Dental care: Access, Use and Cost of Services for Children in hawk-i. Report to the Iowa Department of Human Services

Transcription

1 Health Policy Dental care: Access, Use and Cost of Services for Children in hawk-i. Report to the Iowa Department of Human Services Peter C. Damiano University of Iowa Elizabeth T. Momany University of Iowa Stephen D. Flach University of Iowa Please see article for additional authors. Copyright 2005 the authors Hosted by Iowa Research Online. For more information please contact:

2 Dental care: Access, Use and Cost of Services for Children in hawk-i Report to the Iowa Department of Human Services Peter C. Damiano, DDS, MPH Professor and Director Elizabeth T. Momany, PhD Assistant Research Scientist Stephen D. Flach, MD, PhD Assistant Professor Knute Carter Student Research Assistant Michael P. Jones, PhD Professor Health Policy Research Program Public Policy Center The University of Iowa May 2005 The Iowa Department of Human Services and the Agency for Health Care Research and Quality supported this study. The results and views expressed are the independent products of university research and do not necessarily represent the views of the funding agencies. Any analysis, interpretation, or conclusion based on these data is solely that of the authors.

3

4 Preface This report presents the results of a study about the ability of children to receive dental services, the types of services they receive, how soon they first receive dental services after enrolling, and the costs of dental services in the Iowa hawk-i separate state child health insurance program (S-SCHIP) and the Iowa Medicaid program. It was conducted at the request of the Iowa Department of Human Services as part of their continuing quality assurance activities with health plans participating in hawk-i. Researchers at The University of Iowa Public Policy Center conducted this study with funding provided by the Iowa Department of Human Services and US Agency for Health Care Research and Quality (AHRQ). Information and conclusions presented in this report are the responsibility of the authors and do not represent the views of the Iowa Department of Human Services, the AHRQ, the health plans or the University of Iowa.

5

6 Acknowledgements The authors would like to thank Anita Smith, Shellie Goldman and Anna Ruggle of the Iowa Department of Human Services for their assistance with the completion of this research and for creating an environment conducive to conducting this evaluation. Cindy Brach, project officer with the Agency for Health Care Research and Quality was also helpful with her support and critiques of the report. Special thanks to our colleagues at the Public Policy Center. As always, Professor David Forkenbrock, the Center Director, made the valuable resources of the Center available to us. Jean Willard and Maggie Tyler, research assistants at the Center, Kathy Holeton, administrative assistant, Teresa Lopes, editor, and Peggy Waters, secretary, all provided valuable assistance in this research. Greg Wohlwend and David Svoboda, University of Iowa student research assistants, contributed excellent research support throughout the project.

7

8 Table of Contents Executive Summary... 1 Chapter 1. Overview of the study and dental care in the hawk-i program... 9 Introduction... 9 Background on the hawk-i program... 9 hawk-i and Medicaid dental plans The current hawk-i dental program Chapter 2. Access to dental services for children in hawk-i Introduction Methods Results Chapter 3. Dental utilization rates for children in hawk-i Background Methods Results Chapter 4. Time to first dental visit for children in hawk-i Methods Results Chapter 5. Costs of dental care for children in hawk-i Introduction Methods Results Chapter 6. Programmatic dental costs for the program Introduction Methods Results Chapter 7. Discussion and policy implications...47 Access to dental services...47 Utilization of dental services...47 Time to first dental visit...49 Cost of first episode of dental care...49 Per member/per month costs of dental care...50 Conclusions...50 Chapter 8. Research limitations...53 Appendix A...55 Appendix B...57 Appendix C...61

9

10 Dental care: Access, Use and Cost of Services for Children in hawk-i Executive Summary As part of the quality assurance activities for the hawk-i program for fiscal year 2005, the Iowa Department of Human Services and the hawk-i Clinical Advisory Committee requested that the University of Iowa Public Policy Center conduct a study concerning the use of dental services by children in the hawk-i program. Because each health plan in hawk-i uses a different approach to providing dental services, one of the primary questions for this study was whether the use of dental services varied based on the plan. For these analyses, Medicaid was included as an additional comparison group, since it is a long-established public insurance program with yet another approach to providing dental services. This study investigated the following research questions: 1. Did enrollment in the hawk-i program improve access to dental care for children based on responses to the pre/post evaluation survey? 2. What were the dental utilization rates for children in hawk-i based on the insurance encounter data and how were they affected by the method used to calculate the rates? 3. How soon did children receive dental services after enrolling in hawk-i and what factors were associated with receiving a dental visit sooner after enrolling? 4. What were the costs associated with providing dental services to children in hawk-i, and was there a difference in the costs for children who received care sooner (a measure of pent-up demand)? Also, what was spent on the different types of dental services (diagnosis and prevention vs. routine restorative care vs. extensive restorative care? 1) Access to dental care for children in hawk-i Results of a survey comparing children before and after joining hawk-i found significant improvement in several indicators of access to dental care. Although the percentage reporting need for dental care did not change, the proportion reporting unmet need (Figure ES-1) and delays for needed care declined significantly after a year in the program. The proportion with a regular source of dental care increased to between 80-90% of children and the percentage with any dental visit also improved after joining hawk-i. About three-quarters of children had a dental check-up, however, the percentage declined slightly over time. 1

11 Figure ES-1. Unmet need for dental care before and after enrolling in hawk-i by plan 2) Utilization of dental services Claims and encounter data were used to evaluate the percentage of children with a dental visit during FY 2001 and to explore how different calculation methods affected dental utilization rates. This is important because the utilization rates are greatly affected by partial year enrollment for many children in these programs. The four approaches investigated were: (1) Centers for Medicare and Medicaid Services (CMS) Form 416 Methodology, (2) Health plan Employer Data and Information Set (HEDIS) Methodology, (3) calculating rates as Full-Time Equivalent (FTEs), and (4) including only new enrollees in the program in the calculation. Figure ES-2 indicates that the rates achieved by the different calculation methods produce different results with different implications for comparing use in Medicaid and hawk-i. 100% 80% 60% 40% 20% 0% 34% 35% 45%55% 58% 45% CMS 416 HEDIS FTE All new enrollees Calculation methodology 50% 41% 31% 18% New month Medicaid SCHIP Figure ES-2. Dental utilization rates by program in CY 2001 Dental utilization rates were lowest, using all approaches, for children ages 1-3 and for adolescents over age 12. Almost all children with a dental visit received a diagnostic procedure (e.g., a dental check-up) and about 85% received a preventive procedure (e.g., cleaning or dental sealants). About a third received a restorative procedure such as a filling, and about one in five or six received a more complex procedure (e.g., root canal or crown). Children in Medicaid were slightly more likely to have received a complex procedure. 2

12 An evaluation of how the utilization rates were affected by the number of in which children were enrolled in the program indicated that rates are relatively consistent after children were in the program between 8 and 11. 3) Time to first dental visit The length of time it took children to receive a dental visit after first enrolling in hawk-i and Medicaid was evaluated using claims and encounter data over a three-year period (fiscal years ). About one in four children had a dental visit in the first 6 after enrolling in hawk-i or Medicaid, after adjusting for partial-year enrollments using survival analytic techniques. Rates were highest for children in John Deere and lowest for children in Iowa Health Solutions (IHS) (Figure ES-3). Figure ES-3. Adjusted dental utilization rates using survival analysis for children 6, 1 year and 3 years after first enrolling in each plan (FY ) 4) Costs of the first episode of dental care for newly enrolled children in hawk-i Costs of care for the first episode of dental care for children newly enrolled in hawk-i and Medicaid were evaluated using claims, encounter and enrollment files for FY The first episode of care was defined as the four-month period following the initial dental visit after joining the program. Fees from a 2003 American Dental Association survey of dentists in the West North Central region were used to standardize dentists charges across the plans. Among new enrollees with a dental visit during the three-year period, between one-third and one-half had a dental visit during the first three after enrolling in hawk-i or Medicaid (Table ES-1). Children newly enrolled in John Deere were most likely to have received a visit in the first three, while children in Medicaid and Iowa Health Solutions were least likely. 3

13 Table ES-1. Time to first dental visit after first enrolling, by plan IHS John Deere Wellmark Medicaid Time of first dental visit % Cum. % % Cum. % % Cum. % % Cum. % % 37% 50% 50% 46% 46% 33% 33% % 88% 45% 95% 47% 93% 55% 88% % 100% 5% 100% 7% 100% 13% 100% Children in Medicaid received the highest mean cost worth of dental services during the first episode of care, while the mean costs of the first episode of dental care for children in IHS and Wellmark were similar. Children in all plans who received care soonest (first three ) were most likely to receive more than $500 worth of services. Children in Medicaid were most likely to receive over $500 worth of services in the first episode of care, while children in John Deere were least likely (see Table ES-2). Table ES-2. Costs of all dental care received during the first episode of care by time of first dental visit by plan Medicaid IHS John Deere Wellmark Time of first dental visit after enrollment Time of first dental visit after enrollment Time of first dental visit after enrollment Time of first dental visit after enrollment Cost Categories All Dental Services Mean cost $433 $346 $356 $312 $274 $243 $236 $212 $193 $319 $259 $216 $ % 21% 20% 23% 25% 27% 26% 27% 27% 24% 30% 34% $ % 32% 32% 31% 37% 37% 38% 40% 51% 31% 32% 33% $ % 27% 28% 28% 24% 24% 26% 24% 13% 28% 24% 23% $ % 20% 21% 18% 15% 12% 10% 9% 9% 18% 13% 10% 4) Programmatic dental costs for the hawk-i program Administrative claims, encounter and enrollment files were used to evaluate the total costs of dental services provided by the different hawk-i plans and Medicaid, as well as the per-member per-month (pm/pm) costs. Costs were calculated using both the ADA survey fees (except for John Deere due to data issues) and the amount actually paid to the dentists by the plans. Table ES-3 shows the total amount spent on dental services during FY 2003 by plan. These costs are highly influenced by the number of enrollment during the year. 4

14 Table ES-3. Dental costs by service category for all children in hawk-i and Medicaid, FY 2003 Medicaid* IHS John Deere Wellmark Type of Service Cost basis All Services ADA Fees $14,399,950 $750,977 $686,435 $1,045,992 Paid Claims $8,776,731 $600,555 $638,066 $860,959 *Medicaid costs did not include $2 million spent on orthodontics (not covered by other plans) Amount equals dentist s charges ADA-related costs not available for John Deere To standardize the costs, given differences in the number of children enrolled and the length of time for which they were enrolled, the costs were calculated pm/pm, as shown in Table ES-4. The three hawk-i plans had higher costs worth of services per enrolled member than Medicaid. The costs of dental services paid by the plans pm/pm were highest for children in John Deere and lowest for children in Medicaid. Table ES-4. Dental costs by service category for all children in hawk-i and Medicaid, FY 2003 Medicaid* IHS John Deere Wellmark Type of Service Cost basis All Services ADA Fees $9.52 $15.40 n/a $13.51 Dentist Charges n/a $14.43 $14.24 $11.65 Paid Claims $5.80 $12.32 $13.23 $11.12 *Medicaid costs did not include $2 million spent on orthodontics (not covered by other plans) Amount equals dentist s charges ADA-related costs not available for John Deere Policy implications The survey results clearly indicate that providing children with dental coverage through the hawk-i program has significantly improved their access to dental care services. The slight reduction in the percentage of children with a dental visit from 2001 to 2003 should be monitored to make sure this downward trend does not continue. Dental utilization rates for children in hawk-i and Medicaid are generally more comparable than for children in those programs nationally. The 63-70% of children in hawk-i found to have had a dental visit in their first year in hawk-i from the follow-up survey is slightly lower than data from the National Health Interview Survey (NHIS), which indicated that 73% of children in Medicaid and other public programs had a dental visit in the previous year. 1 Unmet need, while significantly decreased, was still slightly higher than the 6.8% found in the NHIS for publicly insured children. Dental utilization rates based on 1 Day AN, Bloom B Summary health statistics for US children: National Health Interview Survey, National Center for Health Statistics. Vital Health Stat 10(223). 5

