Cost-effectiveness of hearing and vision support for residents with dementia in long-term care in Ireland: A Health Economic Analysis Plan (HEAP) for a simulation model for the SENSE-Cog Care feasibility trial

Mikael Äijälä; Iracema Leroi; Dominic Trépel

doi:10.12688/hrbopenres.13948.1

Home Browse Cost-effectiveness of hearing and vision support for residents with...

ALL Metrics

-

Views

17

Downloads

Get PDF

Get XML

Export

▬

✚

Study Protocol

Cost-effectiveness of hearing and vision support for residents with dementia in long-term care in Ireland: A Health Economic Analysis Plan (HEAP) for a simulation model for the SENSE-Cog Care feasibility trial

[version 1; peer review: awaiting peer review]

Mikael Äijälä ¹, Iracema Leroi¹, Dominic Trépel¹

PUBLISHED 31 Oct 2024

Author details Author details

¹ Trinity College Institute of Neuroscience, Trinity College Dublin School of Medicine, Dublin, Leinster, D02 PN40, Ireland

Mikael Äijälä
Roles: Conceptualization, Data Curation, Formal Analysis, Investigation, Methodology, Project Administration, Validation, Visualization, Writing – Original Draft Preparation

Iracema Leroi
Roles: Conceptualization, Funding Acquisition, Project Administration, Resources, Writing – Review & Editing

Dominic Trépel
Roles: Conceptualization, Data Curation, Funding Acquisition, Methodology, Project Administration, Resources, Software, Supervision, Validation, Writing – Review & Editing

OPEN PEER REVIEW

REVIEWER STATUS AWAITING PEER REVIEW

This article is included in the Dementia Trials Ireland (DTI) and Dementia Research Network Ireland (DRNI) gateway.

Abstract

Introduction

Sensory challenges exacerbate the dementia symptoms of nursing home residents. The Sense-Cog Care trial is piloting a multi-faceted intervention to investigate whether optimising and supporting hearing, vision, and sensory-friendly environments through the “sensory champion” model in Irish nursing homes can improve quality of life and dementia-related outcomes and be cost-effective (Connelly et al., 2023).

A Health Economic Analysis Plan (HEAP) provides a standardised and comprehensive framework by outlining the systematic approach, methodologies, and key considerations involved in assessing the cost-effectiveness of a healthcare intervention. The HEAP presented in this paper describes the alongside-trial decision modelling approach used for preliminary estimates of cost-effectiveness and to inform the design of the final intervention.

Methods

A Markov model will be used to simulate residents' health progression with and without the intervention. Costs will be estimated from the Irish health and social care perspective. A key output of the planned economic analysis will be the Incremental-Cost-Effectiveness Ratio (ICER) which describes the cost over one additional unit of Quality-Adjusted Life Year (QALY). A Probabilistic Sensitivity Analysis (PSA) will be conducted in the form of a Monte Carlo simulation to examine parameter uncertainty and the probability of a sensory champion being cost-effective.

Results

Tables for inputs and outputs are presented, and graphical representations for uncertainty, such as a tornado plot and a Cost-Effectiveness Analysis Curve (CEAC), will be produced. The results will be interpreted in the context of the Irish cost-effectiveness thresholds of €20,000 and €45,000 per QALY.

Discussion

To our knowledge, this is the first study to model the cost-effectiveness of a sensory intervention for nursing home residents with dementia. We aim to develop an early-stage Bayesian model that serves as a prior for further research.

Keywords

Health Economic Analysis plan, Cost-effectiveness analysis, Dementia, Sensory care

Corresponding author: Mikael Äijälä

Competing interests: No competing interests were disclosed.

Grant information: Health Research Board [DIFA-2020-007]
The study was funded by HRB DIFA (2020) award entitled “SENSE-Cog Care: A feasibility study of hearing and vision support to improve quality of life in a Care home residents with dementia”

The funder did not have a say in the design of the model or the decision to publish this HEAP.

The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Copyright: © 2024 Äijälä M et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

How to cite: Äijälä M, Leroi I and Trépel D. Cost-effectiveness of hearing and vision support for residents with dementia in long-term care in Ireland: A Health Economic Analysis Plan (HEAP) for a simulation model for the SENSE-Cog Care feasibility trial [version 1; peer review: awaiting peer review]. HRB Open Res 2024, 7:70 (https://doi.org/10.12688/hrbopenres.13948.1) First published: 31 Oct 2024, 7:70 (https://doi.org/10.12688/hrbopenres.13948.1) Latest published: 31 Oct 2024, 7:70 (https://doi.org/10.12688/hrbopenres.13948.1)

Introduction

In Ireland, around 64,000 individuals are currently living with dementia, with around 34% residing in long-term care facilities. Projections indicate that this number will more than double, reaching over 150,000 by the year 2045 (HSE & National dementia office, 2020).

The Irish National Survey of Dementia in Long-Term Residential Care revealed that approximately 68% of residents in long-term care facilities are living with dementia (Cahill et al., 2014). The survey found that in specialised dementia care units, only about 50% of Long-term care staff and healthcare assistants (HCAs) had received training in dementia care and estimated that only 10% of residents’ symptoms are directly caused by dementia, while the remaining 90% result from the quality of care they receive in inappropriate settings (Cahill et al., 2014). Staff training is essential for managing the complex needs of dementia patients and improving their overall well-being, thereby enhancing their quality of life (Cahill et al., 2014).

Sensory impairments, including hearing and vision loss, are common among these residents, exacerbating cognitive decline and negatively impacting their quality of life (Amieva et al., 2015; Lin et al., 2013; Yeo et al., 2023) The Lancet Commission on dementia prevention identified hearing loss as the highest individual risk factor for developing dementia, with a relative risk of 1.09 (Livingston et al., 2020). Despite this, hearing aid use remains low. The Irish study of longitudinal ageing (TILDA) found that hearing loss is prevalent among older adults in Ireland, affecting 58% of men and 54% of women over the age of 75, yet only 11% of this population uses hearing aids (McGarrigle & Donoghue, 2023). The report linked the underutilisation to poorer mental health, increased loneliness and depression, and lower quality of life, which could be improved through better screening and provision of hearing aids.

