Cost-Effectiveness of Atypical Antipsychotics: Implications for Practice Robert Rosenheck MD New England MIRECC Yale Medical School April 21, 2011 VA Cooperative Study #451: Olanzapine vs. haloperidol Prospective 12 month double-blind trial N=309 veterans 17 VA medical centers Assesed multiple outcomes Symptoms Quality of Life Cognitive functioning Health service use VA and Societal Costs Figure 2: Compliance with Double Blind Study Drug (log rank test p=.25) Cummulative % Remaining in Study 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0 50 100 150 Days 200 250 300 Figure 3. PANSS Total Symptom Score: Intention-to-treat PANSS Total Score 100 75 50 25 0 0 1.5 3 Mixed model analysis: ns 6 9 12 Months Olanzapine Haloperidol Olz. N: 159 121 108 93 90 92 Hal. N: 150 115 105 94 83 85 VACS #451 Results(1) Patients assigned to olanzapine had: slightly less akathisia and superior cognitive functioning (small effect <0.2 -not enough to affect quality of life); Olanzapine showed no advantage in: Symptoms Quality of life Parkinsonian (EPS) side effects Olanzapine incurred greater weight gain and greater VA costs ($3,000 - $9,000/pt/year). Cost Utility of the Latest Antipsychotic Drugs in Schizophrenia Study (CUtLASS)(S Lewis et al. 2005) RCT conducted in the UK for the NHS Health Tecnhnology Assessment Program Recruited: Aug 1999 – April 2002 Randomly assigned patients to “doctor’s choice” of any SGA other than clozpaine or FGA (n=227) Blinded assessments. 59% stayed on original drug for 1 year 81% were followed-up at 1 year Primary outcome was Quality of Life Scale (Heinrichs/Carpenter) CUtLASS Results:No significant differences on… % compl. on original study drug (FGA<SGA) Quality of Life (QOLS) (FGA>SGA) Symptoms (PANSS) (FGA<SGA) Global Assessment (GAF) (SGA>FGA) Depression (Calgary) (FGA < SGA) Compliance (FGA=SGA) EPS (Simpson Angus) (FGA>SGA) Akathisia (Barnes) (FGA>SGA) TD (AIMS) (FGA>SGA) Health costs (even without drugs)(FGA<SGA) Estimate of difference in QLS after imputation of missing data, with 95% c.i. of difference Hypothesis of 5 point advantage for SGA excluded Observed -6 -5 -4 -3 Favors FGA -2 Expected -1 0 1 2 3 Equivalence 4 5 6 Favors SGA Difference in QLS scores Data from Table 3b CATIE Schizophrenia Trial Design Phase 2 Phase 1* Double-blind, random treatment assignment. Participants who discontinue Phase 1 choose either the clozapine or the ziprasidone randomization pathways OLANZAPINE CLOZAPINE (open-label) QUETIAPINE 1460 patients with SCZ Comorbidity Other meds R R OLANZAPINE, QUETIAPINE or RISPERIDONE RISPERIDONE Phase 3 Participants who discontinue Phase 2 choose one of the following open-label treatments •ARIPIPRAZOLE •CLOZAPINE •FLUPHENAZINE DECANOATE •OLANZAPINE •PERPHENAZINE ZIPRASIDONE •QUETIAPINE ZIPRASIDONE R OLANZAPINE, QUETIAPINE or RISPERIDONE •RISPERIDONE •ZIPRASIDONE PERPHENAZINE No one assigned to same drug as in Phase 1 •2 of the antipsychotics above *Phase 1A: participants with TD do not get randomized to perphenazine; phase 1B: participants who fail perphenazine will be randomized to an atypical (olanzapine, quetiapine, or risperidone) before they are eligible for phase 2. Stroup TS et al. Schizophr Bull. 2003;29:15-31. Proportion of Patients without Event Time to Discontinuation for Any Reason 1 Overall p-value = 0.004* 0.8 0.6 0.4 0.2 0 P<0.001 for olanzapine vs quetiapine P=0.002 for olanzapine vs risperidone 0 3 6 9 12 15 Time to Discontinuation for Any Cause (mo) Olanzapine Perphenazine Quetiapine Risperidone 18 Ziprasidone OLZ (n=330) QUET (n=329) RISP (n=333) PER (n=257) ZPR (n=183) Discontinued 210 (64%) 269 (82%) 245 (74%) 192 (75%) 145 (79%) Kaplan-Meier Median (mos) [95%CI] 9.2 [6.9, 12.1] 4.6 [3.9, 5.5] 4.8 [4.0, 6.1] 5.6 [4.5, 6.3] 3.5 [3.1, 5.4] Hazard ratios for Olanzapine --- 0.63 < 0.001* 0.78 0.021 0.76 0.028 0.75 0.002* No advantage of any SGA over perphenazine on… Symptoms (Rosenheck et al., 2006) Neurocognition (Keefe et al., 2007) Violent behvaior (Swanson et al., 2008) Employment (Resnick et al., 2008) Community functioning (Swartz et al., 2006) Family Burden (Perlick et al, in press) EPS/TD/Akithesia (Miller et al., 2008) CATIE Cost-Effectiveness Results (1) COST Service Use and Cost Measures Service Use (Service Use and Resources Form [SURF]) Outpatient Inpatient Mental health Medical Mental Health Substance Abuse Medical Cost Nursing home Residential Medication records Criminal justice, public support, productivity Outpatient All residential Inpatient All health care Experimental medications 2003 Medicaid Discount rates and mandated company rebates VA discount (40%) Sensitivity analysis of price discounts (as funded vs Medicaid vs. VA). Ancillary medication (discounted cost to privately insured Market Scan ® patients). Monthly Costs: All medication costs (experimental drugs and concomitant medications)(actual prices, with discounts)(mean=$499/month)(ITT) $700 $500 $200/mo $2,400/yr $400 $300 $200 $100 OLANZAPINE QUETIAPINE ZIPRASIDONE PERPHENAZINE RISPERIDONE m o. 18 m o. 17 m o. 16 m o. 15 m o. 14 m o. 13 m o. 12 m o. 11 m o. 10 m o. 9 m o. 8 m o. 7 m o. 6 m o. 5 m o. 4 m o. 3 m o. 2 m o. 1 se lin e $0 Ba Monthly expenditures $600 Monthly service use: All inpatient days (mental health and medical/surgical)(mean=0.70/month)(ITT)(p=ns) Monthly inpatient days 3 2 1 Olanzapine Quetiapine Ziprasidone Perphenazine Risperidone m o. 18 m o. 17 m o. 16 m o. 15 m o. 14 m o. 13 m o. 12 m o. 11 m o. 10 m o. 9 m o. 8 m o. 7 m o. 6 m o. 5 m o. 4 m o. 3 m o. 2 m o. 1 Ba se lin e 0 Monthly Costs: All inpatient (mental health and medical/surgical)costs (mean=$363/month)(ITT)(p=ns) $1,500 $1,000 $500 Olanzapine Quetiapine Ziprasidone Perphenazine Risperidone m o. 18 m o. 17 m o. 16 m o. 15 m o. 14 m o. 13 m o. 12 m o. 11 m o. 10 m o. 9 m o. 8 m o. 7 m o. 6 m o. 5 m o. 4 m o. 3 m o. 2 m o. 1 se lin e $0 Ba Monthly expenditures $2,000 Monthly Costs: All outpatient (mental health and medical/surgical)costs (mean=$372/month)(ITT)(p=ns) $700 $500 $400 $300 $200 $100 Olanzapine Quetiapine Ziprasidone Perphenazine Risperidone m o. 18 m o. 17 m o. 16 m o. 15 m o. 14 m o. 13 m o. 12 m o. 11 m o. 10 m o. 9 m o. 8 m o. 7 m o. 6 m o. 5 m o. 4 m o. 3 m o. 2 m o. 1 se lin e $0 Ba Monthly expenditures $600 Monthly health costs (IP and OP) excluding medications (mean=$1,047/month)(ITT)(p=ns) $2,100 $1,400 $700 Olanzapine Quetiapine Ziprasidone Perphenazine Risperidone m o. 18 m o. 17 m o. 16 m o. 15 m o. 14 m o. 13 m o. 12 m o. 11 m o. 10 m o. 9 m o. 8 m o. 7 m o. 6 m o. 5 m o. 4 m o. 3 m o. 2 m o. 1 se lin e $0 Ba Monthly expenditures $2,800 Monthly Costs: All healthcare costs including medications (mean=$1,544/month)(actual drug prices with discounts)(ITT)(p=P<O,Q,R,P)(APA). Monthly expenditures $2,800 $2,100 $1,400 ` $700 Olanzapine Quetiapine Ziprasidone Perphenazine Risperidone m o. 18 m o. 17 m o. 16 m o. 15 m o. 14 m o. 13 m o. 12 m o. 11 m o. 10 m o. 9 m o. 8 m o. 7 m o. 6 m o. 5 m o. 4 m o. 3 m o. 2 m o. 1 Ba se lin e $0 Cost Summary In CATIE average total health care costs were 20%-30% lower for perphenazine than for second generation antipsychotics, because of lower drug cost. $300-$500/month $3,600-$6,000/year