Cost effectiveness of paliperidone palmitate versus oral antipsychotics in patients with schizophrenia and a history of criminal justice involvement.

Journal of Medical Economics

Journal of Medical Economics Downloaded from informahealthcare.com by Yale Dermatologic Surgery on 05/14/15 For personal use only.

1369-6998 doi:10.3111/13696998.2015.1037307

2015, 1–9

Article 0023.R1/1037307 All rights reserved: reproduction in whole or part not permitted

Original research Cost effectiveness of paliperidone palmitate versus oral antipsychotics in patients with schizophrenia and a history of criminal justice involvement

Erik Muser

Abstract

Janssen Scientific Affairs LLC, Titusville, NJ, USA

Chris M. Kozma CK Consulting Associates LLC, Saint Helena Island, SC, USA

Carmela J. Benson Janssen Scientific Affairs LLC, Titusville, NJ, USA

Lian Mao Janssen Research & Development LLC, Titusville, NJ, USA

H. Lynn Starr Larry Alphs John Fastenau Janssen Scientific Affairs LLC, Titusville, NJ, USA Address for correspondence: Erik Muser PharmD MPH, Janssen Scientific Affairs LLC, 1125 Trenton-Harbourton Road, Titusville, NJ 08560, USA. Tel: +1 636-828-4999; [email protected] Keywords: Cost effectiveness – Criminal justice system – Oral antipsychotics – Paliperidone palmitate – Schizophrenia Accepted: 31 March 2015; published online: 6 May 2015 Citation: J Med Econ 2015; 1–9

! 2015 Informa UK Ltd www.informahealthcare.com/jme

Objective: Conduct a cost effectiveness analysis for the Paliperidone palmitate Research In Demonstrating Effectiveness (PRIDE) trial. Research design and methods: PRIDE was a 15 month, prospective, randomized, open-label study in which once monthly paliperidone palmitate significantly delayed the time to first treatment failure (healthcare or criminal justice system [HC/ CJS] events) versus oral antipsychotics in recently incarcerated adults with schizophrenia. The present analysis used a state government perspective and HC/CJS event data that were collected on a resource use questionnaire (RUQ) every 3 months. Main outcome measures: Since cost information was not collected in the trial, cost estimates from published literature and an analysis of multistate Medicaid data for CJS and HC events, respectively, were applied to RUQ event data. Effectiveness and costs were adjusted to 456 days (trial duration). Incremental cost effectiveness was calculated as the adjusted cost difference divided by the adjusted effectiveness difference. Results: Adjusted costs (in US dollars) in the paliperidone palmitate group (n ¼ 198) and the oral antipsychotic group (n ¼ 193), respectively, were: non-drug costs $22,331 and $25,027; drug costs $18,592 and $7833; and total costs $40,923 and $32,860. Adjusted effectiveness differences and corresponding incremental cost effectiveness per event avoided (in parentheses) for paliperidone palmitate versus oral antipsychotics were as follows: 0.33 fewer CJS events ($24,409), 0.13 fewer psychiatric hospitalizations ($60,484), 0.46 fewer psychiatric hospitalizations or CJS events combined ($17,391), and 0.30 fewer incarcerations ($26,754). Conclusions: Costs for HC/CJS events avoided offset 25% of the greater drug cost for the paliperidone palmitate versus oral antipsychotic treatment group in this vulnerable population. Use of a recall-dependent RUQ for event rates and cost estimates instead of actual costs are potential limitations and may make the results conservative from a state government perspective. Indirect costs are likely to be substantial for this population, but were not considered in the analysis.

Cost effectiveness of paliperidone palmitate Muser et al.

1


2015


Introduction Despite a large number of efficacy and effectiveness trials for antipsychotic therapy, the understanding of societal and payer willingness to pay to prevent psychiatric events is limited. A 2006 article reviewed eight clinical trials based on cost effectiveness comparisons for immediate-release second-generation oral antipsychotics versus conventional antipsychotics1. Abstracts for six of the studies reported that the second-generation agent was more cost effective or had equal costs and greater effectiveness, and the other two studies did not demonstrate cost effectiveness for the second-generation agents. Cost effectiveness claims were limited to restricted populations in three of these studies. The designs of the individual studies may have influenced their ability to demonstrate significant cost effectiveness and the findings of the cost effectiveness review were inconclusive. A decision-analytic model of long-acting risperidone injection, oral risperidone, olanzapine, quetiapine, ziprasidone, aripiprazole, and depot haloperidol found that use of long-acting risperidone injection was associated with direct annual medical cost savings ranging from $161 to $8224 versus the other agents2. A decision-analytic model of paliperidone palmitate versus atypical oral antipsychotics showed paliperidone palmitate had lower total medical costs and greater effectiveness (i.e., dominant decision)3. Management of schizophrenia patients can be complicated due to multiple factors, such as unstable living conditions, contact with the criminal justice system, lack of care continuity, and community support after discharge from institutions (hospitals, jails, or prisons). Patients with serious mental illness have a high risk of incarceration and the total cost of care for incarcerated patients with serious mental illness is significantly greater than for non-incarcerated patients with serious mental illness4. Psychiatric relapse and recidivism are both common and costly among patients with schizophrenia after they are released from incarceration5. Insurance eligibility is important in patients with serious mental illnesses such as schizophrenia, and has policy implications beyond healthcare. One study found Medicaid eligibility is a significant predictor of criminal justice system events and costs in schizophrenic adults. After release from incarceration, delays in gaining Medicaid eligibility were associated with a greater hazard for re-incarceration, and greater cumulative duration of Medicaid eligibility was associated with a lower hazard of re-incarceration6. Another study examined arrestees with serious mental illnesses and reported that Medicaid enrollees had higher expenditures overall but a substantially lower proportion of these expenditures were for criminal justice system events compared with those not enrolled in 2


