Reported Experiences with Medicaid Managed Care Models Among Parents of Children

Matern Child Health J (2014) 18:544 553 DOI 10.1007/s10995-013-1270-5 Reported Experiences with Medicaid Managed Care Models Among Parents of Children Allyson G. Hall Amy Yarbrough Landry Christy Harris Lemak Erin L. Boyle R. Paul Duncan Published online: 19 April 2013 Ó Springer Science+Business Media New York 2013 Abstract The proportion of children enrolled in Medicaid managed care arrangements has grown significantly over the past decade. Yet, few studies have attempted to assess differences in parental reports and ratings of care for children enrolled in different types of Medicaid managed care. We examine parental reports and ratings of care to explore whether and how patient and parent experiences vary by child health status and managed care plan type, including provider-sponsored specialized plans serving only children. Parents of children in a Florida Medicaid demonstration project in two counties over 3 years were A. G. Hall (&) Department of Health Services Research, Management and Policy, College of Public Health and Health Professions, PO Box 100195, Gainesville, FL 32610, USA e-mail: hallag@phhp.ufl.edu A. Y. Landry Department of Health Services Administration, School of Health Professions, University of Alabama at Birmingham, 1675 University Blvd., Webb 518, Birmingham, AL, USA e-mail: akyarb@uab.edu C. H. Lemak Department of Health Management and Policy, School of Public Health, University of Michigan, 109 S. Observatory, 3611 SPH II, Ann Arbor, MI 48109, USA e-mail: chrislem@umich.edu E. L. Boyle Health Systems Administration, Georgetown University, Washington, DC, USA e-mail: elb74@georgetown.edu R. P. Duncan Department of Health Services Research Management and Policy, University of Florida, PO Box 100195, Gainesville, FL 32610, USA e-mail: pduncan@phhp.ufl.edu surveyed using Consumer Assessment of Health Providers and Systems surveys (n = 2,741 11,067). Ordered logistic regression models with interaction terms were used to assess relationships between plan type, presence of chronic condition, and measures of patient experience. Parents of children enrolled in provider-sponsored plans that focus on pediatrics were more likely to provide a positive rating for their doctor, health plan, and specialty care compared to parents of children in an health maintenance organization (HMO). Parents of children with a chronic condition were less likely than parents of children without a chronic condition to provide a favorable rating of overall health care, their doctor, or health plan. The interaction term that assessed whether patient experience by plan type was impacted by the child s health status was not statistically significant. Parents of Medicaid children may prefer provider-sponsored arrangements over HMOs. Findings can inform the future development of other integrated models of care involving provider-sponsored arrangements, such as pediatric Accountable Care Organizations and Patient- Centered Medical Homes. Keywords Medicaid Managed care Pediatrics Satisfaction Experiences Introduction Although fee-for-service forms of Medicaid still exist, Medicaid managed care continues to grow [1]. States utilize a variety of managed care arrangements. However, few studies explore the extent to which patients and their families in various managed care plan types experience care differently. Providing patient-centered care is one of the six quality improvement aims set out by the Institute of

Matern Child Health J (2014) 18:544 553 545 Medicine in their 2001 report, Crossing the Quality Chasm [2]. Identifying differences in experiences with care is important because satisfaction with health care is paramount to the continual improvement of health care quality and is a measure of patient centered care. Children might be particularly sensitive to differences between various forms of managed care due to the importance of the physician-patient relationship and need for consistent well-child and prevention services. Children with chronic conditions could be even more susceptible to potential differences as parents must navigate multiple providers and services. For example, parents of children with special health care needs report lower levels of satisfaction with health care. Among these parents, parents of black and Hispanic children are less likely to be satisfied with their experiences [3]. In most states, Medicaid enrollees typically belong to either (1) traditional, fee-for-service Medicaid in which the state contracted directly with providers for the provision of services, or (2) Medicaid health maintenance organizations (HMOs), with capitated arrangements between the state and the managed care plan. Florida s recent Medicaid demonstration project recently migrated all enrollees (excluding the dual eligible population) into managed care plans, and eliminated access to traditional Medicaid in several demonstration counties. Managed care plans offered to enrollees in the demonstration counties include Medicaid HMOs and new forms of provider-sponsored managed care called provider service networks (PSNs). These organizations are not unlike the emerging Accountable Care Organizations (ACOs) where provider