The longitudinal association between patient empowerment and patient-reported outcomes: What is the direction of effect?

Mariela Acuña Mora; Carina Sparud-Lundin; Eva Fernlund; Shalan Fadl; Kazamia Kalliopi; Annika Rydberg; Åsa Burström; Katarina Hanseus; Philip Moons; Ewa-Lena Bratt

doi:10.1371/journal.pone.0277267

. 2022 Nov 10;17(11):e0277267. doi: 10.1371/journal.pone.0277267

The longitudinal association between patient empowerment and patient-reported outcomes: What is the direction of effect?

Mariela Acuña Mora ^1,², Carina Sparud-Lundin ¹, Eva Fernlund ^3,⁴, Shalan Fadl ⁵, Kazamia Kalliopi ⁶, Annika Rydberg ⁷, Åsa Burström ⁸, Katarina Hanseus ⁹, Philip Moons ^1,^10,¹¹, Ewa-Lena Bratt ^1,^12,^*

Editor: James Mockridge¹³

¹Institute of Health and Care Sciences, University of Gothenburg, Gothenburg, Sweden

²Faculty of Caring Science, Work Life and Social Welfare, University of Borås, Borås, Sweden

³Division of Pediatrics, Department of Clinical and Experimental Medicine, Linköping University, Linköping, Sweden

⁴Crown Princess Victoria Children’s Hospital, Linköping University Hospital, Linköping, Sweden

⁵Department of Children and Young Adults, University Hospital Örebro, Örebro, Sweden

⁶Department of Women’s and Children’s Health, Karolinska Institute, Stockholm, Sweden

⁷Department of Clinical Sciences, Umeå University, Umeå, Sweden

⁸Karolinska Institute, Department of Neurobiology, Care Sciences and Society, Stockholm, Sweden

⁹Children’s Heart Center, Skåne University Hospital Lund, Lund, Sweden

¹⁰Department of Public Health and Primary Care, KU Leuven, Leuven, Belgium

¹¹Department of Paediatrics and Child Health, University of Cape Town, Cape Town, South Africa

¹²Department of Pediatric Cardiology, The Queen Silvia Children’s Hospital, Gothenburg, Sweden

¹³PLOS: Public Library of Science, UNITED KINGDOM

Competing Interests: The authors have declared that no competing interests exist.

^✉

* E-mail: ewa-lena.bratt@gu.se

Roles

Mariela Acuña Mora: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Resources, Software, Visualization, Writing – original draft

Carina Sparud-Lundin: Conceptualization, Investigation, Methodology, Project administration, Supervision, Writing – review & editing

Eva Fernlund: Project administration, Resources, Writing – review & editing

Shalan Fadl: Project administration, Resources, Writing – review & editing

Kazamia Kalliopi: Project administration, Resources, Writing – review & editing

Annika Rydberg: Project administration, Resources, Writing – review & editing

Åsa Burström: Project administration, Resources, Writing – review & editing

Katarina Hanseus: Project administration, Resources, Writing – review & editing

Philip Moons: Conceptualization, Funding acquisition, Investigation, Methodology, Project administration, Supervision, Writing – review & editing

Ewa-Lena Bratt: Conceptualization, Funding acquisition, Investigation, Methodology, Project administration, Supervision, Writing – review & editing

James Mockridge: Editor

PMCID: PMC9648754 PMID: 36355855

Abstract

Background

Theoretical literature and cross-sectional studies suggest empowerment is associated with other patient-reported outcomes (PROs). However, it is not known if patient empowerment is leading to improvements in other PROs or vice versa.

Aims

The present study aimed to examine the direction of effects between patient empowerment and PROs in young persons with congenital heart disease (CHD).

Methods

As part of the STEPSTONES-CHD trial, adolescents with CHD from seven pediatric cardiology centers in Sweden were included in a longitudinal observational study (n = 132). Data were collected when patients were 16 (T0), 17 (T1) and 18 ½ years old (T2). The Gothenburg Young Persons Empowerment Scale (GYPES) was used to measure patient empowerment. Random intercepts cross-lagged panel models between patient empowerment and PROs (communication skills; patient-reported health; quality of life; and transition readiness) were undertaken.

Results

We found a significant cross-lagged effect of transition readiness over patient empowerment between T1 and T2, signifying that a higher level of transition readiness predicted a higher level of patient empowerment. No other significant cross-lagged relationships were found.

Conclusion

Feeling confident before the transition to adult care is necessary before young persons with CHD can feel in control to manage their health and their lives. Clinicians interested in improving patient empowerment during the transitional period should consider targeting transition readiness.

Background

An increase in the prevalence of chronic conditions (CC) has led healthcare systems to develop care models that accommodate the treatment of patients who are in need of follow-up, multidisciplinary attention and in some cases frequent medical interventions [1,2]. In recent years, person-centered care has been suggested and introduced. This entails healthcare providers collaborating with the patients, planning mutual care goals and involving them in the decision-making process [3], aspects that are important when treating those with CC. Nevertheless, those who live with CC also need to become autonomous, learn to cope with their CC and develop critical skills to manage their condition [4,5].

