How important is the structure of school vaccine requirement opt-out provisions? Evidence from Washington, DC's HPV vaccine requirement

https://doi.org/10.1016/j.jhealeco.2021.102480Get rights and content

Abstract

Recent increases in vaccine-preventable diseases have led policymakers to reconsider the scope of vaccine requirement exemptions. Yet eliminating these provisions is politically difficult. Beginning in 2009, sixth grade girls in Washington, DC were required to receive the HPV vaccine or submit a one-time opt-out form. In 2014, the requirement was expanded to all students grades 6–12, and those not vaccinating were required to opt-out annually. I show that the movement from a one-time opt-out provision to an annual requirement increased the probability that teen girls in Washington, DC initiated HPV vaccination by 11 percentage points. Teen boys were 20 percentage points more likely to be vaccinated. Back-of-the-envelope calculations suggest 7 fewer cases of cervical cancer and 41 fewer cases of oropharyngeal cancer for the 33,000 enrolled during the 2017/2018 year. Using the initial value of cancer care and the value of a statistical life year, my estimates imply nearly $36 million in savings compared to $1.5 million spent on vaccination. In generalizing these results to other states, effect sizes even one-tenth the size of my most conservative estimate would imply meaningful reductions in the nationwide incidence of HPV-related cancers.

Introduction

Seventy-nine million Americans are infected with human papillomavirus (HPV) making it the most common sexually transmitted infection in the United States (CDC, 2017). Approximately 80 percent of sexually active people will contract HPV during their lives (Cleveland Clinic, 2018). HPV is a group of more than 200 related viruses (National Cancer Institute, 2020a), and HPV types 16 and 18 are responsible for 66 percent of all cervical cancers in the US (CDC, 2018). Over 40,000 people annually are diagnosed with an HPV-related cancer (Van Dyne et al., 2018). Approximately 11 million men are currently thought to have oral HPV (Deshmukh et al., 2017), and the incidence of male oral cancer exceeds the incidence of cervical cancer in women (Mourad et al., 2017).

Unlike most cancers, there is a highly effective vaccine that provides near complete protection against some of the most dangerous strains of HPV (Villa et al., 2005; Villa et al., 2006). Yet in 2018, only 68 percent of teens had initiated HPV vaccination and only 55 percent were fully vaccinated.1 Over the last decade, more than 40 states have introduced legislation on HPV vaccination, many of which sought to leverage the success of other school-entry vaccine requirements by mandating HPV vaccination (Barraza et al., 2016). However, these mandates have proven politically difficult. For example, the New York state Parent Teachers Association announced opposition to a recent bill which would require middle school students to receive the HPV vaccine, despite re-expressing its support for other mandated vaccinations (Times Herald-Record, 2020). Only three states and the District of Columbia have successfully adopted HPV vaccine school requirements, and vaccination proponents argue that their broad opt-out provisions limit their efficacy (Reynolds, 2012).2

In this paper, I provide the first causal evaluation of how Washington, DC's 2014 HPV vaccine school requirement affected vaccine coverage. I find that the requirement increased the probability that a teen was fully vaccinated against HPV by nearly 20 percentage points—a 71 percent increase over the 2013 vaccination rate. I use an event study specification to show that this increase was not driven by pre-existing trends in vaccination. The use of placebo permutation tests confirms that the increase is larger than would be expected by chance. I also show that the estimate is robust to employing a synthetic control design. Back-of-the-envelope calculations imply that this requirement will directly result in 7 fewer cases of cervical cancer and 41 fewer cases of oropharyngeal cancer for the 33,000 students enrolled in Washington, DC schools during the 2017/2018 academic year. After accounting for the initial costs of cervical and oropharyngeal cancer care, as well as the statistical value of the life years lost, my estimates imply nearly $36 million dollars in reduced cancer savings compared to the $1.5 million it cost to vaccinate these students.

