Article
Caregivers’ willingness-to-pay for Alzheimer’s disease medications in Canada
Dementia 2015, Vol. 14(1) 63–79 ! The Author(s) 2013 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/1471301213490709 dem.sagepub.com
Mark Oremus McMaster University, Canada
Jean-Eric Tarride McMaster University, Canada
Eleanor Pullenayegum McMaster University, Canada
Natasha Clayton McMaster University, Canada
Canadian Willingness-To-Pay Study Group Memorial University (St. John’s, NL): Gerry Mugford, Marshall Godwin; McGill University (Montreal, QC): Allen Huang, Yves Bacher; Institut universitaire de ge´riatrie de Montre´al (Montreal, QC): JuanManual Villalpando; Queen’s University (Kingston, ON): Sudeep S. Gill; Sunnybrook Health Sciences Centre (Toronto, ON): Krista L. Lanctoˆt, Nathan Herrmann; McMaster University (Hamilton, ON): David Cowan; University of Western Ontario (London, ON): Robert Petrella; University of Calgary (Calgary, AB): David B. Hogan; University of British Columbia (Vancouver, BC): Philip E. Lee.
Parminder Raina McMaster University, Canada
Abstract We studied caregivers’ willingness-to-pay for Alzheimer’s disease drug therapy. We recruited 216 caregivers of persons with mild or moderate Alzheimer’s disease and presented them with four scenarios describing a hypothetical Alzheimer’s disease medication. The scenarios described the medication as capable of either treating the symptoms of disease or modifying the course of disease. The scenarios also presented two different probabilities of adverse effects occurrence, i.e. 0% or 30%. Most caregivers said they would pay out-of-pocket for the medication, with support for such payment ranging from 68% to 93%, depending on the specific scenario. The highest level of support was for the ‘disease modifying and no adverse effects’ scenario, while the lowest level was for the ‘symptom treatment and 30% chance of adverse effects’ scenario. Corresponding author: Mark Oremus, Department of Clinical Epidemiology and Biostatistics, McMaster University, 50 Main Street East, Room 308, Hamilton, Ontario L8N 1E9, Canada. Email: [email protected]
64
Dementia 14(1)
On average, caregivers’ monthly willingness-to-pay out-of-pocket for the medication ranged from $214 to $277 (Canadian dollars). Dollar amounts were highest for the ‘disease modifying and no adverse effects’ scenario and lowest for the ‘symptom treatment and 30% chance of adverse effects’ scenario. Support for out-of-pocket payment and specific dollar amounts were highest when the medication did not involve adverse effects. Caregivers placed more value on the absence of adverse effects than on drug efficacy. Keywords adverse effects, Alzheimer’s disease, caregiver, efficacy, medication, willingness-to-pay
Introduction Alzheimer’s disease (AD) is a progressive neurological disorder characterized by cognitive impairment, functional limitations and behavior and mood problems (Reitz, Brayne, & Mayeux, 2011). The earliest symptom is usually memory loss; in the later stages of disease, persons with AD become dependent on caregivers to provide the basic needs of life. Medications are often used to treat AD, but existing drugs are not curative. The efficacy of these medications is limited to relieving the symptoms of disease (Raina et al., 2008). AD negatively impacts the life of unpaid caregivers, who are typically close family members of persons with AD. These caregivers report high levels of burden (Zucchella, Bartolo, Pasotti, Chiapella, & Sinforiani, 2012), stress (Lee, McKeith, Mosimann, Ghosh-Nodyal, & Thomas, 2013) and depression (Mausbach, Chattillion, Roepke, Patterson, & Grant, 2013), as well as low quality-of-life (Theofilou, 2012). The impact of AD on caregivers led Zarit, Orr, and Zarit (1985) to refer to caregivers as the ‘hidden victims’ of the disease. Evidence suggests that caregivers of persons who take AD medications may experience lower levels of burden (Feldman et al., 2003; Wilcock et al., 2003) and distress (Cummings, Schneider, Tariot, Kershaw, & Yuan, 2004; Moretti, Torre, Antonello, Cazzato, & Bava, 2002) than caregivers of persons who do not take AD medications. This may occur because the medications relieve disease symptoms and consequently make persons with AD easier to manage. Additionally, caregivers could experience a sense of personal relief because something is being done to help a loved one, or because they feel optimistic that treatment will improve the loved one’s condition. Given the potential benefits of AD medications from the caregiver perspective, we conducted a willingness-to-pay study to obtain the value AD caregivers place on drug therapy. Since research is ongoing to develop medications that may someday slow or halt disease progression (Massoud & Leger, 2011), we elicited caregivers’ willingness-to-pay for two types of drugs: one that only offered symptom treatment (relief) and one that offered disease modification. In addition, we assessed the influence of adverse effects on caregivers’ willingness-to-pay.
