: To determine if the instant approval (IA) process differs from the traditional prior authorization (PA) process in preferred drug channeling, resultant gaps in therapy, and provider dissatisfaction.

: An interrupted time series analysis using pharmacy claims and a retrospective cohort study.

: The study assessed changes in preferred drug use and subsequent cost reductions. A retrospective cohort study determined if the IA process produced fewer gaps in therapy than the PA process. Provider acceptance of the IA process was assessed using a brief survey of 240 randomly selected primary care practices.

: Market share for preferred proton pump inhibitors quadrupled from a range of 17.6% to 19.3% at baseline to 76% in the first month after implementation of the new IA policy. Most practices (81.1%) reported reduced administrative burden with the IA process. The median gaps between medication fills for patients using IA were approximately one-half those of patients using PA (

 <.001) and were one-fourth in a subset of highly adherent, regularly filling patients (

: Instant approval may be more patient friendly and prescriber friendly than PA as assessed by a proxy measure for access (gap in therapy) and physician-reported acceptance. Despite its ease of use, IA does not seem to reduce switching to preferred drugs.

	Instant approval may be a more patient-friendly and prescriber-friendly process than traditional prior authorization as assessed by a proxy measure for access (gap in therapy) and physician-reported acceptance. Despite its ease of use, instant approval does not seem to reduce switching to preferred drugs.

		Administrative burden is not needed to encourage preferred drug use.

		Prescribers prefer the path of least resistance.

		Patient gaps in therapy are substantially reduced by preempting hard edits at the pharmacy.

	Prior authorization (PA) is a widely adopted method of drug utilization management that requires submission of qualifying clinical or drug utilization criteria for review by a third party before acceptance of a prescription drug claim.

 Prior authorization programs have successfully contained drug costs over the past few decades, yet reports of unintended consequences have called into question their overall value.

 As implemented, PA often requires multiple telephone calls and facsimiles between pharmacy, practice, and a third-party administrator to gain resolution. In addition to causing clinician dissatisfaction,

 the process often results in patient attrition from therapy,

 and, in some instances, negative patient outcomes.

	In North Carolina, Medicaid has a long history of considering costcontainment policies that are palatable to clinicians and patients. Community Care of North Carolina (CCNC), an organization with more than 1000 primary care practice members focusing on medical home development and cost-effective care delivery, often works collaboratively with Medicaid administrators and policymakers to develop and implement statewide programs.

 In the case of a new coverage policy for proton pump inhibitors (PPIs) implemented in June 2007, North Carolina Medicaid consulted with CCNC and the North Carolina Physician Advisory Group to develop a modified PA program that allowed for an optional instant approval (IA) process. The IA process was available to all prescribers and could be used in lieu of the traditional third-party submission process. This alternative process allowed prescribers to document PA criteria directly on the prescription so the pharmacist could fill the prescription without delay (

	The North Carolina Medicaid PPI PA program (hereafter, the program) was developed using the best available clinical evidence for determining which PPI medications would require prior approval

 and included the following 3 prescribing options: (1) use of a “preferred” PPI, including omeprazole and Prilosec OTC (manufactured by AstraZeneca, Wilmington DE and distributed by Proctor and Gamble, Cincinnati, OH), without any coverage restrictions; (2) use of a “nonpreferred” PPI and the traditional PA process of submitting criteria to a third party; or (3) use of a nonpreferred PPI with approval criteria written on the prescription, bypassing the PA process. Written approval criteria included the following: (1) “failed omeprazole (40 mg) for 30 days,” (2) “esophagitis grade C” or “esophagitis grade D,” and (3) “cannot swallow tablets” or “cannot swallow capsules.”

	Community Care of North Carolina was charged with educating prescribers and pharmacies about the new process. Facsimile, e-mail, and face-to-face meetings were used before and after initiation of the policy.

	Ultimately, the program was constructed with the following 2 equally important objectives: (1) to generate drug cost savings and (2) to avoid some of the unintended consequences of traditional PA programs such as increased gaps in therapy and clinician dissatisfaction. In this study, we attempt to determine if both program objectives were met.

	Before implementation of the program on June 1, 2007, there were no coverage restrictions on PPI medications for Medicaid enrollees in North Carolina. This offered a unique natural experiment because the optional alternative IA process was initiated simultaneously with the PA start date. Three separate analyses were performed to evaluate program success. First, an interrupted time series analysis was used to evaluate drug cost savings and channeling to preferred PPIs. Second, gaps in therapy were determined as a proxy for patient access to PPIs by the type of override. Third, a survey was administered to primary care practices across North Carolina to assess provider satisfaction with the program. The study was approved by the institutional review board of the University of North Carolina School of Medicine at Chapel Hill.

	For the interrupted time series analysis, all North Carolina Medicaid recipients with paid PPI claims during 2007 were included. Per the PPI PA policy, enrollees younger than 6 years or enrollees having an aid category indicating pregnancy were excluded from the analysis. Dually eligible enrollees having Medicare Part D coverage were also excluded from the time series analysis regardless of age. For the cohort analysis of gaps in therapy, an additional exclusion criterion was added requiring continuous enrollment over the entire evaluation period to remove any potential bias imparted by loss of Medicaid eligibility.

	For the survey, a stratified random sample of 240 practices was selected from approximately 1000 CCNC practices, representing more than 85% of primary care delivery in North Carolina. The following 3 strata, comprising 80 practices each, were created based on practice-level PPI prescribing volume during the calendar year (2006) before the study period: (1) 1 to 9 prescriptions (80 of 386 practices), (2) 10 to 49 prescriptions (80 of 506 practices), and (3) 50 or more prescriptions (80 of 153 practices).