Patient empowerment has been suggested as a way of achieving the aforementioned goals [6]. This construct can be understood as a process or outcome that results from communicating with the healthcare professional, which increases the patient’s sense of control, coping abilities and self-efficacy [7]. Empowerment has been theoretically and empirically associated with improvements on other patient-reported and clinical outcomes [8]. It entails the person becoming the manager of their own care, having the skills to set goals and define ways of achieving them and gaining higher knowledge, which is expected to lead them to making healthier choices [9]. This makes it a relevant outcome for those with a CC. Moreover, patient empowerment is a central concept within nursing science, given that nurses are responsible for helping individuals feel capable of caring for themselves based on their life’s resources [10].

Studies that investigated the association between patient empowerment and clinical outcomes have found that a higher level of patient empowerment is associated with improved glycemic control (e.g. HbA1c, glucose testing, foot care) [11–16] and better disease and pain management [13]. Additionally, studies have also found that patient empowerment is associated with patient-reported outcomes (PROs), including disease-related knowledge [17,18], self-efficacy [19,20], health behaviors [11], transition readiness [21], communication skills [21], self-care [11] and quality of life (QoL) [22–26]. The problem with these studies is that we do not know the direction of effect. For instance, is more knowledge about the condition resulting in a higher level of patient empowerment, or a higher empowerment level triggering people to seek information and consequently build up more knowledge? Such direction of effect can only be investigated by using longitudinal designs. Understanding directionality can clarify how different variables are associated with each other and which one has a predictive effect. This information is important when developing interventions and determining which outcomes should be targeted and possible effects over other variables.

A particular group that can benefit from patient empowerment, is young persons with CC. During adolescence this group will face important physical and psychosocial changes. They will also have to fulfill tasks associated with this developmental stage, as well as prepare for adulthood, all while dealing with the effects of their CC. Increasing their level of empowerment can help this group develop essential skills that will facilitate these changes and tasks. Congenital heart defects (CHD) are the most frequent congenital disorders [27] and represent a heterogeneous group, in terms of defects, complexity, treatment and long-term needs [28]. Patients with CHD therefore serve as a good sample case when evaluating the generalizability of results. The present study thus aimed to examine the longitudinal associations between patient empowerment and PROs in young persons with CHD.

Methods

Design

This study is part of a larger study evaluating the effectiveness of a person-centered transition program to empower young persons with CHD using a hybrid experimental design [29]. This design involves a randomized controlled trial, where patients are assigned to an intervention or comparison group and embedded in a longitudinal, observational study, which comprises a control group from intervention-naive centers. The study evaluates the effectiveness of a nurse-led intervention who received training amongst other things on adolescent health, person-centered care, and sexual and reproductive health. Participants were followed up for a period of two and a half years and were asked to answer a set of questionnaires at three different time points. Given the aim of the present study, only the comparison and control groups were included, as the patients in these groups were not exposed to any intervention that could increase their level of empowerment. Additional information on the hybrid experimental design can be found in a methods paper [29].

Study population

Eligible participants had to fulfill the following criteria: 1) have been diagnosed with a CHD; 2) age 16 years; 3) Swedish speaking; and 4) literate [29]. Young persons were excluded if they had cognitive and/or physical limitations inhibiting them from answering the questionnaires or had a prior heart transplantation.

Given that this study is part of a larger project, the sample size was calculated based on the primary outcome of the larger study, which was patient empowerment. The target was an improvement of 5.25 points (i.e., half a standard deviation). For two-sided tests with α = 0.05 and power = 80%, 63 patients were needed in each arm of the RCT. To compensate for potential drop-out, 70 were recruited in each arm [29]. Combining the patients from the comparison arm (n = 70) and the control arm (n = 70) together, the potential total sample was of 140 participants [29]. A total of 138 individuals answered the questionnaires at T0, 108 at T1 and 101 at T2. Between T0 and T1, 29 participants dropped-out of the study and between T1 and T2 an additional seven patients. However, these were still included in the analyses to increase the sample size. Drop-outs did not differ in gender or disease complexity from participants that remained in the study. Information on how missing data was managed is explained in the data analysis section.

Procedure

Data collection was undertaken by post when the participants were 16 years (T0) and 17 (T1) and 18 ½ years (T2). Eligible participants were sent a package containing information about the study, an informed consent document, a set of questionnaires and pre-addressed return envelopes. To minimize non-response, reminders were sent after 2, 4 and 6 weeks [29].

Measures

Demographic characteristics (age, sex, educational level) and clinical data (primary diagnosis) were collected from a background information questionnaire and the patients’ medical records. Complexity of the heart disease was divided into simple, moderate or complex heart defects [30].

Patient empowerment was measured using the Gothenburg Young Persons Empowerment Scale-Congenital Heart Disease module (GYPES-CHD). This instrument comprises 15 items measured on a five-point Likert scale (strongly disagree to strongly agree). The scale measures five dimensions of patient empowerment: 1) knowledge and understanding; 2) personal control; 3) identity; 4) shared decision-making; and 5) enabling others. It is possible to calculate a subscale score for every dimension or a total score that ranges from 15–75, with higher scores denoting a higher level of empowerment [31]. The total score is used in the present study.

The following PROs to be assessed in relation to patient empowerment were chosen based on previous research that found significant effects between these variables and patient empowerment [21]. Patient-reported health was assessed with the generic and cardiac modules of the Pediatric Quality of Life Inventory 4.0 (PedsQL 4.0) [32]. The PedsQL 4.0 generic module comprises 23 items, measured on a five-point Liker scale (never to always). The scale covers the following dimensions: physical, social functioning, emotional and school functioning and it is possible to calculate subscale scores as well as a total patient-reported health score [32]. For the purpose of this study, the latter is included in the analyses.