The most reported reason for poor use of hearing aids/glasses is related to a lack of maintenance and care procedures and poor adherence support (Leroi et al., 2021). A study in the UK by Cross et al. (2024) reported that only 26.7% of long-term care staff regularly checked or tested hearing aids, providing sensory care to just half of those who needed it. Similarly, the provision of sensory assisting devices such as hearing aids or glasses is often overlooked, contributing to serious comorbidities arising from untreated sensory loss. This oversight is further exacerbated by suboptimal environmental conditions, including loud TVs, inadequate lighting, and poor surface design (Leroi et al., 2021).

The SENSE-Cog Care trial aims to address these issues by implementing a multi-faceted intervention in Irish long-term care homes and is currently conducting a feasibility trial. The intervention will include hearing and vision assessments and the provision of devices for the correction of sensory impairment. It will also include staff training and the election of staff members as ‘sensory champions’ who will receive further training in cognitive health and sensory cognitive support. The intervention will also focus on creating sensory-friendly environments and optimise care pathways to hearing/vision care providers. (Connelly et al., 2023). The trial aims to evaluate the intervention’s impact on key outcomes for residents with dementia and includes a health economics component (Connelly et al., 2023).

In a resource-constrained healthcare system, efficient allocation of limited resources is an important factor in policy decisions, and it has become more common to incorporate economic evaluations into randomised controlled trials (Drummond et al., 2015). Health economics aids these decisions by evaluating the cost-effectiveness of healthcare interventions. An economic analysis weighs the additional value of a particular technology against the opportunity cost of foregoing something else that the health budget could have generated, aiming for efficient allocation to maximise health or welfare outcomes (Briggs et al., 2006).

Writing a Health Economic Analysis Plan is recommended practice in alignment with agreed international guidelines with the Consolidated Health Economic Evaluation Reporting Standards (CHEERS) checklist (Husereau et al., 2024), which requires a Health Economic Analysis Plan to be written and made available prior to publication. A Health Economic Analysis plan establishes the strategy and the methods for the cost-effectiveness analysis, promoting transparency and establishing guidelines and structure for economic assessment. Thorn et al. (2021) conducted an Expert Delphi Consensus Survey on the content of Health Economics Analysis Plans (HEAPs) for trial-based economic evaluations. Based on this content, items for the model-based HEAP were selected, omitting those specific to trial-based evaluations.

This Health Economic Analysis Plan (HEAP) describes the alongside-trial model-based economic analysis used to produce initial estimates of cost-effectiveness based on available evidence. It was created following the guidelines for the Economic Evaluation of Health Technologies in Ireland from the Health Information and Quality Authority (HIQA, 2016). The methodologies selected align closely with those endorsed by reputable entities such as the National Institute for Health and Care Excellence (NICE) and are heavily inspired by seminal works such as the handbook "Decision Modelling for Health Economic Evaluation" by Briggs, Claxton, and Sculpher (2006).

This research aims to contribute to bridging the gap in economic evaluation of non-pharmaceutical interventions in long-term care home populations. As a preliminary assessment, it provides structured estimates of costs and outcomes associated with the intervention, informing the design of the final intervention, and laying the foundation for future research and discussions.

Methods

Study design

The planned economic evaluation will use a purpose-built, Excel-based Markov model to simulate the progression of dementia of long-term care home residents with various sensory conditions. The model will consider the prevalence of the various sensory conditions in the treatment-arm long-term care homes, which receive the SENSE-Cog Care sensory champion intervention, and control arm long-term care homes, which provide care-as-usual.

The extent to which the intervention reduces the prevalence of unaddressed sensory impairment will be modelled and the uncertainty reflected in sensitivity analysis, as well as its impact on the progression of dementia of the residents and their quality of life.

The model will estimate all the benefits and costs of each arm using a lifetime horizon, which means that all future costs and benefits will be considered, and an Incremental Cost-effectiveness Ratio (ICER) will be produced. ICER is defined as the difference in the costs divided by the difference in the effectiveness of alternative courses of action (Drummond et al., 2015). In this case, the comparison will be between the implementation of the SENSE-Cog Care intervention and continuing with care as usual. The ICER essentially quantifies the cost required to generate one additional unit of Quality-Adjusted Life Year (QALY), which provides a measure of the intervention’s value in terms of health outcomes. This figure can then be compared against the willingness-to-pay (WTP) threshold in Ireland, which is typically considered to be between €20,000 and €45,000 per QALY (O’Mahony & Coughlan, 2016). The WTP threshold represents the maximum amount that society is willing to pay for a QALY, providing a benchmark to determine whether the intervention is considered cost-effective (Drummond et al., 2015). The ICER will be calculated using the formula below:

ICER = \frac{\bar{C o s t_{S}} - \bar{C o s t_{C}}}{\bar{Q A L Y_{S}} - {\bar{Q A L Y}}_{C}} 1

, where $\bar{C o s t_{C}}$ describes the total costs of the control group (C), $\bar{C o s t_{S}}$ describes the total costs of the treatment group (S) that receives the sensory care intervention, $\bar{Q A L Y_{C}}$ is the sum of QALYs of the control group, $\bar{Q A L Y_{S}}$ is the sum of QALYs of the treatment group.

The model will consider the long-term care home residents with dementia with four categories (i) of sensory impairment: residents with no sensory impairment (1), residents with unaddressed hearing impairment (2), residents with unaddressed vision impairment (3) and residents with unaddressed dual impairment (4).