Medicaid (34% vs 59%). Thus, expenditures for criminal justice system events were lower in Medicaid enrollees7. Patients with serious mental illnesses who receive antipsychotic treatment after hospitalization or incarceration have a reduced risk of arrest and thus have lower total cumulative costs4,5. To our knowledge, there are no direct comparisons (i.e., head-to-head trials) that address the cost effectiveness of long acting atypical agents versus oral antipsychotics in vulnerable populations, such as persons with criminal justice system contact(s). The odds of a seriously mentally ill person being in jail or prison compared to a hospital across all 50 states is 3.1 to 1, which implies that persons with serious mental illness are at least three times more likely to be in jail than in a hospital8. To encompass the substantial costs associated with these non-healthcare events, an analysis of the cost effectiveness of an antipsychotic therapy would need to consider not only the cost of avoided healthcare but also the cost of avoided criminal justice system events. The Paliperidone palmitate Research In Demonstrating Effectiveness (PRIDE) trial was a comparative effectiveness study of paliperidone palmitate versus oral antipsychotics (aripiprazole, haloperidol, olanzapine, paliperidone, perphenazine, quetiapine, and risperidone) in patients with schizophrenia and a recent history of incarceration9. The primary results of PRIDE found that paliperidone palmitate significantly delayed time to first treatment failure (including both healthcare events and criminal justice system events) versus oral antipsychotics10. The objective of the present study was to conduct a cost effectiveness analysis of once monthly paliperidone palmitate versus once daily oral antipsychotic therapy using the effectiveness data from PRIDE. This cost effectiveness analysis was planned prior to the start of PRIDE with a hypothesis that paliperidone palmitate would have greater effectiveness compared with oral antipsychotics, but greater drug acquisition costs. Greater effectiveness was demonstrated for paliperidone palmitate treated patients in the trial10, further supporting the need for a cost effectiveness analysis (CEA).

Patients and methods Study design A detailed summary of the PRIDE study design has been previously published9. Briefly, PRIDE was a 15 month, prospective, randomized, active-controlled, open-label, flexible-dose study of paliperidone palmitate compared with oral antipsychotic treatment in delaying time to treatment failure (see definition in Effectiveness data) in adults with schizophrenia who had prior contact with the criminal justice system. Study participants were men and women 18 to 65 years of age who had a current www.informahealthcare.com/jme ! 2015 Informa UK Ltd



diagnosis of schizophrenia. Patients were required to have been placed into custody at least twice within the 24 months before study start, with at least one of them leading to incarceration, and were released from the most recent custody within 90 days before the start of the screening period. All patients provided written informed consent to participate in PRIDE and the study was conducted in accordance with the principles in the Declaration of Helsinki. Paliperidone palmitate was administered by intramuscular injection, starting at a dose of 234 mg on Day 1, followed by 156 mg on Day 8, and then flexible dosing between 78 mg and 234 mg once monthly thereafter. In the oral antipsychotic group, the investigator identified possible treatment options from a list of seven medications (aripiprazole, haloperidol, olanzapine, paliperidone, perphenazine, quetiapine, and risperidone). After identifying acceptable potential oral treatments, the patient was randomly assigned to take one of the medications identified as acceptable by the investigator. Antipsychotic monotherapy was encouraged and the dose of the study drug could be adjusted as needed. Concomitant nonantipsychotic therapy (prescription or non-prescription) could be continued or initiated at the discretion of the investigator. Because the hypothesis was that paliperidone palmitate would have greater effectiveness than oral antipsychotics, patients were allowed to switch from paliperidone palmitate to an oral antipsychotic, but they were not allowed to switch from an oral antipsychotic to paliperidone palmitate. Patients who discontinued their assigned study treatment were encouraged to continue in the study for 15 months in order to permit evaluation of events that occurred after early treatment discontinuation. The analysis described herein was not a primary or secondary objective of the study. This was an exploratory analysis that combined effectiveness data from a resource use questionnaire (RUQ) used in PRIDE with cost estimates for criminal justice events from the literature and administrative claims data for health resource use events.

