Peer-Reviewed

Economic Impact of a Novel Naloxone Autoinjector on Third-Party Payers


Richard C. Weiss, MS
Managed Solutions LLC, Mt. Freedom, N.J.
Gary R. Bazalo, MS, MBA
Managed Solutions LLC, Mt. Freedom, N.J.
Heather Thomson, MS
Kaléo Pharma, Richmond, Va.
Eric Edwards, MD, PhD
Kaléo Pharma, Richmond, Va.

Abstract


Background: Patient overdoses on prescription opioid analgesics in the United States continue to rise, resulting in increased emergency department and hospitalization costs. Opioid overdose is readily reversible with naloxone, a fast-acting opioid antagonist. A new naloxone autoinjector (NAI), Evzio, which does not require medical training to use, was approved by the FDA in April 2014. Payers must decide on reimbursement policies for this product.


Purpose: To demonstrate to payer decision makers the costs and potential medical resource cost offsets associated with the utilization of a new NAI.


Design: A deterministic model using matched controls.


Methodology: An Excel-based cost model was developed for a hypothetical health plan with 1 million adult members. Costs of prescription opioid overdose events for patients appropriately dispensed NAI were compared with matched controls.


Results: NAI prescriptions increased from 218 in Year 1 to 2,527 in Year 3. In Year 3, 86 NAI patients (and their matched controls) experienced opioid overdose events. For this period, fatal overdoses in the NAI cohort totaled 11.1 vs. 14.7 for the control group. In Year 3, 2.5 deaths (10.1–7.6) were avoided.


NAI acquisition costs rose from $125,000 in Year 1 (PMPM = $0.01) to nearly $1.5 million in Year 3 (PMPM = $0.12). This cost was offset by medical resource savings of approximately $84,000 in Year 1, increasing to $975,000 in Year 3. The total net cost (NAI less offsets) in Year 3, when NAI uptake was assumed to plateau, was $481,000 (PMPM = $0.04).


Conclusion: A deterministic model demonstrated that NAI acquisition costs can be offset through medical cost reductions with improved timely access to naloxone.


Introduction


As sales of prescription opioid analgesics in the United States have quadrupled in the last decade, there has been an accompanying rise in emergency department (ED) visits and hospitalizations involving adverse events with these medications. In fact, the problem is so serious that for every death caused by prescription opioids, there are 10 treatment admissions for abuse, 26 opioid-related ED visits, 108 people who abuse and are dependent upon opioids, and 733 people using prescription opioids for nonmedical purposes (CDC 2012). Although the public health focus has been on the number of deaths from prescription opioid overdose (nearly 17,000 in 2011, which equates to 46 deaths per day or 1 death every 36 minutes [CDC 2014]), the number of opioid-related poisoning events leading to hospitalization is also of concern. For example, a study based on 2009 data estimated prescription opioids to be responsible for 108,106 admissions following ED care in the U.S. population (Inocencio 2013). This same report estimates the prevalence of prescription opioid overdoses at 130 per 100,000 lives. Therefore, a health plan or payer with 1 million lives can expect over 1,300 prescription opioid analgesic overdoses to occur annually.


Although significant numbers of persons suffer the life-threatening respiratory depression of opioid overdose as a result of abusing opioid medications, the data suggest that the majority of patients experiencing opioid overdose neither have an identified history of substance abuse, nor are they “doctor shoppers,” and they have only 1 prescriber for their opioid medications (Paulozzi 2012). It also has been documented that up to 60% of opioid-related fatalities occur in patients provided prescriptions based on current prescribing guidelines, with 20% of deaths occurring in patients taking <100 mg morphine equivalent daily dose (MEDD) (Manchikanti 2012). Furthermore, a number of concomitant illnesses and other factors besides substance use disorder and high doses of opioids have been identified for increased risk of opioid overdose, including chronic pulmonary disease (Hasegowa 2014), renal or hepatic disease, trauma, depression, sleep apnea, and use of extended-release or long-acting opioids (Boyer 2012, Zedler 2014). Drug-drug interactions involving inhibition of CYP450 metabolic pathways have also been implicated (Tennant 2011).


Opioid analgesic overdose encompasses a range of clinical findings, including the classic toxidrome of apnea, stupor, and miosis. The sine qua non of serious opioid toxicity is respiratory depression (Boyer 2012). Time is a critical factor in successful treatment, as prolonged respiratory depression can lead to hypoxic damage to the central nervous system or to death. The drug of choice for treating a patient with opioid-induced respiratory or central nervous system depression is parenteral (intravenous, subcutaneous, or intramuscular) administration of naloxone (Boyer 2012, Compton 2013). Naloxone is a competitive m opioid receptor antagonist with high affinity for the receptor site. With the timely administration of naloxone, an opioid poisoning event is reversible (SAMHSA 2014, Boyer 2012). With more than 40 years of clinical use, naloxone injection is well documented to be safe and effective. It is standard practice in emergency settings, even when opioid overdose is suspected but unconfirmed because of its rapid action as an opioid antagonist coupled with having no pharmacologic activity in the absence of opioids (SAMHSA 2014, Compton 2013, Boyer 2012).