Cost-effectiveness question: Are benefits (quality of life) or avoidance of risks (TD) worth this public expenditure? Magnitude of cost should = magnitude of benefit CATIE Cost-Effectiveness (2) EFFECTIVENESS PANSS TOTAL SCORE (LS Means from Mixed Models (Adj for site, baseline, exacerbation) 80 PANSS 70 60 50 40 30 Baseline 1 mo. 3 mo. 6 mo. Olanzapine Quetiapine Ziprasidone 9 mo. 12 mo. 15 mo. Perphenazine Risperidone Olanzapine<risperidone, quetiapine (with Hochberg adjustment for multiple comparisons) 18 mo. Quality Adjusted Life Years (LS Means from Mixed Models (Adj for site, baseline, exacerbation) (all group p<.0001; time p<.0001) 0.80 QALYs 0.60 0.40 0.20 0.00 Baseline 1 mo. 3 mo. 6 mo. Olanzapine Quetiapine Ziprasidone 9 mo. 12 mo. 15 mo. Perphenazine Risperidone Perphenazine>risperidone (with Hochberg adjustment for multiple comparisons) 18 mo. Visual Analogue Scale: 0-100 [death - perfect health] (LS Means from Mixed Models (Adj for site, baseline, exacerbation)(all group p= ns, time p<.0002) Monthly expenditures 75 60 45 30 15 0 Baseline 1 mo. 3 mo. 6 mo. Olanzapine Quetiapine 9 mo. 12 mo. 15 mo. Perphenazine Risperidone 18 mo. Aripiprazole vs. Perphenazine in patients. Refractory olz and risp: Kane Meltzer, 2007 6 week run-in trial with olanzapine or ripseridone with <20% PANSS response 334/416 completed open label phase 2% responded Mean PANSS increased by 3.5% on olanz and 3.2% of risperidone 225 (75%) completed random phase 6 weeks outcome Completed DC Adevrse event DC lack of Effic Aripiprazole 71% 14% 6% Perphenazine 79% 8% 5% Aripiprazole vs. Perphenazine in patients. Refractory olz and risp: Kane Meltzer, 2007 Aripip PANSS (mean) -9.8 PANSS Responder (25%) 27% CGI -0.3 QLS responders 35% Serious Adv evnt. 21% EPS 13.7% Akithesia 6.3% EPS meds 27.8% Body weight (loss) -1.5 Perph. -10.5 ns 25% ns -0.3 ns 26% ns 17% ns 19.4% ns 9.0% ns 17.6% ns -2.2 ns A Randomized Clinical Trial of LongActing Injectable Risperidone (CONSTA ®) and Oral Antipsychotics in Unstable Chronic Schizophrenia RA Rosenheck, JH Krystal, R Lew, PG Barnett, L Fiore, D Valley, SS Thwin, JE Vertrees, MH Liang for the CSP555 Investigators VA Cooperative Studies Program NEJM, March 4, 2011 Time on randomized drug (log rank=1.7, df=1. p<0.19) 1 Survival Distribution Function 0.9 0.8 0.7 0.6 0.5 0.4 OralOral 0.3 LAI Risp LAI-Risperidone 0.2 0.1 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Month Time to rehospitalization: p=.39: LAI HR .87 (95% CI .63-1.2) PANSS Follow-up score* Total PANSS 85 80 75 70 Oral 65 LAI 60 55 50 Baseline 3 Mos 6 Mos 9 Mos 12 Mos 15 Mos 18 Mos 21 Mos 24 Mos Oral N= 178 140 122 117 112 99 84 68 50 LAI N= 182 147 132 120 111 100 86 71 57 * (t=.47, df=1944, p=.65) Tardive Dyskinesia (TD) Risk SGA=vs. FGA Annual Incidence (0.8% vs 5.4%)* vs Health Outcomes Six Dimensions of Sensitivity analysis i) severity, ii)duration, iii)treatment with SGAs, iv)QOL, iv)QALYs v) Annual cost * Correll. Leucht and Kane, AJP 161(3): 596-601, 2004 ICERs for TD If ,as per CATIE cost difference is $2,400- $3,600$6,000 cost/case of TD yields the following matrix: D iff in CATIE CATIE Total Effe ct low/long-term risperidone D iff. in Annua lize d H e a tlh Cost vs Pe rph. $2,400 $6,216 Difference in Risk of TD cases 4.6% $52,174 $135,130 Assume all cases are severe 0.14 $372,671 $965,217 Asssume 2/3 mild (QALY loss= 0.07) 0.093 $561,010 $1,453,015 Assume 15% of cases last < 3 months 0.85 $660,012 $1,709,430 Off Label Use Most patients prescribed SGAs do not have indicated treatments. VA 1999: 33% off label Market Scan® 2004: 56% off label