Background Report for the Request for Public Comment on Initial, Recommended Core Set of Children's Healthcare Quality Measures for Voluntary Use by Medicaid and CHIP Programs
Screening to identify health conditions for early intervention with effective treatments is a primary purpose of preventive (well-care) visits. Three screening services were identified as important for children and adolescents: weight assessment documentation, use of standardized screening tools for potential delays in social and emotional development among young children, and Chlamydia screening for females ages 16-20.
- Weight assessment for children/adolescents. This new NCQA measure assesses the number of children ages 2-18 who had at least one outpatient visit with a primary care provider or an obstetrician-gynecologist (OB/GYN) during the measurement year (and was continuously enrolled in the measurement year) whose medical record documents that a Body Mass Index (BMI) assessment was performed during a visit.
This is a process measure with an evidence grade of D. SNAC voting resulted in a rank of 4 for this measure, although members noted that a better measure would actually provide the BMI results. In 2005, the USPSTF graded screening by BMI as "I" for insufficient evidence because at the time there were almost no studies demonstrating that followup interventions would be effective in reducing weight for overweight children.e,33 NCQA specifications and current reporting levels are for health plans. Specifications are available at the health care provider level. NCQA data sources are administrative and medical records. Because 2009 was the first year of reporting on this measure, performance data and information on the number of plans and States with valid reporting rates are not publicly available.
Having a documented BMI is considered by many to be an important first step in preventing overweight and obesity in children. All 2- to 18-year-old children enrolled in Medicaid and CHIP could be eligible for this measure. Publicly covered children are more likely to be overweight (using the 85th percentile).34 According to 2006 AHRQ Medical Expenditure Panel Survey (MEPS) data, 36.9 percent of publicly insured children ages 6-11 are overweight, compared to 30.5 percent of uninsured and 20 percent of privately insured children. Among 12-17 year-olds, 21.5 percent of Medicaid/CHIP children are overweight, compared to 12.4 percent of uninsured children and 11.8 percent of privately insured children.
Because the NCQA measure requires medical chart review, State Medicaid and CHIP programs will face challenges in reporting on these data. A performance measurement code for BMI documentation is available for adults but not for children. As noted above, almost half of Medicaid- and CHIP-enrolled children are not in managed care plans, and an alternative method will be needed for those States and for other publicly insured children in primary care case management (PCCM) and fee-for-service (FFS) programs. In addition, SNAC members professed a preference for measuring quality using outcomes data (e.g., the actual BMIs) tracked over time.
- Screening using standardized screening tools for potential delays in social and emotional development. This measure assesses the extent to which children at various young ages from 0-36 months were screened for social and emotional development with a standardized, documented tool or set of tools.
This is a process measure with an evidence of grade B. SNAC voting resulted in a rank of 11 for this measure. Selected State Medicaid and CHIP programs with Assuring Better Child Health and Development (ABCD) initiative grants are using this measure,35 and staff from additional programs are being trained on the measure as part of a training academy.36 The data sources are administrative, chart review, and other data (e.g., cards filled out by providers). Data are collected in racial/ethnic populations other than non-Hispanic white, but those data were not included in the reports received by the SNAC.
Evidence is scarce that screening for social or emotional problems in young children leads to better health outcomes, primarily due to the paucity of rigorously done treatment intervention studies for these conditions. The conditions may emerge gradually over months or years, and endpoints for successful outcomes may be difficult to define. However, well-done surveys have shown that parents are eager to discuss their child's social and emotional development with doctors and other health care providers. Studies also demonstrate that the use of standardized tools for screening is more likely to result in accurate identification of a child with such problems.
