The following summaries represent a small sample of the many real-world, evidence-based studies presented at the 30th Annual Meeting of the Academy of Managed Care Pharmacy (AMCP), April 23-26, 2018, in Boston, MA.
- Mixed Results with Intranet Informational Program for Opioid Prescribing
- Electronic Submission of Prior Authorization Requests Boosts Efficiency
- Adding Long-Acting Insulin to the Formulary Lowers Diabetes Costs, Preserves Clinical Outcomes
- Adding a New Acute Myeloid Leukemia Option to a Formulary Has Minimal Cost Impact
- More Treatment Options Needed for Patients with Unresectable Stage III Non–Small-Cell Lung Cancer
- Digital Messaging Had Some Impact on Medication Adherence in Patients with Cystic Fibrosis
Mixed Results with Intranet Informational Program for Opioid Prescribing
Physician intranet-based information and education about opioid prescribing had mixed effects on clinician prescribing behavior, according to a survey of participating physicians.
The intervention had a positive impact on physicians’ evaluation of morphine equivalent dosing but a minimal effect on the identification of co-existing health risks associated with opioids, opioid use tapering, and discussion of alternative therapies, according to a study presented at the 2018 AMCP meeting.
In the 2-part survey, physicians who practice within the Kelsey-Seybold system completed the 10-question survey that was designed to assess their familiarity and comfort level with key opioid-related issues. The survey focused on opioid prescribing guidelines, opioid tapering to treatment discontinuation, patient evaluation, and the use of the Texas controlled-substance database.
Answers to the initial survey formed the basis for the development of opioid education materials that were posted to the Kelsey-Seybold physician intranet. The materials included a tutorial about registration and the use of the Texas Prescription Monitoring Program database, an opioid-tapering guidance document, a resource of 90 morphine milligram equivalent doses for various opioid drugs, as well as patient education handouts for physician use in clinical settings.
Physicians who responded to the initial survey were asked to complete the survey a second time to assess the impact of the education materials on their knowledge of and prescribing behaviors related to opioids.
As part of the study, the investigators retrospectively reviewed pharmacy claims data to identify any trends in opioid overuse or misuse. The query sought out patients whose opioid prescriptions exceeded 120 morphine equivalent dosing for more than 90 days in the past year. The records of patients identified by the query were reviewed to determine the appropriateness of opioid use.
Physician participation decreased substantially from the first survey (N = 117) to the second (N = 46). For each of the 10 questions, the participants rated their knowledge or comfort level associated with opioid prescribing on a scale of 1 to 5. Overall, the responses to the follow-up questionnaire demonstrated modest improvement in self-assessment on 8 of the 10 questions (range, +0.02 to +0.82).
The greatest improvement was seen in the participants’ self-rated practice of evaluating the morphine equivalent dosing of patients with noncancer pain (+0.82), identification of patients at high risk for opioid diversion or misuse (+0.61), use of pain management services for assistance with the management or discontinuation of a patient’s opioid regimen (+0.61), and familiarity with opioid prescribing recommendations (+0.56). The scores declined for participants’ rating of their ability to identify the potential risks for opioid harm or misuse (–0.04) and their comfort level with patient discussions about tapering or discontinuing their opioid use (–0.04).
“Overall comfort levels regarding opioids could still be improved in all areas, as average responses did not achieve a [targeted] four-point rating,” concluded Paul Hudacek, PharmD, Kelsey-Seybold Clinic, Houston, TX, and colleagues.
The future goals of the program might include exploration of collaborative methods of information sharing with physicians and periodic pharmacy claims reviews to monitor for opioid overuse or abuse, noted the investigators.
[Source: Hudacek P, Gandhi K, Martinez Jonathan D, et al. Implementation of an opioid monitoring and education program in a large multi-specialty practice group.]
