COST-EFFECTIVENESS ANALYSIS OF AN
mHEALTH APPLICATION (SMART4MD)
AND ANALYSIS OF THE EFFECT OF DIALYSIS
TREATMENTS ON LABOR MARKET
OUTCOMES
HEALTH TECHNOLOGY ASSESSMENT OF TWO TREATMENT METHODS
Zartashia Ghani
Blekinge Institute of Technology
Licentiate Dissertation Series No. 2020:07
Department of Health
Health Technology Assessment is an important factor for decision making in the healthcare sector in Sweden. It helps to curtail the rising costs asso-ciated with the healthcare sector and aids in the efficient allocation of scarce public health resourc-es. This thesis investigates the cost-effectiveness and the effectiveness in general of two health tech-nologies, addressing the following research objec-tives: i) assessing the cost-effectiveness of mobile health (mHealth) interventions designed for older adults diagnosed with mild cognitive impairment, and ii) assessing the effect of peritoneal dialysis (PD) treatment on labor market outcomes in comparison with institutional hemodialysis (IHD) treatment in Swedish settings.
Study I and Study II are related to the first research objective. In Study I, we summarized and critically assessed the current evidence on the cost-effec-tiveness of mHealth interventions focusing on old-er adults; we found some evidence supporting the cost-effectiveness of these interventions. In Study II, we conducted a within-trial cost-effectiveness
analysis of the software application Support, Mon-itoring and Reminder Technology for Mild Demen-tia (SMART4MD) from a healthcare perspective for a period of six months. A total of 345 Swedish dyads (MCI patient and informal caregiver) partic-ipated in this study. For a short time period of six months, we found that SMART4MD is not cost-effective for MCI patients (statistically insignif-icant); however, a trend was observed that indi-cated that it might be cost-effective for informal caregivers, although results remained statistically insignificant (p > 0.05).
Study III is related to the second research objec-tive. In Study III, we investigated the effect of PD on labor market outcomes (employment rate, work income, and disability pension) in comparison to IHD. We found that PD is associated with a treat-ment advantage over IHD in terms of increased employment, work income, and reduced disability pension in the Swedish population after controlling for non-random selection into treatment.
2020:07 ISSN: 1650-2140 ISBN: 978-91-7295-412-0 COST -EFFECTIVENESS ANAL Y SIS OF AN m HEAL TH APPLICA TION (SMAR T4MD) AND ANAL Y SIS OF
THE EFFECT OF DIAL
Y
SIS
TREA
TMENTS ON LABOR MARKET OUTCOMES
Zar
tashia Ghani
2020:07
Cost-Effectiveness Analysis of an mHealth
Application (SMART4MD) and Analysis
of the Effect of Dialysis Treatments on
Labor Market Outcomes
Health Technology Assessment of Two Treatment Methods
Blekinge Institute of Technology Licentiate Dissertation Series
No 2020:07
Cost-Effectiveness Analysis of an mHealth
Application (SMART4MD) and Analysis
of the Effect of Dialysis Treatments on
Labor Market Outcomes
Health Technology Assessment of Two Treatment Methods
Zartashia Ghani
Licentiate Dissertation in
Applied Health Technology
2020 Zartashia Ghani
Department of Health
Publisher: Blekinge Institute of Technology
SE-371 79 Karlskrona, Sweden
Printed by Exakta Group, Sweden, 2020
ISBN: 978-91-7295-412-0
ISSN: 1650-2140
urn:nbn:se:bth-20587
“And whosoever saves a life, it is as if they have saved the whole of humankind.” (Quran 5:32)
Acknowledgements
Writing a thesis was a huge task which would not have been possible without the invaluable support and assistance of several people to whom I would like to express my sincere thanks. First, I would like to express my deepest gratitude towards my supervisors Prof. Peter Anderberg, Prof. Johan Sanmartin Berglund, Dr. Johan Jarl, and Prof. Martin Andersson for giving me the opportunity to contribute to this project and for the support, guidance, professional advice, critiques, and personal encouragement to make this thesis possible. I also extend my acknowledgements to the Support, Monitoring and Reminder Technology for Mild Dementia (SMART4MD) project for its financial support.
I would like to thank my colleagues at the Department of Health for the friendly work environment they facilitated, with special thanks to the current and previous heads of the department, Prof. Lisa Skär and Dr. Doris Bohman, as well as to the current and previous coordinators of the department, Jeanine Osbeck and Ingela Silverflod, for providing the best managerial support whenever needed. I would also like to thank the database manager, Ulrika Isaksson, for answering all queries related to the SMART4MD dataset, Annelie Cedergren from Blekinge Region for providing cost data for the SMART4MD project, Dr. Sanjib Saha and Dr. Ana Luiza Dallora for their guidance on writing the literature review, and Cholada Kittipittayakorn, PhD student in the industrial economics department, for valuable discussions during our combined study hours for our self-paced course on economic evaluation. Thanks also goes to my dear friend Sudabe Tavakoli for the good memories we share at BTH.
It is a pleasure to thank our family friends in Sweden and Denmark particularly Irfan Sakhawat and Ayesha Irfan, Asjad Azeem and Aroob Anwar, Muhammad Tayyab Bashir and Omaira Ashraf, Mian Mubashar Ali and Arshi Amjad, and Muhammad Shahbaz and Saamra Manzoor for the nice companies and wonderful times we shared.
