Cost Considerations When Optimizing Sedation in the ICU John W. Devlin, PharmD, FCCP, FCCM Associate Professor Northeastern University School of Pharmacy Adjunct Associate Professor Tufts University School of Medicine Boston, Massachusetts 1) Optimizing patient comfort is a major goal when caring for patients in the ICU 2) Today, I would like to present a case of a 75 y/o lady I helped care for in the medical ICU at NEMC and who illustrates many of the dilemma that critical care pharmacists are faced with when trying to optimize patient comfort in the ICU 3) This case also provides a nice context for me to present some of the research I have completed in the first year of my fellowship 1

Faculty Disclosure It is the policy of The France Foundation to ensure balance, independence, objectivity, and scientific rigor in all its sponsored educational activities. All faculty, activity planners, content reviewers, and staff participating in this activity will disclose to the participants any significant financial interest or other relationship with manufacturer(s) of any commercial product(s)/device(s) and/or provider(s) of commercial services included in this educational activity. The intent of this disclosure is not to prevent a person with a relevant financial or other relationship from participating in the activity, but rather to provide participants with information on which they can base their own judgments. The France Foundation have identified and resolved any and all conflicts of interest prior to the release of this activity. Dr. Devlin has received grant/research support from Hospira. He has served as a consultant for AstraZeneca and Hospira.

Curriculum Learning Objectives Manage adult patients who need sedation and analgesia while receiving ventilator support according to current standards and guidelines Use validated scales for sedation, pain, agitation and delirium in the management of these critically ill patients Assess recent clinical findings in sedation and analgesia management and incorporate them into the management of patients in the acute care and procedural sedation settings

Activity Learning Objectives Understand the sedation-related factors that may influence patient care costs in the ICU Discuss the influence of sedation protocolization and interruption on care costs Compare the cost-efficiency of available sedation regimens in the ICU Develop a cost-effective regimen for sedation in the ICU

ICU Sedation: The Balancing Act and Ventilatory Optimization Patient Comfort and Ventilatory Optimization Oversedation Prolonged mechanical ventilation Increase length of stay Increased risk of complications - Ventilator-associated pneumonia Increased diagnostic testing Inability to evaluate for delirium G O A L Undersedation Patient recall Device removal Ineffectual mechanical ventilation Initiation of neuromuscular blockade Myocardial or cerebral ischemia Decreased family satisfaction w/care Jacobi J, et al. Crit Care Med. 2002;30:119-141. 5 5

Influence of Sedation Practices on Health Care Costs Outcomes and Costs of Undersedation Tachycardia ⇨ myocardial ischemia Use of neuromuscular blockers Prolonged neuromuscular weakness common Median $66,713 in additional costs (1995 costs) Short and long-term psychological effects Cost of device removal 10 patients removed 42 devices (GI/vascular) 74% were “significantly agitated” within 2 hr Total $7606 or $181/event (1997 costs) Annual cost at 42-bed ICU > $250,000 Dasta JF, et al. Crit Care Clin. 2009;25:571-583. Rudis MI, et al. Crit Care Med. 1996; 24:1749-1756. Fraser GL, et al. Pharmacotherapy. 2001;21:1-6. 6

Oversedation in the ICU Is Common Payen J, et al. Anesthesiology. 2007;106:687-695. Incidence of Inadequate Sedation A recent abstract by Kaplan and Bailey reported observational data when blinded BIS values were collected during standard practice care on patients who were receiving continuous infusions of sedatives and neuromuscular blocking agents. The results suggested that more than 54% of patients were receiving more sedation than was necessary, more than 15% were receiving less sedation than was necessary, and only about 31% were appropriately sedated. References 1. Kress et al. N Engl J Med. 2000;342:1471-1477. 2. Dasta JF, et al. Crit Care Med. 2005;33:1266-1271. N = 274 MICU patients 32% unarousable 21% no spontaneous motor activity Little variation over 24 hours in LOC, motor activity, or drug dose given RNs thought only 2.6% of patients were “oversedated” . Weinert CR, et al. Crit Care Med. 2007;35:393-401.