groups direct the care for a defined population and are potentially more sensitive to community needs. The PSN category also includes specialized pediatric only plans (POPs) serving children with special needs (children with chronic conditions as determined by a clinical screening process, children alleged to have been abused or neglected, and children who are developmentally behind their peers). The siblings of these children are also eligible for participation in the POPs [4]. A division of the Florida Department of Health operates the POP plans. The POPs operate similarly to other PSNs in that the participating PCPs agree to provide or arrange all necessary services to an assigned panel of patients. What is especially unique about the POPs is their focus on ensuring that children receive a coordinated system of care. The plans provide care coordinators who work with physicians, families, and social workers to schedule and follow-up on health care services [4]. Few studies have examined the impact of publically sponsored managed care on children, especially children with special health care needs. Fewer still have examined managed care s impact on satisfaction or health care experiences [5]. A study from Oregon of managed care enrollees found no differences in satisfaction, self-reported utilization, or perceptions of unmet needs between children with and without a disability [6]. Studies comparing the satisfaction ratings reported by parents of children in feefor-service health plans versus those in managed care plans have produced mixed results. In various studies, parents of children in a fee-for-service arrangement provided more negative ratings, similar ratings, or more positive ratings than parents of children in managed care arrangements [7 9]. A more recent study of the Florida demonstration also showed that following the shift into managed care, parental reports and ratings of satisfaction and care experiences did not change. In the one exception to that study, satisfaction with customer service improved with increased enrollment into managed care [10]. None of these studies, however, have examined whether various kinds of managed care arrangements may have a differential impact on reported experiences with care. Our study assesses differences in parental experiences among three different types of Medicaid managed care arrangements. We compare the experiences and satisfaction of parents with children enrolled in a traditional form of managed care (i.e., HMO) with those enrolled in PSNs and the specific plans for children with special health care needs (POPs) in two urban counties participating in the demonstration. Study Setting We focus on the experiences of children and families in Medicaid in two Florida counties participating in Florida s Medicaid demonstration project. By 2008, 23 managed care arrangements were offered to Medicaid enrollees in the two counties. Eight of the participating plans are PSNs, including two POPs. See Appendix for more details on the organizational characteristics of the plans including ownership, geographic reach, and mission. PSNs are managed care arrangements sponsored by a provider organization (e.g., safety net hospital) or group of organizations (e.g., physician network or hospital system). In this model, the state reimburses providers directly on a fee-for-service basis and the sponsor is charged with care coordination, utilization management and health improvement activities. Like the emerging ACOs, if specified expenditure goals are met, the PSNs receive a portion of the savings as additional revenue. Once enrolled in a PSN, beneficiaries must select a primary care provider or have one assigned to them. Many of the primary care physicians and clinics in the PSNs saw a large number of Medicaid patients prior to the demonstration. The PCPs can either be employees of the provider organization or have a contractual relationship with the

546 Matern Child Health J (2014) 18:544 553 PSN, however, they are paid by the state on a fee-forservice basis plus monthly case management fees. PCPs are required to provide or arrange for coverage of services, consultations, or approval for referrals 24 h a day, 7 days a week. While PSNs are eligible for shared savings, they have minimal downside financial risk. The PSNs sponsored by safety-net providers assume no risk, while physician sponsored networks have only limited risk for administrative fees [11]. HMOs participating in the demonstration project are reimbursed by the state on capitated arrangements and thus, assume full financial risk for services provided. In turn, the HMOs reimburse primary care physicians who have an assigned panel of patients. Many of the managed care organizations that operate HMOs for the commercial market utilize their existing provider networks to serve the Medicaid population. However, each plan (PSN or HMO) is required to meet state standards in terms of physician and provider network to be approved for participation in the demonstration. Participating HMOs reimburse providers in a variety of ways, including fee-for-service and capitation payments. All Medicaid enrollees in the participating demonstration counties, including children, have the opportunity to select