. Two interrupted time series were created using administrative pharmacy claims to assess the effectiveness of IA in reducing PPI costs and channeling patients to preferred medications. For the first time series, the market share of preferred PPIs as a percentage of total PPI use was computed and plotted for 12 sequential months (calendar year 2007). This period included 5 months immediately before and 7 months after program implementation. For the second time series, PPI cost totals for each of the 12 months were calculated. North Carolina Medicaid does not have a mandatory formulary and is the single provider of pharmacy benefits for all of its enrollees with no mail service option.

. Gaps in therapy were calculated for 5 patient cohorts distinct in their program experience (described herein). A gap in therapy is calculated by subtracting the total number of days of medication provided from the total number of days between consecutive fills. It represents the number of days a patient is likely without medication. To determine if the method of approval produced differing gap lengths, we selected all patients with consecutive PPI fills that surrounded the program implementation date. The preprogram PPI fill must have occurred in the 5 months before the start of the program, and the postprogram PPI fill must have occurred in the 5 months after the start of the program (

). For enrollees refilling their first postprogram prescription early, the gap was left censored to zero to avoid reporting nonsensical “negative gaps.”

	Because PPIs are customarily used chronically and on an as-needed basis, gap lengths were calculated for the entire Medicaid population and for a subset of highly adherent, regularly filling PPI users. Regular PPI users were defined as patients having at least 5 PPI fills with a total number of medication days exceeding 150 in the 5 months before program implementation.

	Five mutually exclusive cohorts, each representing a different patient-level result, were assembled for the analysis of gaps in therapy (

). An electronic record of pharmacy edits and overrides for IA and PA was used along with pharmacy claims to construct the cohorts. Cohort 1 (reference group) included patients filling preferred PPIs before and after program implementation. Approval was not required for this group at any time during the study period. Cohorts 2 through 5 represent patients using nonpreferred products immediately before the PA start date. Cohort 2 (switch group) included patients who ultimately switched from nonpreferred PPIs to preferred PPIs in their first fill after program implementation. Cohort 3 (IA group) included patients who continued filling nonpreferred PPIs on program implementation through the use of the IA process. Cohort 4 (PA group) included patients who continued filling nonpreferred PPIs on program implementation through the use of the traditional PA process. Cohort 5 (PA switch group) included patients who filled nonpreferred PPIs before program implementation, initiated the traditional PA process, but ultimately filled a preferred drug as their first PPI fill after program implementation. This group represents a cohort of patients who predominantly fall into 1 of the following 2 categories: (1) PA request was rejected by a third-party vendor or (2) PA request caused a delay sufficient to motivate the patient or prescriber to switch to a preferred product.

	After creating boxplots of gap lengths for each cohort, we performed Wilcoxon rank sum exact test (SAS, version 9.2; SAS Institute Inc, Cary, NC) to determine if gap days differed by cohort. For the nested group of regular PPI users, we also determined the number and percentage of patients having more than 10- and 30-day gaps as a proxy measure for quality. The latter is considered a significant gap in therapy as a consensus quality-of-care standard for many classes of chronic drugs.

. We also developed a survey to assess the adoption and satisfaction of the IA process among prescribers. The survey inquired about a practice’s use of IA, including provider and patient burden, as well as experience with and perceptions of IA and PA processes. A focus group of 2 physicians and a pharmacist was used to construct survey content (TKT, BDS). The draft survey was then pilot tested with physicians from 4 unaffiliated practices and was revised based on their responses by a panel familiar with the policy that included a biostatistician, an epidemiologist, a pharmacist, and a physician (TKT, BDS).

	The finalized survey was sent from the North Carolina Physician Advisory Group and included a cover letter, the survey, and a postage-paid addressed return envelope. Community Care of North Carolina staff delivered in person a second copy of the survey to nonresponding practices approximately 6 weeks after the initial distribution. Respondents were instructed to answer on behalf of their practice and were selected based on convenience (any prescriber in the practice could respond). Receipt of completed surveys and responses were entered into a spreadsheet (EXCEL, version 2003; Microsoft, Redmond, WA).

	Responses to the survey were summarized as frequencies (number and percentage) for each stratum and in aggregate with weighted and unweighted results. For each prescribing volume stratum, weighting was based on initial sampling proportions for each group. 

 Goodness-of-fit test was used to determine if responses differed between practices with varying volumes of PPI prescribing. 

 Contingency table test was applied to the Likert-type scale survey item.

	Tying Survey Results to Administrative Claims. Responses to the practice survey were then combined with pharmacy claims using a coded practice identifier maintained for CCNC. In North Carolina, almost three-quarters of all Medicaid enrollees are linked with a CCNC practice to participate in primary care medical home care management. This practice linkage allowed us to track PPI use and trends over time for practices that responded to the survey. Another interrupted time series model was then created to examine if practices claiming any use of IA produced fewer switches to preferred PPIs than practices not claiming any use of IA using traditional PA exclusively for PPI PA overrides.

	Market share for preferred PPIs remained stable at 17.6% to 19.3% in the preprogram period and increased to 76.3% in the first month of the program. Degradation of preferred market share was minimal (7.1%) in the 7 months after the start of the program (

). Subsequently, drug costs dropped from a pre-PA monthly mean of approximately $4 million to a post-PA monthly mean of approximately $1.5 million. Market share changes in the subset of CCNC-linked patients were similar for practices using IA and practices using PA (

	All gap length comparisons between the reference cohort and all other cohorts, as well as between the IA cohort and the PA and PA switch cohorts, were statistically significant (

 <.001, Wilcoxon 2-sample rank sum exact test). The median gap in therapy for patients using PA (26 days) was roughly twice the median gap in therapy for patients using IA (12 days) (

). For the subset of regular users, PA produced a gap (12 days) that was 4 times the gap (3 days) produced by IA (

 <.001). The percentage of regular users with more than a 30-day gap in therapy was significantly less for the users of the IA process (8.7%) compared with the users of the PA process (21.9%) (

 <.001). The ordered rank of fewest to most gap days by group was the same for all users group and regular users group: (Fewest Gap Days) → Reference Group → IA Group → Switch Group → PA Group → PA Switch Group → (Most Gap Days).