Costs of Oversedation in the ICU More days on mechanical ventilation = more of the costly negative sequelae associated with mechanical ventilation: – Ventilator-associated pneumonia: Occurs in 10%-20% of patients receiving mechanical ventilation for > 48 hrs Mean ICU LOS of 6.1 additional days (95% CI, 5.32-6.87) Additional cost of evaluating and treating a patient with VAP = $10,019 – Venous thrombosis: Incremental cost per event = $3,000 (in 2000 $) Costs associated with increased use of head CT scan/neurology consult Costs associated with increased frequency of sedation-related adverse events and drug withdrawal effects Propofol-related infusion syndrome Lorazepam-related propylene glycol toxicity Costs associated with increased neurocognitive sequelae PTSD Delirium Safdar N, et al. Crit Care Med. 2005; 33:2184-2193. Dasta JF, et al. Crit Care Clin. 2009; 25:571-583. Devlin JW, et al. Crit Care Med. 2010 (in press). Gunther ML, et al. Crit Care Clin. 2007; 23:491-506.

Influence of Sedation Practices on Health Care Costs ICU costs account for more than 1/3 of total inpatient costs Incremental cost of mechanical ventilation = $1522 per day Factors Affecting ICU Cost. Health care consumes 14% of the United States’ gross domestic product; inpatient care is responsible for ~1/3 of these health care costs. One of the largest cost drivers in the hospital setting is the intensive care unit (ICU), which despite accounting for 10% of the beds in U.S. hospitals, accounts for nearly 1/3 of total inpatient costs. Daily ICU care costs 3 to 5 times more than care provided on a general medical/surgical floor. Much of this increased cost may be due to interventions such as mechanical ventilation (MV). Patients who require MV represent approximately 33% of all patients admitted to the ICU and incur a disproportionately high share of the total cost of ICU treatment. Studies demonstrate that ICU patients who require >3 weeks of MV account for 50% of ICU cost. In addition to the economic consequences, it is well established that prolonged ICU stays and MV predispose patients to a greater risk of nosocomial infection and death.1 Incorporation of a daily sedation interruption policy into a medical ICU guideline can significantly reduce ICU stays and days of MV. Also, opioid administration and ICU complications are reduced in patients receiving daily sedation interruption2 References 1. Dasta JF, et al. Crit Care Med. 2005;33:1266-1271. 2. Wittbrodt ET. Pharmacotherapy. 2005;25(5 Pt 2):3S-7S. Dasta JF, et al. Crit Care Med. 2005;33:1266-1271.

Influence of Sedation Practices on Health Care Costs Use of a spontaneous awakening–spontaneous breathing trial protocol reduces duration of mechanical ventilation and length of ICU stay Outcome SBT SAT+SBT P-value Ventilator-free days 12 15 0.02 Time-to-event, days Successful extubation 7 5 0.05 ICU discharge 13 9 0.01 Hospital discharge 19 0.04 Factors Affecting ICU Cost. Health care consumes 14% of the United States’ gross domestic product; inpatient care is responsible for ~1/3 of these health care costs. One of the largest cost drivers in the hospital setting is the intensive care unit (ICU), which despite accounting for 10% of the beds in U.S. hospitals, accounts for nearly 1/3 of total inpatient costs. Daily ICU care costs 3 to 5 times more than care provided on a general medical/surgical floor. Much of this increased cost may be due to interventions such as mechanical ventilation (MV). Patients who require MV represent approximately 33% of all patients admitted to the ICU and incur a disproportionately high share of the total cost of ICU treatment. Studies demonstrate that ICU patients who require >3 weeks of MV account for 50% of ICU cost. In addition to the economic consequences, it is well established that prolonged ICU stays and MV predispose patients to a greater risk of nosocomial infection and death.1 Incorporation of a daily sedation interruption policy into a medical ICU guideline can significantly reduce ICU stays and days of MV. Also, opioid administration and ICU complications are reduced in patients receiving daily sedation interruption2 References 1. Dasta JF, et al. Crit Care Med. 2005;33:1266-1271. 2. Wittbrodt ET. Pharmacotherapy. 2005;25(5 Pt 2):3S-7S. Girard TD, et al. Lancet. 2008;371:126-134.