their plan through a Choice Counseling process. The process allows enrollees to directly compare benefits offered by available plans with a trained staff member and make the selection that is best for the individual or family. In each study county, enrollees could choose from multiple HMOs and PSNs (including a single POP in each county). Eligible families who did not participate in Choice Counseling and thus declined to select a plan, were autoassigned through a random process to a participating plan. During the time of this study, the clear majority of eligible families voluntary selected a plan and most enrolled in HMOs. Over the study period, a slight increase in HMO enrollment was observed coupled with a simultaneous decrease in PSN enrollment. The only PSNs that did not experience declining enrollment during this time period were the POPs [12]. Methods For this study we hypothesize that differences in organizational structure, ownership, and mission will translate to differences in experiences and satisfaction with care across plans. More specifically, we suggest that since the PSNs are local provider-based organizations with a safety-net orientation, they may have higher ratings and more favorable reports of care compared to the HMOs. In addition, although parents of children in poorer health may provide lower and less favorable ratings of care, being in a POP may moderate or improve the effects. We analyzed Consumer Assessment of Health Providers and Systems (CAHPS) surveys that were fielded as a part of a larger evaluation of the Florida Medicaid demonstration. Although four rounds of the survey were conducted, this analysis is limited to the three rounds conducted after the demonstration was fully implemented. By this time, most beneficiaries were enrolled in one of the managed care arrangements (i.e., HMO, PSN, POP). The sampling frame included beneficiaries in the Florida Medicaid s Member-Month and Recipient Eligibility files. Beneficiaries who were enrolled in Medicaid in one of the counties participating in the demonstration for at least six consecutive months were randomly selected to participate in the survey. Probability sampling ensured that the survey sample was representative of the populations of each Medicaid managed care plan. Sampled beneficiaries were surveyed by telephone by trained interviewers from the University of Florida Survey Research Center. Parents completed the survey on behalf of their children. Response rates (which included beneficiaries who had a non-working phone number or who declined participation in the denominator) were around 40 % in each of the 3 years. Comparisons using eligibility file data, between respondents and non-respondents showed no statistically significant differences with regard to race/ethnicity, gender, Medicaid eligibility category, type of managed care plan, and county of residence. The actual cooperation rate (i.e., the proportion of respondents contacted who actually completed the survey) ranged from 64 to 76 % depending on the year. The analytic sample was limited to parents of children living in the two urban demonstration counties. Individuals who were dually-eligible for Medicare and Medicaid, who live with a developmental disability, or who are residents of nursing homes were also excluded from the sample. For the 3 years reported in this study, in the two urban counties the survey sample included a total of 18,276 respondents of which 11,853 were parents of children (ages 0 21). Excluded from the analytic sample were beneficiaries who changed health plans within 6 months of the survey, resulting in a sample of 11,067. During the time under study, several health plans left the market and these beneficiaries were moved to other health plans. Outcome Variables CAHPS represents a family of surveys designed by the Agency for Health Care Quality and Research (AHRQ) that ask consumers and patients to assess aspects of their care [13, 14]. To assess consumer experience with care, the CAHPS asks two kinds of questions: ratings of care and reports of care. Ratings of care ask patients to rank their care on a 0 10 scale along a number of dimensions,

Matern Child Health J (2014) 18:544 553 547 including overall rating of health care, overall rating of personal doctor, overall rating of specialty care, and overall health plan rating. For this analysis, ratings were divided into three groups representing a high score (9 10), a medium score (7 80), and a low score (6 and below), as reported in the CAHPS Health Plan Survey Chartbook [15]. We decided to categorize the variable because of the highly skewed nature of the distribution. Reports of care ask about specific health care experiences within the last 6 months, such as finding a personal doctor and getting care in a timely manner. Reports of care use a Likert scale response [13, 14]. Analytic sample (parents of children ages 0 21 in the two urban counties) sizes vary from 2,741 to 11,067 depending on the question asked. Variations are due to skip patterns such that respondents could only provide reports or ratings on aspects of care that they had experienced. Between 63 and 70 % of the population was in an HMO, between 26 and 27 % was in a PSN, while between 4 and 10 % was in