Letters
Substandard Vaccination Compliance and the 2015 Measles Outbreak The ongoing measles outbreak linked to the Disneyland Resort in Anaheim, California, shines a glaring spotlight on our nation’s growing antivaccination movement and the prevalence of vaccination-hesitant parents. Although the index case has not yet been identified, the outbreak likely started sometime between December 17 and 20, 2014.1,2 Rapid growth of cases across the United States indicates that a substantial percentage of the exposed population may be susceptible to infection due to lack of, or incomplete, vaccination. Herein, we attempt to analyze existing, publicly available outbreak data to assess the potential role of suboptimal vaccination coverage in the population.
Figure. Estimated Measles, Mumps, and Rubella (MMR) Vaccination Rates Among the Exposed Population Associated With the 2015 Measles Outbreak 0.90 0.85
Estimated MMR Vaccination Rate
RESEARCH LETTER
0.80 0.75 0.70 0.65 0.60
0.50 0.45
Methods | Without vaccination, measles is a highly contagious disease; estimates of the basic reproductive number (R0) from the prevaccination era range from 11 to 18.3 R0 is the mean number of secondary infections per infectious agent that occurs during the course of the entire infectious period in a population that is 100% susceptible at time 0.3 However, when a portion of the population is immune to a given disease, effective reproductive number (RE) is observed instead of R0.3 When historical values of R0 are available and RE can be approximated, the rate of immunity in the exposed population can be estimated.3 Furthermore, in situations where immunity is primarily conferred via vaccination, rate of immunity (I) and rate of vaccination effectiveness (VE) can be used to estimate rate of vaccination (V) in the exposed population as follows:
V=
1 − (RE /R0) I = VE VE
To estimate the vaccination rate in the context of the 2015 measles outbreak, cumulative incidence data were obtained via the California Department of Public Health and HealthMap media alerts.2,4 We used the incidence decay and exponential adjustment (IDEA) method to approximate the effective reproductive number (RE).5 Historically, the serial interval (SI) associated with measles is 10 to 14 days; thus, this range was used to parameterize the model.6 Results | Using nonlinear optimization, RE was solved for SI= 10, 12, and 14 days at 3.2, 4.1, and 5.8, respectively. Measles, mumps, and rubella (MMR) vaccination rates among the exposed population were then estimated using RE = 3.2, 4.1, and 5.8; prevaccination era values of R0; and a vaccination effectiveness (VE) of 95%.3 Over the range of R0 = [11, 18], the estimated vaccination rates vary from 75% to 86% when RE = 3.2, from 66% to 81% when RE = 4.1, and from 50% to 71% when RE = 5.8 (Figure). 494
Estimated effective reproductive No. 3.2 4.1 5.8
0.55
11
12
13
14
15
16
17
18
Basic Reproductive No.
Vaccination rates were estimated using 3 approximated values for effective reproductive number (3.2, 4.1, and 5.8) and 8 well-established values for basic reproductive number (11-18). A vaccination effectiveness of 95% was used for all scenarios.
Discussion | This preliminary analysis indicates that substandard vaccination compliance is likely to blame for the 2015 measles outbreak. Our study estimates that MMR vaccination rates among the exposed population in which secondary cases have occurred might be as low as 50% and likely no higher than 86%. Given the highly contagious nature of measles, vaccination rates of 96% to 99% are necessary to preserve herd immunity and prevent future outbreaks.3 Even the highest estimated vaccination rates from our model fall well below this threshold. While data on MMR vaccination rates are available, coverage is often calculated at the state or county level and may not be granular enough to assess risk in an outbreak situation; this is especially the case for outbreaks originating at a tourist destination, where vaccination coverage among visitors is highly heterogeneous. Clearly, MMR vaccination rates in many of the communities that have been affected by this outbreak fall below the necessary threshold to sustain herd immunity, thus placing the greater population at risk as well. Maimuna S. Majumder, MPH Emily L. Cohn, MPH Sumiko R. Mekaru, DVM, PhD Jane E. Huston, MPH John S. Brownstein, PhD Author Affiliations: Engineering Systems Division, Massachusetts Institute of Technology, Cambridge (Majumder); Computational Epidemiology Group, Children’s Hospital Informatics Program, Division of Emergency Medicine,
JAMA Pediatrics May 2015 Volume 169, Number 5 (Reprinted)
Copyright 2015 American Medical Association. All rights reserved.