To date, preparations of naloxone were developed for professional administration in supervised medical settings (Throckmorton 2014). In recent years, some communities have developed opioid overdose education programs that distribute naloxone “kits.” The kits usually contain two doses of naloxone plus gloves, needles, and syringes, along with brochures and ancillary medical supplies such as rescue breathing masks (Coffin 2013). Other programs provide kits with prefilled glass syringes for intranasal administration of naloxone utilizing a separate mucosal atomization device (Doe-Simkins 2014), but the intranasal route is not approved by the FDA, and safety or efficacy data with this route of administration are limited (Zedler 2014).


Evzio (naloxone hydrochloride) autoinjector

Evzio (naloxone hydrochloride) autoinjector


Under its priority review program, the FDA recognized the need for broader access of naloxone, especially by family members and other caregivers in a position to administer naloxone to a patient during an opioid overdose event. The agency approved the Evzio (naloxone hydrochloride) autoinjector in April 2014, which rapidly delivers a single dose of naloxone via a handheld autoinjector that can be carried in a pocket or stored in a medicine cabinet (FDA 2014). This is significant because laypersons now have available to them a naloxone product approved for use by non-health care professionals (FDA 2014).


The NAI provides standard fixed dosing for adult and pediatric subjects, eliminating the variability of assembling a needle and syringe and manually drawing up the prescribed dose of naloxone. The inclusion of a fully automatic retractable needle system lessens the potential of inadvertent sharp injury. Functionally, the voice instruction system and light-emitting diode (LED) visual cues help guide a user through the administration steps (Evzio 2014). Finally, each NAI carton contains a trainer for practice. The trainer is part of an educational component in the introduction of NAI that includes ensuring patients, family members, and caregivers have information on the appropriate use of NAI in case it needs to be used in a suspected overdose emergency.


The cost-benefit analysis for a health plan involves balancing the projected annual pharmacy cost of NAI against the potential for medical cost reduction associated with opioid-related morbidity and mortality. Accordingly, an economic model was developed to study the financial impact of the use of the branded NAI compared with the current standard of care (emergency professional administration of naloxone), with particular emphasis on drug costs, costs of emergency department treatment, cost of hospitalization, costs associated with cardiac resuscitation, and costs related to fatalities.


The objectives of the model were to estimate (1) the number of appropriate patients for NAI therapy, (2) the net pharmacy costs for NAI brand prescriptions, and (3) the difference in medical resource utilization and costs between the NAI treatment group and a risk-matched control group administered the current standard of care (i.e., emergency medical service resuscitation and naloxone administration) over a 3-year time horizon in a hypothetical commercial health plan with 1 million adult members.


Key points


Assuming a health plan covering 1 million people and various criteria and assumptions developed by the researchers:


  • 4,356 people will be deemed appropriate for naloxone autoinjector (NAI) prescriptions, according to criteria developed by the researchers
  • Number of people who actually get a prescription will grow from 218 (5% of those deemed appropriate) in Year 1 to 2,527 (50%) in Year 3 as awareness of NAI increases
  • In Year 3, total acquisition cost of the NAI is $1,456,407 (PMPM = $0.12), assuming a wholesale acquisition cost of $575 per prescription
  • In Year 3, the net cost of NAI (the acquisition cost offset by lower ED and hospitalization costs) is $481,000 (PMPM = $0.04)
  • Over a 3-year period, NAI will prevent 3.6 fatal overdoses
  • In Year 3, cost per death avoided is $192,400 but that is highly dependent on assumptions about the incidence of overdoses


Methods


An Excel-based deterministic cost model was developed to estimate the combined pharmacy and medical resource costs to a health plan or other population-based health care organization owing to prescription opioid overdose events for appropriate patients dispensed NAI compared with matched control patients without NAI. The model assumes the latter group will receive some other form of naloxone administration in the event of a prescription opioid overdose. Medical resources included in the model are ED visits, hospitalizations, and cardiac resuscitations, as well as cost associated with deaths due to prescription opioid overdose.


