PEDIATRICS Vol. 105 No. 5 May 2000, pp. 1020-1028

Quality of Care and Use of the Medical Home in a State-Funded Capitated Primary Care Plan For Low-Income Children

Allison Kempe, MD, MPH^*, , Brenda Beaty, MSPH^*, Brian P. Englund, MS, CHA/PA-C^*, Robert J. Roark, MS, CHA/PA-C^*, Nellie Hester, BA^*, and John F. Steiner, MD, MPH^{, §}

From the Departments of ^* Pediatrics,  Preventive Medicine and Biometrics, and ^§ Internal Medicine, University of Colorado Health Sciences Center, Denver, Colorado.

	ABSTRACT

Top Abstract Methods Results Discussion References

Objective.  To evaluate the quality of care and use of the medical home in a state-funded capitated insurance plan for low-income childrenthe Colorado Child Health Plan (CCHP).

Design.  A retrospective cohort study using medical record review at pediatric and family practice offices in 4 geographic areas of Colorado. At each practice, CCHP-enrolled children (6 months to 6.5 years) and 2 controls were selected, 1 with Medicaid (MK) and 1 with private insurance (PI), matched by date of birth to the CCHP-enrolled child (N = 596). CCHP-enrolled children with a diagnosis of asthma, aged 3 to 18 years, and asthmatic children with MK and PI, matched by age, were also selected from each practice (N = 139).

Results.  Quality of preventive services were comparable in the 3 groups. CCHP-enrolled children made more health maintenance visits than MK-enrolled children (1.3 CCHP vs .9 MK vs 1.1 PI) and were more frequently screened for lead (8.1% CCHP vs 3.4% MK vs 1.2% PI) and anemia (5.0% CCHP vs 4.4% MK vs 2.4% PI) than children in either control group. Documented immunization rates were similar in the 3 groups, but a shift in location of immunization from public health clinics to the primary care site was seen in the CCHP group. CCHP-enrolled children made more office visits for acute care than did MK-enrolled children (4.1 CCHP vs 3.1 MK vs 3.4 PI), but a higher proportion of these visits took place at the medical home rather than the emergency department for the CCHP group (.04) as compared with the MK (.07) or PI (.06) groups. Asthmatic children in the CCHP group made more preventive office visits for maintenance therapy and more frequently used the primary care site rather than the emergency department for acute exacerbations than did children with PI (mean ratio of emergency department visits to total acute visits .04 CCHP vs .06 MK vs .19 PI).

Conclusions.  Despite capitated reimbursement for primary care services, CCHP provided children from low-income families with preventive, acute, and chronic care services of comparable quantity and quality to those received by children with MK or PI. The program was associated with a shift of immunization location to the primary care site and increased health maintenance care for new enrollees. CCHP-enrolled children used their medical home for the majority of acute health needs and were not high utilizers of emergency department or hospital services.  Key words:  indigent care, health insurance, quality of care.

Approximately 10.5 million children under 19 years old in the United States are currently uninsured.¹ Despite continued growth in employment and expansions of Medicaid (MK) coverage, the numbers of uninsured and underinsured children are projected to grow.² The majority of uninsured children in the United States are from families of the working poor, with at least 1 parent working in 83% of cases.² Many of these families have incomes too high to qualify for MK coverage, but too low to afford employer-based insurance.¹ The Balanced Budget Act of 1997 established the State Children's Health Insurance Program (SCHIP), which allocated $24 billion to states over a 5-year period to provide health insurance to children who would otherwise be uninsured.³ Currently, 36 states have either expanded MK coverage beyond federal requirements or have developed programs with at least some state assistance to cover uninsured children in low-income families.² A major policy objectives of such state programs is to establish a primary care site or medical home for uninsured patients with the aim of shifting health care utilization patterns away from episodic care in expensive settings and toward continuity of care at the primary care site. To date, although state programs have reportedly expanded coverage to more than half a million children,² there has been little evaluation of the quality of health services received by children enrolled in these programs or of their success in increasing use of the medical home. In addition, few data are available regarding utilization of services by previously uninsured children to aid in the planning of health care services.

The Colorado Child Health Plan (CCHP), which began enrollment in 1992, is a health care reimbursement plan for low-income children in Colorado families living below 185% of the federal poverty level who do not qualify for MK. The CCHP currently provides outpatient care to 12 051 children under the age of 18 throughout Colorado. We compared the care received by children enrolled in the CCHP program with that received by children in the same practices with MK or private insurance (PI) with respect to the quality of care, use of the medical home, and patterns of health care utilization for 1) preventive care services, 2) care for acute illness, and 3) care for children with chronic illness.

