A Public Health of Consequence—February 2016

In this issue of AJPH, several authors tackle issues of sentinel contemporary public health concern that we felt can inform and inflect the broader health conversation.

Perhaps no issue has taken up more public discussion, and academic health discussion, than the implementation of the Patient Protection and Affordable Care Act (ACA; Pub L No. 111–148). The ACA implementation is directly relevant, first and foremost, to clinical care, and with full roll out of the ACA the nature of medical care in the country will change, in many ways for the better. Since the introduction of the ACA in October 2013, nearly 18 million Americans have gained insurance coverage and nearly 11 million are eligible for coverage in the current enrollment period.¹ Universal health coverage (UHC) has been called, by World Health Organization Director Margaret Chan, “the most powerful idea public health has to offer.”²

While the ACA falls far short of UHC, the engagement of the public health community around issues of insurance coverage linked to the ACA seems apposite and timely. To that end, we enjoyed the article by August et al. in this issue of AJPH.³ August et al. estimate that states’ current Medicaid expansion plans are likely to meet about half the need for coverage among women of reproductive age in need. Centrally though, the authors point out that 2.5 million women in need will remain uncovered, pointing to one of the core points relevant to public health when dealing with enthusiasm about the ACA: there remain marginalized populations in need, and as we do achieve coverage for many, we widen health gaps as hard-to-reach groups fall further behind, becoming “health have-nots.” This suggests the importance of applauding the very real achievements of the ACA but also a redoubling of efforts to ensure equitable availability to health resources for all through the innovative availability of programs—in this case, as the authors correctly suggest, publicly funded contraceptive programs. We also applaud the authors for using existing data and transparent simulation methods across several sensible scenarios to provide estimates of coverage, as well as their transparency in the enumeration of the limitations of their approach and the threats to validity.

One of the tremendous missed opportunities in capitalizing on the full benefits possible with the ACA is the number of states who have refused Medicaid expansion, mainly for political and ideological reasons. There are currently 19 states that have not adopted Medicaid expansion and one state that continues to debate adoption. Callaghan and Jacobs present a fascinating analysis of the competing political forces that have informed these decisions.⁴ They find an important influence of lobbying on state decisions to resist Medicaid, counteracted somewhat by the influence of public interest advocates. We would have liked to better understand how these two forces interact and, more importantly, what the data suggest might be lever points for public interest advocacy to counter the special interest pull that is diminishing the potential of the ACA. We thought that the analytic approach could have been improved, given the distribution of the Medicaid expansion outcome variable provided in Figure 1 of their article (i.e., that > 50% of states were already at the highest level). In their Table 1, Callaghan and Jacobs provide parameter estimates from the application of ordinal logistic regression relating several structural political variables (i.e., lobbying density, political party control) to level of Medicaid expansion that, when exponentiated by the reader, result in point estimates of odds ratios (ORs) of 73.6 and nearly 1400. The presentation of ln(OR)s from an ordinal logistic regression (i.e., the parameter estimates in Table 1) is less intuitive to the reader than would be the OR, and ORs of that magnitude (regardless of significance) should have alerted the authors that perhaps a different analytic approach would have been more appropriate. Overall, we found the simple observation, grounded in data, of countervailing influences on the implementation of health coverage that will save lives, bracing and important.

In keeping with analyses that explore how structural factors influence health around important dimensions, two other articles in this issue tackle policy approaches to mitigating two important epidemics of our time: the smoking epidemic and the more recent opioid epidemic.

Vuolo et al. demonstrate the complementary effects of excise taxes and smoking bans on young adults.⁵ They show that the former is directly linked to decreased regular smoking while the latter influences current, perhaps social, smoking. This observation makes intuitive sense, and we would argue that it suggests the need for both approaches in order both to mitigate initiation of smoking and to reduce transition to chronic heavy smoking in this population. They provide an exemplary description of their analytic framework with careful explication of all variables and choice of models, and the figures made the potentially complex results easily digestible. There is relatively little work such as the article by Vuolo et al. that offers compelling empiric justification for complementary policy approaches, and we see this as an important application of public health analysis to the task of guiding the implementation of policies that can shift the population health curve.

Kennedy-Hendricks et al. consider the impact of state laws and law enforcement in Florida, suggesting that more than 1000 lives were saved with the complementary application of these two approaches to reduce the impact of pill mills in that state.⁶ Importantly, and convincingly, the authors show this through a comparison with a state—North Caroline—that did not have the same programs implemented, providing perhaps robust encouragement for other states to consider these programs as an effective approach to improving the health of the public. The application of multivariate adaptive regression spline (MARS) models, while appropriate, is likely unknown to many public health researchers, with the inner workings daunting if not impenetrable for most of us. Use of this method in appropriate contexts suggests that population health science would benefit from familiarity with this approach and from the development of easily digestible explanations for what analytic hurdles these (and other) methods overcome, and how they do so.

We comment in closing on what is, in our assessment, one of the most pressing public health issues of our time: firearm violence. More people die from firearm injury than from motor vehicle deaths in this country, and comparison with peer countries such as Canada readily suggest that with the adoption of comparable policies the United States could see 25 000 lives a year saved. Loughran et al. tackle this issue through a study of juvenile offenders showing the cognitive fallacy that so often characterizes public debate around this issue.⁷ The authors show that owning a gun increases one’s risk of injury and death, even if it may confer a sense of protection from the same. This is an important message and a key piece to the public health message that we need to ensure enters the broader national discussion, moving us toward a set of policies and regulations that mitigate this preventable epidemic.