My deepest gratitude goes to my family—to my late parents, Abdul Ghani and Munawar Sultana, who always believed in me; to my sister, Arshia Ghani; to my cousin, Baji Samina;
8
Table of Contents
Acknowledgements ... 7 Table of Contents ... 9 Abbreviations ... 10 Abstract ... 11 Publications ... 12 Introduction ... 13 Study aims ... 15 Theoretical Background ... 17Cost effectiveness of mHealth interventions for older adults ... 17
1. mHealth ... 17
2. Gerontechnology ... 18
3. Economic evaluation in healthcare ... 18
Effect of dialysis treatments on labor market outcomes in Sweden ... 20
Dialysis treatments in Sweden ... 20
Summary ... 23
Discussion ... 25
Strengths and Limitations... 28
Future Work ... 31
10
Abbreviations
CBA
Cost-Benefit Analysis
CCA
Cost-Consequence Analysis
CEA
Cost-Effectiveness Analysis
CEAC
Cost-Effectiveness Acceptability Curve
CMA
Cost Minimization Analysis
CUA
Cost-Utility Analysis
HTA
Health Technology Assessment
ICER
Incremental Cost-Effectiveness Ratio
IHD
Institutional Hemodialysis
MCI
Mild Cognitive Impairment
MMSE
Mini-Mental State Examination
PD
Peritoneal Dialysis
QALY
Quality-Adjusted Life Years
QoL-AD
Quality of Life-Alzheimer’s Disease Scale
SMART4MD
Support, Monitoring and Reminder Technology for
Mild Dementia
Abstract
Health Technology Assessment is an important factor for decision making in the healthcare sector in Sweden. It helps to curtail the rising costs associated with the healthcare sector and aids in the efficient allocation of scarce public health resources. This thesis investigates the cost-effectiveness and the effectiveness in general of two health technologies, addressing the following research objectives: i) assessing the cost-effectiveness of mobile health (mHealth) interventions designed for older adults diagnosed with mild cognitive impairment, and ii) assessing the effectiveness of peritoneal dialysis (PD) treatment on labor market outcomes in comparison with institutional hemodialysis (IHD) treatment in Swedish settings.
Study I and Study II are related to the first research objective. In Study I, we summarized and critically assessed the current evidence on the cost-effectiveness of mHealth interventions focusing on older adults; we found some evidence supporting the effectiveness of these interventions. In Study II, we conducted a within-trial cost-effectiveness analysis of the software application Support, Monitoring and Reminder Technology for Mild Dementia (SMART4MD) from a healthcare perspective for a period of six months. A total of 345 Swedish dyads (MCI patient and informal caregiver) participated in this study. For a short time period of six months, we found that SMART4MD is not cost-effective for MCI patients (statistically insignificant); however, a trend was observed that indicated that it might be cost-effective for informal caregivers, although results remained statistically insignificant (p > 0.05).
Study III is related to the second research objective. In Study III, we investigated the effect of PD on labor market outcomes (employment rate, work income, and disability pension) in comparison to IHD. We found that PD is associated with a treatment advantage over IHD in terms of increased employment, work income, and reduced disability pension in the Swedish population after controlling for non-random selection into treatment.
12
Publications
This licentiate thesis is based on the following three research studies:
Study I:
Ghani Z, Jarl J, Sanmartin Berglund J, Andersson M, Anderberg P. The Cost-Effectiveness of Mobile Health (mHealth) Interventions for Older Adults: Systematic Review.
International Journal of Environmental Research and Public Health. 2020;17(15).
Study II:
Ghani Z, Jarl J, Sanmartin Berglund J, Andersson M, Anderberg P. Economic evaluation of
a software application (Support, Monitoring and Reminder Technology for Mild Dementia [SMART4MD]) designed for older adults with mild cognitive impairment. 2020 (In
manuscript).
Study III:
Ghani Z, Rydell H, Jarl J. The Effect of Peritoneal Dialysis on Labor Market Outcomes Compared with Institutional Hemodialysis. Peritoneal Dialysis International. 2019;39(1):59–65. doi:10.3747/pdi.2017.00236
Introduction
Health technologies used in healthcare are diverse in nature and can be divided into several categories, including drugs, medical equipment, medical and surgical procedures, and support systems in the form of clinical laboratory, electronic health record, or telemedicine systems, among others. The technological innovations and advancements of existing technologies over the years have not only improved the healthcare delivery process in many ways but have also raised healthcare costs substantially. Thus, efficient and effective use of healthcare resources requires frequent health technology assessment (HTA) in healthcare (1).
The HTA glossary defines HTA as:
“The systematic evaluation of the properties and effects of a health technology, addressing the direct and intended effects of this technology, as well as its indirect and unintended consequences, and aimed mainly at informing decision making regarding health technologies.”(2)
HTA is a multidisciplinary process that assesses the added value of a new technology as compared with the existing one. The multidisciplinary nature of HTA means that these assessments are based on many factors, including the economic, legal, social, and ethical aspects of new health technologies. Medical equipment, drugs, diagnostic methods, treatment and prevention methods, and rehabilitation are all examples of health technologies. Therefore, HTA provides evidence to assist decision makers in making safe, patient-centered, and cost-effective decisions (3).
The importance of HTA in Sweden is highlighted by the fact that the Swedish Agency for Health Technology Assessment and Assessment of Social Services (SBU) is one of the oldest HTA agencies in the world, established in 1987 (4). From an economic perspective, there are many ways to perform HTA of a technology, including literature reviews to summarize the current state of knowledge and economic evaluation to establish the cost-effectiveness of a technology. HTA also includes cost-effectiveness studies, as it is common
14
an increased burden on informal caregivers; aside from this, dementia disorders negatively impact the quality of life of demented patients. Therefore, a rising concern is finding cost-effective methods of supporting older adults with chronic conditions so that they might live independently and that the burden on informal caregivers might be reduced. mHealth is a possible solution to these problems, as it aims to improve the healthcare delivery process and efficacy of healthcare providers and to reduce healthcare expenditures (6, 7). In addition to that, the use of software applications on smartphones or tablets has been touted as a solution to allow older adults with chronic conditions to live independently. However, little is known about the overall cost-effectiveness of these software applications designed for older adults, particularly with respect to dementia.