15 surveys are frequently higher than those found using administrative claims and encounter data, as was true in this study. Dental utilization rates in hawk-i and Iowa Medicaid are generally higher than the available national estimates based on administrative data. The 35% utilization rates found using the CMS methodology is higher than the 27% found using a similar approach for Medicaid nationally. The 45% and 55% rates found using the HEDIS approach in this study are also higher than the 37% found for Medicaid health plans nationally using the HEDIS methodology. Differences in costs for the first episode of care were found for children by the health plan in which they were enrolled. Costs, based on the ADA survey fees, were found to be highest for children in Medicaid ($376 per initial episode) as compared to all hawk-i plans. John Deere ($223 per initial episode) had the lowest costs for this initial episode of care among hawk-i plans, while costs for Wellmark ($284 per initial episode) and IHS ($285 per initial episode) were very similar. There were significant differences in costs related to how soon the child received their first visit as well. Children receiving care sooner had higher average costs for the initial episode of care than children receiving care later. This tends to indicate some degree of pent-up demand among the population, but whether there is a higher proportion of earlier utilizers with higher needs compared to other populations is unclear. There will always be some children with higher needs, but these results indicate that a larger proportion with greater needs tend to seek care sooner after enrolling. The programmatic cost analyses were an attempt to evaluate the total expenditures paid by the plans for dental care in FY 2003, and the pm/pm costs, which take into account differences in the number of children enrolled in the different plans. The plans are paid approximately $15 pm/pm for dental services, however, this is bundled in one total payment for all required services provided under contract with the Department of Human Services. Based on the paid claims, the actual amount paid by the plans ranged from about $11 for Wellmark to about $13.25 for John Deere. Because of a number of possible factors, including reimbursement rates being lower than the ADA survey fees, children reaching the annual maximum or needing provision of uncovered services, the paid amount was $2 to $3 less than the costs based on the ADA survey fees for children in Wellmark and IHS, respectively. It is also about $2 to $3 less than the $15.50 pm/pm for which the plans are paid to provide dental care. Medicaid reimbursed costs pm/pm were about $3.70 less than the cost estimated from the ADA survey fees. The difference between the amount paid to the managed care plans and the amount they pay in claims to the dentists covers the administrative costs of the plans to operate the program, as well as operating profit. The appropriateness of the amount that should be available for the administrative expenses and profit is a matter for policymakers to determine. Overall, children in John Deere, the open-access plan paying full charges, were more likely to receive their initial visit sooner than children in the other plans yet had lower costs for the initial episode of care. The higher utilization rates for new enrollees in John Deere may be related to several factors. First, John Deere operates an open access dental panel, where children can receive dental care from any dentist who will see them. The dentist is not required to submit paperwork to become a part of their provider network. Second, John Deere pays the dentists their full private practice charge. While it may take some dentists 6

16 longer to understand how to participate in John Deere, since they are not signing up with a network, once they figure it out, being paid their full charges is an obvious positive incentive to treat children in this program. Third, John Deere is located primarily in metro areas. This may provide better access to dentists due to the availability/distribution of dentists in Iowa than some of the more rural areas of the state. As becomes more apparent in the cost section, the new enrollees in John Deere may also have fewer treatment needs (if received service is related to need) since their average costs for the first episode of care are much lower. Children in John Deere also had lower costs per enrolled child for the first episode of care. The lower costs for the initial episode could be related to several factors: (a) John Deere is covering a population with lower needs in general or (b) John Deere is providing better access to care for children, allowing a higher proportion of children with less severe needs to receive care (diluting the average cost for the initial episode of care). The greater access hypothesis is supported by other analyses in a report to the State of Iowa Department of Human Services, indicating that the pm/pm paid dental claims were highest for John Deere ($13.23) compared to the IHS closed panel plan ($12.32) or the Wellmark indemnity plan ($11.12). Even at $13.23 paid out, the dental care portion of the capitation rate paid to the plans for providing services to hawk-i-enrolled children is approximately $15.50, allowing over $2 for administrative expenses/profit per child per month for all plans, even with John Deere paying full dentist charges. John Deere seems to have found a dental delivery system model that provides improved access to dental care for children in a public insurance program, easier participation on the part of dentists at reimbursement rates equal to their usual charges while still managing to recoup administrative expenses in the range of $2-3 per child per month. Further investigation into using a similar model for other programs should be evaluated. 7

17 8

18 Chapter 1. Overview of the study and dental care in the hawk-i program Introduction As part of the quality assurance activities for the hawk-i program for fiscal year 2005, the Iowa Department of Human Services and the hawk-i Clinical Advisory Committee requested that the University of Iowa Public Policy Center conduct a study concerning the use of dental services by children in the hawk-i program. Because each health plan in hawk-i uses a different approach to providing dental services, one of the primary questions for this study was whether the use of dental services varied based on the plan. For these analyses, Medicaid was included as an additional comparison group since it is a long-established public insurance program with yet another approach to providing dental services. This study was reviewed and approved by the University of Iowa Committee for the Protection of Human Subjects Internal Review Board (IRB). This study investigated the following research questions: 1. Did enrollment in the hawk-i program improve access to dental care for children based on responses to the pre/post evaluation survey? 2. What were the dental utilization rates for children in hawk-i based on the insurance encounter data and how were they affected by the method used to calculate the rates? 3. How soon did children receive dental services after enrolling in hawk-i and what factors were associated with receiving a dental visit sooner after enrolling? 4. What were the costs associated with providing dental services to children in hawk-i, and was there a difference in the costs for children who received care sooner (a measure of pent-up demand)? Also, what was spent on the different types of dental services (diagnosis and prevention vs. routine restorative care vs. extensive restorative care)? Background on the hawk-i program The hawk-i program is part of Iowa s State Child Health Insurance Program (also known as SCHIP or Title XXI), which is designed to provide health insurance coverage for uninsured children in Iowa whose incomes fall between 134% and 200% of the poverty level. The Iowa legislature authorized the creation of a two-part combination SCHIP program (Figure 1-1). The first part is a Medicaid expansion (M- SCHIP) for children with family incomes up to 133% of the federal poverty level (FPL). The second component is hawk-i, the separate state child health insurance program (S-SCHIP). hawk-i provides 9

19 health insurance for children with family incomes ranging from 134% to 200% of the FPL (the upper eligibility limit was raised from 185% to 200% of the FPL on July 1, 2000). In this program, the State of Iowa contracts with private health plans to provide covered services to enrolled children in the program. In hawk-i, families with incomes from 134% to 150% of the FPL have no premiums or copayments, while those with household incomes from 151% to 200% of the FPL pay a premium of $10 per child per month up to a maximum of $20 per family per month. For those above 150% of the FPL, there is also a $25 fee for non-emergent care provided in an emergency room (non-emergent is defined following the prudent layperson standard). The first recipients were enrolled in hawk-i in January As of January 31, 2005, there were 18,854 children enrolled in hawk-i. Figure 1-1. Income eligibility guidelines for hawk-i Medicaid For the period of time included in this study, John Deere Health Plan and/or Iowa Health Solutions Health Plan were the managed care plans available to enrollees in 37 Iowa counties (Figure 1-2). These are the counties in which the health plans contracted to provide all covered services including dental care. Wellmark is an indemnity plan that operates in all counties that do not have a managed care plan. 10

20 Figure 1-2. Distribution of hawk-i health plans by county during the time period of these studies (up until January 30, 2005) hawk-i and Medicaid dental plans The hawk-i program pays each health plan a per member/per month (pm/pm) payment for all covered services including dental care. There are some differences in the dental services covered and in the annual maximum payment for dental services in the hawk-i plans and Medicaid. All plans cover most routine preventive and primary restorative dental services, including root canals (Table 1-1). Table 1-1. Dental benefits for children in ALL hawk-i and Medicaid plans All plans cover these services Preventive & Diagnostic Primary Services Examinations Fillings Cleanings Root Canals Fluoride treatment for children Periodontics X-rays routine bitewing Oral Surgery X-rays full mouth Sealants 11

21 Differences in covered dental services and annual maximum dental benefits are detailed in Table 1-2. All plans except Wellmark covered crowns, bridges and dentures. Medicaid is the only plan that provides limited orthodontic coverage. Iowa Health Solutions had the highest annual maximum dental benefit among the hawk-i plans, but required dentists to submit preauthorization for treatment plans over $200. Medicaid has no annual maximum dental benefit. Table 1-2. Differences in dental benefit packages for children in hawk-i and Medicaid Iowa Health Solutions Wellmark John Deere Health Medicaid Preventive & Diagnostic Preventive & Diagnostic Preventive & Diagnostic Preventive & Diagnostic Space maintainers up to age 18 Space maintainers (no age limit) Space maintainers up to age 18 Space maintainers (no age limit) Major Services covered Major Services covered Major Services covered Major Services covered Crowns Crowns Crowns Crowns Dentures Dentures Dentures Bridgework Bridgework Bridgework Limited Orthodontics Non-covered services Non-covered services Non-covered services Non-covered services Orthodontics Dentures Orthodontics None Bridgework Orthodontics Annual maximum dental benefit Annual maximum dental benefit Annual maximum dental benefit Annual maximum dental benefit $1,500 per calendar year Preauthorization required for treatment plans over $200 $1,000 per benefit period No preauthorization required $1,000 per calendar year No preauthorization required Unlimited No preauthorization required Each hawk-i health plan is responsible for organizing their own dental provider networks and paying the dentists in a way that allows children to have access to dental care. While the three health plans have chosen different approaches to establishing their dental programs, all plans pay dentists on a fee-forservice (FFS) basis (Table 1-3). Table 1-3. Dental plan structures for hawk-i and Medicaid Plan attributes Wellmark IHS John Deere Medicaid Payment method 90 th percentile FFS Reduced FFS Provider panel type Any participating dentist Closed panelparticipating dentists only Full dentist charges FFS Any willing dentist no sign up required Reduced FFS Any participating dentist At the time of the study, Wellmark operated an indemnity dental insurance program using the same provider network available to their privately insured population. They reimbursed participating dentists 12