a POP. Key Independent Variables Children were categorized into two groups based on parental response to the CAHPS Item Set for Children with Chronic Conditions (ISCCCs), which was added to the surveys of parents of children. The ISCCC includes five items which are used to classify children with and without one or more chronic conditions. The ISCCC asks about use or need of prescription medication; above average use or need for medical, mental health, or educational services; functional limitations compared with others of the same age; use of specialized therapies; and treatment or counseling for emotional, behavioral, or developmental problems [16]. Health plan type was derived from the Member-Month and Recipient Eligibility Files from the state Medicaid agency. Plans were coded as provider-sponsored networks (PSNs), the pediatric-only-plan (POP), or HMO. Control Variables Other variables thought to impact parental reports of care were also included in the analyses. These variables came from the survey data and included county (Broward or Duval), ethnicity of the child (Hispanic, not-hispanic), race of the child (white, black, other), educational attainment of the parent (no high-school diploma, high-school graduate, more than high-school), and language spoken at home (English, language other than English). Since we thought plan performance and parental ratings and reports of care might improve over time, we also included calendar year of the survey (2007, 2008, 2009 2010). Analytic Approach Both ratings of care (high, medium, low) and reports of care (never, sometimes, usually, always or big problem, small problem, not a problem) are considered ordered as the distances between the adjacent categories are unknown. Consequently, ordered logistic regression was used to assess the relationship between an outcome variable, plan type, and presence of chronic condition [17]. To specifically address whether ratings or reports by health plan type are moderated by the health status of the child, an interaction term of the two variables was also included in the models. Regression models controlled for race of the child, Hispanic ethnicity of the child, educational attainment of the parent, county, language spoken at home, and year of survey. We also calculated and report the marginal effect for each odds ratio. Marginal effects are the difference in the probability of an outcome occurring between a given category and the reference group. Analyses were performed using STATA version 10. Data were weighted to reflect plan share in each county and survey non-response. This study was deemed exempt from review by the University of Florida Institutional Review Board. Results Table 1 compares the characteristics of the sample across the three plan types. With the exception of ethnicity, characteristics of respondents are significantly different across plan types. As would be expected, a greater proportion of children in a POP arrangement have a chronic condition compared to children in an HMO or PSN. Over half of PSN enrollees are in Duval County, while over onethird of HMO and POP enrollees live in that county. Over half of parents indicated that their child was black, with slightly higher proportions in an HMO and PSN. A greater proportion of parents of children in a POP had more than a high school diploma compared to parents of children in an HMO or PSN. Over 80 % of families spoke English at home. Overall, the highest proportion of beneficiaries was surveyed during the first year of the demonstration followed by the 3rd year and 2nd year. Tables 2 and 3 show the percent distribution for the ratings and reports of care. Over 50 % of the population provided either a 9 or 10 for their ratings of aspects of their care. However, the proportion of individuals providing their health plan with the highest rating was about 57 % compared to 75 % of respondents who rated their doctor a 9 or 10. Reports of care tended to be favorable with the exception of getting an appointment with a specialist. Only about 50 % of respondents thought getting an appointment

548 Matern Child Health J (2014) 18:544 553 Table 1 Characteristics of Florida Medicaid beneficiaries. Percent distribution of analytic sample POP HMO PSN Total p value Weighted N 10,043 176,409 69,818 256,272 Unweighted n 1,396 6,542 3,129 11,067 Have a chronic condition Yes 54.96 7.423 9.99 9.914 0.0000 No 45.04 92.58 90.01 90.09 County of residence Broward 60.61 61.94 42.74 56.65 0.0000 Duval 39.39 38.06 57.26 43.35 Ethnicity Not Hispanic 73.29 75.82 77.48 76.18 0.1584 Hispanic 24.04 21.76 20.36 21.47 Unknown 2.67 2.41 2.16 2.35 Race White 30.43 24.33 26.14 25.05 0.0070 Black 50.26 56.59 56.17 56.24 Unknown/other 19.31 19.08 17.69 18.71 Education of respondent parent No high school 20.72 22.93 23.61 23.03 0.0135 diploma High school 34.8 39.13 40.47 39.33 graduate More than High School 44.48 37.94 35.91 37.63 Language spoken at home English 81.31 85.74 86.51 85.78 0.0079 Another language 18.69 14.26 13.49 14.22 Calendar year surveyed 1st year of 29.08 40.37 34.4 38.32 0.0000 demonstration 2nd year of 29.09 29.23 27.33 28.71 demonstration 3rd year of demonstration 41.82 30.39 38.28 32.97 higher doctor or plan rating compared to parents of HMO enrollees. For