Default values were provided for model parameters from the published literature where available. For modeling purposes, variables used for appropriate patient selection are shown in Table 1 with the values used in the base case analysis. To be deemed appropriate for NAI, two criteria were used in this sample representative patient population. First, patients had to be evaluated as either “chronic opioid users,” defined as patients with long-term prescription opioid use >120 days of supply over 6 months (Leider 2011), or as “at risk” acute or short-term opioid users. Examples of “at risk” short-term users are those with a prior opioid overdose episode or with comorbid respiratory disease such as asthma, sleep apnea, or chronic obstructive pulmonary disease who may be exposed to opioids. Second, patients were included based on a selection of risk factors that were most associated with opioid overdose in a Veterans Health Administration (VHA) study (Zedler 2014). These included having a history of drug dependence, minimum of one hospital admission in the past 6 months, and history of >100 mg MEDD (see Table 1 for prevalence of each of the selected risk factors). Upon meeting these criteria, 5% of these patients were assumed to receive an NAI prescription in Year 1, 20% in Year 2, and 50% in Year 3.


















Table 1 NAI patient selection
Variable category Model variable Default (base case) value Source
Potential patient and prescription volume/cost estimation US Census 2013 estimate of adult (18+) population 241,838,562 US Census Bureau 2013
Number of adult members in health plan 1,000,000 Base case assumption
Payer type Commercial Base case assumption
Prevalence of chronic and at-risk acute opioid users 2.4% (= 1.3% chronic + 5% of 21.4% acute) Leider 2011
Refill failure rate (failure to refill after NAI shelf-life expiry) 20% Base case assumption
NAI prescriptions per patient during 1 year 1.1 (= 90% x 1) + (10% x 2) Base case assumption
NAI shelf life (months) 20 Base case assumption
% of chronic and at-risk acute opioid users receiving NAI prescription Based on prevalence of drug dependence, hospitalization, and MEDD >100mg/day Base case assumption
Probability of appropriate patient getting prescription and filling (uptake) 5%, 20%, 50% for Years 1, 2, and 3 Base case assumption
WAC per NAI prescription $575 Kaléo
Member copayment per prescription $51 KFF 2012
Risk factors for appropriate patients History of drug dependence prevalence 0.6% Zedler 2014
≥1 inpatient stay in last 6 months prevalence 12.6% Zedler 2014
Morphine equivalent daily dosage >100 mg/day prevalence 6.0% Zedler 2014

These uptake figures were considered reasonable, given the novel nature of NAI. The model assumes that 90% of the appropriate patients receive one NAI prescription and 10% receive two NAI prescriptions (on the basis of such factors as being prescribed opioids such as buprenorphine that may require >2 doses of naloxone or living in rural locations with extended emergency medical service response times). Additionally, the model was built so that all but 20% of NAI prescriptions are replaced after 20 months owing to shelf-life expiration. NAI prescription costs to the health plan and member copayments are provided in Table 1.


Model variables for medical resource utilization are shown in Table 2 with the values used in the base case analysis. The incidence (events per year) of opioid overdose in the appropriate NAI patient population was 3.4% (Inocencio 2013, Kaléo data on file). The probability of resource use (probability of use per overdose event) for the control group was 100% for ED visits, 57.6% for hospital (inpatient) admission (Hasegowa 2014), 2.2% for cardiac resuscitation (Sporer 1996), and 11.8% for death (Dunn 2010). The probability of resource use for the NAI group was assumed to be 75% of the control group probability for all resources except ED, which was assumed to be the same as the control group.






















Table 2 Medical resource utilization
Variable category Model variable Default (base case) value Source
Overdose incidence Incidence of prescription opioid overdose among appropriate patients 3.4% Inocencio 2013
Medical resources: probability of use per opioid overdose event ED visits: NAI group 100% Same as control group for base case
ED visits: control group 100% Hasegawa 2014
Hospital admission: NAI group 43.2% 75% of control group for base case
Hospital admission: control group 57.6% Hasegawa 2014
Cardiac resuscitation: NAI group 1.7% 75% of control group for base case
Cardiac resuscitation: control group 2.2% Sporer 1996
Fatalities: NAI group 8.8% 75% of control group for base case
Fatalities: control group 11.8% Dunn 2010
Medical resources: cost per use ED visits: NAI group $2,021 75% of control group for base case
ED visits: control group $2,695 Inocencio 2013
Hospital admission: NAI group $21,816 56% of control group for base case
Hospital admission: control group $38,784 Inocencio 2013
Cardiac resuscitation: NAI group $94,916 Same as control group for base case
Cardiac resuscitation: control group $94,916 Paulsen 2012
Fatalities: NAI group $3,112 Same as control group for base case
Fatalities: control group $3,112 Paulsen 2012
NAI administration Probability NAI is administered in the event of opioid overdose for patient with dispensed NAI prescription 80% Coffin 2013