Communication skills was measured with a subscale from the PedsQL 4.0 cardiac module [33]. The items are measured with the same Likert scale as the generic version. A higher score indicates fewer problems communicating with the healthcare providers and others about their CC [33].

QoL was measured with a linear analog scale (0–100), where a higher scoring indicates a better self-perceived QoL [34].

Transition readiness was assessed with two items from the Readiness for Transition Questionnaire (RTQ) [35]. These two items address readiness for taking over responsibility for their care and readiness for the transfer to adult care. The items are measured with a four-point Likert scale (not at all ready to completely ready) and the total score ranges from 2–8, with a higher score indicating the person is more ready for transition.

Data analysis

Descriptive statistics were expressed in absolute numbers, percentages, means and standard deviations. To test mean differences between time points, one-way repeated measures ANOVAs were undertaken. Effect sizes were reported through partial eta squared values. Additionally, minimal clinically important differences (MCIDs) for all the scales were determined by a one standard error of measurement, which is a distribution-based method to define MCIDs [36].

To explore the longitudinal associations between patient empowerment and other PROs, random intercepts cross-lagged panel models (RI-CLPM) were used. This type of structural equation modelling is useful when analysing longitudinal, observational data to understand reciprocal relationships between variables [37,38]. RI-CLPM also allows to differentiate between and within person effects, that otherwise are not possible to be determine with a classic CLPM [39]. Between person effects focus on the differences in the variables of interest in between persons (individuals who report a higher/lower level of empowerment also report a higher/lower level of the PRO compared to their peers), whereas the within person effects provide evidence on the relationship between variables (individuals who score a higher/lower level of empowerment than their expected score also tend to score higher/lower in their PRO score). Including a random intercept in the CLPM helps reduce biases in the estimates that are caused by uncontrolled between-person effects in variances and p-values, but not over fixed effects [39]. Additionally, by differentiating between and within person effects it is possible to identify who needs an intervention and modifiable targets, respectively. In the S1 File, a comparison between the model fit indices of the CLPM and the RI-CLPM is provided.

Fig 1 provides an example of the RI-CLPM used. This model follows the procedures described by Hamaker et al [39]. The figure shows that each patient empowerment score and the corresponding PRO score was decomposed into a stable between-person part and a within-person varying part. The between-person part is represented by the two random intercepts included in the figure (one for each construct). The model includes three types of relations: within-time relations, carry-over stability paths and cross-lagged relations. The latter are the ones of most interest in this study, they indicate how the variables influence each other, i.e. the extent to which changes from the PROs are predicted by patient empowerment and viceversa [39]. The coefficient (β) from the cross-lagged relations indicates the extent to which a persons’ change in deviation from their expected score in one PRO are predicted by deviations from their expected score on the other PRO from the previous measurement.

To ensure that all variation was captured by the between and within person effects, all error variances of the observed scores were constrained to zero [40]. Given that there were no strong theoretical or empirical reasons to constrain the cross-lagged and carry-over relations, these are calculated fully free.

The aforementioned model was replicated four times, as we individually assessed the relationships between patient empowerment and transition readiness, patient-reported health, communication skills and QoL. For each of the four models, goodness-of-fit indices were reported: chi-square, comparative fit index (CFI), root mean square error of approximation (RMSEA) and standardised root mean square residual (SRMR). Models with a CFI >0.090 were considered to have an acceptable fit and good fit if it was >0.95. RMSEA and SRMR values of <0.08 were acceptable and <0.05 were a good fit [41]. Missing data was managed by the Full-Information Maximum Likelihood method (FIML), which is the recommended approach in structural equation modeling [42]. FIML produces estimates based on the assumption that missing data is at random, to confirm this assumption, Little’s test for missing completely at random (MCAR) was undertaken. Results from this analysis indicated that the data was MCAR (p = 0.128).

Statistical analyses were performed using IBM SPSS Statistics for Windows version 27 and the Lavaan package in R. All tests were two-sided and the significance level was established at p≤0.05.

Ethical approval and informed consent

Ethics approval to conduct the study was received from the Ethics board of Gothenburg, Sweden (no.931 15). According to Swedish regulations, persons between the ages of 15–18 years are able to provide assent in order to participate in research studies, independently on whether the parents give their approval [43]. In this project, in line with the aforementioned regulation, the young persons were the ones who provided assent and their parents were not required to provide consent on behalf of their child. All participants were asked to sign an informed consent document and were informed they could withdraw from the study at any time point.

Results

Demographic and clinical characteristics

The sample comprised of 59.4% (n = 82) boys. Overall, 13.8% (n = 19) had a mild CHD, 61.6% (n = 85) a moderate CHD and 24.6% (n = 34) a complex CHD.

Temporal relationships

Longitudinal assessment of the study variables showed significant mean differences over the three time points (See Table 1).

Table 1. Longitudinal assessment of study variables.

	T0 Mean (±SD)	T1 Mean (±SD)	T2 Mean (±SD)	p value	Partial eta squared	MCIDs	Increase (%)^*	Decrease (%)^*
Empowerment	52.63 (10.31)	53.71 (11.81)	54.96 (12.64)	0.107	0.02	12.99	14.1	3.0
Patient-reported health	81.35 (14.57)	81.62 (17.42)	82.84 (12.87)	0.500	0.00	17.51	9.0	5.0
Communication skills	78.90 (22.62)	80.05 (20.57)	83.60 (18.43)	0.073	0.03	31.22	11.0	3.0
Quality of life	81.15 (20.47)	77.06 (18.49)	80.60 (15.16)	0.044	0.04	25.31	10.8	16
Transition readiness	5.12 (1.64)	5.59 (1.63)	6.38 (1.47)	<0.001	0.23	2.38	22.4	3.0

Open in a new tab

MCIDs: Minimal clinically important difference.

