A Model to Assess the Impact of Digital Technologies on the Health-related Quality of Life

Objectives Health-related quality of life (HRQoL) is a vital instrument to account for individuals’ well-being in various settings. However, no model of HRQoL allows examining the effect of digital technology on HRQoL. Therefore, we extend an established HRQoL model by adding a digital technology-related construct. We refer to this extension as the TA-HRQoL. Methods We investigate the extended TA-HRQoL model through a survey. In the survey, we exemplify the use of digital technology through a device for self-managing bladder dysfunction. Hence, we explore whether the model extension proposed is valid and how determinants of the HRQoL affect patients with bladder dysfunction. Results The results indicate that the use of digital technology improves the HRQoL. In our exemplary use scenario, the digital technology decreases bladder-related functional impairments and increases well-being and life satisfaction directly. Conclusion Our study may provide evidence for the influence of digital technologies on the HRQoL, thus supporting our model extension. We consider our proposed TA-HRQoL model as valid and as useful to account for the influence of digital technology on an individual’s HRQoL. With the TA-HRQoL model, the impact of a digital technology on an individual’s HRQoL can be assessed.

140 Finally, the questionnaire was composed of reflective measurement models for the constructs 141 Symptoms due to Bladder Dysfunction (SYMP; 9), Functional Impairments due to Bladder 142 Dysfunction (FUNC; e.g., 10 Table 2 shows the exemplary items for the 153 construct TECH.
155 At the start of the survey participant information provided potential respondents with details 156 concerning the goal of the survey, its procedure, and the privacy policy. After agreeing to 157 participate in the survey, the privacy policy and the status of the respondent (patient or assistant 158 of such), participants completed the survey.
159 Sample 160 A prerequisite to participate in our study was suffering from any kind of bladder dysfunction 161 oneself (patient questionnaire) or knowing and supporting someone with bladder dysfunction 211 to report the statistical probability of inappropriately disapproving a true null hypothesis.

213
Insert Figure 1 here.    By offering interventions (e.g., using mobile devices), digital technology can be applied to support facing and improving these chronic conditions and thus help improve individuals' health conditions (2). Examples of such interventions are mental health apps helping patients cope with milder psychological disorders (e.g., deprexis) or diet apps to change eating and moving behaviors toward a healthier lifestyle and ultimately lose weight (e.g., oviva).
However, the impact of such digital technologies is often difficult to measure, particularly regarding intangible outcomes, such as HRQoL. Following this difficulty, technology assessment in health care often does not account for such perspectives yet, although studies demonstrate the influence of technology on the quality of life (3,4). As a result, research calls for a stronger inclusion of these intangible and humanistic outcomes when investigating the effects of digital technologies on individuals (5).
Early research demonstrates how to make use of HRQoL models in technology assessment (4).
However, to the best of our knowledge, we are unaware of any HRQoL model that particularly accounts for the influence of digital technologies in health care. Hence, we posit the following research question: How can the HRQoL-model be extended to assess the influence of digital technologies within inContAlert is an AI-enabled digital technology that allows individuals suffering from incontinence to monitor the filling status of their urinary bladder and proactively give notice when to best empty the bladder. We investigated our collected data by applying partial least squares structural equation modeling (PLS-SEM) to empirically assess and validate the relationships proposed for our structural model (8).

Method Measurements and Procedure
Before conducting our research, the ethical board of the University of Bayreuth authorized the study proposal. To test our TA-HRQoL model, we conducted an online survey assessing all constructs of our model and complementary information such as participants' demographics.
We operationalized the constructs by collecting indicators from pre-validated questionnaires, if available. For constructs that so far do not have pre-validated items developed, we derived items from existing literature on HRQoL generally and specifically on HRQoL of patients with bladder dysfunction. Finally, we selected the indicators according to the best-fit principle and adapted them as well as the declarative information preceding the indicators to the study context (i.e., patients suffering from bladder dysfunction) if necessary 1 .
The initial questionnaire consisted of reflective and formative measurement models (each reflective model was also assessed with a formative one), and single-item measures. As recommended by Hair et al. (8), we conducted a pretest to test them in terms of reliability and validity. None of the 42 pre-test responses indicated suspicious response patterns or issues due to missing data. The results of the reliability and validity analysis led to the deletion of all formative measurements. We analyzed the convergent validity of the reflective measurement models. Thereby, we investigated the outer loadings and trimmed down the questionnaire according to established procedures (8). Following their approach, we retained each indicator with outer loadings of ≥ .700, whereas indicators with outer loadings of < .400 were deleted.
An outer value ≥ .700 indicates that the construct explains half or more of the item's variance, thus, it is widely regarded as a suitable threshold. Items with outer loadings <.400 are considered to have only little explanatory power, thus, we removed them. Indicators with outer 1 The different scales and items had the following specifications: SOCI1-9, TECH1-6 and FORM1 from 1 (strongly disagree) to 5 (strongly agree). FORM3 from 1 (minimal) to 4 (large). DEPR1-4 from 1 (as much as ever) to 5 (not at all). DEPR5-7 and FUNC7-10 from 1 (never) to 5 (always). PHYS1 from 1 (not at all) to 5 (extremely). PHYS2-4 and FORM2 from 1 (very dissatisfied) to 5 (very satisfied). SYMP1 from 1 (Zero -I [She/he] would not have urine leakage) to 5 (more than once a day). SYMP2-3 from 1 (Zero -I [She/he] would not have urine leakage) to 5 (large (clothes/pads are soaked)). SYMP4 from 1 (never) to 4 (many times a day).  At the start of the survey participant information provided potential respondents with details concerning the goal of the survey, its procedure, and the privacy policy. After agreeing to participate in the survey, the privacy policy and the status of the respondent (patient or assistant of such), participants completed the survey.