Cost per resource use for the control group was $2,695 for ED visits (Inocencio 2013), $38,784 for hospital (inpatient) admission (Inocencio 2013), $94,916 for cardiac resuscitation (Paulsen 2012), and $3,112 for death (Paulsen 2012). The cost for the NAI group was assumed to be 75% of the control group cost for ED visits and 56% for hospital admissions (75% of control group daily cost and reduction in length of stay from 4 to 3 days), based on the assumption that patients who are administered NAI in a timely fashion are likely to have fewer severe opioid overdose symptoms by the time they reach the ED. With the recent availability of NAI, data specific to interventions with NAI are not yet available. Therefore, the cost for the NAI group for cardiac resuscitation and death was assumed to be the same as the control group cost.


The probability of administering the NAI device in the event of an opioid overdose was assumed to be 80% based on previous models and associated data of naloxone use in suspected opioid overdose emergencies (Coffin 2013).


In addition to the base case analysis using the values in Tables 1 and 2, sensitivity analyses were conducted varying the probability of use and cost of medical resources for the NAI group (as a percentage of control group probability and cost), the incidence of opioid overdose, and the probability that NAI is administered in the event of an opioid overdose.


Results


The results of the base case analysis are shown in Tables 3 and 4. In Table 3, the number of at-risk opioid users (all chronic users plus a subpopulation of acute opioid users) in each year for a health plan with 1 million adult members was estimated to be 23,700. Of these, 18.4% met the requirement of having at least 1 of the 3 criteria for NAI selection (history of drug dependence, hospitalization in last 6 months, or history of >100 mg MEDD), resulting in 4,356 appropriate patients each year. Based on the assumed NAI uptake over the 3-year period, this resulted in 218 patients with NAI prescriptions in Year 1, 929 in Year 2, and 2,527 in Year 3. Table 3 shows the number of expected overdose events for these patients (in both the NAI and control groups) and the number of NAI patients in which the product is administered. In Year 3, the 2,527 NAI patients (and their matched controls) were expected to experience approximately 86 opioid overdose events. The bottom of Table 3 shows the number of patients in each group who access each type of medical resource. For the 3-year period, fatal overdoses in the NAI cohort totaled 11.1 vs. 14.7 for the control group. In Year 3, 2.5 deaths (10.1 – 7.6) were estimated to be avoided.















Table 3 Patient counts for base case: NAI group vs. control groupa
  Year 1 Year 2 Year 3
At-risk opioid users 23,700 23,700 23,700
Number of appropriate NAI patients 4,356 4,356 4,356
Patients dispensed NAI 218 929 2,527
Patients with opioid overdose event 7.4 31.6 85.9
Patients with opioid overdose event where NAI is utilized… 5.9 25.3 68.7
  Reduction with NAI NAI Control NAI Control NAI Control
…resulting in emergency department visits 0% 7.4 7.4 31.6 31.6 85.9 85.9
…resulting in hospital admission 25% 3.2 4.3 13.6 18.2 37.1 49.4
…resulting in cardiac arrest (survival) 25% 0.1 0.2 0.5 0.7 1.4 1.9
…resulting in fatal overdose 25% 0.7 0.9 2.8 3.7 7.6 10.1
aFor hypothetical health plan with 1 million adult members

Table 4 shows the NAI net acquisition costs and medical resource costs for NAI and control groups for the base case of 75% of ED and 56% of hospital costs for NAI as compared to controls, 80% likelihood of use of NAI in the case of opioid overdose, and incidence of overdose of 3.4%. The estimated total acquisition cost of NAI rises from approximately $125,000 in Year 1 (per member per month [PMPM] = $0.01) to nearly $1.5 million in Year 3 (PMPM = $0.12). This cost is offset by estimated medical resource savings of approximately $84,000 in Year 1, increasing to approximately $975,000 in Year 3. The resulting total net cost (NAI less offsets) in Year 3, which is when NAI uptake is assumed to plateau, is approximately $481,000 (PMPM = $0.04). The cost of medical resources to treat the estimated 86 opioid overdose events for the control group in Year 3 is approximately $2.4 million.
