*Proportion of patients that have an increase or decrease higher than the MCID between T0 and T2.

Patient empowerment, patient-reported health, communication skills and transition readiness appeared to increase. On the other hand, QoL decreased between T0 and T1 and increased later on at T2. The changes between measurements were significant for QoL and transition readiness. Additionally, the effect size for the temporal changes for transition readiness was large. Altogether, between 3%-22% of the participants reported increases or decreases larger than the MCIDs between T0 and T2. Transition readiness was the PRO with the largest proportion of participants whose score increased.

Results from the four models tested are shown in Fig 2 and their corresponding model fit indexes are shown in Table 2. All fit indexes of the models had a good fit, with the exception for the one tested between QoL and patient empowerment. The latter had a RMSEA value above the expected range.

Table 2. Model fit statistics of the four random intercepts cross-lagged panel models tested.

Fit indexes	Quality of life	Transition readiness	Communication skills	Patient-reported health
Comparative fit index (CFI)	0.981	1.000	0.999	0.998
Root mean square error of approximation (RMSEA)	0.180	0.000	0.044	0.061
Standardized root mean square residual (SRMR)	0.037	0.012	0.023	0.028
Chi square test of model fit Degrees of freedom P value Normed chi² index (x²/df)	5.464 1 0.019 5.464	0.412 1 0.521 0.412	1.263 1 0.261 1.263	1.519 1 0.0218 1.519

Open in a new tab

A CFI >0.090 was considered to have an acceptable fit and good fit if it was >0.95. RMSEA and SRMR values of <0.08 were acceptable and <0.05 were a good fit.

The between person associations (not shown in Fig 2) were only significant in the model between patient empowerment and communication skills, indicating that individuals with a higher level of empowerment across the time points reported better communication skills than individuals with lower patient empowerment (β = 0.46, p<0.01). On the within person level, there was a significant cross-lagged effect from transition readiness to patient empowerment between T1 and T2 (β = 0.39, p<0.05). This significant effect is interpreted as persons with deviations from their expected patient empowerment score at T2 were predicted by their level of transition readiness at T1 (Fig 2, panel B). None of the other models had significant cross-lagged effects (Fig 2, panels A-D).

There was a significant carry-over effect in QoL from T0 to T1 (β = 0.62, p<0.01), indicating that within person deviations from the expected QoL score at T0 predicted deviations from the expected QoL score at T1 (Fig 2, panel A). Additionally, a significant within time association was found between patient empowerment and transition readiness at T1 (β = 0.47, p<0.05). This can be translated as changes in the mean score of the patient empowerment at T1 were associated with changes in transition readiness at this same time point.

Discussion

To the best of our knowledge, the present study is the only study that uses a longitudinal design and RI-CLPM to test the longitudinal associations between patient empowerment and other PROs. Through these analyses and the nature of the data, it is possible to determine which variable has a predictive effect over the other one and to separate between and within person effects, obtaining pure estimates from the relationships of these variables. Our results showed that in young persons with CHD, transition readiness predicted the score of patient empowerment when the participants were 18 ½ years, indicating the direction of effect.

Bravo and colleagues [8] provide one of the most thorough conceptual models on patient empowerment. In the model, they propose that taking an active role and having perceived control are indicators of patient empowerment, aspects that in our study are captured by transition readiness. Indeed, transition readiness in our study is defined as the “adolescents’ readiness to assume complete responsibility for their care and their readiness to transfer to adult medical care” [35]. Unlike in Bravo and colleagues’ conceptual model, our findings support the idea that transition readiness, rather than being an indicator of patient empowerment, it is actually a determinant of it.

The association between patient empowerment and transition readiness has been investigated before in a cross-sectional study, which found a significant association [44]. The current study adds to available evidence by providing details on the direction of effects between these two variables. Research that has focused on the association between empowerment and transition readiness is limited and research that focuses on variables associated with transition readiness are also mostly cross-sectional [45]. Transition readiness has previously been found to be significantly associated with several psychosocial factors such as self-efficacy, motivation and patient activation [45]. Therefore, it is reasonable for it to also be associated with patient empowerment.

The predictive effect of transition readiness over patient empowerment can be potentially explained by the fact that young persons with CHD need to have the confidence or feel they are ready to assume the new role in front of them, before they can take a more active role in their health and lives and become empowered. A higher level of transition readiness can be associated with the adolescents having a certain level of awareness about their possibilities and what is required from them once they are transferred to adult care. This awareness is a prerequisite for choosing their own agenda (i.e. becoming empowered) [46].

Interestingly, the predictive effect of transition readiness was between T1 and T2, when the participants are 17 and 18 ½ years, respectively. Previous studies have found that transition readiness is associated with age and that it increases with age [45,47,48]. It is plausible that it is up to this point when their level of transition readiness is high enough to lead to an increment on patient empowerment. This fact is perhaps of relevance in understanding the longitudinal effects of transition readiness and the potential point when improvements over patient empowerment and perhaps other outcomes can be expected and therefore, measured. Future studies could evaluate the association between age, transition and patient empowerment, more specifically determining whether age is a mediating factor between these two variables.