The rationale behind studying the effect of dialysis treatments on labor market outcomes was that, in the case of the working-age group, chronic illnesses, such as end-stage renal disease (ESRD), have an even greater impact on society in terms of decreased employment levels, the cost of disability pensions, and loss of work income in addition to a substantial increase in healthcare costs. These substantial direct and indirect costs differ between different treatment modalities. Therefore, it is important to study the effectiveness of different dialysis treatments in order to improve these patients’ participation in the labor force and to conduct cost-effectiveness studies for informed decision making.
Study aims
The overall aim of this thesis was to investigate the HTA of two treatment methods (mHealth intervention for MCI patients and PD vs IHD treatment for ESRD patients).
Research setting (i) was composed of two studies, the individual aims of which are defined as follows:
• The aim of Study I was to perform a systematic literature review (SLR) on the cost-effectiveness of mHealth interventions designed for older adults to present and critically appraise the state-of-the-art evidence and to identify gaps in the research. • The aim of Study II was to conduct cost-effectiveness analysis of a software
application (SMART4MD) designed for older adults diagnosed with MCI. For research setting (ii), Study III was conducted with the following aim:
• The aim of Study III was to examine the impact of PD on labor market outcomes (employment, work income, disability pension) in comparison with institutional dialysis treatment in Sweden while controlling for non-random selection into treatment.
The following chapters will provide the theoretical background and summaries of the research studies, as well as a discussion of the importance and relevance of these results.
Theoretical Background
This chapter presents important concepts related to the cost effectiveness of mHealth interventions for older adults and the effects of dialysis treatments on labor market outcomes in Sweden.
Cost effectiveness of mHealth interventions for older adults
Our studies in the first setting lie beneath the umbrella of applied health technology, which deals with the direct or indirect impacts of technology interventions on an individual’s health or healthcare systems. At Blekinge Institute of Technology, research within this area focuses on how health sciences can be combined with engineering sciences to improve quality of life. Gerontechnology is one of the important areas of applied health technology; it mainly focuses on the use of health technology by older people. The aim of gerontechnology is to facilitate good social participation, better health, and independent living by designing products and services to improve the quality of life of older adults. HTA of these technologies is necessary to help decision making. These concepts are discussed in detail below.1. mHealth
Opportunities to use mobile phones and apps in healthcare increased with the release of the first smartphone in 2007, followed by the introduction of mobile phone apps in 2008 (8). During the third quarter of 2019, about 2.47 and 1.8 million apps were available on Google Play and the Apple App Store, respectively (9). A report on mHealth app use reported that a quarter of the total apps available focus on disease prevention and management of chronic diseases. During 2015, 6% of the most-downloaded apps were related to medication reminders and related information. Moreover, it also reported the increasing interest of
18
For this thesis, we defined mHealth as encompassing any delivery of healthcare and/or dissemination of healthcare information using portable devices possessing an internet connection and/or software applications. Portable devices may include mobile phones, smart phones, tablets, and wireless devices. Furthermore, this definition also covers any mHealth intervention aims to improve the healthcare delivery process (reminders, appointment attendance, disease management, tele-homecare) or efficacy (screening, diagnosis, monitoring, etc.) or to deliver disease-management programs (13, 14).
2. Gerontechnology
Gerontechnology is a multidisciplinary area of research that aims to integrate engineering sciences with other social sciences dealing with ageing. The idea behind gerontechnology is to provide optimal technological solutions for aged individuals within their physical, economic, and social living environment. These aged individuals are those people who are living an active life of their own choosing (15). Therefore, ageing in place is one of the important themes within gerontechnology. However, one of the great challenges associated with aging in place is that older adults are at great risk of developing dementia due to cognitive decline at later stages of their lives. Dementia affects approximately 6% of the EU population aged 60 years or older. This poses difficulties for older adults to live independently while having this condition, which might also substantially increase healthcare burden. Integrated and innovative technical solutions have shown to be useful in facilitating aging in place in the presence of cognitive impairment. Moreover, these user-friendly technologies might be able to delay the onset of dementia (16) and provide support for older adults to age at home in a familiar environment (15). Thus, the use of smart user interfaces, such as smartphones and first, second, and third generation technologies, is expected to become an integral part of healthcare systems.
Besides, older adults generally account for a large share of healthcare utilization. A report by the European Commission suggested the use of technological and digital solutions in providing healthcare to older adults as a way to deal with the rising consumption of healthcare resources by older adults. These technical solutions are expected to increase the efficiency of the delivery of care and can enable older people to stay in their homes longer. The opportunity to produce more personalized solutions for elders may improve their quality of life. The use of new technical solutions in healthcare is expected to increase the quality of care, enhance efficiency, and reduce fiscal pressure on the healthcare system (16).
3. Economic evaluation in healthcare
Economic evaluation can be defined as the process of comparing the costs and effects of two or more alternative programs (17). It includes various steps such as identification, measurement, valuation, and comparison of both the costs and consequences of alternative programs under investigation. An economic evaluation is strictly comparative and is generally undertaken to be the basis for a decision; thus, it has a tendency to influence the decisions made. Through economic evaluation, the procedure of decision making becomes transparent and accountable with regard to the social choices made on the behalf of others.
An economic evaluation can be performed with different perspectives, including societal and healthcare perspectives. The selection of a perspective depends upon the decision makers who commissioned the evaluation to use in their decision making. A healthcare perspective intends to maximize health outcomes within the given healthcare budget. Thus, an analysis from a healthcare perspective includes only those costs that can directly influence the healthcare budget, whereas an analysis from a societal perspective includes both costs related to healthcare and those generated outside the healthcare budget, such as patients’ out-of-pocket expenses and other considerations.