Pharmacist-Enforced Sedation Protocol 156 MICU patients prescribed continuous sedation Protocol encouraged 25% down-titration when patients more sedated than goal Before/after design evaluating impact of pharmacist promoting protocol on at least a daily basis. P < 0.00101 Median Days of Mechanical Ventilation 6.9 5.2 Marshall, J, et al. Crit Care Med. 2008;36 (2):427-433.

Influence of Sedation Practices on Health Care Costs: Delirium Choice of sedative influences incidence of delirium Delirium is associated with increased health care costs $ 14,730 $50,000 $41,836 $40,000 All P ≤ 0.002 $ 9,014 $30,000 $27,106 Median Cost $22,346 $20,000 $13,332 $10,000 Never Delirium (n = 41) Ever Delirium (n = 183) $0 ICU Cost Hospital Cost Milbrandt EB, et al. Crit Care Med. 2004;32:955-962.

Characteristics of the Ideal ICU Pharmacoeconomic Study Decision tree that is consistent with standard ICU care Developed by multidisciplinary panel of experts Includes all possible alternative therapies (eg, choice of sedative) as well as the most important positive and negative clinical outcomes Use of rigorous RCT(s) to define efficacy Cost estimates (ie, drug costs, patient charges etc) should be collected prospectively Costs related to drug dispensing, administration, monitoring and adverse events should be included Comprehensive sensitivity analysis Hay JW, et al. Value Health. 2009 Oct 28. [epub ahead of print] Dasta JF, Kane-Gill SL. Crit Care Clin. 2009;25:571-583. 13

Cost-Effectiveness Analysis Compares cost and outcomes of two different interventions All costs are related to a single, common outcome Duration of mechanical ventilation, mortality Helps answer the question: Is a new therapy with a higher acquisition cost worth it? A useful tool to help improve health care resource utilization at both the bedside and the institutional (eg, guidelines) level Hay JW, et al. Value Health. 2009 Oct 28. [epub ahead of print] 14

Cost-Effectiveness Analysis Outcomes Intermediate effectiveness outcomes/surrogate endpoints mm Hg blood pressure reduction % serum cholesterol reduction Duration at the targeted level of sedation Final effectiveness outcomes Mortality, years of life gained, presence of disease Length of mechanical ventilation or ICU stay Preferred but not easily measured What is a reasonable common effect of measurement when comparing treatments for anxiolysis?   Frequency of side effects Duration at target sedation score** Nurse satisfaction with titrability Duration of treatment with anxiolytic 15

Cost Minimization Analysis Compares two or more treatments demonstrated to have equal efficacy Cost is measured in dollars A measure of efficacy (eg, duration of mechanical ventilation) is not measured since it is assumed to be equivalent between 2 groups Helps the clinician(s) choose the least costly treatment Hay JW, et al. Value Health. 2009 Oct 28. [epub ahead of print] 16

Drug Acquisition Cost (70 kg patient, per day) Lorazepam 3 mg/hr: $35 Midazolam 5 mg/hr $42 Propofol 30 mcg/kg/min: $150 Dexmedetomidine 0.5 mcg/kg/hr: $274 Tufts Medical Center 2010 Pricing 17

Propofol vs Continuous Midazolam Randomized, unblinded pharmacoeconomic study Considered sedation drug costs and costs associated with ICU care (eg, MD consult, RN care, testing, etc) but did not measure actual ICU LOS or charges associated with it Propofol resulted in faster extubation time LOS in the ICU and total cost per patient did not differ Analysis of Covariance* Extubation Time Sedative Drug Cost Coefficient P-value Constant 0.58 0.02 4.61 0.001 Treatment effect (TE) Propofol vs midazolam -1.43 1.27 Sedation time (ST) 0.003 0.04 0.007 Interaction TE x ST -0.005 *Both extubation time and sedative drug cost were log-transformed 2 Anis A, et al. Anesthesiology. 2002;96:196-201.

Midazolam vs Lorazepam vs Propofol Hypothetical cohort study Compared the expected costs of short, intermediate and long-term sedation with propofol, lorazepam and midazolam Preparation, administration, adverse drug reactions (ADRs), therapeutic failures Relied on estimates from already published studies MacLaren R, et al. Pharmacotherapy. 2005;1319-1328.