Effectiveness data Patients in PRIDE were followed for 15 months with monthly visits. Data collection for health resource use and criminal justice system events used the RUQ at the following times: at scheduled visits approximately every 3 months (i.e., months 3, 6, 9, 12, and 15); at any unscheduled visit (e.g., for safety, tolerability, or treatment failure); and at early study termination. Specific dates were recorded for hospitalizations, emergency room visits, and contacts with the criminal justice system. Treatment failure events, as defined in the clinical trial, included criminal justice system events (arrest or incarceration), psychiatric hospitalization, suicide, discontinuation of ! 2015 Informa UK Ltd www.informahealthcare.com/jme

2015

antipsychotic treatment due to inadequate efficacy, treatment supplementation with another antipsychotic due to inadequate efficacy, discontinuation of antipsychotic treatment due to safety or tolerability, and increase in the level of psychiatric services to prevent imminent psychiatric hospitalization. Effectiveness variables that were included in this analysis were the number of psychiatric hospitalization events, the number of criminal justice system events, the number of incarceration events, the number of incarcerations with a duration 42 days, and the number of patients with at least one incarceration event. The primary measure for this exploratory analysis was the average number of events per patient for psychiatric hospitalizations and criminal justice system events, combined.

Cost estimates PRIDE was conducted prospectively but did not include collection of event costs; therefore, cost estimates were assigned to event rates collected during the trial. All costs were provided in US dollars and a US state government perspective was used for assignment of costs. Event costs were obtained from a review of criminal justice system costs reported in the published literature and analysis of claims from a large multistate Medicaid database to estimate health resource use event costs11. Median values from that analysis were used to estimate costs in the current analysis. Costs identified through the literature review are provided in Supplementary Table 1. Cost estimates obtained from administrative claims included procedurebased healthcare costs and hospitalization costs from the Truven Multistate Medicaid data for patients with schizophrenia diagnoses (Supplementary Table 2). A cost analysis that used these values and the baseline PRIDE data to estimate health resource and criminal justice system costs12 showed no significant difference in total costs for 1 year prior to trial enrollment (p ¼ 0.1819; independent two sample t-test) between the paliperidone palmitate group (n ¼ 218; mean $31,314; SD $21,347; 95% confidence interval [CI] $28,464 to $34,163) and the oral antipsychotic group (n ¼ 213; mean $28,664; SD $19,733; 95% CI $25,999 to $31,330). For the cost effectiveness analysis described herein, the cost estimates were applied to the observed rates post-baseline for any criminal justice system event, probation/parole/jail, hospitalization, emergency room visits, outpatient events, and emergency medical services. Study drug costs were based on start and stop dates reported for each study drug in PRIDE, combined with daily cost estimates from paid claims data in a nationwide administrative claims database not associated with PRIDE (Truven 2011 Multistate Medicaid Database). Cost effectiveness of paliperidone palmitate Muser et al.

3



2015

Statistical analysis

Results

This cost effectiveness analysis included effectiveness data for patients who received at least one dose of study treatment (the intent-to-treat population) and who had at least one response recorded in the RUQ database, both at baseline and at least once post-baseline. Statistical comparisons between the paliperidone palmitate and oral antipsychotic groups were conducted for baseline values by using independent two-sample t-tests for continuous variables and chi-square tests for categorical variables. Statistical comparisons between the paliperidone palmitate and oral antipsychotic groups for unadjusted costs and number of events were conducted by using Mann– Whitney U (Wilcoxon rank-sum) tests. Event rates were compared by using chi-square tests. No adjustment was made for multiplicity. Statistical analyses were conducted with SAS 9.4 for Windows (Cary, NC, USA). Events were recorded on the RUQ at each 3 month visit. In addition, sites were asked to contact patients (and their designee), and to search online county jail records in an effort to obtain criminal justice system event data for patients lost to follow-up. Thus, to adjust for different observation durations for the different variables, unadjusted effectiveness and cost data were prorated to 456 days (i.e., the planned 15 months of study followup). The mean value for each variable was divided by the mean duration of follow-up for that variable, and that result was then multiplied by 456 days. Incremental cost effectiveness was calculated for events avoided as follows13, where the costs and effectiveness were based on the adjusted values prorated to 456 days:

Patient characteristics

Deltacost Deltaeffect ¼

Costpaliperidone palmitate group Costoral antipsychotic group Effectivenesspaliperidone palmitate group Effectivenessoral antipsychotic group

