Abstract
Video Abstract
OBJECTIVE
To assess nationally endorsed claims-based quality measures in pediatric sickle cell anemia (SCA).
METHODS
Using data from the Sickle Cell Data Collection programs in California and Georgia from 2010 to 2019, we evaluated 2 quality measures in individuals with hemoglobin S/S or S/β-zero thalassemia: (1) the proportion of patients aged 3 months to 5 years who were dispensed antibiotic prophylaxis for at least 300 days within each measurement year and (2) the proportion of patients aged 2 to 15 years who received at least 1 transcranial Doppler ultrasound (TCD) within each measurement year. We then evaluated differences by year and tested whether performance on quality measures differed according to demographic and clinical factors.
RESULTS
Only 22.2% of those in California and 15.5% in Georgia met or exceeded the quality measure for antibiotic prophylaxis, with increased odds associated with rural residence in Georgia (odds ratio 1.61; 95% confidence interval 1.21–2.14) compared with urban residence and a trend toward increased odds associated with a pediatric hematologist prescriber (odds ratio 1.28; 95% confidence interval 0.97, 1.69) compared with a general pediatrician. Approximately one-half of the sample received an annual assessment of stroke risk using TCD (47.4% in California and 52.7% in Georgia), with increased odds each additional year in both states and among younger children.
CONCLUSIONS
The rates of receipt of recommended antibiotic prophylaxis and annual TCD were low in this sample of children with SCA. These evidence-based quality measures can be tracked over time to help identify policies and practices that maximize survival in SCA.
What’s Known on This Subject:
Children with SCA are at an increased risk of early mortality. Few nationally endorsed measures exist to track the quality of care received by children with SCA.
What This Study Adds:
This study extends previous evaluations of 2 claims-based quality measures in SCA with similar evaluations that include only children with laboratory-confirmed diagnoses of SCA.
Sickle cell anemia (SCA) is an inherited form of severe anemia associated with acute painful episodes, end-organ damage, and early mortality. It affects ∼100 000 individuals in the United States,1 including 1 in 400 African Americans and 1 in 19 000 Latinos.2 Over the past several decades, longevity has increased, but only to the fifth decade of life.3,4 Among children, the most common causes of death include sepsis and acute chest syndrome, followed by multiorgan failure syndrome and stroke.5 Evaluation of the Dallas Newborn Cohort links the recent significant reduction in deaths due to bacterial sepsis to the delivery of high-quality care, including prophylactic penicillin6 and pneumococcal and Haemophilus influenzae type b vaccines.5 Despite clinical guidelines that proactively support such care,7 poor outcomes persist.8,9 Evidence-based quality indicators can be tracked over time to help identify policies and practices to maximize survival in SCA.
There were no nationally endorsed indicators of quality in pediatric SCA until 2016.10 There are now 2 such measures for children with SCA, the most common and severe form of sickle cell disease.11,12 The first quality indicator assesses the proportion of children aged 3 months to 5 years with SCA who are dispensed prophylactic antibiotics on 300 or more days in a given year. The authors of a randomized controlled trial documented an 84% decrease in pneumococcal sepsis with daily preventive use of penicillin in children with SCA.6 The second nationally vetted quality indicator assesses the proportion of children aged 2 to 15 years with SCA who received their annual screening for stroke risk using transcranial Doppler ultrasonography (TCD). Children with abnormal TCD screens have a 40% risk of stroke within 3 years.13 The initiation of chronic blood transfusions reduces the risk of stroke by 92% among children at the highest risk of stroke, as identified by TCD screening.13,14 In this study, we aimed to assess these nationally endorsed claims-based quality measures among children with SCA and to identify demographic and clinical characteristics (eg, age, sex, and provider type) that may facilitate targeted quality improvement efforts.
Methods
Study Design and Data Sources
We analyzed data from the Sickle Cell Data Collection (SCDC) programs in California and Georgia from 2010 to 2019. The SCDC is a longitudinal, population-based surveillance system that includes linked data from multiple sources to identify individuals living with SCA.15,16 SCDC is overseen by the Division of Blood Disorders of the Centers for Disease Control and Prevention. This project was reviewed by the Centers for Disease Control and Prevention and was determined to be a non-research, public health practice activity. The Georgia Public Health Department and the Georgia State University Institutional Review Board declared the project exempt from review, whereas the California Committee for the Protection of Human Subjects reviewed and approved the study with a waiver of consent for individual data collection. In addition, specialty hemoglobinopathy treatment center institutional review boards similarly exempted the project from review or approved with a waiver of consent. State data requests were reviewed by the appropriate agency, assuring that data privacy safeguards were in place.
Study Population
We included all individuals with hemoglobin (Hb) S/S or Hb S/β0-thalassemia who were reported by newborn screening or one of the participating hemoglobinopathy specialty treatment centers in Georgia and California from 2004 to 2019 with a laboratory-confirmed diagnosis based on the results of clinical laboratory evaluation that included quantitative Hb identification by Hb electrophoresis, high-performance liquid chromatography, or DNA analysis. These individuals were then linked to 2010 to 2019 Medicaid eligibility and claims files by using probabilistic methods in Georgia and deterministic methods based on social security numbers in California. Linkage processes and sources are further detailed elsewhere.15 Participants who were Medicaid beneficiaries for all 12 months of the study year were included.
