Top Evidentiary Gaps in Managed Care Pharmacy: A Research Agenda.

Research agendas play valuable roles in clearly identifying high-priority topics that reflect potential to improve health care quality. The purpose of this report is to present work completed by the Academy of Managed Care Pharmacy (AMCP) and AMCP Foundation Joint Research Committee. This committee set forth to develop a research agenda for our 2 organizations that focuses on critical evidence needs in managed care pharmacy. This document reviews results from 2 surveys that were conducted to better understand unmet research needs within managed care pharmacy and to inform professional efforts of managed care pharmacists. The first survey collected qualitative data from key opinion leaders (KOLs) regarding the top evidentiary gaps in managed care pharmacy and barriers to closing those gaps. The second survey was sent to AMCP members and AMCP Foundation stakeholders, used a mixed methods quantitative-qualitative design, and incorporated concepts from initial KOL responses. The key outcome from these proceedings is the research agenda, which identifies and prioritizes 4 evidentiary gaps in managed care pharmacy: (1) real-world evidence to inform managed care pharmacy decision making, (2) value-based models in managed care pharmacy to address total cost of care, (3) impact of benefit design or utilization management strategies on patient outcomes, and (4) impact of direct patient care services provided by managed care pharmacy on patient outcomes. The agenda was intended to be broad and will evolve over time. AMCP and the AMCP Foundation hope that this research agenda inspires the AMCP membership, researchers, and funding agencies to close these gaps in knowledge and understanding. DISCLOSURES: Chairs and members of the Joint Research Committee oversaw and conducted the work outlined in this report with the support of AMCP and the AMCP Foundation. No outside funding was received. Gembarski, Couto, Wilson, and Eichenbrenner declare no conflicts of interest, real or apparent, with any product or service mentioned in this report. Gembarski is employed by BCBS Michigan; Couto is employed by Cigna; Wilson is employed by HealthCore, a wholly owned subsidiary of Anthem; and Eichenbrenner is employed by the AMCP Foundation.

Usefulness of pharmacy claims for medication reconciliation in primary care.

OBJECTIVES: Methods for efficient medication reconciliation are increasingly important in primary care. Aggregated pharmacy data within the native electronic health record (EHR) may create a new opportunity for efficient and systematic medication reconciliation in practice. Our objective was to identify the prevalence and predictors of medication discrepancies between pharmacy claims data and the medication list in a primary care EHR. STUDY DESIGN: Retrospective cohort study. METHODS: We conducted a retrospective cohort study of patients prescribed a new antihypertensive in a large primary care practice network between January 2011 and September 2012. We compared patients' active medications recorded in the practice EHR with those listed in pharmacy claims data available through the EHR. The primary outcome was the presence of a medication discrepancy. RESULTS: Of 609 patients, 468 (76.9%) had at least 1 medication discrepancy. Significant predictors of discrepancies included the total medication count (odds ratio [OR], 2.18; 95% CI, 1.85-2.57) and having a recent emergency department visit (OR, 2.58; 95% CI, 1.03-6.45). The identified discrepancies included 171 patients (28.1%) with 229 controlled substance discrepancies. CONCLUSIONS: Our study revealed a high rate of discrepancies between pharmacy claims data and the provider medication list. Aggregated pharmacy claims data available through the EHR may be an important tool to facilitate medication reconciliation in primary care.

Using aggregated pharmacy claims to identify primary nonadherence.

OBJECTIVES: Aggregate pharmacy claims available within an electronic health record (EHR) provide an opportunity to understand primary nonadherence in real time. The objective of this study was to use pharmacy claims data available within the EHR to identify the prevalence and predictors of primary nonadherence to antihypertensive drug therapy in a multi-payer primary care network. STUDY DESIGN: We conducted a retrospective cohort study of patients prescribed a new antihypertensive medication in a large primary care practice network between January 2011 and September 2012. METHODS: We matched prescriptions for the new antihypertensive to pharmacy claims listed in the EHR. The primary outcome was the presence of a fill for the new medication within 30 days of the prescription. RESULTS: Of 791 patients in our study cohort, two-thirds (522; 66%) filled their prescription within 30 days. The majority (409; 78.4%) of that group filled the prescription on the day it was issued. Lower diastolic blood pressure and Medicare coverage increased the probability of nonadherence. CONCLUSIONS: Medication fill data within the provider EHR can identify primary nonadherence in clinical practice. As adoption of this technology increases, it provides an opportunity to identify nonadherence, allowing for the effective design of interventions to improve adherence to therapy.