The sensory conditions are each assumed to have a different progression of dementia and to have an impact on health state related quality of life. This leads to differences in lifetime QALY’s and costs of residents in each category. They will be weighted (W%) according to their prevalence in the treatment and control arm nursing homes.

\bar{Q A L Y_{k}} = W i_{k} % * {QALY}_{k}^{i} 2

\bar{C o s t_{k}} = W i_{k} % * {COST}_{k}^{i} 3

i = {\begin{matrix} No unaddressed impairment (1), \\ Unaddressed hearing impairment (2), \\ Unaddressed vision impairment (3), \\ Unaddressed dual impairment (4) \end{matrix}}, k = {treatment arm (S), control arm (C)}

In summary, the planned CEA simulation will estimate the incremental costs and benefits of the intervention by examining how effectively it reduces the prevalence of unaddressed sensory impairments, which can impact the progression of dementia and improve the quality of life for residents in Irish long-term care homes.

Markov model

The Markov model (see Figure 1) will have four states: mild, moderate and severe stages of dementia, and a fourth absorbing state which is death. A one-way model of dementia is used where returning to a less severe state is not possible.

Figure 1. Markov model structure.

Figure 1 illustrates the key health states (represented by ovals) related to stages of dementia and transitions between these states (represented by arrows) indicating the likelihood of staying or moving between stages (Mild, Moderate and Severe) or the likelihood of death (an absorbing state) from each state.

The findings from Effective interventions for potentially modifiable risk factors for late-onset dementia: A costs and cost-effectiveness modelling study by Mukadam et al. (2020) will be used to determine transition probabilities and utilities. In this model, the probabilities will be adjusted by relative risk factors to capture the potential accelerated impact of sensory impairments on dementia progression. This will be done by multiplying the probabilities by a relative risk variable (RR) greater than 1 to account for the potential accelerated impact of poor sensory health on the natural progression of the disease (see Table 9 and Table 6). Population death rates will be imported from Irish data and will affect the transition probabilities according to age and sex (see Table 4).

The model cycle length will be quarterly, as the SENSE-Cog care will also conduct their follow-up quarterly. This cycle length also takes into account the small probability of progressing through multiple health states in a single year. The horizon will be a lifetime horizon to capture all the costs and health benefits of the intervention. This will be set at 10 years, which is 40 cycles.

Costs

The direct costs of the intervention, which are hypothetical since the model-based evaluation precedes the trial, along with the cost consequences, will be considered from an Irish health and social care perspective.

Direct costs will include a sensory champion cost, which considers any training-related staff costs of the intervention, purchases of hearing aids and glasses, which will be estimated based on the market price of these devices in Ireland, and hearing and vision screening costs, which will be estimated based on provider quotes (see Table 2). For cost consequences, the weekly institutional cost of care, along with medical, community, and informal care costs, will be considered (see Table 3). These will be estimated based on the best available evidence.

Costs and QALYs will be discounted at 4% to present value as per HIQA recommendations (HIQA, 2016). This reflects a societal preference for immediate benefits over future benefits and is an approach to express results in present value often found in guidelines, such as those by NICE (YHEC, 2016). Inflation is adjusted using the Irish Consumer Price Index (CPI) to standardize all costs to the equivalent value in 2024 euros.

QALY

Quality Adjusted Life Year (QALY) will be the effectiveness outcome value as recommended by HIQA. It captures an individual’s health-related quality of life and length of life and allows for comparability of economic evidence (Drummond et al., 2015). The QALY is calculated by multiplying a utility value by the duration spent in that state (e.g., one year in a health state with a utility value of 0.5 equates to 0.5 QALYs). Utility values measure preference for a health state, ranging from 0.0 (death) to 1.0 (perfect health) (Drummond et al., 2015).

The CEA will incorporate utility values from the study by Mukadam et al. (2020) for the health states of mild, moderate, and severe dementia (see Table 5). In the model, changes to the progression of dementia influence QALYs. For example, slowing down dementia progression results in fewer residents spending time in more severe, lower-utility states, thus increasing overall QALYs by extending time spent in higher-utility states.

For simplicity, sensory health will be categorised as either no impairment or unaddressed impairment, and its impact on utility will be similarly binary. The model incorporates the impact of sensory impairment on utility values by modelling it as a binary variable. It either describes normal utility or lower utility due to sensory impairment (see Formula 13 & Formula 14). This approach simplifies the parameter by creating an overall estimate of changes in health-related quality of life due to unaddressed sensory impairments.

Prevalence weights of sensory impairment

Assessing the prevalence of unaddressed sensory impairment before and after the intervention is a key assessment in evaluating the potential cost-effectiveness of the treatment.

The weights are determined based on the probability of a resident being in a particular category, considering both the presence of sensory impairment and whether it is addressed. The model considers the progression of dementia and associated health-state utilities to be the same for residents who have addressed their impairments as for those with no impairments, treating both groups as having ‘no unaddressed impairment’. Figure 2 shows a flow diagram that represents the part of the model that determines these weights.

Figure 2. Flow chart for the prevalence of unaddressed sensory impairment.

The flow chart determines the prevalence of dual impairment, hearing impairment, vision impairment and no impairment, depending on the answers given. Questions are represented by ovals; answers are indicated with arrows. If a resident has both vision and hearing impairments and neither is addressed, they are categorised as having unaddressed dual impairment. Addressing either hearing or vision impairment or having no impairment at all, classifies an individual as having no unaddressed impairment. Therefore, if a resident has unaddressed vision impairment but either does not have hearing impairment or has addressed hearing impairment, they are considered vision impaired only. Conversely, if a resident has unaddressed hearing impairment but either does not have vision impairment or has addressed vision impairment, they are categorised as having unaddressed hearing impairment.

Whether a sensory impairment is addressed or unaddressed depends on whether the impairment can be corrected with glasses or hearing aid, if the resident owns these devices, and if they use them (also referred to as adherence in the literature [Leroi et al., 2021]). The flowcharts in Figure 3 and Figure 4 illustrate how the model incorporates these considerations, by breaking down further the Is hearing addressed- and Is vision addressed ovals in figure 1.