An estimated 9.5 million Medicaid- and CHIP-enrolled preschool children are eligible for screening (Table 1). In the United States, 17 percent of children (12 million children) were found to have a behavioral disability such as autism, mental retardation, or attention-deficit/hyperactivity disorder. Medicaid serves more than 25 percent of all children in the United Sates (and more than half of all poor and low-income children). Children from poor families are at greater risk than those from non-poor families for poorer outcomes, including those related to mental development.37 The 2007 National Survey of Children's Health (NSCH) found that publicly insured children were 1.9 times as likely as privately insured children (18.3 percent versus 9.7 percent, respectively) to have one or more of six specified learning, developmental, or behavioral conditions.38
Currently, information on performance is only available from several State programs engaged in the ABCD program. In the program, States select practices to test strategies for the delivery of developmental services to young children at risk for or with social or emotional developmental delays. In each State, the average screening rate using an objective screening tool improved considerably after program implementation.35 The greatest increases were in California (two managed care plans) and Minnesota (from 0 percent pre-implementation to 93-94 percent post-implementation for both). Referral rates improved among most States; however, the States experienced difficulty in assuring receipt of followup services.
An advantage of this measure is that claims can be entered to document providers' use of objective screening tools. Some States also provided a cross-walk to match diagnoses to the DSM-IV (Diagnostic and Statistical Manual of Mental Disorders) codes required for claims payment. However, participating ABCD States have used differing codes and denominator populations. Standardization across States will be needed if information is to be valid and comparable across States and the Medicaid and CHIP populations they serve.
- Chlamydia screening for women. This measure assesses the extent to which at least one Chlamydia test is given during the measurement year to sexually activef women ages 16-20 years of age as of December 31 of the measurement year and with no more than one gap in enrollment of up to 45 days during the measurement year.39
This is a process measure with an evidence grade of B. SNAC voting resulted in a ranking of 14 for this measure. Measure specifications are available at health plan and provider levels, and reports are made by health plans participating in Medicaid managed care. In 2008, 129 health plans across 30 States reported valid data to NCQA. The data source is administrative records, an advantage from a State and health plan perspective.
Chlamydia screening is important for adolescents. Chlamydia is a sexually transmitted infection that causes pelvic inflammatory disease, a condition that can result in sterility. According to the CDC's 2007 Youth Risk Behavior Surveillance System survey, 47.8 percent of students (9th-12th grade) have ever had sexual intercourse.40 Further, 7.1 percent did so before age 13; 14.9 percent did so with four or more individuals; and 35.0 percent did so with at least one person in the last 3 months. Of the 35.0 percent currently sexually active students, 61.5 percent said they had used a condom during their last sexual intercourse. In 2007, 13,848 cases of Chlamydia were reported among 5-14 year-olds and 779,280 cases among 15-24 year-olds.31 (Data specific to Medicaid and CHIP enrollees are not available.)
In 2007, the national mean performance rate across Medicaid managed care health plans reporting to NCQA with valid data was 48.56 percent. The average rate for plans at the 10th percentile was 31.62 percent, and at the 90th percentile the average rate was 65.26 percent. Medicaid health plan performance was, however, better than performance among commercial health plans (34 percent).
Well-care visits (WCVs)
Current American Academy of Pediatrics (AAP) guidelines suggest that all children receive a WCV at a specific periodicity depending on age because the visits are the gateway to immunizations and early identification of problems, and they provide opportunities to discuss developmental issues with parents and deliver evidence-based and other recommended specific preventive services. The three well-child visit measures currently specified by NCQA are recommended.
10-12. Well-child Visits (WCV)—three NCQA measures: 1) WCVs in the first 15 months of life; 2) WCVs in the 3rd, 4th, 5th, and 6th years of life; 3) Adolescent WCVs. These measures assess, for each age group, the number of children who received a well-child or preventive care visit from a primary care practitioner (including, for adolescents, an obstetrician-gynecologist) during the measurement year.
For the youngest group, children who turn 15 months during the measurement year and are continuously enrolled from 31 days after birth to 15 months of age are in the measure denominator. The number of visits is counted (0, 1, 2, 3, 4, 5, 6 or more visits) for this age group. For 3-, 4-, 5- and 6-year-olds, the criterion is at least one well-child visit with a primary care practitioner during the measurement year. Children must be ages 3, 4, 5, or 6 at the end of the measurement year and be continuously enrolled for the measurement year. For adolescents, those who are 12-21 years of age must have at least one comprehensive well-care visit during the measurement year to meet the criterion. Adolescents must also be enrolled continuously for the measurement year.
