits (r = 0.44), childhood well visits (r = 0.46), and adolescent well visits (r = 0.42). Reimbursement levels were especially strongly related to the rates of adolescent varicella vaccination (r = 0.53). When partial correlations were examined to control for pediatrician concentration, the correlations were reduced by 0.09 on average, suggesting that pediatrician supply may serve as an intermediary of the reimbursement relationship.
Conclusions. Immunization and well-visit rates for infants, children, and adolescents were positively linked with physician reimbursement rates for those services. Although methodologic limitations suggest caution when interpreting these findings, more attention should be given to physician reimbursement levels as a possible predictor of immunization and well-visit rates as measures of quality of care and to the importance of reimbursement levels for pediatrician recruitment.
In some areas of the country, physician reimbursement levels for immunizations and health-supervision visits may not be adequate to cover both vaccine delivery and overhead costs.1 This may serve as a barrier that would limit pediatricians and family physicians from administering childhood vaccines and adhering to the recommended health-supervision periodicity schedule.2,3 Instead, these doctors may refer their patients to health departments and public clinics for immunizations and health-supervision visits, resulting in fragmented care and breaking up their medical home.4,5 The unintended consequences of fragmented care may be lower immunization and health-supervision visit rates or duplication of services and higher costs.
Physician reimbursement levels may also impact immunization and well-child visit outcomes indirectly through the effect on physician recruitment and geographic distribution. A strong correlation has been found between immunization rates and the concentration of pediatricians in an area.6 Because potential income is a strong determinant used by pediatricians when considering the location of their practice, more pediatricians may locate to better-paying states, leading to higher immunization and well-visit rates.
Several factors have been linked to higher immunization and well-child rates,7–12 but to date no data have been collected that examine the relationship of reimbursement levels with immunization and health-supervision visit rates. The hypothesis that higher-reimbursing states would also have higher immunization and well-child visit rates was examined in this study, focusing on patients with commercial health insurance.
Information was collected from several secondary data sources to examine possible associations between commercial reimbursement levels and adherence to American Academy of Pediatrics (AAP) well-visit and immunization recommendations. Adherence rates were assessed by using the Health Plan Employer Data and Information Set (HEDIS) measures based on the AAP recommendations. The measures were purchased from the National Committee for Quality Assurance (NCQA) for 2002. NCQA reports that HEDIS data collection is a tool used by >90% of America’s health plans, including both health maintenance organization and point-of-service plans. Data reported from commercial health plans were specifically considered in this study by using the state-level HEDIS summary file. This analysis focused on infant, child, and adolescent HEDIS measures for immunizations and health-supervision visits.
The combined immunization measure for children included 4 diphtheria-pertussis-tetanus, 3 polio virus, 1 mumps-measles-rubella (MMR), 2 hemophilus influenza type B (Hib), and 3 hepatitis B immunizations by the age of 2 years. A second combined measure was also computed that included 1 varicella-zoster virus (VZV) immunization in addition to the other immunizations. The combined adolescent measure included a second dose of MMR and 3 hepatitis B immunizations by the adolescent’s 13th birthday. A second combined measure was also created that included 1 VZV immunization.
NCQA has taken several actions to ensure the integrity of the state-level HEDIS measures. First, a measure was considered reliable and included in the data set only when ≥5 health plans within a state reported data. The state-level summaries also included data from some commercial health plans that elected not to have their plan-level data reported publicly. The state-level averages provided by NCQA represent simple averages of health plans that reported results rather than weighted averages. Second, all data in this analysis passed the NCQA compliance audit. This audit is an independent check of the health plan’s information practices and control procedures, their sampling methods, their compliance with measurement specifications, their analytic file production, and their reporting and documentation. Additionally, HEDIS immunization measures were correlated with the National Immunization Survey 2002 results to provide a separate check of data quality. Correlations ranged from r = 0.66 (P < .001) to r = 0.71 (P < .001) for the combined childhood immunization measures. This agreement is consistent with other comparisons of HEDIS measures and non-HEDIS survey measures.13,14
Commercial reimbursement levels for pediatric medical care was the primary predictor variable examined in this study. Commercial reimbursement levels were obtained from the AAP Medical Cost Model.15 This model was developed by Reden & Anders, Ltd (San Francisco, CA; formerly Tillinghast Health) and based on actuarial analysis of the AAP’s recommended children’s health benefit package. By using the Reden & Anders national actuarial database, the monthly cost of providing a child with a comprehensive benefit package on a state-by-state basis was estimated.
The model provided utilization, unit cost, and per member per month (PMPM) data by state. The PMPM values reflected the estimated amount insurers paid for an array of benefits and coverage policies. In other words, they were the net medical costs paid by the insurer. This is essentially physician reimbursement for services, with the exception that some independent practice association plans or other administrative middle organizations may not pass along all reimbursement to physicians. The PMPM values should not be confused with patient premiums.