Figure 3. Flow chart for unaddressed hearing impairment.

The flow chart determines how much of hearing impairment is addressed or unaddressed depending on the answers given. Questions are represented by ovals; answers are indicated with arrows.

Figure 4. Flow chart for unaddressed vision impairment.

The flow chart determines how much of vision impairment is addressed or unaddressed depending on the answers given. Questions are represented by ovals; answers are indicated with arrows.

Since the planned CEA will be conducted before the trial, the effects on the proportion of residents with glasses and hearing aids, and their usage, will be hypothetical (see Table 8 and Table 7). The simulation will assume that all residents will be screened, and those with addressable impairments in the intervention group will accept these devices. Due to the novelty of the intervention and uncertainty about its effectiveness in overcoming barriers to device use, the simulation will run different levels of adherence improvement—10%, 30%, and 70%—to assess the potential impact on the ICER.

The planned cost-effectiveness simulation results will be influenced by both the intervention’s ability to mitigate unaddressed sensory impairment, and the extent to which sensory impairment impacts Costs and QALY of residents with dementia in long-term care homes.

Probability sensitivity analysis

To conduct probabilistic sensitivity analysis (PSA) a Monte Carlo simulation is conducted to explore uncertainties in our model's input variables. Monte Carlo simulation treats model parameters as random variables, using probabilistic distributions to generate 1,000 ICER results by randomly sampling from these distributions. As recommended in Decision Modelling for Health Economics (Briggs et al., 2006) costs will be assumed to have a gamma distribution, utilities and transition probabilities will be distributed along a beta distribution, and the relative risk parameter will be assigned a lognormal distribution. Where the standard deviation is not known a 20% standard deviation is applied, with a 10% standard deviation being used for the utilities and transition probabilities from the Mukadam et al. (2020) study.

Deterministic sensitivity analysis

Scenario analysis will be performed on the treatment effect of improvement in the share of sensory aids used by those who have them, by +10%, +30%, and +70% (see Formula 23 to Formula 28).

In the absence of robust evidence, the analysis will vary key uncertain parameters by several orders of magnitude to assess their influence on the incremental cost-effectiveness ratio (ICER), without attempting to determine the most probable scenario.

Results

Since this is a plan, no results have been generated yet. Table 1 shows the expected presentation of results based on the methods described in this document.

Table 1. Deterministic results.

Scenario	ICER (€/QALY)
+10%	x
+30%	x
+70%	x

Table 2. Unit costs related to treatment.

Item description	Unit cost (€)	Source
Sensory champion cost	x	x
Hearing aid cost	x	x
Glasses cost	x	x
Hearing aid screening cost	x	x
Glasses screening cost	x	x

Table 3. Per resident weekly costs.

Item description	PP weekly Cost (€)	Source
Mild
Medical cost	x	x
Community cost	x	x
Informal care cost	x	x
Institutional care cost	x	x
Moderate
Medical cost	x	x
Community cost	x	x
Informal care cost	x	x
Institutional care cost	x	x
Severe
Medical cost	x	x
Community cost	x	x
Informal care cost	x	x
Institutional care cost	x	x

Table 4. The population death rates.

Age	Male	Female
n_t
n_t+1	x	x
n_t+2	x	x
n_t+3	x	x
n_t+4	x	x
n_t+5	x	x
n_t+6	x	x
n_t+7	x	x
n_t+8	x	x
n_t+9	x	x
n_t+10	x	x

Table 5. Other input items.

Item description	Value	Source
Health utility mild	x	x
Health utility moderate	x	x
Health utility severe	x	x
starting distribution mild $(n_{0, k}^{i} m i l d)$	x	x
starting distribution moderate $(n_{0, k}^{i} m o d e r a t e)$	x	x
starting distribution severe $(n_{0, k}^{i} s e v e r e)$	x	x
Irish CPI multiplier	x	x

Table 6. Relative risk for each category of sensory impairment.

Item description	Value	Source
RR unaddressed hearing loss	x	x
RR unaddressed vision loss	x	x
RR unaddressed dual impairment	x	x

Table 7. Care-as-usual inputs for weights.

Input item	P(%)	Source
P(Vision impairment)	x	x
P(VI which can be addressed by glasses)	x	x
P_C(has glasses)	x	x
P_C(uses glasses)	x	x
P(hearing impairment)	x	x
P(HI which can be addressed by hearing aids)	x	x
P_C(has hearing aids )	x	x
P_C(uses hearing aids)	x	x

Table 8. Intervention outcomes and also inputs for weights.

Input item	P(%)	Source
P_S(uses glasses)	x	Formula 23, Formula 25 & Formula 27
P_S(uses hearing aids)	x	Formula 24, Formula 26 & Formula 28
P_S (has glasses)	x	Formula 29
P_S(has hearing aids)	x	Formula 30

Table 9. Transition probabilities for the Markov model.

From/To	mild	moderate	severe	death
mild	1 – X * RRⁱ – pop rate	X * RRⁱ	0	pop rate
moderate	-	1 – Y * RRⁱ – Z – pop rate	Y * RRⁱ	Z + pop rate
severe	-	-	1 – Z – pop rate	Z + pop rate
death	-	-	-	1

Pop-rate stands for yearly population death rate and RRⁱ stands for relative risk for each category of impairment i. X is the transition probability from mild to moderate. Y is the transition probability from moderate to severe. Z is the increased probability of death due to dementia. Adapted from Effective interventions for potentially modifiable risk factors for late-onset dementia: a costs and cost-effectiveness modelling study, by Mukadam N, Anderson R, Knapp M, et al Lancet Healthy Longev 2020 with modifications by the author.