These are process measures with an evidence grade of B. SNAC scoring resulted in a rank of 5 for this measure. Measure specifications are available at the health plan and provider levels and are reported at the health plan level. The measures are currently reported to NCQA by about 160 MMC health plans across 32 States. (The level of reporting does not vary by age group.) The data sources are administrative and medical records.
In 2007, the national mean performance across health plans for WCV in the first 15 months of life was 52.9 percent for six or more visits, and the 10th and 90th percentiles were 28.9 percent and 73.7 percent, respectively. Only 5.68 percent of these young children had no visits, according to data reported by NCQA. For annual WCVs for 3-6 year olds, the national mean was 65.11 percent, and the 10th and 90th percentiles were 50.9 percent and 78.9 percent, respectively. For adolescent WCVs, the national mean was 41.88 percent, and the 10th and 90th percentiles were 26.2 percent and 56.7 percent, respectively.
While this measure was ranked high by the SNAC, there is increasing interest in measuring the content of well child care visits rather than the mere fact of a visit.41 Measures of specific content are being field-tested by NCQA. The future challenge for State Medicaid and CHIP programs is that the more specific a measure is, the more difficult it may be to obtain from claims data. The relatively low number of States reporting may reflect the limited number of States with wide use of Medicaid managed care plans. In order for the current measure to be applicable across all Medicaid and CHIP States and populations, new specifications will need to be identified and tested. The same will be true of the content-specific measures being tested by NCQA, should they show promise during field testing.
Dental services are a required service for most Medicaid-eligible individuals under the age of 21, as a component of the EPSDT benefit,42 and with the advent of CHIPRA, dental services are a required benefit for CHIP enrollees.
- Total eligibles receiving preventive dental services (EPSDT measure Line 12B). This is an EPSDT measure that assesses the percent of unduplicated children who received a dental preventive service (defined by Healthcare Common Procedure Coding [HCPC] codes D1000-D1999) as a function of the number of children eligible for EPSDT services as shown on line 1 of the CMS-416 form. "Unduplicated" means that each child is counted only once for purposes of this line even if multiple services were received.
This is a process measure for which there was not enough evidence to specify an evidence level grade. The U.S. Preventive Services Task Force concluded that the evidence is insufficient to recommend for or against routine risk assessment of preschool children by primary care clinicians for the prevention of dental disease, but they did recommend that primary care clinicians prescribe oral fluoride supplementation at currently recommended doses to preschool children older than 6 months of age whose primary water source is deficient in fluoride.43 The recommendation about routine risk assessment did not apply specifically to dental providers. SNAC voting resulted in a ranking of 6 for this measure. As one of the EPSDT measures, this measure is reported by CMS in national and State-level reports, based on State self-report data.44
Preventive visits are recommended to provide dental hygiene, fluoride applications, and to identify dental caries and other oral health problems. In 2006, more than two-thirds of low-income children in the United States (69 percent) received dental coverage through Medicaid and CHIP during at least part of the year.45 That number will rise with the CHIPRA requirement. However, according to the national EPSDT report, only 32 percent of eligibles in 2008 (9,920,468 children) received preventive dental services under EPSDT.
This measure, in combination with the EPSDT measure on dental treatment services, was preferred to the NCQA HEDIS measure of annual dental visits because the HEDIS measure is reported by NCQA only for children in managed care, and because the NCQA HEDIS measure assesses the total number of visits annually, rather than the total number of children served, and the nature of their visits (preventive or treatment). Using the EPSDT report would prevent States from having to collect an additional component of dental care from a different source (CMS-416 also reports the number of children receiving any dental treatment, a summation of the preventive and treatment visit data). Medicaid and CHIP officials also raised concerns about variation in EPSDT reporting across States, which could affect the validity and reliability of cross-State comparisons.
e The USPSTF is currently doing an update of their evidence review.
f Two methods identify sexually active women: pharmacy data and claim/encounter data. The organization must use both methods to identify the eligible population; however, a member only needs to be identified in one method to be eligible for the measure.
