Some individuals on Medicaid receiving treatment for opioid use disorder might have received unnecessary or dangerous opioid prescriptions outside of their treatment programs, according to an audit issued by State Comptroller Thomas P. DiNapoli. Auditors found treatment programs were not checking whether patients were getting other opioid prescriptions or coordinating care with health care providers.

This audit is part of a broader initiative by the State Comptroller's office to look at opioid use across New York state. Auditors, in a series of reports, will be looking to identify at-risk patient/provider behavior and gaps in the effectiveness of government programs.

"New York and the rest of the country are facing an opioid addiction epidemic, and people's lives are at stake," DiNapoli said. "Programs designed to get individuals off highly addictive opioids can only be effective with proper vigilance. The state Department of Health should take steps to help treatment programs and health care providers work together to prevent overdoses that could lead to hospitalizations or death."

Overdose deaths involving prescription opioids in New York state have risen from 754 in 2010 to 1,382 in 2015, a jump of 83 percent, according to state and federal authorities.

The state maintains a database to monitor prescription drug use, known as the Internet System for Tracking Over-Prescribing (I-STOP), which contains records of all controlled substances that were dispensed in the state and reported by a pharmacy or dispenser. I-STOP is a tool to help prescribers determine the most appropriate prescriptions for an individual. Opioid treatment programs are not required to disclose the medications they provide to patients, but should check for a patient's other opioid use on I-STOP, and in certain instances they are required to do so.

If a treatment program becomes aware of outside prescriptions, the program can ask for patient consent to consult with other health care providers on the patient's medical condition and situation. Treatment programs are required to obtain consent from the patient prior to sharing treatment information with other practitioners.

DiNapoli's auditors examined state Department of Health (DOH) records from Oct. 1, 2013 to Sept. 30, 2017 and identified 18,786 Medicaid recipients who received 208,198 prescriptions for opioids through the Medicaid program while also receiving opioids, usually methadone, as part of a treatment program for opioid use disorder.

Auditors found 33 percent of Medicaid recipients in treatment programs received prescription opioids outside of their program. Three percent of these individuals received medical care for an overdose within 30 days of obtaining a prescription. Specifically, 493 needed medical care as a result of 691 opioid or narcotic overdoses that occurred within a month of receiving a prescription opioid. Twelve of those individuals died during the time of their medical care involving the overdose.

DiNapoli's auditors reviewed the medical records for a sample of 25 Medicaid recipients from three treatment programs. Auditors also analyzed Medicaid data and determined these individuals received 1,065 Medicaid opioid prescriptions while they were in treatment for opioid abuse. On average, these recipients were in their treatment programs for nearly 3.5 years. However, auditors found treatment programs did not check I-STOP for every recipient. For the 25 individuals, I-STOP was checked only 18 times. Auditors also found that I-STOP was not checked every time a medication-assisted opioid was dispensed for take-home use, as required by state law.

Auditors found that 13 of the 25 recipients were asked at least once during treatment to sign a consent form to coordinate care with the recipient's opioid prescriber. However, three of the 13 recipients declined to sign the form. Treatment programs were aware of 566 of the 1,065 opioid prescriptions (53 percent) provided to the 25 patients. Auditors found that consent to coordinate care was obtained regarding just 85 of the 1,065 prescriptions (8 percent) and care coordination occurred for only 59 of the 1,065 prescriptions (6 percent).

DiNapoli recommended DOH:

• Evaluate the benefits of the following actions to improve scrutiny over opioid prescriptions for Medicaid recipients who are being treated for opioid use disorder:
  • Developing a report that can be used to notify treatment programs when I-STOP indicates recipients are receiving potentially dangerous prescriptions;
  • Taking steps to ensure treatment programs are aware of the option to upload patient information when querying I-STOP;
  • Taking steps to ensure Medicaid managed care organizations have controls requiring medical appropriateness reviews prior to dispensing opioids to recipients with opioid use disorder consistent with Medicaid fee-for-service controls; and
  • Including a risk assessment that is specific to individuals receiving medication-assisted treatment for opioid use disorder concurrently with opioid prescriptions.

DiNapoli also recommended DOH:

• Issue guidance to remind treatment programs of the statutory and regulatory requirement to check I-STOP when they dispense take-home doses of opioid medications. Evaluate the benefits of establishing additional guidance for treatment programs to make other checks of I-STOP when clinically appropriate; and
• Formally remind treatment program providers of the importance of seeking to coordinate care with prescribers of opioids outside of the treatment programs.

DOH disagreed with some of the conclusions of the audit but generally agreed that actions would be taken to address the recommendations. The department's complete response is included in the audit.

Read the report, or go to: https://www.osc.state.ny.us/audits/allaudits/093019/sga-2019-17s66.pdf

Find out how your government money is spent at Open Book New York. Track municipal spending, the state's 150,000 contracts, billions in state payments and public authority data. Visit the Reading Room for contract FOIL requests, bid protest decisions and commonly requested data.