Deterministic results

Deterministic scenario analysis will demonstrate changes to the ICER based on improvements (of 10%, 30% and 70%) in the rate of use of hearing aids and glasses for residents who have them. Multiple tables will be produced if orders of magnitude analysis on key uncertain parameters is deemed necessary.

PSA results

To represent uncertainty in our model, probabilistic sensitivity analysis (PSA) will be conducted in line with methods commonly used in health economic evaluations ability to depict uncertainty and the robustness of the model’s outcomes (Briggs et al., 2006; Drummond et al., 2015). Results from the PSA will be presented in a scatterplot, showing the distribution of incremental costs and Quality-Adjusted Life Years (QALYs) on the cost-effectiveness plane. This will allow for a visual representation of the uncertainty surrounding the Incremental Cost-Effectiveness Ratio (ICER). Additionally, a cost-effectiveness acceptability curve (CEAC) will be plotted to demonstrate the probability of the intervention being cost-effective at various willingness-to-pay thresholds. To further explore the sensitivity of the ICER to individual parameters, a tornado diagram will be generated, ranking parameters by their impact on the ICER, with the most influential variables displayed at the top and the least influential at the bottom. These three steps of sensitivity analysis will be repeated for each deterministic order of magnitude scenario analysis.

Conclusion

Publishing a HEAP promotes transparency and establishes a structure for the economic assessment. Establishing a similar consensus to Thorn et al. (2021) around model-based HEAPs would be beneficial.

This Health Economic Analysis Plan presents an approach to simulating the cost-effectiveness of sensory aids for residents with dementia in long-term care settings in Ireland. To our knowledge, this is the first study to model the cost-effectiveness of a sensory intervention for nursing home residents with dementia. We aim to develop an early-stage Bayesian model that serves as a prior for further research.

This is HEAP version 1. Any modifications to this plan will be documented and reported in subsequent HEAPs, if changes occur.

Other relevant information

Reporting deviations: The following items from the Thorn et al. (2021) expert consensus on essential trial-based evaluation elements were excluded: Trial HEAP version, Trial protocol version, Trial SAP version, Trial design, Trial start and end dates, data cleaning for analysis, sampling uncertainty and resource use data collected.

The above items from Thorn et al. (2021) expert consensus of essential items were omitted as they were specific to trial-based evaluations.

Patients and the public will not participate in the design, execution, reporting, or dissemination of the planned model-based economic evaluation. PPI is involved in the wider study.

Formulas

To ensure full transparency and facilitate easy replication, the following sections provide a detailed description of the formulas, input variables, and the model’s process for generating the desired outputs. This thorough documentation also serves as a form of model validation, ensuring that the methodology is robust and can be independently verified by others.

Markov formulas

Mathematically the Markov model can be described as the multiplication of two matrices. This approach is also described in the appendix to the study by Mukadam et al. (2020).

1. A matrix that contains the number of residents in each state at any time t:
$N_{k, t}^{i} = [\begin{matrix} n_{k, t}^{i} m i l d \\ n_{k, t}^{i} m o d e r a t e \\ n_{k, t}^{i} s e v e r e \end{matrix}] 4$
$N_{k, 0}^{i}$ is the starting distribution at t=0, which will be determined from our Irish long-term care home dataset.
2. And a transition probability matrix:
$P^{i} = [\begin{matrix} m i l d t o m i l d & m i l d t o m o d e r a t e & m i l d t o s e v e r e & m i l d t o d e a d \\ m o d e r a t e t o m i l d & m o d e r a t e t o m o d e r a t e & m o d e r a t e t o s e v e r e & m o d e r a t e t o d e a d \\ s e v e r e t o m i l d & s e v e r e t o m o d e r a t e & s e v e r e t o s e v e r e & s e v e r e t o d e a d \\ d e a d t o m i l d & d e a d t o m o d e r a t e & d e a d t o s e v e r e & d e a d t o d e a d \end{matrix}] 5$
The distribution of residents at time t is given by the multiplication of the two matrices:
$N_{k, t}^{i} = N_{k, t - 1}^{i} \times P^{i} 6$

Cost formulas

Treatment arm (S) Cost

\begin{array}{l} {COST}_{S}^{i} = t_{0} C o s t o f I n t e r v e n t i o n \\ + \sum_{t = 1}^{40} ((c o s t c o n s e q u e n c e s m i l d * n_{i, S, t}^{t} m i l d + c o s t c o n s e q u e n c e s m o d e r a t e \\ * n_{i, S, t}^{t} m o d e r a t e + c o s t c o n s e q u e n c e s s e v e r e * n_{i, S, t}^{i} s e v e r e) * 13 / (1 \\ + d r / 4)^(t - 1)) \end{array} 7

, where the cost consequences for mild, moderate and severe states are calculated weekly and then adjusted to a quarterly basis by multiplying by 13. Similarly, the annual discount rate (dr) is divided by 4 for a quarterly discount rate.

Control arm (C) Cost

The formula for the control arm ${COST}_{C}^{i}$ is the same, except it does not include the cost of the intervention:

\begin{array}{l} COS T_{C}^{i} = \sum_{t = 1}^{40} ((c o s t c o n s e q u e n c e s m i l d * n_{C, t}^{i} m i l d + c o s t c o n s e q u n c e s m o d e r a t e \\ * n_{C, t}^{i} m o d e r a t e + c o s t c o n s e q u e n c e s s e v e r e * n_{C, t}^{i} s e v e r e) * 13 / (1 + d r / 4)^(t \\ - 1)) \end{array} 8

Direct cost formula

\begin{array}{l} Cost of intervention \\ = s e n s o r y c h a m p i o n c o s t + h e a r i n g a i d & g l a s s e s s c r e e n i n g c o s t s \\ + h e a r i n g a i d c o s t * P (n o f n e w h e a r i n g a i d s) + g l a s s e s c o s t \\ * P (n o f n e w g l a s s e s) \end{array} 9

, where P(n of new hearing aids) = P_S (has hearing aids) – P_C(has hearing aids) and stands for the number of new hearing aids purchased as part of the intervention and P(n of new glasses) = P_S(has glasses) – P_C(has glasses) which accounts for the number of new glasses purchased as part of the intervention.