There are five types of full economic evaluations: i) cost minimization analysis (CMA); ii) cost-benefit analysis (CBA); iii) cost-consequence analysis (CCA); iv) cost-effectiveness analysis (CEA); and v) cost-utility analysis (CUA). All are similar in terms of the valuation of costs borne by a program or intervention. However, differences occur in terms of the valuation of health gains received from a program or intervention. The selection of the type of evaluation might depend upon the type of health gains expected and the method used for their valuation and measurement (17).
The usability of mHealth apps has been discussed in previous studies (18); however, there is a lack of information on the cost effectiveness of mHealth apps (10). Thus, besides usability, it is also important to assess the cost effectiveness of mHealth apps to optimally utilize healthcare resources. This can also be explained in terms of opportunity cost. The opportunity cost is the value or health benefits (e.g., QALY gained) of the next best alternative (in terms of technical equipment) that must be given up when another item or piece of technical equipment is adopted. Since resources are scarce relative to need, the use of resources in one way prevents their use in another. Therefore, information on cost effectiveness is important for decision makers to help them make trade-offs between different programs while allocating the healthcare budget.
a) ICER
The incremental cost-effectiveness ratio (ICER) is a ratio of the difference in average costs per patient and average QALYs per patient (Δ𝑐𝑜𝑠𝑡𝑠
Δ𝑄𝐴𝐿𝑌𝑠) between intervention and usual care. b) Cost-effectiveness plane
The cost-effectiveness plane is used in economic evaluations to graphically present the bootstrapped ICERs and point estimates. Figure 1 shows a four-quadrant cost-effectiveness (CE) plane, where the x-axis measures health gain differences and the y-axis measures cost
20
ICERs in the SW quadrant show the possibility of an intervention being less effective and less costly compared to its comparator. The cost effectiveness of an intervention in the SW quadrant depends on a willingness-to-accept threshold; that is, how much savings we are willing to accept to forgo an additional health benefit. While, ICERs in the NE quadrant reflect an intervention that is more effective and more costly than its comparator. The cost-effectiveness of an intervention in the NE quadrant depends on a willingness-to-pay threshold; that is, how much we are willing to pay to receive an additional health benefit (17).
Effect of dialysis treatments on labor market outcomes in
Sweden
Dialysis treatments in Sweden
The proportion of patients who have undergone renal transplant (59.4%) is high in Swedish patients with prevalent end-stage renal disease (ESRD), but 41% are still dependent on dialysis. Almost 77% of the dialysis patients in Sweden use institutional hemodialysis (IHD); meanwhile, peritoneal dialysis (PD) is used by 20% of the patients. Home hemodialysis (HHD) is a marginal treatment in Sweden, used by only 3% of dialysis patients in 2018 (19).
The overall employment rate has been found to be substantially lower among ESRD patients in general (20–22) and among dialysis patients in particular (21). Several reasons for this low employment rate among ESRD patients have been reported in the literature, such as frequent visits to a healthcare facility and long treatment hours in the case of IHD, reluctance of employers to employ persons with a chronic disease, poor physical health, access to social insurance benefits, anxiety, fatigue, and low self-confidence (20, 22–27). Some earlier studies have shown a positive effect of PD treatment on the likelihood of employment
compared with IHD treatment (22, 27), while others found no effect (28). Most prior studies fail to account for possible treatment selection bias, have no follow-up or follow up only for a short period, and lack important information, such as education, employment status before treatment, and income level (21–23, 29).
Treatment selection bias is one of the main challenges associated with the estimation of treatment effects. This problem occurs when individuals who received treatment differ from those who did not received treatment for reasons other than the allocation of treatment. For example, one previous study utilized two models (two-stage probit model vs naïve model) to find the relationship between choice of dialysis treatment and employment level. The first model controlled for causality (endogeneity) between treatment choice and employment level, whereas the second model did not. The results for the two-stage probit model showed a much lower positive effect on employment of PD vs. IHD (46.6% vs. 40.6% in absolute terms and 14.7% in relative terms) compared to the naïve model (51.3% vs. 32.2% in absolute terms and 59.2% in relative terms) (27). This highlights the importance of controlling for the endogenous relationship between employment and treatment choice; that is, those patients who are more likely to work are more likely to choose PD. Indeed, it is likely that other factors, such as level of education and age, affect the treatment choice as well as the treatment outcome. Therefore, it is beneficial to study whether a particular dialysis treatment has an advantage over the other treatment in terms of increase in employment rate and work income and decrease in disability pension while controlling for treatment selection bias.
Propensity score matching
Propensity score matching is a technique used to balance the treatment groups over a number of observed covariates in the absence of a randomized controlled trial. This method deals with the issues of systematic differences found in the baseline characteristics of treatment and control groups in observational studies. In this method, a binary variable of treatment and control group is regressed on baseline covariates (characteristics) to generate the propensity score (conditional probability) for patients to receive a treatment given pre-treatment characteristics. Then the patients from the pre-treatment and control groups are matched based on their propensity score to calculate the average treatment effect. Matching can be done using various methods, including kernel, nearest neighbor, radius, and stratification (30).
Summary
The studies included in this thesis investigate the HTA of mHealth interventions (first research question: Publication I and II) and dialysis treatment (second research question: Publication III).
Study I reviewed 11 articles on mHealth interventions designed for older adults. The
objective of this study was to critically assess and review available empirical evidence on the cost effectiveness of mHealth interventions for older adults. Data were extracted on methods and empirical evidence (costs, effects, incremental cost-effectiveness ratio), and we assessed if this evidence supported the reported findings in terms of cost-effectiveness. The consolidated health economic evaluation reporting standards (CHEERS) checklist was used to assess the reporting quality of the included studies.
It was difficult to compare the results of one study with others because of the differences in methodologies (CEA, CUA, or CBA) and the types of costs included (cost of intervention, direct costs, non-medical costs, etc.). We found that the cost-effectiveness of complex smartphone communication interventions cannot be judged due to the lack of information. Some evidence was found for the cost effectiveness of interventions related to simple text-based communications.