Measures
Definition of Quality Indicators: Antibiotic Prophylaxis
The numerator is the number of children aged 3 months to 4 years with SCA who were dispensed appropriate antibiotic prophylaxis for at least 300 days within the measurement year (ie, must not have a fifth birthday within the measurement year). The age range begins at 3 months of age to account for any lag in confirmation of the SCA status of the child. Each measurement year extends from January 1 to December 31. Although the guidelines specifically recommend penicillin for antibiotic prophylaxis, some children may have penicillin sensitivity. Therefore, this measure uses a definition of antibiotic prophylaxis beyond penicillin (ie, aminopenicillins, β-lactamase inhibitors, macrolides, natural penicillins, erythromycin-sulfisoxazole, and vancomycin). The drugs were identified in the data by the National Drug Code. The denominator is the number of children aged 3 months to 4 years with SCA and continuous enrollment in Medicaid for 12 months of the measurement year. The resulting proportion may be interpreted such that higher proportions indicate better quality.
Definition of Quality Indicators: TCD
The numerator is the number of children aged 2 to 15 years with SCA who received at least 1 TCD screening within the measurement year (ie, must not have a second or 16th birthday within the measurement year). The receipt of TCD screening is identified as the presence of at least 1 current procedural terminology code for any of 5 acceptable ultrasonography tests (ie, 93886, 93888, 93890, 93892, and 93893) within the measurement year among children in the target population. The denominator is the number of children aged 2 to 15 years with SCA and continuous enrollment in Medicaid for 12 months of the measurement year. As before, the resulting proportion may be interpreted such that higher proportions indicate better quality.
Demographic and Clinical Measures
Age was defined as of January 1 of the year of study. Sex, race, and state of residence were defined as of the last entry date in the Medicaid files during the year of study. Rural or urban status was defined by using 2013 USDA Urban Rural Continuum Codes (https://www.ers.usda.gov/data-products/rural-urban-continuum-codes.aspx) and was based on the patient’s residence county as reported in Medicaid enrollment data. California’s SCA population is heavily concentrated in areas deemed to be “urban;” counts of those in rural areas are too small to be reported, so urban and rural analyses are provided for Georgia data only. Because children are administered medications while receiving inpatient care, antibiotic prescription days were defined as the total days of filled prescriptions plus the number of days hospitalized during the year of study. To categorize the prescribing provider type, we used the National Provider Identifier file from the Centers for Medicaid and Medicare Services’ Web site (https://download.cms.gov/nppes/NPI_Files.html), which lists the provider’s primary and up to 14 additional taxonomy codes, and linked it to the prescription claims. Some of the children had >1 prescribing provider type. If the child had at least 1 antibiotic prescription from a pediatric hematologist, pediatric hematologist was tallied. If the child did not have a prescribing pediatric hematologist but did have an antibiotic prescription from a pediatrician, pediatrician was tallied. In the case of any other provider type, the tally for prescriber type was “other.” Of note, the available data did not provide a link between provider type and TCD events. Further discussion of the link between provider type and claims data may be found in previous work.17
Statistical Analysis: Quality Indicator Performance in California and Georgia
We examined SCDC data in strata of state and year with summary counts of persons meeting and not meeting the antibiotic prophylaxis and TCD quality indicators. Within each of the state, year, and quality indicator strata, analyses were further stratified by variables in 5 separate data tables of summary counts within (1) state, year, and age strata, (2) state, year, and sex strata, (3) state, year, and race (African American, not African American, unknown) strata, (4) state, year, and residence (urban, rural) strata, and (5) state, year, and provider type (pediatric hematologist, pediatrician, other) strata (for antibiotic prophylaxis indicators only). Antibiotic prophylaxis data were categorized into age groups of 3 months to 1 year, 1 year, 2 years, and 3 years. TCD data were categorized into age groups of 2 to 4 years, 5 to 7 years, 8 to 10 years, and 11 to 14 years. Within each state, we calculated the proportion of persons meeting antibiotic prophylaxis and TCD quality indicators by year with 95% confidence intervals (CIs). Differences by year in proportions meeting quality indicators were tested by using a χ-square test and a test for trend by year using the Cochran–Armitage test for trend. We used logistic regression to test whether the achievement of quality indicators differed by age, sex, race, urban or rural residence, or provider type. Because data were summarized by separate strata of these independent variables (as described above), we were unable to combine all the variables in a single regression model; each of these variables was modeled separately, with the inclusion of state and year. Given the small numbers of rural residents in the California data, urban or rural residence was analyzed by using only Georgia data. The dependent binary variable indicated whether a child met the performance metric in a given state or measurement year. A robust variance–covariance parameter matrix was used to account for model misspecification due to children contributing data to multiple measurement years.18 Measurement year was modeled as an indicator variable, with associations estimated as odds ratios (ORs) with 95% CIs, relative to the 2010 referent year. Measurement year was also modeled as a single variable to assess trends. Interactions of each variable (age, sex, etc) with year were also included to evaluate if these associations differed by year. See the Supplemental Information for further specification of the models. All statistical tests were performed by using 2-sided P < .05. Analyses were performed by using SAS 9.4 and Stata 17.0.19,20
Results
Demographic Characteristics
Approximately one-third of the sample was represented in claims data from California, and the remainder was from Georgia. Most children resided in urban areas, and more than one-half relied on pediatric hematologists to prescribe prophylactic antibiotics. General pediatricians were the next most common prescribers of antibiotics, serving ∼1 in 5 children with filled antibiotic prescriptions. More than 90% of the sample reported African American race. The distribution across age and sex groups was approximately even. These demographic trends were observed in the overall sample and in both the antibiotic prophylaxis and TCD ultrasound subsamples (Tables 1 and 2).
TABLE 1.