Electronic Submission of Prior Authorization Requests Boosts Efficiency
The development of multiple-vendor partnerships facilitated a 10-fold increase in electronic prior authorization (e-PA) submissions for high-cost prescription drugs in 1 year, according to a study. The increased use of e-PA was associated with a reduction in the average turnaround time from 2.5 days with faxed authorization requests to 0.87 days. Significantly fewer (P <.0001) cases submitted by e-PA required additional information from providers, but the approval and denial rates did not change.
“Requests using a drug-specific decision tree were less likely to require pharmacist review and were more likely to be approved by a technician compared to requests that used a default decision tree,” reported Leann C. McDowell, PharmD, and Susan Cooper, RPh, of HealthPartners, Bloomington, MN. “Providers within HealthPartners’ care system were reporting that pre-prescribe data for HealthPartners members appeared to be more accurate compared to other payers.”
A growing list of approved complex, high-cost drugs will likely lead to an increase in PAs, which will prove challenging for providers, health plan members, and health plans. The increased use of e-PAs can help address the challenges by increasing the efficiency of the PA process. Minnesota has required support for e-PAs as of January 1, 2016, but implementation and adoption have been slowed by issues related to system technology and integration, Dr McDowell and colleagues noted.
HealthPartners, a health insurance plan and care delivery organization, manages its own customized formulary and PA program. Preparation for e-PA began with the development of drug-specific decision trees that included clinical review questions and information needed to process approvals.
HealthPartners made access to e-PA available to providers after partnering with an electronic health record (EHR) vendor that had existing relationships with multiple healthcare systems. Access expanded as a result of partnerships with 2 additional vendors. At the same time, HealthPartners improved pre-prescribe data and alternative medication information within the EHR, including medications that require a PA and alternate drugs that do not.
From January 1, 2016, through March 31, 2018, HealthPartners received 114,315 PA requests. At the start of the implementation period, e-PA submissions accounted for less than 5% of all PA requests, but this increased to more than 50% by early 2018 as a result of the vendor partnerships.
As of April 1, 2018, HealthPartners had more than 300 new drug-specific decision trees, which were used in more than 60% of PA requests. The number of requests that required additional provider information declined by almost 50% with e-PA submission (6.0% vs 15.5% for fax submissions; P <.0001). The need for additional information declined by approximately 33% with e-PA submissions versus PAs submitted by fax that used a default decision tree with general clinical questions (10.4% vs 15.5%, respectively; P <.0001).
Requests involving drug-specific decision trees required significantly less pharmacist review and were more likely to be approved by a technician than requests involving a default decision tree (27.1% vs 47.5%, respectively; P <.0001).
[Source: McDowell LC, Cooper S. Improving prior authorization efficiencies with electronic prior authorization (ePA).]
Adding Long-Acting Insulin to the Formulary Lowers Diabetes Costs, Preserves Clinical Outcomes
The placement of long-acting insulin agents on a health plan preferred formulary tier led to lower costs and similar or superior clinical results for preferred versus nonpreferred drugs, a retrospective analysis showed.
The use of the preferred drugs insulin glargine (Basaglar) and insulin degludec (Tresiba) was associated with almost a $200 monthly saving per user compared with nonpreferred drugs. Adherence to the prescribed therapy was similar between the different tier categories, and data for a small subgroup of patients suggested at least comparable hemoglobin (Hb) A1c levels for patients using preferred or nonpreferred antidiabetes drugs.
Currently, 5 long-acting insulin drugs are approved by the US Food and Drug Administration for the treatment of types 1 and 2 diabetes mellitus. The researchers noted that Basaglar has demonstrated biologic similarity to another insulin glargine, Lantus, but at a lower cost. They also pointed out that Tresiba has a longer duration of action that permits once-daily dosing, which encourages patient adherence and has demonstrated lower rates of clinically significant hypoglycemia.
Beginning June 1, 2017, Passport Health Plan made a formulary change, designating Basaglar and Tresiba as the preferred long-acting insulin agents for diabetes.