While important for policymakers, generalizing these estimates to the broader US requires caution. For one, the HPV vaccine initiation rate in the US in 2018 was higher than Washington, DC's initiation rate immediately prior to the policy change (68 percent vs. 62 percent). Moreover, vaccine initiation rates between girls and boys have converged. In 2018, 70 percent of girls and 67 percent of boys had received at least one shot of the HPV vaccine. As a result, school requirements may no longer induce larger increases in take-up for teen boys than teen girls. Yet even subject to these caveats, considering how these estimates could generalize is a useful exercise. There are 30 million 6–12th grade students in the US (National Center for Education Statistics, 2018). Applying my most conservative estimated increase in vaccine initiation (10.9 percentage points) still yields approximately 3.27 million more vaccinated students and over 6000 fewer cases of cervical cancer.

In addition to learning about ways to improve HPV vaccination, the Washington, DC policy change offers broader insights into the importance of how vaccine mandates are implemented. While respondents view vaccine school requirements more favorably if they contain opt-out provisions, these provisions likely reduce the mandates’ efficacy (Calo et al., 2016). Indeed, there is a positive association between the ease of opting-out of vaccination and the number of exemptions granted in a state (Blank et al., 2013) and repealing non-medical vaccine exemptions is associated with greater vaccine coverage (Nyathi et al., 2019; Richmine, Dor, and Moghtaderi, 2019). Beginning in 2009, sixth grade girls in DC were required to (i) receive the HPV vaccine or (ii) submit a one-time opt-out form. In 2014, the requirement was expanded to 6th grade boys and all students up through 12th grade. Additionally, all those not vaccinated were required to opt-out annually. As such, the treatment for teen girls was not a movement from “no requirement” to an “HPV vaccine requirement,” but rather a change from a one-time opt-out in 6th grade to an annual opt-out requirement all the way through 12th grade.

For teen girls, I find a 11-percentage point increase in HPV vaccine initiation and a 20-percentage point increase in vaccine completion. This pattern suggests that the annual reminder induced girls who had previously opted-out of HPV vaccination to receive their first shot, while also encouraging girls who had initiated vaccination to complete the vaccine series. In support of this pathway, I show that DC's 2014 HPV vaccine requirement reduced the probability that a teen girl had initiated but not completed the HPV vaccine. Additionally, I find that the 2014 requirement increased the probability that teen girls completed HPV vaccination within the recommended timeframe conditional on initiation. In contrast to these results for teen girls, I find that the increase in vaccine completion for teen boys is fully explained by an increase in vaccine initiation.

In Section 2, I discuss the history of the HPV vaccine, state HPV vaccine school requirements, and the existing literature on policies promoting HPV vaccination. In Section 3, I provide an overview of the NIS-Teen data and show descriptively that DC experienced a dramatic increase in HPV vaccination in the post-school requirement period. I then discuss my identification strategies, as well as the difficulties of conducting statistical inference with a single treated unit. In Section 4, I show that DC's 2014 HPV vaccine school requirement led to a large statistically significant increase in HPV vaccination, and I explore how the relationship varied by sex, grade level, race/ethnicity, and mother's educational attainment. In Section 5, I use my estimates to project the number of cancers prevented due to the 2014 mandate, and I estimate the cost savings associated with these reductions. Finally, I conclude in Section 6 by discussing the policy implications of my estimates and areas for future research.

Section snippets

Policy background & existing literature

In this section, I provide a history of the HPV vaccine within the United States. I discuss when various age and sex groups were eligible to receive the vaccine, as well as state HPV vaccine school requirements. Next, I summarize the literature on vaccine mandates with a focus on the structure of these requirements. Finally, I discuss existing work on ways to improve HPV vaccination.

Data & methodology

In this section, I describe the specifics of the NIS-Teen data. I show descriptively that the share of DC teens vaccinated against HPV increased dramatically concurrent with the 2014 HPV vaccine school requirement. The average vaccination rate in all other states was unchanged. I then describe my two empirical strategies intended to test whether this is a causal effect—difference-in-differences and synthetic control.

