In the past two years, the opioid epidemic in the US has reached a tipping point. The New York Times reports that in 2016, roughly 64,000 people were killed by opioid-related overdoses, including from prescription painkillers and heroin. How can analysis of real-time longitudinal data play a role in fighting this crisis?
Government Takes Notice
In recent months, the federal government has sat up and begun to take notice of the magnitude of this problem. In October 2017, the President declared a 90-day public health emergency that would “urgently mobilize the federal government to tackle the opioid epidemic,” reports Politico. And while there’s much controversy over how much the declaration has actually accomplished in the months since, HHS announced on January 19 of this year that it would be extending the public health emergency “as a result of the continued consequences of the opioid crisis affecting our nation.” The designation “gives federal health agencies the authority to quickly hire more treatment specialists and reallocate money to strengthen the response to the epidemic,” reports Politico. But even as early as 2012, articles were published reporting on over a 20 year history of an “opioid epidemic with adverse consequences.” If longitudinal databasis were available with clinical and financial data, and had the analytical power to recognize patterns of care and costs, would we be able to recognize adverse issues long before they hit crisis status?
On February 25, new HHS secretary Alex M. Azar II said in a talk given to the National Governors Association that the FDA planned to expand medication-assisted treatment (MAT) for those addicted to opioids. The Substance Abuse and Mental Health Services Administration (SAMHSA) defines MAT as “the use of medications with counseling and behavioral therapies to treat substance use disorders and prevent opioid overdose.” Medication prescribed as part of MAT “operates to normalize brain chemistry, block the euphoric effects of alcohol and opioids, relieve physiological cravings, and normalize body functions without the negative effects of the abused drug,” says SAMHSA.
In his remarks to the National Governor’s Association, Mr. Azar said the FDA planned to “correct a misconception that patients must achieve total abstinence in order for MAT to be considered effective.” In fact, as the Times reports, “the FDA has approved three drugs for opioid treatment — buprenorphine (often known by the brand name Suboxone), methadone, and naltrexone (known by the brand name Vivitrol) — and says they are safe and effective combined with counseling and other support.”
On February 27, multiple senators across party lines introduced the Comprehensive Addiction and Recovery Act 2.0, legislation that would increase funding authorization levels for CARA programs and implement additional policy reforms to combat the opioid crisis. The bill would authorize $1 billion in dedicated resources to evidence-based prevention, enforcement, treatment and recovery programs and impose a three-day limit on initial opioid prescriptions for acute pain. And on February 28, the House Energy and Commerce Health Subcommittee held the first of three planned hearings to examine the opioid crisis and possible legislative solutions. The American Hospital Association reports that the hearings “reviewed several bills pertaining to the Controlled Substances Act.”
But, Everyone is Doing It
“Nearly two decades ago, we were encouraged to be more aggressive about treating pain, often without enough training and support to do so safely,” Surgeon General Vivek H. Murthy, MD, MBA, wrote last April. “This coincided with heavy marketing of opioids to doctors. Many of us were even taught – incorrectly – that opioids are not addictive when prescribed for legitimate pain.”
Today, we know better. According to the Department of Health and Human Services, more people died from drug overdoses in 2014 than in any previous year, with over six out of ten involving an opioid. And the resulting economic impact is equally significant — $55 billion per year in health and social costs related to opioid abuse and $20 billion per year in emergency room and inpatient care for opioid poisonings.
Cornell University recently published an article, “Opioid abuse drops when doctors check patients’ drug history,” about an upcoming study to be published in American Economic Journal: Economic Policy. The study’s co-author Colleen Carey, assistant professor of policy analysis and management in the College of Human Ecology, writes, “The main issue is getting providers to change their prescribing behavior. The majority of opioids that people abuse start in the medical system as a legitimate prescription.” Carey’s co-author is Thomas Buchmueller of the University of Michigan.
According to the Cornell article about the study, until recently, Medicare has had very few legislative tools to curtail the epidemic, and insurance companies have little incentive, because opioids are relatively cheap, costing about $1.60 per day. And opioids don’t hit Medicare insurers in the bottom line, making up only 3 percent of their total drug costs, Carey said.
“If people had been abusing hepatitis C drugs that cost $84,000 for a course of treatment,” she said, “there would have been some action.”
The opioid crisis has also served to heighten awareness about other highly-prescribed medications. There are other drugs that are just as important — or as detrimental — to the entire population when it comes to the misuse of medication (such as antibiotics). While it would be inappropriate to say that this particular “epidemic” of overdoses is related to the observer phenomenon, it is a reminder that we have a responsibility to manage and improve the lives of those who, unfortunately, have become habituated or addicted to any medication.
The Role Longitudinal Data Can Play
In my experience over the past 14 years, I’ve gained a deep appreciation for the true value of a longitudinal dataset. Longitudinal patient data (data from many patients, over many years) allows for the analysis of very specific population cohorts to determine relevant courses of treatment. This is vastly superior to a snapshot view of a patient based on data available in a single episode.
When hospital pharmacy leaders can identify and model drug cost reduction opportunities, evaluate the comparative efficacy of alternative treatments, and analyze utilization patterns to optimize treatment intensity for procedures and diagnoses, healthcare delivery changes. It’s not about penalties, but about having the data to reflect, longitudinally, a pattern that shows insights and outlines potential results so that decision makers can make the best treatment choices. The purpose of understanding the pattern of use and identifying inappropriate or appropriate use of medication is to say, “On a scale where the majority of patients are huddled around the mean, there are also outliers. As a decision maker or combination of professional individuals, who delivers those outliers of care, both good and bad?”
Longitudinal databases also allow institutions to pick their peer groups and measure themselves against like organizations. This exercise may result in the ultimate decision to not use a particular drug (such as a certain type of pain-killer, or antibiotics or any other medication) because it doesn’t really change the outcome of care. Alternatively, it may show that appropriate use of pain medication early on, followed by appropriate follow-up and transition to physical therapy, might be a better approach to managing that pain.
Having the opportunity to look at the larger picture at this more granular level via longitudinal data sets means that clinicians can define across an entire continuum of care what combinations of medications are leading to the best outcomes and what factors are being used to define those baselines, from lower lengths of stay to reduced readmissions.
The Value of Effective Interventions
In light of this crisis, it’s more important than ever to recognize patterns of care and measure medication options against the resources necessary to effect change over time. If one can’t do that, why bother? It’s like screening for a disease that no one can treat.
Narcotics (as well as antibiotics) are often prescribed when they aren’t the best option for patients and may do more harm than good, a recent survey of U.S. physicians suggested. The survey, conducted by The American College of Physicians (ACP) and published in the Annals of Internal Medicine, asked doctors to identify treatments that they see routinely used, despite guidelines recommending against the interventions and little or no value to patients. One likely reason for this overprescribing is that doctors assume, even if the prescription won’t help alleviate the patient’s condition, it won’t hurt them, so they adopt a “better safe than sorry” mindset and prescribe anyway. Having physicians justify the use of the drug on the medical record might help them stop and think before writing that prescription.
“We need to start concentrating on high-value care,” said lead ACP study author Dr. Amir Qaseem, vice president of clinical policy and chair of the high-value care task force for the American College of Physicians.
It's time that health IT and Big Data join with providers and the government in the fight against opioid addiction and many other critical issues contributing to both the financial and clinical crisis of the health delivery system. Knowledge is power—and the knowledge lies in the data.