“With the short-term economic clinical observations trending in favor of the plan’s two preferred long-acting insulin products, additional studies evaluating the long-term impact of this formulary change could demonstrate more significant benefits,” noted Cavan O’Reilly, PharmD, Sullivan University College of Pharmacy, Louisville, KY, and colleagues. “Daily dose monitoring would provide a useful mechanism for more accurate adherence evaluation.”
O’Reilly and colleagues conducted a retrospective review to assess the clinical and financial impacts of the formulary change and to assess patient adherence rates for all long-acting insulin drugs.
Plan members with diabetes who require insulin therapy were divided into 2 cohorts based on the formulary tier status of the long-acting insulin agent they used (preferred or nonpreferred). They also were stratified into 5 cohorts defined by the specific long-acting insulin drugs.
The data analysis included drug and medical spending, member adherence as determined by medication possession ratio (MPR; adherent = MPR ≥0.8), and change in HbA1c. The results were compared for formulary cohort status and by agent-specific cohorts.
Of the 805 plan members, 741 received the 2 preferred long-acting insulin drugs and 64 received the nonpreferred drugs. The monthly overall pharmacy and medical costs expenditures averaged $933.51 for members using preferred long-acting insulin drugs and $1114.78 for nonpreferred long-acting insulin drugs (P = .48).
The data on average change in HbA1c were based on 34 members (29 in the preferred drugs cohort and 5 from the nonpreferred drugs group). The between-group comparison showed a decrease of 0.44% in mean HbA1c for patients using the preferred drugs compared with a reduction of 0.3% with nonpreferred drugs (P = .48).
Approximately 66% of patients in both formulary tier cohorts met the definition for treatment adherence. There was a small numerical advantage for members who used preferred versus nonpreferred long-acting insulin drugs (66% vs 64%, respectively; P = .86).
[Source: O’Reilly C, Armstrong CM, McClenton K. Financial and clinical impact of health plan formulary changes to long-acting insulin agents in a Medicaid population.]
Adding a New Acute Myeloid Leukemia Option to a Formulary Has Minimal Cost Impact
Adding a new treatment option for relapsed or refractory acute myeloid leukemia (AML) to a formulary would have minimal cost impact for a large health plan, according to the results of a decision-analytic model.
Treating 15 patients with the oral isocitrate dehydrogenase (IDH) 2 inhibitor enasidenib (Idhifa) would add $0.002 to the per-member per-month (PMPM) costs during the first year in a hypothetical 5-million-member health plan. The PMPM cost would rise to $0.006 during the second year and reach a plateau of $0.011 in the third year.
The recent US Food and Drug Administration approval of enasidenib, a first-in-class oral IDH2 inhibitor, posed questions for managed care plans regarding the economic impact of adding the drug to a formulary. In a pivotal trial involving patients with relapsed or refractory AML with IDH2 mutation, treatment with enasidenib led to a complete response in 23% of patients and to conversion from red blood cell or platelet transfusion dependence in 34% of patients.
Christopher N. Graham, RTI Health Solutions, Research Triangle Park, NC, and colleagues developed a decision-analytic model to estimate the costs associated with drug acquisition, drug administration, transfusion, diagnostic testing, and adverse events. They compared the estimated costs for enasidenib versus standard high- and low-intensity chemotherapy regimens for relapsed or refractory AML. Previous studies suggested 15 patients within a 5-million-member health plan would require treatment for relapsed or refractory AML with IDH2 mutation.
“Two scenarios were examined: market without enasidenib and market with enasidenib,” the investigators noted. “The difference in costs between the 2 scenarios is the budget impact of enasidenib.”
The model incorporated the costs for treatment with enasidenib for 7 chemotherapy regimens (including current market share) and for best supportive care. The estimated drug cost per cycle for enasidenib was $23,214, and ranged from $3 to $4320 for chemotherapy. The costs for inpatient administration of high-intensity chemotherapy regimens ($25,187) were compared with $0 for enasidenib administration and $752 for other treatment regimens; the cost for molecular pathology was $68 and $1608 per blood transfusion.
