HCP Advisory Board: Reflecting on Research and Planning Ahead

November 10, 2016

Ateev Mehrotra Presents

October 2016—Every six months, the HCP Advisory Board meets to discuss critical problems in health policy and recent intriguing research that the faculty have published. This meeting was no exception.

Richard Frank kicked off the proceedings with a discussion of the role that the Office of the Assistant Secretary for Planning and Evaluation (ASPE) played in policy analysis of key issues during the Obama administration. Richard served as the Assistant Secretary at ASPE for two years.  The discussion was moderated by Leonard Schaeffer, chair of the USC Price School of Public Policy (pictured, below right, with Joseph Newhouse).

Frank detailed how his teams at ASPE supported decision-making related to Medicaid expansion, the King vs. Burwell Supreme Court case and whether federal Marketplaces should offer Joe Newhouse, Leonard Schaeffersubsidies to low income individuals, and work on the HHS opioid strategy.

Sherri Rose outlined her work on health policy data science and machine learning, presenting to the Board for the first time her foundational research in a new area she dubs “Computational Health Economics.” One notable project discussed was her machine learning framework for plan payment risk adjustment, which has the potential to mitigate incentives to health plans to “upcode” reported diagnoses of patients.

Another discussion covered Rose's new study, which demonstrated that some prescription drug classes are predictive of unprofitable enrollees. These classes may need special attention from regulators in health insurance market design.

Rose closed her presentation with a recent project from another component of her portfolio: computational health outcomes research. In joint work with Sharon-Lise Normand, Rose is developing statistical techniques to study the comparative effectiveness of multiple drug-eluting stents. Their early results show that even relatively similar devices from the same manufacturer may have disparate effectiveness with regard to a composite safety outcome. 

Ateev Mehrotra presented recent research on price transparency. His work has attempted to answer the question: why are price transparency efforts not being used? The answer may be, in part, because the underlying systems are still too complex for patients to navigate.

Mehrotra has examined two tools, Truven and Castlight, to see whether consumers would utilize the tool to save money on upcoming care. The work to date has demonstrated that few people sign up for such tools and even lower consistent usage—and, most importantly, there is no reduction in spending.

Nicole Maestas recounted new research on physician prescribing of opioid pain relievers. While medical use of opioids has risen dramatically over the last two decades, the population prevalence of pain has increased very little. Maestas and colleagues undertook a comprehensive review of the medical indications used by physicians to support treatment with opioids. They sought to answer the question: when is opioid therapy appropriate?

Data from two representative surveys were used to identify ambulatory care office visits where an opioid was prescribed. They found that at 35 percent of such visits, the opioid was prescribed with no recorded pain diagnosis, implying that at as many 121 million doctor visits between 2006 and 2011, an opioid was prescribed without any documentation of pain.

At present, physicians are not required to provide justification for medical treatment with opioids. But without clear documentation, it is impossible to assess whether the treatment is appropriate or not.

Finally, Tom McGuire told the Board about an opportunity for him and Tim Layton to work with policymakers at The Center for Consumer Information & Insurance Oversight on high-cost enrollees in Marketplaces.

For 2014-16, the Marketplaces included a reinsurance mechanism that subsidized Marketplace plans for the extremely high cost individuals they enrolled. This mechanism is set to expire at the end of 2016. Many insurers have cited its expiration as a key reason for premium hikes and their exit of the market.

Layton and McGuire worked with policymakers at CCIIO to develop an alternative way to protect Marketplace issuers from high cost cases via the existing permanent risk adjustment system. The methods are both politically feasible without new legislation and in some ways superior to the current reinsurance system.

The researchers showed via simulation that the alternative methods are quite effective at reducing the risk faced by small issuers. CMS is planning on implementing these new methods in the coming years.

Enjoy content like this? Sign up for our newsletter