	Of 240 selected practices, 169 (70.4%) completed and returned IA practice surveys. Almost half (49.7%) of respondents reported using IA, with greater percentages of medium- volume (57.4%) and high-volume (67.9%) prescribers opting to use this method (

 <.001). Of those practices using IA, 81.1% believed that it reduced the administrative burden to the practice (

). Approximately three-fourths (73.4%) of respondents indicated that IA was better or much better than traditional PA, with only 7.0% reporting that traditional PA was better or much better. No differences were found between practices with high, medium, or low volumes of PPI prescribing with respect to IA or PA preference or reduced burden of IA use. The use of IA was pronounced, comprising 68.2% (4959 IA and 2308 PA) of all PA overrides in the first PPI fill after policy implementation.

	Despite the ease of use of IA, switching to preferred PPIs was similar to or exceeded that of PPI PA programs in other states and countries.

 Instant approval was a popular alternative to traditional PA, with survey responses indicating that almost half of all primary care practices in North Carolina used the optional IA process and that those practices preferred IA over PA more than 10 to 1 (73.4% believed that IA was better or much better, and 7.0% believed that PA was better or much better). Most importantly, offering an alternative IA did not seem to diminish the ability of PA to produce substantial cost savings.

	Many states are testing various modalities for drug utilization management that attempt to be more patient friendly and prescriber friendly. One emerging method is SmartPA, a system that reviews a patient’s claims history to determine eligibility for an instant override at the pharmacy.

 Another evolving method deployed alongside e-prescribing would allow the prescriber to affirm or submit PA criteria with real-time links to formulary and eligibility. Nonelectronic methods include collaborative practice agreements such as the Therapeutic Interchange Program used in Washington state.

	The IA method used for North Carolina Medicaid was unique in its simplicity of administration. The prescriber was given the authority to override the PA on his or her own form (the prescription pad) at the ideal time (the point of prescribing). Subsequently, practices reported a median time savings of 15 minutes per patient using IA. As such, IA may reduce or totally eliminate the indirect PA submission cost to the practice (estimated as high as $17.77).

 Similarly, IA may reduce implementation costs to the payer (estimated at $20-$25 per enrollee per submission).

	This study is limited by its quasi-experimental design, although an experimental design would have been impractical because PA programs are universally applied to enrolled populations. It is possible that factors may have been present that influenced patient selection into the type of override process used by the prescriber. However, the likelihood that observed differences in therapy gaps were due to drug or patient character stics is low because the net result (extended gap length) is intrinsic to the override process itself, and prescriber choice of IA versus PA was not likely influenced by drug or patient. Therefore, in the absence of large differences in baseline adherence between patients of prescribers using PA versus IA, a causal link may be inferred between override type and gap length.

	We did not attempt to make a pre—post comparison of gaps in therapy but rather choose to create a cross-sectional view of gaps using the first post-PA fill as a patient-specific index date. This is the most relevant date because it represents the day when rejection occurs at the pharmacy in the absence of an override. Although gap measurements are a convenient proxy measure for access,8 the approach lacks specificity and precision because of baseline “noise” caused by well-known and widespread nonadherence among all drug classes. Fortunately, the inclusion of a reference group of PPI users not subject to the effects of the PA policy was available for comparison to mitigate this limitation. In addition, we analyzed a nested cohort of highly adherent users for gaps after PA implementation to “zero out” poor baseline adherence. We found identical ordering of cohorts, differing only in gap magnitude, but with similar relative gap lengths in this highly adherent group.

	The generalizability of results may be limited by the drug class (PPIs) and the payer (Medicaid). It is unclear whether the results of this study would be replicated if IA was used for other drug classes that maintain stronger indications for chronic use or more serious or immediate consequences from delayed access. With respect to the survey, the available unit of analysis was the practice; therefore, a single prescriber was asked to respond on behalf of each practice, imparting the possibility of response bias based on self-selection. Furthermore, the sampling frame was limited to primary care practices within the CCNC membership. Although CCNC represents more than 85% of all primary care practices in North Carolina, future work should make every attempt to include specialists, subspecialists, and institutionally based prescribers.

	No formal time series hypothesis testing was performed. The magnitude and timing of the market share changes immediately after the policy implementation were so strong that it removed the need for inferential or other statistical analyses in time series. We did not attempt to measure attrition from therapy because a comparative population of nonpreferred PPI users subject to a PA policy without the IA process would have been required. We also made no attempt to perform a time-in-motion analysis or other continuous quality improvement exercise that should be considered if IA processes proliferate. Our emphasis herein was to determine the perceived palatability of the IA method from the prescriber perspective and its effect on medication access from the patient perspective.

Articles

WEB EXCLUSIVE

A policy allowing prescribers to write prior authorization criteria directly on the prescription led to decreased gaps in therapy without diminishing drug cost savings.

A Physician-Friendly Alternative to Prior Authorization for Prescription Drugs

: To assess the feasibility of using medical and prescription drug claims data to develop models that identify patients at risk for prescription opioid abuse or misuse.

: Deidentified prescription drug and medical claims for approximately 632,000 privately insured patients in Maine from 2005 to 2006 were used. Patients receiving prescription opioids were divided into 2 mutually exclusive groups, namely, prescription opioid abusers and nonabusers.