Even when there is literature suggesting patient empowerment is associated with other variables in the long-term [8,49], there is still a lack of longitudinal research [50], meaning the available evidence supporting current models on patient empowerment is relatively low. This study is one of the first ones in providing longitudinal evidence that informs the potential predictive value of patient empowerment. Our findings are an initial step in more thoroughly understanding the theoretical framework of patient empowerment and how this construct associates with other variables.

The lack of significant cross-lagged effects with the other evaluated variables, such as QoL, might be associated with the fact that there could be a mediating variable that is not accounted for in the models. It is plausible that the effect of patient empowerment is mediated by a third variable. According to Palumbo [51], the variables that mediate the relationship between patient empowerment and other PROs have been overlooked. Palumbo [51] suggests that self-efficacy, self-management, health literacy and even patient activation could be some of the variables that strengthen the association between patient empowerment and other outcomes.

Another possible explanation for why we did not find significant cross-lagged effects could be that the level of patient empowerment of the participants was not sufficiently high to be associated with improvements of other PROs later on. There is no available information on what constitutes a high level of patient empowerment or even the necessary amount of patient empowerment to trigger improvements in other variables. There is evidence that empowering interventions have led to improvement in different outcomes, such as QoL [23,24,52]. Perhaps in the case of our participants, they did not have a sufficiently high level of patient empowerment or even a sufficiently high level of the other PROs to achieve this.

It is also worth noting that cross-lagged analyses are dependent on the length of the time interval between measurements, meaning that other researchers who study the same models can potentially obtain different estimates, merely due to the time interval they use [53]. This particular aspect could also explain why we only found a significant cross-lagged effect and why perhaps other researchers in the future may find opposite results.

Our findings are the first from a longitudinal study that provide evidence on the associations between patient empowerment and other PROs across time. However, more longitudinal research is necessary to understand the predictive effect (or lack thereof) of patient empowerment. By doing so, current theoretical and conceptual models can be revised and interventions aiming to improve patient empowerment can be better designed.

Methodological issues

This study has several strengths: 1) longitudinal data on patient empowerment are scarce and therefore not many studies have been capable of assessing longitudinal associations in relation to this construct; 2) all the questionnaires used had been previously validated in young persons with chronic conditions; 3) data collections were done within the same time frame for all participants; and 4) the study involved data from 7 different hospitals, which increases the generalizability of the results.

However, some limitations ought to be considered when interpreting the results. First, we have a small sample size. While literature on RI-CLPM and structural equation modeling has not come to an agreement on the required sample in order to convey proper models, a minimum of 200 participants has been suggested as the threshold [41]. Second, considering the sample size, it was not possible to test models that are more complex and control the effect of different variables. Third, our sample only included young persons who were between 16–18 ½ years old, which means these results might not provide evidence beyond this period in adolescence. Fourth, our study includes three different time points, which is the minimum for RI-CLPM, and this increases the risk that our study is underpowered. Fifth, dropouts are always a possibility when undertaking several data collections. We started with 138 patients at T0, and unfortunately, had only 101 by T1, despite our best efforts to promote participation.

Conclusions

Our study is the first longitudinal study to attempt to describe the longitudinal associations between patient empowerment and other PROs, as well as the potential predictive value of empowerment. The present study provides evidence on the predictive value of transition readiness over patient empowerment. This means increments in transition readiness are associated with positive changes over patient empowerment. Researchers or clinicians working in transitional care can reflect on whether patient empowerment should be an outcome to be measured, in light of the predictive effect that transition readiness has over this outcome.

Supporting information

S1 File. Model comparison of the CLPM and RI-CLPM models.

(DOCX)

Click here for additional data file.^{(15.4KB, docx)}

Acknowledgments

We would like to thank Koen Raymaekers for his support and comments throughout the data analysis and revision of the manuscript.

Data Availability

Data cannot be shared publicly because of the sensitive nature of it. However, requests to access the data set from qualified researchers trained in human subject confidentiality protocols can be made by sending an email to the University of Gothenburg at karin.dejke@gu.se.

Funding Statement

This study is supported by research grants from the Swedish Research Council for Health, Working Life and Welfare-FORTE (grant STYA-2015/0003; http://forte.se/; PM); Swedish Heart-Lung Foundation (grant 20150535; https://www.hjart-lungfonden.se/; PM); Swedish Research Council (grant 2015-02503; https://www.vr.se/; PM); Swedish Children Heart Association (http://www.hjartebarn.se/; ELB). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