	METHODS

Top Abstract Methods Results Discussion References

The CCHP

Covered services of the CCHP included outpatient primary care and acute care services, subspecialty care, outpatient surgery, laboratory, radiology, and limited pharmacy benefits. CCHP contracted directly with designated providers throughout the state rather than with large health maintenance organizations, as distribution of health maintenance organizations is not uniform in Colorado. As of December 1997, CCHP had recruited >1000 primary care pediatricians and family practitioners and >1000 specialists to its provider network. The program was administered by the University of Colorado Health Sciences Center and was funded from a combination of state appropriations and private donations.

Reimbursement and Utilization Management by CCHP, MK, and PI

During the study period, the CCHP reimbursed all primary care services at a capitated rate, while emergency department visits and subspecialty care by designated subspecialists were reimbursed on a fee-for-service basis. There was a $2 copay for all types of visits. Emergency services required approval by the primary care physician. CCHP did not cover hospitalization costs, but if CCHP patients were hospitalized another program for the uninsured paid the majority of the bill, with the patient's family being responsible for a portion determined by their yearly income.

In the practices enrolled in this study, capitated reimbursement plans covered 0% (6 practices), 10% (1 practice), and 50% (1 practice) of children covered by PI and none of the children covered by MK. No copayment was required for MK-enrolled patients and copayment requirements varied for the PI groups. For the majority of emergency department visits by MK or PI patients, prior approval was not required. Hospitalization costs were generally fully covered for children with MK or PI.

Study Populations

To generate a representative sample of children throughout the state of Colorado, we conducted the evaluation in 4 geographic areas of the state, Mesa County (Grand Junction), La Plata County (Durango), Larimer County (Fort Collins), and El Paso County (Colorado Springs). In these 4 areas we reviewed medical records in the 8 pediatric and family practice physician offices and community health centers with the largest number of CCHP enrollees. Because approximately half of the children in these areas are cared for by family physicians, we attempted to enroll practices so that we would have roughly equivalent numbers of children with family practice and pediatric providers. The majority of the population in these 4 areas was white, ranging from 85.1% in Colorado Springs to 94.4% in Grand Junction. Median household income ranged from $22 996 in Durango to $28 610 in Colorado Springs. The percentage of children 5 years old or younger below the federal poverty level ranged from 13.5% in Fort Collins to 28.6% in Durango.⁴

In assessing quality of care and utilization of services we compared groups of CCHP-enrolled children and, within the same practices, children of the same age with MK and PI. The comparison groups were selected because CCHP-enrolled children were expected to be intermediate between MK and PI populations with respect to sociodemographic variables that are correlated with health care utilization. Thus, the comparison groups were approximations of both ends of the sociodemographic spectrum in the practices and were reflective of usual care for children in these insurance strata.

For the review of preventive and acute ambulatory services, children 6 months to 6.5 years of age enrolled in the CCHP program for a minimum of 6 months and enrolled in the study practice for at least 6 months were randomly selected from computerized office records. To ensure sufficient numbers from each age category, children were then randomly selected in the 2 age categories of 6 to 24 months and >24 months up to 6.5 years, up to a maximum of 25 subjects in each age category per practice. Each CCHP subject was then matched to a MK and a PI child enrolled in the same practice for a minimum of 6 months with the closest date of birth. These study populations were designated the preventive services study group (total N = 596). Of the 221 CCHP children, 155 had 2 matches and 67 had 1 match.

In reviewing quality of care for children with chronic illness, we selected CCHP-enrolled children 3 years to 18 years of age with CCHP coverage for a minimum of 6 months with a diagnosis of asthma or reactive airway disease. Each case was then matched to an asthmatic child with MK and an asthmatic child with PI with the closest date of birth. Children were identified as asthmatic from computerized office billing data if they 1) were 3 to 18 years old, 2) had a billing diagnosis of asthma or reactive airway disease at least once (International Classification of Diseases, Ninth Revision, Clinical Modification codes 493.0-493.9)⁵ or a diagnosis of wheezing on 2 or more occasions. They were retained as a case if, after medical record review, their medical record substantiated a diagnosis of asthma or reactive airway disease or at least 2 documented episodes of wheezing having occurred with -adrenergic therapy given. These groups were designated the Asthma Study Groups (total N = 139).

Children were excluded from the study only if the patient had not been seen in the practice for 2 or more years (N = 26 from all groups; 6 CCHP, 13 MK, and 6 PI), if the patient died during the period of review (N = 1) or if the medical record documented that the patient had moved or gone elsewhere for care (N = 43). If a CCHP patient was excluded, the MK and PI matches for this child were also excluded. If a MK or PI match was excluded, the child with the next closest date of birth from the same insurance category was enrolled. Medical records were located for all children selected by these criteria.