: Potential risk factors for prescription opioid abuse were incorporated into logistic models to identify their effects on the probability that a prescription opioid user was diagnosed as having prescription opioid abuse. Different models were based on data available to prescriptionmonitoring programs and managed care organizations. Best-fitting models were identified based on statistical significance (

 ≤.05), parsimony, clinical relevance, and area under the receiver operating characteristic curve.

: The drug claims models found that the following factors (measured over a 3-month period) were associated with risk for prescription opioid abuse: age 18 to 34 years, male sex, 4 or more opioid prescriptions, opioid prescriptions from 2 or more pharmacies, early prescription opioid refills, escalating morphine sulfate dosages, and opioid prescriptions from 2 or more physicians. The model integrating drug and medical claims found that the following factors (measured over a 12-month period) were associated with risk for prescription opioid abuse or misuse: age 18 to 24 years, male sex, 12 or more opioid prescriptions, opioid prescriptions from 3 or more pharmacies, early prescription opioid refills, escalating morphine dosages, psychiatric outpatient visits, hospital visits, and diagnoses of nonopioid substance abuse, depression, posttraumatic stress disorder, and hepatitis.

: Using drug and medical claims data, it is feasible to develop models that could assist prescription-monitoring programs, payers, and healthcare providers in evaluating patient characteristics associated with elevated risk for prescription opioid abuse.

)It is feasible to use claims data to develop models that estimate risk of prescription opioid abuse. Prescription-monitoring programs and managed care organizations have data that may be used to implement such models. Consequently, there is potential for early identification of patients at risk, which may improve patient outcomes through earlier detection and treatment.

Age 18 to 34 years and male sex were significant demographic risk factors.

Filling opioid prescriptions at multiple pharmacies and refilling opioid prescriptions early were significant drug use—related risk factors.

Clinicians should be cautious in applying such statistical models to individual patients, who should each be considered individually in light of such risk factors.

Prescription opioid abuse or misuse (ie, intentional exposure to rescription opioids in a manner inconsistent with the use they were prescribed for) has become a major societal issue.

 The negative health consequences resulting from prescription opioid abuse or misuse include emergency department visits, hospitalizations, and substance abuse treatment admissions. The National Survey on Drug Use and Health estimates that from 2002 to 2007 an annual mean of 11.8 million Americans used prescription pain relievers for nonmedical purposes.

 The number of patients admitted for inpatient treatment of prescription opioid abuse or misuse has increased from approximately 46,000 in 2002 to more than 81,000 in 2007.

 Furthermore, the number of emergency department visits related to nonmedical use of opiates or opioid analgesics has jumped from fewer than 180,000 in 2004 to more than 250,000 in 2006.

 These increases in prescription opioid abuse or misuse underscore its importance as a major public health issue.

Programs that combat the rise of prescription opioid abuse or misuse (such as prescription-monitoring programs [PMPs]) exist in various forms.

 As of December 2007, a total of 26 states had launched PMPs; 9 more had legislation requiring their creation, while several others were considering similar legislation.

 Many speculate about the  benefits of PMPs, but little empirical evidence exists.

Hypothesized advantages include identification of potentially inappropriate prescribing or dispensing practices and effective ongoing monitoring of patients receiving prescription opioid therapy, including identification of at-risk patients to physicians via warning letters. Other benefits may include the ability to provide better care and reduce costs associated with abuse-related events (eg, emergency department visits). Prescription-monitoring programs have recognized the limitations inherent in interpreting their data and have set up external professional committees to limit unintended negative consequences (eg, false positives). Nevertheless, some concern exists that PMPs may produce a “chilling effect” as discussed by Wastila and Bishop.

 For example, further research by Simoni-Wastila et al

 demonstrated that the requirement by New York that physicians use state-monitored prescription forms for benzodiazepines led to reduced benzodiazepine use.

Although risk factors associated with drug abuse have been studied, few published studies have used data similar to those available to PMPs and managed care organizations (MCOs) to develop models that identify patients at risk for prescription opioid abuse or misuse.

 While some PMPs use basic algorithms or thresholds to identify potential at-risk patients, combining more detailed claims-based algorithms with clinical measures of abuse may enhance their ability to identify those at risk. Our objective was to assess the feasibility of using medical and prescription drug claims data to develop models that identify patients at risk for prescription opioid abuse or misuse using observed outcomes measures (ie, based on 

International Classification of Diseases, Ninth Revision, Clinical Modification

This study used deidentified medical and prescription drug claims data provided by the Maine Health Data Organization

 for approximately 632,000 privately insured residents of Maine (from 2005 to 2006). Data on the prescribing physician were only available beginning in September 2006. Patients aged 12 to 64 years with at least 1 claim for a prescription opioid and at least 1 medical claim from 2005 to 2006 were identified (ie, the prescription opioid user population). This population was further divided into 2 mutually exclusive groups, namely, prescription opioid abusers and nonabusers. Prescription opioid abusers were defined as the subset of all prescription opioid users having at least 1 medical claim associated with any of the following 

 codes from 2005 to 2006: 304.0 (opioid- type dependence), 304.7 (combinations of opioid type with any other), 305.5 (opioid abuse), or 965.0 (poisoning by opiates or related narcotics but excluding 965.01 [heroin poisoning]). The nonabusers consisted of the remaining prescription opioid users, who had no claims associated with the 4 abuse codes.

 codes to identify opioid abusers is consistent with prior research.

 However, there are other instances of abuse-related phenomena (eg, “nonmedical use of prescription drugs” as measured by the National Survey on Drug Use and Health) that this study does not identify and which may occur in the nonabuser sample.

Potential risk factors for prescription opioid abuse were incorporated into models based on prior analyses of prescription opioid utilization patterns and clinical knowledge of comorbidities associated with prescription opioid abuse.