Management of Acute Conditions
Upper respiratory conditions
Upper respiratory infections (URIs) are among the most common reasons for children's acute care encounters with health care providers. Health care quality is an issue for URIs because URIs present opportunities for the overuse of antibiotics. Most URIs are time-limited and, if viral, cannot be cured with antibiotics.
- Pharyngitis-appropriate testing. This NCQA measure assesses whether a strep test was administered in the 7-day period from 3 days prior through 3 days after the first eligible episode date. An eligible episode is an outpatient visit with a diagnosis of pharyngitis at which an antibiotic was dispensed. The measure counts children ages 2-18 who were continuously enrolled 30 days prior to the episode date through 3 days after the episode date.
The purpose of the measure is to reduce the unnecessary use of antibiotics by ensuring that antibiotics were not given without a diagnosis of strep. Without the presence of a bacterial infection, antibiotics are ineffective in treating pharyngitis.
This is a process measure with an evidence grade of A. SNAC voting resulted in a rank of 18 for this measure. The measure is currently specified at the health plan and provider levels, and it is currently reported at the health plan level. Administrative data are used for reporting. The data source is administrative records. The measure is currently reported to NCQA by 108 MMC health plans in 28 States.
Overuse of antibiotics continues to be a problem, and opportunities for inappropriate use abound. Acute pharyngitis was the leading diagnosis for 6.4 million visits to physician offices and hospital outpatient departments for all children under age 15 in 2006 and 6.2 million visits in 2001.46 Improvement is still needed in this measure. In 2007, the national mean performance among plans reporting to NCQA was 58.65 percent, with 10th and 90th percentiles of 31.75 percent and 77.31 percent, respectively.
- 15. Otitis Media with Effusion (OME)—avoidance of inappropriate use of systemic antimicrobials. This AMA PCPI measure assesses the extent to which children ages 2-12 years presenting with otitis media with effusion (fluid in the middle ear) were not prescribed systemic antimicrobials. The rationale for the measure is the same as that for the measure assessing antibiotic use for pharyngitis.
This is a process measure with an evidence grade of A. SNAC voting resulted in a ranking of 18 for this measure (it tied with the pharyngitis measure). According to the nominator, the measure can be collected using computerized prescription order entry technology. The current data source is administrative records using Current Procedural Terminology (CPT) Category II codes.
As with pharyngitis, opportunities for inappropriate antibiotic use are frequent. About 90 percent of children have OME at some time before school age, most often between the ages of 6 months and 4 years. In the first year of life, more than 50 percent of children will experience OME, increasing to more than 60 percent by age 2 years. Many episodes resolve spontaneously within 3 months, but about 30 percent to 40 percent of children have recurrent OME, and 5 percent to 10 percent of episodes last 1 year or longer.
Performance data for this measure specified by the American Medical Association Physician Consortium for Performance Improvement (AMA PCPI) were not reported because the measure as nominated currently is not yet in use. Availability of data from administrative data sources increases its feasibility for use by Medicaid and CHIP programs.
As noted above, dental services are a required EPSDT service for most Medicaid enrollees under age 21 and are now a requirement under CHIPRA. EPSDT captures access to dental treatment, as well as preventive services, on CMS Form 416 Line 12C.
- Total Early Periodic Screening, Diagnosis, and Treatment (EPSDT) eligibles who received dental treatment services (EPSDT CMS Form 416 Line 12C). This is an EPSDT measure that assesses the percent of unduplicated children who received a dental treatment service (defined by CMS' HCPC codes D2000-D9999) as a function of the number of children eligible for EPSDT services as shown on line 1 of the CMS-416 form. "Unduplicated" means that each child is counted only once for purposes of this line even if multiple services were received.
The measure has an evidence grade of D. States report data to CMS using form CMS-416, based on specifications provided by CMS.47
The need for performance improvement on this measure is difficult to assess, given the denominator used, which does not reflect the number of children in need of dental treatment services. In 2008, 18 percent of eligible children (5,654,499) received dental treatment services44 Roughly one-third of low-income children ages 6-19 have untreated tooth decay, compared with 15 percent of children at or above twice the poverty level.48 Eighty percent of tooth decay is found in 25 percent of children ages 5 to 17, mostly from low-income and other vulnerable groups.