The reasons behind differences in the author’s reported results and our assessment are summarized below:
- None of the included studies made any rigorous justification to reason about the societal valuation of used outcomes (unified Parkinson’s disease rating scale [UPDRS], metabolic equivalent-hour of walking per week, or depression-free days) other than QALY.
- It is important to report 95% confidence interval or similar tests (e.g., cost-effectiveness planes and cost-cost-effectiveness acceptability curve [CEAC]) to establish
24
Study II evaluated the cost-effectiveness of a software application (Support, Monitoring and
Reminder Technology for Mild Dementia [SMART4MD]) combined with usual care (UC) versus UC alone to provide support for patients with mild cognitive impairment (MCI) from a healthcare perspective. Primary data was collected on 345 Swedish dyads (MCI patient as a primary participant and an informal caregiver) from the SMART4MD trial. In the homecare setting, participants were provided with the mHealth software application on a tablet in order to set up reminders to take their medication and for doctoral appointments (n = 173) or UC (n = 172), with a follow-up period of six months. The effectiveness measures for MCI patients were quality-adjusted life years (QALYs), quality of life-Alzheimer’s disease scale (QoL-Ad), and mini-mental state examination (MMSE); QALY and Zarit burden were used for informal caregivers. Results were stated as an incremental cost effectiveness ratio (ICER). Non-parametric bootstrapping was used to assess sampling uncertainty. SMART4MD has higher costs and lower QALYs compared with the control group for MCI patients, which means SMART4MD is dominated by usual care (statistically insignificant). The cost per composite QoL-AD was 384 SEK, and the cost per MMSE for MCI patients was 608 SEK. In the case of informal caregivers, the SMART4MD group dominated the control group, meaning that intervention was less costly and more effective (statistically insignificant). The CEA showed an ICER of 23,982 SEK for the Zarit burden scale, meaning that an additional 23,982 SEK was needed to reduce the burden of the caregiver by one point at six-month follow-up compared to the control group. Short run results suggested that SMART4MD is not significantly better than usual care in the case of MCI patients; however, it might yield benefits for informal caregivers (statistically insignificant).
Study III compared the impact of peritoneal dialysis (PD) and institutional hemodialysis
(IHD), the two most common dialysis modalities, on employment, work income, and disability pension in Sweden. A total of 4,734 patients were included in this study; they were aged 20–60 years, they started treatment in Sweden during the period between 1995 and 2012, and survived the first year of dialysis therapy. Both “intention to treat” and “on treatment” analyses were performed by including transplant patients into the former and censoring them at the date of transplant in the latter analysis. A reduced bias treatment effect of PD vs IHD on labor market outcomes was estimated. In this study, the propensity score matching technique was used to account for the non-random selection into treatment. Peritoneal dialysis was found to be associated with a 4-percentage-point increased probability of employment compared with IHD in the “on treatment” analysis. Also, PD was associated with a reduced disability pension by 6 percentage points, as well as increased work income (EUR 3,477 for employed) compared with IHD during the first year of treatment. The “intention to treat” analysis tended to show greater effects compared with the “on treatment” analysis. The results indicate that PD is associated with a treatment advantage over IHD in terms of increased employment, work income, and reduced disability pension in the Swedish population after controlling for non-random selection into treatment.
Discussion
The main contributions of this research can be divided into two settings. In the first setting, the work contributes to the knowledge by gathering and critically evaluating the current evidence on the cost-effectiveness of mHealth interventions focusing on older adults. This SLR identified a gap in the research of cost-effectiveness studies on mHealth interventions for older adults diagnosed with MCI.
We then performed within-trial economic evaluation of a software application (SMART4MD) designed for older adults diagnosed with MCI. The purpose of the SMART4MD trial was to study whether the use of the mHealth application could be beneficial for participants with MCI and their informal caregivers. Our results showed that SMART4MD has statistically insignificant higher costs and lower QALYs compared to the control group (control group dominated SMART4MD group) for MCI patients. However, in the case of informal caregivers, the SMART4MD group dominated the control group, meaning that intervention was less costly and more effective (statistically insignificant). Due to insignificant results, these findings suggest that the SMART4MD intervention is neither better nor worse compared to usual care alone for MCI patients. It is important to state that the time horizon of Study II was six months, and as such, it might not fully account for those benefits that are spread over more extended periods. This may imply an under-estimation of the cost-effectiveness of the SMART4MD intervention.
Some of the issues identified in Study I on cost-effectiveness of mHealth interventions are discussed below:
First, there is no consensus on an acceptable willingness-to-pay threshold to establish the cost-effectiveness of an intervention in the literature. For instance, a threshold range of £20,000–£30,000/QALY gained is used in the UK (31), while American (32) and Australian (33) researchers used the amount of 50,000/QALY gained in their respective currencies. Despite the lack of consensus on a precise value, it is still easier to establish the cost-effectiveness for QALY against the well-established threshold levels. However, the lack of
26
Second, it is important to find out whether the alternatives used in an economic evaluation differ in terms of costs and effects before considering cost-effectiveness. However, confidence intervals or similar tests were not reported in many studies (34–38). Moreover, CEAC was reported in few studies to deal with uncertainties around costs and effects (39– 41). Including CEAC in economic evaluations can be used as an alternative to produce confidence intervals around ICER. However, there is no consensus among researchers as to which value a CEAC should be referred to consider an intervention cost-effective. It is also important to note that decisions regarding implementation of an intervention should not be entirely based on findings from CEAC, but rather considered in the light of the uncertainty around the estimated cost-effectiveness ratio (42).