Table 4 Costs for base case: NAI group vs. control group
    Net drug acquisition cost Total cost of ED visits Total hospital cost based on NAI patients Total cost of cardiac arrest (survival) Total cost of fatal overdoses Cost of medical resources Total costs (drug acquisition plus medical resources)
Year 1 NAI used $125,552 $15,967 $88,850 $12,372 $2,169 $119,358 $244,910
Control group   $19,959 $165,302 $15,465 $2,711 $203,437 $203,437
Difference $125,552 ($3,992) ($76,452) ($3,093) ($542) ($84,079) $41,473
%   (20%) (46%) (20%) (20%) (41%) 20%
Year 2 NAI used $535,690 $68,127 $379,092 $52,786 $9,255 $509,260 $1,044,950
Control group   $85,158 $705,288 $65,983 $11,569 $867,998 $867,998
Difference $535,690 ($17,032) ($326,196) ($13,197) ($2,314) ($358,738) $176,952
%   (20%) (46%) (20%) (20%) (41%) 20%
Year 3 NAI used $1,456,407 $185,219 $1,030,657 $143,513 $25,162 $1,384,551 $2,840,958
Control group   $231,524 $1,917,501 $179,391 $31,453 $2,359,869 $2,359,869
Difference $1,456,407 ($46,305) ($886,844) ($35,878) ($6,291) ($975,318) $481,089
%   (20%) (46%) (20%) (20%) (41%) 20%
3-year total NAI used $2,117,649 $269,313 $1,498,599 $208,671 $36,586 $2,013,169 $4,130,818
Control group   $336,642 $2,788,091 $260,838 $45,733 $3,431,304 $3,431,304
Difference $2,117,649 ($67,328) ($1,289,492) ($52,168) ($9,147) ($1,418,135) $699,515
%   (20%) (46%) (20%) (20%) (41%) 20%
Mean NAI used $705,883 $89,771 $499,533 $69,557 $12,195 $671,056 $1,376,939
Control group   $112,214 $929,364 $86,946 $15,244 $1,143,768 $1,143,768
Difference $705,883 ($22,443) ($429,831) ($17,389) ($3,049) ($472,712) $233,172
%   (20%) (46%) (20%) (20%) (41%) 20%

Sensitivity analyses


The incremental cost of NAI treatment, taking into account both drug acquisition cost and potential medical cost offsets, is highly dependent on the reduction in medical resource costs due to NAI treatment. For example, the NAI acquisition cost is offset completely when medical resource costs for the NAI group are 46% below control group costs. NAI acquisition cost is also completely offset if the incidence of overdose increases from 3.4% to 5.1% due to more frequent indicated uses of NAI with resulting medical resource savings. Table 5 shows the estimated cost offsets when the probability of NAI administration is reduced from 80% to 50%. Net costs (NAI less offsets) in Year 3 rise from approximately $481,000 (PMPM = $0.04) in the base case to approximately $847,000 (PMPM = $0.07).
























Table 5 Costs for base case analysis with 50% probability of NAI administration
    Net drug acquisition cost Total cost of ED visits Total hospital cost based on NAI patients Total cost of cardiac arrest (survival) Total cost of fatal overdoses Cost of medical resources Total costs (drug acquisition + medical resources)
Year 1 NAI used $125,552 $17,464 $117,519 $13,532 $2,373 $150,888 $276,440
Control group   $19,959 $165,302 $15,465 $2,711 $203,437 $203,437
Cost differential $125,552 ($2,495) ($47,783) ($1,933) ($339) ($52,549) $73,003
As a %   (13%) (29%) (13%) (13%) (26%) 36%
Year 2 NAI used $535,690 $74,514 $501,416 $57,735 $10,123 $643,787 $1,179,477
Control group   $85,158 $705,288 $65,983 $11,569 $867,998 $867,998
Cost differential $535,690 ($10,645) ($203,872) ($8,248) ($1,446) ($224,211) $311,479
As a %   (13%) (29%) (13%) (13%) (26%) 36%
Year 3 NAI used $1,456,407 $202,584 $1,363,223 $156,967 $27,521 $1,750,295 $3,206,702
Control group   $231,524 $1,917,501 $179,391 $31,453 $2,359,869 $2,359,869
Cost differential $1,456,407 ($28,941) ($554,278) ($22,424) ($3,932) ($609,574) $846,833
As a %   (13%) (29%) (13%) (13%) (26%) 36%
3 year total NAI used $2,117,649 $294,561 $1,982,158 $228,233 $40,016 $2,544,969 $4,662,619
Control group   $336,642 $2,788,091 $260,838 $45,733 $3,431,304 $3,431,304
Cost differential $2,117,649 ($42,080) ($805,932) ($32,605) ($5,717) ($886,334) $1,231,315
As a %   (13%) (29%) (13%) (13%) (26%) 36%
Mean NAI used $705,883 $98,187 $660,719 $76,078 $13,339 $848,323 $1,554,206
Control group   $112,214 $929,364 $86,946 $15,244 $1,143,768 $1,143,768
Cost differential $705,883 ($14,027) ($268,644) ($10,868) ($1,906) ($295,445) $410,438
As a %   (13%) (29%) (13%) (13%) (26%) 36%

Table 6 summarizes the results of the base case and the sensitivity analyses for medical resource cost reduction, incidence of opioid overdose, and probability of NAI administration.