For all groups, medical records were reviewed for the time period during which CCHP had been active in the area until the end of the study period, July 1, 1997. The start dates at the 4 locations were as follows: October 1, 1992 for Mesa County; September 1, 1993 for La Plata County; and June 1, 1996 for Larimer and El Paso Counties; therefore, the period of review varied between approximately 1 and 4.5 years. The chart reviewers were 2 pediatric nurses and 2 child health associates who had received training in the use of a standardized chart review tool.

Outcome Measures

In assessing health care utilization rates for each CCHP patient, the analytic period extended from the time of the subject's enrollment in the program, termed the index date to the end of the chart review period. For each of the 2 matched control subjects, the analytic period began at the age corresponding to the age of the CCHP subject at enrollment (termed the index age) and ended at the end of the chart review.

Quality of Care of Preventive Services

We chose outcome measures for quality of preventive services that were accessible through review of the child's medical record and reflected the guidelines of major professional pediatric and family practice societies. We used the Clinical Assessment Software Application (CASA), developed by the Centers for Disease Control, to assess immunization status.⁶ We also examined whether catch-up in immunization up-to-date levels occurred by comparing the number of documented immunizations given in each insurance group at 2 months and 6 months after the index date for children who still needed immunizations at the index date. Additional measures used in assessing quality of preventive health incorporated recommendations of the Bright Futures project⁷ including: the number of health maintenance visits, the percent of health maintenance visits at which developmental assessments were performed, injury prevention counseling was done and growth parameters were measured and plotted on growth charts. For children 9 months or older, we also determined the percent who had been tested for anemia and who had been tested for or were documented in the chart to be in a low-risk category for lead exposure or tuberculosis. These criteria reflected the stated screening policies of the participating practices at the time of the study.

Utilization of Acute Care Services

To assess utilization of acute health care services we compared: 1) total number of outpatient visits for acute illness, 2) the total number of emergency department visits and the rate of emergency department utilization per child, 3) the number of hospitalizations and the rate of hospitalization per child and 4) the percentage of all visits for acute illness that occurred in a primary care setting, rather than in an emergency department, for children in the different insurance categories. Because chronic medical conditions might increase the need for emergency department visits or hospitalization, we controlled for the presence of any such condition in our analyses. Although the definition of chronic conditions has been controversial in the pediatric literature,^8-14 we classified patients as having a chronic condition if they had a diagnosis that was expected to last >3 months^12,14 and might result in functional impairment or the use of more medical services than expected for a child of the same age.¹⁴ We did include the diagnoses of recurrent otitis media or chronic otitis media, although these are sometimes classified as recurrent acute conditions^13,14 but required that included cases meet the criteria of having 3 or more episodes of otitis within 6 months or 3 or more months of unresolved otitis.¹⁵

Quality of Care of Services for Chronic Illness

We examined quality of care in asthmatic children because asthma is the most prevalent chronic disease in children and is an illness for which preventive care can reduce severity and morbidity. Based on recommendations of the National Heart, Lung and Blood Institute¹⁶ and our assessment of outcomes that could be found in medical records, we examined the following quality of care outcome measures in asthmatic children: 1) the number of office visits specifically for maintenance or preventive care for asthma when there was not an acute exacerbation, termed preventive asthma visits, 2) the number of emergency department visits or hospitalizations for asthma, and 3) the percentage of all visits for acute exacerbations that took place in an emergency department (rather than a primary care setting) per child. In addition, we examined and compared the following process of care measures in the different insurance groups: 1) the use of preventive inhaled antiinflammatory medications, 2) the use of home or office monitoring measures (peak expiratory flow rate) and, 3) the rates of subspecialty referrals for asthma.

Data Analysis

Univariate group comparisons were performed using t tests for continuous variables and ² tests for dichotomous variables. Analysis of variance using General Linear Models was used for continuous outcomes and logistic regression was used for dichotomous outcomes in multivariate analyses. In all analyses we controlled for chronic condition, practice type, age, and whether the child was seen in the practice before the index date. These variables were chosen because they might independently affect the outcomes of interest. McNemar's test for paired proportions was used for the analysis of location of immunizations within insurance groups before and after the index date. All tests were 2-sided and were considered to be significant if P was < .05. CASA was used to determine immunization up-to-date rates at 12 and 24 months of age for all groups. For all other analyses, SAS version 6.12 (SAS Institute, Cary, NC) was used.