References

1.Nuño R, Coleman K, Bengoa R, Sauto R. Integrated care for chronic conditions: the contribution of the ICCC Framework. Health Policy. 2012;105(1):55–64. doi: 10.1016/j.healthpol.2011.10.006 [DOI] [PubMed] [Google Scholar]
2.Hajat C, Stein E. The global burden of multiple chronic conditions: a narrative review. Prev Med Rep. 2018;12:284–93. doi: 10.1016/j.pmedr.2018.10.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
3.Ekman I, Swedberg K, Taft C, Lindseth A, Norberg A, Brink E, et al. Person-centered care—ready for prime time. Eur J Cardiovasc Nurs. 2011;10(4):248–51. doi: 10.1016/j.ejcnurse.2011.06.008 [DOI] [PubMed] [Google Scholar]
4.Aujoulat I, d’Hoore W, Deccache A. Patient empowerment in theory and practice: polysemy or cacophony?. Patient Educ Couns. 2007;66(1):13–20. doi: 10.1016/j.pec.2006.09.008 [DOI] [PubMed] [Google Scholar]
5.Ambrosio L, Senosiain García JM, Riverol Fernández M, Anaut Bravo S, Díaz De Cerio Ayesa S, Ursúa Sesma ME, et al. Living with chronic illness in adults: a concept analysis. J Clin Nurs. 2015;24(17–18):2357–67. doi: 10.1111/jocn.12827 [DOI] [PubMed] [Google Scholar]
6.Anderson RM, Funnell MM. Patient empowerment: reflections on the challenge of fostering the adoption of a new paradigm. Patient Educ Couns. 2005;57(2):153–7. doi: 10.1016/j.pec.2004.05.008 [DOI] [PubMed] [Google Scholar]
7.Small N, Bower P, Chew-Graham CA, Whalley D, Protheroe J. Patient empowerment in long-term conditions: development and preliminary testing of a new measure. BMC Health Serv Res. 2013;13:263. doi: 10.1186/1472-6963-13-263 [DOI] [PMC free article] [PubMed] [Google Scholar]
8.Bravo P, Edwards A, Barr PJ, Scholl I, Elwyn G, McAllister M, et al. Conceptualising patient empowerment: a mixed methods study. BMC Health Serv Res. 2015;15(1):252. doi: 10.1186/s12913-015-0907-z [DOI] [PMC free article] [PubMed] [Google Scholar]
9.Falk-Rafael AR. Empowerment as a process of evolving consciousness: a model of empowered caring. Adv Nurs Sci. 2001;24(1):1–16. [DOI] [PubMed] [Google Scholar]
10.Meleis A. The discipline of nursing: perspective and domain. Theoretical nursing: development and progress. 6th edition ed. China: Wolters Kluwer; 2018. [Google Scholar]
11.Yang S, Hsue C, Lou Q. Does patient empowerment predict self-care behavior and glycosylated hemoglobin in chinese patients with type 2 diabetes?. Diabetes Technol Ther. 2015;17(5):343–8. doi: 10.1089/dia.2014.0345 [DOI] [PubMed] [Google Scholar]
12.Kuo CC, Lin CC, Tsai FM. Effectiveness of empowerment-based self-management interventions on patients with chronic metabolic diseases: a systematic review and meta-analysis. Worldviews Evid Based Nurs. 2014;11(5):301–15. doi: 10.1111/wvn.12066 [DOI] [PubMed] [Google Scholar]
13.Wong CKH, Wong WCW, Wan EYF, Wong WHT, Chan FWK, Lam CLK. Increased number of structured diabetes education attendance was not associated with the improvement in patient-reported health-related quality of life: results from Patient Empowerment Programme (PEP). Health Qual Life Outcomes. 2015;13:126. doi: 10.1186/s12955-015-0324-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
14.Rossi MC, Lucisano G, Funnell M, Pintaudi B, Bulotta A, Gentile S, et al. Interplay among patient empowerment and clinical and person-centered outcomes in type 2 diabetes. The BENCH-D study. Patient Educ Couns. 2015;98(9):1142–9. doi: 10.1016/j.pec.2015.05.012 [DOI] [PubMed] [Google Scholar]
15.Cortez DN, Macedo MML, Souza DAS, Dos Santos JC, Afonso GS, Reis IA, et al. Evaluating the effectiveness of an empowerment program for self-care in type 2 diabetes: a cluster randomized trial. BMC Public Health. 2017;17(1):41. doi: 10.1186/s12889-016-3937-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Baldoni NR, Aquino JA, Sanches-Giraud C, Di Lorenzo Oliveira C, de Figueiredo RC, Cardoso CS, et al. Collective empowerment strategies for patients with Diabetes Mellitus: a systematic review and meta-analysis. Prim Care Diabetes. 2017;11(2):201–11. doi: 10.1016/j.pcd.2016.09.006 [DOI] [PubMed] [Google Scholar]
17.Akhter K, Bunn C, Graffy J, Donald S, Ward C, Simmons D. Empowerment-based education for established type 2 diabetes in rural England. Pract Diabetes. 2017;34(3):83–8. [Google Scholar]
18.van den Berg JJ, Neilands TB, Johnson MO, Chen B, Saberi P. Using path analysis to evaluate the healthcare empowerment model among persons living with HIV for antiretroviral therapy adherence. AIDS Patient Care STDS. 2016;30(11):497–505. doi: 10.1089/apc.2016.0159 [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Grealish A, Tai S, Hunter A, Emsley R, Murrells T, Morrison AP. Does empowerment mediate the effects of psychological factors on mental health, well-being, and recovery in young people?. Psychol Psychother. 2017;90(3):314–35. doi: 10.1111/papt.12111 [DOI] [PubMed] [Google Scholar]