Demographic Characteristics: Antibiotic Prophylaxis Population
| 2010 | 2011 | 2012 | 2013 | 2014 | 2015 | 2016 | 2017 | 2018 | 2019 | |
|---|---|---|---|---|---|---|---|---|---|---|
| State | ||||||||||
| California | 110 (30.0%) | 121 (31.2%) | 125 (35.4%) | 121 (35.1%) | 122 (36.9%) | 122 (39.6%) | 116 (37.4%) | 114 (37.1%) | 103 (31.7%) | 107 (33.3%) |
| Georgia | 256 (70.0%) | 267 (68.8%) | 228 (64.6%) | 224 (64.9%) | 209 (63.1%) | 186 (60.4%) | 194 (62.6%) | 193 (62.9%) | 222 (68.3%) | 214 (66.7%) |
| Sex | ||||||||||
| Female | 167 (45.6%) | 184 (47.4%) | 182 (51.6%) | 175 (50.7%) | 170 (51.4%) | 156 (50.6%) | 145 (46.8%) | 139 (45.3%) | 152 (46.8%) | 150 (46.7%) |
| Male | 199 (54.4%) | 204 (52.6%) | 171 (48.4%) | 170 (49.3%) | 161 (48.6%) | 152 (49.4%) | 165 (53.2%) | 168 (54.7%) | 173 (53.2% | 171 (53.3%) |
| Age | ||||||||||
| 3 mo to <1 y | 75 (20.5%) | 59 (15.2%) | 63 (17.8%) | 65 (18.8%) | 58 (17.5%) | 44 (14.3%) | 56 (18.1%) | 54 (17.6%) | 75 (23.1%) | 57 (17.8%) |
| 1 to <2 y | 96 (26.2%) | 106 (27.3%) | 83 (23.5%) | 87 (25.2%) | 88 (26.6%) | 81 (26.3%) | 72 (23.2%) | 87 (28.3%) | 80 (24.6%) | 91 (28.3%) |
| 2 to <3 y | 111 (30.3%) | 105 (27.1%) | 101 (28.6%) | 85 (24.6%) | 96 (29.0%) | 88 (28.6%) | 93 (30.0%) | 77 (25.1%) | 93 (28.6%) | 88 (27.4%) |
| 3 to <4 y | 84 (23.0%) | 118 (30.4%) | 106 (30.0%) | 108 (31.3%) | 89 (26.9%) | 95 (30.8%) | 89 (28.7%) | 89 (29.0%) | 77 (23.7%) | 85 (26.5%) |
| Race | ||||||||||
| African American | 342 (93.4%) | 361 (93.0%) | 330 (93.5%) | 324 (93.9%) | 309 (93.4%) | 287 (93.2%) | 296 (95.5%) | 292 (95.1%) | 305 (93.8%) | 300 (93.5%) |
| Not African American | 20 (5.5%) | 22 (5.7%) | 20 (5.7%) | 20 (5.8%) | 20 (6.0%) | 19 (6.2%) | 11 (3.5%) | 12 (3.9%) | 18 (5.5%) | 18 (5.6%) |
| Unknown | 4 (1.1%) | 5 (1.3%) | 3 (0.8%) | 1 (0.3%) | 2 (0.6%) | 2 (0.6%) | 3 (1.0%) | 3 (1.0%) | 2 (0.6%) | 3 (0.9%) |
| Residencea | ||||||||||
| Urban | 207 (80.9%) | 223 (83.5%) | 186 (81.6%) | 182 (81.2%) | 176 (84.2%) | 155 (83.3%) | 161 (83.0%) | 168 (87.0%) | 196 (88.3%) | 190 (88.8%) |
| Rural | 49 (19.1%) | 44 (16.5%) | 42 (18.4%) | 42 (18.8%) | 33 (15.8%) | 31 (16.7%) | 33 (17.0%) | 25 (13.0%) | 26 (11.7%) | 24 (11.2%) |
| Provider type | ||||||||||
| Pediatric hematologist | 214 (75.6%) | 227 (71.4%) | 215 (71.4%) | 225 (74.3%) | 216 (71.5%) | 210 (74.5%) | 224 (76.4%) | 240 (85.1%) | 264 (87.1%) | 248 (84.3%) |
| Pediatrician | 37 (13.1%) | 53 (16.7%) | 62 (20.6%) | 53 (17.5%) | 52 (17.2%) | 37 (13.1%) | 19 (6.5%) | 25 (8.9%) | 25 (8.3%) | 27 (9.2%) |
| Other | 32 (11.3%) | 38 (11.9%) | 24 (8.0%) | 25 (8.2%) | 34 (11.3%) | 35 (12.4%) | 50 (17.1%) | 17 (6.0%) | 34 (4.6%) | 19 (6.5%) |
Urban/rural residence analysis limited to Georgia.
TABLE 2.