Medicaid and CHIP officials raised concerns about variation in EPSDT reporting across States, which could affect the validity and reliability of cross-State comparisons.
Emergency Department Use
Emergency departments are a critical feature of the U.S. health care delivery system. Sometimes, however, their availability relative to other settings of care means that they may be used when traditional ambulatory settings would be more appropriate and less costly.
- Emergency Department (ED) Utilization—average number of emergency room visits per member per reporting period. This measure, in use by the State of Maine's MaineCare program, assesses the number of visits per member per year as a function of all child and adolescent members enrolled and eligible during the measurement year.
The intent of using this measure is to reduce unnecessary ED visits.
This is a proxy outcome measure with an evidence grade of B. SNAC voting resulted in a ranking of this measure as 2. The data source is administrative records.
The measure is potentially important given the proportion of ED visits paid for by Medicaid:; 61.65 percent of all ED visits of children <1 and 41.87 percent of all ED visits of children 1-17 yrs in 2006.49,g
Some Medicaid and CHIP officials expressed concerns about this measure, given the multitude of reasons for which children come in contact with the ED. As with other ambulatory care-sensitive condition measures, data from this measure are probably best used to raise potential red flags about the quality and accessibility of ambulatory care, with in-depth studies being conducted when plans or providers vary from the average rate.
Inpatient Patient Safety
In 2007, children ages 0-17 accounted for 6.3 million inpatient community hospital discharges, half of which had Medicaid as an expected payer, for total aggregate charges to Medicaid of $42 billion. Health care-associated infections are a major public health concern, for which HHS has a multifaceted, cross-entity action plan.50 Routine surveillance is a major component of the plan.
- Pediatric catheter-associated blood stream infection rates (intensive care and high risk nursery patients). This measure assesses the number of central line-associated blood stream infections (CLABSI) identified during periods selected for surveillance as a function of the number of central line catheter days selected for surveillance in pediatric and neonatal intensive care units. Children at risk are patients in pediatric intensive care units (PICUs), neonatal intensive care units (NICUs), and other intensive care units (ICUs).
This is an outcome measure with an evidence grade of B. SNAC voting resulted in a ranking of 17 for this measure. The data source for the measure is medical records; data are collected by hospital infection control staff.
Health care-associated infections are regarded as "never events," that is, adverse events that should never occur during patient care. Data on the number of children at risk are difficult to come by because at the current time, PICU and NICU placements are not recorded in hospital discharge data. The Healthcare Cost and Utilization Project's (HCUP) Kids Inpatient Database (KID) procedures data suggest that 108,000 hospitalized children with central lines may have been at risk for CLABSI in 2006. However, experts in hospital discharge data note that other procedures for the most severely ill children (e.g., children undergoing chemotherapy, brain surgery) may dominate the available procedure code fields, meaning that the HCUP number is certainly an underestimate. A SNAC member with expertise in this area estimated that there are between 400,000 and 600,000 PICU admissions in a year. National Healthcare Safety Network (NHSN) data for 2007 report a pooled mean of CLABSI rates and central line utilization ratios of 2.9 in pediatric medical/surgical units and 1.0 in pediatric medical ICUs.51 In level III NICUs, the rate is a gradient by birth weight category, ranging from a pooled mean of 3.7 for infants <750 g to 2.0 for infants >2500 g.
Medicaid and CHIP officials on the SNAC reported that many States are working to try to include this measure in their quality monitoring, but that the data have been difficult to obtain. Current national reporting (i.e., NHSN reports) does not include reporting of patients by source of insurance. However, that information is typically available in hospital discharge records and could be reported. According to one analysis, the information has the potential downside of reflecting "expected" rather than actual source of payment,52 so some modifications might be needed. Nineteen States require reporting of hospital-associated infections performance across all populations.53
Management of Chronic Conditions
Almost 20 percent of children ages 0-17 have a special health care need, defined as a chronic condition.38,54 Among the leading chronic conditions are asthma, attention-deficit/hyperactivity disorder (ADHD), and mental health conditions. Long considered a disease of middle age and older people, diabetes is also emerging as an important chronic condition, in large part due to the epidemic of overweight and obesity among children. Asthma has a substantially higher prevalence among black children and somewhat higher among Hispanics. Given the racial and ethnic disparities in overweight and obesity, diabetes is also likely to be more prevalent among these groups.