Design and methods Participants We recruited unpaid caregivers of persons with a diagnosis of mild or moderate AD. Diagnoses were based on criteria enumerated in the Diagnostic and Statistical Manual of
Oremus et al.
65
Mental Disorders – Fourth Edition (Diagnostic and Statistical Manual of Mental Disorders, 2000) or the National Institute of Neurological and Communicative Disorders and Stroke – Alzheimer’s Disease and Related Disorders Association (McKhann et al., 1984). Classification of AD into mild or moderate stages was done using the Functional Assessment Staging (FAST) in Alzheimer’s disease scale (Sclan & Reisberg, 1992). Recruitment took place in nine Canadian memory or geriatric clinics between November 2008 and August 2011. The clinics were located in St. John’s, Montre´al (two clinics), Kingston, Toronto, Hamilton, London, Calgary and Vancouver. We excluded caregivers who could not communicate in English or French, as well as caregivers who were caring for someone with severe AD.
Data collection The data collection instrument was a two-part questionnaire, which research assistants administered in one-on-one interviews held at the clinics or in the home. The first part of the questionnaire contained questions on sample characteristics (e.g. age, sex), caregiver attitudes (e.g. satisfaction with healthcare services), the type of care provided (e.g. hours per day spent caregiving) and care recipient information (e.g. living arrangements). Caregivers also completed the AD Knowledge Test (ADKT) (Dieckmann, Zarit, Zarit, & Gatz, 1988), which contains five true or false questions. One point is awarded for each correct answer and higher scores indicate a better knowledge of AD. Caregivers used the Dementia Behavior Disturbance Scale (DBDS) (Baumgarten, Becker, & Gauthier, 1990) to rate the extent of care recipients’ behavior problems in 28 areas. The DBDS captures the frequency of each behavior on a five-point scale from 0 (never) to 4 (all of the time). Total scores range from 0 to 112 and higher scores indicate greater levels of behavior disturbance. The second part of the questionnaire presented caregivers with four scenarios describing a hypothetical new AD medication. Figure 1 depicts these scenarios. For every scenario, caregivers indicated whether they would support paying out-of-pocket for the medication (yes/no). Caregivers who answered ‘yes’ were presented with three bids ($75, $150 and $225) representing monthly out-of-pocket costs. These bids were anchored on the average daily cost of an AD medication, which is $5.00 ($150 per month) (Alzheimer’s Society of Canada, 2013). Caregivers accepted or rejected each bid. If all three bids were accepted, then we asked caregivers to specify their maximum willingness-to-pay amount. We did not assign an upper limit to the amount they could specify. Caregivers who rejected all three bids were invited to provide an amount between $0 and $74. We randomized caregivers to the order of scenarios and bids to minimize ordering effects. Caregivers were expressly told in the scenarios that the government would not pay for the hypothetical new medication. Caregivers’ maximum willingness-to-pay (MWTP) for each scenario was their highest bid for that scenario. For caregivers who accepted or rejected all three bids, MWTP was their maximum specified amount or the amount between $0 and $74. If a caregiver would not support paying out-of-pocket for a medication, then MWTP for the scenario in question was assigned a value of $0. We used every caregiver’s MWTP to calculate a mean MWTP for each scenario. We excluded one caregiver from the computation of mean MWTP, and from all regression analyses related to MWTP, because this individual provided an outlier response of $5,000 for each scenario.
66
Dementia 14(1)
Disease Modification
Symptom Treatment
Medication halts the progression of
Medication treats the symptoms of
cognitive impairment
cognitive impairment without halting the progression of impairment
Scenario A
Scenario B
Scenario C
Scenario D
DM medication
DM medication has
ST medication does
ST medication has a
does not have
a 30% chance of
not have adverse
30% chance of
adverse effects
adverse effects
effects
adverse effects
Figure 1. Alzheimer’s disease medication scenarios. Adverse effects were described as primarily gastrointestinal, e.g. nausea, vomiting. We chose gastrointestinal adverse effects to mimic the most common problems associated with existing Alzheimer’s disease medications. DM: disease modifying; ST: symptom treatment.
