Why wait? (Cost)effectiveness of an online transdiagnostic therapy for patients and their loved ones

Below, parameters (A1, A2, A3, A4, A5) for assessing (cost) effectiveness will

be described. While preliminary effectiveness of the eHealth PPI will already

be rated in the RSCD, (cost) effectiveness will be determined rigorously in the

RCT.Please note, that although we see (cost) effectiveness as our primary study

objective and parameters, only the first parameter (positive mental health) is

the primary outcome variable (baseline to 1 year follow-up post outpatient

psychological treatment). The sample size calculation is based on this primary

outcome variable.



A1. Positive mental health

In order to examine the primary objective, the frequency of positive mental

health symptoms will be defined as the main study parameter. This parameter

will be measured, using the Mental Health Continuum Short Form (MHC-SF), which

is a clinically validated instrument aimed at assessing positive mental health

across a range of emotional, psychological and social domains (Lamers et al.,

2011).



A2. Self-reported psychological symptoms

Self-reported psychological symptoms will be defined as the second parameter

for examining the primary objective. This parameter will be operationalized,

using the validated Brief Symptom Inventory (BSI), measuring the number of

self-reported psychological symptoms across a range of domains that are

clinically relevant in adults and adolescents (De Beurs & Zitman, 2005).



A3. Patient dropout

A parameter for patient dropout will be measured in all three treatment

conditions, since patient dropout is an important clinical marker of treatment

success (Arntz et al., 2023).



A4. Quality of life

First, general and objective quality of life is often considered an important

parameter for health gains in the field of health economics. This parameter

will be employed, using quality-adjusted Life Years (QALYs). In general, QALYs

can be estimated, using utility scores from the EQ-5D-5L questionnaire (Feng et

al., 2021; van Dongen et al., 2021), which is a prominent instrument in the

field of cost effectiveness. However, the clinical meaning of generic EQ-5D-5L

utility scores in patients with mental complaints has been criticized in the

past (Pietersma et al., 2013). In order to address these concerns, adjusted

utility scores (for computing quality of life in patients with mental

complaints) will be measured as well. This will be done, using the recently

developed mental health quality-of-life questionnaire (Van Krugten et al.,

2021) and the ICE-CAP-A questionnaire (Al-Janabi et al., 2012) for assessing

positive capabilities in adults with accompanying Dutch tariffs (Rohrbach,

Dingemans, Groothuis-Oudshoorn, et al., 2022).



A5. Societal costs

Second, a parameter for societal costs will be measured, including cost

categories within the Dutch healthcare system and beyond. In general, 1-month

consumption of health services in psychiatric patients will be measured, using

the Tic-P Midi (Timman et al., 2015). In order to obtain the corresponding

healthcare costs, consumption of health service use will be multiplied with the

corresponding reference prices, using the ZIN guidelines (ZIN, 2024a, 2024b).

Costs beyond the healthcare sector mainly include productivity losses in the

workspace due to absenteeism (i.e. not being able to work) and presenteeism

(i.e. reduced productivity at work due to health complaints). In order to

measure these productivity losses within our research population, the

Productivity Cost Questionnaire (PCQ) will be used (Bouwmans et al., 2015).

Using a 1-month recall period in the PCQ, participants need to report their

hours of absence (i.e. absenteeism) and the additional hours they needed to

finish work due to health complaints (i.e. presenteeism). Estimated healthcare

costs will be extrapolated to an extended time horizon that is not covered by

the recall period. In chapter 8, we will elaborate on all statistical

techniques to quantify the aforementioned cost categories and utility scores.

Below, parameters (B1, B2, B3, B4) will be described regarding the added

clinical effects of including loved ones. Loved ones are included in both the

RSCD and RCT.



B1. Positive mental health

B2. Self-reported psychological symptoms

B3. Dropout

B4. Loved ones assessing patient recovery



Similar to the clinical parameters mentioned in (A), positive mental health,

self-reported psychological symptoms and dropout (B1, B2, B3, B4) will be

measured among patients and their loved ones in the second arm (eHealth PPI for

patients and loved ones). While dropout will be measured statistically in this

treatment arm, positive mental health and self-reported psychological symptoms

will be measured, using the MHC-SF among loved ones. In this arm, loved ones

will receive a different set of questionnaires to rate aspects of positive

mental health in their own life (scenario 1) and to evaluate the psychological

well-being of the patient they are related to (scenario 2).

In the former scenario, loved ones will receive distinct questionnaires to rate

their own beliefs regarding self-compassion, savouring and optimism. In light

of this purpose, the Self Compassion Scale - Short Form (SCS-SF), Savouring

Beliefs Inventory (SBI) and Life Orientation Test - Revised (LOT-R) will be

used, respectively (Babenko & Guo, 2019; Ford et al., 2017; Hinz et al., 2017).

In the latter scenario, loved ones will serve as a proxy for assessing a

patient*s psychological well-being, which may be considered an important marker

of treatment success. Loved ones will conduct this proxy rating for the patient

they are related to, using the validated Patient Health Questionnaire (PHQ-9).