20.Sit JW, Chair SY, Choi KC, Chan CW, Lee DT, Chan AW, et al. Do empowered stroke patients perform better at self-management and functional recovery after a stroke? A randomized controlled trial. Clin Interv Aging. 2016;11:1441–50. doi: 10.2147/CIA.S109560 [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Acuña Mora M, Saarijarvi M, Moons P, Sparud-Lundin C, Bratt EL, Goossens E. The scope of research on transfer and transition in young persons with chronic conditions. J Adolesc Health. 2019;65(5):581–9. doi: 10.1016/j.jadohealth.2019.07.014 [DOI] [PubMed] [Google Scholar]
22.Shin S, Park H. Effect of empowerment on the quality of life of the survivors of breast cancer: The moderating effect of self-help group participation. Jpn J Nurs Sci. 2017;14(4):311–9. doi: 10.1111/jjns.12161 [DOI] [PubMed] [Google Scholar]
23.Lian J, McGhee SM, So C, Chau J, Wong CKH, Wong WCW, et al. Five-year cost-effectiveness of the Patient Empowerment Programme (PEP) for type 2 diabetes mellitus in primary care. Diabetes Obes Metab. 2017;19(9):1312–6. doi: 10.1111/dom.12919 [DOI] [PubMed] [Google Scholar]
24.Moattari M, Ebrahimi M, Sharifi N, Rouzbeh J. The effect of empowerment on the self-efficacy, quality of life and clinical and laboratory indicators of patients treated with hemodialysis: a randomized controlled trial. Health Qual Life Outcomes. 2012;10:115. doi: 10.1186/1477-7525-10-115 [DOI] [PMC free article] [PubMed] [Google Scholar]
25.Pibernik-Okanovic M, Prasek M, Poljicanin-Filipovic T, Pavlic-Renar I, Metelko Z. Effects of an empowerment-based psychosocial intervention on quality of life and metabolic control in type 2 diabetic patients. Patient Educ Couns. 2004;52(2):193–9. doi: 10.1016/s0738-3991(03)00038-7 [DOI] [PubMed] [Google Scholar]
26.Wang J, Wang J, Cao Y, Jia S, Wu B. Perceived empowerment, social support, and quality of life among chinese older residents in long-term care facilities. J Aging Health. 2018;30(10):1595–619. doi: 10.1177/0898264318795724 [DOI] [PubMed] [Google Scholar]
27.Benziger CP, Stout K, Zaragoza-Macias E, Bertozzi-Villa A, Flaxman AD. Projected growth of the adult congenital heart disease population in the United States to 2050: an integrative systems modeling approach. Popul Health Metr. 2015;13:29. doi: 10.1186/s12963-015-0063-z [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Mandalenakis Z, Rosengren A, Skoglund K, Lappas G, Eriksson P, Dellborg M. Survivorship in children and young adults with congenital heart disease in Sweden. JAMA Intern Med. 2017;177(2):224–30. doi: 10.1001/jamainternmed.2016.7765 [DOI] [PubMed] [Google Scholar]
29.Acuña Mora M, Sparud-Lundin C, Bratt EL, Moons P. Person-centred transition programme to empower adolescents with congenital heart disease in the transition to adulthood: a study protocol for a hybrid randomised controlled trial (STEPSTONES project). BMJ Open. 2017;7(4):e014593. doi: 10.1136/bmjopen-2016-014593 [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Warnes CA, Liberthson R, Danielson GK, Dore A, Harris L, Hoffman JI, et al. Task force 1: the changing profile of congenital heart disease in adult life. J Am Coll Cardiol. 2001;37(5):1170–5. doi: 10.1016/s0735-1097(01)01272-4 [DOI] [PubMed] [Google Scholar]
31.Acuña Mora M, Luyckx K, Sparud-Lundin C, Peeters M, van Staa A, Sattoe J, et al. Patient empowerment in young persons with chronic conditions: psychometric properties of the Gothenburg Young Persons Empowerment Scale (GYPES). PLoS One. 2018;13(7):e0201007. doi: 10.1371/journal.pone.0201007 [DOI] [PMC free article] [PubMed] [Google Scholar]
32.Petersen S, Hägglöf B, Stenlund H, Bergström E. Psychometric properties of the Swedish PedsQL, Pediatric Quality of Life Inventory 4.0 generic core scales. Acta Paediatr. 2009;98(9):1504–12. doi: 10.1111/j.1651-2227.2009.01360.x [DOI] [PubMed] [Google Scholar]
33.Sand P, Kljajić M, Sunnegårdh J. The reliability of the Pediatric Quality of Life Inventory 3.0 Cardiac Module™ for Swedish children with congenital heart defects. Nord Psychol. 2013;65(3):210–23. [Google Scholar]
34.Moons P, Van Deyk K, De Bleser L, Marquet K, Raes E, De Geest S, et al. Quality of life and health status in adults with congenital heart disease: a direct comparison with healthy counterparts.Eur J Cardiovasc Prev Rehabil. 2006;13(3):407–13. doi: 10.1097/01.hjr.0000221864.19415.a0 [DOI] [PubMed] [Google Scholar]
35.Gilleland J, Amaral S, Mee L, Blount R. Getting ready to leave: transition readiness in adolescent kidney transplant recipients. J Pediatr Psychol. 2012;37(1):85–96. doi: 10.1093/jpepsy/jsr049 [DOI] [PubMed] [Google Scholar]
36.Crosby RD, Kolotkin RL, Williams GR. Defining clinically meaningful change in health-related quality of life. Journal of Clinical Epidemiology. 2003;56(5):395–407. doi: 10.1016/s0895-4356(03)00044-1 [DOI] [PubMed] [Google Scholar]
37.Raymaekers K, Luyckx K, Moons P. A guide to improve your causal inferences from observational data. Eur J Cardiovasc Nurs. 2020;19(8):757–62. doi: 10.1177/1474515120957241 [DOI] [PMC free article] [PubMed] [Google Scholar]
