Ryan M. Conrad, PhD

 We measure the impact of Medicaid formulary restrictions (FRs) on the rate at which patients who previously failed a drug therapy for schizophrenia are returned to that therapy.

 We collect drug-level information on FRs in state Medicaid programs and examine claims of noninstitutionalized Medicaid enrollees with schizophrenia.

A difference-in-differences technique is used to compute the change in the probability of adverse outcomes before and after a state adopts an FR. This change is compared with the change in failure probabilities in states with no FRs.

 Regardless of FRs, patients tend to resume the same drug after an adverse medical event. In 2005, 69% of inpatient mental health—related admissions resulted in patients resuming the same therapy within 6 months of the event, and 63% of patients resumed the same drug after a mental health–related emergency department admission. In states where FRs limit access to all atypicals, the likelihood of a patient resuming the same atypical after having ceased treatment for at least 30 days increases by 20.1% relative to patients in states without restrictions. Additionally, patients in states that impose FRs on all atypicals are 11.6% more likely to discontinue all treatments.

 FR may increase the likelihood that patients will return to failed treatments or cease treatment altogether. Although formularies are designed to reduce drug spending, an unintended consequence may be an increase in the use of other services needed to treat patients with schizophrenia.

We measure the impact of Medicaid formulary restrictions (FRs) on the rate at which patients who previously failed on a particular drug therapy for schizophrenia are returned to that same therapy. Our findings suggest that FRs increase the likelihood that patients will return to failed treatments, or totally cease treatment.

		In 2005, 69% of all schizophrenia patients’ mental health—related admissions resulted in patients resuming the same schizophrenia treatment within 6 months

		Patients in states that impose FRs on all atypical antipsychotics are 20.1% more likely to resume use of the same atypical after having ceased treatment for at least 30 days and 11.6% more likely to discontinue all treatments relative to patients in states without restrictions.

Between 1995 and 2005, Medicaid spending on drugs increased an average of 13% per year—an increase due in part to physicians favoring newer, more expensive drugs.

 To attempt to control costs, many state Medicaid programs responded by imposing policies that encourage the use of lower-cost drugs.

Prior authorization, for example, is designed to reduce pharmacy spending by requiring physicians to seek approval from the state’s Medicaid program before prescribing certain high-cost drugs to patients. These controls are designed to reduce the demand for expensive drugs when lowercost alternatives exist, which ultimately results in reduced Medicaid spending. But these policies may affect healthcare spending and utilization through other channels, especially if patients do not respond well to the preferred drugs. Physicians match patients to the drug and dosage that is optimal for each patient based on various practice guidelines, recommendations, and formulary practices.

On average, several treatments might be equally effective at the population level, but heterogeneous patients will respond differently to each. Removing a drug from coverage could adversely affect those who would have most benefited from it. If the use of these drugs leads to deteriorating health status, the result may be higher nonpharmaceutical medical spending.

An important example of this trade-off is the second generation of drugs for the treatment of schizophrenia, called atypical antipsychotics, which account for almost 15% of Medicaid spending.

 For this reason, they are increasingly popular targets of state Medicaid formulary restrictions.

 Because a number of patients may respond only to 1 drug, formulary restrictions can place patients with idiosyncratic treatment responses at greater risk of uncontrolled symptoms.

 Uncontrolled schizophrenia has dramatic costs for patients (about 10% of whom commit suicide), and for society in terms of increased social service use and crime.

 Although lower-cost alternatives to atypicals do exist, they are problematic. The first generation of schizophrenia drugs resulted in improvement of symptoms but with potentially debilitating side effects.

 Atypical antipsychotics were introduced in the 1990s and have largely replaced first-generation treatments because atypicals control schizophrenia better and have fewer negative side effects. For example, about 30% to 40% of patients relapse with first-generation drugs (relative to 80% without treatment), but with second-generation drugs, relapse rates fall to about 25%.

Formulary restrictions appear to alter doctors’ prescribing behavior. Previous studies have examined the impact of utilization management programs and found that formulary restrictions reduced the use of atypicals by about 5%.

 After prior authorization of atypicals became a requirement in Maine, the risk of discontinuation rose by 29%.

 Studies in Georgia and New Hampshire found that prior authorization reduced the use of targeted drugs and increased the use of other health services.

 Previous studies of the impact of restrictions on treatment have focused on individual states. This study, to the best of our knowledge, is the first nationally representative study of formulary restrictions’ impact on treatment. In this study, we examine how formulary restrictions of atypicals affect the likelihood that a patient who has been diagnosed with schizophrenia, and who has had some marker of an unsuccessful response to a particular drug therapy, is continued on that same drug therapy, changes the type of drug, or discontinues all treatment.

Policy information on the restrictions placed on atypical antipsychotics from 1999 to 2005 was collected from 24 state Medicaid programs via a mailed survey and personal communications with policy directors. States reported whether prior authorization, quantity limits, step therapy (fail-first), and other policies applied to a list of drugs identified by US brand name. To supplement this information, additional formulary information was collected using publicly available documents. Regulation data were collected at the individual drug level. The survey included 10 atypicals. All major atypical antipsychotic agents are addressed in the survey, except for paliperidone (brand: Invega) and olanzapine/fluoxetine combined therapy (brand: Symbax), both of which were not yet commercially available during the study period. States reported if each drug was placed under prior authorization, quantity limits, step therapy, dose restriction, or duplicate therapy restrictions in each year.