Table 6 Summary of base case and sensitivity analysis
Scenario 3-year average incremental cost for NAI group compared with control group
Base case (25% cost reduction, LOS reduction from 4 to 3 days, and 3.4% incidence of opioid overdose, 80% NAI administration probability) $233,172
Base case with 50% NAI administration probability $410,438
Base case with 46% medical resource cost offset Zero (Break-even)
Base case with 5.1% incidence of opioid overdose Zero (Break-even)

Cost per death avoided


In the base case, the cost per death avoided is approximately $192,400 (2.5 deaths avoided in Year 3 for an incremental cost of $481,000). The cost per death avoided is highly sensitive to the incidence of overdose. Raising the overdose incidence from 3.4% in the base case to 4.0% lowers the cost per death avoided to $103,940 (three deaths avoided in Year 3 for an incremental cost of $308,974). An overdose incidence of 5% yields a cost per death avoided of $5,952 (3.7 deaths avoided in Year 3 for an incremental cost of $22,116). Break-even (NAI acquisition cost is offset by medical resource savings) occurs with an overdose incidence of just under 5.1%.


Discussion


The FDA and other federal, state, and local entities have made deaths due to opioid overdose, and in particular prescription opioid overdose, a high priority. In response to this public health emergency, government agencies are beginning to produce recommendations relating to naloxone coprescription for certain high-risk patients on opioids, and a growing number of municipalities are supplying their first responders, including police departments, with training and naloxone. For example, the Department of Health and Human Services and SAMHSA recently released an update to their Opioid Overdose Toolkit recommending the prescription of FDA-approved naloxone, including NAI, to certain patients on opioids at high risk, such as those “taking high doses of opioids for long-term management of chronic malignant or nonmalignant pain” and “those who are on certain preparations that may increase risk for opioid overdose such as extended release/long-acting preparations,” among others (SAMHSA 2014). In addition, states are enacting laws to provide immunity to laypersons who administer naloxone (Network 2014, Davis 2014). Additionally, the U.S. attorney general recently called overdoses from heroin and prescription medications an “urgent health crisis” and is recommending that all first responders carry naloxone.


The feasibility of caregivers or other laypersons responding to opioid overdose in a community setting rests on two key concepts. First, the ability to recognize life-threatening respiratory depression does not require medical training. Second, the safety profile of naloxone favors administration even if an opioid overdose is only suspected but not confirmed. This is because if a patient is taking opioids but there is not an overdose event, the risk associated with naloxone administration is limited to withdrawal, which in nonpostoperative settings has not been shown to be life-threatening with the exception of the neonate patient population (SAMHSA 2014, Compton 2013, Evzio 2014).


The limiting factor in naloxone reaching its full public health potential benefit has been its availability only in parenteral formulations requiring manual administration by a trained professional. With the FDA’s recent approval of Evzio, the first NAI that includes labeling for family member or caregiver use in nonmedically supervised settings, the landscape has changed dramatically (FDA 2014).


Although it is not the responsibility of third-party payers to solve the nation’s opioid overdose public health crisis, a payer’s decision to reimburse NAI may depend on these beliefs: (1) the provision of a naloxone preparation to patients for use by family, caregivers, friends, or coworkers is essential to saving lives in opioid overdose situations, (2) use of FDA-approved products is preferable over formulations that are not approved by the FDA, and (3) medical cost offsets based upon credible data or conservative assumptions justify the drug purchase costs and reimbursement. These considerations must all be put into the context that there continues to be significant growth in opioid overdose adverse events and their resulting costs.


With respect to the economics, the base case analysis in the model indicates that net NAI acquisition cost for a hypothetical health plan with 1 million adult members is likely to plateau at approximately $1.5 million annually. The net costs of NAI use (NAI acquisition cost less medical resource offsets) are primarily dependent on (1) the expected reduction in the probability of use and cost per use of medical resources, (2) the expected overdose incidence in the NAI-prescribed population, and (3) the probability that NAI will actually be administered following an overdose.


A further reduction in probability of resource use (for hospitalization, cardiac resuscitation and death) and cost per use (for ED and hospitalization) from 25% to 46% resulted in the net NAI acquisition cost being completely offset by medical resource savings. Similarly, increasing the expected overdose rate in these high-risk patients from 3.4% to 5.1% resulted in a complete offset of NAI acquisition cost. In studies of heroin users, the probability that a distributed naloxone kit was used each year was 13.6%, which indicates that our estimates of opioid overdose may be low (Coffin 2013).