Demographic Characteristics: TCD Population
| 2010 | 2011 | 2012 | 2013 | 2014 | 2015 | 2016 | 2017 | 2018 | 2019 | |
|---|---|---|---|---|---|---|---|---|---|---|
| State | ||||||||||
| California | 317 (29.5%) | 332 (29.2%) | 357 (29.6%) | 364 (30.0%) | 388 (30.9%) | 415 (32.4%) | 432 (33.7%) | 421 (32.7%) | 423 (33.3%) | 385 (31.0%) |
| Georgia | 759 (70.5%) | 806 (70.8%) | 847 (70.4%) | 850 (70.0%) | 869 (69.1%) | 866 (67.6%) | 850 (66.3%) | 867 (67.3%) | 849 (66.7%) | 855 (69.0%) |
| Sex | ||||||||||
| Female | 516 (48.0%) | 552 (48.5%) | 588 (48.8%) | 599 (49.3%) | 629 (50.0%) | 638 (49.8%) | 637 (49.7%) | 621 (48.2%) | 597 (46.9%) | 585 (47.2%) |
| Male | 560 (52.0%) | 586 (51.5%) | 616 (51.2%) | 615 (50.7%) | 628 (50.0%) | 643 (50.2%) | 645 (50.3%) | 667 (51.8%) | 675 (53.1%) | 655 (52.8%) |
| Age | ||||||||||
| 2 to <5 y | 291 (27.0%) | 305 (26.8%) | 318 (26.4%) | 306 (25.2%) | 293 (23.3%) | 278 (21.7%) | 275 (21.4%) | 259 (20.1%) | 261 (20.5%) | 256 (20.7%) |
| 5 to <8 y | 278 (25.8%) | 291 (25.6%) | 295 (24.5%) | 284 (23.4%) | 321 (25.5%) | 345 (26.9%) | 328 (25.6%) | 315 (24.5%) | 288 (22.6%) | 289 (23.3%) |
| 8 to <11 y | 220 (20.5%) | 254 (22.3%) | 278 (23.1%) | 304 (25.0%) | 299 (23.8%) | 291 (22.7%) | 296 (23.1%) | 317 (24.6%) | 336 (26.4%) | 308 (24.8%) |
| 11 to <15 y | 287 (26.7%) | 288 (25.3%) | 313 (26.0%) | 320 (26.4%) | 344 (27.4%) | 367 (28.7%) | 383 (29.9%) | 397 (30.8%) | 387 (30.4%) | 387 (31.2%) |
| Race | ||||||||||
| African American | 973 (90.4%) | 1039 (91.3%) | 1108 (92.0%) | 1121 (92.3%) | 1175 (93.5%) | 1192 (93.0%) | 1199 (93.1%) | 1208 (93.8%) | 1200 (94.3%) | 1168 (94.2%) |
| Not African American | 60 (5.6%) | 64 (5.6%) | 64 (5.3%) | 63 (5.2%) | 60 (4.8%) | 68 (5.3%) | 68 (5.3%) | 65 (5.0%) | 63 (5.0%) | 63 (5.1%) |
| Unknown | 43 (4.0%) | 35 (3.1%) | 32 (2.7%) | 30 (2.5%) | 22 (1.7%) | 21 (1.6%) | 20 (1.6%) | 15 (1.2%) | 9 (0.7%) | 9 (0.7%) |
| Residencea | ||||||||||
| Urban | 627 (82.6%) | 672 (83.4%) | 704 (83.1%) | 705 (82.9%) | 721 (83.0%) | 730 (84.3%) | 721 (84.8%) | 736 (84.9%) | 722 (85.0%) | 739 (86.4%) |
| Rural | 132 (17.4%) | 134 (16.6%) | 143 (16.9%) | 145 (17.1%) | 148 (17.0%) | 136 (15.7%) | 129 (15.2%) | 131 (15.1%) | 127 (15.0%) | 116 (13.6%) |
Urban/rural residence analysis limited to Georgia.
Antibiotic Prophylaxis
Regardless of demographic characteristics, few children had antibiotic prescriptions filled for 300 or more days per year from 2010 to 2019 (Fig 1). Over this time period, the overall proportion of children meeting or exceeding this quality measure was 22.2% in California and 15.5% in Georgia. The odds of appropriate antibiotic prophylaxis were not dependent on age, sex, or year. Only rural residence in Georgia (OR 1.61; 95% CI 1.21–2.14) compared with urban residence was associated with increased odds of appropriate antibiotic prophylaxis; pediatric hematologist prescriber (OR 1.28; 95% CI 0.97–1.69) compared with pediatrician prescriber revealed a trend toward increased antibiotic prophylaxis (Table 3).
FIGURE 1.
Antibiotic prophylaxis. Percentage of children aged 3 months to 4 years old with SCA who were dispensed appropriate antibiotic prophylaxis for at least 300 days within each measurement year.
TABLE 3.
Association With Antibiotic Prophylaxis ≥300 Days per Year
| Variable | OR (95% CI) | Wald P | Predicted Probability (95% CI): Prophylaxis ≥300 d per y |
|---|---|---|---|
| Model 1: sex, y, state | |||
| Sex | |||
| Male | 1.0 (referent) | .37 | 0.184 (0.166–0.202) |
| Female | 0.92 (0.77–1.10) | — | 0.172 (0.154–0.192) |
| Year (vs 2010) | — | .36 | — |
| 2010 | 1.0 (referent) | — | 0.189 (0.148–0.229) |
| 2011 | 1.01 (0.70–1.46) | — | 0.191 (0.151–0.230) |
| 2012 | 1.06 (0.73–1.55) | — | 0.198 (0.156–0.240) |
| 2013 | 1.07 (0.74–1.56) | — | 0.200 (0.158–0.242) |
| 2014 | 1.00 (0.68–1.47) | — | 0.189 (0.147–0.231) |
| 2015 | 0.66 (0.43–1.01) | — | 0.134 (0.096–0.171) |
| 2016 | 0.76 (0.50–1.14) | — | 0.150 (0.110–0.189) |
| 2017 | 0.88 (0.59–1.32) | — | 0.171 (0.129–0.212) |
| 2018 | 1.01 (0.69–1.48) | — | 0.190 (0.147–0.232) |
| 2019 | 0.83 (0.56–1.24) | — | 0.163 (0.123–0.203) |
| Model 2: age, y, state | |||
| Age | |||
| 3mo to <1 y | 1.0 (referent) | .67 | 0.181 (0.150–0.212) |
| 1 to <2 y | 1.07 (0.82–1.40) | — | 0.191 (0.165–0.217) |
| 2 to <3 y | 0.94 (0.71–1.22) | — | 0.171 (0.147–0.195) |
| 3 to <4 y | 0.94 (0.72–1.23) | — | 0.172 (0.148–0.196) |
| Year | — | .38 | — |
| Model 3: residence, y (conducted in Georgia only)a | |||
| Residence | |||
| Urban | 1.0 (referent) | .001 | 0.144 (0.128–0.160) |
| Rural (vs urban) | 1.61 (1.21–2.14) | — | 0.213 (0.170–0.255) |
| Year | — | .63 | — |
| Model 4: provider, y, state | |||
| Provider | |||
| Pediatrician | 1.0 (referent) | — | 0.164 (0.130–0.199) |
| Pediatric hematologist | 1.28 (0.97–1.69) | .08 | 0.201 (0.184–0.217) |
| Other | 0.96 (0.64–1.45) | .86 | 0.159 (0.117–0.202) |
| Year | — | .25 | — |
All models use logistic regression, including state and year as covariates; —, the indicated statistical analysis was not performed.