Medicaid claims for 24 states (12 states that adopted formulary restrictions and 12 states that did not) for a 5-year period (2001-2005) were used to measure medical and pharmacy use at the individual level. We selected the states shown in 

 based on the size of the Medicaid population; the fraction of enrollees in fee-for-service Medicaid (with complete claims); and whether the state Medicaid program had imposed restrictions on selected classes of medications over the study period. The years included in this study (2001-2005) captured the widespread introduction and use of prior authorization requirements in many state Medicaid programs, particularly for atypical antipsychotics.

The sample consisted of Medicaid enrollees 18 years and older who have schizophrenia in each of the study states. These patients were identified by (1) the presence of an 

International Classification of Diseases, 9th Revision, Clinical Modification

) code beginning with 295 in any diagnosis position for inpatient, outpatient, or long-term care claims, and (2) a confirmatory schizophrenia diagnosis at least 30 days after the initial diagnosis. Thus, we eliminated individuals who were diagnosed only once without a subsequent diagnosis. This eliminated individuals with a socalled “rule-out” diagnosis: a physician may make an initial diagnosis of suspected schizophrenia that is later ruled out. The requirement of a subsequent “confirmatory” diagnosis excluded “ruled out” cases from the sample. This criterion also imposed a 30-day minimum window of observation for inclusion in the study. In addition, to be included, a patient must have filled at least 1 prescription for one of the drugs included in our sample. In practice, this restriction eliminated very few patients because those with a subsequent confirmatory diagnosis are almost always given a prescription.

presents a consort diagram for the sample construction: 944,609 patients over the age of 18 years who have had at least 1 schizophrenia diagnosis. Of these, about 77% had a confirmatory diagnosis after 30 days. Of the remaining population, 63%, or 460,991 patients, were prescribed an atypical.

To estimate the impact of prior authorization on treatment failure and “replication of failure,” a difference-in-differences method (DD) was used. In the DD method, 2 groups—patients in states with restrictions and patients in states without restrictions&mdash; were observed in 2 different time periods: 2001 and 2005. Patients living in states that adopted formulary restrictions after 2001 were, in the terminology of DD models, exposed to “treatment” in 2005, while those in nonformulary states were not exposed in either period. This method first calculates the change in the average likelihood of the outcome variable in the treated group, and then calculates the corresponding change in the nontreated group. The excess change in the treated group, above and beyond the change in the nontreated group, represents the estimated DD effect. For example, in states with formulary restrictions, we calculated the change in the likelihood of a patient resuming the same medication after a 30-day gap before and after the implementation of the restriction. We then compared this with the corresponding difference in likelihood within states that never adopted restrictions. The approach removes the effects of trends due, for example, to general changes in atypical use between 2001 and 2005. Because formulary restrictions were measured at the state level, we clustered the standard errors at the state level. We utilized an alpha of 0.10, although the tables included alternative alpha values.

To examine the replication of failure, all patients currently with schizophrenia who are taking an atypical antipsychotic for their condition were identified. Then, patients were classified as failing on a particular therapy if they either were admitted to a hospital for their disease or had a medication gap greater than 30 days (in a baseline year). Patients who failed treatment were followed over time, and their claims were examined for resumption of treatment with the same drug after at least a 30-day gap. As a robustness check, we examined longer gaps, but faced serious sample size limitations due to the limited number of people observed restarting after longer gaps. Finally, we also examined whether individuals on an atypical discontinued drug therapy altogether, switched from an atypical to a typical (first generation) antipsychotic, or switced from one atypical to another.

We included an indicator variable equal to 1 in the year 2005 and an indicator for states that adopt formulary restrictions. The interaction between the 2 indicators, year 2005 and formulary restrictions, is the difference-in-differences estimate discussed above (labeled DD in the tables).

 shows summary statistics measuring atypical antipsychotic usage along with emergency department (ED) and inpatient admissions related to mental health causes, stratified by the number of atypicals under a formulary restriction in a given state-year. The observed population with schizophrenia increased slightly over time, from just fewer than 500,000 in 2001 to about 541,000 in 2005. The number of patients with schizophrenia receiving atypical antipsychotics increased from nearly 195,000 in 2001 to over 320,000 in 2004, and then fell to about 273,000 in 2005.

Each patient, on average, experienced 0.22 inpatient admissions related to mental health issues and 0.06 ED visits per year. Approximately 2.8% of the sample was admitted to the ED and 7.9% was admitted on an inpatient basis for issues related to mental health.

 summarizes the replication of failure by year, and the restrictiveness of Medicaid formulary in that year. The variable of interest is the number of times a patient goes off a particular medication for 30 days and then resumes that same medication. For example, a value of 1 indicates that the average patient had one 30-day gap in their medication. Patients in states with formulary restrictions tend to have slightly more replication of failure, measured as stoppages and restarts. For example, in 2005, individuals in states with all atypicals under restriction experienced an average of 1.04 stoppages and subsequent restarts of the same atypical in the same year, while those in states with no restrictions experienced an average of 0.99 stoppages and restarts per year.