However, these cost offsets may potentially be reduced by a lower NAI administration rate in the event of an opioid overdose. The base case offsets of $975,000 in Year 3 were reduced to $610,000 when the probability of administration was reduced from 80% in the base case to 50% (Table 5, appendix). In support of the 80% base case figure, this value was reported by Wagner in a study of heroin users (Coffin 2013). In other studies, 85% of heroin overdoses were witnessed, indicating that caregivers or other laypersons are an important factor in reversing overdoses using naloxone (Coffin 2013). Studies are in progress to determine actual usage of NAI in the event of a suspected opioid overdose.


The more practical issue involves providing NAI to patients at increased risk for opioid overdose, as the model’s incidence of overdose events in the NAI-prescribed population is highly dependent on patient selection. The 1.1% of patients on long-acting or extended-release opioids are easily identified, but the percentage of patients on short-acting opioids at increased risk is not as easily identified from an administrative claims database, nor can recommendations for prescribers be as straightforward.


The clinical implications of the modeling suggest that the expanded utilization of naloxone using NAI is likely to save lives and result in less severe morbidity of patients entering the ED. Opioid overdose patients receiving NAI prior to an ED visit are expected to use fewer ED services and have fewer hospitalizations. The economic implication is that utilization of NAI will result in medical cost offsets that can achieve break-even or better, negating the acquisition cost of the drug.


This study aims to provide a basis for further investigations and, pending real-world experience, serves to prompt additional research and analysis into this growing public health concern and the role of naloxone.


Limitations


The major limitation of this study is the reliance on naloxone experience in the community setting, primarily among heroin users. Additionally, the risk factors chosen for this sample representative patient population are not inclusive of all the risk factors associated with opioid overdose. As it is unclear how much overlap there is among all the possible conditions in the Zedler study (Zedler 2014), to include all risk factors would have resulted in a substantial overstatement of chronic and “at-risk” opioid users deemed appropriate for NAI (i.e., >90%). Although VHA provides a large national database from which to sample, the population comprises primarily older white males who receive most of their health care within a single closed system. Therefore, more work needs to be completed to identify risk factors associated with nonfatal and fatal overdose in a more demographically generalized population.


Studies, however, are currently ongoing to gather additional information that will help to inform the model. For example, the probability of NAI administration in the event of an opioid overdose needs to be determined after obtaining an appropriate amount of reliable data from the community setting.


This is a first-generation analysis based on the evidence available at this time, and as payers collect data on the use of NAI, the projected cost of various formulary policies will become more accurate. Furthermore, the objective of this study was to determine the cost efficiency of the use of NAI vs. the current standard of care based on product acquisition cost and reasonable assumptions for medical cost offsets. Therefore, we did not assess quality-adjusted life-years (QALYs) and other quality-of-life measures.


Conclusion


Opioid overdose is a serious and growing health concern in the United States that is largely preventable. The introduction of Evzio, the first FDA-approved naloxone autoinjector labeled for administration by non-health care professionals, such as family members or caregivers, allows for earlier intervention during a suspected opioid overdose emergency. Third-party payers must decide whether or not, and how, to reimburse for this potentially life-saving intervention. With this decision in mind, an economic model was developed to estimate the economic burden to a health plan. A base case analysis through a representative simulation indicates the potential for NAI administration to reduce fatalities, as well as to reduce medical resource costs due to less intensive interventions when NAI is used vs. current standard of care.


Actual savings in total pharmacy and medical costs depend on (1) the method used to identify appropriate NAI patients, (2) the probability that NAI will be administrated during a prescription opioid overdose event, (3) the incidence of opioid overdose events among patients prescribed NAI, and (4) the extent of the reduction in medical resource use and health care costs for patients who are administered NAI in the event of a prescription opioid overdose. The model demonstrates that anticipated medical cost offsets can be achieved to cover the cost of NAI.


Funding source: Kaléo Pharma, Richmond, Va.


Conflict disclosures: Kaléo Pharma is the maker of Evzio, the naloxone autoinjector approved by the FDA in April 2014. Heather Thomson and Eric Edwards are employees of Kaléo Pharma, and Edwards reports holding shares of the company. Richard Weiss and Gary Bazalo report receiving a fee from Kaléo Pharma for creating the model discussed in this article and for drafting the manuscript.


Corresponding author:

Richard C. Weiss

PO Box 526

Mt. Freedom, NJ 07869

rich@managedsolutionsllc.com


References


Boyer EW. Management of opioid analgesic overdose. N Engl J Med. 2012;367:146–155.