A composite test across costs was not conducted because the duration of the data collection for the various cost components could vary over time; PRIDE was a pragmatic study and patients continued to be followed for observation even if treatment failure occurred. Therefore, mean costs were estimated per component (i.e., criminal justice system contacts, hospitalization, emergency transportation, emergency room visits, outpatient health resource utilization, parole/probation/jail diversion or mental health court participation and drugs) and ramped up to 456 days independently. A two-way sensitivity analysis was conducted on the incremental cost effectiveness per psychiatric hospitalization or criminal justice system event (combined) by varying total cost by 20% and effectiveness by 20%. 4


A total of 450 patients (230 in the paliperidone palmitate group and 220 in the oral antipsychotic group) completed the RUQ at baseline. A total of 59 patients (28 in the paliperidone palmitate group, 27 in the oral antipsychotic group, and 4 that were not randomized to either group) were excluded from the study for the following reasons: no post-baseline records for 40 patients; no contact with the criminal justice system recorded at baseline for 8 patients; issues with criminal justice system dates where self-evident corrections were not possible for 6 patients; and no study medication received (i.e., not part of the intent-to-treat population) for 5 patients. A total of 391 patients (198 in the paliperidone palmitate group and 193 in the oral antipsychotic group) were included in the cost effectiveness analysis; characteristics of these patients are summarized in Table 1. In the group randomized to oral antipsychotic treatment, 21.8% received paliperidone, 16.1% received risperidone, 15.5% received olanzapine, 15.0% received aripiprazole, 14.5% received quetiapine, 9.3% received perphenazine, and 7.8% received haloperidol. No statistically significant differences were observed between the groups at baseline as follows (paliperidone palmitate and oral antipsychotics, respectively): mean age (37.4 [SD 10.8] and 38.8 [SD 10.5] years); sex (85.4% and 87.0% male); race (64.1% and 61.2% black or African American; 32.3% and 32.6% white; 3.5% and 6.2% other race); and medical insurance eligibility (38.4% and 45.6% none; 35.4% and 29.5% Medicaid; 13.6% and 15.5% Medicare; 10.6% and 8.8% other insurance).

Costs Costs are provided in Table 2. Unadjusted component costs based on event data from the study were greatest for criminal justice system events and study drug, followed by outpatient care. Component event costs were not significantly different between the paliperidone palmitate group and the oral antipsychotic group for criminal justice system events, probation, parole, jail diversion program, mental health court, hospitalization, emergency room, EMS, and outpatient care; however, study drug costs were significantly greater in the paliperidone palmitate group than in the oral antipsychotic group. Average days of observation for each cost component are provided in Table 2. Duration of observation was not significantly different between the treatment groups for any component event. Criminal justice system events had slightly longer observation times than the other component events in both treatment groups. www.informahealthcare.com/jme ! 2015 Informa UK Ltd


Costs adjusted to 456 days are also provided in Table 2. Total adjusted non-drug costs were $2696 lower in the paliperidone palmitate group compared with the oral antipsychotic group ($22,331 and $25,027, respectively). Total adjusted study drug costs were $10,759 greater in the paliperidone palmitate group compared with the oral antipsychotic group ($18,592 and $7833, respectively). Thus, the total 456 day estimated costs with study drug

2015

were $8063 greater in the paliperidone palmitate group compared with the oral antipsychotic group ($40,923 and $32,860, respectively).

Effectiveness Unadjusted and adjusted effectiveness measures are provided in Table 3. There were fewer criminal justice


Table 1. Patient characteristics. Variable

Paliperidone Palmitate (n ¼ 198)

Oral Antipsychotic (n ¼ 193)

p-value for unadjusted cost*

37.4 (10.8)

38.8 (10.5)

0.182

169 (85.4) 29 (14.6)

168 (87.0) 25 (13.0)

0.628

127 (64.1) 64 (32.3) 7 (3.5)

118 (61.2) 63 (32.6) 12 (6.2)

0.451

70 (35.4) 27 (13.6) 5 (2.5) 3 (1.5) 76 (38.4) 21 (10.6)

57 (29.5) 30 (15.5) 5 (2.6) 2 (1.0) 88 (45.6) 17 (8.8)

0.194 0.622 0.980z 1.000z 0.173 0.527

Age in years, mean (SD) Sex, n (%) Male Female Race, n (%) Black or African American White Other Medical insurance after jail, n (%)y Medicaid Medicare Commercial Veteran’s or Military Medical Benefits None Other

*Independent two-sample t-test for continuous variables and chi-square (within category) for categorical variables. ySubjects could report more than one type of insurance (i.e., subjects can be in more than one group). zFisher’s Exact Test for cell sizes 55.