Apart from dental caries, asthma is the most prevalent chronic physical condition among children. As of 2007, 12.9 percent of children covered by Medicaid reported they currently had asthma (N=2,699,000, equal to one-third of all U.S. children with asthma in that year).55 Hospitalization rates for black children with asthma are two to three times as high as they are for non-Hispanic white children.56-57
- Annual number of asthma patients (>1 year old) with >1 asthma-related emergency room (ER) visit. This State of Alabama Medicaid program measure is similar to the overall measure of ED utilization (measure 17 above), except that the patient population is children with asthma, and the denominator is all children >1 year of age diagnosed with asthma or treatment with at least two short-acting beta adrenergic agents during the measurement year.
This is an outcome measure with an evidence of grade C. SNAC voting resulted in a rank of 3 for this measure. The data source is administrative records.
ED visits are important to measure because they are expensive and provide a gateway to hospitalization. A relatively large body of literature suggests a relationship between the quality and accessibility of ambulatory care and use of the emergency department, although definitive causal evidence linking specific ambulatory care practices and ED visits is lacking. Asthma is the most prevalent chronic physical condition among children. As of 2007, 12.9 percent of children covered by Medicaid reported they currently had asthma (N=2,699,000, equal to one-third of all U.S. children with asthma in that year).
The measure was nominated by the State of Alabama Medicaid Medical Director. Its use by other State programs is not known. However, it seems feasible to use with administrative data, and the State of Alabama specifications can provide a model for implementation.
About 4.5 million children 5-17 years of age had ever been diagnosed with ADHD as of 2006. Children with Medicaid were more likely than uninsured children or privately insured children to have a diagnosis. The 2007 NSCH estimates that 4.2 percent of children nationally take medication for ADHD.38 The rate is higher (6.5 percent) among publicly insured children than among children with private insurance and uninsured children (3.6 percent and 1.5 percent, respectively). Sixty-three percent of parents of publicly insured children ages 2-17 report that their child's ADHD is moderate or severe, which is 30 percent higher than among parents of privately insured or uninsured children.38
- Followup care for children prescribed attention-deficit/hyperactivity disorder (ADHD) medication (continuation and maintenance phase). This measure assesses the extent to which children ages 6-12 diagnosed with ADHD as of the index prescription start date (IPSD) had, in addition to the followup visit in the initiation phase, at least two followup visits with a practitioner within 270 days (9 months) of the end of the initiation phase. To be eligible, children must have remained on the medication for 210 days and be continuously enrolled for 120 days prior to IPSD through 30 days after the IPSD.
This is a process measure with an evidence grade of D. SNAC voting resulted in a ranking of 12 for this measure. Measure specifications are available for reporting by health plans and providers. The measure is currently reported to NCQA by 73 MMC health plans across 24 States. The data source is administrative records.
Followup care for children prescribed ADHD medications is critical, and performance improvements are needed, at least among health plans reporting data to NCQA. In 2007, the national mean among reporting plans with valid data was 38.83 percent, with the 10th percentile plans at 13.21 percent and the 90th percentile plans at 58.33 percent.14
Concerns about this measure come from several perspectives. Health care providers express concerns that only followup visits in person count, rather than by followup by phone. Reporting levels and rates may vary across plans and States, depending on whether children's ADHD care is covered in a carve-out plan. From a health outcomes perspective, the measure lacks any information on the nature of the followup visit and whether appropriate adjustments in care were made, if needed. Balancing all of these perspectives in a single measure would be difficult, although the prevalence of ADHD makes the topic a good candidate for measure enhancement or development.