In deriving PMPM cost estimates for the children’s population, Reden & Anders relied on data sources contained in their database and other publicly available information. Their commercial database includes ∼1.5 million member years of medical and pharmacy claim data from patients aged 0–21. For the current study analyses, the costs were tailored to the care provided in ambulatory office visits by limiting the model to 4 cost categories: office visits, preventive care, immunizations/injections, and hearing examinations. In determining average plan costs, Reden & Anders combined fee-for-service payments with deaggregated capitated payments based on utilization estimates and local payment levels determined by the Centers for Medicare and Medicaid Services Geographic Practice Cost Index.
Pediatrician supply, relative to the number of children in each state, was examined as a possible mediator of any correlations between reimbursement levels and the adherence measures. Per child pediatrician supply was computed by the Dartmouth Center for the Evaluative Sciences, using American Medical Association (AMA)/American Orthopaedic Association Masterfile data for physician supply and US census data for number of children in each state.16 A ratio of the number of pediatricians per 100 000 children was computed for our analyses. Thus, higher values represent a greater relative supply of pediatricians in the state.
Reimbursement, physician supply, and HEDIS adherence data were available for 32 states. Simple correlations between reimbursement levels and the adherence measures were conducted. Partial correlations with the adherence measures were then computed for physician reimbursement levels that controlled for the effect of pediatrician supply. If pediatricians chose to practice in higher-reimbursing and more affluent states, then pediatrician supply would emerge as a mediator of possible relationships of reimbursement levels and the adherence measures. A P value ≤.05 was used for all statistical tests. This study was reviewed by the AAP Institutional Review Board and was granted an exemption from additional review.
The HEDIS immunization measures for children provide the percentage of children who received the AAP-recommended number of vaccines by age 2 (Table 1). For the combined measures, an average of 70% of children were fully immunized (excluding the VZV vaccine), and 60% were fully immunized when including the VZV vaccine. These values show that there is still room for improvement in childhood immunization rates. The SDs for these immunization measures ranged from 3% to 7%, showing that a variation in rates exists across states. Immunization rates for adolescents were lower, especially for the hepatitis B and VZV vaccines. SDs were higher for the adolescent immunization measures (12–15%) indicating greater state-to-state variability.
Similarly, there was a large gap between the adolescent and child health-supervision visit rates. The health-supervision visit rate was 34% for adolescents compared with 59% and 60% for 3- to 6-year-olds and infants, respectively (Table 2). SDs between states for these measures ranged from 9% to 11%.
On average, physicians were reimbursed $22.19 PMPM for ambulatory care activities: office visits, preventive care, immunizations/injections, and hearing examinations. Reimbursement values ranged from a low of $16.88 in Missouri to $32.06 in Indiana. Those states with higher reimbursement rates consistently showed higher immunization and well-visit rates than did states with lower reimbursement rates. Specific values by state are identified in the scatter plots for the child combined immunization (Fig 1) and for the age-specific health-supervision visit (Figs 2–4) measures. The Pearson correlation values corresponding to each of these relationships are also presented in Tables 1 and 2. Statistically significant positive correlations were found for 8 of the 16 measures; correlation values ranged from r = 0.01 to r = 0.53 and were positive for all measures. The strongest correlation was for the adolescent VZV immunization, which contributed to a significant correlation for the combined adolescent immunization measure that included VZV (r = 0.43).
When partial correlations were computed to control for pediatrician supply, on average the correlation coefficients dropped in value by 0.09. For most measures, this drop resulted in a loss of statistical significance for the association. The 3 exceptions were adolescent varicella, infant well visits, and well visits for ages 3 to 6 years. These results suggest that pediatrician concentration may serve as an important mediator of the various relationships between physician reimbursement levels for pediatric ambulatory services and immunization and well-visit rates.
Positive relationships were consistently found between physician reimbursement for pediatric services and the various HEDIS immunization and well-visit measures examined in this study. Although these results are clear in showing that states with higher reimbursement levels have higher immunization and well-visit rates, there are a number of factors that may be involved in producing these associations. A direct relationship between reimbursement and HEDIS outcomes is understandable. Higher reimbursement levels may allow physicians to better staff their offices and support a well-organized system that could remind parents when their child is due for immunizations or well-child visits. Higher reimbursement levels may also increase the likelihood that immunizations would be administered on-site rather than referring patients to public health departments and public clinics.
Pediatrician supply as a mediator of reimbursement levels and HEDIS relationships was examined also. Partial correlations were consistent in showing reliable shared variability, with pediatricians more likely to locate in states with higher reimbursement levels. A greater concentration of pediatricians rather than family physicians may increase immunization and well-visit rates, because pediatricians are less likely to refer patients for immunizations and other services17,18 and may be more likely to adhere to the AAP well-visit schedule. Although a consistent drop in the various partial correlations was seen when controlling for pediatrician supply, this did not account completely for the associations, because the values remained positive and many remained statistically significant. It is very likely that other factors may be involved in the relationship between reimbursement levels and child health outcomes.
