Healthcare Analytics, Population Health Management, Healthcare Big Data

Population Health News

Big Data Analytics Cuts Medicaid Opioid Abuse, Ups Patient Safety

- Opioid abuse is often described as an “epidemic” sweeping the nation, killing more than 15,000 patients each year and affecting the lives of a further two million Americans who abuse or depend on painkilling drugs.

Big data analytics and patient safety

Like other chronic conditions that have reached epidemic proportions, such as diabetes or heart disease, the responsibility for reducing and preventing prescription drug abuse is split between the patient himself and the healthcare system that may be inadvertently contributing to the problem through inaction, lax monitoring, and insufficient population health management.

Forty-nine out of the fifty states have attempted to address this massive public health concern by implementing prescription monitoring programs: centralized databases or online registries that detect prescribing abnormalities, track dispensing patterns, and ideally prevent patients from accessing dangerous amounts of opioids and other drugs. 

While many of these programs enlist sophisticated big data analytics techniques to manage a statewide population, not all are optimally effective.  And for the one state in the nation that does not have such a program in place, the task of safeguarding the region’s Medicaid patients is even more difficult.

In Missouri, Steve Calloway, Director of Pharmacy at MO HealthNet, the state’s Medicaid program, is using big data analytics to close some of the gaps created by the lack of a statewide prescription monitoring program. 

READ MORE: Maine HIE Expands Real-Time Healthcare Big Data Analytics

With Xerox technology and a person-centered approach to patient safety and population health management, MOHealthNet has achieved dramatic results in a relatively short time frame – and they’re just getting started.

“We have about 947,000 participants in our Missouri HealthNet program,” Calloway told HealthITAnalytics.com.  “Obviously, not all of those persons are on controlled substances or opioids, but Missouri, along with many other states, faces big problems with prescription drug abuse.  That problem is compounded by the fact that Missouri is the only state without a prescription drug monitoring program in place.”

“So Missouri winds up being a good place to be for people who are thinking about abusing drugs,” he continued.  “We're surrounded by nine states, so we become a catch-all for folks from our neighboring states who do have prescription drug monitoring programs."

"A Vicodin-type product is the top prescription that we dispense and pay for.  When you have something like that hit the top of your list in terms of what you're paying for month after month, quarter after quarter, it behooves you to do something about that.”

Pain medication is a tricky area for prescribers.  One patient may react so differently from the next to an injury or illness that gauging the effectiveness of the drugs is a highly individualized process.  Simply prohibiting prescribers or pharmacists from dispensing opioids above a certain threshold would be ineffective and potentially harmful.

READ MORE: Familiar Faces Win Best in KLAS for EHRs, Big Data Analytics

“While we want to make sure patients have appropriate access to medications that they absolutely need, the nature of pain medications is such that you've got some patients who rely on these medications improperly,” Calloway explained.

Fortunately, MO HealthNet doesn’t have to take a black-and-white approach to opioid dispensing.  Because HealthNet has access to integrated medical and pharmacy benefit claims, “we can correlate the diagnosis codes with the pharmacy claims, so we can see if there’s a reason in the electronic medical record for the patient to keep getting these controlled substances,” said Calloway.  “We might be looking for certain codes that would allow or not allow the prescriptions to go through.” 

“Then you have simply the pharmacy software itself that actually adjudicates the claims that allows you to know when a prescription was last billed, know what quantity was last dispensed, and that sort of thing.  That's not a unique software there, but the fact that MO HealthNet integrates the medical claims with the prescription claims is what really created the opportunity for us.”

MO HealthNet started to address the problem by establishing benchmarks for prescription drug activity, said Joshua Moore, Account Manager for Government Healthcare Solutions at Xerox, who worked with Calloway to implement the prescription clinical authorization tool. 

“We started by looking at the trends of what we thought fraud looked like, and then built our algorithm around that to catch it a little bit better every time,” Moore said.  “At this point we think we've got a pretty good system in place where we know that we can minimize the number of false positives, but still help out those people that need a little bit of help.”

READ MORE: Healthcare Big Data Analytics Suffers from “Fragmented” Approach

Using the FDA’s recommended guidelines for painkillers, the analysts installed an upper limit for how much medication a patient should be receiving daily, and then flagged cases that significantly exceeded those amounts.

For the Vicodin-like products – that would be hydrocodone and Tylenol products – about 10 percent of the claims prior to implementation were above the FDA recommended dosages,” Moore explained.  “At the beginning of this year, only 0.33 percent were above that mark.  For Percocet-type prescriptions with both oxycodone and acetaminophen combinations, about 7.9 percent were over the FDA recommended dosage before implementation, and now we're down to 2.6 percent as of January 2015.”

It was important for the program to pay attention to acetaminophen levels, because patients aren’t often aware of just how much they are taking.  In 2014, the FDA urged providers to stop prescribing combination painkillers that contained more than 325mg of acetaminophen per dose, and warned that exceeding the limit could result in severe liver damage.