	RESULTS

Top Abstract Methods Results Discussion References

Preventive and Acute Care Services

Of the 596 medical records reviewed, 221 children (37.1%) were covered by CCHP, 205 (34.4%) by MK, and 170 (28.5%) by a PI plan. Demographic characteristics of the preventive services study group in the 3 insurance categories are listed in Table 1. Because of the matched design, there were no statistically significant differences in mean age or the proportion of children >2 years old and <2 years old between CCHP children and those covered by MK or PI. Only 6.4% of the individual matches differed by >2 months in age. The percentage of children with chronic conditions was very similar in the CCHP and MK groups and slightly lower, although not significantly, in the PI group. Differences in the mean total duration of time seen in the practice did not differ statistically, although there was a trend for PI patients to have been in the practice for a longer time. The percentage of each group that had been seen within the practice before the index date for CCHP enrollment was similar in the 3 groups. The clinical specialties of the providers surveyed for this study were also similar among insurance groups, with 37% to 50% of children receiving care from family physicians.

Table 2 shows documented immunization data for children in the 3 insurance groups. There were no significant differences in the percentage of CCHP, MK, and PI children with documented immunizations up to date at 12 months or 24 months. Documented immunization up-to-date rates at the time of enrollment for CCHP children and at the corresponding age for the MK and PI were also similar. The rates of receiving immunizations after the index age, assessing catch-up, were similar at 2 months and at 6 months in the 3 groups. The proportion of children who received at least 1 documented immunization at a public health clinic site before the index date was higher in the CCHP group than in the PI group (P < .01) and similar in the CCHP and MK groups. This proportion decreased substantially within the CCHP and MK groups and remained unchanged in the PI group after the index date (P < .01 for before vs after within group comparisons for CCHP and MK groups by McNemar's test). The proportion receiving an immunization at the current primary care site was significantly higher after the index date than before the index date for CCHP children (16% increase; P < .01 for before vs after within group comparison for CCHP group by McNemar's test) and was less evident for children with MK (3%, P = not significant [NS]) or PI (11%, P = NS).

The CCHP group had a significantly higher mean number of health maintenance visits over the period of review than the MK group, with an intermediate number for the PI group (Table 3). The quality of preventive services provided during health maintenance visits was generally similar between the groups. However, CCHP children were more frequently screened for anemia and lead than their counterparts in either control group.

As Table 4 demonstrates, use of the primary care site for acute problems was significantly higher for CCHP children than for MK children and intermediate for PI children. The ratio of acute visits that took place in the emergency department to total outpatient acute visits, however, was significantly lower for CCHP children, compared with both control groups. The mean number of hospitalizations was significantly higher in the PI group compared with the CCHP group.

To assess any change in health care utilization as a result of enrollment into the CCHP program by patients who had been seen in the study practices before enrollment into the program, we also compared use of the medical home for acute and preventive services before and after the index date between the CCHP group and the other 2 groups. As shown in Table 5, the mean number of office visits for acute care did not differ between the CCHP group and either comparison group either before or after the enrollment date, but the mean number of health maintenance visits per child was significantly higher in the CCHP group than for the MK group after enrollment into the CCHP program. Within the CCHP group itself, the adjusted rate of acute visits per month after enrollment was not significantly different among those children with and without a previous visit to the practice (.29 vs 9.26, respectively, P = .99). However, children without a previous visit to the practice had a higher adjusted rate of health maintenance visits per month after enrollment (.11 vs 1.09, respectively, P < .01) when compared with CCHP children who had been seen at the practice before enrollment.

Services for Children With Asthma

Table 6 describes the demographic characteristics of the asthma study group by insurance type. No significant differences were seen in the age or gender of children enrolled in CCHP when compared with children enrolled in MK or PI. Of the asthmatics in our sample, 89% or more were cared for by pediatricians rather than family practitioners. Because the practices were selected partially based on size of the CCHP population, this figure may not be reflective of the distribution of care for asthmatics throughout the region.

Health care services utilization and measures of process of care in the asthma study group by type of insurance are presented in Table 7. CCHP-enrolled children received a significantly higher number of preventive visits for asthma and a higher mean number of office visits for acute exacerbations than did PI children, while data for CCHP and MK groups were similar. The ratio of emergency department to total outpatient visits for an acute exacerbation of asthma was also significantly higher for PI children than for children enrolled in CCHP and similar for MK children. All other utilization and process measures for asthmatics were similar among the 3 groups.

	DISCUSSION

Top Abstract Methods Results Discussion References

Almost 14% of children in the United States lack health insurance and many more have gaps in coverage for periods of 6 months or more.^17-19 With the resources allocated by SCHIP, state programs have been rapidly expanding coverage for previously uninsured children.¹ The quality of health care services delivered to children enrolled in these state programs has not been extensively evaluated. In addition, although there is ample evidence that uninsured children see physicians less often than do insured children, particularly for preventive health services,^20-29 health care utilization and use of a medical home by the previously uninsured after barriers to access are removed have not been well-studied.

