Article Content


Machine learning helps reduce drug errors

Clinical decision support (CDS) alerting tools can identify and reduce medication errors, but they typically identify only errors previously programmed into their alerting logic. New research indicates that machine learning systems can generate clinically valid alerts that CDS systems may miss. An application of artificial intelligence, machine learning gives a system the ability to automatically learn and improve from experience rather than from explicit programming.


In the study, over 10,600 alerts were generated on outpatient data from two academic medical centers between 2009 and 2013. When alerts generated by the machine learning system were compared with those generated by an existing CDS system, researchers found that over 68% of machine learning alerts would not have been generated by the CDS system. Based on a random sample of the chart-reviewed alerts, 92% of these alerts were accurate and nearly 80% were clinically valid. The estimated cost of adverse events potentially prevented was more than $60 per drug alert.


Although these results are promising, machine learning also has risks and drawbacks. For example, due to the complexity of how the system arrives at a decision to generate an alert, clinicians may misinterpret an alert's significance. The authors write that "the true value of such alerts is highly contingent on whether and how clinicians respond to such alerts and their potential to prevent actual patient harm."


Source: Rozenblum R, Rodriguez-Monguio R, Volk LA, et al. Using a machine learning system to identify and prevent medication prescribing errors: a clinical and cost analysis evaluation. Jt Comm J Qual Patient Saf. 2020;46(1):3-10.



Response to early ART in neonates investigated

Some adults with HIV infection who start antiretroviral therapy (ART) soon after infection have achieved periods of remission when ART is withdrawn. A study was conducted to determine whether HIV-infected neonates who initiate ART within 14 days of birth and are maintained on ART would also be able to maintain viral suppression when ART was withdrawn.


Seventy-three HIV-infected neonates at one hospital in Johannesburg, South Africa, were initiated on ART within 14 days of birth and maintained on ART while viral load decline and immune recovery were tracked. In 46 infants, ART was initiated within less than 2 days of birth; in the remaining 27 infants, ART was initiated between 2 and 14 days of birth.


Three infants died and nine were lost to follow-up before age 48 weeks.Of the remaining infants, too few met eligibility requirements for proceeding with the proposed interruption of treatment as originally planned. However, the researchers found that the rates of achieving and sustaining viral suppression were similar among infants who initiated ART within less than 2 days of birth (51%) and within 2 to 14 days of birth (54%), suggesting that starting ART immediately after birth is not essential for a good result.


"Our results should be reassuring for clinicians that speed is not everything," said lead researcher Louise Kuhn, PhD, MPH, in an interview. "Although undue delay should not be introduced, good outcomes for neonates with HIV can be obtained if ART is started in the first 2 weeks of life."


Sources: Kuhn L, Strehlau R, Shiau S, et al. Early antiretroviral treatment of infants to attain HIV remission. EClinicalMedicine. [e-pub January 2020] Outcomes similar among infants who start ART at 2 vs. 14 days. January 8, 2020.



Customizable dosage controller approved

An interoperable automated glycemic controller device, the Tandem Diabetes Care Control-IQ Technology, is the first insulin controller that can be used with other diabetes devices that are meant to be integrated into a diabetes management system for automated insulin delivery. The device automatically adjusts insulin delivery through connection to an alternate controller-enabled insulin pump (ACE pump) and integrated continuous glucose monitor (iCGM). Patients with type 1 diabetes can use the new controller with compatible iCGMs and ACE pumps to automatically increase, decrease, and suspend delivery of basal insulin based on insulin delivery history, iCGM readings, and predicted glucose values. The controller can also automatically deliver insulin when the glucose value is predicted to exceed a predefined number.


The device was shown to be safe and effective in a clinical trial of 168 patients with type 1 diabetes. It was also assessed for its ability to communicate with the entire system with appropriate reliability, cybersecurity, and fail-safe modes. However, despite these assessments, a risk for delayed insulin delivery still exists. The FDA also cautions that loss of communication between connected devices or from cybersecurity issues could also cause delays.


Source: US Food and Drug Administration. FDA authorizes marketing of automated insulin dosing controller. News release. December 19, 2019.



Few teens receive drug therapy after overdose

In a cohort study of 4,039,216 Medicaid-enrolled adolescents and young adults ages 13 to 22 (median age, 18), 3,791 experienced a nonfatal opioid-related overdose. Of 3,606 youths with opioid-related overdose and continuous enrollment for at least 30 days after overdose, less than a third received timely, evidence-based addiction treatment, defined as a claim for behavioral health services; for buprenorphine, methadone, or naltrexone prescription or administration; or for both behavioral health services and pharmacotherapy within 30 days of overdose. About 29% received only behavioral therapy and 69% received no therapy at all. Less than 2% received pharmacotherapy with buprenorphine, naltrexone, or methadone. Youths with heroin overdose were significantly less likely than those who overdosed on other opioids to receive any treatment. "Interventions are urgently needed to link youths to treatment after overdose, with priority placed on improving access to pharmacotherapy," the authors write.


Source: Alinsky RH, Zima BT, Rodean J, et al. Receipt of addiction treatment after opioid overdose among Medicaid-enrolled adolescents and young adults. JAMA Pediatr. [e-pub Jan. 6, 2020]



New approval for acute migraine treatment

The FDA has approved ubrogepant (Ubrelvy) tablets for the acute treatment of migraine with or without aura in adults. It is not indicated for the preventive treatment of migraine. Ubrogepant is the first drug in the class of oral calcitonin gene-related peptide receptor antagonists approved for the acute treatment of migraine.


In two randomized, double-blind, placebo-controlled trials of 1,439 adult patients with a history of migraine (with and without aura), patients received the approved doses of ubrogepant to treat an ongoing migraine. In both studies, the percentages of patients achieving freedom from pain and other symptoms such as nausea within 2 hours was significantly greater than for those receiving placebo. Patients were allowed to take their usual acute treatment of migraine at least 2 hours after taking ubrogepant and 23% of patients were taking a preventive medication for migraine. The most common adverse reactions were nausea, fatigue, and dry mouth. Ubrogepant should not be used concurrently with strong CYP3A4 inhibitors such as ketoconazole, itraconazole, and clarithromycin.


The FDA describes ubrogepant as a novel treatment that represents "an important new option for the acute treatment of migraine in adults." In the US, migraine affects an estimated 37 million people.


Source: US Food and Drug Administration. FDA approves new treatment for adults with migraine. News release. December 23, 2019.