example, parents of POP enrollees who were 15 % more likely (OR 1.93, 95 % 1.63, 2.38) and 9.1 % more likely (OR 1.52, 95 % CI 1.22, 1.92) to rate their health plan and specialty care respectively, a 9 or 10. Interaction effects between chronic condition and plan type were not statistically significant. However, chronic condition by itself was found to be a significant predictor for three of the four ratings. Parents of children with a chronic condition were less likely than parents of children without a chronic condition to provide a more favorable rating of overall health care, their doctor, or health plan. For example, parents of children with a chronic were 12 % less (OR 0.61, 95 % CI 0.53, 0.70) likely to give their health plan a 9 or 10 compared to parents of children without a chronic condition. However, differences in specialty care ratings between the two groups were not observed. Table 5 presents the odds of providing a favorable assessment of a report of care by plan type and chronic condition. Parents of children who were in a PSN were more likely to provide a positive report of finding a personal doctor, getting the help or advice needed, and filling a prescription through a health plan relative to the parents of children in an HMO. For example, parents of the POP enrollees were 9.5 % (OR 1.46, 95 % CI 1.18, 1.81) more likely to report never having problem getting an appointment with a specialist compared to parents of children in HMOs. In addition to finding a personal doctor, parents of children in POP were also more likely to provide favorable reports associated with securing an appointment with a specialist, obtaining the care and tests needed, being able to fill a prescription, and arranging an appointment as soon as needed. Again, as with the ratings of care, interactions between chronic condition and plan type were not significant. However, with the exception of getting an appointment as soon as wanted and getting care right away, parents of children with a chronic condition were less likely to provide favorable reports of care. with a specialist was always easy. In contrast, about 80 % reported that they often always got care right away. Table 4 provides the adjusted odds with the marginal effects of providing a more favorable rating of care by plan type. The relationship between plan type and ratings of care is shown to be statistically significant for doctor rating, plan rating, and specialty care rating (for POP enrollees). Odds ratios show that parents of children enrolled in a POP were more likely to provide a positive rating for their doctor, health plan, and specialty care compared to parents whose children were in an HMO. Parents of PSN enrollees (including children in a POP) were more likely to provide a Discussion We compared the experience of care across different types of managed care arrangements for children in Medicaid with the underlying premise that differences in organization type could result in variation in reports and experiences of care for parents of children in Medicaid. Findings provide some evidence and support for our hypothesis that the nature of the PSNs might result in better experiences and greater satisfaction with care. The higher doctor ratings and plan ratings among PSN and POP parents relative to HMO parents may partially be

Matern Child Health J (2014) 18:544 553 549 Table 2 Weighted percent distributions: ratings of care Overall health care a Doctor rating b Plan rating c Specialty care rating d Unweighted n 9,015 10,347 11,067 3,392 Weighted N 206,299 240,117 256,273 65,844 Rating Low rating 9.8 7.6 15.9 12.6 Medium rating 26.6 17.6 27.3 23.0 High rating 63.6 74.8 56.8 64.4 a Among those with a visit to the doctor b Among those with a regular provider c Among all respondents d Among those with a visit to the specialist Table 3 Weighted percent distributions: reports of care How much of a problem finding a personal doctor a How often got help or advice needed when called during regular office hours b How easy getting an appointment with a specialist c How often was it easy to get care and tests needed d How often was it easy to get prescriptions needed through a health plan e How often got appointment as soon as wanted f How often got care right away g Unweighted n 10,355 4,482 2,741 4,191 5,710 8,397 4,025 Weighted N 240,694 96,189 53,064 90,064 122,024 189,959 87,202 Report Never 3.9 (big 5.10 13.7 7.8 2.9 2.7 2.5 problem) Sometimes 7.5 (small 11.5 19.5 17.5 13.1 14.4 10.0 problem) Usually 88.6 (no 12.3 17.3 18.7 14.7 11.7 8.1 problem) Always 71.1 49.5 56.0 69.3 71.2 79.4 a Among those with a personal doctor. Scale for this variable is big problem, small problem, not a problem b Among those who needed care right away c Among those who needed a specialist d Among those who needed care or tests, or treatment e Among those who needed a prescription f Among those needing care but not right away g Among those who needed care right away due to the composition of the provider network. Many of the HMOs utilized their commercial provider networks, rather than networks developed specifically to address the needs of the Medicaid population. Access issues for Medicaid enrollees may have resulted from the utilization of these networks. Two