The aims of the CCHP, a state-based primary care reimbursement program for low-income children, were to establish a medical home for previously uninsured children for the delivery of high-quality preventive and acute care with controlled access to expensive emergency and hospital services. Having a regular source of care or a medical home is 1 of the 4 critical features of primary care^30,31 and has been associated with more complete immunization, higher rates of well-child visits, lower rates of visits for illness care, fewer emergency department visits, and 25% lower total costs of care than children with no regular source of care.^32-36 The current study examined quality of health services received and utilization of health services by CCHP-enrolled children as well as the use of the medical home. Our data show that despite capitation of all primary care services, children enrolled in CCHP made more health maintenance visits to their primary care site than did children enrolled in the uncapitated MK program and received preventive services of comparable quality to children with MK or PI. Slightly over half of CCHP-enrolled children appeared to be using the practices for primary care as frequently as children in the other insurance groups before enrollment into CCHP, suggesting that for this subgroup enrollment into CCHP may have reflected a change in payor status rather than entry into the primary care system. However, the shift in documented immunization location from public health clinics to the child's primary care site and the increased rate of health maintenance visits in children not previously seen in the practice, suggested that the program increased use of the medical home for preventive care. Our data show some pent-up demand for both preventive and acute services in previously uninsured children. However, the proportion of acute care visits that took place in an emergency department rather than the primary care site was substantially lower in CCHP children than in MK or PI children.

Few state-based plans for uninsured children have been evaluated formally. Martin et al³⁷ examined health care utilization of new members of a program in Washington State for uninsured people with incomes <200% of the federal poverty level. Their data show some pent-up demand on the part of adults who had been uninsured for >1 year, but, overall, enrollees in the program were not high users of care and total expenditures were comparable to those of traditionally insured families and lower than those for MK recipients. This study did not address quality or content of health services and did not report data for children separately. Lave et al³⁸ examined access to care and barriers to care in new enrollees into 2 Pennsylvania programs providing coverage for children in families with incomes <235% of the federal poverty level. Families reported that access to a regular source of medical care improved and the proportion seeing a physician in an office setting increased, while the proportion reporting any unmet needs or delayed care or visiting an emergency department decreased after enrollment into the programs. However, these findings were based on the recollection of interviewed parents and self-reported utilization, unconfirmed by objective sources. Rodewald et al³⁹ examined content of preventive care for children before and after implementation of the New York State insurance program, Child Health Plus, which covers ambulatory care and immunization services for children of families earning <222% of the federal poverty level. They reported an increase in immunization up-to-date rates of 7% and a shift in both the receipt of immunizations and health maintenance care from public health clinic sites to private primary care provider sites. The most notable effects were seen among previously uninsured children or those with a gap in coverage of 6 months or more.

The present study demonstrates that the quality of preventive and chronic health services received by CCHP-enrolled children was comparable to services received by children with MK or PI coverage in the same practices. The rates of screening for anemia and lead were actually higher in the CCHP group, possibly because providers were more cognizant of the need for catch-up in this group of children. The limited data available from medical record review did not allow us to identify risk criteria to assess the appropriateness of screening. Our review demonstrated that only 47% to 53% of 1-year-olds and 25% to 28% of 2-year-olds in all 3 insurance groups were documented to be up-to-date with immunizations in the 4 surveyed areas of Colorado. These figures underestimate the true rates of immunizations because the CASA reporting system⁶ requires complete documentation of dates for all immunizations.⁴⁰ Previous estimates of immunization rates have ranged widely (between 11% and 82%) depending on the populations and data collection methods used.^41-45 Our finding that CCHP-enrolled children had comparable documented immunization rates to children with other insurance before they enrolled in the program differs from previous reports showing lower immunization rates in uninsured children than in children with insurance.^46,47 The rates in our study may have been similar because of prior MK enrollment in some children or, in younger children, beneficial effects of the federal Vaccines For Children program that made free vaccine available to the uninsured in Colorado's private sector after 1995. In addition, initial enrollees in a program such as CCHP may be a more motivated portion of the uninsured population and not necessarily those with the highest medical need.

Our data also show that children enrolled in the CCHP used their medical home for their preventive and acute health care needs as much as or more than children in the same practices with MK or PI. Capitated reimbursement to private practitioners for primary care services was not a barrier to preventive services for healthy children or for those with asthma. In addition, despite a copay that was considered too small to discourage overutilization by some state legislators, we did not observe overutilization of acute outpatient services by CCHP enrollees. As other studies have shown,⁴⁸ some pent-up demand for acute care services in previously uninsured children was suggested by a higher rate of acute care visits in the CCHP group. However, this acute care generally was received at the primary care site, rather than in the emergency department. In fact, the ratio of emergency department visits to total outpatient visits was significantly higher for the 2 control groups than for the CCHP group, and for asthmatic children with PI than for asthmatic children with CCHP insurance. These differences could reflect higher severity of disease in the control group populations or, more likely, the effect of gatekeeping procedures that were implemented by the CCHP for emergency department use. Despite a rate of chronic illness that was comparable to the MK population, hospitalization rates were not significantly different in the 2 groups, with a trend toward lower rates in the CCHP group. PI children did have a higher mean number of hospitalizations compared with CCHP children, despite a rate of chronic illness that was slightly lower. These findings are in agreement with other recent data^37,38 suggesting that, overall, previously uninsured low-income children do not use health services at a higher rate than do children in other insurance groups. Our data also suggest they do not use services in a medically inappropriate manner.