Third, it is vital to perform sensitivity analyses to test the robustness of results concerning changes in assumptions and the values of the input variables. However, only two studies (40, 41) provided detailed justifications for their choice of parameters or ranges of parameter values in the analysis. Furthermore, four studies did not perform any sensitivity analysis (34, 37, 38, 43).
Study I highlighted the fact that none of the identified studies discussed costs and benefits related to informal caregivers, who are considered an important part of those interventions directed towards older adults in the homecare setting. We have included informal caregivers in Study II to fill this gap in the literature.
In the second setting, the work contributes to the knowledge by studying the effect of two dialysis treatments (PD vs IHD) on labor market outcomes. Our findings suggest that PD treatment is associated with a higher employment rate (61% vs 57%), work income (EUR18720 vs EUR15242 for employed), and reduced disability pension (40% vs 46%) in working-age patients in comparison to IHD.
Hirth et al. found higher estimated employment in PD compared to our findings in Study III (6 vs 4 percentage points). We may have found a lower estimated effect of the treatment in Study III because we controlled for other factors, such as gender, age, and education, in addition to employment status before start of dialysis. On the other hand, Hirth et al. controlled only for the endogenous relationship between employment and treatment modalities (27).
In Greece, 73% of IHD patients are reported to be on disability pension (24). A Dutch study found that the ratio of patients who were getting disability insurance benefits increased from 35%, for both IHD and PD, to 49% for IHD and 47% for PD patients after starting treatment (28). In comparison to these studies, Study III finds lower rates of disability pension in both treatment groups (IHD = 45% and PD = 40%) after controlling for treatment selection. However, PD is associated with a lower rate of disability pension (full- or part-time)
compared to IHD. Since a patient can be on disability pension while working part-time, they might be employed at the same time. The current analysis thus only captures changes between any vs. no disability pension, and additional studies are required to investigate the full effect on disability pension (i.e., also potential reduction in level of disability pension). The implication of these results with regard to promoting PD largely depends on goodness of fit. This approach is limited to only those variables included in the treatment selection regression despite the fact that the balancing property and the map of overlap shows that our
selection equation was well balanced. It is possible that treatment choice might be influenced by other, unobserved factors. In the literature, treatment choice has been shown to be associated with many other factors, such as general health status, patient awareness of the availability of different treatment modalities, the number of patients at a particular facility center, and distance from patient’s home to facility center (29, 44, 45). Lacking this information in Study III, the results should be interpreted with caution and rather as estimates that reduce bias due to observed selection into treatment. However, we have conducted several sensitivity analyses to investigate the risk of unobserved factors influencing the results.
The selection equation can be interpreted in terms of factors associated with choice of a certain type of treatment. These estimates suggest that patients who are highly educated, married, and employed before starting dialysis treatment are more likely to have PD compared to IHD. These findings are in line with Chanouzas et al., except for the factor of
higher education (46). We also find a positive association between being a woman and receiving PD treatment, whereas other studies found no association between gender and treatment choice (46, 47). Study III shows that young patients have a higher probability of getting PD, which is in line with previous univariate analyses (46, 47). However, the results of a multivariate analysis performed by Keating et al. showed no effect of age (47). Both settings highlight the importance of HTA in healthcare in that it provides systemic evidence to aid informed decision making.
28
Strengths and Limitations
Regarding the first setting, the cost effectiveness of mHealth intervention was based on the pragmatic randomized controlled trial, which is considered the best source of data for economic evaluations (48) (Study II).
A few limitations related to the first setting are listed below:
• The findings of the SLR are based on the studies that were identified through the search strategy used for this review. Our search strategy may have missed important studies, although we find this unlikely (Study I).
• In the SLR, the assessment of the quality of the studies using the CHEERS list is subjective and may result in disagreement with each study’s scores (Study I). • The economic evaluation of SMART4MD was conducted over a short follow-up
period of six months. Further effects of this intervention might become visible beyond the scope of the SMART4MD trial due to the nature of this intervention (Study II).
• Although the importance of using a societal perspective was highlighted in Study I, we performed the economic evaluation in Study II using the healthcare perspective due to the lack of information on out-of-pocket expenses and other costs related to the societal perspective.
Regarding the second setting, the strength of Study III is that we reduced the estimation bias caused by non-random selection by balancing the covariates over treatment groups. Furthermore, the data set covers the entire Swedish dialysis population, resulting in a relatively large sample.
Conclusions
• In Study I, no evidence of cost effectiveness was found for interventions related to complex smartphone communication, while some evidence of cost effectiveness was found for simple, text-based communication-related interventions. Comprehensive economic evaluations of mHealth interventions designed for older adults over a more extended time period may allow one to draw more robust conclusions.
• The results of Study II suggest that the SMART4MD intervention might not be cost effective for older adults with MCI in the short run; however, it might help to reduce the burden on the informal caregiver.
• A dialysis method that helps patients to become socially and professionally active will have a positive impact on their quality of life as well as decrease the burden on society (49). Despite the fact that the national insurance policy covers dialysis treatment in Sweden and patients also receive sickness benefits, gainful employment is likely to improve the quality of life of these patients. It is therefore important to select a dialysis method that helps these patients to engage in job-related activities to the extent allowed by their health condition. Based on the findings in Study III, PD is expected to achieve this to a greater extent than IHD, being associated with higher employment, work income, and reduced disability pension.
Future Work
Regarding the first setting, the economic evaluation conducted in Study II was based on six months of data from the Blekinge site, whereas the total duration of the SMART4MD trial was 18 months. Therefore, future work will primarily focus on including full data into our analysis to complement Study II and conducting economic evaluations on the other sites of the SMART4MD project.
Regarding the second setting, it could be promising to investigate the effect of PD on healthcare and pharmaceutical cost compared to IHD and to use this future study along with Study III to conduct an economic evaluation to determine the cost effectiveness of PD treatment in comparison to IHD.