 Stepwise logistic regression models were used to evaluate the effects of independent variables corresponding to these risk factors. Each model began with a core set of demographic variables that categorized patients by sex and age group. Then, additional variables related to drug use, comorbidities, and medical services were added based on their statistical or clinical significance. Independent variables are listed and described in more detail in 

The models took the following general form:

 are variables of the model to be estimated, and 

 patients. Note that the general model describes associations rather than causation.

Three general logistic models were developed, namely, 2 based on drug claims alone and 1 that integrates data from drug and medical claims. The drug claims models serve as approximations for the data available to a PMP director, whereas the integrated model represents the data available to MCOs. For the models based only on drug claims, drug utilization behavior was assessed during the 3-month period following the index date (because many PMPs review data on a quarterly basis). The first variation excludes information on prescribing physicians, while the second variation incorporates such information but requires the use of an abbreviated time frame because the Maine data did not identify prescribing physicians until the last quarter of 2006. The third (integrated) model builds on the first by also including information on healthcare services and comorbidities from medical claims to estimate a more comprehensive model. Patient characteristics were measured over the 12-month period following the index date. To identify the best-fitting models, the following 4 criteria were used:

Area Under the Receiver Operating Characteristic Curve. The performance of different algorithms was evaluated by comparing the area under the receiver operating characteristic (ROC) curves using SAS (Version 9.1, SAS Institute, Inc, Cary, NC). The ROC curves were used to assess overall fit (eg, a model that has a greater area under the ROC curve will have a better fit). This study sought to optimize the tradeoff between sensitivity and specificity. That is, a curve with the highest sensitivity (ie, the most correct identifications of abuse) for any given level of specificity (ie, false positives) was preferred. The C statistic was used to evaluate the area under the ROC curve.

. A model with a smaller set of relevant independent variables was desired. A parsimonious model is more practical and places lesser data burden on potential users.

 <.05 were preferred so that the model would focus on variables with the most statistical power. The 5% level was used in initial model building to identify potentially relevant risk factors. However, demographic variables were included regardless of their 

. The model focused on independent variables that were clinically relevant based on existing literature and expert opinion. We generally adopted a 

 <.05 criterion but considered all 4 selection criteria in arriving at a final decision. Therefore, there were times when the model included variables with an odds ratio (OR) close to 1 because we believed a priori that the variable was clinically relevant.

Alternative sets of results are presented that correspond to the 2 models evaluating only drug claims data. The first describes results for the full sample of 116,382 prescription opioid users (alternative A), while the second summarizes results for the subset of 8592 patients for whom sufficient information was available to construct a “physician shopping” variable (alternative B). Because of the short duration of the data, no eligibility criteria were imposed, so patients could have potentially cycled in and out of the data.

There were 116,382 patients who received at least 1 prescription opioid (approximately 15% of the privately insured population of Maine). Of the prescription opioid user sample, 875 patients were diagnosed as being prescription opioid abusers from 2005 to 2006 (ie, 0.8% of users were abusers based on 

). The prescription opioid user population was predominantly female (55.7%), while the abuser population was predominantly male (59.1%). Most prescription opioid users (26.7%) were aged 45 to 54 years. In contrast, most prescription opioid abusers (24.6%) were aged 25 to 34 years. Patients aged 18 to 34 years accounted for 48.5% of the abuser population but only 25.0% of the overall sample. Several utilization behaviors were significantly more prevalent among abusers than nonabusers, including filling prescriptions at 2 or more pharmacies (39.4% vs 7.7%), having 1 or more early refills (36.0% vs 3.7%), and demonstrating 2 or more consecutive months of dose escalation (4.7% vs 0.4%).

The best-fitting model for alternative A included the following variables: age, sex, number of pharmacies, number of prescriptions, early prescription opioid refills, and dose escalation (

). The ROC curve had a C statistic of 0.840, and the pseudo 

 was 0.21. All variables were significant at the 5% level except for age 35 to 44 years (included for purposes of comparison). Variables indicating that a patient was aged 18 to 24 years, received 4 or more opioid prescriptions, filled opioid prescriptions at 2 or more pharmacies, or exhibited early refill behavior all had ORs above 2.00, meaning that each factor at least doubled the odds of a patient being at risk for prescription opioid abuse. Because the model controls for the number of opioid prescriptions, the ORs for the other variables indicate that those measures were associated with prescription opioid abuse diagnoses and were not merely statistical consequences of long-term opioid therapy.

As already mentioned, because Maine did not require third-party payers to record the prescribing physician’s identifier until June 30, 2006, this analysis was limited to the final quarter of 2006.

 This subsample of patients with a prescriber identifier for a prescription opioid included 8592 patients (7.4% of the total patients from alternative A). The demographic characteristics (

) were similar to those in alternative A. Alternative B covers a shorter period to include a metric on physician shopping.

The best-fitting model for alternative B included the following variables: age, sex, number of pharmacies, early prescription opioid refills, and number of physicians (

). The physician shopping variable was statistically significant at the 10% level but not at the 5% level. The ROC curve had a C statistic of 0.774, and the pseudo 

 was 0.15. The lower C statistic and pseudo 

 values are likely indicative of the shorter interval and more limited data available for this model. All variables were significant at the 10% level (and at the 5% level) except for ages 25 to 34 years and 35 to 44 years, as well as the physician shopping variable. Early prescription opioid refills (OR, 3.31; 95% confidence interval [CI], 2.56-4.28) and the age bracket of 18 to 24 years (OR, 3.53; 95% CI, 1.64-7.61) seemed particularly important. In addition, variables indicating that a patient was male, filled opioid prescriptions at 2 or more pharmacies, or was aged 25 to 34 years all had ORs above 1.50. The number of opioid prescriptions was considered for this model as well but was omitted from the best-fitting model because it was highly correlated with the physician shopping variable. It is difficult to compare the C statistics of alternatives A and B given the much smaller sample size of the latter (116,382 vs 8592).