Geographic variation in medication adherence in commercial and Medicare part D populations.

BACKGROUND: Previous literature has emphasized the importance of cost sharing, health literacy, socioeconomic status, cognitive function, disease burden, and polypharmacy as some of the determinants of medication adherence. Little research has been published examining disparities in adherence rates when comparing different regions of the United States. OBJECTIVE: To examine the impact of geography, socioeconomic status, and other demographic variables on medication adherence rates in a large national sample of Medicare Part D and commercially insured beneficiaries. METHODS: This study focused on users of oral antidiabetic, antihypertensive, and/or antilipidemic medications. Beneficiaries who had at least 2 antidiabetic, antihypertensive, or antilipidemic prescription fills in 2010, 2011, or 2012 and who were enrolled in a large commercial or Medicare Part D prescription drug plan for at least 80% of one of these years (9.6 months) were included in this study. Results were stratified by year and by benefit type. Logistic regression was used to test for the adherence differences among the 9 U.S. regions as defined by the U.S. Census Bureau. Additional variables included in the model to control for population differences were age, gender, socioeconomic status, and yearly out-of-pocket medication expenses. RESULTS: After meeting all inclusion and exclusion criteria, 379,533 beneficiaries were in the 2012 Medicare cohort, and 659,553 beneficiaries were in the 2012 commercial cohort. New England was statistically the most adherent geographic region in both cohorts (Medicare odds ratio [OR] = 1.512, CI = 1.399-1.635); commercial OR = 1.193, CI = 1.109-1.284). Younger age beneficiaries, lower income beneficiaries, and females were less adherent in both groups. CONCLUSIONS: In the commercial and Medicare populations, geography, socioeconomic status, age, and gender all impact the likelihood of a beneficiary being adherent to chronic medications for hypertension, diabetes, and hyperlipidemia. While this study does not elucidate the specific factors (i.e., health literacy, disease severity) driving geographic and other differences in medication adherence observed between groups, it does highlight the limitations of quality metrics and wellness initiatives that assume relative homogeneity in beneficiary characteristics across the United States.

The use of urine drug monitoring in chronic opioid therapy: an analysis of current clinician behavior.

OBJECTIVES: The purpose of this study was to better quantify how urine drug monitoring (UDM) is used in clinical practice. Little is known about which patients are monitored, how often patients are monitored, which substances are important to detect, and under what circumstances clinicians modify the frequency of monitoring. DESIGN: An online survey was developed based on qualitative phone interviews with eight clinicians who use UDM as a routine component of clinical practice. PARTICIPANTS: One thousand fourteen randomly selected clinicians known to order urine toxicology screenings were invited by mail in June 2011 to respond to the online survey assessing their clinical needs and preferences regarding UDM. RESULTS: Of the 93 respondents, 76 percent (n = 72) require all new patients to have UDM performed when they enter their clinic. The majority administer UDM to patients four times a year. The most common reasons cited by clinicians for a change in the frequency of monitoring are patient history of substance abuse and aberrant behaviors. Overall, the respondents showed broad support to test patients consistently for the most common illicit drugs, the majority of opioids, and a handful of prescription medications associated with abuse. CONCLUSION: Despite a lack of agreement between guidelines informing the use of UDM, there appears to be a general consensus among practitioners that use UDM on: which patients to monitor, how often to monitor, and which substances are most important to detect.

Measuring migraine-related quality of care across 10 health plans.