Downloaded From: http://archpedi.jamanetwork.com/ by a Penn State Milton S Hershey Med Ctr User on 05/22/2015
jamapediatrics.com
Letters
Boston Children’s Hospital, Boston, Massachusetts (Majumder, Cohn, Mekaru, Huston, Brownstein); Department of Pediatrics, Harvard Medical School, Boston, Massachusetts (Brownstein). Corresponding Author: Maimuna S. Majumder, MPH, Computational Epidemiology Group, Children’s Hospital Informatics Program, Division of Emergency Medicine, Children’s Hospital Boston, 1 Autumn St, Boston, MA 02215 ([email protected]). Published Online: March 16, 2015. doi:10.1001/jamapediatrics.2015.0384. Author Contributions: Mss Majumder and Cohn had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Majumder, Cohn, Mekaru, Brownstein. Acquisition, analysis, or interpretation of data: All authors. Drafting of the manuscript: All authors. Critical revision of the manuscript for important intellectual content: All authors. Statistical analysis: Majumder. Obtained funding: Brownstein. Administrative, technical, or material support: Cohn, Mekaru, Huston. Study supervision: Brownstein. Conflict of Interest Disclosures: None reported. Funding/Support: This work was supported by grant R01LM010812 from the National Library of Medicine. Role of the Funder/Sponsor: The funder had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication. 1. Centers for Disease Control and Prevention. US multi-state measles outbreak, December 2014–January 2015. http://emergency.cdc.gov/HAN/han00376.asp. Accessed January 28, 2015. 2. California Department of Public Health. Measles. http://www.cdph.ca.gov /HealthInfo/discond/Pages/Measles.aspx. Accessed January 28, 2015. 3. Plans-Rubió P. Evaluation of the establishment of herd immunity in the population by means of serological surveys and vaccination coverage. Hum Vaccin Immunother. 2012;8(2):184-188. 4. HealthMap. HealthMap website. http://www.healthmap.org/en/. Accessed January 28, 2015. 5. Fisman DN, Hauck TS, Tuite AR, Greer AL. An IDEA for short term outbreak projection: nearcasting using the basic reproduction number. PLoS One. 2013;8 (12):e83622. 6. Fine PEM. Herd immunity: history, theory, practice. Epidemiol Rev. 1993;15(2): 265-302.
Effect of Dependent Coverage Expansion of the Affordable Care Act on Health and Access to Care for Young Adults Beginning September 23, 2010, the dependent coverage expansion in the Affordable Care Act allowed adults younger than 26 years to obtain insurance through their parents. In a JAMA Pediatrics article, Kotagal and colleagues1 concluded that this provision had “limited impact on health and access.” We reassessed this question using the same data set but a larger sample size and different methods. Methods | Our study used deidentified publically available data from the 2005 to 2012 Behavioral Risk Factor Surveillance System and was deemed exempt from review by the Institutional Review Board of the Harvard School of Public Health. Following prior work,2-4 we used a difference in differences approach to compare coverage, health, and access for individuals aged 19 to 25 years (treatment group) and 26 to 34 years (control group). jamapediatrics.com
Outcomes included having any health insurance, selfreported health (excellent/very good/good vs fair/poor), access to a usual source of care, a routine checkup in the past year, and the inability to see a physician in the past year because of cost. Multivariate linear probability models controlled for a monthly trend; state, year, and calendar month; and an indicator for the enactment period (March 30, 2010-September 22, 2010). The postexpansion period began September 23, 2010. Because economic conditions may have affected coverage rates differentially for younger and older adults, we controlled for state unemployment rates. We also controlled for race/ethnicity, sex, income, marital status, education, employment, and cellular telephone use because cellular telephones were added to the Behavioral Risk Factor Surveillance System in 2011 and the use may have differed by age. Analyses used Behavioral Risk Factor Surveillance System survey weights and accounted for the complex survey design. Two-sided P values less than .05 indicated statistical significance. Analyses were performed with Stata 13 (StataCorp). Results | The sample (n = 456 966) was 50.5% male and 70.8% white. Group demographics were similar across the intervention and control groups except for marital status and education (Table 1). Compared with the control group, the dependent coverage provision was associated with an increase of 6.6 percentage points (95% CI, 5.4-7.7) in the probability of insurance coverage among individuals aged 19 to 25 years (P < .001) and a decrease of 0.8 percentage points (95% CI, 0.0-1.6) in fair or poor self-reported health (P = .04). Compared with the control group, the proportion of young adults with a usual source of care increased by 2.4 percentage points (P < .001; 95% CI, 1.1-3.6) while the proportion of young adults unable to see a physician because of cost declined by 1.9 percentage points (P = .001; 95% CI, 0.8-3.0). There was no statistically significant change in the percentage of young adults who reported a routine checkup in the previous year (Table 2). Comparison of trends before 2010 in these outcomes showed no significant differences by age group. Discussion | We found that the dependent coverage provision was associated with improved self-reported health and access to health care among young adults aged 19 to 25 years compared with a control group of older adults who experienced worsening coverage and access to care during this period. Another recent analysis of the Behavioral Risk Factor Surveillance System reported that the provision did not significantly affect health and access; however, we found different results primarily owing to increased power from a larger sample. Our findings were consistent with previous evidence from several data sources that the Affordable Care Act’s dependent coverage expansion improved self-reported health and access to care.2,5,6 A key limitation of this study was that any factors changing differentially across time for younger adults vs the control group could have biased our findings; however, our ad(Reprinted) JAMA Pediatrics May 2015 Volume 169, Number 5
Copyright 2015 American Medical Association. All rights reserved.
Downloaded From: http://archpedi.jamanetwork.com/ by a Penn State Milton S Hershey Med Ctr User on 05/22/2015
495
Substandard Vaccination Compliance and the 2015 Measles Outbreak The ongoing measles outbreak linked to the Disneyland Resort in Anaheim, California, shines a glaring spotlight on our nation’s growing antivaccination movement and the prevalence of vaccination-hesitant parents. Although the index case has not yet been identified, the outbreak likely started sometime between December 17 and 20, 2014.1,2 Rapid growth of cases across the United States indicates that a substantial percentage of the exposed population may be susceptible to infection due to lack of, or incomplete, vaccination. Herein, we attempt to analyze existing, publicly available outbreak data to assess the potential role of suboptimal vaccination coverage in the population.
Figure. Estimated Measles, Mumps, and Rubella (MMR) Vaccination Rates Among the Exposed Population Associated With the 2015 Measles Outbreak 0.90 0.85
Estimated MMR Vaccination Rate
RESEARCH LETTER
0.80 0.75 0.70 0.65 0.60
0.50 0.45
Methods | Without vaccination, measles is a highly contagious disease; estimates of the basic reproductive number (R0) from the prevaccination era range from 11 to 18.3 R0 is the mean number of secondary infections per infectious agent that occurs during the course of the entire infectious period in a population that is 100% susceptible at time 0.3 However, when a portion of the population is immune to a given disease, effective reproductive number (RE) is observed instead of R0.3 When historical values of R0 are available and RE can be approximated, the rate of immunity in the exposed population can be estimated.3 Furthermore, in situations where immunity is primarily conferred via vaccination, rate of immunity (I) and rate of vaccination effectiveness (VE) can be used to estimate rate of vaccination (V) in the exposed population as follows:
V=
1 − (RE /R0) I = VE VE
To estimate the vaccination rate in the context of the 2015 measles outbreak, cumulative incidence data were obtained via the California Department of Public Health and HealthMap media alerts.2,4 We used the incidence decay and exponential adjustment (IDEA) method to approximate the effective reproductive number (RE).5 Historically, the serial interval (SI) associated with measles is 10 to 14 days; thus, this range was used to parameterize the model.6 Results | Using nonlinear optimization, RE was solved for SI= 10, 12, and 14 days at 3.2, 4.1, and 5.8, respectively. Measles, mumps, and rubella (MMR) vaccination rates among the exposed population were then estimated using RE = 3.2, 4.1, and 5.8; prevaccination era values of R0; and a vaccination effectiveness (VE) of 95%.3 Over the range of R0 = [11, 18], the estimated vaccination rates vary from 75% to 86% when RE = 3.2, from 66% to 81% when RE = 4.1, and from 50% to 71% when RE = 5.8 (Figure). 494
Estimated effective reproductive No. 3.2 4.1 5.8
0.55
11
12
13
14
15
16
17
18
Basic Reproductive No.