Measuring replication of failure using inpatient or ED admission provides a more explicit measure of failure. Repeated failure here signifies that the patient was admitted for a mental health issue while on a particular drug, but then was put back on that atypical within 6 months of the admission. Approximately 65% of inpatient admissions result in the individual resuming the same atypical they were on before the admission, within 6 months. There is no significant difference over time. Among ED visits, 59% result in the same atypicals being resumed within 6 months in both restrictive and unrestrictive states.

) show the estimates from difference-in-differences regressions comparing outcomes from 2001 with those from 2005. The first outcome measured in Table 4 is for patients who resumed the same atypicals after experiencing at least a 30-day gap in stock of the drug. The sample is restricted to those patients who filled at least 1 atypical prescription in the given year. States that place some drugs on prior authorization are compared with states with no restrictions (Table 4A). The difference-in-differences coefficient is calculated to be 0.062 (standard error [SE] 0.009): formulary restrictions on some atypicals are associated with a 6.2% increase in the probability that an individual resumes the same atypical after a stoppage, relative to the control group. In the regressions that use the states that have placed all atypicals under a formulary restriction by 2005 as the treatment group (Table 4B), the difference-in-differences coefficient is found to be 0.201 (SE 0.011). This shows a substantial effect of placing all atypicals onto a formulary restriction, as the probability of resuming the same atypical after a stoppage of at least 30 days increased by 20.1% relative to the control group states.

The next outcomes measured were resumptions of the same atypical antipsychotic after an inpatient and ED admission with a mental health diagnosis. The sample was further restricted to patients meeting our inclusion criteria who were also admitted as an inpatient, or who visited the ED in the given year with a mental health condition. When states with no formulary restrictions are compared with those with only some atypicals restricted, the results suggest no statistical impact of these restrictions on resuming the same atypical after an admission. In states that restricted all atypicals, the policy was not shown to have any effect on resuming the same atypical after an ED visit.

Table 5 (A and B) shows difference-in-differences results for the outcome of switching from one atypical to another. The control group includes patients in states with no formulary restrictions, while the 2 treatment groups consist of patients in (1) states that placed some atypicals on a formulary restriction, or (2) states that placed all atypicals on formulary restrictions. Individuals in restrictive states who were admitted as inpatients experienced a reduction of about 4.3% in the number of switches between atypicals relative to those in nonrestrictive states. In this case, a triggering event reduces the likelihood that a patient switches from one atypical to another. Similar effects are observed for ED admissions; the reduction in number of switches following an ED admission is 3.2%.

The results in Table 6 (A and B) examine the relationship between the introduction of formulary restrictions on atypicals and patients stopping the use of atypicals in favor of typical antipsychotics that are not on formulary restrictions. For states that placed all atypicals onto formulary restrictions, the number of patients stopping atypical usage in favor of typical antipsychotics is shown to have increased by less than 3.5%.

Finally, Table 7 shows estimates for when atypical users stopped the use of all mental health drugs. Among those in states that restrict all atypicals, the likelihood of discontinuation increased by 11.6%.

Although formulary restrictions hold the promise of reduced Medicaid outlays for pharmaceuticals, they may also have unintended effects that raise Medicaid costs in other areas. In this study, we examined the impact of formulary restrictions on the treatment of schizophrenia. Schizophrenia is very responsive to psychopharmacologic treatment, but the treatment response is highly heterogeneous across patients, and some patients respond only to a single medication. Moreover, the process of finding the appropriate medication is often one of trial and error. Because of this, treatment failures would be expected to cause a switch to a different medication.

We examined several different treatment failures, such as a patient going off medication, a patient being admitted to the hospital or visiting the ED with a psychological episode, or a patient terminating drug treatment altogether. These rates of replication of failure are relatively large, with 65% of inpatient admissions and 59% of ED admissions resulting in a resumption of the same atypical, regardless of the formulary policies in place. Such rates suggest possible quality-of-care issues, but a deep examination of quality of care lies beyond the scope of this study.

We find that replication of failure increases when states adopt formulary restrictions. These findings are also suggestive of a causal effect. We found that formulary restrictions on atypical antipsychotics increase the likelihood that a patient who has stopped taking medication will end up back on the same drug—an indication that the drug failed, but was restarted. This likelihood increases by 6.2% if the patient resides in a state that has implemented formulary restrictions on some atypicals in its Medicaid program, and by 20.1% if the state of residence’s formulary restrictions cover all atypicals. This suggests that formulary restrictions cause doctors to return patients to treatments for which there is evidence that the treatment is not effective.