of the PSNs were founded by networks comprised of mostly Hispanic or African American physicians. The organizations placed significant emphasis on working with primary care providers to ensure patient-centered care [18, 19]. The other PSNs were formed by safety-net hospitals with significant experience with Medicaid populations. The POP plans specifically target children with chronic conditions and through its care coordination activities likely pays greater attention to meeting both the general and specific health care needs of these children. Consequently, the POP plan does a better job than both the regular PSN and the HMO in areas such as specialty care access, getting care and tests needed, and getting prescriptions. While not surprising, of concern is the finding that parents of children with a chronic condition have consistently lower ratings and less favorable reports of care. We expected that since the POPs were designed specifically for sick children, being in this arrangement would result in

550 Matern Child Health J (2014) 18:544 553 Table 4 Odds of providing the highest rating (9 10) among parents of children enrolled in Florida s Medicaid reform plans (95 % CI) Overall health care a Doctor rating b Plan rating c Specialty care rating d OR ME e (%) OR ME (%) OR ME (%) OR ME (%) Unweighted n 9,015 10,347 11,067 3,392 Weighted N 206,299 240,117 256,273 65,844 Plan type PSN (HMO) 1.09 (0.97, 1.22)?2.1 1.21 (1.07, 1.37)** POP (HMO) 1.18 (0.99, 1.40)?3.8 1.41 (1.15, 1.74)** Chronic condition present No (yes) 0.70 (0.68, 0.93)*** -8.2 0.67 (0.57, 0.99)*?3.5 1.27 (1.15, 1.40)***?6.0 1.93 (1.63, 2.38)*** -7.8 0.61 (0.53, 0.70)***?5.8 0.99 (0.81, 1.22)?0.01?15.0 1.53 (1.22, 1.92)**?9.1-12.2 0.88 (0.70, 1.08) -3.03 Adjusted for: Hispanic ethnicity, race, education, language parent speaks at home, educational attainment of parent, year of survey, and county of residence * p B.05; ** p B.01; *** p B.001 a Among those with a visit to the doctor b Among those with a regular provider c Among all respondents d Among those with a visit to the specialist e Marginal effects represent the difference in the probability of an outcome occurring between a given category and the reference group more favorable patient experiences for these children. However, the interaction term that assessed this relationship was not statistically significant in any of the regression models. Therefore, contrary to our expectation, there was no direct evidence to support the assertion that being in the POP may actually improve ratings and reports of care among parents of children with chronic conditions or in poorer health. Although reports and ratings among POP parents are more favorable than HMO ratings for all children, being in a POP plan with its additional attention to the unique needs of special needs children does not differentially improve the health care experiences of sicker children. In order to make appropriate changes to the delivery system, future research needs to understand these relationships. In particular, as more provider-driven financial and organizational models (e.g. ACOs) and delivery system approaches (e.g. Patient-Centered Medical Homes) are developed to serve the unique needs of children, including individuals with special needs, their experiences must be explored in greater detail. While a managed care plan focused on chronically ill children is fairly unusual, we think that this form of managed care arrangement might be a viable option for future ACOs developed around chronically ill children. A provider driven approach to care coordination makes sense moving forward in the new environment of health reform. Our findings call for more in-depth examination of the impact of specific programs and mechanisms used by health plans and ACO-like structures on the children s health and experiences of care. This study has some limitations. For example, the study is not longitudinal and we are therefore unable to determine if parents reports or ratings of care changed over time. The study is also limited to two counties in Florida. These counties, however, collectively represent 256,272 pediatric Medicaid beneficiaries. Other areas of the state had not yet begun to experience managed care in the same way. Finally, the measure of chronic care is based on parental self-report and not based on a clinical assessment. This limitation is mitigated somewhat by the fact that the measure has been tested and shown to be effective in identifying children with special needs [20, 21]. These limitations affect our ability to definitively conclude that the organizational characteristics of the plans are the cause of differences in experiences. However, we do think our findings provide good clues about the relationship between how a plan is organized and patient experience. Findings suggest that a managed care arrangement similar to a PSN may be an appropriate organizational form for Medicaid beneficiaries. Our analysis indicates that the parents of these children may prefer a PSN arrangement to that of an HMO. Further research is needed to explore exactly how the financial and structural arrangements be they HMO, PSN, POP, or others models such as ACOs or impact the costs and quality of care for children.