Indirect cost formulas

\begin{array}{l} c o s t c o n s e q u e n c e s f o r m i l d \\ = a v e r a g e w e e k l y c o s t o f c a r e p e r r e s i d e n t f o r I r e l a n d \\ + m e d i c a l c o s t e u r o p e a n a v e r a g e f o r m i l d \\ + c o m m u n i t y c a r e c o s t e u r o p e a n a v e r a g e f o r m i l d \end{array} 10

\begin{array}{l} c o s t c o n s e q u e n c e s f o r m o d e r a t e \\ = a v e r a g e w e e k l y c o s t o f c a r e p e r r e s i d e n t f o r I r e l a n d \\ + m e d i c a l c o s t e u r o p e a n a v e r a g e f o r m o d e r a t e \\ + c o m m u n i t y c a r e c o s t e u r o p e a n a v e r a g e f o r m o d e r a t e \end{array} 11

\begin{array}{l} c o s t c o n s e q u e n c e s f o r s e v e r e \\ = a v e r a g e w e e k l y c o s t o f c a r e p e r r e s i d e n t f o r I r e l a n d \\ + m e d i c a l c o s t e u r o p e a n a v e r a g e f o r s e v e r e \\ + ​ c o m m u n i t y c a r e c o s t e u r o p e a n a v e r a g e f o r s e v e r e \end{array} 12

QALY formulas

The formula for QALY¹, which is the QALY for non-impairment, is:

\begin{array}{l} Q A L Y_{k}^{1} = \sum_{t = 1}^{40} ((H e a l t h u t i l i t y m i l d * n_{k, t}^{1} m i l d + H e a l t h u t i l i t y m o d e r a t e * n_{k, t}^{1} m o d e r a t e \\ + H e a l t h u t i l i t y s e v e r e * n_{k, t}^{1} s e v e r e / (1 + d r / 4)^(t - 1)) \end{array} 13

The formula for QALY^2,3,4, which are the QALY’s for unaddressed sensory impairment, are:

\begin{array}{l} Q A L Y_{k}^{2, 3, 4} = \sum_{t = 1}^{40} (((H e a l t h u t i l i t y m i l d - β) * n_{k, t}^{2, 3, 4} m i l d + (H e a l t h u t i l i t y m o d e r a t e - β) \\ * n_{k, t}^{2, 3, 4} m o d e r a t e + (H e a l t h u t i l i t y s e v e r e - β) * n_{k, t}^{2, 3, 4} s e v e r e) / (1 \\ + d r / 4)^(t - 1)) \end{array} 14

, where β is a loss in health utility from sensory impairment.

Prevalence formulas

P_k(HI = 1) is the notation that is used for the probability that an individual has an unaddressed hearing impairment, and P_k(VI = 1) is the notation used for the probability that an individual has unaddressed vision impairment in the treatment and control arm. Conversely, P_k(HI = 0) and P_k(VI = 0) represent the probabilities that an individual does not have an unaddressed hearing or vision impairment, respectively.

The probabilities for the weights $(W_{k}^{i})$ of the model are the following:

1. No sensory impairment:
$W_{k}^{1} = P_{k} (HI = 0) * P_{k} (VI = 0) 15$
2. Unaddressed hearing impairment
$W_{k}^{2} = P_{k} (HI = 1) * P_{k} (VI = 0) 16$
3. Unaddressed vision impairment
$W_{k}^{3} = P_{k} (HI = 0) * P_{k} (VI = 1) 17$
4. Unaddressed dual impairment
$W_{k}^{4} = P_{k} (HI = 1) * P_{k} (VI = 1) 18$

The probabilities for having or not having a particular unaddressed sensory impairment are described below:

\begin{array}{l} P_{k} (VI = 1) = P (VI) * P (VI which can be addressed by glasses) * P_{k} (has glasses) * P_{k} (uses glasses) \\ + 1 - P (VI which can be addressed by glasses) \end{array} 19

\begin{array}{l} P_{k} (HI = 1) = P (HI) * P (HI which can be addressed by hearing aids) * P_{k} (has hearing aids) \\ * P_{k} (uses hearing aids) + 1 - P (HI which can be addressed by hearing aids) \end{array} 20

\begin{array}{l} P_{k} (VI = 0) = P (VI) * P (VI which can be addressed by glasses) * P_{k} (has glasses) \\ * P_{k} (uses glasses) + 1 - P (VI) \end{array} 21

\begin{array}{l} P_{k} (HI = 0) = P (HI) * P (HI which can be addressed by hearing aids) * (P_{k} (has hearing aids) \\ * P_{k} (uses hearing aids) + 1 - P (HI) \end{array} 22

How the model assumes that all residents who can benefit from glasses and hearing aids will possess them in the intervention arm:

P_{S} (has glasses) = P (VI) * P (VI which can be addressed by glasses) 29

P_{S} (has hearing aids) = P (HI) * P (HI which can be addressed by hearing aids) 30

Considering not all residents will use the sensory aids due to obstacles identified in the literature, scenario analysis will be conducted on the rate of adherence for glasses P_S(uses glasses) and rate of adherence for hearing aids P_S(uses hearing aids) in the treatment arm:

Scenario for 10% increase

P_{S} (uses glasses) = P_{C} (uses glasses) * (100 % + 10 %) 23

P_{S} (uses hearing aids) = P_{C} (uses hearing aids) * (100 % + 10 %) 24

Scenario for 30% increase

P_{S} (uses glasses) = P_{C} (uses glasses) * (100 % + 30 %) 25

P_{S} (uses hearing aids) = P_{C} (uses hearing aids) * (100 % + 30 %) 26

Scenario for 70% increase

P_{S} (uses glasses) = P_{C} (uses glasses) * (100 % + 70 %) 27

P_{S} (uses hearing aids) = P_{C} (uses hearing aids) * (100 % + 70 %) 28

Ethics and consent

Ethical approval and consent were not required.