Moreover, we found that the likelihood that a patient who was initially on an atypical switches to an older and less expensive medication (a typical/first generation) increases by 3.5% if the state’s formulary restricts all atypicals, again relative to the control group of states without restrictions. This again suggests that state formulary restrictions are pushing doctors to alter their treatment choices in ways that they did not think benefited the patient before the formulary restriction. Finally, we found that when states impose formulary restrictions on atypicals, the likelihood that a patient on atypicals discontinues all treatment with either typicals or atypicals increases by 11.6% if the state restricts all atypicals under its formulary restrictions. These findings indicate that state formulary restrictions have an impact on the likelihood that doctors will try an alternative therapy when markers of failure present themselves, and that doctors are induced by formulary restrictions to switch to alternative therapies even when there are no indicators that the current treatment is ineffective.

 University of Southern California, Leonard D. Schaeffer Center, Los Angeles, CA (DPG); Johnson & Johnson, West Chester, PA (RD); Janssen Pharmaceuticals LLC, Titusville, NJ (JF); University of Southern California, Los Angeles, CA (RMC).

This research was funded by Janssen Scientific Affairs, LLC, Titusville, NJ.

Dr Goldman is a partner of Precision Health Economics. Dr Conrad was an employee of Precision Health Economics at the time of study completion. Mr Fastenau is an employee of Janssen Scientific Affairs, LLC, which funded this study, and a Johnson & Johnson stockholder. Dr Dirani was an employee of Janssen Scientific Affairs, LLC at the time of study completion.

 Concept and design (DPG, RD, JF, RMC); analysis and interpretation of data (DPG, RD, JF, RMC); drafting of the manuscript (DPG, RD, JF, RMC); critical revision of the manuscript for important intellectual content (JF); statistical analysis (RD, RMC); supervision (DPG).

Dana P. Goldman, PhD, 3335 South Figueroa St, Unit A, Los Angeles, CA 90089-7273. E-mail: dpgoldma@ healthpolicy.usc.edu.1. Consumer Health Sourcebook. National health expenditures by type of service and source of funds: calendar years 1960 to 2009. http://www .chsourcebook.com/docs/nhe_aggregate_2009.pdf. Published 2009.

2. Soumerai SB, Zhang F, Ross-Degnan D, et al. Use of atypical antipsychotic drugs for schizophrenia in Maine Medicaid following a policy change. Health Aff (Millwood). 2008;27(3):w185-w195.

3. Mello MM, Studdert DM, Brennan TA. The pharmaceutical industry versus Medicaid: limits on state initiatives to control prescription-drug costs. N Engl J Med. 2004;350(6):608-613.

Articles

This study measures the impact of Medicaid formulary restrictions on the rate at which patients who failed on a drug therapy for schizophrenia return to the same therapy.

Do Strict Formularies Replicate Failure for Patients With Schizophrenia?

To compare Medicaid spending among patients with type 2 diabetes mellitus (T2DM) receiving exenatide or other add-on therapies.

Medicaid data in patients with T2DM were compared among those who initiated exenatide, thiazolidinediones, dipeptidyl peptidase-4 inhibitors, or basal insulin. Patients were on a regimen of metformin and/or sulfonylurea for 30 days and continued the next-line therapy for at least 90 days.

Total inpatient, outpatient, prescription, and total annual health expenditures were compared for 1 year following treatment initiation. Regression analyses were conducted to compare spending; analyses controlled for patient characteristics, year of initiation, and prior therapy. Propensity score matching was used to match patients receiving exenatide with those receiving other therapies, and analyses were repeated with matched cohorts.

 Of 23,966 eligible patients, 1345 initiated exenatide and 22,621 initiated other therapies. In the unmatched analysis, medical spending was significantly lower in those given exenatide compared with those given other therapies for inpatient ($1945 vs $3893), prescription ($4505 vs $5714), and total costs ($11,830 vs $15,459) (

 <.01 for all); outpatient spending was not significantly different ($5380 vs $5853, 

 = .15). In the matched analysis (n = 1345 for exenatide, n = 1345 non-exenatide), patients receiving exenatide had significantly lower spending in all 4 categories: inpatient ($1945 vs $4242), outpatient ($5380 vs $6826), prescription ($4505 vs $5878), and total ($11,830 vs $16,945) (

 Patients with T2DM receiving exenatide had lower annual Medicaid claims costs compared with patients receiving other therapies.

In 2010, more than 34 million adults were enrolled in Medicaid; it has been estimated that 11.3% of American adults have diabetes.

 In 2007, total annual US direct medical spending for diabetes was estimated to exceed $116 billion.

 Annualized, this translates to adjusted mean medical expenditures of $11,744 for patients with diabetes, compared with $5095 for those without diabetes.

 The Congressional Budget Office projects adult Medicaid enrollment to exceed 43 million people in 2014, and more than 55 million people in 2020.

Clinical guidelines recommend that patients with newly diagnosed type 2 diabetes mellitus (T2DM) receive initial treatment with metformin unless contraindicated. The addition of other therapies is recommended if target glycated hemoglobin (A1C) levels are not reached. As diabetes is a progressive disease, most patients eventually require combination therapy to maintain A1C goals, although guidelines do not stipulate a clear therapeutic preference vis-à-vis glycemic control.