CDC (Centers for Disease Control and Prevention). Prescription drug abuse and overdose: public health perspective. Oct. 24, 2012. http://www.cdc.gov/primarycare/materials/opoidabuse/docs/pda-phperspecti.... Accessed Jan. 21, 2015.


CDC. Prescription drug overdose in the United States: fact sheet. July 3, 2014. https://www.cdc.gov/homeandrecreationalsafety/overdose/facts.html. Accessed Jan. 21, 2015.


Coffin PO, Sullivan SD. Cost-effectiveness of distributing naloxone to heroin users for lay overdose reversal. Ann Intern Med. 2013;158:1–9.


Compton WM, Volkow ND, Throckmorton DC, Lurie P. Expanded access to opioid overdose intervention: research, practice, and policy needs. Ann Intern Med. 2013;158:65–66.


Davis C. Naloxone Overdose Prevention Laws Map. Robert Wood Johnson Foundation Law Atlas. 2014. http://lawatlas.org/query?dataset=laws-regulating-administration-of-naloxone. Accessed Jan. 21, 2015.


Doe-Simkins M, Quinn E, Xuan Z, et al. Overdose rescues by trained and untrained participants and change in opioid use among substance-using participants in overdose education and naloxone distribution programs: a retrospective cohort study. BMC Public Health. 2014;14:297.


Dunn KM, Saunders KW, Rutter CM, et al. Opioid prescriptions for chronic pain and overdose: a cohort study. Ann Intern Med. 2010;152:85–92.


Evzio [package insert]. Richmond, Va.: Kaléo, Inc. April 2014.


FDA (U.S. Food and Drug Administration). FDA approves new hand-held auto-injector to reverse opioid overdose [press release]. April 3, 2014. Accessed Jan. 21, 2015.


Hasegawa K, Brown DF, Tsugawa Y, Camargo CA Jr. Epidemiology of emergency department visits for opioid overdose: a population-based study. Mayo Clin Proc. 2014;89:462–471.


Inocencio TJ, Carroll NV, Read EJ, Holdford DA. The economic burden of opioid-related poisoning in the United States. Pain Med. 2013;14:1534–1547.


Kaléo. Data on file.


KFF (Kaiser Family Foundation). 2012 Employer Health Benefits Survey. http://kff.org/private-insurance/report/employer-health-benefits-2012-annual-survey. Accessed Jan. 21, 2015.


Leider HL, Dhaliwal J, Davis EJ, et al. Healthcare costs and nonadherence among chronic opioid users. Am J Manag Care. 2011;17:32–40.


Manchikanti L, Helm S 2nd, Fellows B, et al. Opioid epidemic in the United States. Pain Physician. 2012;15(3 suppl):ES9–ES38.


Network for Public Law Health. Legal interventions to reduce overdose mortality: naloxone access and overdose good Samaritan laws. March 2013. https://www.networkforphl.org/_asset/qz5pvn/network-naloxone-10-4.pdf. Accessed Jan. 21, 2015.


Paulozzi L. CDC grand rounds: prescription drug overdoses—a U.S. epidemic. MMWR Morb Mortal Wkly Rep. 2012;61:10–13.


Paulsen MJ, Haddock AJ, Silbergleit R, et al. Empirical hospital and professional charges for patient care associated with out of hospital cardiac arrest before and after implementation of therapeutic hypothermia for comatose survivors. Resuscitation. 2012;83:1265–1270.


SAMHSA (Substance Abuse and Mental Health Services Administration). SAMHSA Opioid Overdose Prevention Toolkit. HHS Publication No. (SMA) 14-4742. Rockville, Md.: Substance Abuse and Mental Health Services Administration, 2014. http://store.samhsa.gov/shin/content/SMA14-4742/Overdose_Toolkit.pdf. Accessed Jan. 21, 2015.


Sporer KA, Firestone J, Isaacs SM. Out-of-hospital treatment of opioid overdoses in an urban setting. Acad Emerg Med. 1996;3:660–667.


Tennant F. Making practical sense of cytochrome P450. Practical Pain Management. 2011;10(4).


Throckmorton DC. Opioid auto-injector can help prevent overdose deaths. FDA Voice. April 3, 2014. https://www.fda.gov/newsevents/newsroom/fdavoices/default.htm. Accessed Jan. 21, 2015.


U.S. Census Bureau. https://www.census.gov/quickfacts/. Accessed Jan. 21, 2015.


Zedler B, Xie L, Wang L. Risk factors for serious prescription opioid-related toxicity or overdose among Veterans Health Administration patients. Pain Med. 2014;15:1911–1929.