VA 2007: 60.2% off label Recent review by AHRQ found weak evidence of effectiveness of off-label uses CATIE AD Trial: Quality Adjusted Life Years from the Health Utilities Index Mark III 0.4 Monthlyy QALY 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 Baseline 3 mo. 6 mo. Olanzapine (N=100, 91, 80, 74) Quetiapine (N=94, 83, 74, 68) Risperidone (N=85, 115, 74, 70) Placebo (N=142, 123, 115, 109) 9 mo. CATIE AD Trial: Monthly Costs: Experimental Medication costs Monthly expenditures $200 $100 $0 1 mo. 2 mo. 3 mo. 4 mo. 5 mo. 6 mo. 7 mo. Olanzapine (mean=$106) Quetiapine (mean=$102) Risperidone (mean = $92) Placebo (mean = $45) 8 mo. 9 mo. CATIE AD Trial: Average Monthly Costs: All healthcare costs including medications* Monthly expenditures $1,600 $1,200 $800 $400 $0 Baseline 1 2 3 4 5 6 7 Olanzapine (mean=$1,118) Quetiapine (mean=$1,215) Risperidone (mean=$1,092) Placebo (mean=$1,023) 8 9 How Could This Be? (1) Leucht meta-analysis of 150 short-term studies (2009) showed cloazpine (ES=.52), amisuplridide (ES=-.31), olanzapine (ES=-.28) and risperidone (ES=-.13) to be superior to FGAs. “ Most previous studies addressed pure efficacy and safety, whereas in the CATIE and CUtLASS studies the investigators focused on real-world effectiveness. “ How Could This Be? (2) “A strength of CATIE and CUtLASS was the use of comparator drugs that are less potent than haloperidol. A major limitation of our metaanalysis is that haloperidol was the comparator drug in most of the studies, and the number of studies of mid-potency first-generation drugs was insufficient.” Leucht et al. 2009, p. 40. Four groups of APS drugs Risperidone or any FGA Clozapine (2 or 3 failures) Generic available Weight gain risk is of concern and some patients may not tolerate the required blood monitoring Aripiprazole, ziprasidone or quetiapine Olanzapine: greatest weight gain Implementation Two principals (not a protcol: flexibility) Do not create disincentive to take medication Individualize treatment when needed Apply preferential sequence when: a) initiating therapy or b) change is required Self-authortization of group 3 and 4 drug use (aripiprazole, ziprasidone, quetiapine, olanzapine) Mandatory review form before Rx is filled Review form sociodemographic characteristics diagnoses risk factors for TD and metabolic disease previous antipsychotic medication failures, clinical reasons for selecting the proposed agent. Consent for FGAs (TD) clozpaine, quetiapine, olanzapine (weight gain, DM risk) All important criteria for selecting medications. Incentives to Change Academic Detailing; Education/Information Monitor prescribing decision making with a structured form when NEW SGAs are prescribed Very weak disincentive (time burden for the form) Feedback information in further academic detailing Caveat: Astra-Zeneca complained that we were not treating all SGAs equally as per VA agreement so we had to add risperidone to level 2 VA CT: Monitoring Form Data On All new Starts of Atypical Antipsychotics FY 20082009 (N=1,721): Diagnosis Schizophrenia or BP Schizophrenia Bipolar disorder Sleep Other Diagnoses Other affective PTSD 44% 18.1% 26.6% 18% 38% 24.0% 23.4% Monitoring Form Data (N=1,721): Reason for new medication Efficacy Sleep Patient preference Less EPS Less TD risk Less akathisia Less sedation Treatment of TD Other 39.3% 30.0% 27.0% 12.5% 8.9% 4.7% 5.8% 0.8% 20.7% Monitoring Form Data (N=1,721): Co-Morbidity TD EPS Akathisia Diabetes Hyperlipidemia Obesity Hypertension - ASCVD 4.3% 3.7% 3.0% 14.9% 29.0% 20.5% 34.0% 10.6% (DK-17.0%) (DK-11.3%) (DK-10.2%) (DK-24.5%)