Sample
A prerequisite to participate in our study was suffering from any kind of bladder dysfunction oneself (patient questionnaire) or knowing and supporting someone with bladder dysfunction (assistant questionnaire). Participants had to be at least 18 years old and have sufficient knowledge of either English or German language. To obtain a diverse and extensive sample we applied convenience sampling. The survey was distributed in a four-month period (18 th November 2020 to 15 th February 2021) in associations, social media groups, forums, and information portals addressing bladder dysfunction or diseases commonly associated it.
To determine the minimum required sample size, we followed

Results
To examine our research question, we extended the TA-HRQoL model by applying PLS-SEM (18) using the statistical software SmartPLS3.

Measurement Models
Prior to the evaluation of the structural model, we assessed the measurement models. We refrained from evaluating the single-item measures in terms of their reliability and validity.
Only five indicators of the reflective measurement models displayed outer loadings smaller than .700. We examined the impact of omitting the items on the composite reliability. Based on the results, we dropped three items. The Table 3 shows the strength of each path coefficient and their reflective significance for each hypothesized relationship in the structural model. Figure 1      combinations, the independent variables have no effect on their dependent variables. The effect sizes show that the relationships between the different constructs are substantial with two relationships even having medium effect sizes and on relationship showing even a large effect size.

Structural Model
In sum, the results demonstrate that the data provides evidence for the applicability of our model in further research.

Discussions and Conclusion
The aim of our study was to assess the influence of digital technology on the HRQoL. To analyze that influence, we propose the TA-HRQoL model, which we evaluated through an Our data revealed significant positive relationships for both hypotheses, thus indicating a substantial influence of digital technologies on HRQoL. To just name one, e. g. patients with diabetes need to constantly check their blood sugar levels to avoid negative consequences of the disease. Digital technologies, such as inContAlert, enable these tracking functions within health care. Thus, we consider the TA-HRQoL to also be applicable to digital technologies used within the contexts of other diseases. Our TA-HRQoL model contributes to the ongoing discussion about the impact of technologies in the health care domain (19). Our study shows that it is important to consider the influence of digital technologies when assessing HRQoL. To the best of our knowledge, no previous study proposed digital technology as an individual construct to explain the dependent variables of HRQoL. Our result -digital technology significantly influences the HRQoL -supports earlier research (e.g., 20,21), which suggested that digital technology can positively affect HRQoL.

Theoretical and Practical Contribution
Thus, we contribute to research on technology assessment in health care and validate existing research in this domain (4). Our study is a starting point to assess the impact of digital technologies on humanistic outcomes, such as HRQoL.
However, it must be considered that certain disparities concerning the availability, knowledge, and use of digital technologies have an influence on how patients use them. These disparities also indirectly affect the patients' HRQoL in the context of the use of digital technology.
Consequently, the consideration of such inequalities among the target group is critical to the design and development of digital technology-based interventions (22). Since our study demonstrated the positive effect of technology on the HRQoL, we recommend governments and health insurers to guarantee access to adequate health technology. Digital technology that succeeds in improving a patient's health status can ultimately reduce the cost of health care by lowering the need for traditional treatments in the future (23).
Also, the TA-HRQoL model sets the scene to assess the medical and economic value of digital technologies in health care. So far, an indicator to evaluate health services or programs is the health-adjusted expectancy of life (HALE). HALE is a measurement that includes life expectancy, mortality, and quality of life (24). More precisely, it is the life expectancy adjusted for HRQoL (25). Our validated TA-HRQoL model could integrate the influence of digital technology into the HALE model. Hence, we posit that digital technologies also have the potential to improve life expectancy.
Thus, we contribute to research on HRQoL as well as on the impact and effects of digital technologies in health care. Research calls to address the missing link between the impact of digital technologies and so-called humanistic outcomes that are intangible results of the use of digital technologies (5). We address this call for future research by demonstrating the potential of digital technologies to improve humanistic outcomes of their use for individuals (5).

Limitations and Future Research
Even though we highlight the theoretical contribution of our work, we acknowledge its limitations. First, our survey design indeed included an introduction of the digital technology, but it was rather brief. Respondents may not have understood all aspects of the technology correctly, although we applied control questions to ensure a basic level of understanding.
Second, expectations of patients regarding the usability of the technology and its integration into daily routines may have differed from experiences they would gain from actual use. Hence, researchers ought to repeat the study once the technology is freely available for use. Third, even though we applied all standard procedures to ensure the validity and reliability of our findings, we cannot entirely rule out empirical biases (8).
These limitations in mind our study provides various fruitful avenues for further research on TA-HRQoL. To overcome the empirical limitations, we suggest the use of a longitudinal study design and multiple methods for data collection (e.g., surveys, experiments, medical records, patient examination). Doing so allows for (dis-)confirming and expanding the insights we obtained based on the expected use of digital technologies. Further, the TA-HRQoL model should also be tested in various study settings and among different patient groups. Research in this area has the potential to unfold an understanding of the relationship between digital technology and a patient's quality of life. Our study thereby expands the body of knowledge on the influence of technology on HRQoL, while the results are valuable for academia and practice, such as for physicians and health care providers, simultaneously. Finally, future research could use our TA-HRQoL model to evaluate the impact of digital technologies on life expectancy, e.g., using HALE as a measurement.

Conclusion
In health care, digital technology offers a wide range of application scenarios. Targeting selfmanagement of chronic disease patients, the use of digital technology aims to increase their