 While comparative effectiveness research studies have addressed clinical outcomes more broadly, it is not fully understood which therapeutic combinations are associated with lower total costs.

The availability of antidiabetic medications for patients on Medicaid varies from state to state, and formulary restrictions often limit use of drugs. Many drugs have limitations for use in various settings due to restricted Medicaid formulary coverage. The use of pioglitazone, insulin analogues, and newer antidiabetic medications recently approved by the US Food and Drug Administration (FDA) (glucagon-like-peptide-1 [GLP-1] receptor agonists and dipeptidyl peptidase-4-inhibitors [DPP4Is]) is limited to varying degrees based on state and local policies. In the case of exenatide, a GLP-1 receptor agonist, Medicaid boards in several states approved its use soon after FDA approval. For example, in 2005, exenatide was placed on the preferred drug list in West Virginia and Iowa.

 However, Medicaid boards in other states have delayed the inclusion of exenatide in their formularies.

 In contrast, glucose-lowering agents that have been on the market longer, such as metformin and sulfonylureas (SUs), are widely available and covered by Medicaid preferred drug lists.

 Antidiabetic therapies differ in their costs, with 30-day average wholesale prices in this study population ranging from $236 for DPP4Is to $513 for exenatide (2011 dollars).

To address the value of adding exenatide to the treatment regimen of patients on existing therapy consisting of metformin and/or SUs, this study compared Medicaid-enrolled patients with T2DM who had exenatide added to their therapeutic regimen with those who had thiazolidinediones (TZDs), basal insulin, or DPP4Is added. Four spending outcomes—inpatient, outpatient, prescription, and total&mdash;were measured for the year following initiation of each of these treatments. Regression analyses were conducted to control for patient characteristics. All comparisons with patients given exenatide were performed with the entire study cohort and again with propensity score—matched groups.

This retrospective cohort analysis applied Medicaid claims data from January 1, 2005, to December 31, 2009, from the Thomson Reuters MarketScan Multi-State Medicaid Database. This database contains inpatient, outpatient, and pharmacy claims of approximately 880,000 Medicaid enrollees with multiple claims with diabetes diagnoses. Also included in this database are demographic information, diagnosis codes, costs, dates of occurrence, and other administrative information. Clinical information, however, such as body mass index, blood pressure measurements, and laboratory results, such as A1C values, are not available. This database contained Medicaid claims information from 12 geographically dispersed states; however, state-level information or identifiers for geographic location were not available.

Patients between 18 and 65 years of age were eligible for the study if they had T2DM, confirmed by 2 diagnoses of T2DM (

International Classification of Diseases, 9th Revision, Clinical Modification codes 250.x0 or 250.x2

) at least 30 days apart, or at least 1 T2DM diagnosis and a filled prescription for a 30-day supply of antidiabetic medication. These requirements helped to ensure the stability and accuracy of the diabetes diagnosis. Patients were required to have received an SU and/or metformin for at least 30 days, followed by the addition of exenatide, a TZD, basal insulin (long-acting or intermediate-acting, including detemir, glargine, and neutral protamine Hagedorn), or a DPP4I for at least 90 days within 12 months of initiating these add-on therapies. Patients who received treatment with more than 1 of the 4 specified add-on therapies during the study period were excluded. Those who used rapid-acting insulin prior to, or in combination with, the addition of a new therapy were excluded.

The index date was defined as the first date of add-on therapy initiation. For study inclusion, patient data had to be available for 12 consecutive months prior to the index date (ie, pre-period) and 12 consecutive months after the index date (ie, follow-up period). The pre-period served as a washout to ensure that the study captured new T2DM treatment initiation. Data obtained during the pre-period also enabled identification of comorbidities and calculation of the Charlson Comorbidity Index (CCI). The 12-month follow-up period allowed for an adequate observation window for outcomes.

Spending outcomes were evaluated for the 365 days subsequent to add-on treatment initiation. Outcomes were inpatient, outpatient, and pharmacy claims, and total spending (consisting of these 3 cost components). Costs were converted to 2011 US dollars based on the medical care component of the Consumer Price Index.

Two potential confounders that could affect study outcomes were identified prospectively. First, physicians make treatment decisions based on a variety of factors, some of which cannot be fully observed in claims data (eg, comorbid disease severity, treatment adherence likelihood, and formulary restrictions).

 In addition, patient preferences and other factors can influence treatment initiation. These unobserved factors could confound the effect of therapy on measured outcomes.

To address measured potential confounders, propensity score matching was used to match patients who were given exenatide with patients using 1 of the other 3 therapies (TZD, basal insulin, or DPP4I). Matching was based on patient age, gender, race, pre-period CCI, comorbidities (including congestive heart failure [CHF], hyperlipidemia, hypertension, stroke, chronic obstructive pulmonary disease [COPD], and end-stage renal disease [ESRD]), year of treatment initiation, and prior therapy (ie, an SU and/or metformin used in the 90 days prior to initiation). Predicted probabilities from a multivariate logistic regression analysis of exenatide use were used to identify and match the “nearest neighbor” who received exenatide to each patient who did not.