“Many patients don’t realize that the dosage is quite high, especially when you start taking five, six, eight tablets a day,” Moore said.  “So we looked at what the pre-program dose of Tylenol was of these tablets per day for these patients, and it was about 2,300 milligrams per day.  And we've been able to reduce that to – as recently as January – to about 1,780 milligrams a day.  So you're looking at just over 500 milligrams of Tylenol being removed from these patients.”

Patient safety is paramount to MO HealthNet’s program, but Calloway and Moore recognize that prescription painkillers are not a one-size-fits-all proposition.

“I used to work in the retail pharmacy world, and one of the good things about this program is that we never tell anybody that they absolutely cannot have this medication,” Moore said.  “There's always a pathway for exceptions for those false positives that I was talking about earlier, where the pharmacy can pick up the phone and explain the situation or provide additional details.”

“Say that the patient went to the ER on day one, and two days later they realize that medication is not working, and they really need something stronger,” he added.  “Our program would stop that from being filled automatically.  But if the pharmacist talked to the patient, and talked to the ER doctor, they could call into the help desk and the help desk could then put in an override for them.”

“On the flip side of that, we also have those denials in place that say, ‘This patient got this medication two days ago.  They probably shouldn't be on it again without that conversation.’  So it gives the pharmacist more backing, if you will, to say, ‘We really don't feel comfortable dispensing this medication.’  And at the end of the day, it's all about helping the patient get the best care.”

In these exceptional situations, pharmacists and physicians still have the final say.  But should they agree that more medication is warranted, “the whole process from the time the pharmacy hits enter on the claim to receive a message back is about half a second, so you're not going to see much of a delay,” Moore said.  “And I know in the retail pharmacy world, time is very much of the essence.  A delay of even a few minutes can be quite trying for a pharmacist and a patient, so we're really trying to eliminate any of that wait.”

Painkiller problems don’t just happen when patients get hooked and try to illicitly obtain more pills than they need from pharmacies and physicians.  Sometimes the prescribers themselves can be unaware of their prescribing habits, and may be inadvertently sending patients home with too much medication, which may lead to dependence or abuse.

Using the data to keep a watchful eye on providers’ habits is the flip side of the coin, Calloway said, though it is not always easy to pick out fraud when many providers have valid reasons for working the way they do.

“We require prescribers to be identified in the system as part of the prescribing and adjudication of the claim, which can help us identify trends when it comes to a high prescriber of a particular medication,” he said.  “That may be totally legitimate – you would expect a pain management practitioner to prescribe opioids a lot more frequently than a pediatrician.  But the data can still be useful for flagging outliers who may not be adhering to prescribing guidelines.”

Those guidelines are under review as the prescription drug abuse epidemic continues to claim an unacceptable number of lives.  “One of the things that's going on in the pain management world and the opioid prescribing is – and many states are implementing this as well as third parties as well – something we call morphine equivalents,” said Calloway. 

“Morphine is kind of the standard drug.  Now, not everybody gets morphine, so when you talk about oxycodone or hydrocodone or something like that – the potency of it and the amount of the can be expressed in terms of morphine equivalents.”

“The provider community has largely agreed that 120 morphine equivalents is a reasonable daily amount,” he said.  “So the next step for us is to use the software to perform some calculations on the total amount of opioids for an individual patients so that we can limit the amount of morphine equivalents to 120 per day.  We want to get a closer look at those prescriptions that exceed that amount, and see if there’s a way to help these patients stay closer to the guidelines.”

“There are some legitimate reasons why a patient would need more than that,” he stressed.  “Cancer pain is extremely difficult to manage, for example - but we’d like to give providers feedback on patients who might be exceeding the recommended dosages, and we’d like to understand why these things are occurring in our patient population.”

The system can identify when oncological pain may be the cause of a high opioid dose, Moore added. “We can actually pull out specific diagnosis codes for patients, so we can separate something like cancer-related pain from those patients who shouldn’t require more than 120 morphine equivalents to manage their conditions. And of course, we always have a place for exceptions.  There will always be exceptions in patient care.  The guideline isn’t exact yet, and there's a lot of gray area for individuals.”

“But it will always help patient care when the prescriber has to be accountable to someone about why the patient is on that level of medication,” he continued.  “Then that doctor or that care team can go back and say, ‘You know what? Maybe we can do something else, instead of being at double the number of morphine equivalents that everybody else is using.’"  

By reexamining the underlying reasons why and how patients are prescribed painkillers in Missouri, HealthNet hopes to improve the process for providers, pharmacists, and patients alike.  Preventing residents from falling into dependence isn’t just about reducing costs or illegal activity within the Medicaid system.

“This is not just about trying to do something about drug abuse,” said Calloway.  “It really is about patient safety.  Some of the overdoses and adverse events related to opioid use stem from abuse, but many others can be accidental or inadvertent.  We want to make sure our participants are safe, and that’s what this process is allowing us to do.”

X

Join 25,000 of your peers

Register for free to get access to all our articles, webcasts, white papers and exclusive interviews.

Our privacy policy

no, thanks