22 up to the 90 th percentile of usual and customary fees. John Deere operated an open access plan through which they reimbursed dentists their full charges for care provided. Iowa Health Solutions (IHS) had a closed panel of providers in most counties, only allowing dentists who had been accepted through a credentialing process to provide care for IHS enrollees. In a few counties where they did not have an adequate provider panel, enrollees could go to any dentist if approved by IHS in advance. IHS paid dentists based on an established fee schedule that was at a level between that paid by Medicaid and the other hawk-i plans. The Medicaid program uses a traditional, statewide, fee-for-service model dental program operated by the Iowa Department of Human Services. Any dentist willing to sign up and accept Medicaid-enrolled children at the Medicaid fee schedule can participate. Medicaid is required to provide comprehensive dental care for children under the Early and Periodic Screening, Diagnosis and Treatment program (EPSDT) with no annual maximum amount of coverage. Lack of provider participation has been found to be a barrier to children receiving dental services; however, Iowa appears to be doing better than many states regarding the level of dentist participation (Table 1-4). Internal analyses of Medicaid claims data by the Iowa Department of Human Services for fiscal year 2004 are indicated in the table. Table 1-4. Dentist participation in the Iowa Medicaid program No. of dentists enrolled with program No. of dentists with at least one paid claim No. of dentists with paid claims > $10,000 (considered significant providers) Percent of counties without an enrolled Medicaid dentist Percent of counties without an enrolled Medicaid dentist with paid claims > $10, pvt practice 1071 pvt practice /99 counties (5%) 125 U of I faculty & student 114 out-of-state providers 66 out-of-state providers 16 out-of-state providers Source: Cathy Coppes, Iowa Department of Human Services The current hawk-i dental program Beginning on February 1, 2005, Iowa Health Solutions ceased to be a hawk-i participating health plan and Delta Dental Plan of Iowa became an option for dental care for hawk-i enrollees in counties with the John Deere managed care plan. The distribution of counties and plan options by county is shown in Figure

23 Figure 1-3. Distribution of hawk-i health plans by county beginning February 1,

24 Chapter 2. Access to dental services for children in hawk-i The primary research questions on access to dental services were: 1. Was access to dental services improved by providing children with hawk-i insurance? 2. Did the need for dental care and unmet need for dental care change over time? Children included in this study: All children enrolled in the hawk-i program for whom a baseline and followup survey was received during FY Introduction One of the primary aims of the hawk-i program was to increase access to care by providing improved financial access to services, including dental care. Previous studies have indicated that dental care is the area of highest unmet need for low-income children. The purpose of this study is to evaluate if the hawk-i program improved access to dental care for enrolled children, and whether the need and unmet need for such care changed over time. Methods A longitudinal pretest-posttest panel study design is being used to evaluate the effect of hawk-i health insurance on the access to care and health status of enrollees. The study process is ongoing. Surveys are conducted with each household upon entry into the hawk-i program and again at the one-year enrollment anniversary. A mixed-mode data collection process is used, starting with a mailed survey and continuing with a telephone data collection process for nonrespondents. Survey instruments Questionnaires for the baseline and follow-up surveys are nearly identical. The 60-item baseline survey instrument asks about the 12 prior to joining hawk-i, and includes questions on the child s health status, presence of chronic conditions, physical and behavioral/emotional limitations, and access to health care, including medical care, dental care, mental health care, prescription medicine, and vision care. The follow-up questionnaire asks the same questions with the addition of a few questions concerning the hawk-i plan in which respondents were enrolled. A question about the length of time since receiving a dental check-up is included only in the follow-up survey. Demographic sections were included in both questionnaires. Only results for the dental questions are presented in this report. 15

25 Both survey instruments were developed by researchers at the University of Iowa Public Policy Center at the request of and in consultation with the hawk-i Clinical Advisory Committee (see previous reports). The questions were developed after review of existing documents such as the National Health Interview Survey (NHIS), 2 the Consumer Assessment of Health Plan Study (CAHPS), 3, 4 the SCHIP Program Evaluation Guidelines established by the American Academy of Pediatrics, 5 and enrollee surveys used to evaluate the Iowa Medicaid program. 6 Questions were also added about children with special health care needs; these were developed by the Children with Special Health Care Needs Subcommittee of the hawki program. Survey process A modified Dillman method 7 is used for both the baseline and one-year follow-up mailed questionnaires. The process is as follows: (1) a prenotification postcard is sent, addressed to the parent or guardian of one randomly chosen child per household, (2) about one week later, a cover letter is sent along with a questionnaire and business reply envelope, (3) one week after the letter and questionnaire, a reminder postcard is sent to each household, and (4) three weeks after the postcard, a second letter, questionnaire and business reply envelope are sent to nonrespondents. Telephone surveys are conducted with families that do not respond to the mailed survey. Calls are made until: (a) a telephone questionnaire is completed, (b) a refusal is obtained, or (c) 10 failed contact attempts have been made. MAXIMUS, the third party administrator (TPA) for enrollment and claims processing for the hawk-i program, has conducted the data collection for the surveys presented in this report. Sample selection Each household with a child enrolled in the hawk-i program was asked to participate in this study. To reduce respondent burden for families with more than one enrolled child, the survey vendor (MAXIMUS) randomly selected one child per household to be the subject of the survey. The parent or guardian of this child was asked to fill out the survey as it related to the child s health and health care. Data used to evaluate hawk-i In order to assess the effect of hawk-i on comparable populations, only data for children from whom both a baseline and follow-up survey were received were used in this analysis. This allows for a more accurate 2 National Health Interview Survey, National Center for Health Statistics, US Department of Health and Human Services. Available at Accessed most recently November 7, CAHPS Survey and Reporting Kit, Agency for Health Care Research and Quality. Washington, DC: US Department of Health and Human Services, Public Health Service. 4 CAHPS Products. CAHPS-Survey Users Network. Available at Accessed most recently November 7, SCHIP Evaluation Tool. American Academy of Pediatrics. Available at Accessed most recently November 7, Damiano PC, Tyler MA, Momany ET Evaluating Iowa Medicaid Managed Care Plans: A Consumer Perspective. Final report to the Iowa Department of Human Services. University of Iowa: Public Policy Center, Iowa City, IA. 7 Dillman DA Mail and Internet Surveys: The tailored design method, 2nd Edition. New York: Wiley and Sons. 16

26 evaluation of the impact of the program. The data used in this report were for children who initially enrolled in the program between July 1, 2000, and June 30, This approximates children who were enrolled during the second, third, and fourth years of operation of the hawk-i program. Data were not yet available from the follow-up survey for children who first enrolled during FY There were a total of 5,143 cases with data for both the baseline and follow-up survey. The process of identifying children for whom there were complete data began with the receipt of data for completed follow-up surveys from the survey vendor through June 30, Results Five areas concerning dental care that were evaluated from the survey: 1. Need for dental care 2. Unmet need for dental care and reason why 3. Delays for dental care and reason why 4. Regular source of dental care 5. Time of last dental visit 6. Time of last dental check-up Need for dental care The results for the baseline and follow-up survey regarding need for dental care are presented in Figure 2-1. There was little difference in reported need for dental care between baseline and follow-up survey in any of the three years, however, there was a slight drop in the percentage reporting need for dental care between FY 2001 and Figure 2-1. Need for dental care, FY by plan 17

27 Unmet need for dental care Reported unmet need for dental care (those who were stopped from receiving dental care at some point during the year) from the baseline and follow-up surveys is presented in Figure 2-2. There was a significant decrease in unmet need for dental care from the year before having insurance to the year after being in hawk-i in all three fiscal years. The percentage with unmet need is relatively constant in the baseline survey over time, while the rates of unmet need for dental care after being in hawk-i for a year declined moderately from FY 2001 to FY Figure 2-2. Unmet need for dental care, FY by plan Delays in receiving dental care Similar to unmet need for dental care, delays in getting care are defined as times when the child was not able to receive care as soon as the parent thought they should. Figure 2-3 shows that in all three fiscal years there was a significant decrease from the year before to the year after being in hawk-i in the percentage of children who had to wait longer to receive needed dental care than they thought they should have. The percentage with delays declined slightly at baseline from FY 2001 to FY 2003, as did the percentage with delays in John Deere. The percentage with delays in Wellmark remained relatively constant. 18

28 Regular source of dental care Figure 2-3. Delays in receiving dental care, FY by plan Having a regular source of dental care is an important aspect of access to comprehensive care. Figure 2-4 shows the percentage of children who had one place where they could go if they needed dental care. There was a slight increase in the percentage of children who had a regular source of dental care from the year before to the year after being in hawk-i in all three fiscal years. The percentage with a regular source of care declined slightly at baseline from FY 2001 to FY 2003 for children entering John Deere. Figure 2-4. Regular source of dental care, FY by plan 19

29 Time of last dental visit Having had a dental visit in the past six to 12 is another indication that a child is receiving adequate access to dental care. Figure 2-5 shows the percentage of children that had had a dental visit in the past 12. There was a significant increase in the percentage of children who had a dental visit in the 12 after being in hawk-i for a year. This improvement was consistent for all three fiscal years. The percentage with a dental visit in the previous 12 declined slightly, especially for the follow-up period from FY 2001 to FY Figure 2-5. Percentage with a dental visit in previous 12, FY by plan Time of last dental check-up Having a dental check-up in the past six to 12 is an indication that a child is receiving comprehensive dental care in addition to being able to access services. Figure 2-6 shows the percentage of children that had a dental check-up during their first year in hawk-i. In general, about three-quarters of children in hawk-i had a dental check-up during the year. The rates declined in all plans from FY 2001 to FY IHS had the lowest rates of receiving a dental check-up in all three years (under 65%). 20

30 Figure 2-6. Percentage with a dental check-up in previous 12, FY by plan 21

31

32 Chapter 3. Dental Utilization Rates for Children in hawk-i The primary research questions regarding the use of dental services were: 1. What are the dental utilization rates for children in hawk-i based on insurance claims/encounter data? (i.e. what percentage of children in hawk-i receives a dental visit during the year)? 2. Are these dental utilization rates and the comparison between plans affected by the approach used to calculate these rates? Children included in this study: All children enrolled in hawk-i and Medicaid during calendar year (CY) Background Public insurance programs like Medicaid and SCHIP are an important way of improving financial access to dental services for children. Most children below the poverty level who receive dental care do so through the Medicaid program. 8 Although Medicaid is the largest payer of dental services for children in poverty, long-standing concerns about access to dental care for children enrolled in Medicaid persist. 9,10 Less is known about the impact of SCHIP programs on utilization of dental services. As shown in the previous chapter, surveys have indicated that unmet need for dental care has declined for enrollees in hawk-i. However, few studies have used insurance claims data to document dental utilization rates for children in SCHIP programs. Previous studies of children in Medicaid have found low yet varying levels of dental utilization. One of the most frequently cited analyses of children s dental utilization in Medicaid is the 1996 report by the U.S. Department of Health and Human Services Office of the Inspector General (OIG). 11 The OIG report found that in 1993, less than 20 percent of Medicaid-enrolled children throughout the U.S. received a preventive dental service. More recent data from the Centers for Medicare and Medicaid Services (CMS) 8 US Congress, Office of Technology Assessment Children s dental services under the Medicaid program-background paper. Washington, DC: US Government Printing Office. 9 Oral health: Factors contributing to low use of dental services by low-income populations. GAO/HEHS April 12, United States General Accounting Office. Available at: Last accessed March 1, U.S. General Accounting Office (GAO) Dental disease is a chronic problem in low-income populations. Washington, DC: United States General Accounting Office, GAO/HEHS Office of the Inspector General (OIG), U.S. Department of Health and Human Services Children s dental services under Medicaid: access and utilization. San Francisco, CA: Office of Evaluation and Inspection; OE