Baseline characteristics, including demographics (age, race, sex), CCI, comorbidities (from CCI), prior diabetes therapy, and year of treatment initiation, were compared among patients who received exenatide and those who did not.

Mean medical spending in patients receiving exenatide was compared with that in patients receiving other add-on therapies. Additionally, results in patients not receiving exenatide were analyzed by individual treatment groups (ie, TZD, basal insulin, and DPP4I). Analyses were first completed using all patients who did not receive exenatide, then limited to the propensity score—matched sample.

Regression-adjusted models controlled for patient characteristics including age, sex, race, comorbidities (CHF, hypertension, stroke, COPD, hyperlipidemia, and ESRD), CCI, year of drug initiation, and prior therapy. With a large presented as regression-adjusted mean outcomes.

 Analyses were conducted using Stata version 11.1 software; results with a 

A total of 23,966 patients met the criteria defining the population of interest. Of these, 1345 had exenatide added as a next-line treatment. Patients who did not receive exenatide (n = 22,621) were predominantly initiated on TZDs (n = 18,208), followed by basal insulin (n = 3823) or DPP4Is (n = 590). 

 shows how the analysis sample was determined as the inclusion and exclusion criteria were applied.

Once matched, data were included for 1345 patients who received exenatide, and 1345 patients who received a TZD (n = 1061), basal insulin (n = 224), or DPP4I (n = 60), for a total sample size of 2690. Matched results including baseline characteristics, unadjusted outcomes, and adjusted mean outcomes were consistent with the unmatched results. Only unmatched results are presented below (matched results are presented in the 

Baseline characteristics of the study populations are shown in 

. Results in patients not receiving exenatide are also reported by individual therapy (eg, TZD, basal insulin, and DPP4I). Compared with patients who initiated another therapy (n = 22,621), those who initiated exenatide therapy (n = 1345) were a mean of 3.6 years younger (51.5 vs 47.9 years, respectively; 

 <.01). The exenatide group had a higher proportion of female patients compared with the non-exenatide group (79% vs 66%, respectively; 

 <.01). There were also differences in the racial composition of the 2 groups; 68% of patients receiving exenatide were white, compared with 51% of patients receiving other therapy (

The exenatide group had a mean CCI score of 2.0, while the non-exenatide group had a mean score of 1.5 (

 <.01). Generally, patients receiving exenatide were significantly more likely than those receiving other treatments to have comorbid conditions: hypertension (49% vs 45%), hyperlipidemia (33% vs 24%), and COPD (23% vs 17%) (

 <.01 for all). However, compared with patients receiving other therapies, fewer patients receiving exenatide had ESRD (0.9% vs 0.3%, respectively, 

The baseline analysis also examined prescribing patterns of patients’ physicians. A significantly lower percentage of patients receiving exenatide were given an SU and/or metformin compared with those not receiving exenatide (SU, 58% vs 68%, respectively; metformin, 58% vs 63%; SU and metformin, 31% vs 39%; 

 <.05 for all). Of the 22,621 patients in the non-exenatide group, 58% visited physicians who had prescribed exenatide for any of their Medicaid patients (42% visited physicians who had never prescribed exenatide).

, patients receiving exenatide averaged $11,830 (standard deviation [SD] = $15,986) in total annual healthcare expenditures, with a median of $7082 (data not shown). In comparison, patients receiving other add-on therapies averaged $15,459 (SD = $26,012) (

 <.01). Compared with the non-exenatide group, the exenatide group had lower inpatient costs ($3893 vs $1945, 

 <.01) and prescription costs ($5714 vs $4505, 

 <.01); outpatient costs were not significantly different ($5853 vs $5380). Focusing on specific therapies, the TZD and basal insulin groups had higher expenditures than the exenatide group. For example, total annual healthcare spending averaged $14,779 among those receiving a TZD compared with $11,830 for those receiving exenatide (

Differences in utilization were more muted. Mean total days supplied within a year of initiation of therapy were not significantly different in the exenatide, DPP4I, and basal insulin groups (233.2 days, 235.9 days, and 229.1 days, respectively). Mean TZD use was higher (249.0 days, 

The results reported here correspond to the full sample; results for the propensity-score matched sample were similar, and are reported in the Appendix.

Regression-adjusted mean medical spending in all patients who met the inclusion criteria (n = 23,966) is shown by physician prescribing status for exenatide in 

. After adjusting for covariates, inpatient claims for patients of physicians who prescribed exenatide were $560 less per year than those for patients of physicians who prescribed other therapies (

 = .05) (Figure 2A). Compared with patients whose physicians prescribed other add-on therapies, those with physicians that prescribed exenatide had lower outpatient spending ($158 less) and lower total spending ($561 less), although these differences were not significant (Figures 2B and 2D). However, adjusted prescription medication spending for patients of physicians who prescribed exenatide was significantly greater ($157 more [2.7% higher]) than that for patients of physicians who prescribed other therapies (