Currently, 47% of patients with AML undergo IDH2 mutation analysis. The second scenario assumed that every patient with AML would have the analysis. The average cost per adverse event was $7828, and the estimated frequency of grade 3/4 adverse events was 0.60 with enasidenib to 6.00 with the widely used 7+3 chemotherapy regimen (ie, 7 days of cytarabine therapy followed by 3 days of idarubicin therapy).
Over a 5-year period, the introduction of enasidenib to a formulary would decrease drug administration costs by $875,938, transfusion costs by $90,849, and adverse events costs by $651,386. The costs would increase for drug acquisition ($4,097,026) and for diagnostic testing ($2748).
The estimated postintroduction total cost differences ranged from $103,000 in the first year to $686,000 in the fifth year. Sensitivity analyses yielded a minimum PMPM cost increase of $0.001 and a maximum of $0.013 across 5 years.
“Adding enasidenib to the formulary for patients with IDH2-mutated relapsed/refractory AML has a minimal impact on the PMPM cost of a US health plan while providing physicians with another treatment option for patients in a disease with limited treatment options,” the investigators concluded.
[Source: Graham CN, McGuire M, Knox H, Ung B. Budget-impact analysis of enasidenib in patients with relapsed and refractory acute myeloid leukemia with an isocitrate dehydrogenase-2 mutation.]
More Treatment Options Needed for Patients with Unresectable Stage III Non–Small-Cell Lung Cancer
Approximately 25% of patients with unresectable stage III non–small-cell lung cancer (NSCLC) received no treatment during a second or third disease progression interval, and treated patients who did receive treatment usually only received chemotherapy, according to the results of a real-world retrospective review of almost 500 patients.
Based on medical records of patients treated in the community settings, 27% of patients received no treatment after first disease progression (second progression interval). After second relapse, 26% of patients received no treatment during the third progression interval.
Approximately 78% of the patients received guideline-recommended chemoradiation (CRT) as initial therapy, and 20% of these patients died before disease progression. Approximately 58% of additional patients who initially received CRT had disease progression.
The anti–PD-1 agent durvalumab (Imfinzi) received a new US Food and Drug Administration indication in February 2018 for the treatment of patients with unresected stage III NSCLC after concurrent CRT. Additional studies are needed to evaluate the impact of durvalumab on treatment practices and patient outcomes.
The National Comprehensive Cancer Network recommends concurrent CRT as initial treatment for patients with unresectable stage III NSCLC. However, not all patients are healthy enough for concurrent treatment and instead receive sequential CRT or monotherapy, the researchers noted. Limited data exist on real-world treatment patterns for unresected stage III NSCLC in the community setting.
A search of medical records from a network of 10 community oncology practices identified 478 patients with newly diagnosed stage III NSCLC between January 2011 and March 2016. The records showed no evidence of surgery within 6 months of diagnosis. The investigators used descriptive methods to evaluate the treatment practices.
The analysis included data collected from diagnosis through the end of the third disease progression interval. The first progression interval was defined as the interval from diagnosis until first disease progression. Subsequent intervals were defined as the time between disease progressions.
Overall, 438 (91.6%) patients received therapy during the first progression interval, including concurrent CRT in 340 (77.6%) patients; 145 patients received 2 regimens during the first progression interval, 23 received 3 regimens, and 2 patients received 4 regimens. Chemotherapy without radiation was the most common second and third regimens used during the first progression interval.
In the first progression interval, the median treatment duration was 2 months and the median progression-free survival (PFS) was 10 months.
Of 278 patients who had a first disease progression, 202 (72.7%) received treatment during the second progression interval, which was chemotherapy without radiation in 40.3% of cases. Of those receiving treatment, 32 patients received 2 regimens, 7 received 3 regimens, and 1 patient received 4 regimens during the second progression interval. Approximately 27% of patients received radiation only as the first treatment during the interval.