Comorbidities were initially identified based on prior research.

 The best-fitting integrated model included the following variables: age, sex, nonopioid substance abuse, depression, posttraumatic stress disorder, hepatitis, cancer, mental health outpatient facility visits, hospital visits, number of pharmacies, number of prescriptions, and early prescription opioid refills (

). The ROC curve had a C statistic of 0.926. With a pseudo 

 of 0.37, this model provided a much better fit to the data than either drug claims model. All  variables listed in eAppendix A were considered in the stepwise regression analysis. Early prescription opioid refills was most important (OR, 6.52; 95% CI, 5.39-7.89), followed by nonopioid substance abuse (OR, 5.83; 95% CI, 5.03-6.75). The categorization for the number of pharmacies where opioid prescriptions were filled differs slightly because the data period in Table 5 differs from that in Tables 1 through 4.

This study demonstrates the feasibility of developing models derived from claims data variables to identify specific characteristics associated with elevated risk for prescription opioid abuse. Several identified risk factors increased the likelihood that a patient may be at risk for abusing or misusing prescription opioids. The ability of the models to precisely identify patients who received an 

 diagnosis for abuse or misuse was less robust, which may be explained by the few patients with such clinical diagnoses, as well as other limitations of the data used and of claims data in general for assessing patient risk levels.

Nonetheless, such models may serve as a valuable tool to assist clinicians in monitoring their patients. As Katz et al state, “healthcare providers will feel more comfortable prescribing and dispensing opioids when they are able to better identify patients at risk for abuse, and will be able to intervene in problematic cases to minimize risk.”

 For instance, PMPs could automatically screen all patients receiving prescription opioids using this type of model and generate reports of patients with certain characteristics (eg, physician shopping and multiple opioid prescriptions), or they could notify providers if patients meet certain key risk factors such as those identified in this study (eg, pharmacy shopping and early refills). Several state PMPs have indicated an interest in using drug data as a clinical tool.

In practice, clinicians must carefully balance reducing prescription opioid abuse with supporting appropriate prescription opioid therapy for patients in pain. It is important to avoid falsely identifying patients as having increased risk who do not have prescription opioid abuse problems (the difficulty experienced by physicians in identifying patients with prescription opioid abuse has been discussed in the literature

). If claims-based models are used to accomplish this goal, cutoffs with high specificity (low false-positive rates) should be chosen. For example, using the best-fitting alternative A model set at 95% specificity, the sensitivity (proportion of individuals with prescription opioid abuse identified by the model) was 54%. Based on the Maine data population, a PMP director sending out “warning letters” using this cutoff would send 8792 letters, of which only 473 would be true positives. The implication of this is that almost 95% of the letters (8319 of 8792) sent out would be false positives. Drug data alone (all that are available to PMPs) could identify more than half of the abusers in this population. Although the number of false positives may seem large in absolute terms, it represents only about 5% of all prescription opioid users (some of whom may be abusers or misusers who have not received an 

 diagnosis or patients potentially at future risk). Low base rates of diagnosed abuse or misuse (only 0.8% of this population was diagnosed as having prescription opioid abuse) also contribute to the high absolute number of false positives.

By comparison, the integrated model illustrates the potential for improving these models by constructing and validating more refined algorithms. It also highlights the potential role for MCOs (which have access to patient medical claims in addition to drug claims) in the monitoring process. The increase in sensitivity of the integrated model best illustrates the improved accuracy in identifying at-risk patients produced by combining medical and drug data. At 95% specificity, the integrated model yields a sensitivity of 71%, a substantial improvement over the 54% sensitivity achieved using drug claims alone. Compared with the drug claims model, a program administrator using the same 95% specificity would correctly identify an additional 143 abusers (a 31% improvement). Therefore, combining medical and drug claims is significantly more effective at identifying prescription opioid abuse or misuse than using drug claims alone. Still, from the perspective of a PMP, creating algorithms based on drug claims data may be an important first step toward identifying potential prescription opioid abusers or misusers.

In any event, physicians using such algorithms would need to take great care to incorporate this information into the full clinical picture to avoid taking inappropriate action. From a clinician’s perspective, reports generated based on claims data could serve as an additional data point for consideration in light of other sources (eg, patient histories, physical examinations, laboratory test results, and ongoing monitoring).

 Physicians would still require diagnostic confirmation through a comprehensive clinical assessment. Used in conjunction with these other assessment tools, claims-based models have the potential to aid in detecting prescription opioid abuse or misuse and allow physicians to prescribe prescription opioids to patients more confidently.

There are some limitations to our study and other directions for future research. Although they are the only option for claims data analysis, the use of 

 codes to identify prescription opioid abuse has important limitations. First, these codes are likely to identify a mixture of individuals with heroin and prescription opioid abuse, with an unknown proportion of each. To mitigate this problem, the analysis required that all patients have at least 1 prescription opioid claim; however, heroin abusers may also be prescription opioid users. Second, the true extent of prescription opioid abuse is likely understated because many patients who are abusers may not receive an abuse diagnosis for several reasons (such as the stigma associated with the condition).

 Moreover, PMPs always understate the true prevalence of prescription opioid abuse or misuse because they focus on practitioners and patients and do not characterize the many other sources of illicit use of prescription drugs.