OBJECTIVES: To refine a previously published standardized quality and utilization measurement set for migraine care and to establish performance benchmarks. STUDY DESIGN: Retrospective application of the migraine measurement set to health plan data in order to assess patterns of health service utilization. METHODS: Measurement specifications were applied to data from 10 health plans for measurement year 2009. RESULTS: Of the 2.9 million continuously enrolled members of the health plans, 138,004 (4.7%) met inclusion criteria for the migraine population. Of these, 26% did not have a migraine diagnosis, but were utilizing migraine drugs; 12% had a computed tomography scan within the year (range 8%-25% across plans); and 8% had magnetic resonance imaging (range 6%-11%). Nearly 18% of the migraineurs had 1 or more visits to an emergency department/urgent care center for migraine; few (6%) were followed up with primary care visits. Approximately one-fourth of the migraineurs were not being routinely monitored by a physician. Medication utilization also was examined for members of the migraine population with pharmacy benefits. A significant proportion (42%) were given a migraine preventive, 38% had at least 1 prescription for a triptan, and 2% of those on triptans were potentially overutilizing the medication. Among patients aged 18 to 49 years who were given triptans, 3% had a cardiac contraindication; this percentage rose to 7% for patients aged 50 to 64 years. CONCLUSIONS: This study demonstrates the value of standardized measures in identifying potential quality issues for migraine care, including underdiagnosis, overutilization of imaging, and underutilization of preventive drugs.

Recommendations for urine drug monitoring as a component of opioid therapy in the treatment of chronic pain.

OBJECTIVE: Several prominent guidelines recommend that patients on long-term opioid therapy have periodic urine drug monitoring (UDM) for appropriate use; however, none address the specific questions of which patients to test, which substances to test for, how often to test, and how to act on the results. DESIGN: In the absence of adequate scientific evidence in the literature, a panel of experts in the field of pain and addiction medicine was convened to develop consensus UDM recommendations. The panel met three times between March 2010 and April 2011, and reviewed several drafts of the recommendations document between meetings. RESULTS: The group was able to achieve consensus on a set of UDM recommendations addressing test selection, test frequency, interpretation of results, and how to handle discrepancies based on specific results. CONCLUSION: While the participating panel members recognize that there currently is a limited evidence base to support the expert panel's recommendations, primary care providers and pain specialists are largely acting today based on anecdote, intuition, and individual experience. The recommendations are meant to begin to provide a framework for standardizing practices for UDM in the treatment of chronic pain, and to serve as a catalyst to advance research that quantifies the effects of UDM on opioid therapy management and patient outcomes.

Relationship between potential opioid-related adverse effects and hospital length of stay in patients receiving opioids after orthopedic surgery.

STUDY OBJECTIVE: To determine whether there is an association between opioid-related adverse effects and postoperative hospital length of stay (p-LOS). DESIGN: Retrospective medical record review. SETTING: Large academic medical center. PATIENTS: Random sample of 402 patients (mean age 60.2 yrs, 50.3% female) who underwent orthopedic spine, hip, knee, or shoulder surgery during 2007 and received opioids during or after the procedure. MEASUREMENTS AND MAIN RESULTS: Potential opioid-related adverse effects were identified by using established criteria. Bivariate and multivariate analyses (generalized linear regression model, log transformed) were used to identify predictors of p-LOS. The model also estimated the effect of specific types of adverse effects and adverse-effect combinations on p-LOS. Mean ± SD p-LOS was 3.0 ± 2.1 days; median oral morphine equivalent postoperative dose was 60 mg/day. More than half of the patients (54.2%) experienced one or more adverse effects, 25.6% experienced two or more adverse effects, and 7.2% experienced three or more adverse effects. The composite of nausea and vomiting was experienced by 36.1% of study patients, and 12.6% had at least one emesis episode. Constipation and confusion were documented in 6.5% and 3.7% of patients, respectively. Constipation (p<0.0001), emesis (p<0.001), and confusion (p<0.01) were associated with increased p-LOS after adjusting for other significant variables. Patients with constipation had an adjusted 49% (95% confidence interval [CI] 25-77%) longer p-LOS (additional 1.4 days) compared with patients without constipation. Emesis and confusion significantly increased p-LOS by 25% (95% CI 10-42%) and 38% (95% CI 11-72%), respectively. Incremental increases in p-LOS for patients with two adverse effects (p=0.02), three adverse effects (p<0.001), and four adverse effects (p<0.001) versus patients with no adverse effects were 15%, 40%, and 82%, respectively. CONCLUSION: Constipation, emesis, and confusion were associated with increased p-LOS in patients receiving opioids after orthopedic surgery. In addition, there was a significant linear relationship between the number of adverse effects/patient and increased p-LOS, and the strength of the association increased as the number of adverse effects increased. Although the opioid dosages and adverse-effect rates were typical, these findings reinforce the need to balance pain management with risk of events.