Estimates of ORs and probabilities by year are limited to the sex–year model for brevity (estimates are consistent across models).
Residence model limited to Georgia data.
TCD Screening
Approximately one-half of children with SCA received an annual assessment of stroke risk using TCD, with 47.4% in California and 52.7% in Georgia. There was an increased odds of receiving TCD screening in both California (OR 1.04; 95% CI 1.01–1.06 per year) and Georgia (OR 1.09; 95% CI 1.07–1.11 per year) with each additional year of observation (Fig 2). In contrast, age was associated with decreased receipt of TCD screening, such that older children were less likely to undergo TCD screening. As observed for antibiotic prophylaxis, sex was not associated with performance rates (Table 4).
FIGURE 2.
TCD screening. Percentage of children aged 2 to 15 years old with SCA who received at least 1 TCD screening within each measurement year.
TABLE 4.
Association With ≥1 TCD Screening per Year
| Variable | OR (95% CI) | Wald P | Predicted Probability (95% CI): ≥1 Screening per y |
|---|---|---|---|
| Model 1: sex, y, state | |||
| Sex | |||
| Male | 1.0 (referent) | .26 | 0.505 (0.493–0.517) |
| Female | 1.04 (0.97–1.12) | — | 0.515 (0.503–0.528) |
| Year | |||
| 2010 | 1.0 (referent) | <.001 | 0.459 (0.429–0.489) |
| 2011 | 1.07 (0.91–1.27) | — | 0.477 (0.448–0.506) |
| 2012 | 0.99 (0.84–1.16) | — | 0.456 (0.428–0.484) |
| 2013 | 1.00 (0.85–1.18) | — | 0.459 (0.431–0.487) |
| 2014 | 0.86 (0.73–1.02) | — | 0.423 (0.396–0.450) |
| 2015 | 1.31 (1.11–1.54) | — | 0.527 (0.499–0.554) |
| 2016 | 1.49 (1.27–1.75) | — | 0.558 (0.531–0.585) |
| 2017 | 1.41 (1.20–1.66) | — | 0.545 (0.518–0.572) |
| 2018 | 1.63 (1.38–1.92) | — | 0.580 (0.553–0.607) |
| 2019 | 1.78 (1.51–2.10) | — | 0.602 (0.574–0.629) |
| Model 2: age, y, state | |||
| Age | |||
| 2 to <5 y | 1.0 (referent) | <.001 | 0.575 (0.557–0.593) |
| 5 to <8 y | 0.87 (0.79–0.97) | — | 0.542 (0.525–0.560) |
| 8 to <11 y | 0.73 (0.66–0.81) | — | 0.498 (0.480–0.516) |
| 11 to <15 y | 0.57 (0.52–0.63) | — | 0.439 (0.422–0.455) |
| Year | — | <.001 | — |
| Model 3: residence, y (conducted in Georgia only) | |||
| Residence | |||
| Urban | 1.0 (referent) | .36 | 0.525 (0.513–0.536) |
| Rural | 1.06 (0.94–1.19) | — | 0.538 (0.512–0.565) |
| Year (per y) | 1.07 (1.05–1.10) | <.001 | — |
All models use logistic regression, including state and year as covariates; —, the indicated statistical analysis was not performed.
Estimates of ORs and probabilities by year are limited to the sex–year model for brevity (estimates are consistent across models).
Residence model limited to Georgia data.
Discussion
We used Medicaid claims data to assess the quality of care for children with SCA in California and Georgia from 2010 to 2019. In doing so, we determined that the majority of children with SCA in our sample did not receive appropriate amounts of recommended antibiotic prophylaxis, and approximately one-half received annual TCDs to assess stroke risk. We also examined key demographic and clinical characteristics as potential predictors of performance on these quality measures. We found that children with SCA had a greater likelihood of receiving appropriate antibiotic prophylaxis when they lived in rural areas of Georgia; there was also a trend toward increased antibiotic prophylaxis with pediatric hematologist prescribers. Those who were <5 years old had a greater likelihood of receiving annual TCDs. Therefore, in these preliminary analyses, age, residence, and provider type appeared to influence the quality of care in pediatric SCA.
Performance rates on both quality measures were similar to those in previous studies. A recent evaluation of children 3 months to 5 years of age with SCA in Florida, Illinois, Louisiana, Michigan, South Carolina, and Texas (2005–2012) revealed that only 18% of children had 300 or more filled antibiotic prescriptions each year.21 The odds of receiving appropriate antibiotic prophylaxis increased with the number of well child visits and SCA-related outpatient visits, but not age, sex, year, or state. We similarly found that age, sex, and year did not predict the rates of antibiotic prophylaxis. Analysis of claims data from the same states (2005–2010) for 2- to 16-year-old children with SCA showed annual TCD screening rates of 22% to 44%.22 The previous receipt of TCD and number of well child visits predicted greater odds of TCD screening, whereas older age predicted lower odds. Older children in our sample and another cohort15 were also less likely to receive timely TCD screening.