Midazolam vs Lorazepam vs Propofol Total Cost ($) Probability of Being Least Costly (%) Short Lorazepam 497 (13.5 hrs) Midazolam 294 14 Propofol 272 86 Intermediate 932 (14.3 hrs) 587 97.5 674 2.5 Long 1604 84 (119.8 hrs) 1737 16 2003 MacLaren R, et al. Pharmacotherapy. 2005;1319-1328.

Carson SS, et al. Crit Care Med. 2006;34:1326-32. Scheduled Intermittent Lorazepam vs Propofol with Daily Interruption in MICU Patients Lorazepam (N = 64) Propofol (N = 68) P-value Ventilator days (median) 8.4 (4.6 to 14.7) 5.8 (3.5 to 10.3) 0.04 ICU LOS 10.4 (6.7 to 16.8) 8.3 (5.2 to 15.2) 0.20 APACHE II 22.9 ± 7.7 20.7 ± 7.3 0.05 Daily sedation dose 11.5 (3.8 to 22.7) mg 24.4 ± 16.3 mcg/kg/min - Morphine dose (mg/day) 10.7 (5.4 to 25.8) 31.6 (16.2 to 71.6) 0.001 Use of haloperidol 12% 9% 0.80 Carson SS, et al. Crit Care Med. 2006;34:1326-32. 21

Intermittent Lorazepam vs Propofol (DIS in both arms) Comprehensive MacLaren and colleagues also consider administration and prep time, ADEs such as VAP, post sedation agitation, neurologic changes…. NO one considered delirium AND delirium has gained substantial attention recently. In fact, BNZ use is an independent risk factor for delirium. DIS= Daily Interruption of Sedation Cox CE, et al. Crit Care Med. 2008;36:706-714. 22

Propofol Is More Cost-Effective Than Intermittent Lorazepam Use of propofol associated with lower total costs ($45,631 vs $52,009) Three more ventilator-free days with propofol Propofol was less costly or more effective in 94% of the 1000 simulations Cox C, et al. Crit Care Med. 2008;36:706-714.

Propofol is More Cost-Effective Than Intermittent Lorazepam Propofol less expensive 2 ` 0.5 Ratio of propofol to lorazepam MV days Average duration of MV Cost of ICU day Hospital mortality Cost of hospital ward day Probability of propofol intolerance Probability of lorazepam intolerance Crossover group from propofol Physician costs Cost of propofol Daily propofol dose, mg Cost of lorazepam Daily lorazepam dose, mg Low High Lorazepam more effective $1,825 $9,488 75% 5% $1,892 $631 20% 0%c 20% 0% Lorazepam Midazolam High Low $11.37 $60.77 4,347 949 $0.81 $7.82 23 4 -$35 -$30 -$25 -$20 -$15 -$10 -$5 $0 $5 $10 $15 $20 $25 $30 $35 Cost Difference Between Lorazepam and Propofol ($ Thousands) Cox CE, et al. Crit Care Med. 2008;36:706-714.

Addition of Dexmedetomidine to Midazolam and/or Propofol in Cardiac Surgery Patients Reduced Mean Total Treatment Charges P < 0.05 12-month retrospective administrative claims database analysis Comparison of patients receiving either midazolam plus propofol (M+P) or dexmedetomidine plus M+P (D+M+P) The D+M+P cohort showed significant reductions in per patient total charges Pharmacoeconomic Analysis: Reduced Mean Total Treatment Charges. The primary outcomes for the D+M+P cohort showed significant reductions in per patient total charges (approximately $18,000, P<.05), total LOS (0.6 days, P<.0001), ICU/CCU days (3.87, P<.0001), and mortality (2%, P=.0142). While pharmacy charges were higher (approximately $4000 per patient), lower ICU/CCU, operating room, room/board, and respiratory charges were observed in the D+M+P cohort. Mechanical ventilation was also shorter by approximately 0.5 days in D+M+P (P<.0001). Conclusions: These initial findings of a real-world assessment of dexmedetomidine use with other agents suggest favorable hospital clinical and economic outcomes. Further research to better understand the patient-mix, medication dosing, and causality is warranted through randomized clinical trials. Reference Dasta JF, et al. Pharmacotherapy. 2006;26:798-805. M+P, n = 9996 D+M+P, n = 356 Dasta JF, et al. Pharmacotherapy. 2006;26:798-805.