33 show that 27% of children enrolled in Medicaid had a dental visit during Reports based on data collected by the National Committee on Quality Assurance (NCQA), in contrast, found that 37% of children in Medicaid had a dental visit during Some of the differences in these rates (especially between the CMS and NCQA figures) stem from differences in the methodologies used to calculate utilization rates; in particular, which criteria were used to determine whether to include or exclude enrolled children in the analyses based on the length and continuity of their enrollment in the year of record. Other plausible factors underlying the differences are the proportion of Medicaid enrollees included in the respective databases and the enrollment characteristics of these children. Several approaches for calculating utilization rates from administrative data for Medicaid and SCHIP enrollees are described below. 1) Centers for Medicare and Medicaid Services (CMS) Form 416 Methodology All states are required to submit utilization data for services covered by the Medicaid Early and Periodic Screening Diagnosis and Treatment (EPSDT) benefit including dental services annually to CMS. The CMS-416 rates are calculated using all children enrolled at any point during the year of record as the denominator. The strength of this approach is that it considers all children enrolled in the program during a given year, and thus can be compared to other national measures (e.g., self-reported utilization rates obtained from surveys) that usually include children regardless of length or type of coverage. 2) Health plan Employer Data and Information Set (HEDIS) Methodology HEDIS is a set of measures developed by NCQA for evaluating quality in health plans. With this approach, the annual utilization rate includes only those who have been enrolled for in the year of record. Thus the HEDIS approach provides a comparison of utilization for individuals with a similar enrollment period or opportunity to schedule and receive services, albeit a relatively lengthy period. 3) Full-Time Equivalent (FTE) Methodology Similar to calculations of FTEs in the workplace, this approach includes all children enrolled at any point in the year, but adjusts the denominator of the rates by summing the total number of for all children and dividing by 12. This measure includes all children enrolled in the program at any point during the year and then adjusts globally for partial-year enrollment. 12 American Dental Association State Innovations to Improve Access to Oral Health Care for Low Income Children: A Compendium. Chicago: American Dental Association. 13 National committee for Quality Assurance (NCQA). Medicaid HEDIS 2002 Audit Means, Percentiles & Ratios: Annual Dental Visits. Available at: Accessed November 5,

34 4) New Enrollee Methodology In this study, we also investigate an approach that calculates utilization rates for only those children who were newly enrolled in the program (i.e., who had not been in the same program during the previous 12 ). This rate can be important when high enrollee turnover and/or access to care is of particular concern, and there is interest in assessing the ease with which children can enter the dental care system after obtaining coverage. This method was used both for all children and for those enrolled for during the year (as in the HEDIS approach). Which children are included in each of the 4 approaches used to determine dental utilization rates: 1. CMS 416 reporting approach: All children enrolled at any point in the year 2. HEDIS approach: All children enrolled for during the year 3. FTE approach: All children enrolled at any point in the year the total number of of eligibility during the year is added together and divided by 12 to determine the number of full time equivalent children enrolled during the year 4. New enrollees approach: Only children who had not been enrolled in the program in the previous 12 Methods Data Sources Iowa hawk-i and Medicaid enrollment and claims/encounter data for all children ages 1 to 18 for calendar year (CY) 2001 were used in these analyses. All children in this age range at the end of 2001 were identified in the eligibility files, and select demographic and enrollment period data were assimilated. All claims for dental services provided for children enrolled in Iowa s Medicaid and hawk-i programs during 2001 were also obtained from the respective claims files. The enrollment and claims data were then matched for each enrolled child. 25

35 Calculation of Dental Utilization Rates Dental utilization rates for the four different approaches were calculated by dividing the following numerators by the given denominators: The CMS 416 Methodology Number of children with a dental visit during the year Number of children enrolled at any point during the year The HEDIS Methodology Number of children enrolled for with a dental visit during the year Number of children enrolled for during the year The Full-Time Equivalent Methodology Number of children enrolled at any point with a dental visit during the year Total number of children are enrolled during the year / 12 The New Enrollee Methodology Number of children newly enrolled at any point in the year with a dental visit Number of children newly enrolled at any point in the year Types of services received were also categorized into diagnostic, preventive, restorative and complex restorative categories using CDT codes. Results Descriptive characteristics of the children enrolled in hawk-i and Medicaid in calendar year (CY) 2001 are shown in Table 3-1. There were about seven times as many Iowa children enrolled in Medicaid as in hawk-i during the year. Children enrolled in hawk-i were more likely to have been in the program for shorter periods of time than Medicaid enrollees. The hawk-i-enrolled population was also older than the Medicaid enrollees, with about one in four Medicaid-enrolled children being in the 1-3 age category compared to 16% for hawk-i. Figure 3-1 shows the relative impact of using each approach to measure dental utilization on the proportion of all children in hawk-i and Medicaid included in the analysis. The CMS 416, using data for all children enrolled at any point during the year, by definition includes all children from both plans. Evaluating only newly enrolled children who were in the program for incorporates the lowest proportion of children eligible during the year. The different enrollment patterns for children in Medicaid and hawk-i also create differences in the proportion of children included in the other approaches. For example, since children tend to be enrolled longer in Medicaid, a larger proportion of children were included in the HEDIS and FTE approaches, while a larger proportion of hawk-i enrollees were included in the approach evaluating only new enrollees. 26

36 Table 3-1. Demographics of children in Medicaid and hawk-i, CY2001 Characteristic Medicaid hawk-i Enrolled children by approach CMS: Enrolled at any point in year 141,154 18,961 HEDIS: Enrolled for ,641 5,289 FTE: Total FTEs 106,290 11,366 Newly enrolled: At any point in year 20,076 9,726 Newly enrolled: Enrolled for ,439 1,031 Gender (% female) 50% 50% Age (enrolled any point) % (35,496) 16.4% (3,104) % (27,164) 18.5% (3,503) % (46,338) 37.6% (7,134) % (32,159) 27.5 (5,220) Less than 50% of FPL 47% Not applicable Figure 3-1. Proportion of all children in Medicaid and hawk-i during CY2001 that are included in each approach to measuring dental utilization About one-third of all children in both Medicaid and hawk-i were identified as having had a dental visit during 2001 using the CMS method of including all enrolled children, regardless of length of enrollment (Figure 3-2). The resultant percentages of children identified as having a dental visit were higher using the HEDIS approach for children in both programs. Differences in utilization rates between the two public insurance programs also increased when the HEDIS method was used i.e., 10% more hawk-i enrollees than Medicaid enrollees were identified as having had a dental visit. Use of the FTE methodology produced rates similar to the HEDIS methodology, although the hawk-i rate was 3% higher. The analyses based only on new enrollees produced the lowest utilization rates, especially for Medicaid enrollees. When including only children who were newly enrolled for the year and were enrolled for 11-12, the rates for both Medicaid and SCHIP were closer to the HEDIS and FTE rates (although the number of children who met the criteria to be included in these analyses were quite low). 27

37 100% 80% 60% 40% 20% 0% 34% 35% 45%55% 58% 45% CMS 416 HEDIS FTE All new enrollees Calculation methodology 50% 41% 31% 18% New month Medicaid SCHIP Figure 3-2. Dental utilization rates calculated using different approaches CY 2001 Dental utilization rates were lowest for children ages 1-3 with all approaches, and highest for children ages 4-12 (Table 3-2). Rates for adolescents were slightly lower than for 4-12 year olds. With the CMS approach, utilization rates for children ages 4 and over were higher in Medicaid than in hawk-i. For all other methods, the dental utilization rates for children of all ages were higher in hawk-i than in Medicaid. The highest dental utilization rates identified were for children ages 7-9 in hawk-i (69.2%) using the FTE methodology. Table 3-2. Dental utilization rates by age and approach Age 1-3 Age 4-6 Age 7-12 Age Methodology N % N % N % N % CMS 416 method Medicaid % 11,570 43% 20,389 44% 11,850 37% hawk-i % % % % HEDIS method Medicaid % % 14,735 56% % hawk-i % % % % FTE method Medicaid % 11,570 56% 20,389 58% 11,850 50% hawk-i % % % % New enrollee method Medicaid 282 7% % % % hawk-i % % % % New enrollee method Medicaid 40 16% % % % hawk-i 30 17% % % % 28

38 Figures 3a-3c demonstrate the effects of including children who are enrolled for different lengths of time (numbers of ) when calculating dental utilization rates. Figure 3a is comparable to the CMS approach, where each succeeding increment represents the inclusion of children enrolled for an additional month, and culminating with all children enrolled at any point during the entire year (i.e., children enrolled anywhere from 1 to 12 during the year of record). Calculated utilization rates increase with the length of enrollment from 3% for children enrolled for only 1 month up to 34%-35% for children enrolled for a full 12, undoubtedly reflecting a greater likelihood of obtaining services with a longer window of opportunity. Figure 3-3a. Figure 3-3b. Figure 3-3c. 29

39 Figure 3-3b shows dental utilization rates for children who were enrolled for a set number of (1 month, 2, etc.) during the year rather than up to 12 as in 3-3a. The percentage of children receiving at least one dental visit increases relatively consistently until the 12-month increment, where there is a 10% jump in utilization for children who were enrolled for the entire year (compared to the rate for those enrolled for 11 ). Figure 3-3c, shows the effect of including children who are enrolled in Medicaid and hawk-i for varying minimum lengths of enrollment (i.e., at least 11, 10, etc.) throughout the year of record on reported dental utilization rates. These are the utilization rates that would result if the HEDIS criteria were more inclusive, going beyond just those who have been enrolled for For example, if the HEDIS criteria were changed to include all children enrolled for at least 6 (i.e., enrolled in the program for 6-12 during the year), the reported dental utilization rates would be 40 percent for Medicaid and 47 percent for hawk-i. Narrowing the inclusion criterion to include all children enrolled for at least 8 during the year of record would yield reported utilization rates of 42% and 50% for Medicaid and SCHIP, respectively. The figure thus shows the effect of inclusion criteria that reflect the range of enrollment periods between the current CMS-416 criteria (of being enrolled for any length of time during the year of record) and the HEDIS criteria (of being enrolled for at least 11 or 12 during the year of record). Among children with a dental visit during the year (Figure 3-4), almost all children in both programs (93%) received a diagnostic procedure, and the vast majority (84%) received a preventive procedure. About one in three children received routine restorative care, and approximately one in six received a complex restorative procedure. Figure 3-4. Services received by children with a dental visit 30