The examination of similar Medicaid claims in New York and Michigan (2011–2018) revealed annual prophylactic antibiotic rates of 16% to 22% and TCD screening rates of 39% to 45%.23 Performance on these measures remained the same or decreased over the evaluation period. In contrast, trend analyses of TCD screening rates in our data from 2010 to 2019 increased in both California and Georgia. There were measurement differences in that we limited our claims data to those with laboratory-confirmed cases of SCA, but several other factors potentially contributed to these outcomes. For instance, the average annual TCD screening rate for eligible patients in a large pediatric sickle cell center was 45%, but there was an average cancellation rate by patients of 20%.24 It was also found that providers who had at least 1 opportunity per year to order TCD screening did so 65% to 80% of the time. One study links low rates of TCD completion to low provider expectations that families will have access to testing or agree to a chronic transfusion protocol if a significant stroke risk is detected.25 In fact, provider-perceived low patient adherence to TCD appointments and distance to a vascular laboratory have been identified as barriers to TCD screening.26 However, providers, especially primary care physicians, may also lack the self-efficacy and awareness of TCD recommendations to make the necessary orders.25,26 Further study is needed to disentangle the factors that affect TCD screening rates and devise approaches to improve care.
Importantly, scholars have identified enhanced national surveillance systems and access to care as targets to decrease the health inequities experienced by individuals with SCA.27 However, as recently as 2018, there was an opportunity to include filled antibiotic prescriptions and TCD screenings into the core set of 25 pediatric quality measures that track the performance of the Medicaid program. Despite expert recommendations, the Centers for Medicare and Medicaid Services did not include these measures in the core set provided to states that opt to track quality in cohorts with complex chronic disease.28 Since then, isolated statewide evaluations of Medicaid data continue to reveal the same poor outcomes: low likelihood that a child with SCA in the United States will receive appropriate antibiotic prophylaxis and a coin toss that he or she will receive regular screenings to assess stroke risk.
This analysis represents an opportunity to systematically track outcomes in pediatric SCA with some limitations. Unlike previous evaluations of nationally endorsed quality measures in pediatric SCA, we did not use claims data to identify cases. Instead, we included only children with laboratory-confirmed diagnoses of Hb SS disease or Hb S/β0-thalassemia. This increased our accuracy in identifying cases, but may not be replicable in states without access to such data. Notably, users of the claims-based approach to identify children with SCA have reported methods with high accuracy.29–31 Such data have also been successfully used to track the receipt of TCDs.32 Our findings were consistent with those based exclusively on claims data and may be reinforced by analyses that compare our findings with those based on administrative data alone.
Similar to previous evaluations, we report quality measures based on Medicaid claims, which exclude children with SCA who have other means of health insurance. Studies suggest that those with private insurance are more likely to receive regular antibiotic prophylaxis and TCD screens.24,33 However, the majority of health care for pediatric SCA is supported by the Medicaid program,34 so our findings are expected to reflect the reality faced by most patients. In addition, this analysis was based on aggregated data stratified by year and demographic characteristics; therefore, we cannot report performance rates by combinations of those characteristics. The data use agreement does not allow the patient-level analysis that would be required. Finally, the data revealed few rural residents in California because large urban counties capture most residents but do not necessarily equate to urban-based care for all residents. As a result, the comparison of rural versus urban residence was not a reliable marker of rural versus urban care in California.
In sum, despite support from high-quality studies and explicit clinical guidelines, many children with SCA do not receive adequate antibiotic prophylaxis or annual screenings for stroke risk. This may reflect gaps in provider knowledge and practice alongside challenges families face accessing care, as evidenced by suboptimal rates of health supervision visits, hematologist visits, and pneumococcal vaccination.17,35,36 There remains a critical need to know where we stand in the delivery of care to children with SCA to implement quality improvement programs, compare with appropriate benchmarks, and track progress over time. The next steps may include further inquiry into statewide approaches to care, focus groups on provider expectations and patient experience, and surveys of general pediatrician knowledge of care in SCA. In the absence of such quality assessments, we risk failing this vulnerable population and squandering the opportunity to learn from high-performing systems of care.
Supplementary Material
Acknowledgments
We would like to express our gratitude for the advisory contributions provided by Sarah Hunter of the RAND Corporation during the developmental and execution phases of the present analysis. We are also grateful for the contributions provided by the children affected by sickle cell disease in this analysis and the providers who care for them.
Glossary
- CI
confidence interval
- Hb
hemoglobin
- OR
odds ratio
- SCA
sickle cell anemia
- TCD
transcranial Doppler ultrasonography
Footnotes
Dr Anderson conceptualized and designed the study and drafted the initial manuscript; Dr Mack cleaned the data, performed and summarized all statistical analyses, including tables and figures, and drafted portions of the methods and results sections of the manuscript; Ms Horiuchi provided access to and interpretation of the California data and conducted the descriptive analysis of California data; Ms Paulukonis provided access to and interpretation of the California data and provided counsel during study conceptualization, design, and data analysis; Ms Zhou provided access to and interpretation of the Georgia data and conducted the descriptive analysis of Georgia data; Dr Snyder provided access to and interpretation of the Georgia data and provided counsel during study conceptualization, design, and data analysis; Drs Doctor and Kipke contributed to the conception and design of the work and the interpretation of the data; Dr Coates provided counsel during study conceptualization and design; Dr Freed assisted in design of the study and data analysis; and all authors reviewed and revised the manuscript, approved of the final manuscript as submitted, and agree to be accountable for all aspects of the work.
FUNDING: This article was supported by grant UL1TR001855 from the National Center for Advancing Translational Science (NCATS) of the US National Institutes of Health. The contributions from the Sickle Cell Data Collection teams from California and Georgia were supported by a cooperative agreement from the Centers for Disease Control and Prevention Sickle Cell Data Collection Program (CDC-RFA-DD20-2003). The funder did not participate in the work.