Table 2. Component costs (in US dollars). Unadjusted cost

Paliperidone Palmitate (n ¼ 198) Mean (SD)

Criminal justice system Probation, parole, jail diversion program, or mental health court Hospitalization Emergency room Outpatient EMS Study drug Adjusted to 456 days Criminal justice system Probation, parole, jail diversion program, or mental health court Hospitalization Emergency room Outpatient EMS Total non-drug Study drug Total

$9602 ($18,875) $1811 ($3718) $1828 ($4721) $168 ($352) $3835 ($6571) $51 ($159) $12,911 ($7349)


p-value*

Obs Days

Mean (SD)

Obs Days

397 307

$11,970 ($19,708) $2007 ($4402)

408 320

0.039 0.910

321 321 307 321 317

$2586 ($7131) $218 ($360) $3377 ($6457) $54 ($217) $5542 ($6687)

329 328 320 327 323

0.364 0.067 0.910 0.553 50.001

Averagey

Averagey

Difference

$11,024 $2693

$13,394 $2858

$2370 $165

$2597 $239 $5702 $75 $22,331 $18,592 $40,923

$3585 $304 $4808 $77 $25,027 $7833 $32,860

$989 $65 $894 $2 $2696 $10,759 $8063

EMS ¼ emergency medical services; Obs Days ¼ mean number of observation days. *p-values are based on Mann–Whitney U (Wilcoxon rank-sum) tests. yCalculated as (mean cost 7 observation days) 456 days.



5


2015

Table 3. Effectiveness. Unadjusted events


No. of events, mean (SD) Psychiatric hospitalizations Criminal justice system events Incarcerations Incarcerations 42 days Patients incarcerated, n (%) Events adjusted to 456 days No. of events, average Psychiatric hospitalizations Criminal justice system events Psychiatric hospitalizations or criminal justice system events Incarcerations Incarcerations 42 days Patients incarcerated, %

Paliperidone Palmitate (n ¼ 198)


Mean (SD) or n (%)

Obs Days

Mean (SD) or n (%)

Obs Days

0.20 (0.59) 0.69 (1.10) 0.52 (0.90) 0.44 (0.84) 69 (34.8)

321 397 397 397 397

0.32 (1.06) 1.01 (1.42) 0.80 (1.29) 0.56 (0.90) 87 (45.1)

329 408 408 408 408

p-value*

0.433 0.011 0.017 0.122 0.039

Averagey

Averagey

Difference

0.23 0.79 1.02

0.36 1.12 1.48

0.13 0.33 0.46

0.59 0.50 40.4

0.89 0.63 50.8

0.30 0.13 10.4

Obs days ¼ observation days. *p-values are based on Mann–Whitney U (Wilcoxon rank-sum) tests for continuous variables and chi-square for categorical variables. yCalculated as (mean number of events or event rate 7 mean observation days) 456 days.

Table 4. Incremental cost effectiveness ratio (in US dollars per event avoided) in the paliperidone palmitate group compared with the oral antipsychotic group. Event Avoided Cost per psychiatric hospitalization avoided Cost per criminal justice system event avoided Cost per psychiatric hospitalization or criminal justice system event avoided Cost per incarceration avoided Cost per incarceration 42 days avoided Cost per patient that avoided any incarceration

Ratio* $60,484 $24,409 $17,391 $26,754 $63,258 $77,731

*Cost ($8063 from Table 2) divided by the effectiveness difference (from Table 3). Cost and effectiveness differences are presented as rounded values; ratios are based on the actual cost difference and effectiveness differences, prior to rounding.

system and incarceration events in the paliperidone palmitate group compared with the oral antipsychotic group and a smaller percentage of patients had any incarceration event in the paliperidone group compared with the oral antipsychotic group. Differences between the treatment groups were statistically significant for the mean number of criminal justice system events, the mean number of incarcerations, and the number of patients with any incarceration event. The adjusted differences over 456 days (15 months) were 0.13 fewer psychiatric hospitalizations, 0.33 fewer criminal justice system events, 0.46 fewer psychiatric hospitalizations or criminal justice system events combined, 0.30 fewer incarcerations, 0.13 fewer incarcerations with a duration 42 days, and 10.4% fewer 6


patients with at least one incarceration event in the paliperidone palmitate group compared with the oral antipsychotic group.

Incremental cost effectiveness ratios The adjusted total cost difference ($8063; Table 2) was divided by the adjusted difference in events (Table 3) to determine the incremental cost effectiveness ratio per event avoided (Table 4). Cost effectiveness ratios ranged from $17,391 per psychiatric hospitalization or criminal justice system event avoided to $77,731 per patient that avoided any incarceration for the paliperidone palmitate group compared with the oral antipsychotic group. In the sensitivity analysis of the incremental cost effectiveness ratio per psychiatric hospitalization or criminal justice system event avoided (Figure 1), results ranged from $11,594 (20% lower costs and 20% greater efficacy) to $26,086 (20% greater costs and 20% lower efficacy) for the paliperidone palmitate group compared with the oral antipsychotic group.