40 Chapter 4. Time to First Dental Visit for Children in hawk-i The primary research questions regarding the time until a first dental visit for children in hawk-i were: 1. How soon after enrolling in hawk-i did children receive their dental visit (an important indicator of access to care)? 2. What factors were associated with how soon they received their first dental visit? Children included in this study: All children with a dental visit who were enrolled in hawk-i and Medicaid during fiscal years (FY) 2001, 2002 and Little has been written concerning the timing of the first dental visit once a child has been enrolled in a public insurance program such as SCHIP or Medicaid. Though we know that children with insurance are more likely to have a dental visit, how soon they receive such a visit after first enrolling in a program, especially a public insurance program where needs may be greater, is unclear. Evaluating utilization rates for a single year (i.e., cross sectional approaches), as in the previous chapter, combines information for children who have been in a program for a long time with children who are newly enrolled during the year. Enrollees with experience in a program are more likely to know how to use public insurance appropriately and how to locate a participating provider. This particular study is designed to determine the timing of the first dental visit for children newly enrolled in the hawk-i program and the factors associated with how soon after enrollment their first dental visit occurs. The timing of a dental visit for new enrollees is an important indicator of access to dental care for children in hawk-i. Methods Data Iowa hawk-i and Medicaid enrollment and claims/encounter data for all children ages 1 to 18 for fiscal years (FY) 2001, 2002 and 2003 were used in these analyses. Only newly enrolled children and adolescents were included. To be considered newly enrolled, a child or adolescent had to be enrolled during the study period with no evidence of enrollment in the 12 preceding the study period in the same program (hawk-i or Medicaid). This resulted in a final study population in both programs of 96,704 children and adolescents. The first dental visit was evaluated in two ways. The first analysis looked at the time until any dental visit, defined as any visit with services rendered by a dentist with an acceptable CDT code. The second analysis evaluated time to any preventive dental visit, defined as a visit with a dentist rendering a 31

41 diagnostic or preventive service with a CDT code from and Receiving preventive dental care may be more indicative of a child beginning comprehensive dental treatment. Once a child or adolescent was identified for inclusion in the study, five outcomes were possible during the three-year study period (FY ). Four of these outcomes (1) being no longer eligible or enrolled, (2) becoming 19 years of age, (3) switching plans, and (4) not receiving dental care when the study ended resulted in the child being removed from the study at that point (i.e., censoring as it is referred to in survival analysis). The possible associations between the time to a dental visit (i.e., the dependent variable) and the factors of age, gender, race and location (the covariates) were evaluated using survival analysis techniques. Survival analysis is a helpful technique because it allows the information for all children with a dental visit during the study period to be included in the analysis regardless of whether they were in for the entire three-year period. The associations between time to a dental visit and each covariate separately (the bivariate relationships) were estimated using the life table method, due to the large sample size. This method allows for the definition of time intervals to be used in the estimation procedure and provides essentially the same results as the more widely utilized Kaplan-Meier method. To evaluate the association of the factors as a group to the time to a dental visit (the multivariate relationships), Cox proportional hazard modeling was completed for the two different types of dental visits. The same model was utilized for each analysis including covariates. The full model was utilized for these analyses with no selection method employed. Results Table 4-1 provides descriptive information about the children with a dental visit. A much larger proportion of these children were enrolled in Medicaid (74%), which is reasonable given its larger size; the remaining children and adolescents were distributed among the three hawk-i plans (7% in Iowa Health Solutions-IHS, 7% in John Deere and 12% in Wellmark). Children in all four plans were enrolled for an average of about 11 during the three-year period, however, the enrollment distributions were different. Medicaid-enrolled children were significantly more likely to be enrolled for both the shorter (1-6 ) and longer (25-36 ) periods of time compared to children in hawk-i. Children in Medicaid were also more likely to be adolescents. John Deere was by far the most urban plan, with 88% of enrollees living in a metropolitan area. 32

42 Table 4-1. Demographics of children in Medicaid and hawk-i with a dental visit (FY ) Characteristic Medicaid IHS John Deere Wellmark Enrollment No. Enrolled at any point 71,313 7,252 6,610 11,529 Months enrolled (mean) ,131 (41%) 1,955 (27%) 2,030 (31%) 3,328 (29%) ,028 (25%) 3,365 (46%) 3,043 (46%) 5,149 (45%) month 16,233 (23%) 1,488 (21%) 1,231 (19%) 2,202 (19%) ,931 (11%) 444 (6%) 306 (5%) 850 (7%) Gender % Female 52% 49% 49% 49% Age (enrolled any point) ,367 (20%) 1,787 (25%) 1,661 (25%) 2,494 (22%) 4-7 1,718 (24%) 1,683 (26%) 2,719 (24%) ,322 (27%) 2,070 (29%) 1,877 (28%) 3,416 (30%) ,130 (28%) 1,677 (23%) 1,389 (21%) 2,900 (25%) Race/ethnicity White 40,693 (57%) 4,562 (63%) 4,127 (62%) 7,648 (66%) Race other than white 8,694 (12%) 472 (7%) 364 (6%) 403 (4%) Unspecified 21,926 (31%) 2,218 (30%) 2,119 (32%) 3,478 (30%) Location Metro 37,275 (52%) 3,582 (49%) 5,792 (88%) 2,367 (21%) Urban adjacent to metro 14,838 (21%) 2,231 (31%) 428 (7%) 2,355 (20%) Urban not-adjacent metro 14,175 (20%) 1,026 (14%) 4 (<1%) 5,283 (46%) Rural 5,026 (7%) 413 (6%) 386 (6%) 1,524 (13%) The survival analysis produced estimates of the percentage of children with a dental visit after being enrolled for 6, 12 and 36, taking into account the children who were only eligible for a portion of this time period (Figure 4-1). After 6 in the program, about one-fourth of all children had any dental visit. There were significant differences in the proportion with a dental visit by plan, however. Children in John Deere were most likely to have had a dental visit within the first 6 after enrollment (36%), while children in IHS were least likely to have had a visit (21%). Similar patterns were found for preventive visits, with about 2 percent less children receiving a preventive visit than any visit at each time interval (see Appendix B). After being in the program for a year, the proportion of enrollees with a visit increased to an average of 44%, ranging from 39% of children in IHS to 56% in John Deere after one year in the program, and from 73% in Medicaid to 88% in John Deere after three years. Similar age patterns were found as in the last study, with children ages 1-3 being least likely to have had a visit during any time period, but with the percentages increasing significantly as children got older and were in the program longer. 33

43 Figure 4-1. Adjusted dental utilization rates using survival analysis for children 6, 1 year and 3 years after first enrolling in each plan (FY ) The relationship between time to the first dental visit and health plan remained consistent with the bivariate analysis in the multivariate analysis, after controlling for the age, race, gender and location of the children (Appendix B Tables B2 and B3). As compared to Medicaid, children in IHS were similarly likely to have had any dental visit or a preventive dental visit in the same amount of time, while children in John Deere and Wellmark were statistically significantly more likely to have had either dental visit sooner after enrollment. 34

44 Chapter 5. Costs of Dental Care for Children in hawk-i The primary research questions regarding the costs of dental care for children in hawk-i were: 1. What were the dental costs for children newly enrolled in the hawk-i program during their initial episode of care? 2. Did the costs differ based on how soon they had a dental visit after enrolling in the program? (One potential indicator of some degree of pent-up demand.) 3. What factors were associated with the costs of care received? Children included in this study: All children with a dental visit who were newly enrolled in hawk-i and Medicaid during fiscal years (FY) 2001, 2002 and Introduction Concern has been raised about the costs of services for new enrollees in public insurance programs and the potential for high dental need and pent-up demand for dental care among such enrollees. High need could increase utilization and drive up costs either early in the program or overall and thus increase the capitation rates paid to managed care plans. Few studies, if any, have detailed the dental costs for children newly enrolled in insurance programs, especially SCHIP and Medicaid programs. The purpose of this study was to evaluate the costs of dental care for newly enrolled children in Medicaid and hawk-i, and to determine whether the length of time to when they received their first dental visit was related to the cost of the care received (one potential indicator of pent-up demand), as well as the factors that affect the costs of care received by children. Methods Data Iowa hawk-i and Medicaid enrollment and claims/encounter data for all children ages 1 to 18 for fiscal years (FY) 2001, 2002 and 2003 were used in these analyses. Only newly enrolled children and adolescents with a dental visit during fiscal years were included. To be considered newly enrolled, a child or adolescent had to be enrolled during the study period with no evidence of enrollment in the same program in the 12 preceding the study period (FY 2000). In addition, children and adolescents with a dental visit had to be enrolled for at least 4 following the month of the initial dental visit, (i.e., a total of at least 5 ) to be included in the study. This 5-month period represents the initial episode of dental care. To develop this definition of the initial episode of care, we used a combination of (1) a statistical analysis of the distribution of total dental costs following the initial 35

45 visit and (2) dental judgment concerning the length of time required to complete a typical set of diagnostic and therapeutic interventions in real-world settings after first seeing a dentist. Ninety-two percent of the costs for a child during the 6 following the initial dental visit were incurred in the first four. In addition, going beyond 6 would potentially extend into a second episode of care for children who received a dental check-up every six. Thus, 120 days from the day of the initial dental visit was considered the initial episode of dental care for all children in this analysis. Time of the first dental visit was considered important when evaluating pent-up demand. Children were placed into one of three groups based on how soon after enrollment they received their first dental care: (1) early utilizers (those who had a dental claim within three of enrolling), (2) middle utilizers (those whose first dental claim was from three to twelve after enrolling), and (3) late utilizers (those whose first dental claims was more than 12 from enrolling). Determination of dental costs After determining that four from the initial visit represented a comprehensive and reasonable episode of care, we collected cost data for each individual who had a dental visit. Each dental claim was classified by CDT codes into one of our four categories of dental procedures, and a cost was associated with each procedure code based on the 2003 American Dental Association (ADA) survey of dental fees for the West North Central region of the United States. 14 The ADA survey fees were used to approximate the Usual and Customary private practice fees in Iowa. All CDT codes no longer in use in 2003 were mapped onto appropriate 2003 codes to ensure complete collection of cost data. The ADA fee schedule was chosen because it represents a commonly accepted metric of costs that can be used to compare resource utilization across different health plan settings, as in this study. After assigning costs to each claim, we summed up total costs for a four-month period from the initial visit, and calculated the distribution of costs by the four subcategories. We examined measures of the central tendency of costs (mean, median) and the distribution of costs to test for the possibility of a skewed distribution (i.e., a small percentage of subjects with very high costs). The costs were tallied for all dental care as well as categorized into four areas of dental services (preventive/diagnostic, routine restorative, complex restorative and other). Analysis plan These analyses explored the relationship between the dental costs incurred in the first episode of care and demographic characteristics of the population receiving them. The main explanatory variables we investigated were the effect of the patient s plan (Medicaid, or one of the three hawk-i plans) and the time period after enrollment when children had their first dental visit. It was hypothesized that both plan type and length of time between enrollment and the first dental visit would be significant predictors of total costs in this first episode of care, after controlling for other factors. The associations between the costs of the initial episode of dental care (i.e., the dependent variable) and the explanatory factors of health plan, time of first dental visit, age, gender, percent of the federal poverty level, race and location (the covariates) were evaluated first using bivariate measures of association between the mean total dental cost for the initial treatment episode and categorical variables representing: (1) health plan and (2) the subject s timing of first dental visit Survey of Dental Fees Chicago, IL: American Dental Association. 36