In the second progression interval, the median duration of treatment was 1.4 months and the median PFS was 4.2 months.
In total, 131 patients had a second disease progression, and 97 (74.0%) of them received anticancer treatment during the third progression interval. Approximately 52% of the treated patients received chemotherapy alone, and 32% received a nonplatinum agent. In the third progression interval, the median duration of treatment was 1.4 months and median PFS was 3.1 months.
A “significant unmet need still exists in patients with unresected stage III NSCLC,” Rajeshwari S. Punekar, PhD, Vector Oncology, Memphis, TN, and colleagues concluded. “Novel therapies are needed to address unmet need in unresected stage III NSCLC patients.”
[Source: Punekar RS, Ryan KJ, Fernandes AW, et al. Real-world treatment patterns among patients with unresected stage III non-small cell lung cancer (NSCLC).]
Digital Messaging Had Some Impact on Medication Adherence in Patients with Cystic Fibrosis
The implementation of a secure digital clinical messaging system for cystic fibrosis helped to improve medication adherence, but the impact did not reach statistical significance.
The results of a retrospective analysis showed that medication adherence, as defined by medication possession ratio (MPR), was 6.56% (but not significantly) higher among patients who received secure messages over a digital messaging system than among those who did not receive these messages. Several other metrics of adherence improved to varying degrees during the 3-month evaluation period, but only a difference in the interval to first fill drop-off achieved statistical significance.
Historically, engagement between a specialty pharmacy and patients has occurred by telephone or mail. Advances in digital technology increasingly afford opportunities for message-based engagement.
CVS Specialty implemented a clinical secure messaging system on a digital platform for patients with select diseases, including chronic myeloid leukemia, rheumatoid arthritis, multiple sclerosis, and cystic fibrosis. Ruchik Patel, PharmD, and Christine Sawicki, RPh, of CVS Specialty, Providence, RI, reported their findings of the messaging system’s impact on adherence among patients filling a first prescription for cystic fibrosis from June 28, 2017, to August 31, 2017.
A propensity-matched analysis included 174 patients who were enrolled in the digital messaging program. During a 3-month follow-up period, approximately 50% of the patients received secure messages and 50% did not.
The primary end points were MPR and the proportion of patients reaching optimal adherence, which was defined as an MPR of ≥85%. The analysis also covered drivers of adherence, including first-fill drop-off rate, days of medication on hand, and on-time refill.
The patients who received secure messages did not significantly exceed the primary adherence outcomes compared with the patients who were enrolled in digital messaging and did not receive secure messages. The rate of optimal adherence was 6.9% higher in the group that received messages than in the group that did not (P = .35).
An analysis of the adherence drivers showed a trend in favor of the group that received secure messages during the follow-up period, including a reduction in first-fill drop-off of 13.95% (P = .04), an increase in medication on hand of 7.63 days (P = .15), and an increase in the percentage of on-time refills of 5.23% (P = .50).
With clinical secure messaging, 8.78% more patients demonstrated a lift in MPR, and 12.2% more patients were optimally adherent during the 3-month follow-up period. In addition, 42.8% fewer patients dropped their treatment after the first fill (P = .04).
“Adherence outcomes are impacted positively in patients receiving cystic fibrosis agents, although the impact was not statistically significant,” stated Dr Patel and Ms Sawicki. “Further studies with longer enrollment windows and a 6- or 12-month follow-up are warranted.”
“Overall, the impact of clinically focused programs may vary between disease states,” the researchers added. “Studies across several disease states should be considered before determining the impact of these programs.”
“First fill drop-off is a contributor to adherence,” concluded the investigators. “Days on hand and refill on time may partially contribute to an increase in adherence. Adherence and persistency curves continue regressing beyond 3 months. Therefore, the observed outcomes may split the adherence curves with statistical significance.”
[Source: Patel R, Sawicki C. Investigation of clinical secure messaging on adherence outcomes within a specialty pharmacy population.]