Data availability

No data are associated with this article.

Faculty Opinions recommended

References

Amieva H, Ouvrard C, Giulioli C, et al.: Self-Reported hearing loss, hearing aids, and cognitive decline in elderly adults: a 25-year study. J Am Geriatr Soc. 2015; 63(10): 2099–2104. PubMed Abstract | Publisher Full Text
Briggs A, Claxton K, Sculpher M: Decision modelling for health economic Evaluation. Oxford University Press. 2006. Reference Source
Cahill S, O’Nolan C, O’ Caheny D, et al.: An Irish National survey of Long-term residential care. 2014. Reference Source
Connelly JP, Boland E, Buckley B, et al.: SENSE-Cog residential care: hearing and vision support for residents with dementia in long-term care in Ireland: protocol for a pilot cluster randomised controlled trial. (Preprint), 2023. Publisher Full Text
Cross H, Armitage CJ, Dawes P, et al.: Capabilities, opportunities and motivations of staff to provide hearing support to long-term care home residents with dementia. Int J Audiol. 2024; 63(7): 519–526. PubMed Abstract | Publisher Full Text | Free Full Text
Drummond MF, Sculpher MJ, Claxton K, et al.: Methods for the Economic evaluation of health care programmes. Oxford University Press, 2015. Reference Source
Health Information and Quality Authority: A guide to Health Technology Assessment at HIQA. 2016. Reference Source
Health Service Executive and National Dementia Office. 2020. Reference Source
Husereau D, Drummond M, Petrou S, et al.: Consolidated Health Economic Evaluation Reporting Standards (CHEERS) 2024: updated reporting guidance for health economic evaluations. Value Health. 2024.
Leroi I, Chauhan N, Hann M, et al.: Sensory health for residents with dementia in care homes in England: a knowledge, attitudes, and practice survey. J Am Med Dir Assoc. 2021; 22(7): 1518–1524.e12. PubMed Abstract | Publisher Full Text
Lin FR, Yaffe K, Xia J, et al.: Hearing loss and cognitive decline in older adults. JAMA Intern Med. 2013; 173(4): 293–299. PubMed Abstract | Publisher Full Text | Free Full Text
Livingston G, Huntley J, Sommerlad A, et al.: Dementia prevention, intervention, and care: 2020 report of the Lancet commission. Lancet. 2020; 396(10248): 413–446. PubMed Abstract | Publisher Full Text | Free Full Text
McGarrigle C, Donoghue O: Impact of hearing loss and hearing aid use on psychosocial health and well-being and healthcare cover in the older population in Ireland. TILDA The Irish Longitudinal Study on Ageing. Trinity College Dublin, 2023. Publisher Full Text
Mukadam N, Anderson R, Knapp M, et al.: Effective interventions for potentially modifiable risk factors for late-onset dementia: a costs and cost-effectiveness modelling study. Lancet Healthy Longev. 2020; 1(1): e13–e20. PubMed Abstract | Publisher Full Text
O’Mahony JF, Coughlan D: The Irish cost-effectiveness threshold: does it support rational rationing or might it lead to unintended harm to Ireland’s health system? Pharmacoeconomics. 2016; 34(1): 5–11. PubMed Abstract | Publisher Full Text
Thorn JC, Davies CF, Brookes ST, et al.: Content of Health Economics Analysis Plans (HEAPs) for trial-based Economic evaluations: expert Delphi consensus survey. Value Health. 2021; 24(4): 539–547. PubMed Abstract | Publisher Full Text | Free Full Text
Yeo BSY, Song HJJMD, Toh EMS, et al.: Association of hearing aids and cochlear implants with cognitive decline and dementia: systematic review and meta-analysis. JAMA Neurol. 2023; 80(2): 134–141. PubMed Abstract | Publisher Full Text | Free Full Text
York Health Economics Consortium: Discount Rate. 2016. Reference Source

Comments on this article Comments (0)

Version 1

VERSION 1 PUBLISHED 31 Oct 2024

Author details Author details

¹ Trinity College Institute of Neuroscience, Trinity College Dublin School of Medicine, Dublin, Leinster, D02 PN40, Ireland

Mikael Äijälä
Roles: Conceptualization, Data Curation, Formal Analysis, Investigation, Methodology, Project Administration, Validation, Visualization, Writing – Original Draft Preparation

Iracema Leroi
Roles: Conceptualization, Funding Acquisition, Project Administration, Resources, Writing – Review & Editing

Dominic Trépel
Roles: Conceptualization, Data Curation, Funding Acquisition, Methodology, Project Administration, Resources, Software, Supervision, Validation, Writing – Review & Editing

Competing interests

No competing interests were disclosed.

Grant information

Health Research Board [DIFA-2020-007]
The study was funded by HRB DIFA (2020) award entitled “SENSE-Cog Care: A feasibility study of hearing and vision support to improve quality of life in a Care home residents with dementia”

The funder did not have a say in the design of the model or the decision to publish this HEAP.

The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Article Versions (1)

version 1

Published: 31 Oct 2024, 7:70

https://doi.org/10.12688/hrbopenres.13948.1

Copyright

© 2024 Äijälä M et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Download

Export To

metrics

VIEWS

245

downloads

17

Citations

0

SEE MORE DETAILS

CITE

how to cite this article

Äijälä M, Leroi I and Trépel D. Cost-effectiveness of hearing and vision support for residents with dementia in long-term care in Ireland: A Health Economic Analysis Plan (HEAP) for a simulation model for the SENSE-Cog Care feasibility trial [version 1; peer review: awaiting peer review]. HRB Open Res 2024, 7:70 (https://doi.org/10.12688/hrbopenres.13948.1)

NOTE: If applicable, it is important to ensure the information in square brackets after the title is included in all citations of this article.

track

receive updates on this article

Track an article to receive email alerts on any updates to this article.