32
References
1. Goodman CS. HTA 101: Introduction to Health Technology Assessment: National Information Center on Health Services Research and Health Care Technology (NICHSR) National Library of Medicine; [updated July 22, 2019; cited 2020 20200709]. Available from: https://www.nlm.nih.gov/nichsr/hta101/ta10103.html. 2. Glossary H. Health Technology Assessment (HTA): HTAi, INAHTA; [Available
from: http://htaglossary.net/health+technology+assessment+%28HTA%29. 3. European Commission. Health Technology Assessment.
4. Jonsson E. History of health technology assessment in Sweden. International Journal
of Technology Assessment in Health Care. 2009;25 Suppl 1:42–52.
5. World Health Statistics 2016: World Health Organization (WHO); [Available from:
http://www.who.int/gho/publications/world_health_statistics/2016/en/.
6. Khosravi P, Ghapanchi AH. Investigating the effectiveness of technologies applied to assist seniors: A systematic literature review. International Journal of Medical
Informatics. 2016;85(1):17–26.
7. Free C, Phillips G, Watson L, Galli L, Felix L, Edwards P, et al. The effectiveness of mobile-health technologies to improve health care service delivery processes: A systematic review and meta-analysis. PLoS Medicine. 2013;10(1):e1001363.
8. Lupton D. Digital health: Critical and cross-disciplinary perspectives. 2017;1–170. 9. Statista. Number of apps available in leading app stores 2019. 2019.
10. IMS. Patient Adoption of mHealth--Use, Evidence and Remaining Barriers to Mainstream Acceptance. 2015.
11. Khosravi P, Ghapanchi AH. Investigating the effectiveness of technologies applied to assist seniors: A systematic literature review. International journal of medical informatics. 2016;85(1):17–26.
12. World Health Organization. mHealth: new horizons for health through mobile technologies. 2011.
13. mHealth: IMT - Learn, Connect, Innovate; [Available from:
https://innovatemedtec.com/digital-health/mhealth.
14. Free C, Phillips G, Felix L, Galli L, Patel V, Edwards P. The effectiveness of M-health technologies for improving health and health services: A systematic review protocol. BMC research notes. 2010;3:250.
15. Kwon S. Gerontechnology: Research, Practice, and Principles in the Field of
Technology and Aging. Springer Publishing Company. 2016.
17. Drummond MF, Sculpher MJ, Claxton K, Stoddart GL, Torrance GW. Methods for the
Economic Evaluation of Health Care Programmes. Oxford University Press. 2015.
18. Torkamani M, McDonald L, Aguayo IS, Kanios C, Katsanou M-N, Madeley L, et al. A randomized controlled pilot study to evaluate a technology platform for the assisted living of people with dementia and their carers. Journal of Alzheimer's Disease. 2014;41(2):515–23.
19. SNR. Svenskt njurregister årsrapport 2019. Sweden: SNR; 2019.
20. Molsted S, Aadahl M, Schou L, Eidemak I. Self-rated health and employment status in chronic haemodialysis patients. Scand J Urol Nephrol. 2004;38(2):174–8.
21. Law MC, Chow KM, Fung JSF, Szeto CC, Li PKT. Employment status in peritoneal-dialysis patients. Hong Kong Journal of Nephrology. 2016;18:11–4.
22. Helantera I, Haapio M, Koskinen P, Gronhagen-Riska C, Finne P. Employment of patients receiving maintenance dialysis and after kidney transplant: A cross-sectional study from Finland. Am J Kidney Dis. 2012;59(5):700–6.
23. Murray PD, Dobbels F, Lonsdale DC, Harden PN. Impact of end-stage kidney disease on academic achievement and employment in young adults: A mixed methods study.
The Journal of adolescent health: official publication of the Society for Adolescent Medicine. 2014;55(4):505–12.
24. Kaitelidou D, Maniadakis N, Liaropouls L, Ziroyanis P, Theodorou M, Siskou O. Implications of hemodialysis treatment on employment patterns and everyday life of patients. Dialysis & Transplantation. 2005;34(3):138–47.
25. Godoy MR, Balbinotto Neto G, Ribeiro EP. Earnings and chronic renal disease.
Transplant Proc. 2007;39(2):378–80.
26. Kutner NG, Zhang R, Huang Y, Johansen KL. Depressed mood, usual activity level, and continued employment after starting dialysis. Clin J Am Soc Nephrol. 2010;5(11):2040–5.
27. Hirth RA, Chernew ME, Turenne MN, Pauly MV, Orzol SM, Held PJ. Chronic illness, treatment choice and workforce participation. International Journal of Health Care
Finance and Economics. 2003;3(3):167–81.
28. van Manen JG, Korevaar JC, Dekker FW, Reuselaars MC, Boeschoten EW, Krediet RT. Changes in employment status in end-stage renal disease patients during their first year of dialysis. Perit Dial Int. 2001;21(6):595–601.
34
33. Taylor CJ, Jan S. Economic evaluations of medicines. Australian Prescriber [Internet]. 2017 2019-05-16; 40(2):[76–8 pp.]. Available from:
https://doi.org/10.18773/austprescr.2017.014.
34. Cubo EM, Mariscal N, Solano B, Becerra V, Armesto D, Calvo S, Arribas J, Seco J, Martinez A, Zorrilla L, Heldman D. Prospective study on cost-effectiveness of home-based motor assessment in Parkinson’s disease. Journal of Telemedicine and Telecare.
2016;23(2):328–38.
35. Choi Yoo SJ, Nyman JA, Cheville AL, Kroenke K. Cost effectiveness of telecare management for pain and depression in patients with cancer: Results from a randomized trial. General Hospital Psychiatry. 2014;36(6):599–606.