Addition of Dexmedetomidine to Midazolam and/or Propofol in Cardiac Vessel Surgery Patients Mean Total Charges Mean Length of Stay Mortality Rate P < 0.05 P < 0.01 P = 0.1074 Pharmacoeconomic Analysis: Reduced Charges, Hospitalization, and Mortality in Patients With Cardiac Vessel. Procedures. Similar to the general findings, cardiac surgery D+M+P patients appeared to have better outcomes than their M+P peers, averaging a half day shorter total length of stay, fewer days in the ICU/CCU, and a significantly lower mortality rate. Reference Dasta JF, et al. Pharmacotherapy. 2006;26:798-805. M+P, n = 7577 D+M+P, n = 293 Dasta JF, et al. Pharmacotherapy. 2006;26:798-805.

MENDS: Dexmedetomidine Versus Lorazepam Dexmedetomidine resulted in more days alive without delirium or coma (P = 0.01) and a lower prevalence of coma (P < 0.001) than lorazepam Dexmedetomidine resulted in more time spent within sedation goals than lorazepam (P = 0.04) Differences in 28-day mortality and delirium-free days were not significant 12 10 8 Days 6 4 MENDS: Dexmedetomidine Versus Lorazepam Effect on Acute Brain Dysfunction1 Investigators compared dexmedetomidine with lorazepam in 106 adult mechanically ventilated medical and surgical ICU patients to determine whether dexmedetomidine reduces the duration of delirium and coma in mechanically ventilated ICU patients while providing adequate sedation compared with lorazepam. The main outcome measures were days alive without delirium or coma and percentage of days spent within 1 Richmond Agitation-Sedation Scale (RASS) point of the sedation goal. Sedation with dexmedetomidine resulted in more days alive without delirium or coma (median days, 7.0 versus 3.0; P=.01) and a lower prevalence of coma (63% versus 92%; P <.001) Patients sedated with dexmedetomidine spent more time within 1 RASS point of their sedation goal compared those sedated with lorazepam (median percentage of days, 80% versus 67%; P=-.04). The 28-day mortality in the dexmedetomidine group was 17% versus 27% in the lorazepam group (P =.18), and the cost was similar between both groups. The investigators concluded that use of dexmedetomidine resulted in more days alive without delirium or coma and more time at the targeted level of sedation than with a lorazepam infusion. Reference 1. Pandharipande PP, et al. JAMA. 2007;298:2644-2653. 2 Dexmedetomidine n = 52 Lorazepam n = 51 Delirium/Coma-Free Days Delirium-Free Days Coma-Free Days Pandharipande PP, et al. JAMA. 2007;298:2644-2653.

MENDS Trial: Cost of Care Outcome Lorazepam Dexmedetomidine P-value Pharmacy 20.6 (10,42) 27.4 (16,46) 0.15 Respiratory 2.9 (2,6) 3.5 (2,7) 0.35 ICU cost 59.5 (36,83) 61.4 (37,108) 0.32 $ – Costs represented in thousands, US dollars (Median, IQR) MENDS Trial: Cost of Care Maximizing Efficacy of Targeted Sedation and Reducing Neurological Dysfunction (MENDS) Trial This study was designed to determine whether dexmedetomidine reduced the duration of delirium and coma in mechanically ventilated ICU patients while providing adequate sedation, compared with lorazepam. The median calculated cost for drug was $4675 in the dexmedetomidine group, and was $2335 in the lorazepam group. Across all charges, the median total hospital cost was $22 500 higher in the dexmedetomidine group, but this difference was not statistically significant. Reference Pandharipande PP, et al. JAMA. 2007;298:2644-2653. Pandharipande PP, et al. JAMA. 2007;298:2644-2653.