Vaccination rates were estimated using 3 approximated values for effective reproductive number (3.2, 4.1, and 5.8) and 8 well-established values for basic reproductive number (11-18). A vaccination effectiveness of 95% was used for all scenarios.
Discussion | This preliminary analysis indicates that substandard vaccination compliance is likely to blame for the 2015 measles outbreak. Our study estimates that MMR vaccination rates among the exposed population in which secondary cases have occurred might be as low as 50% and likely no higher than 86%. Given the highly contagious nature of measles, vaccination rates of 96% to 99% are necessary to preserve herd immunity and prevent future outbreaks.3 Even the highest estimated vaccination rates from our model fall well below this threshold. While data on MMR vaccination rates are available, coverage is often calculated at the state or county level and may not be granular enough to assess risk in an outbreak situation; this is especially the case for outbreaks originating at a tourist destination, where vaccination coverage among visitors is highly heterogeneous. Clearly, MMR vaccination rates in many of the communities that have been affected by this outbreak fall below the necessary threshold to sustain herd immunity, thus placing the greater population at risk as well. Maimuna S. Majumder, MPH Emily L. Cohn, MPH Sumiko R. Mekaru, DVM, PhD Jane E. Huston, MPH John S. Brownstein, PhD Author Affiliations: Engineering Systems Division, Massachusetts Institute of Technology, Cambridge (Majumder); Computational Epidemiology Group, Children’s Hospital Informatics Program, Division of Emergency Medicine,
JAMA Pediatrics May 2015 Volume 169, Number 5 (Reprinted)
Copyright 2015 American Medical Association. All rights reserved.
Downloaded From: http://archpedi.jamanetwork.com/ by a Penn State Milton S Hershey Med Ctr User on 05/22/2015
jamapediatrics.com
Letters
Boston Children’s Hospital, Boston, Massachusetts (Majumder, Cohn, Mekaru, Huston, Brownstein); Department of Pediatrics, Harvard Medical School, Boston, Massachusetts (Brownstein). Corresponding Author: Maimuna S. Majumder, MPH, Computational Epidemiology Group, Children’s Hospital Informatics Program, Division of Emergency Medicine, Children’s Hospital Boston, 1 Autumn St, Boston, MA 02215 ([email protected]). Published Online: March 16, 2015. doi:10.1001/jamapediatrics.2015.0384. Author Contributions: Mss Majumder and Cohn had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Majumder, Cohn, Mekaru, Brownstein. Acquisition, analysis, or interpretation of data: All authors. Drafting of the manuscript: All authors. Critical revision of the manuscript for important intellectual content: All authors. Statistical analysis: Majumder. Obtained funding: Brownstein. Administrative, technical, or material support: Cohn, Mekaru, Huston. Study supervision: Brownstein. Conflict of Interest Disclosures: None reported. Funding/Support: This work was supported by grant R01LM010812 from the National Library of Medicine. Role of the Funder/Sponsor: The funder had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication. 1. Centers for Disease Control and Prevention. US multi-state measles outbreak, December 2014–January 2015. http://emergency.cdc.gov/HAN/han00376.asp. Accessed January 28, 2015. 2. California Department of Public Health. Measles. http://www.cdph.ca.gov /HealthInfo/discond/Pages/Measles.aspx. Accessed January 28, 2015. 3. Plans-Rubió P. Evaluation of the establishment of herd immunity in the population by means of serological surveys and vaccination coverage. Hum Vaccin Immunother. 2012;8(2):184-188. 4. HealthMap. HealthMap website. http://www.healthmap.org/en/. Accessed January 28, 2015. 5. Fisman DN, Hauck TS, Tuite AR, Greer AL. An IDEA for short term outbreak projection: nearcasting using the basic reproduction number. PLoS One. 2013;8 (12):e83622. 6. Fine PEM. Herd immunity: history, theory, practice. Epidemiol Rev. 1993;15(2): 265-302.