CONFLICT OF INTEREST DISCLOSURES: The authors have indicated they have no potential conflicts of interest relevant to this article to disclose.
COMPANION PAPER: A companion to this article can be found online at https://www.pediatrics.org/cgi/doi/10.1542/peds.2023-064284.
References
- 1. Brousseau DC, Panepinto JA, Nimmer M, Hoffmann RG. The number of people with sickle-cell disease in the United States: national and state estimates. Am J Hematol. 2010;85(1):77–78 [DOI] [PubMed] [Google Scholar]
- 2. National Human Genome Research Institute. Learning about sickle cell disease. Available at: https://www.genome.gov/10001219. Accessed November 17, 2022
- 3. Lanzkron S, Carroll CP, Haywood CJr. Mortality rates and age at death from sickle cell disease: U.S., 1979–2005. Public Health Rep. 2013;128(2):110–116 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Hamideh D, Alvarez O. Sickle cell disease related mortality in the United States (1999-2009). Pediatr Blood Cancer. 2013;60(9):1482–1486 [DOI] [PubMed] [Google Scholar]
- 5. Quinn CT, Rogers ZR, Buchanan GR. Survival of children with sickle cell disease. Blood. 2004;103(11):4023–4027 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Gaston MH, Verter JI, Woods G, et al. Prophylaxis with oral penicillin in children with sickle cell anemia. A randomized trial. N Engl J Med. 1986;314(25):1593–1599 [DOI] [PubMed] [Google Scholar]
- 7. Expert Panel. Evidence-based management of sickle cell disease: expert panel report. Available at: www.nhlbi.nih.gov/health-pro/guidelines/sickle-cell-disease-guidelines/. Accessed August 2, 2023
- 8. Davis H, Gergen PJ, Moore RMJr. Geographic differences in mortality of young children with sickle cell disease in the United States. Public Health Rep. 1997;112(1):52–58 [PMC free article] [PubMed] [Google Scholar]
- 9. Kanter J, Phillips S, Schlenz AM, et al. Transcranial Doppler screening in a current cohort of children with sickle cell anemia: results from the DISPLACE study. J Pediatr Hematol Oncol. 2021;43(8):e1062–e1068 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Wang CJ, Kavanagh PL, Little AA, et al. Quality-of-care indicators for children with sickle cell disease. Pediatrics. 2011;128(3):484–493 [DOI] [PubMed] [Google Scholar]
- 11. National Quality Forum. Antibiotic prophylaxis among children with sickle cell anemia. NQF# 3166. Available at: https://www.qualityforum.org/QPS/QPSTool.aspx#qpsPageState=%7B%22TabType%22%3A1,%22TabContentType%22%3A2,%22SearchCriteriaForStandard%22%3A%7B%22TaxonomyIDs%22%3A%5B%5D,%22SelectedTypeAheadFilterOption%22%3A%7B%22ID%22%3A0,%22FilterOptionLabel%22%3A%222797%22,%22TypeOfTypeAheadFilterOption%22%3A1,%22TaxonomyId%22%3A0,%22FilterOptionLabe%22%3A%222797%22%7D,%22Keyword%22%3A%222797%22,%22PageSize%22%3A%2225%22,%22OrderType%22%3A3,%22OrderBy%22%3A%22ASC%22,%22PageNo%22%3A1,%22IsExactMatch%22%3Afalse,%22QueryStringType%22%3A%22%22,%22ProjectActivityId%22%3A%220%22,%22FederalProgramYear%22%3A%220%22,%22FederalFiscalYear%22%3A%220%22,%22FilterTypes%22%3A0,%22EndorsementStatus%22%3A%22%22,%22MSAIDs%22%3A%5B%5D%7D,%22SearchCriteriaForForPortfolio%22%3A%7B%22Tags%22%3A%5B%5D,%22FilterTypes%22%3A0,%22PageStartIndex%22%3A1,%22PageEndIndex%22%3A25,%22PageNumber%22%3Anull,%22PageSize%22%3A%2225%22,%22SortBy%22%3A%22Title%22,%22SortOrder%22%3A%22ASC%22,%22SearchTerm%22%3A%22%22%7D,%22ItemsToCompare%22%3A%5B%5D,%22SelectedStandardIdList%22%3A%5B%5D,%22StandardID%22%3A1519,%22EntityTypeID%22%3A1%7D. Accessed August 2, 2023
- 12. National Quality Forum. Transcranial Doppler ultrasonography screening for children with sickle cell anemia. NQF# 2797. Available at: https://www.qualityforum.org/QPS/QPSTool.aspx#qpsPageState=%7B%22TabType%22%3A1,%22TabContentType%22%3A2,%22SearchCriteriaForStandard%22%3A%7B%22TaxonomyIDs%22%3A%5B%5D,%22SelectedTypeAheadFilterOption%22%3A%7B%22ID%22%3A3166,%22FilterOptionLabel%22%3A%223166%22,%22TypeOfTypeAheadFilterOption%22%3A4,%22TaxonomyId%22%3A0%7D,%22Keyword%22%3A%223166%22,%22PageSize%22%3A%2225%22,%22OrderType%22%3A3,%22OrderBy%22%3A%22ASC%22,%22PageNo%22%3A1,%22IsExactMatch%22%3Afalse,%22QueryStringType%22%3A%22%22,%22ProjectActivityId%22%3A%220%22,%22FederalProgramYear%22%3A%220%22,%22FederalFiscalYear%22%3A%220%22,%22FilterTypes%22%3A0,%22EndorsementStatus%22%3A%22%22,%22MSAIDs%22%3A%5B%5D%7D,%22SearchCriteriaForForPortfolio%22%3A%7B%22Tags%22%3A%5B%5D,%22FilterTypes%22%3A0,%22PageStartIndex%22%3A1,%22PageEndIndex%22%3A25,%22PageNumber%22%3Anull,%22PageSize%22%3A%2225%22,%22SortBy%22%3A%22Title%22,%22SortOrder%22%3A%22ASC%22,%22SearchTerm%22%3A%22%22%7D,%22ItemsToCompare%22%3A%5B%5D,%22SelectedStandardIdList%22%3A%5B%5D,%22StandardID%22%3A1732,%22EntityTypeID%22%3A1%7D. Accessed August 2, 2023
- 13. Adams RJ, McKie VC, Carl EM, et al. Long-term stroke risk in children with sickle cell disease screened with transcranial Doppler. Ann Neurol. 1997;42(5):699–704 [DOI] [PubMed] [Google Scholar]