Although we report median MWTP in addition to mean MWTP, our results focus on mean MWTP. Mean values are consistent with measuring the strength of stakeholder preferences (Dolan, 2000), which meshes with welfare economics and maps onto our desire to estimate the value caregivers place on drug therapy. Also, willingness-to-pay methodology is consistent with the tenets of welfare economics (Gafni, 1991). All monetary values are reported in Canadian dollars. Based on the average exchange rate in 2012, $1.00 (Canadian) was worth an average of E0.78 (Euro) or $1.00 (United States) (Bank of Canada, 2013).
Statistical analysis Based on the results of willingness-to-pay research in the general public (Oremus et al., 2012), we developed two hypotheses to examine: (1) caregiver support for out-of-pocket payment and MWTP would be greater in disease modifying versus symptom treatment scenarios; and (2) support and MWTP would also be greater in scenarios without adverse effects versus scenarios with a 30% chance of adverse effects. To test these hypotheses, support for out-of-pocket payment (yes/no) and MWTP were treated as dependent variables in separate regression (generalized estimating equation) models. These models each contained two scenario ‘grouping variables’ as independent variables: (1) disease modification versus symptom treatment and (2) no adverse effects versus 30% chance of adverse effects.
Oremus et al.
67
To elicit the determinants of supporting out-of-pocket payment and MWTP, we utilized a two-step regression approach. Regarding support for payment, the first step involved simple logistic regression models wherein support for payment was regressed onto each variable shown in Table 1. The second step involved multivariable logistic regression modeling to regress support for payment onto all variables that met a significance threshold of ¼ 0.05 in the simple regression models. For the determinants
Table 1. Sample characteristics (n ¼ 216). Characteristic Age (years) Sex Female Male Education High school or less At least some technical/community college (includes completed technical/community college) At least some university (includes completed university undergraduate degree) At least some post-graduate school (includes completed post-graduate school) Annual household income
Caregivers’ willingness-to-pay for Alzheimer’s disease medications in Canada
Dementia 2015, Vol. 14(1) 63–79 ! The Author(s) 2013 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/1471301213490709 dem.sagepub.com
Mark Oremus McMaster University, Canada
Jean-Eric Tarride McMaster University, Canada
Eleanor Pullenayegum McMaster University, Canada
Natasha Clayton McMaster University, Canada
Canadian Willingness-To-Pay Study Group Memorial University (St. John’s, NL): Gerry Mugford, Marshall Godwin; McGill University (Montreal, QC): Allen Huang, Yves Bacher; Institut universitaire de ge´riatrie de Montre´al (Montreal, QC): JuanManual Villalpando; Queen’s University (Kingston, ON): Sudeep S. Gill; Sunnybrook Health Sciences Centre (Toronto, ON): Krista L. Lanctoˆt, Nathan Herrmann; McMaster University (Hamilton, ON): David Cowan; University of Western Ontario (London, ON): Robert Petrella; University of Calgary (Calgary, AB): David B. Hogan; University of British Columbia (Vancouver, BC): Philip E. Lee.