46 After determining that there were significant differences in our primary measure of costs and the key explanatory variables, we constructed multivariate ordinary least squares regression models. We used two definitions of total cost as the dependent variable: total costs for the initial episode of dental care, and the natural logarithm of total costs, after descriptive analyses indicated the cost distribution was skewed to the right, that is, a small number of patients had very high costs. More details about the methodology used in the regression analyses can be found in Appendix C. Results The population of children newly enrolled during FY and the proportion of children meeting the selection criteria for being included in these analyses is shown in Table 5-1. About two-thirds of the children met the criteria of having been enrolled for at least 5 (study population). Between one-quarter and one-third of children had both a dental visit and were enrolled for the next 4, allowing for all services during that episode of care to be captured in the administrative data as the costs of the initial episode of care. Table 5-1. Study population for cost analyses Characteristic Medicaid IHS John Deere Wellmark N % N % N % N % Study population Dental visit 18,241 26% 1,909 26% 2,285 35% 3,585 31% No dental visit 28,932 41% 3,254 45% 2,241 34% 4,404 38% Excluded Dental visit only in last 4 of FY 2003 Less than 5 of total enrollment Enrolled in last 4 of FY ,487 6% 560 8% 567 9% 890 8% 12,422 17% % % 1,323 12% 7,227 10% % % 1,327 12% Total 71, % 7, % 6, % 11, % Table 5-2 shows the percentage of children who received a dental visit by the number of they were eligible during the three-year study period. Children in John Deere were most likely to have had a dental visit regardless of the length of enrollment. Children in Medicaid and IHS were least likely to have had a dental visit. Table 5-2. Dental visit by length of enrollment Characteristic Medicaid IHS John Deere Wellmark Length of enrollment Visit No visit Visit No visit Visit No visit Visit No visit % 81% 26% 74% 39% 61% 33% 67% % 47% 55% 45% 72% 28% 61% 39% % 29% 76% 24% 88% 12% 85% 15% 37

47 The demographic characteristics of children enrolled in FY are presented in Table 5-3. Children with longer enrollment periods were more likely to have had a dental visit at some point during their enrollment. Children ages 7-12 at the time they enrolled were the most likely to have had a dental visit. Table 5-3. Demographics of children in hawk-i and Medicaid Characteristic Medicaid IHS John Deere Wellmark Visit No visit Visit No visit Visit No visit Visit No visit Gender Female 39% 61% 37% 63% 52% 48% 46% 54% Male 38% 62% 37% 63% 49% 51% 44% 56% Age at enrollment % 77% 21% 80% 27% 73% 19% 81% % 54% 46% 54% 61% 39% 52% 48% % 53% 45% 55% 63% 37% 56% 44% % 65% 35% 65% 48% 52% 48% 53% FPL <50% 37% 63% n/a n/a n/a n/a n/a n/a % 41% 59% n/a n/a n/a n/a n/a n/a 101% 133% 42% 58% n/a n/a n/a n/a n/a n/a % n/a n/a 37% 63% 48% 52% 43% 57% % n/a n/a 41% 59% 53% 47% 49% 51% Race/ethnicity White 43% 57% 38% 62% 52% 48% 46% 54% African American 36% 64% 46% 54% 42% 58% 50% 50% Latino 32% 68% 33% 68% 44% 56% 30% 70% Other/unspecified 31% 69% 35% 65% 49% 51% 43% 57% Urban/rural Metro 39% 61% 38% 62% 50% 50% 41% 59% Urban adjacent to metro 39% 61% 35% 65% 59% 41% 44% 56% Urban not-adjacent to metro 38% 62% 36% 64% 46% 54% 47% 53% Rural 39% 61% 38% 62% 51% 50% 44% 56% Table 5-4 shows that among the children with a dental visit in FY , children in John Deere were most likely to receive a dental visit during the first three after enrollment, while enrollees in IHS and Medicaid were least likely to have had a visit in the first three. Table 5-4. Time after enrollment in which children received first dental visit by plan IHS John Deere Wellmark Medicaid Time of first dental visit % Cum. % % Cum. % % Cum. % % Cum. % % 37% 50% 50% 46% 46% 33% 33% % 88% 45% 95% 47% 93% 55% 88% % 100% 5% 100% 7% 100% 13% 100% Table 5-5 presents the distribution of the cost of dental care by the time period when children had their first dental visit, by the type of dental service (e.g., preventive/diagnostic) and by the health plan in which they 38

48 were enrolled. A higher percentage of children who received their initial dental visit in the first three after enrollment had dental costs for their first episode of care over $500. There was little difference in the proportion with high preventive/diagnostic or routine restorative costs, however, the difference was most apparent in the complex restorative care. It was also most apparent for children in Medicaid and IHS. Table 5-5. Dental costs by period of enrollment and plan Medicaid IHS John Deere Wellmark Time of first dental visit after enrollment Time of first dental visit after enrollment Time of first dental visit after enrollment Time of first dental visit after enrollment Cost Categories All Dental Services $ % 21% 20% 23% 25% 27% 26% 27% 27% 24% 30% 34% $ % 32% 32% 31% 37% 37% 38% 40% 51% 31% 32% 33% $ % 27% 28% 28% 24% 24% 26% 24% 13% 28% 24% 23% $ % 20% 21% 18% 15% 12% 10% 9% 9% 18% 13% 10% Preventive/Diagnostic $0 6% 5% 6% 8% 5% 3% 5% 2% 4% 9% 8% 6% $ % 32% 31% 32% 32% 35% 36% 34% 36% 34% 39% 46% $ % 50% 48% 45% 52% 49% 53% 58% 57% 43% 41% 39% $ % 13% 16% 15% 10% 12% 5% 5% 4% 14% 12% 9% $501+ 0% 0% 0% 0% 0% 1% 0% 0% 0% 0% 0% 0% Routine RestorativeServices $0 55% 61% 61% 64% 68% 70% 66% 69% 72% 61% 69% 68% $ % 8% 5% 9% 9% 8% 14% 14% 15% 9% 8% 8% $ % 10% 10% 10% 8% 8% 9% 8% 4% 11% 10% 11% $ % 14% 14% 12% 11% 11% 9% 8% 8% 13% 10% 11% $ % 6% 7% 6% 4% 3% 2% 1% 2% 7% 4% 3% Complex Restorative Services $0 75% 81% 82% 80% 86% 90% 83% 87% 89% 83% 85% 89% $ % 6% 5% 6% 3% 4% 6% 4% 5% 5% 6% 5% $ % 4% 3% 4% 4% 2% 4% 5% 2% 5% 3% 2% $ % 5% 5% 3% 5% 3% 5% 3% 4% 4% 4% 2% $501+ 8% 5% 4% 4% 3% 1% 2% 1% 1% 3% 3% 2% 39

49 Most children (95%) who received any dental care received a preventive/diagnostic service, however, relatively few received over $200 worth of these services. Those receiving a dental visit in the first three were most likely to not have received any preventive services. Regarding routine restorative services, between one-half and two-thirds of children did not receive any routine restorative care during their first episode of care (Table 5-5). Between 75% and 90% of children did not receive any complex restorative care. The total costs spent for new enrollees during the initial episode of care, as well as the average cost for each new enrollee for this initial episode, is shown in Table 5-6. While the average cost of the initial episode of care for new enrollees who made a visit was lower for John Deere than IHS, when averaged across all new enrollees, the cost was higher for John Deere than IHS, since John Deere had a higher percentage of new enrollees who made a dental visit. Costs for Medicaid enrollees were highest for all plans. Table 5-6. Costs of care for the initial episode of care for new enrollees by plan Medicaid IHS John Deere Wellmark Total costs $6,858,616 $544,065 $509,555 $1,018,140 Costs per new enrollee $145 $105 $113 $127 Costs per new enrollee with a visit $376 $285 $223 $284 The distribution of total costs by service category and health plan is shown in Figure 5-1. John Deere had the highest proportion of costs being spent on preventive and diagnostic services. Medicaid had the highest proportion of all costs being used for complex restorative care. Almost one in four dollars spent on dental care during this initial episode of care was used for care such as root canals, extractions and crowns. Figure 5-1. Percentage of dental costs by type of service and health plan for initial episode of care Table 5-7 shows the average dental costs by time of first dental visit after enrollment by health plan. Medicaid had the highest average costs for the first episode of care, using the costs from the ADA fee survey, regardless of how soon they made their first dental visit. The costs for children who visited in the first three after enrollment were higher for all plans. Among hawk-i plans, John Deere had the lowest average costs for the first dental episode, while average costs for children in IHS and Wellmark were very similar. 40

50 Table 5-7. Average dental costs for the first episode of care, by time of the first dental visit after enrollment and health plan Time of first dental visit after enrollment Medicaid IHS John Deere Wellmark 1-3 $433 $312 $236 $ $346 $274 $212 $ $356 $243 $193 $216 All children $376 $285 $223 $284 Table 5-8 provides the distribution of the per-member costs for the initial-four-month episode of care for children in hawk-i and Medicaid. Average costs were higher for children in Medicaid and for those who had their first dental visit in the first three. The distributions are skewed, with half of the children having costs of $200 or less and a few having very high costs of over $8,000 in Medicaid and $3,800 in Wellmark. Table 5-8. Distribution of dental costs for the first episode of care, by time of first visit after enrollment and health plan Medicaid IHS John Deere Wellmark Time of first dental visit after enrollment Time of first dental visit after enrollment Time of first dental visit after enrollment Time of first dental visit after enrollment All dental care (mean) $433 $346 $356 $312 $274 $243 $236 $212 $193 $319 $259 $ th percentile $113 $105 $106 $101 $101 $91 $93 $91 $94 $101 $87 $85 50 th percentile $233 $189 $196 $179 $127 $127 $142 $127 $124 $178 $131 $ th percentile $540 $424 $439 $370 $327 $360 $274 $241 $194 $370 $304 $ th percentile $1459 $1177 $1168 $1085 $869 $756 $711 $652 $709 $1068 $875 $736 Maximum $8139 $8366 $5428 $3676 $3111 $2647 $2597 $1417 $943 $3801 $2786 $1365 Ordinary Least Squares (OLS) Regression modeling was used to determine the factors associated with the costs of all dental care during the initial episode of care in the program using services and costs based on the ADA fee survey. After controlling for age, gender, race, and urban/rural location, the cost of care for the initial episode of dental treatment was related to the plan in which children were enrolled and the time in which they received their first dental visit. All hawk-i plans had lower total dental costs for the initial episode of dental treatment per person with a visit than did Medicaid. Children in John Deere had the lowest costs, IHS the second lowest, and Wellmark spent the most. Children who received a visit in the first three after enrolling had the highest costs in all four plans. (See Table C-4 in Appendix C for regression results for cost analysis). 41