Matern Child Health J (2014) 18:544 553 551 Table 5 Odds of providing the most favorable report of care among parents of children enrolled in Florida s Medicaid reform plans (95 % CI) Not a problem finding a personal Always got help or advice needed when called Never a problem getting an appointment Always easy to get care and tests doctor a during regular office hours b with a specialist c needed d OR ME e (%) OR ME OR ME (%) OR ME (%) Unweighted n 10,355 4,482 2,741 4,191 Weighted N 240,694 96,189 53,064 90,064 Plan type PSN (HMO) 1.25 (1.05, 1.48)*?2.1 1.24 (1.04, 1.49)*?4.4 0.99 (0.79, 1.21)?0.3 1.14 (0.97, 1.36)?3.4 POP (HMO) 1.30 (1.04, 1.64)*?2.4 1.24 (0.94, 1.65)?4.2 1.46 (1.18, 1.81)***?9.5 1.26 (1.05, 1.53)**?5.7 Chronic condition No (yes) 0.50 (0.42, 0.61)* -8.0 0.77 (0.63, 0.95)* -5.5 0.75 (0.610, 0.92)*** -7.0 0.61 (0.51, 0.74)*** -15.4 Always easy to get prescriptions needed through health plan f Always got appointment as soon as wanted g Always got care right away h OR ME e (%) OR ME (%) OR ME (%) Unweighted n 5,710 8,397 4,025 Weighted N 122,024 189,959 87,202 Plan type PSN (HMO) 1.50 (1.28, 1.76)***?8.3 1.00 (0.88, 1.13)?0.2 1.01 (0.82, 1.25)?0.2 POP (HMO) 1.60 (1.31, 1.96)***?9.0 1.24 (1.03, 1.49)**?4.2 1.26 (0.96, 1.66)?3.5 Chronic condition No (yes) 0.69 (0.58, 0.82)*** -8.1 0.97 (0.81, 1.16) -2.0 0.95 (0.75, 120) -0.8 Adjusted for: Hispanic ethnicity, race, education, language parent speaks at home, educational attainment of parent, year of survey, and county of residence * p B.05; ** p B.01; *** p B.001 Among those with a personal doctor a b Among those who needed care right away Among those who needed a specialist c d Among those who needed care or tests, or treatment Marginal effects represent the difference in the probability of an outcome occurring between a given category and the reference group Among those who needed a prescription Among those needing care but not right away Among those who needed care right away e f g h

552 Matern Child Health J (2014) 18:544 553 Appendix See Table 6. Table 6 Characteristics of Medicaid reform HMO and PSN plans Plan type Area served Ownership Mission Multi-state Florida only Publicly traded For profit (not public) Not-for-profit Diversified Medicaid only Govt. payers Broward county HMO 1 X X X HMO 2 X X X HMO 3 X X X HMO 4 X X X HMO 5 X X X HMO 6 X X X HMO 7 X X X HMO 8 X X X HMO 9 X X X HMO 10 X X X HMO 11 X X X PSN 1 X X X PSN 2 X X X PSN 3 X X X PSN 4 X X X POP 1 X X X Duval county HMO 12 X X X HMO 13 X X X HMO 14 X X X HMO 15 X X X PSN 5 X X X PSN 6 X X X POP 2 X X X References 1. Landon, B. E., Schneider, E. C., Normand, S. L., Scholle, S. H., Pawlson, L. G., Epstein, A. M., et al. (2007). Quality of care in Medicaid Managed care and commercial health plans. JAMA, 298(14), 1674 1681. 2. Institute of Medicine. (2001). Crossing the quality chasm: A new health system for the 21st century. Washington, D.C.: National Academy Press. 3. Ngui, E. M., & Flores, G. (2006). Satisfaction with care and ease of using health care services among parents of children with special health care needs: The roles of race/ethnicity, insurance, language, and adequacy of family-centered care. Pediatrics, 117(4), 1184 1196. 4. Children s Medical Services (CMS). www.cms-kids.com. Accessed September 19, 2012. 5. Huffman, L. C., Brat, G. A., Chamberlain, L. J., & Wise, P. H. (2010). Impact of managed care on publicly insured children with special health care needs. Academic Pediatric, 10(1), 48 55. 