Parminder Raina McMaster University, Canada
Abstract We studied caregivers’ willingness-to-pay for Alzheimer’s disease drug therapy. We recruited 216 caregivers of persons with mild or moderate Alzheimer’s disease and presented them with four scenarios describing a hypothetical Alzheimer’s disease medication. The scenarios described the medication as capable of either treating the symptoms of disease or modifying the course of disease. The scenarios also presented two different probabilities of adverse effects occurrence, i.e. 0% or 30%. Most caregivers said they would pay out-of-pocket for the medication, with support for such payment ranging from 68% to 93%, depending on the specific scenario. The highest level of support was for the ‘disease modifying and no adverse effects’ scenario, while the lowest level was for the ‘symptom treatment and 30% chance of adverse effects’ scenario. Corresponding author: Mark Oremus, Department of Clinical Epidemiology and Biostatistics, McMaster University, 50 Main Street East, Room 308, Hamilton, Ontario L8N 1E9, Canada. Email: [email protected]
64
Dementia 14(1)
On average, caregivers’ monthly willingness-to-pay out-of-pocket for the medication ranged from $214 to $277 (Canadian dollars). Dollar amounts were highest for the ‘disease modifying and no adverse effects’ scenario and lowest for the ‘symptom treatment and 30% chance of adverse effects’ scenario. Support for out-of-pocket payment and specific dollar amounts were highest when the medication did not involve adverse effects. Caregivers placed more value on the absence of adverse effects than on drug efficacy. Keywords adverse effects, Alzheimer’s disease, caregiver, efficacy, medication, willingness-to-pay
Introduction Alzheimer’s disease (AD) is a progressive neurological disorder characterized by cognitive impairment, functional limitations and behavior and mood problems (Reitz, Brayne, & Mayeux, 2011). The earliest symptom is usually memory loss; in the later stages of disease, persons with AD become dependent on caregivers to provide the basic needs of life. Medications are often used to treat AD, but existing drugs are not curative. The efficacy of these medications is limited to relieving the symptoms of disease (Raina et al., 2008). AD negatively impacts the life of unpaid caregivers, who are typically close family members of persons with AD. These caregivers report high levels of burden (Zucchella, Bartolo, Pasotti, Chiapella, & Sinforiani, 2012), stress (Lee, McKeith, Mosimann, Ghosh-Nodyal, & Thomas, 2013) and depression (Mausbach, Chattillion, Roepke, Patterson, & Grant, 2013), as well as low quality-of-life (Theofilou, 2012). The impact of AD on caregivers led Zarit, Orr, and Zarit (1985) to refer to caregivers as the ‘hidden victims’ of the disease. Evidence suggests that caregivers of persons who take AD medications may experience lower levels of burden (Feldman et al., 2003; Wilcock et al., 2003) and distress (Cummings, Schneider, Tariot, Kershaw, & Yuan, 2004; Moretti, Torre, Antonello, Cazzato, & Bava, 2002) than caregivers of persons who do not take AD medications. This may occur because the medications relieve disease symptoms and consequently make persons with AD easier to manage. Additionally, caregivers could experience a sense of personal relief because something is being done to help a loved one, or because they feel optimistic that treatment will improve the loved one’s condition. Given the potential benefits of AD medications from the caregiver perspective, we conducted a willingness-to-pay study to obtain the value AD caregivers place on drug therapy. Since research is ongoing to develop medications that may someday slow or halt disease progression (Massoud & Leger, 2011), we elicited caregivers’ willingness-to-pay for two types of drugs: one that only offered symptom treatment (relief) and one that offered disease modification. In addition, we assessed the influence of adverse effects on caregivers’ willingness-to-pay.
Design and methods Participants We recruited unpaid caregivers of persons with a diagnosis of mild or moderate AD. Diagnoses were based on criteria enumerated in the Diagnostic and Statistical Manual of
Oremus et al.
65
Mental Disorders – Fourth Edition (Diagnostic and Statistical Manual of Mental Disorders, 2000) or the National Institute of Neurological and Communicative Disorders and Stroke – Alzheimer’s Disease and Related Disorders Association (McKhann et al., 1984). Classification of AD into mild or moderate stages was done using the Functional Assessment Staging (FAST) in Alzheimer’s disease scale (Sclan & Reisberg, 1992). Recruitment took place in nine Canadian memory or geriatric clinics between November 2008 and August 2011. The clinics were located in St. John’s, Montre´al (two clinics), Kingston, Toronto, Hamilton, London, Calgary and Vancouver. We excluded caregivers who could not communicate in English or French, as well as caregivers who were caring for someone with severe AD.