- 14. Adams R, McKie V, Nichols F, et al. The use of transcranial ultrasonography to predict stroke in sickle cell disease. N Engl J Med. 1992;326(9):605–610 [DOI] [PubMed] [Google Scholar]
- 15. Snyder AB, Lakshmanan S, Hulihan MM, et al. Surveillance for sickle cell disease — Sickle Cell Data Collection Program, two states, 2004–2018. MMWR Surveill Summ. 2022;71(9):1–18 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Centers for Disease Control and Prevention. Sickle cell data collection (SCDC) program. Available at: https://www.cdc.gov/ncbddd/hemoglobinopathies/scdc.html. Accessed August 5, 2023
- 17. Horiuchi SS, Zhou M, Snyder A, Paulukonis ST. Hematologist encounters among Medicaid patients who have sickle cell disease. Blood Adv. 2022;6(17):5128–5131 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Huber PJ. The behavior of maximum likelihood estimates under nonstandard conditions. In: Proceedings of the Fifth Berkeley Symposium on Mathematical Statistics and Probability. Berkeley: University of California Press; 1967. p. 221–233 [Google Scholar]
- 19. SAS Institute, Inc. SAS Enterprise Miner 9.4, Cary, NC. Available at: https://www.sas.com/en_us/software/enterprise-miner.html
- 20. StataCorp, LLC. Stata statistical software: release 17, College Station, TX. Available at: https://www.stata.com/
- 21. Reeves SL, Tribble AC, Madden B, et al. Antibiotic prophylaxis for children with sickle cell anemia. Pediatrics. 2018;141(3):e20172182. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Reeves SL, Madden B, Freed GL, Dombkowski KJ. Transcranial Doppler screening among children and adolescents with sickle cell anemia. JAMA Pediatr. 2016;170(6):550–556 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Reeves SL, Freed GL, Madden B, et al. Trends in quality of care among children with sickle cell anemia. Pediatr Blood Cancer. 2022;69(2):e29446. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Raphael JL, Shetty PB, Liu H, et al. A critical assessment of transcranial doppler screening rates in a large pediatric sickle cell center: opportunities to improve healthcare quality. Pediatr Blood Cancer. 2008;51(5):647–651 [DOI] [PubMed] [Google Scholar]
- 25. Cabana MD, Kanter J, Marsh AM, et al. Barriers to pediatric sickle cell disease guideline recommendations. Glob Pediatr Heal. 2019;6:2333794X19847026. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Reeves SL, Fullerton HJ, Dombkowski KJ, et al. Physician attitude, awareness, and knowledge regarding guidelines for transcranial Doppler screening in sickle cell disease. Clin Pediatr (Phila). 2015;54(4):336–345 [DOI] [PubMed] [Google Scholar]
- 27. Lee L, Smith-Whitley K, Banks S, Puckrein G. Reducing health care disparities in sickle cell disease: a review. Public Health Rep. 2019;134(6):599–607 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Freed GL. A missed opportunity to address a national shame: the case of sickle cell disease in the United States. JAMA Pediatr. 2019;173(8):715–716 [DOI] [PubMed] [Google Scholar]
- 29. Reeves SL, Madden B, Wu M, et al. Performance of ICD-10-CM diagnosis codes for identifying children with sickle cell anemia. Health Serv Res. 2020;55(2):310–317 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Snyder AB, Zhou M, Theodore R, et al. Improving an administrative case definition for longitudinal surveillance of sickle cell disease. Public Health Rep. 2019;134(3):274–281 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Reeves S, Garcia E, Kleyn M, et al. Identifying sickle cell disease cases using administrative claims. Acad Pediatr. 2014;14(5 Suppl):S61–S67 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Neunert CE, Gibson RW, Lane PA, et al. Determining adherence to quality indicators in sickle cell anemia using multiple data sources. Am J Prev Med. 2016;51(1 Suppl 1):S24–S30 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Teach SJ, Lillis KA, Grossi M. Compliance with penicillin prophylaxis in patients with sickle cell disease. Arch Pediatr Adolesc Med. 1998;152(3):274–278 [DOI] [PubMed] [Google Scholar]
- 34. Mvundura M, Amendah D, Kavanagh PL, et al. Health care utilization and expenditures for privately and publicly insured children with sickle cell disease in the United States. Pediatr Blood Cancer. 2009;53(4):642–646 [DOI] [PubMed] [Google Scholar]
- 35. Raphael JL, Dietrich CL, Whitmire D, et al. Healthcare utilization and expenditures for low income children with sickle cell disease. Pediatr Blood Cancer. 2009;52(2):263–267 [DOI] [PubMed] [Google Scholar]
- 36. Nero AC, Akuete K, Leasure Reeves S, Dombkowski KJ. Pneumococcal vaccination rates in children with sickle cell disease. J Public Health Manag Pract. 2014;20(6):587–590 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.