Open Peer Review

Comments on this article Comments (0)

Version 1

VERSION 1 PUBLISHED 31 Oct 2024

Open Peer Review

Reviewer Status

AWAITING PEER REVIEW

Comments on this article

All Comments(0)

Add a comment

Sign up for content alerts

[1] Amieva H, Ouvrard C, Giulioli C, et al.: Self-Reported hearing loss, hearing aids, and cognitive decline in elderly adults: a 25-year study. J Am Geriatr Soc. 2015; 63(10): 2099–2104. PubMed Abstract | Publisher Full Text

[2] Briggs A, Claxton K, Sculpher M: Decision modelling for health economic Evaluation. Oxford University Press. 2006. Reference Source

[3] Cahill S, O’Nolan C, O’ Caheny D, et al.: An Irish National survey of Long-term residential care. 2014. Reference Source

[4] Connelly JP, Boland E, Buckley B, et al.: SENSE-Cog residential care: hearing and vision support for residents with dementia in long-term care in Ireland: protocol for a pilot cluster randomised controlled trial. (Preprint), 2023. Publisher Full Text

[5] Cross H, Armitage CJ, Dawes P, et al.: Capabilities, opportunities and motivations of staff to provide hearing support to long-term care home residents with dementia. Int J Audiol. 2024; 63(7): 519–526. PubMed Abstract | Publisher Full Text | Free Full Text

[6] Drummond MF, Sculpher MJ, Claxton K, et al.: Methods for the Economic evaluation of health care programmes. Oxford University Press, 2015. Reference Source

[7] Health Information and Quality Authority: A guide to Health Technology Assessment at HIQA. 2016. Reference Source

[8] Health Service Executive and National Dementia Office. 2020. Reference Source

[9] Husereau D, Drummond M, Petrou S, et al.: Consolidated Health Economic Evaluation Reporting Standards (CHEERS) 2024: updated reporting guidance for health economic evaluations. Value Health. 2024.

[10] Leroi I, Chauhan N, Hann M, et al.: Sensory health for residents with dementia in care homes in England: a knowledge, attitudes, and practice survey. J Am Med Dir Assoc. 2021; 22(7): 1518–1524.e12. PubMed Abstract | Publisher Full Text

[11] Lin FR, Yaffe K, Xia J, et al.: Hearing loss and cognitive decline in older adults. JAMA Intern Med. 2013; 173(4): 293–299. PubMed Abstract | Publisher Full Text | Free Full Text

[12] Livingston G, Huntley J, Sommerlad A, et al.: Dementia prevention, intervention, and care: 2020 report of the Lancet commission. Lancet. 2020; 396(10248): 413–446. PubMed Abstract | Publisher Full Text | Free Full Text

[13] McGarrigle C, Donoghue O: Impact of hearing loss and hearing aid use on psychosocial health and well-being and healthcare cover in the older population in Ireland. TILDA The Irish Longitudinal Study on Ageing. Trinity College Dublin, 2023. Publisher Full Text

[14] Mukadam N, Anderson R, Knapp M, et al.: Effective interventions for potentially modifiable risk factors for late-onset dementia: a costs and cost-effectiveness modelling study. Lancet Healthy Longev. 2020; 1(1): e13–e20. PubMed Abstract | Publisher Full Text

[15] O’Mahony JF, Coughlan D: The Irish cost-effectiveness threshold: does it support rational rationing or might it lead to unintended harm to Ireland’s health system? Pharmacoeconomics. 2016; 34(1): 5–11. PubMed Abstract | Publisher Full Text

[16] Thorn JC, Davies CF, Brookes ST, et al.: Content of Health Economics Analysis Plans (HEAPs) for trial-based Economic evaluations: expert Delphi consensus survey. Value Health. 2021; 24(4): 539–547. PubMed Abstract | Publisher Full Text | Free Full Text

[17] Yeo BSY, Song HJJMD, Toh EMS, et al.: Association of hearing aids and cochlear implants with cognitive decline and dementia: systematic review and meta-analysis. JAMA Neurol. 2023; 80(2): 134–141. PubMed Abstract | Publisher Full Text | Free Full Text

[18] York Health Economics Consortium: Discount Rate. 2016. Reference Source

Cost-effectiveness of hearing and vision support for residents with dementia in long-term care in Ireland: A Health Economic Analysis Plan (HEAP) for a simulation model for the SENSE-Cog Care feasibility trial

Abstract

Introduction

Methods

Results

Discussion

Keywords

Introduction

Methods

Study design

Markov model

Figure 1. Markov model structure.

Costs

QALY

Prevalence weights of sensory impairment

Figure 2. Flow chart for the prevalence of unaddressed sensory impairment.

Figure 3. Flow chart for unaddressed hearing impairment.

Figure 4. Flow chart for unaddressed vision impairment.

Probability sensitivity analysis

Deterministic sensitivity analysis

Results

Table 1. Deterministic results.

Table 2. Unit costs related to treatment.

Table 3. Per resident weekly costs.

Table 4. The population death rates.

Table 5. Other input items.

Table 6. Relative risk for each category of sensory impairment.

Table 7. Care-as-usual inputs for weights.

Table 8. Intervention outcomes and also inputs for weights.

Table 9. Transition probabilities for the Markov model.

Deterministic results

PSA results

Conclusion

Other relevant information

Formulas

Markov formulas

Cost formulas

QALY formulas

Prevalence formulas

Ethics and consent

Data availability

References

Comments on this article Comments (0)

Open Peer Review

Comments on this article Comments (0)

Open Peer Review

Reviewer Status

Comments on this article

Competing Interests Policy

Stay Updated

Are you a HRB-funded researcher?

Thank you!