Discussion This analysis included 391 recently incarcerated adults with schizophrenia who received paliperidone palmitate or an oral antipsychotic in a randomized, controlled trial with up to 15 months of follow-up. Based on data collected from the RUQ, patients in the paliperidone palmitate group had fewer healthcare or criminal justice events, www.informahealthcare.com/jme ! 2015 Informa UK Ltd


Estimated Cost per Psychiatric Hospitalization or Criminal Justice System Event Avoided


20% Lower Cost

Estimated Cost

2015

20% Greater Cost

$30,000 $26,086 $25,000 $20,869

$21,738 $20,000 $17,391

$15,000

X

$17,391

$17,391 $14,492

$13,912 $10,000

Estimated cost per event avoided from the primary analysis

$11,594

$5,000

$0 −20% −15% −10% −5% 0% +5% +10% +15% +20% (0.37) (0.39) (0.42) (0.44) (0.46) (0.49) (0.51) (0.53) (0.56) Difference from Predicted Incremental Efficacy (0.46 was the difference in events per person in the primary analysis)

Figure 1. Sensitivity analysis of incremental cost effectiveness ratio (in US dollars) in the paliperidone palmitate group and the oral antipsychotic group.

leading to numerically lower (but not statistically significant) estimated costs compared with patients in the oral antipsychotic group for both types of events. Study drug costs were significantly higher for paliperidone palmitate than for the oral antipsychotics. The lower non-drug costs for paliperidone palmitate treated patients offset approximately 25% of the increased drug costs ($2696/$10,759 per patient); however, total costs (with study drug) remained approximately $8063 greater per patient over the 15 month period (i.e., $6450 greater annually) in the paliperidone palmitate group compared with the oral antipsychotic group. The primary trial results of PRIDE found that paliperidone palmitate significantly delayed time to first treatment failure versus oral antipsychotics (median, 416 days versus 226 days; hazard ratio, 1.43 [95% CI, 1.09 to 1.88]; p ¼ 0.011) among patients with schizophrenia and a history of criminal justice system involvement10. Those findings provided further support for conducting this cost effectiveness analysis, which was planned before the study began. In this analysis, the cost effectiveness ratio per psychiatric hospitalization or criminal justice system event avoided was $17,391, and the results for this endpoint in the sensitivity analysis ranged from $11,594 to $26,086. There are limited data on willingness to pay to prevent events associated with serious mental illness. One study on how much a community might be willing to pay to prevent incarceration suggested that the lower end of the range was $23,000 for a burglary-related incarceration, with amounts increasing as the seriousness of the offense increased14. In the present cost effectiveness analysis, the ! 2015 Informa UK Ltd www.informahealthcare.com/jme

incremental cost effectiveness ratio per criminal justice system event avoided (i.e., direct non-healthcare costs) was $24,409 for a 15 month period, which is consistent with what a community might be willing to pay. PRIDE enrolled patients who normally would be excluded from clinical trials, such as those with history of incarcerations or substance abuse, or those who were already taking an antipsychotic at study entry. Thus, it included more pragmatic (i.e., real world) evaluations than are seen with traditional clinical trials. PRIDE also examined endpoints, such as arrest/incarceration, that reflect real-world treatment failure considerations for schizophrenia treatments. Limitations of this analysis should be noted. Given the wide range in enrollment duration, results were adjusted to 456 days based on mean incarceration and medical resource utilization values, as opposed to individual values, to avoid making a constant event rate assumption. The adjusted values represent the best estimate. Summing across components was not feasible without additional assumptions because the data collection period could vary by incarceration or medical resource utilization component. Therefore, inferential statistics were not evaluated for the adjusted values and no statement about the statistical significance of differences could be made. This is a limitation imposed by the nature of the data and the method by which the data were collected. Event data came from RUQ data that relied on patient recall at baseline and every 3 months during the trial. Costs were assigned to these events based on information collected about contacts with the criminal justice and healthcare Cost effectiveness of paliperidone palmitate Muser et al.