Effect of Dependent Coverage Expansion of the Affordable Care Act on Health and Access to Care for Young Adults Beginning September 23, 2010, the dependent coverage expansion in the Affordable Care Act allowed adults younger than 26 years to obtain insurance through their parents. In a JAMA Pediatrics article, Kotagal and colleagues1 concluded that this provision had “limited impact on health and access.” We reassessed this question using the same data set but a larger sample size and different methods. Methods | Our study used deidentified publically available data from the 2005 to 2012 Behavioral Risk Factor Surveillance System and was deemed exempt from review by the Institutional Review Board of the Harvard School of Public Health. Following prior work,2-4 we used a difference in differences approach to compare coverage, health, and access for individuals aged 19 to 25 years (treatment group) and 26 to 34 years (control group). jamapediatrics.com
Outcomes included having any health insurance, selfreported health (excellent/very good/good vs fair/poor), access to a usual source of care, a routine checkup in the past year, and the inability to see a physician in the past year because of cost. Multivariate linear probability models controlled for a monthly trend; state, year, and calendar month; and an indicator for the enactment period (March 30, 2010-September 22, 2010). The postexpansion period began September 23, 2010. Because economic conditions may have affected coverage rates differentially for younger and older adults, we controlled for state unemployment rates. We also controlled for race/ethnicity, sex, income, marital status, education, employment, and cellular telephone use because cellular telephones were added to the Behavioral Risk Factor Surveillance System in 2011 and the use may have differed by age. Analyses used Behavioral Risk Factor Surveillance System survey weights and accounted for the complex survey design. Two-sided P values less than .05 indicated statistical significance. Analyses were performed with Stata 13 (StataCorp). Results | The sample (n = 456 966) was 50.5% male and 70.8% white. Group demographics were similar across the intervention and control groups except for marital status and education (Table 1). Compared with the control group, the dependent coverage provision was associated with an increase of 6.6 percentage points (95% CI, 5.4-7.7) in the probability of insurance coverage among individuals aged 19 to 25 years (P < .001) and a decrease of 0.8 percentage points (95% CI, 0.0-1.6) in fair or poor self-reported health (P = .04). Compared with the control group, the proportion of young adults with a usual source of care increased by 2.4 percentage points (P < .001; 95% CI, 1.1-3.6) while the proportion of young adults unable to see a physician because of cost declined by 1.9 percentage points (P = .001; 95% CI, 0.8-3.0). There was no statistically significant change in the percentage of young adults who reported a routine checkup in the previous year (Table 2). Comparison of trends before 2010 in these outcomes showed no significant differences by age group. Discussion | We found that the dependent coverage provision was associated with improved self-reported health and access to health care among young adults aged 19 to 25 years compared with a control group of older adults who experienced worsening coverage and access to care during this period. Another recent analysis of the Behavioral Risk Factor Surveillance System reported that the provision did not significantly affect health and access; however, we found different results primarily owing to increased power from a larger sample. Our findings were consistent with previous evidence from several data sources that the Affordable Care Act’s dependent coverage expansion improved self-reported health and access to care.2,5,6 A key limitation of this study was that any factors changing differentially across time for younger adults vs the control group could have biased our findings; however, our ad(Reprinted) JAMA Pediatrics May 2015 Volume 169, Number 5
Copyright 2015 American Medical Association. All rights reserved.
Downloaded From: http://archpedi.jamanetwork.com/ by a Penn State Milton S Hershey Med Ctr User on 05/22/2015
495