51

52 Chapter 6. Programmatic dental costs for the hawk-i program The primary research questions regarding the costs of dental care for children in hawk-i were: 1. What was total amount of money paid for dental care by the managed care plans in a given year? 2. What were the per-member/per-month costs for the managed care plans for dental care? Children included in this study: All children who were enrolled in hawk-i and Medicaid during fiscal year (FY) Introduction Dental reimbursement rates are generally included in a bundled rate that includes all required services to be provided by managed care plans. It is unclear whether dental care is typically a costly component of the entire package or an area in which the managed care plans can generate a profit. The purpose of this study was to evaluate the costs of dental care for all children enrolled in hawk-i and Medicaid in FY 2003, to determine how much the plans spent for dental care, and to translate these expenses into pm/pm costs for the health plans. Methods Data Iowa hawk-i and Medicaid enrollment and claims/encounter data for all children ages 1 to 18 for fiscal year (FY) 2003 were used in these analyses. Determination of dental costs Cost data for each individual who had a dental visit was collected from the claims encounter data. Each dental claim was classified by CDT codes into one of four categories of dental procedures (as defined previously), and a cost was associated with each procedure code based on the 2003 American Dental Association (ADA) survey of dental fees for the West North Central region of the United States. 15 The ADA survey fees were used to approximate the Usual and Customary private practice fees in Iowa. All CDT codes no longer in use in 2003 were mapped onto appropriate 2003 codes to ensure complete collection of cost data. The ADA fee schedule was chosen because it represents a commonly accepted metric of costs that can Survey of Dental Fees Chicago, IL: American Dental Association. 43

53 be used to compare resource utilization across different health plan settings, as in this study. The amount submitted for reimbursement for each CDT code and the amount actually paid by the plan were also collected from the claim/encounter record. Analysis plan The total cost of dental services as calculated using ADA fees, the dentist charges and the amount paid by the plan were tabulated by plan. The total number of of enrollment was also tabulated for each month. A ratio of costs on a pm/pm basis was then calculated. Results The total dollars spent on children s dental service in FY 2003 is shown in Table 6-1. Note the difference in the estimated value of the dental services based on the ADA survey fees and what was actually paid by the plans. Costs based on the ADA survey are a conservative estimate of dental costs, since the ADA survey only includes fees for the majority of the most common dental procedures. The paid claims may include more services than in the ADA fee total. Table 6-1. Dental costs by service category for all children in hawk-i and Medicaid, FY 2003 Medicaid* IHS John Deere Wellmark Type of Service Cost basis All Services ADA Fees $14,399,950 $750,977 $686,435 $1,045,992 Paid Claims $8,776,731 $600,555 $638,066 $860,959 Diagnostic ADA Fees $2,821,608 $156,544 $215,684 Paid Claims $1,579,977 $118,939 $177,375 Preventive ADA Fees $2,637,243 $184,176 $252,751 Paid Claims $1, $147,515 $208,947 Routine restorative ADA Fees $4,538,010 $219,922 $336,464 Paid Claims $2,524,895 $162,427 $259,384 Complex restorative ADA Fees $2,873,073 $109,899 $130,785 Paid Claims $1,760,715 $85,658 $108,971 Other ADA Fees $1,530,016 $80,436 $110,308 Paid Claims $1,292,939 $86,016 $106,281 *Medicaid costs did not include $2 million spent on orthodontics (not covered by other plans) Amount equals dentist s charges ADA-related costs not available for John Deere The Medicaid program spent by far the most on dental services almost $9 million was paid for services valued at approximately $14 million. The highest amount in all plans was for routine restorative care. The difference between what was paid and the ADA survey fees could be due to several factors. For hawk-i: (1) fees paid by the plans could be lower than the ADA survey fees, (2) children may have received non-covered services (other than orthodontics which were excluded from this analysis) that were not paid for by the plans), or (3) children may have received services in excess of their annual maximum that were not reimbursed by the plans. For Medicaid, since there are virtually no non-covered services for children and no 44

54 annual maximum, the difference is primarily due to reimbursement rates that are lower than the ADA survey fees. Table 6-2 shows the pm/pm dental costs based on ADA survey fees and claims paid by the plans. John Deere paid the highest total amount on a pm/pm basis over $13. IHS paid a little over $12, and Wellmark, just over $11. Medicaid paid about half as much pm/pm at just under $6. Table 6-2. Dental costs pm/pm by service category for all children in hawk-i and Medicaid, FY 2003 Medicaid* IHS John Deere Wellmark Type of Service Cost basis All Services ADA Fees $9.52 $15.40 n/a $13.51 Dentist Charges n/a $14.43 $14.24 $11.65 Paid Claims $5.80 $12.32 $13.23 $11.12 Diagnostic ADA Fees $1.87 $3.21 n/a $2.79 Dentist Charges n/a $2.85 n/a $2.33 Paid Claims $1.04 $2.44 n/a $2.29 Preventive ADA Fees $1.74 $3.78 n/a $3.26 Dentist Charges n/a $3.38 n/a $2.74 Paid Claims $1.07 $3.02 n/a $2.70 Routine restorative ADA Fees $3.00 $4.51 n/a $4.35 Dentist Charges n/a $3.90 n/a $3.57 Paid Claims $1.67 $3.33 n/a $3.35 Complex restorative ADA Fees $1.90 $2.25 n/a $1.69 Dentist Charge n/a $2.09 n/a $1.51 Paid Claims $1.16 $1.76 n/a $1.41 Other ADA Fees $1.01 $1.65 n/a $1.42 Dentist Charge n/a $2.20 n/a $1.50 Paid Claims $0.86 $1.76 n/a $1.37 *Medicaid costs did not include $2 million spent on orthodontics (not covered by other plans) Amount equals dentist s charges-ada related costs not available for John Deere 45

55

56 Chapter 7. Discussion and policy implications The policy implications of the results for the hawk-i program (and Medicaid where relevant) for each of the different analyses in this report are discussed below. 1) Access to dental services As in previous analyses, having hawk-i insurance has been found to greatly improve access to dental care. The percentage of children who reported unmet need and delays for dental care declined significantly in their first year in the hawk-i program as compared to the year before they entered the program. There was a slight yet significant increase in the percentage with a regular source of dental care, and a significant increase in the percentage with a dental visit and a preventive dental visit in the previous year. There were few significant differences in the trends of these measures between FY 2001 and 2003, except for the decline in the percentage receiving a check-up in the first year of hawk-i. This trend should be monitored given the elimination of IHS from the program and the addition of Delta Dental as an option in some counties. Perceived need for dental care also declined slightly in both the baseline and follow-up measures; however, perceived need for dental care varies from need for a check-up to need for emergency care and should thus be interpreted carefully. 2) Utilization of dental services It is clear from the findings in Chapter 3 that reported dental utilization rates for children in public programs could be greatly affected by the methodology used to determine the rates. The CMS methodology, which includes all children enrolled for any length of time during the year, produced the lowest dental utilization rates, as well as the most similar rates for children in Medicaid and hawk-i. The inclusion of only newly enrolled Medicaid and hawk-i children also produced lower utilization rates than when all children were included. The HEDIS and FTE approaches produced higher rates overall, and resulted in a significant differential whereby rates for children in hawk-i were much higher than for those in Medicaid. The FTE approach produced the highest rates, in part due to the potential for having more than one visit counted in the numerator. For example, two children each enrolled in Medicaid for six during the year are considered one FTE (the equivalent of one person-year). If each then had a dental visit, they would both be counted in the numerator of the rate calculation (i.e., a score of two). In contrast, a child can only be counted once in the numerator for each of the other methodologies. Longer enrollment intervals were associated with higher utilization rates. This observation is not surprising since the longer a person is enrolled in the program, the more opportunity he or she will have to identify the need for dental care, locate a participating provider, schedule an appointment and obtain services. 47

57 This is one of the first studies to report on the utilization of dental services for children in Medicaid and hawk-i. Children in Iowa s hawk-i program had rates similar to those in Medicaid when averaged across all children, but had higher rates as the length of enrollment in the program increased. The higher hawk-i dental utilization rates with longer enrollment intervals may be due to a number of reasons. First, children in hawk-i have a higher socio-economic status, and thus their parents may have a greater awareness of the need for dental care. Dentist participation in hawk-i may also be higher than in Medicaid, although no data from Iowa are available yet. The utilization rates derived from the various approaches used in this study are lower than commonly reported national annual dental utilization rates based on survey data which typically are about 75-80% for the general population. However, our rates generally are in the range of rates derived from administrative data for public dental programs for children. The dental utilization rate for publicly insured children in the National Health Interview Survey was also much higher (68%). Differences between rates determined by surveys and administrative data could occur for a number of reasons, including: (a) social response bias, with people responding more positively in a survey, (b) enrollees receiving dental services during periods when not enrolled in Medicaid (which would be missed by administrative data), or (c) providers not submitting all claims for services provided to Medicaid/hawk-i enrollees because they feel that low reimbursement levels are not worth the hassle of filing claims. Ideally, dental utilization rates for children in Medicaid and hawk-i would be calculated using an enrollment period of sufficient length to allow enrollees a reasonable opportunity to identify the need for the service, attempt to access the service, schedule an appointment, and ultimately, receive services during one or more visits. Typically, one would expect this process to occur within 6-8. This raises the question of which methodology is most appropriate for evaluating utilization in state Medicaid and SCHIP programs, or in commercial health plans. It is our belief that an approach that allows for a more consistent comparison between two entities (i.e., states or plans) by controlling for the length of enrollment (in a manner analogous to HEDIS, although not necessarily using the HEDIS of enrollment criteria) may provide a better indicator of access to dental care than currently used methods. An alternative approach to both the CMS and HEDIS methods would be to include all children enrolled for an interval that is somewhere between the two extremes of 1 month (CMS) and (HEDIS). Perhaps, for example, those enrolled for at least 8 could be included when calculating dental utilization rates. As an aside, the same might be advisable for other types of services, such as well-child visits, where there is an expectation that children will have regular, periodic visits within a given year. This would allow for a reasonable time period in which children could be expected to receive services, yet would include more enrolled children in the analysis than the current HEDIS approach. Determination of this type of rate should be relatively easy for states or health plans using administrative data. 48

58 3) Time to first dental visit This is one of the first studies to evaluate the length of time before a child first receives a dental visit after joining a public insurance program for the first time. How soon a child receives dental care is an important indicator of access to care for several reasons. First, it may indicate how clear the process is for seeking dental services from the enrollee s perspective. Second, it may indicate the ease with which a parent is able to find a dentist who participates in the program. Third, an earlier visit may indicate a higher parental awareness and knowledge about the need for annual or biannual preventive dental care for their child, and fourth, it may indicate that other cues about the need for annual or biannual preventive dental care are functioning properly, such as the health education and anticipatory guidance that is supposed to be provided at annual well-child visits. A child s total length of enrollment in a program, an important factor in how well they learn the nuances of seeking care in a program, varied between hawk-i and Medicaid. Although the average length of enrollment during the three-year study was similar in all plans, the distributions were different. Children in Medicaid were more likely to be enrolled for both shorter and longer periods of time, averaging about 11 during the three years, producing an average similar to that for children in hawk-i. Another demographic difference in the populations of relevance is that children enrolled in John Deere were considerably more likely to be located in metro areas than children in the other plans, and children in Wellmark were more likely to be located in non-metro areas. Using the survival analytic techniques, which adjust the calculations for children who were enrolled for only a portion of the three-year period, indicated that about one-quarter of newly enrolled children had received a dental visit after 6, 44% after one year and 75% after three years in the program. Newly enrolled children in John Deere had the highest adjusted utilization rates. Children in IHS had the lowest rates after 6, while Medicaid had the lowest rates after three years. These differences were found to be true even after adjusting for the other factors such as age, gender and location in the survival analysis. 4) Cost of first episode of dental care This is one of the first studies to evaluate costs of dental care for children newly enrolled in public insurance programs. It is also one of the first to try and define costs for the initial episode of care and to determine if how soon an enrollee seeks care after beginning the program affects the amount of treatment received. In general, the cost of the care received is a proxy for need, however, this is an imperfect measure of need since many will never receive all treatment identified in an initial dental check-up. It also cannot assess need in children who never receive a check-up. Differences in costs for the first episode of care were found for children by the health plan in which they were enrolled. Costs, based on the ADA survey fees, were found to be highest for children in Medicaid ($376 per initial episode) as compared to all hawk-i plans. John Deere ($223 per initial episode) had the lowest costs for this initial episode of care among hawk-i plans, while costs for Wellmark ($284 per initial episode) and IHS ($285 per initial episode) were very similar. 49