7



2015

system in the RUQ, and average event costs were applied since actual costs for criminal justice system interactions and healthcare utilization were not collected in the trial. Hospitalization costs were based on the reported number of hospitalizations; no adjustment was made for length of stay. Drug costs for oral antipsychotic therapy were based on paid claims from Truven Multi-State Medicaid by drug and strength multiplied by the number of pills dispensed. This could have led to a lower cost difference for the paliperidone palmitate group relative to the oral antipsychotic group because Truven Multi-State Medicaid paid claims do not include supplemental rebates. Also, many of the oral antipsychotics are available generically and thus no rebates would be provided for these drugs. Including these rebates would change the cost difference between the paliperidone palmitate and oral antipsychotic treatment groups. Cost effectiveness ratios were calculated by combining mean values, and thus were calculated without regard for statistical significance. Small changes in incremental effectiveness led to large changes in cost effectiveness ratios. While somewhat arbitrary, the 20% used for the two-way sensitivity analysis was thought to represent a reasonable range for the maximum variation from the results observed in the trial. While indirect costs related to productivity or quality of life were not examined in this analysis, they remain important public health considerations. The loss of functioning that often occurs during a relapse can impact productivity and quality of life, and is likely to improve gradually as relationships and job skills are re-established. Thus, their impact likely extends beyond the 15 month duration of this trial. Costs of non-study drugs were not included. Costs associated with the study effectiveness endpoints were all lower in the paliperidone palmitate group but did not show statistical significance; if there was no difference in effectiveness then the oral antipsychotic group would be the more cost effective option. Since statistical significance can be driven by sample size, the assumption in this analysis was that the means represented the best estimate of the population value.

Conclusions The paliperidone palmitate treatment group had greater total drug costs than the oral antipsychotic treatment group, but lower healthcare and criminal justice system event costs in the paliperidone palmitate group partially offset this difference. In assessing product value for antipsychotic treatment, payers and government officials should consider not only product cost but also the potential cost savings of avoiding events. This analysis highlights the importance of considering direct nonhealthcare costs such as criminal justice system costs in 8


addition to direct healthcare costs in cost effectiveness analyses for antipsychotic treatment in persons with schizophrenia.

Transparency Declaration of funding This work was supported by Janssen Scientific Affairs LLC. Declaration of financial/other relationships E.M., C.J.B., L.M., H.L.S., L.A., and J.F. have disclosed that they are employees of Janssen Scientific Affairs LLC or Janssen Research & Development, which are members of the Johnson & Johnson family of companies, and they hold stock in Johnson & Johnson. C.M.K. has disclosed that he is an employee of C.K. Consulting Associates LLC, which received funding from Janssen Scientific Affairs LLC to conduct the research and prepare the publication. JME peer reviewers on this manuscript have no relevant financial or other relationships to disclose. Acknowledgments Jonathan Latham of PharmaScribe LLC provided medical writing assistance with funding from C.K. Consulting Associates LLC. Previous presentation: Portions of this work were presented at the 27th Annual US Psychiatric and Mental Health Congress, 20–23 September 2014, Orlando, FL, USA.

References 1. Polsky D, Doshi JA, Bauer MS, et al. Clinical trial-based cost-effectiveness analyses of antipsychotic use. Am J Psychiatry 2006;163:2047-56 2. Edwards NC, Locklear JC, Rupnow MF, et al. Cost effectiveness of long-acting risperidone injection versus alternative antipsychotic agents in patients with schizophrenia in the USA. Pharmacoeconomics 2005;23(Suppl 1):75-89 3. Edwards NC, Muser E, Doshi D, et al. The threshold rate of oral atypical antipsychotic adherence at which paliperidone palmitate is cost saving. J Med Econ 2012;15:623-34 4. Van Dorn RA, Desmarais SL, Petrila J, et al. Effects of outpatient treatment on risk of arrest of adults with serious mental illness and associated costs. Psychiatr Serv 2013;64:856-62 5. Lin I, Muser E, Munsell M, et al. Economic impact of psychiatric relapse and recidivism among adults with schizophrenia recently released from incarceration: a Markov model analysis. J Med Econ 2015;18:219-29 6. Benson C, Dickson WM, Kozma CM, et al. Relationship between Medicaid eligibility and time to re-incarceration of previously incarcerated schizophrenia diagnosed subjects [abstract]. Value Health 2014;17:A151 7. Petrila J, Andel R, Constantine R, et al. Public expenditures related to the criminal justice system and to services for arrestees with a serious mental illness. Psychiatr Serv 2010;61:516-19 8. Torrey EF, Kennard AD, Eslinger D, et al. Treatment Advocacy Center. More Mentally Ill Persons Are in Jails and Prisons Than Hospitals: A Survey of the States. 2010. Available at: http://www.treatmentadvocacycenter.org/storage/ documents/final_jails_v_hospitals_study.pdf [Last accessed 8 April 2015] 9. Alphs L, Mao L, Rodriguez SC, et al. Design and rationale of the Paliperidone Palmitate Research in Demonstrating Effectiveness (PRIDE) study: a novel comparative trial of once-monthly paliperidone palmitate versus daily oral antipsychotic treatment for delaying time to treatment failure in persons with schizophrenia. J Clin Psychiatry 2014;75:1388-93