For those children who had been seen in the practice before enrollment into CCHP, enrollment was associated with a higher rate of health maintenance visits, suggesting this group may have deferred preventive services until they had insurance coverage. This pent-up demand for preventive services was higher in the subset of CCHP enrollees who had not been seen in the practice before enrollment into the plan. Because our data collection was limited to the study practices, we could not assess preprogramatic utilization by CCHP-enrolled children who were never seen in the practice before enrollment. The shifts in immunization location after enrollment in CCHP-enrolled children is also an indication that the program may have increased continuity of care by not uncoupling receipt of immunizations from other preventive care, a concept that has been stressed in other recent studies.^49,50

There are important limitations to the health care utilization data reported in this study. The duration of enrollment was available for all CCHP-enrolled children, but the limited records of the practices studied did not permit us to determine the duration of enrollment of children in MK and PI or gaps in coverage for children in these groups. During a 1-year period, roughly half of children enrolled in MK in Colorado will be discontinuously enrolled.¹⁸ In addition, over all insurance categories, one-quarter of children have gaps in insurance coverage during their first 3 years of life, over half of these for 6 months or longer.¹⁷ Therefore, the insurance status of children with MK or PI is intrinsically in flux. In our study, children in the comparison groups may have shifted coverage during the study and experienced some gaps in coverage that could not be measured. However, the fact that the mean length of time in the practice did not differ among the groups suggests any systematic biases in utilization data are unlikely. Our assessments of quality of care were limited by reliance on documentation of care in the medical record. There were no systematic differences in the way utilization or content of care information was recorded for any insurance group and, therefore, data limitations should have been uniform for all insurance groups.

Interpretation of our findings regarding immunization up-to-date rates is complex, as it relies on services delivered at the primary care site and on the completeness of records from other locations. The low rate of documented up-to-date immunization may reflect poor documentation rather than true underimmunization in all groups. Differences in the completeness of immunization documentation in the 3 groups may have existed and could bias intergroup comparisons. Newly enrolling CCHP families may have been more conscientious about bringing in immunization records at the time of enrollment. Alternatively, the lower utilization of public health immunization sites by children with PI might result in a bias toward not reporting immunizations in the CCHP or MK groups. In addition, this study did not evaluate other important components of health care quality, including satisfaction of clients and clinicians, self-perceived access to care, or costs. Finally, as recently suggested,⁵¹ the effect of implementation of a low-income insurance program will vary substantially based on characteristics of the state in which it is introduced, most notably the extent of the functioning safety net for uninsured children. Therefore, there are limitations to the generalizability of our results.

Our study has several policy implications. Implementation of managed care practices as part of a SCHIP program appeared to be effective in our private practice settings. In our study, capitation was not a deterrent to receipt of primary care. Further, the primary care CCHP-enrolled children received was comparable in quality to that provided by MK and PI plans in the area. Gatekeeping was associated with more acute visits at the primary care site rather than the emergency department in new CHIP enrollees. Our data suggest that implementation of SCHIP in an area in which uninsured children are receiving care in private practices without a large safety net results in shifts in the payor status of patients; the effectiveness of the program in bringing uninsured children into care is less clear-cut. Our results suggest that enrollees in the first few years of CCHP may have been a motivated portion of the uninsured population who had managed to receive care by some other means before enrollment. Nonetheless, enrollment increased their use of their medical home for both preventive and acute services. This study adds to the growing evidence that the legislative intent of the 1997 Balanced Budget Act through the SCHIPs can be converted into viable programs to provide accessible, high quality care to previously uninsured children. Additional research should assess strategies to best reach that portion of the uninsured population in greatest need, the optimal structuring of SCHIP programs in different practice settings, and determining sufficient reimbursement levels to encourage physician participation.

	ACKNOWLEDGMENTS

This project was supported by a grant from the Blue Cross/Blue Shield of Colorado Foundation.

We wish to thank the participating practices for allowing us to review medical records at their clinical facilities and helping with data collection. We also thank Stephen Berman, MD, for his thoughtful review of this manuscript and Barbara Stucky for preparing the manuscript.

	FOOTNOTES

Received for publication Jan 25, 1999; accepted Aug 2, 1999.