59 There were significant differences in costs related to how soon children received their first visit as well. Children receiving care sooner had higher average costs for the initial episode of care than children receiving care later. This tends to indicate some degree of pent-up demand among the population, but the relative amount compared to the population as a whole is unclear. There will always be some children with higher needs, but our results do indicate that some of them tend to seek care sooner after enrolling. 5) Per member/per month costs of dental care These analyses were an attempt to evaluate the total expenditures for dental care by the managed care plans in a year, and the pm/pm costs after adjusting for enrollment in the plans. The plans are paid approximately $15 pm/pm for dental services. This amount, however, is bundled in one total payment for all services that a program is required to provide under contract with the Department of Human Services. Based on the paid claims, the actual amount paid by the plans ranged from about $11 for Wellmark to about $13.25 for John Deere. Because of a number of possible factors, including reimbursement rates being lower than the ADA survey fees, children reaching annual maximum, or the provision of uncovered services, the paid amount was $2 to $3 less than the costs based on the ADA survey fees for children in Wellmark and IHS respectively. It was also about $2 to $3 less than the $15 pm/pm for which they are paid to provide dental care. Medicaid reimbursed costs were about $3.70 less than the ADA survey fees. The difference between the amount paid to the managed care plans and the amount they pay in claims to the dentists covers the administrative costs of the plans to operate the program, as well as operating profit. The appropriateness of the amount that should be available for the administrative expenses and profit is a matter for policymakers to determine. 6) Conclusions Overall, children in John Deere, the open access plan paying full charges, were more likely to receive their initial visit sooner than children in the other plans, yet had lower costs for the initial episode of care. The higher utilization rates for new enrollees in John Deere may be related to several factors. John Deere operates an open access dental panel, where children can receive dental care from any dentist who will see them. The dentist is not required to submit paperwork to become a part of the provider network. Second, John Deere pays the dentists their full private practice charge. While it may take some dentists longer to understand how to participate in John Deere, since they are not signing up with a network, once they figure this out, being paid their full charges is an obvious positive incentive to treat children in this program. Third, John Deere is located primarily in metro areas. This may provide better access to dentists due to the availability/distribution of dentists in Iowa than some of the more rural areas of the state. As becomes more apparent in the cost section, the new enrollees in John Deere may also have fewer treatment needs (if received service is related to need) since their average costs for the first episode of care are much lower. Children in John Deere also had lower costs per enrolled child for the first episode of care. The lower costs for the initial episode could be related to several factors: (a) John Deere is covering a population with lower needs in general or (b) John Deere is providing better access to care for children, allowing a 50

60 higher proportion of children with less severe needs to receive care (diluting the average cost for the initial episode of care). The greater access hypothesis is supported by other analyses in a report to the State of Iowa Department of Human Services indicating that the pm/pm paid dental claims were highest for John Deere ($13.23) compared to the IHS closed panel plan ($12.32) or the Wellmark indemnity plan ($11.12). Even at $13.23 paid out, the dental care portion of the capitation rate paid to the plans for providing services to hawk-i-enrolled children is approximately $15.50, allowing over $2 for administrative expenses/profit per child per month for all plans, even John Deere that is paying full dentist charges. John Deere seems to have found a dental delivery system model that provides improved access to dental care for children in a public insurance program, easier participation on the part of dentists at reimbursement rates equal to their usual charges, and yet is able to recoup administrative expenses in the range of $2-3 per child per month. Further investigation into using a similar model for other programs should be evaluated. 51

61 52

62 Chapter 8. Research limitations As with all research, there are limitations to the interpretation of the results and other issues that need to be considered when trying to generalize these analyses to broader issues of interest. The following is a discussion of some of these issues. 1) Survey evaluating access to dental care As with all survey-based research, there are limitations to the interpretation of the results. Although the survey was pilot tested prior to implementation, differing interpretation of questions may influence the response to some items. In addition, although the overall response rates were reasonable, the inclusion of only those children for whom we had both baseline and follow-up information may have introduced a bias. Non-response bias tests indicated that the bias should have been minimal and the results conservative, but this type of bias is difficult to evaluate unequivocally. The survey is based on self-report and does not evaluate issues such as health status from a clinical perspective. Additionally, since no control group was used, some secular trends could have affected care for all children. However, we are unaware of any such trends in Iowa that would have had this effect on access and health status. Although these results are encouraging for Iowa s S-SCHIP program, they may or may not be applicable to all SCHIP programs. There may be differences related to the model of SCHIP program (i.e., M-SCHIP vs. S-SCHIP), the dental delivery systems used, and the demographic characteristics and previous dental experiences of the enrollees. 2) Use of administrative data generally for conducting research All studies using administrative data have limitations, in large part because the data were designed and submitted for reimbursement purposes rather than for research. The accuracy is a function of the provider submitting the correct procedure codes for billing, as well as correct processing of the claim throughout the billing process. The results reported here also are derived from data on a single state. The design and administration of Medicaid and SCHIP programs, as well as the characteristics of enrollees, vary significantly across states, and thus the utilization rates observed in this study should not be construed as typical of utilization rates for programs in other states. 3) Cost analyses Since the health plans each pay dentists in a slightly different manner, the costs needed to be standardized for consistent comparison. There are no standardized fee schedules available, thus the fees from the ADA survey of dentists for the region of the country that includes Iowa were used as a proxy for usual and customary fees in Iowa. The ADA survey did not ask about every CDT code, and thus the amount of spending reported for each plan associated with the ADA fees are a slightly lower estimate, although they do include the most commonly used procedures. 53

63

64 Appendix A. Definition of codes defining type of dental procedures Table A-1. Codes defining type of dental procedures Type of code Code Description Diagnostic 0120 Periodic oral evaluation 0150 Comprehensive oral evaluation 0210 Intraoral-complete series bitewings 0220 Intraoral-periapical first film 0230 Intraoral-periapical each additional film 0240 Intraoral-occlusal film 0270 Bitewing-single film 0272 Bitewing-two films 0274 Bitewing-four films 0330 Panoramic film Preventive 1110 Prophylaxis-adult 1120 Prophylaxis-child 1201 Topical fluoride with prophylaxis-child 1203 Topical fluoride w/o prophylaxis-child 1205 Topical fluoride with prophylaxis-adult 1204 Topical fluoride w/o prophylaxis-adult 1351 Sealant-per tooth 1330 Oral hygiene instructions 4355 Full mouth debridement to enable exam Restorative 2140 Amalgam- 1 surface, primary or permanent 2150 Amalgam-2 surface, primary or permanent 2160 Amalgam-3 surface, primary or permanent 2161 Amalgam-4-6 surface, primary or permanent 2330 Resin-base composite, 1 surface anterior 2331 Resin-base composite, 2 surface anterior 2332 Resin-base composite, 3 surface anterior 2335 Resin-base composite, 4 or more surface anterior Complex restorative 3220 Therapeutic pulpotomy 2930 Prefab stainless steel crown-primary tooth 3310 Root canal-anterior 3320 Root canal-bicuspid 3330 Root canal-molar 2740 Crown-porcelain/ceramic substrate 2750 Crown-porcelain fused to high noble metal 2751 Crown-porcelain fused to base metal 2752 Crown-porcelain fused to noble metal 2790 Crown-full cast high noble metal 2791 Crown-full cast base metal 55

65 Type of code Code Description Complex restorative (cont.) 2792 Crown-full cast noble metal 1515 Space maintainer-fixed bilateral 1520 Space maintainer-removable unilateral 1525 Space maintainer-removable-bilateral 9420 Hospital call 4210 Gingivectomy or gingivoplasty more than 3 contiguous teeth per quadrant 4211 Gingivectomy or gingivoplasty more than 1-3 teeth per quadrant 4240 Gingival flap procedure more than 3 contiguous teeth per quadrant 4241 Gingival flap procedure 1-3 teeth per quadrant 4245 Apically positioned flap 4249 Clinical crown lengthening-hard tissue 4260 Osseous surgery more than 3 contiguous teeth per quadrant 4261 Osseous surgery 1-3 contiguous teeth per quadrant 4263 Bone replacement graft-first site in quadrant 4264 Bone replacement graft-each additional site in quadrant 4265 Biologic materials to aid in soft and osseous tissue regeneration 4266 Guided tissue regeneration-resorbable barrier 4267 Guided tissue regeneration-non-resorbable barrier 4268 Surgical revision procedure 4270 Pedicle soft tissue graft 4271 Free soft tissue graft 4273 Subepithelial connective tissue graft 4274 Distal or proximal wedge procedure 4275 Soft tissue allograft 4276 Combined connective tissue and double pedicle graft 4320 Provisional splinting-intracoronal 4321 Provisional splinting-extracoronal 4341 Periodontal scaling and root planing-more than 3 teeth per quadrant 4342 Periodontal scaling and root planning-1-3 teeth per qudrant 4355 Full mouth debridement to enable comprehensive evaluation and diagnosis 56

66 Appendix B. Tables and graphs related to time to a dental visit Table B-1. Bivariate relationships between covariates and survival time to a dental visit Variable Level N (%) Time to any dental visit Time to a preventive visit 6-month survival estimate 1-year survival estimate 3-year survival estimate 6-month survival estimate 1-year survival estimate 3-year survival estimate All children Health plan Medicaid 71,313 (74%) IHS 7,252 (8%) John Deere 6,610 (7%) Wellmark 11,529 (12%) Age Ranges ,309 (21%) ,614 (24%) ,685 (28%) ,096 (27%) Gender Male 47,480 (49%) Female 49,224 (51%) Race White 57,030 (59%) Race other than White 9,933 (10%) Race unspecified 29,741 (31%) Location Metro 49,016 (51%) Urban adjacent to metro 19,852 (21%) Urban not-adjacent to metro 20,487 (21%) Rural 7,349 (8%) Note: All bivariate relationships were statistically significantly different at the p<0.001 level due to the large sample size 57

67 Survival and hazard curves are shown in Figure B 1. These indicate the proportion of children with a dental visit relative to the number of they have been in the program. The survival curve indicates the proportion of the population who were without a dental visit while the hazard curve shows the inverse (i.e., the proportion who did receive a dental visit over time). Figure B-1. Hazard curves showing the proportion of children with any dental visit as compared to their length of time in the program by health plan 58