Patient and family characteristics, for example, may influence the reimbursement and HEDIS relationships. That is, both reimbursement levels and pediatrician concentration may covary with patient or family characteristics. The patient factor most strongly associated with child health indicators is insurance status. However, this factor is not likely to have played as important a role in this study, because the data were based exclusively on patients who had commercial insurance. Thus, all patients in this study should have had access to and payment for health care services. Other family characteristics, such as parental adherence to medical advice, could not be accounted for but may have influenced the child health performance indicators. The correlational study design did not allow for the identification of all factors that could be related to immunization and well-visit rates but did reveal that physician reimbursement is a factor that may be deserving of increased attention.
Efforts to increase reimbursement levels may be among several solutions that might help increase immunization and health-supervision rates. Although the correlational relationships were generally positive across the full range of reimbursement levels found in our study, states with reimbursement levels above $25 PMPM showed the highest performance levels. Immunization rates still need to be improved from the levels shown in this study to prevent the rise of diseases that may claim the lives of numerous children. Likewise, health-supervision visit schedules also need to be adhered to more consistently to reduce the number of avoidable hospitalizations and emergency department visits.19,20
Especially, there is a need to improve immunization and health-supervision visits rates for adolescents. Pediatricians should continue to strive to provide continuous care to children as they progress through adolescence. Additionally, greater efforts are needed to coordinate care with other primary care physicians and gynecologists to ensure that adolescents receive scheduled immunizations and well visits. Although the adolescent age group will always be a difficult group for which to attain high health-supervision rates, the current levels are not acceptable.
These results should be interpreted with caution because they are state-level analyses. Within each state, there may be areas with higher or lower immunization or well visit rates. Similarly, reimbursement levels vary by location. The ideal study would be to conduct similar analyses at the plan or local-patient levels. Perhaps future researchers will be able to surmount the barriers that exist to obtaining physician reimbursement data from commercial plans at these levels.
Other limitations of this study should also be noted. Reliable HEDIS data were available for only 32 states, limiting our power to detect correlations. If data were available for more states, perhaps additional correlations and partial correlations may have reached statistical significance. Also, Medicaid reimbursement levels could not be considered in the current study, because there were not enough Medicaid plans reporting to NCQA to allow reliable Medicaid state-level summaries. As HEDIS data collection efforts continue to grow in the future, these analyses may be possible.
Immunization and well-visit rates for infants, children, and adolescents were positively linked with physician reimbursement for services. Although methodologic limitations suggest caution when interpreting these findings, at the very least, more attention should be paid to physician reimbursement as a predictor of the quality of care that children and adolescents receive. States also need to consider the effect that reimbursement levels may have directly on physician recruitment and indirectly on patient outcomes. Additional research, ideally at the plan or patient levels, is needed to examine these relationships for other child health measures in addition to immunizations and health-supervision visits.
- Accepted August 9, 2004.
- Address correspondence to William L. Cull, PhD, Division of Health Services Research, American Academy of Pediatrics, 141 Northwest Point Blvd, Elk Grove Village, IL 60007. E-mail:
No conflict of interest declared.
- ↵American Academy of Pediatrics, Committee on Infectious Diseases. Recommended childhood and adolescent immunization schedule—United States, 2005. Pediatrics.2005;115 :182– 186
- ↵American Academy of Pediatrics, Committee on Practice and Ambulatory Medicine. Recommendations for preventative pediatric health care. Pediatrics.2000;105 :645– 646
- ↵LeBaron CW, Massoudi M, Stevenson J, Lyons B. Vaccination coverage and physician distribution in the United States, 1997. Pediatrics.2001;107(3) . Available at: www.pediatrics.org/cgi/content/full/107/3/e31
- ↵Koepke CP, Vogel CA, Kohrt AE. Provider characteristics and behaviors as predictors of immunization coverage. Am J Prev Med.2001;2 :250– 255
- Bordley WC, Margolis PA, Lannon CM. The delivery of immunizations and other preventive services in private practices. Pediatrics.1996;97 :467– 473
- Wood D, Donald-Sherbourne C, Halfon N, et al. Factors related to immunization status among inner-city Latino and African-American preschoolers. Pediatrics.1995;96 :295– 301
- Yu SM, Bellamy HA, Kogan MD, Dunbar JL, Schwalberg RH, Schuster MA. Factors that influence receipt of recommended preventive pediatric health and dental care. Pediatrics.2002;110(6) . Available at: www.pediatrics.org/cgi/content/full/110/6/e73
- ↵American Academy of Pediatrics. 2002 Pediatric medical cost model. Available at: www.aap.org/research/pedmedcostmodel.cfm. Accessed December 16, 2004
- ↵American Academy of Pediatrics. Mapping health care delivery for America’s children. Available at: www.aap.org/mapping. Accessed December 16, 2004
- ↵Hakim RB, Bye BV. Effectiveness of compliance with pediatric preventive care guidelines among Medicaid beneficiaries. Pediatrics.2001;108 :90– 97
- Copyright © 2005 by the American Academy of Pediatrics