SEDCOM Trial: Efficacy Outcomes Midazolam (n = 122) Dexmedetomidine (n = 244) P-value Time in target sedation range, % 75.1 77.3 0.18 Patients completing daily arousal test 103 (84.3%) 225 (92%) 0.09 Patients requiring sedation interruption 112 (91.8%) 222 (91%) 0.85 Duration of sedation, days 4.1 3.5 0.01 Time to extubation, days 5.6 3.7 ICU length of stay, days 7.6 5.9 0.24 Delirium prevalence 93 (76.6%) 132 (54%) 0.001 Delirium-free days 1.7 2.5 0.002 Patients receiving open-label midazolam 60 (49%) 153 (63%) 0.02 Midazolam dose, mg/kg 0.11 0.65 Patients receiving fentanyl 97 (79.5%) 180 (73.8%) 0.25 Fentanyl dose, µg/kg 9.6 6.4 0.27 SEDCOM Trial: Efficacy Outcomes There was no significant difference between groups in the primary end point of the percentage of time during which patients were within the targeted sedation range of -2 to +1 on the RASS. Analyses of several secondary end points revealed various significant differences between the midazolam and dexmedetomidine treatment groups. Patients receiving dexmedetomidine had a shorter sedation duration and a shorter time to extubation. Additionally, patients receiving dexmedetomidine were less likely to experience delirium and had more delirium-free days than midazolam-treated patients. In addition to the differences listed on the slide, patients treated with dexmedetomidine exhibited better communication, cooperation, and tolerance score than did patients receiving midazolam. These measures were associated with greater tolerance of the ICU and mechanical ventilation. Reference Riker RR, et al. JAMA. 2009;301:489-499. Riker RR, et al. JAMA. 2009;301:489-499.

SEDCOM Trial Time to Extubation ICU Length of Stay P = 0.01 P = 0.24 Riker, RR, et al. JAMA. 2009;301:489-499.

SEDCOM Trial: Prevalence of Delirium Midazolam Dexmedetomidine dexmedetomidine versus midazolam, P < 0.001 Long-Term Sedation: Dexmedetomidine Versus Midazolam Delirium was assessed with the CAM-ICU scale. At baseline, 60.3% of patients receiving dexmedetomidine and 59.3% of patients receiving midazolam were CAM-ICU positive. Analysis of delirium through a generalized estimating equation revealed a significant difference between groups; patients receiving dexmedetomidine exhibited a 24.9% decrease in delirium. Reference Riker RR, et al. JAMA. 2009;301:489-499. Treatment Day Sample Size 118 229 109 206 92 175 77 134 57 92 42 60 44 34 Riker RR, et al. JAMA. 2009;301:489-499.

A Cost Minimization Analysis of a Clinical Trial Dexmedetomidine vs Midazolam Assumed equal sedation efficacy Cost minimization analysis Compared costs of care between groups and selected the therapy generating the lowest cost Economic analysis performed post-hoc and from the institutional perspective Costs were estimated from multiple databases and published literature Dasta JF, et al. Crit Care Med. 2010;38:497-503. Riker RR, et al. JAMA. 2009;301:489-499. Dasta JF, et al. Crit Care Med. 2005;33:1266-1271.

A Cost Minimization Analysis of a Clinical Trial Dexmedetomidine vs A Cost Minimization Analysis of a Clinical Trial Dexmedetomidine vs. Midazolam Components of total ICU costs Cost of ICU stay Cost of mechanical ventilation Cost of adverse drug reactions probably or possibly related to study drug Acquisition cost of study drugs Median regression approach was used to compare two treatment regimens Dasta JF, et al. Crit Care Med. 2010;38:497-503.

SEDCOM Cost of Care Median drug costs Dex $1166 Midazolam $60 Total ICU patient savings with Dex: $9679 Reduced ICU stay Reduced MV Dasta JF, et al. Crit Care Med. 2010;38:497-503.

A Cost Minimization Analysis of a Clinical Trial Dexmedetomidine vs Midazolam Dasta JF, et al. Crit Care Med. 2010;38:497-503.

Summary Choice of sedation practices have a substantial influence on the cost of care in the ICU Maintaining patients at the sedation goal through sedation protocolization (that may include daily sedation interruption) will avoid the costly sequelae of both under- and oversedation Both propofol and dexmedetomidine are more cost- beneficial than benzodiazepines Future studies need to evaluate the cost-effectiveness of propofol vs dexmedetomidine in non-cardiac surgery ICU patients