Improved cardiac management with a disease management program incorporating comprehensive lipid profiling.

The objective of this study was to evaluate the improved effectiveness of a disease management treatment protocol incorporating comprehensive lipid profiling and targeted lipid care based on lipid profile findings in patients with ischemic heart disease (IHD) or congestive heart failure (CHF) enrolled in a managed care plan. This retrospective cohort study, conducted over a 2-year period, compared outcomes between patients with a standard lipid profile to those evaluated with a comprehensive lipid profile. All adult members of the WellMed Medical Management, Inc. managed care health plan diagnosed with IHD or CHF, and continuously enrolled between July 1, 2006 and June 30, 2008, were included in the study. Cases were defined as those who had at least 1 comprehensive lipid test (the VAP [vertical auto profile] ultracentrifuge test) during this period (n=1767); they were compared to those who had no lipid testing or traditional standard lipid testing only (controls, n=289). Univariate statistics were analyzed to describe the groups, and bivariate t tests or chi-squares examined differences between the 2 cohorts. Multivariate regression analyses were performed to control for potential confounders. The results show that the case group had lower total costs ($4852.62 vs. $7413.18; P=0.0255), fewer inpatient stays (13.1% vs. 18.3% of controls; P=0.0175) and emergency department visits (11.9% vs. 15.6% of controls; P=0.0832). Prescription use and frequency of lipid measurement suggested improved control resulting from a targeted approach to managing specific dyslipidemias. A treatment protocol incorporating a comprehensive lipid profile appears to improve care and reduce utilization and costs in a disease management program for cardiac patients.

Prevalence of obesity, type II diabetes mellitus, hyperlipidemia, and hypertension in the United States: findings from the GE Centricity Electronic Medical Record database.

This study analyzed GE Centricity Electronic Medical Record (EMR) data to examine the effects of body mass index (BMI) and obesity, key risk factor components of metabolic syndrome, on the prevalence of 3 chronic diseases: type II diabetes mellitus, hyperlipidemia, and hypertension. These chronic diseases occur with high prevalence and impose high disease burdens. The rationale for using Centricity EMR data is 2-fold. First, EMRs may be a good source of BMI/obesity data, which are often underreported in surveys and administrative databases. Second, EMRs provide an ideal means to track variables over time and, thus, allow longitudinal analyses of relationships between risk factors and disease prevalence and progression. Analysis of Centricity EMR data showed associations of age, sex, race/ethnicity, and BMI with diagnosed prevalence of the 3 conditions. Results include uniform direct correlations between age and BMI and prevalence of each disease; uniformly greater disease prevalence for males than females; varying differences by race/ethnicity (ie, African Americans have the highest prevalence of diagnosed type II diabetes and hypertension, while whites have the highest prevalence of diagnosed hypertension); and adverse effects of comorbidities. The direct associations between BMI and disease prevalence are consistent for males and females and across all racial/ethnic groups. The results reported herein contribute to the growing literature about the adverse effects of obesity on chronic disease prevalence and about the potential value of EMR data to elucidate trends in disease prevalence and facilitate longitudinal analyses.

Comparison of GE Centricity Electronic Medical Record database and National Ambulatory Medical Care Survey findings on the prevalence of major conditions in the United States.