Data collection The data collection instrument was a two-part questionnaire, which research assistants administered in one-on-one interviews held at the clinics or in the home. The first part of the questionnaire contained questions on sample characteristics (e.g. age, sex), caregiver attitudes (e.g. satisfaction with healthcare services), the type of care provided (e.g. hours per day spent caregiving) and care recipient information (e.g. living arrangements). Caregivers also completed the AD Knowledge Test (ADKT) (Dieckmann, Zarit, Zarit, & Gatz, 1988), which contains five true or false questions. One point is awarded for each correct answer and higher scores indicate a better knowledge of AD. Caregivers used the Dementia Behavior Disturbance Scale (DBDS) (Baumgarten, Becker, & Gauthier, 1990) to rate the extent of care recipients’ behavior problems in 28 areas. The DBDS captures the frequency of each behavior on a five-point scale from 0 (never) to 4 (all of the time). Total scores range from 0 to 112 and higher scores indicate greater levels of behavior disturbance. The second part of the questionnaire presented caregivers with four scenarios describing a hypothetical new AD medication. Figure 1 depicts these scenarios. For every scenario, caregivers indicated whether they would support paying out-of-pocket for the medication (yes/no). Caregivers who answered ‘yes’ were presented with three bids ($75, $150 and $225) representing monthly out-of-pocket costs. These bids were anchored on the average daily cost of an AD medication, which is $5.00 ($150 per month) (Alzheimer’s Society of Canada, 2013). Caregivers accepted or rejected each bid. If all three bids were accepted, then we asked caregivers to specify their maximum willingness-to-pay amount. We did not assign an upper limit to the amount they could specify. Caregivers who rejected all three bids were invited to provide an amount between $0 and $74. We randomized caregivers to the order of scenarios and bids to minimize ordering effects. Caregivers were expressly told in the scenarios that the government would not pay for the hypothetical new medication. Caregivers’ maximum willingness-to-pay (MWTP) for each scenario was their highest bid for that scenario. For caregivers who accepted or rejected all three bids, MWTP was their maximum specified amount or the amount between $0 and $74. If a caregiver would not support paying out-of-pocket for a medication, then MWTP for the scenario in question was assigned a value of $0. We used every caregiver’s MWTP to calculate a mean MWTP for each scenario. We excluded one caregiver from the computation of mean MWTP, and from all regression analyses related to MWTP, because this individual provided an outlier response of $5,000 for each scenario.
66
Dementia 14(1)
Disease Modification
Symptom Treatment
Medication halts the progression of
Medication treats the symptoms of
cognitive impairment
cognitive impairment without halting the progression of impairment
Scenario A
Scenario B
Scenario C
Scenario D
DM medication
DM medication has
ST medication does
ST medication has a
does not have
a 30% chance of
not have adverse
30% chance of
adverse effects
adverse effects
effects
adverse effects
Figure 1. Alzheimer’s disease medication scenarios. Adverse effects were described as primarily gastrointestinal, e.g. nausea, vomiting. We chose gastrointestinal adverse effects to mimic the most common problems associated with existing Alzheimer’s disease medications. DM: disease modifying; ST: symptom treatment.
Although we report median MWTP in addition to mean MWTP, our results focus on mean MWTP. Mean values are consistent with measuring the strength of stakeholder preferences (Dolan, 2000), which meshes with welfare economics and maps onto our desire to estimate the value caregivers place on drug therapy. Also, willingness-to-pay methodology is consistent with the tenets of welfare economics (Gafni, 1991). All monetary values are reported in Canadian dollars. Based on the average exchange rate in 2012, $1.00 (Canadian) was worth an average of E0.78 (Euro) or $1.00 (United States) (Bank of Canada, 2013).
Statistical analysis Based on the results of willingness-to-pay research in the general public (Oremus et al., 2012), we developed two hypotheses to examine: (1) caregiver support for out-of-pocket payment and MWTP would be greater in disease modifying versus symptom treatment scenarios; and (2) support and MWTP would also be greater in scenarios without adverse effects versus scenarios with a 30% chance of adverse effects. To test these hypotheses, support for out-of-pocket payment (yes/no) and MWTP were treated as dependent variables in separate regression (generalized estimating equation) models. These models each contained two scenario ‘grouping variables’ as independent variables: (1) disease modification versus symptom treatment and (2) no adverse effects versus 30% chance of adverse effects.
Oremus et al.
67
To elicit the determinants of supporting out-of-pocket payment and MWTP, we utilized a two-step regression approach. Regarding support for payment, the first step involved simple logistic regression models wherein support for payment was regressed onto each variable shown in Table 1. The second step involved multivariable logistic regression modeling to regress support for payment onto all variables that met a significance threshold of ¼ 0.05 in the simple regression models. For the determinants
Table 1. Sample characteristics (n ¼ 216). Characteristic Age (years) Sex Female Male Education High school or less At least some technical/community college (includes completed technical/community college) At least some university (includes completed university undergraduate degree) At least some post-graduate school (includes completed post-graduate school) Annual household income