Address correspondence to Allison Kempe, MD, MPH, Children's Hospital, 1056 E 19th Ave, B032, Denver, CO 80218. E-mail: kempe.allison{at}tchden.org

	ABBREVIATIONS

MK, Medicaid; SCHIP, State Children's Health Insurance Program; CCHP, Colorado Child Health Plan; PI, private insurance/privately insured; CASA, Clinical Assessment Software Application.

	REFERENCES

Top Abstract Methods Results Discussion References

1. Gauthier AK, Schrodel SP. Expanding children's coverage: lessons from state initiatives in health care reform. In: State Initiatives in Health Care Reform. Princeton, NJ: The Robert Wood Johnson Foundation; 1997
2. Thorpe KE Incremental strategies for providing health insurance for the uninsured. JAMA. 1997; 278:329-333 [Abstract]
3. American Academy of Pediatrics, Committee on Child Health Financing Implementation principles and strategies for Title XXI (State Children's Health Insurance Program). Pediatrics. 1998; 101:944-948 [Abstract/Free Full Text]
4. Colorado Census Data, 1990
5. US Public Health Service. International Classification of Diseases, Ninth Revision, Clinical Modification. 2nd ed. Washington, DC: US Public Health Service; 1980:1. US Dept of Health and Human Services Publication No. PHS 801260
6. Centers for Disease Control, National Immunization Program. Clinic/Provider Assessment Software Application for Windows (Win CASA) Version 1.0. Initial release November 1, 1997
7. Green M, ed. Bright Futures: Guidelines for Health Supervision of Infants, Children, and Adolescents. Arlington, VA: National Center for Education in Maternal and Child Health; 1994
8. Perrin JM, Kuhlthau K, Walker DK, Monitoring health care for children with chronic conditions in a managed care environment. Matern Child Health J. 1997; 1:15-23 [CrossRef][Medline]
9. Gortmaker S, Sappenfield W Chronic childhood disorders: prevalence and impact. Pediatr Clin North Am. 1984; 31:3-18 [Medline]
10. Stein RE, Bauman LJ, Westbrook LE, Coupey SM, Ireys HT Framework for identifying children who have chronic conditions: the case for a new definition. J Pediatr. 1993; 122:342-347 [Medline]
11. Newacheck PW, Stoddard JJ, McManus M Ethnocultural variations in the prevalence and impact of childhood chronic conditions. Pediatrics. 1993; 91:1031-1039 [Abstract/Free Full Text]
12. Newacheck PW, Taylor WR Childhood chronic illness: prevalence, severity, and impact. Am J Public Health. 1992; 82:364-371 [Abstract/Free Full Text]
13. Gay JC, Muldoon JH, Neff JM, Wing LJ Profiling the health service needs of populations: description and uses of the NACHRI classification of congenital and chronic health conditions. Pediatr Ann. 1997; 26:655-663 [Medline]
14. Perrin EC, Newacheck P, Pless B, Issues involved in the definition and classification of chronic health conditions. Pediatrics. 1993; 91:787-793 [Abstract/Free Full Text]
15. Berman S Current concepts: otitis media in children. N Engl J Med. 1995; 332:1560-1565 [Free Full Text]
16. National Asthma Education and Prevention Program. Expert Panel Report: Guidelines for the Diagnosis and Management of Asthma. Bethesda, MD: National Heart, Lung, and Blood Institute; 1991. National Institutes of Health Publication No. 91-3642
17. Kogan MD, Alexander GR, Teitelbaum MA, The effect of gaps in health insurance on continuity of a regular source of care among preschool-aged children in the United States. JAMA. 1995; 274:1429-1435 [Abstract]
18. Berman S, Byrns P, Roark R, Bondy J The effect of discontinuous vs continuous Medicaid enrollment on the availability of primary care services. Arch Pediatr Adolesc Med. 1995; 149:86 [Abstract]
19. Berman S Uninsured children: an unintended consequence of health care system reform efforts. JAMA. 1995; 274:1472-1473 [CrossRef][Medline]
20. Rosenbach ML The impact of Medicaid on physician use by low-income children. Am J Public Health. 1989; 79:1220-1226 [Abstract/Free Full Text]
21. Davis K, Gold M, Markus D Access to health care for the poor: does the gap remain? Annu Rev Public Health. 1981; 2:159-187 [CrossRef][Medline]
22. Wilenski GR, Berk ML. The health care of the poor and the role of Medicaid. Health Aff (Millwood). 1982;Fall:94-100
23. Access to Health Care in the United States: Results of a 1986 Survey. Princeton, NJ: The Robert Wood Johnson Foundation; 1987. Special Report No. 2