The study objective was to facilitate investigations by assessing the external validity and generalizability of the Centricity Electronic Medical Record (EMR) database and analytical results to the US population using the National Ambulatory Medical Care Survey (NAMCS) data and results as an appropriate validation resource. Demographic and diagnostic data from the NAMCS were compared to similar data from the Centricity EMR database, and the impact of the different methods of data collection was analyzed. Compared to NAMCS survey data on visits, Centricity EMR data shows higher proportions of visits by younger patients and by females. Other comparisons suggest more acute visits in Centricity and more chronic visits in NAMCS. The key finding from the Centricity EMR is more visits for the 13 chronic conditions highlighted in the NAMCS survey, with virtually all comparisons showing higher proportions in Centricity. Although data and results from Centricity and NAMCS are not perfectly comparable, once techniques are employed to deal with limitations, Centricity data appear more sensitive in capturing diagnoses, especially chronic diagnoses. Likely explanations include differences in data collection using the EMR versus the survey, particularly more comprehensive medical documentation requirements for the Centricity EMR and its inclusion of laboratory results and medication data collected over time, compared to the survey, which focused on the primary reason for that visit. It is likely that Centricity data reflect medical problems more accurately and provide a more accurate estimate of the distribution of diagnoses in ambulatory visits in the United States. Further research should address potential methodological approaches to maximize the validity and utility of EMR databases.

High rates of inappropriate drug use in the chronic pain population.

Chronic opioid treatment is a highly effective method to treat chronic pain; however, the prevalence of abuse of opioids can make treating patients with these agents difficult for clinicians. The objective of this study was to describe rates of inappropriate utilization, abuse, and diversion in a population of patients who were prescribed chronic opioids, as measured by urine drug testing in the clinical setting. A retrospective analysis was conducted of results from all urine drug tests conducted by Ameritox, Ltd. between January 2006 and January 2009, for patients whose physicians ordered the test in order to screen for noncompliance. Data from 938,586 patient test samples showed that 75% of patients were unlikely to be taking their medications in a manner consistent with their prescribed pain regimen. Thirty-eight percent of patients were found to have no detectable level of their prescribed medication, 29% had a nonprescribed medication present, 27% had a drug level higher than expected, 15% had a drug level lower than expected, and 11% had illicit drugs detected in their urine. Note that all categories add to a total greater than 100% as each category is not mutually exclusive, and a single patient could fall into multiple categories. The high observed rate of noncompliance demonstrates a significant clinical concern and confirms the importance of periodic urine drug screening for the population prescribed long-term opioid therapy.

58th annual meeting american society of human genetics, 2008.

The 2008 ASHG meeting, which took place from November 11 to 15, brought almost 6,500 attendees to Philadelphia to learn about the latest advances in human genetics and genomics research. This conference is considered the world's largest gathering of human genetics researchers, academicians, clinicians, genetic counselors, and nurses. This article reviews the 1,000 Genomes Project, Gaucher's disease, warfarin dosing, and Huntington's disease.

Use of an algorithm applied to urine drug screening to assess adherence to an oxycontin regimen.

OBJECTIVE: This study examined the ability of an algorithm applied to urine drug levels of oxycodone in healthy adult volunteers to differentiate among low, medium, and high doses of OxyContin. PARTICIPANTS AND INTERVENTIONS: Thirty-six healthy volunteers were randomized to receive 80, 160, or 240 mg of daily OxyContin to steady state while under a naltrexone blockade. During days 3 and 4 of the study, urine samples of all participants were collected, and oxycodone levels detected in the urine were obtained using a liquid chromatography-mass spectrometry (LC-MS-MS) assay. OUTCOME MEASURES: The concordance was calculated for raw and adjusted LC-MS-MS urine oxycodone values within each study participant between their third and fourth day values. Also, an analysis of medians was calculated for each of the dosage groupings using Bonett-Price confidence intervals for both raw and adjusted LC-MS-MS values. RESULTS: The concordance correlation coefficient for the raw LC-MS-MS values between days 3 and 4 was 0.689 (95% confidence intervals = 0.515, 0.864), whereas the concordance correlation coefficient for the LC-MS-MS values using the algorithm (ie, normalized values) was 0.882 (95% confidence intervals = 0.808, 0.956). Because of greater variability in the raw values, some overlap was observed in the confidence intervals of the various OxyContin doses, whereas no overlap was observed in the normalized confidence intervals regardless of the application of a Bonferroni adjustment. CONCLUSIONS: In contrast to raw LC-MS-MS values, an algorithm that normalizes oxycodone urine drug levels for pH, specific gravity, and lean body mass discriminates well among all three of the daily doses of OxyContin tested (80, 160, and 240 mg), even with correcting for multiple analyses.