In addition, the integrated model could be further improved. The duration of the data was short (2-year maximum). Furthermore, the prescribing physician identifier was only available for 3 months. Algorithms based on several years of claims data may have more explanatory power. Also, drug data were only available for prescriptions that were paid for by private insurance; individuals who paid using cash or by other payers were not included. Linking medical claims data with more comprehensive prescription data could yield a more robust algorithm. Moreover, given the short period covered by the data, it was not possible to assess the sequencing of medical comorbidities (eg, to test the hypothesis that a specific prior  diagnosis such as alcohol abuse would lead to a future opioid abuse diagnosis). Such enhancements would further improve our understanding of the risk factors. Finally, the ability to generalize the data to other geographic areas is unknown. The behavioral patterns of prescription opioid abusers may vary across geographic and demographic groups.

A possible interpretation of the data is that the patients classified as prescription opioid abusers also have pain that is undertreated and are engaging in aberrant behaviors because of pseudoaddiction. Given the notable comorbidity of pain and psychiatric disorders (including addiction and abuse), it is possible that high-risk behaviors such as early refills and physician shopping are palliation attempts rather than true abuse or dependence. This scenario would increase the likelihood of a false-positive identification using these data. Such a consideration is beyond the scope of the present research.

Models such as those used in this study can help successfully identify patients at risk for prescription opioid abuse or misuse. Identifying characteristics of patients who abuse or misuse prescription opioids may help true abusers get treatment sooner, which could yield substantial improvements in patient outcomes and reduced costs. In this way, PMPs and MCOs can have an important role in assisting healthcare providers and their patients to minimize risk of prescription opioid abuse or misuse.

: From Analysis Group, Inc (AGW, HGB, MS, JT), Boston, MA; Department of Anesthesiology (NPK), Tufts University, Boston, MA; Analgesic Research (NPK), Needham, MA; and Inflexxion, Inc (NPK), Newton, MA.

: This research project was funded by Alpharma Pharmaceuticals, LLC.

: Drs White and Birnbaum and Messrs Schiller and Tang are employees of Analysis Group, Inc, which received payment from Alpharma Pharmaceuticals, LLC, for its involvement in the development and preparation of this study. Dr Katz reports serving as a paid consultant to Alpharma Pharmaceuticals, LLC, and Analysis Group, Inc.

: Concept and design (AGW, HGB, MS, NPK); acquisition of data (AGW, HGB, MS); analysis and interpretation of data (AGW, HGB, MS, JT, NPK); drafting of the manuscript (AGW, HGB, MS, NPK); critical revision of the manuscript for important intellectual content (AGW, HGB, MS); statistical analysis (AGW, MS, JT); provision of study materials or patients (AGW); obtaining funding (AGW, HGB); administrative, technical, or logistic support (AGW, HGB, MS, JT); and supervision (AGW, HGB).

: Alan G. White, PhD, Analysis Group, Inc, 111 Huntington Ave, 10th Fl, Boston, MA 02199. E-mail: awhite@analysisgroup.com.

1. Substance Abuse and Mental Health Services Administration

Results From the 2004 National Survey on Drug Use and Health: National Findings

. Rockville, MD: Substance Abuse and Mental Health Services Administration, Office of Applied Studies; 2005. NSDUH series H-28, DHHS publication SMA 05-4062.

2. Substance Abuse and Mental Health Services Administration

Drug Abuse Warning Network, 2004: National Estimates of Drug-Related Emergency Department Visits

. Rockville, MD: Substance Abuse and Mental Health Services Administration, Office of Applied Studies; 2006. DAWN series D-28, DHHS publication SMA 06-4143.

. Opioid analgesics and rates of fatal drug poisoning in the United States. 

4. Johnston LD, O’Malley PM, Bachman JG, Schulenberg JE

Monitoring the Future: National Results on Adolescent Drug Use: Overview of Key Findings, 2005. Bethesda, MD: National Institute on Drug Abuse; 2006

. Prevalence of opioid analgesic injection among rural nonmedical opioid analgesic users. 

. Chronic opioid therapy for nonmalignant pain in patients with a history of substance abuse: report of 20 cases. 

7. Substance Abuse and Mental Health Services Administration

Results From the 2007 National Survey on Drug Use and Health: National Findings

. Rockville, MD: Office of Applied Studies; 2008. NSDUH series H-34, DHHS publication SMA 05-4343.

8. Substance Abuse and Mental Health Services Administration Web site

Quick Statistics From the Drug and Alcohol Services Information System: Treatment Episode Data Set (TEDS)

. Rockville, MD: Substance Abuse and Mental Health Services Administration, Office of Applied Studies. http://wwwdasis.samhsa.gov/webt/tedsweb/tab_year.choose_year_web_table?t_state=US. Accessed January 29, 2009.

9. Substance Abuse and Mental Health Services Administration

Drug Abuse Warning Network, 2006: National Estimates of Drug-Related Emergency Department Visits

. Rockville, MD: Substance Abuse and Mental Health Services Administration, Office of Applied Studies; 2008. DAWN series D-30, DHHS publication SMA 08-4339.

. Major increases in opioid analgesic abuse in the United States: concerns and strategies. 

11. Katz NP, Adams EH, Benneyan JC, et al

. Foundations of opioid risk management. 

12. Drug Enforcement Administration Web site, Office of Diversion Control

Questions & answers: state prescription monitoring programs

. Updated November 2008. http://www.deadiversion.usdoj.gov/faq/rx_monitor.htm. Accessed October 8, 2009.

. Maximizing the value of electronic prescription monitoring programs. 

14. Fishman SM, Papazian JS, Gonzalez S, Riches PS, Gilson A

. Regulating opioid prescribing through prescription monitoring programs: balancing drug diversion and treatment of pain. 

15. Alliance of States With Prescription Monitoring Programs Web site

. http://www.nascsa.org/AlliancePDF/Goals.pdf. Accessed December 10, 2007.