24. Weissman JS, Stern R, Fielding SL, Epstein AM Delayed access to health care: risk factors, reasons, and consequences. Ann Intern Med. 1991; 114:325-331
25. Moscovice I, Davidson G Health care and insurance loss of working AFDC recipients. Med Care. 1987; 25:413-425 [CrossRef][Medline]
26. Cunningham PJ, Monheit AC Insuring the children: a decade of change. Health Aff (Millwood). 1990; 9:76-90 [CrossRef][Medline]
27. Spillman BC The impact of being uninsured on utilization of basic health care services. Inquiry. 1992; 29:457-466 [Medline]
28. Monheit AC, Cunningham PJ Children without health insurance. Future Child. 1992; 2:154-170 [CrossRef]
29. Cornelius LJ Barriers to medical care for white, black, and Hispanic children. J Natl Med Assoc. 1993; 85:281-288 [Medline]
30. Reichel W The continuity imperative. JAMA. 1981; 246:2065 [CrossRef][Medline]
31. Starfield B. Primary Care: Concept, Evaluation, and Policy. New York, NY: Oxford University Press; 1992
32. Holl JL, Szilagyi PG, Rodewald LE, Profile of uninsured children in the United States. Arch Pediatr Adolesc Med. 1995; 149:398-406 [Abstract]
33. Orr S, Charney E, Straus J, Bloom B Emergency room use by low income children with a regular source of medical care. Med Care. 1991; 29:283-286 [CrossRef][Medline]
34. Butler J, Winter W, Singer J, Wenger M Medical care use and expenditure among children and youth in the United States: analysis of a national probability sample. Pediatrics. 1985; 76:495-507 [Abstract/Free Full Text]
35. Alpert J, Robertson L, Kosa J, Haegerty M, Haggerty R Delivery of health care for children: report of an experiment. Pediatrics. 1976; 57:917-930 [Abstract/Free Full Text]
36. McDaniel D, Patton E, Mather J Immunization activities of private practice physicians: a record audit. Pediatrics. 1975; 56:504-507 [Abstract/Free Full Text]
37. Martin DP, Diehr P, Cheadle A, et al. Health care utilization for the "newly insured": results from the Washington Basic Health Plan. Inquiry. 1997;34(Summer):129-142
38. Lave JR, Keane CR, Lin CJ, Impact of a children's health insurance program on newly enrolled children. JAMA. 1998; 279:1820-1825 [Abstract/Free Full Text]
39. Rodewald LE, Szilagyi PG, Holl J, Health insurance for low-income working families: effect on the provision of immunizations to preschool-age children. Arch Pediatr Adolesc Med. 1997; 151:798-803 [Abstract]
40. Ball TM. Immunizations: toddler documentation gap. Clin Pediatr. 1995;Jan:54-56
41. Zell ER, Dietz V, Stevenson J, Cochi S, Bruce RH Low vaccination levels of US preschool and school-age children. Retrospective assessments of vaccination coverage, 1991-1992. JAMA. 1994; 271:833-839 [Abstract]
42. Cutts FT, Zell ER, Mason D, Monitoring progress toward US preschool immunization goals. JAMA. 1992; 267:1952-1955 [Abstract]
43. Griffin MR, Daugherty J, Reed GW, Immunization coverage among infants enrolled in the Tennessee Medicaid program. Arch Pediatr Adolesc Med. 1995; 149:559-564 [Abstract]
44. Bobo JK, Gale JL, Thapa PB, Wassilak SGF Risk factors for delayed immunization in a random sample of 1163 children from Oregon and Washington. Pediatrics. 1993; 91:308-314 [Abstract/Free Full Text]
45. Bordley WC, Margolis PA, Lannon CM The delivery of immunizations and other preventive services in private practices. Pediatrics. 1996; 97:467-473 [Abstract/Free Full Text]
46. Aday LA, Anderson RM The national profile of access to medical care: where do we stand? Am J Public Health. 1984; 74:1331-1346 [Abstract/Free Full Text]
47. Wood DL, Hayward RA, Corey CR, Access to medical care for children and adolescents in the United States. Pediatrics. 1990; 86:666-673 [Abstract/Free Full Text]
48. Bograd H, Ritzwoller DP, Calonge N, Extending health maintenance organization insurance to the uninsured. A controlled measure of health care utilization. JAMA. 1997; 277:1067-1072 [Abstract]
49. Rodewald LE, Szilagyi PG, Shiuh T, Is underimmunization a marker for insufficient utilization of preventive and primary care? Arch Pediatr Adolesc Med. 1995; 149:393-397 [Abstract]
50. American Academy of Pediatrics, Ad Hoc Task Force on Definition of the Medical Home The medical home. Pediatrics 1992; 90:774 [Abstract/Free Full Text]
51. Long SH, Marquis MS Geographic variation in physician visits for uninsured children. JAMA. 1999; 281:2035-2040 [Abstract/Free Full Text]