6. Mitchell, J. B., Khatutsky, G., & Swigonski, N. L. (2001). Impact of the Oregon Health Plan on children with special health care needs. Pediatrics, 107(4), 736 743. 7. Mitchell, J. M., & Gaskin, D. J. (2007). Caregivers ratings of access: Do children with special health care needs fare better under fee-for-service or partially capitated managed care? Medical Care, 45(2), 146 153. 8. Mitchell, J. M., & Gaskin, D. J. (2004). Do children receiving Supplemental Security Income who are enrolled in Medicaid fare better under a fee-for-service or comprehensive capitation model? Pediatrics, 114(1), 196 204. 9. Ganz, M. L., & Sisk, J. E. (2000). Evaluation of Medicaid managed care for children: Access and satisfaction. American Journal of Public Health, 90(12), 1947 1948. 10. Knapp, C., Madden, V., Sloyer, P., & Shenkman, E. (2012). Effects of an Integrated Care System on quality of care and satisfaction for children with special health care needs. Maternal and Child Health Journal, 16(3), 579 586.

Matern Child Health J (2014) 18:544 553 553 11. Agency for Health Care Administration. (2007). Attachment II. Medicaid reform fee-for-service, provider service network model contract. http://www.fdhc.state.fl.us/medicaid/medicaid_reform/ provider/draftcontracts.shtml. Accessed January 27, 2012. 12. Lemak, C. H., Landry, A. K., Bilello, L., Bell, L., & Van Wert, P. (2010). Medicaid reform organizational analyses: April 2009 March 2010: Department of Health Services Research Management and Policy at the University of Florida. 13. Agency for Health Care Research and Quality. (2010). Consumer assessment of healthcare providers and systems. Developing a CAHPS clinician and group survey to measure the medical home. Rockville, MD. https://www.cahps.ahrq.gov/. Accessed January 27, 2012. 14. Agency for Health Care Research and Quality. (2012). CAHPS survey tools to advance patient-centered care. Rockville, MD. https://www.cahps.ahrq.gov/. Accessed January 27, 2012. 15. Agency for Health Care Research and Quality. (2011). CAHPS health plan survey chartbook. Rockville, MD. https://www.cahps. ahrq.gov/cahpsidb/public/chartbookoverview.aspx. Accessed January 25, 2011. 16. Agency for Health Care Research and Quality. (2008). About the item set for children with chronic conditions. Rockville, MD: Agency for Health Care Quality and Research. https://www. cahps.ahrq.gov/. Accessed January 25, 2012. 17. Long, J. S. (1997). Regression models for categorical and limited dependent variables. London: Sage. 18. Johnson, C. E., Lemak, C. H., Hall, A. G., Harman, J. S., Zhang, J., Duncan, R. P., et al. (2010). Outsourcing administrative functions: Service organization demonstrations and Florida Medicaid PCCM program costs. J Health Care Finance, 37(1), 1 12. 19. Lemak, C. H., Hall, A. G., Johnson, C. E., Saxena, P., Aftuck, C., & Johnson, C. (2004). Evaluation of Florida s Minority Physician Network (MPN) Program: Florida Center for Medicaid and the Uninsured. 20. Bethell, C. D., Read, D., Stein, R. E., Blumberg, S. J., Wells, N., & Newacheck, P. W. (2002). Identifying children with special health care needs: Development and evaluation of a short screening instrument. Ambulatory Pediatrics, 2(1), 38 48. 21. Bethell, C. D., Read, D., Neff, J., Blumberg, S. J., Stein, R. E., Sharp, V., et al. (2002). Comparison of the children with special health care needs screener to the questionnaire for identifying children with chronic conditions Revised. Ambulatory Pediatrics, 2(1), 49 57.