Although the PHQ-9 is often used for self-administration, it can also be used

as a proxy instrument to rate well-being and depressive symptoms in other

people (Rooney et al., 2013). Patients and loved ones are allowed to interact

while moving through the eHealth PPI, which may enhance protective resources

and may positively affect dropout rates. Lastly, relational satisfaction will

be measured among patients and their loved ones in the second arm, using the

Relational Satisfaction Scale (RSS) (Anderson et al., 2001).



(C). Parameters for the third objective (moderation effects)

Below, parameters (C1, C2, C3, C4, C5) for assessing moderation effects will be

described. These effects will be assessed in both the RSCD and RCT.



C1. Demographics (age, gender, socioeconomic status)

C2. Clinical profiles (diagnosis, symptom severity, comorbidity)

C3. Type of received treatment at Lionarons GGZ

C4. Treatment expectations

C5. Engagement



To thoroughly explore the moderating effects on our eHealth PPI, a

comprehensive understanding of moderating parameters (C1, C2, C3, C4, C5) is

essential. The objective is to understand how different moderating parameters *

such as diagnosis, age, gender, type of received mental health care, and

treatment expectations*affect the relationship between the three intervention

arms and the primary outcomes mentioned in A (i.e. positive mental health and

cost-effectiveness). Socioeconomic status will be determined, using information

on income, labour type and education.



(D). Parameters for the fourth objective (modelling of personalized treatment):

By examining which patient characteristics are most likely associated with

positive treatment effects and reduced dropout, personalization of the eHealth

PPI will be possible. This procedure may inform beneficial adjustments

regarding the eHealth PPI aimed at enhancing its effectiveness for individual

users. The abovementioned parameters on clinical characteristics and

demographics will be used again to assess implications for personalized eHealth

PPI (D1, D2, D3, D4, D5). The efficacy of the eHealth PPI will always depend on

certain variables, which holds true for most psychological treatments. The PPI

is transdiagnostic and can be applied to various mental health problems, as it

focusses on enhancing general protective factors that can be helpful for

everyone, instead of decreasing disorder specific risk factors. We agree that

the effectiveness of the PPI may differ between subgroups (e.g., based on

demographic factors, type of mental health symptoms), and will explore for whom

the PPI is most beneficial. This analysis will only take place in the RCT study.

D1. Demographics (age, gender, socioeconomic status)

D2. Clinical profiles (diagnosis, symptom severity, comorbidity)

D3. Type of received treatment at Lionarons GGZ

D4. Treatment expectations

D5. Engagement



To facilitate personalization of the eHealth PPI, an exploration of potentially

different treatment effects for different levels of the variables mentioned

above (D1, D2, D3, D4, D5) will be performed. This will be done, using

multilevel modelling techniques. These techniques allow for modelling of

treatment effects based on individual characteristics. This analytical process

involves a detailed examination of the variables mentioned above. In general,

all modeling analyses aim to explore whether and how the efficacy of the

eHealth PPI depends on demographic factors (such as age, gender, and

socioeconomic status), clinical profiles (including diagnosis, symptom

severity, and comorbid conditions), treatment expectations and engagement with

the eHealth platform.



(E). Parameters for the fifth objective (process evaluation)

Below, the parameters used in our process evaluation (E1, E2, E3, E4, E5) will

be explained. Process evaluation is especially relevant in the RSCD. To a lower

extent, process evaluation will also take place in the RCT.



E1. Acceptability

E2. Feasibility

E3. Adherence

E4. Engagement

E5. Accessibility



The implementation of the transdiagnostic eHealth PPI (in the RSCD and

RCT) necessitates a detailed evaluation of several key process parameters (E1,

E2, E3, E4, E5) among patient participants. Prior to organizing the RSCD, two

stakeholder meetings will take place aimed at rating the eHealth PPI from a

patient perspective, based on the mentioned process parameters. These

parameters are crucial for understanding the intervention's practicality, user

experience, and overall impact. Feasibility, which will be operationalized

through specific items in process questionnaires like the System Usability

Scale (SUS), assesses the practicality of applying and accessing the

intervention. Acceptability, gauged through scales measuring user satisfaction

within process questionnaires, evaluates how the intervention is received by

its intended audience. Usability, a critical parameter that can be directly

measured by the SUS, focuses on the ease with which users can navigate and

engage with the digital platform. Adherence measures the extent to which

participants follow the intervention guidelines. Engagement, assessable through

analytics of user interactions recorded in the eHealth platform and

supplemented by questionnaire data, examines the depth of participant

interaction with the intervention's content. Lastly, accessibility, crucial for

understanding the intervention's reach, can be assessed through questions

related to participants' ease of access based on their technological literacy

and resources. In the scheduled RCT, the mentioned SUS will be used to capture

relevant dimensions of user experience regarding the eHealth PPI.