Compared with patients receiving other add-on therapies, Medicaid programs paid lower claims for patients receiving exenatide across all spending categories. After adjusting for covariates, inpatient spending for patients given exenatide was $1939 (

 <.01) less than that for patients given other therapies. Similarly, adjusted outpatient spending was $478 less per year with exenatide (

= .08); this difference represented an 8.2% reduction in outpatient costs compared with other therapies. Adjusted prescription medication spending and total spending were also significantly lower with exenatide compared with other add-on therapies ($397 less, 

 shows spending based on adjusted data and according to the specific add-on therapy that patients received. Spending outcomes for patients given exenatide were less than those for patients using other therapies. Regression analysis showed that outpatient spending in patients receiving exenatide was $390 less (

 = .16) than for patients receiving a TZD, $660 less (

 = .04) than for patients receiving basal insulin, and $1056 less (

 = .045) than for patients receiving a DPP4I (Figure 3B). Prescription medication spending with exenatide was $347 less than for a TZD (

 = .04), $499 less than for basal insulin (

 <.01) (Figure 3C), and $733 less than for a DPP4I (

 <.01). Total spending in patients receiving exenatide was $2303 less than for those receiving a TZD (

 <.01), $5317 less than for those receiving basal insulin (

 <.01), and $2540 less than for those receiving a DPP4I (

Drug expenditures were significantly higher, while inpatient spending was significantly lower, for Medicaid patients with T2DM whose physicians prescribed exenatide at least once during 2005 to 2009. Compared with other add-on therapies, patients given exenatide had significantly lower inpatient, outpatient, prescription medication, and total spending.

Together, these results suggest that patients on Medicaid who are cared for by physicians with the flexibility to prescribe exenatide had lower healthcare spending than those who are cared for by physicians who never prescribed exenatide. These findings add to some recent evidence that exenatide is associated with lower healthcare costs than other T2DM therapies (these earlier studies also found that exenatide was not associated with poor clinical outcomes).

There are several limitations to this study. One limitation was that the data did not contain any state identifiers. Since Medicaid policies vary from state to state,

 similar patients who live in different states will have varying levels of access to antidiabetic medications. This difference could lead to biased results if left uncontrolled. To address this concern, each physician in the data set was identified as to whether or not they have prescribed exenatide to patients. Each patient was categorized as having a physician who prescribed or never prescribed exenatide.

Because insulin is generally initiated later in the progression of T2DM compared with other therapies, patients receiving insulin may have poorer health outcomes and other comorbidities, and thus higher medical costs, compared with those receiving other treatments.

 Certain unobservable factors may account for some of the differences in outcomes observed between exenatide and insulin. Therefore, to better understand the outcomes in patients initiating various treatments, an analysis was performed that stratified patients by specific treatment received. In all categories of medical spending, costs were lower in those given exenatide compared with those given insulin.

Likewise, any study approach that directly compares outcomes between insulin and non-insulin therapies may be biased. This is because patient and physician resistance to insulin initiation is well documented, and can be predicated on concerns about weight gain, hypoglycemia, patient adherence, or regimen complexity.

 For these reasons, insulin may not be prescribed until other treatments fail, resulting in a group of patients that have T2DM of longer duration. To mitigate these unobserved differences, propensity scores were used in this study to create a matched sample. Propensity score matching on characteristics likely to be correlated with these unobservable differences, such as the presence of comorbidities, can help reduce potential biases.

 Finally, it is important to note that these models may not necessarily measure how the specific type of medication produces differences in outcomes.

Another study limitation was that the DPP4I group had the fewest patients from the overall sample (n = 590), and once the propensity score—matched groups were established, this small size was amplified (n = 60). Given the small sample size, results in those given DPP4Is are more likely to have been skewed by outliers (eg, a few patients with very high medical spending due to medication or procedures) and may be less reliable. Results in this group should be validated in larger studies if future investigators are interested in tracking the costs associated with these specific drugs.

The current data suggest that patients given exenatide had reduced costs compared with those given a TZD, basal insulin, or a DPP4I. Additionally, these results indicated that patients who are cared for by exenatide prescribers experienced lower spending outcomes. Medicaid policies that provide exenatide coverage may result in fewer inpatient, outpatient, and prescription claims, and lower total medical spending. These results also suggest that patients who visit physicians who use a wide array of diabetes treatments may be more likely to be given the treatment that is best suited to maintain their health.

Comparative effectiveness research continues to build interest around the efficacy, safety, and value of existing and new T2DM therapies. As further evidence is generated, the role of these medications in clinical guidelines and algorithms for diabetes management is being reassessed. The long-term clinical efficacy and safety profile of exenatide has been established in a series of successful clinical trials.

 The current analysis indicates that there is also potential economic value in positioning exenatide as a preferred second-line therapy. 

The authors thank Amy Blickensderfer for input on study design.

Amylin Pharmaceuticals, Inc, San Diego, California (JHB); Precision Health Economics, LLC, Los Angeles, California (RMC); Price School of Public Policy and Leonard D. Schaeffer Center for Health Policy and Economics, University of Southern California, Los Angeles, CA (DPG); Division of Endocrinology, Keck School of Medicine and Clinical Diabetes Program, University of Southern California, Los Angeles, CA (ALP); Price School of Public Policy, University of Southern California, Los Angeles, CA (JAR). 