www.informahealthcare.com/jme ! 2015 Informa UK Ltd


12. Kozma CM, Muser E, Benson C. Baseline health resource and criminal justice system costs for clinical trial subjects with schizophrenia and prior incarceration. Value Health 2014;17:A218 13. Briggs A, Claxton K, Sculpher M. Decision Modelling for Health Economic Evaluation (Handbooks in Health Economic Evaluation). Oxford: Oxford University Press, 2006 14. Cohen MA, Rust RT, Steen S. National Institute of Justice (US). Measuring public perceptions of appropriate prison sentences: excutive summary. Washington, DC, 2003. Available at: https://www.ncjrs.gov/pdffiles1/nij/ grants/199364.pdf [Last accessed 8 April 2015]


10. Alphs L, Benson C, Cheshire-Kinney K, et al. Real-world outcomes of paliperidone palmitate compared to daily oral antipsychotic therapy in schizophrenia: a randomized, open-label, review boardblinded 15-month study. J Clin Psychiatry. 2015. Prepublished on April 14, 2015 as doi:10.4088/ JCP.14m09584 11. Muser E, Benson C, Kozma C, et al. Estimating costs related to health resource utilization and contacts with the criminal justice system for patients diagnosed with schizophrenia [abstract and poster 308]. In: 26th Annual US Psychiatric and Mental Health Congress. Las Vegas, NV, USA, 2013

2015



9

Impact of paliperidone palmitate versus oral atypical antipsychotics on healthcare outcomes in schizophrenia patients.

Intramuscular long-acting paliperidone palmitate in acute patients with schizophrenia unsuccessfully treated with oral antipsychotics.

Comparative effectiveness of injectable paliperidone palmitate versus oral atypical antipsychotics: early postmarketing evidence.

Cost-Effectiveness of Long-Acting Injectable Paliperidone Palmitate Versus Haloperidol Decanoate in Maintenance Treatment of Schizophrenia.

Comparing Injection Site Pain with Paliperidone Palmitate Versus First-Generation Depot Antipsychotics in Subjects with Schizophrenia.

Costs and Resource Utilization Among Medicaid Patients with Schizophrenia Treated with Paliperidone Palmitate or Oral Atypical Antipsychotics.

Cost Effectiveness of Paliperidone Long-Acting Injectable Versus Other Antipsychotics for the Maintenance Treatment of Schizophrenia in France.

Paliperidone palmitate use in pregnancy in a woman with schizophrenia.

Healthcare utilization and costs of Veterans Health Administration patients with schizophrenia treated with paliperidone palmitate long-acting injection or oral atypical antipsychotics.

Manic Symptoms during a Switch from Paliperidone ER to Paliperidone Palmitate in a Patient with Schizophrenia.

Pharmacoeconomic analysis of paliperidone palmitate versus olanzapine pamoate for chronic schizophrenia in Norway.

Depot Typical Antipsychotics versus Oral Atypical Antipsychotics in Relapse Rate Among Patients with Schizophrenia: A Five -Year Historical Cohort Study.

Once-monthly paliperidone palmitate in recently diagnosed and chronic non-acute patients with schizophrenia.

Paliperidone palmitate for refractory and clozapine-resistant schizophrenia.

Prolactin Levels After Switching to Paliperidone Palmitate in Patients with Schizophrenia.

Cost-effectiveness of aripiprazole once-monthly compared with paliperidone palmitate once-monthly injectable for the treatment of schizophrenia in the United States.

Paliperidone palmitate in non-acute patients with schizophrenia previously unsuccessfully treated with risperidone long-acting therapy or frequently used conventional depot antipsychotics.

A prospective flexible-dose study of paliperidone palmitate in nonacute but symptomatic patients with schizophrenia previously unsuccessfully treated with oral antipsychotic agents.

Randomized, placebo-controlled, double-blind study assessing the efficacy and safety of paliperidone palmitate in Asian patients with schizophrenia.

Adherence, healthcare resource utilization and Medicaid spending associated with once-monthly paliperidone palmitate versus oral atypical antipsychotic treatment among adults recently diagnosed with schizophrenia.

The effect of long-acting paliperidone palmitate once-monthly on negative and depressive symptoms in patients with schizophrenia switched from previous unsuccessful treatment with oral aripiprazole.

Effectiveness of long-acting antipsychotics in clinical practice : 1. A retrospective, 18-month follow up and comparison between paliperidone palmitate, risperidone long-acting injection and zuclopenthixol decanoate.

Switching from oral atypical antipsychotic monotherapy to paliperidone palmitate once-monthly in non-acute patients with schizophrenia: A prospective, open-label, interventional study.

Efficacy, safety, and impact on hospitalizations of paliperidone palmitate in recent-onset schizophrenia.