16. Joranson DE, Carrow GM, Ryan KM, et a

l. Pain management and prescription monitoring. 

. Balancing diversion control and medical necessity: the case of prescription drugs with abuse potential. 

POLICY

This study evaluates the feasibility of using claims data to evaluate risk factors for prescription opioid abuse among patients in a privately insured population.

Analytic Models to Identify Patients at Risk for Prescription Opioid Abuse

: To evaluate the effect on adherence and medical care expenditures of a pharmacy benefit change that included free generic drugs and higher copayments for brand-name drugs. 

: Quasi-experimental pre—post study of patients with ischemic heart disease (1286 control and 555 intervention) and patients with diabetes mellitus (4089 control and 1846 intervention).

: Medical and pharmacy claims data were analyzed for  continuously enrolled members from January 1, 2005, through December 31, 2008. A generalized linear model was used to predict costs as adherence changed. 

: The rate of switching from brand-name drugs to generic drugs in the intervention group was not statistically different from that in the control group. The net change in adherence was higher only for the intervention group patients taking statins who switched to generic drugs, a 6.2% increase compared with an 8.5% decrease in the control group. The estimate of medical cost savings attributable to this benefit change was significant for only the metformin class of diabetes drugs. Improved adherence independent of this benefit change was estimated to reduce all-cause medical costs for patients taking sulfonylureas, metformin, and thiazolidinediones.

: Altering copayments for pharmaceuticals may affect the rate of conversion to generic drugs but is unlikely in and of itself to result in complete conversion. However, increasing adherence can result in net savings for specific diabetic drug classes, as savings from all-cause medical costs offset the increase in pharmacy costs.

A pharmacy benefit design change increasing copayments on brand-name drugs and simultaneously introducing a zero copayment on generic drugs had little influence on patients’ switching to generic drugs or on improving adherence to medication regimens.

Altering copayments for generic drugs and brand-name drugs may affect the rate of conversion to generic drugs but is unlikely in and of itself to result in complete conversion to generic drugs.

Small increases and decreases in pharmacy copayments have little effect on adherence or medical costs.

Improving adherence lowered all-cause medical costs for 3 diabetes drug classes but raised pharmacy costs for all 5 drug classes studied.

The increasing cost of medical care in the United States is an ongoing concern, consuming more than 16.2% of the nation’s 2006 gross national product.

 Government, payers, business leaders, consumers, and healthcare providers are seeking to control healthcare costs without sacrificing quality of care. One such effort is improving the management of chronic conditions. Caring for chronic conditions in the United States consumes 78% of all healthcare dollars,

 meaning that such an effort has the potential for significant effect.

Payers and providers have implemented several approaches to improve the management of chronic conditions. Disease management programs emphasize one-on-one contact with health coaches and nurses.

 Physician practice models that incorporate aspects of a medical home

 focus on increased involvement of healthcare providers in a team approach. Using insurance benefit design to lower barriers to therapies that effectively manage chronic conditions is the focus of this analysis. Value-based plan design promotes the use of care when the benefits exceed the costs and discourages the use of care when the costs exceed the benefits.

Value-based plan design that focuses on reducing out-of-pocket costs for prescription drugs taken by patients with chronic diseases is premised on a relationship between improved medication adherence and better condition management and between better condition management and lower costs. Most research has focused on the relationship between prescription drug out-of-pocket costs and adherence or on the relationship between adherence and medical cost or utilization, although some studies

 considered all 3 factors. Studies of cost and adherence have demonstrated the following mixed results: increasing pharmacy cost sharing had no effect on adherence,

 increasing pharmacy cost sharing decreased adherence,

 and decreasing pharmacy cost sharing increased adherence. 

 This apparent contradiction may be explained by the conditions studied, the adherence measures used, or the sample size.

Studies of the relationship between adherence and health outcomes or costs and utilization have had more conclusive findings. For example, lower adherence has been associated with worse outcomes, including mortality, in populations with diabetes mellitus or acute myocardial infarction.

 A relationship between increased adherence and fewer emergency department visits and hospitalizations has been observed.6,7,18-20 While some studies

 have shown overall medical cost savings as a result of adherence, other studies

 have not in part because the increase in pharmacy costs exceeded the medical cost savings.

This study relies on a natural experiment created when Blue Cross Blue Shield of Minnesota introduced a free generic drug benefit in July 2006 to groups with fewer than 51 employees. The new benefit eliminated the $5 copayment for generic drugs and increased by $5 the copayment for brandname and nonformulary drugs to $35 and $50, respectively. We hypothesized that adherence would increase as member costs for generic drugs decreased and that adherence would decrease as member costs for brand-name drugs increased. We also hypothesized that improved adherence would lead to better condition management and to lower medical costs. This study extends existing research by quantifying the net effect on medical and pharmacy costs of a net change in adherence due to a simultaneous increase and decrease in pharmacy copayments.

Small increases and decreases in pharmacy copayments have little influence on patients’ switching to generic drugs or on improving adherence to medication regimens.

December 2009

A Physician-Friendly Alternative to Prior Authorization for Prescription Drugs

Analytic Models to Identify Patients at Risk for Prescription Opioid Abuse

Plan Designs That Encourage the Use of Generic Drugs Over Brand-Name Drugs: An Analysis of a Free Generic Benefit

Effect of Antihypertensive Medication Adherence Among Employees With Hypertension

Managed Care Quality and Disenrollment in New York SCHIP

Can Managed Care Plans Reliably Infer the Quality of Cardiac Surgeons' Outcomes?

Variation in Outpatient Antibiotic Prescribing in the United States

Pharmaceuticals From Development to Practice: Learnings From an Online Curriculum