36. Maddison R, Pfaeffli L, Whittaker R, Stewart R, Kerr A, Jiang Y, Kira G, Leung W, Dalleck L, Carter K, Rawstorn J. A mobile phone intervention increases physical activity in people with cardiovascular disease: Results from the HEART randomized controlled trial. European Journal of Preventive Cardiology. 2015;22(6):701–9. 37. Whittaker F, Wade V. The costs and benefits of technology-enabled, home-based
cardiac rehabilitation measured in a randomised controlled trial. Journal of
telemedicine and telecare. 2014;20(7):419–22.
38. Barnett TE, Chumbler NR, Vogel WB, Beyth RJ, Ryan P, Figueroa S. The cost-utility of a care coordination/home telehealth programme for veterans with diabetes. Journal
of Telemedicine and Telecare. 2007;13(6):318–21.
39. Cui Y, Doupe M, Katz A, Nyhof P, Forget EL. Economic evaluation of Manitoba Health Lines in the management of congestive heart failure. Healthcare Policy /
Politiques de Sante. 2013;9(2):36–50.
40. Udsen FW, Lilholt PH, Hejlesen O, Ehlers L. Cost-effectiveness of telehealthcare to patients with chronic obstructive pulmonary disease: Results from the Danish 'TeleCare North' cluster-randomised trial. BMJ Open. 2017;7(5).
41. Stoddart A, van der Pol M, Pinnock H, Hanley J, McCloughan L, Todd A, Krishan A, McKinstry B. Telemonitoring for chronic obstructive pulmonary disease: A cost and cost-utility analysis of a randomised controlled trial. Journal of Telemedicine and
Telecare. 2015;21(2):108–18.
42. Fenwick E, Byford S. A guide to cost-effectiveness acceptability curves. The British
Journal of Psychiatry: The Journal of Mental Science. 2005;187:106–8.
43. Katalenich B, Lizheng S, Liu S, Shao H, McDuffie R, Carpio G, Thethi T, Fonseca V. Evaluation of a Remote Monitoring System for Diabetes Control. Clinical
Therapeutics. 2015;37(6):1216–25.
44. Prakash S, Coffin R, Schold J, Lewis SA, Gunzler D, Stark S, et al. Travel distance and home dialysis rates in the United States. Peritoneal Dialysis International. 2014;34(1):24–32.
45. Pernilla Olsson JO. En studie av ekonomiska och sociala faktorer för olika behandlingsmetoder inom svensk dialyssjukvård. 2015.
46. Chanouzas D, Ng KP, Fallouh B, Baharani J. What influences patient choice of treatment modality at the pre-dialysis stage? Nephrology Dialysis Transplantation. 2012;27(4):1542–7.
47. Keating PT, Walsh M, Ribic CM, Brimble KS. The impact of patient preference on dialysis modality and hemodialysis vascular access. BMC Nephrology. 2014;15(1):38. 48. Ramsey SD, Willke RJ, Glick H, Reed SD, Augustovski F, Jonsson B, et al.
Cost-effectiveness analysis alongside clinical trials II-An ISPOR Good Research Practices Task Force report. Value in Health: The Journal of the International Society for
Pharmacoeconomics and Outcomes Research. 2015;18(2):161–72.
49. Takaki J, Yano E. The relationship between coping with stress and employment in patients receiving maintenance hemodialysis. J Occup Health. 2006;48(4):276–83.
COST-EFFECTIVENESS ANALYSIS OF AN
mHEALTH APPLICATION (SMART4MD)
AND ANALYSIS OF THE EFFECT OF DIALYSIS
TREATMENTS ON LABOR MARKET
OUTCOMES
HEALTH TECHNOLOGY ASSESSMENT OF TWO TREATMENT METHODS
Zartashia Ghani
Blekinge Institute of Technology
Licentiate Dissertation Series No. 2020:07
Department of Health
Health Technology Assessment is an important factor for decision making in the healthcare sector in Sweden. It helps to curtail the rising costs asso-ciated with the healthcare sector and aids in the efficient allocation of scarce public health resourc-es. This thesis investigates the cost-effectiveness and the effectiveness in general of two health tech-nologies, addressing the following research objec-tives: i) assessing the cost-effectiveness of mobile health (mHealth) interventions designed for older adults diagnosed with mild cognitive impairment, and ii) assessing the effect of peritoneal dialysis (PD) treatment on labor market outcomes in comparison with institutional hemodialysis (IHD) treatment in Swedish settings.
Study I and Study II are related to the first research objective. In Study I, we summarized and critically assessed the current evidence on the cost-effec-tiveness of mHealth interventions focusing on old-er adults; we found some evidence supporting the cost-effectiveness of these interventions. In Study II, we conducted a within-trial cost-effectiveness
analysis of the software application Support, Mon-itoring and Reminder Technology for Mild Demen-tia (SMART4MD) from a healthcare perspective for a period of six months. A total of 345 Swedish dyads (MCI patient and informal caregiver) partic-ipated in this study. For a short time period of six months, we found that SMART4MD is not cost-effective for MCI patients (statistically insignif-icant); however, a trend was observed that indi-cated that it might be cost-effective for informal caregivers, although results remained statistically insignificant (p > 0.05).
Study III is related to the second research objec-tive. In Study III, we investigated the effect of PD on labor market outcomes (employment rate, work income, and disability pension) in comparison to IHD. We found that PD is associated with a treat-ment advantage over IHD in terms of increased employment, work income, and reduced disability pension in the Swedish population after controlling for non-random selection into treatment.
2020:07 ISSN: 1650-2140 ISBN: 978-91-7295-412-0 COST -EFFECTIVENESS ANAL Y SIS OF AN m HEAL TH APPLICA TION (SMAR T4MD) AND ANAL Y SIS OF
THE EFFECT OF DIAL
Y
SIS
TREA
TMENTS ON LABOR MARKET OUTCOMES
Zar
tashia Ghani
2020:07