 This supplement was supported by Amylin Pharmaceuticals, Inc.

Dr Best reports employment and stock ownership with Amylin Pharmaceuticals, Inc. Dr Conrad reports consultancy with Amylin Pharmaceuticals, Inc. Dr Goldman reports consultancy with Amylin Pharmaceuticals, Inc. Dr Peters reports consultancy with Amylin Pharmaceuticals, Inc, Boehringer Ingelheim, CellNovo, Lilly, Novo Nordisk, and sanofi-aventis, and honoraria from Dexcom, Medtronic MiniMed, Merck, Medscape, and Perrigo. She also reports membership in speakers’ bureaus for Amylin Pharmaceuticals, Inc, Takeda, Lilly, and Novo Nordisk. Dr Romley reports consultancy with Amylin Pharmaceuticals, Inc.

Concept and design (JHB, RMC, DPG, ALP, JAR); acquisition of data (RMC); analysis and interpretation of data (JHB, RMC, DPG, ALP, JAR); drafting of the manuscript (RMC, DPG, JAR); ALP); statistical analysis (RMC); obtaining funding (JHB); supervision (JHB, DPG, ALP, JAR).

 Jennie H. Best, PhD, Amylin Pharmaceuticals, Inc, 9360 Towne Centre Dr, San Diego, CA 92121. E-mail: 

. Spending and Enrollment Detail for CBO’s March 2011 Baseline: Medicaid. 

www.cbo.gov/sites/default/files/cbofiles/attachments/Medicaid_0.pdf

. Published 2011. Accessed July 27, 2012.

Centers for Disease Control and Prevention

. National diabetes fact sheet: national estimates and general information on diabetes and prediabetes in the United States. 

www.cdc.gov/diabetes/pubs/pdf/ndfs_2011.pdf

. Published 2011. Accessed September 20, 2011.

. Consumer Price Index Detailed Report Tables: Annual Average Index, 2011. www.bls.gov/cpi/cpid11av.pdf. Published 2011. Accessed July 27, 2012.

Handelsman Y, Mechanick JI, Blonde L, et al

. American Association of Clinical Endocrinologists Medical Guidelines for Clinical Practice for developing a diabetes mellitus comprehensive care plan. 

. Application of comparative effectiveness research in evaluation of treatments for type 2 diabetes mellitus. 

. Comparative Effectiveness and Safety of Oral Diabetes Medications for Adults With Type 2 Diabetes. Rockville, MD: 2007. 

www.effectivehealthcare.ahrq.gov/repFiles/OralFullReport.pdf

www.dhhr.wv.gov/bms/Pharmacy/dur/Documents/Preferred%20Drug%20List/pdl_44.pdf

. Published 2005. Accessed September 20, 2011.

. Approved PDL for Iowa Medicaid Program: Iowa Medicaid Enterprise: Iowa Medicaid PDL. 

www.iowamedicaidpdl.com/uploads/1H/fA/1HfAqUmB7I401LBkOi9TIQ/report_1010_pdl_a1_pdl_drug_list_vers_x9e_desi_cilostazol.pdf

. Published 2005. Revised November 16, 2005. Accessed September 20, 2011.

Nevada Department of Health and Human Services

. Division of Health Care Financing and Policy. Nevada Medicaid Preferred Drug List (Effective July 1, 2010). 

www.medicaid.nv.gov/Downloads/provider/NV_PDL_20100701.pdf

Colorado Department of Health Care Policy and Financing

. Preferred Drug List (PDL): Effective October 1, 2010. 

www.colorado.gov/cs/Satellite?blobcol=urldata&blobheader=application%2Fpdf&blobheadername1=Content-Disposition&blobheadername2=MDT-Type&blobheadervalue1=inline%3B+filename%3D578%2F142%2FMicrosoft+Word+-+PDL+effective+0--200_2_doc+-111.pdf&blobheadervalue2=abinary%3B+charset%3DUTF-8&blobkey=id&blobtable=MungoBlobs&blobwhere=1251655552502&ssbinary=true

. Published 2010. Accessed September 20, 2011.

. Preferred Drug List/Prior Authorization List. www.lamedicaid.com/provweb1/forms/rxpa/PDL_NPDL_10_1.pdf. Published 2011. Accessed September 20, 2011.

. Wisconsin Medicaid Preferred Drug List. 

https://www.forwardhealth.wi.gov/WIPortal/Tab/42/icscontent/provider/medicaid/pharmacy/pdl/pdfs/pdl_list.pdf.spage

Ross JS, Jackevicius C, Krumholz HM, et al

. State Medicaid programs did not make use of prior authorization to promote safer prescribing after rosiglitazone warning. 

Ryan M. Conrad, PhD

Articles

Do Strict Formularies Replicate Failure for Patients With Schizophrenia?

Expenditures for Medicaid Patients Treated With Exenatide Compared With Other Diabetes Management Regimens