Results

In this section, I show that DC's 2014 HPV vaccine school requirement led to a large increase in the probability that teens were vaccinated against HPV. Using an event study specification, I show that this relationship was not driven by a pre-existing trend in immunization. I also document an increase in HPV vaccine initiation, intentions to vaccinate, and the likelihood of having been recommended the vaccine. Finally, I show that this pattern is robust to utilizing a synthetic control

Implied reductions in cancer and health care costs

To conceptualize the economic and public health benefits of the HPV vaccine school requirement, it is worth considering how many cases of cancer this policy change may have prevented. The American Cancer Society (2020a) indicates that the lifetime risk of developing cervical cancer is 0.63 percent, and the CDC (2020b) estimates that HPV may be responsible for more than 90 percent of these cancers.20

Discussion

Throughout this paper, I have shown that Washington, DC's 2014 HPV vaccine school requirement increased the probability that a teen was vaccinated against HPV by nearly 20 percentage points. Only 28 percent of DC teens had completed the HPV vaccine in 2013, so the estimated effect is large in both absolute magnitude and as a percentage change from the pre-period level. Using an event study specification, I show that this increase was not due to pre-existing trends in the probability of

Declaration of Competing Interest

None.

References (63)

  • American Cancer Society. (2020a). Lifetime risk of developing or dying from cancer. Accessed at:...
  • American Cancer Society. (2020b). Key statistics for cervical cancer. Accessed at:...
  • American Cancer Society. (2020c). Survival rates for cervical cancer. Accessed at:...
  • American Cancer Society (2021a). Key statistics for oral cavity and oropharyngeal cancer. Accessed at:...
  • American Cancer Society (2021b). Survival rates for oral cavity and oropharyngeal cancer. Accessed at:...
  • L. Barraza et al.

    Human papillomavirus and mandatory immunization laws: what can we learn from early mandates?

    Public Health Rep.

    (2016)
  • N.R. Blank et al.

    Exempting schoolchildren from immunizations: states with few barriers had highest rates of nonmedical exemptions

    Health Aff.

    (2013)
  • T.C. Buchmueller et al.

    The effect of an employer health insurance mandate on health insurance coverage and the demand for labor: evidence from Hawaii

    Am. Econ. J.

    (2011)
  • W.A. Calo et al.

    Parents’ support for school-entry requirements for human papillomavirus vaccination: a national study

    Cancer, Epidemiol., Biomark. Prevent.

    (2016)
  • C.S. Carpenter et al.

    Direct and spillover effects of middle school vaccination requirements

    Am. Econ. J.

    (2019)
  • CDC. (2012). Principles of Epidemiology in Public Health Practice. Accessed at:...
  • CDC. (2017). Genital HPV infection - CDC fact sheet. Accessed at: https://www.cdc.gov/std/hpv/HPV-FS-print.pdf (July...
  • CDC. (2018). Sexually Transmitted Disease Surveillance. Accessed at:...
  • CDC. (2020a). Recommended child and adolsecent immunization schedule for 18 years or younger. Accessed at:...
  • CDC. (2020b). HPV-Associated cancer statistics. Accessed at:...
  • B. Churchill

    Insurance Coverage, Provider Contact, and Take-Up of the HPV Vaccine

    (2020)
  • HPV (Human Papilloma Virus)

    (2018)
  • S. Cunningham et al.

    Decriminalizing indoor prostitution: implications for sexual violence and public health

    Rev. Econ. Stud.

    (2018)
  • DC Office of the State Superintendent of Education. (2018). 2017-2018 School Year Enrollment Audit Report and Data....
  • K.S. Deshmukh et al.

    Differences in prevalence between sexes and concodance with genital human papillomavirus infection, NHANES 2011 to 2014

    Ann. Intern. Med.

    (2017)
  • FDA News Release: FDA Licenses New Vaccine For Prevention of Cervical Cancer and Other Diseases in Females Caused By Human Papillomavirus

    (2006)
  • Cited by (7)

    • Effect of vaccine recommendations on consumer and firm behavior

      2024, Journal of Policy Analysis and Management
    View all citing articles on Scopus
    View full text