38.Hamaker EL, Mulder JD, van Ijzendoorn MH. Description, prediction and causation: Methodological challenges of studying child and adolescent development. Dev Cogn Neurosci. 2020;46:100867. doi: 10.1016/j.dcn.2020.100867 [DOI] [PMC free article] [PubMed] [Google Scholar]
39.Hamaker EL, Kuiper RM, Grasman RP. A critique of the cross-lagged panel model. Psychol Methods. 2015;20(1):102–16. doi: 10.1037/a0038889 [DOI] [PubMed] [Google Scholar]
40.Masselink M, Van Roekel E, Hankin BL, Keijsers L, Lodder GMA, Vanhalst J, et al. The longitudinal association between self-esteem and depressive symptoms in adolescents: separating between-person effects from within-person effects. Eur J Pers. 2018;32(6):653–71. doi: 10.1002/per.2179 [DOI] [PMC free article] [PubMed] [Google Scholar]
41.Rex Kline. Principles and practices of strucutal equation modeling. 4th ed. United States: Guilford Publications; 2016. [Google Scholar]
42.Bowen NK. Structural Equation Modeling: Structural Equation Modeling: Cary: Oxford University Press, Incorporated; 2011. [Google Scholar]
43.Law on ethical review of research concerning people (2003:460) 18 §.
44.Acuña Mora M, Sparud-Lundin C, Burstrom A, Hanseus K, Rydberg A, Moons P, et al. Patient empowerment and its correlates in young persons with congenital heart disease. Eur J Cardiovasc Nurs. 2019;18(5):389–98. doi: 10.1177/1474515119835434 [DOI] [PubMed] [Google Scholar]
45.Varty M, Popejoy LL. A Systematic Review of Transition Readiness in Youth with Chronic Disease. West J Nurs Res. 2020;42(7):554–66. doi: 10.1177/0193945919875470 [DOI] [PMC free article] [PubMed] [Google Scholar]
46.Tengland P-A. Empowerment: A goal or a means for health promotion?. Med Health Care Philos. 2006;10(2):197. doi: 10.1007/s11019-006-9027-1 [DOI] [PubMed] [Google Scholar]
47.Lazaroff SM, Meara A, Tompkins MK, Peters E, Ardoin SP. How do health literacy, numeric competencies, and patient activation relate to transition readiness in adolescents and young adults with rheumatic diseases?. Arthritis Care Res. 2019;71(9):1264–9. [DOI] [PubMed] [Google Scholar]
48.McColl J, Semalulu T, Beattie KA, Alam A, Thomas S, Herrington J, et al. Transition readiness in adolescents with juvenile idiopathic arthritis and childhood-onset systemic lupus erythematosus. ACR Open Rheumatol. 2021;3(4):260–5. doi: 10.1002/acr2.11237 [DOI] [PMC free article] [PubMed] [Google Scholar]
49.Castro EM, Van Regenmortel T, Vanhaecht K, Sermeus W, Van Hecke A. Patient empowerment, patient participation and patient-centeredness in hospital care: a concept analysis based on a literature review. Patient Educ Couns. 2016;99(12):1923–39. doi: 10.1016/j.pec.2016.07.026 [DOI] [PubMed] [Google Scholar]
50.Acuña Mora M, Sparud-Lundin C, Moons P, Bratt E-L. Definitions, instruments and correlates of patient empowerment: a descriptive review. Patient Educ Couns. 2021;105(2):346–355. doi: 10.1016/j.pec.2021.06.014 [DOI] [PubMed] [Google Scholar]
51.Palumbo R. The bright side and the dark side of patient empowerment: co-creation and co-destruction of value in the healthcare environment. Switzerland Springer Nature; 2017. [Google Scholar]
52.Anderson RM, Funnell MM, Aikens JE, Krein SL, Fitzgerald JT, Nwankwo R, et al. evaluating the efficacy of an empowerment-based self-management consultant intervention: results of a two-year randomized controlled trial. Ther Patient Educ. 2009;1(1):3–11. doi: 10.1051/tpe/2009002 [DOI] [PMC free article] [PubMed] [Google Scholar]
53.Kuiper RM, Ryan O. Drawing conclusions from cross-lagged relationships: re-considering the role of the time-interval. Struct Equ Modeling. 2018;25(5):809–23. [Google Scholar]

PLoS One. doi: 10.1371/journal.pone.0277267.r001

Decision Letter 0

Kathleen Finlayson

17 Mar 2022

PONE-D-21-31558The longitudinal association between patient empowerment and patient-reported outcomes: what is the direction of effect?PLOS ONE

Dear Dr. Acuña Mora,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please address the queries raised by the reviewers on the analysis methods, and the need for further clarity and detail in the objectives and some of the methods.

Please submit your revised manuscript by May 01 2022 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Kathleen Finlayson

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. You indicated that you had ethical approval for your study. In your Methods section, please ensure you have also stated whether the research ethics committee or IRB specifically waived the need for their consent.

3. In your Data Availability statement, you have not specified where the minimal data set underlying the results described in your manuscript can be found. PLOS defines a study's minimal data set as the underlying data used to reach the conclusions drawn in the manuscript and any additional data required to replicate the reported study findings in their entirety. All PLOS journals require that the minimal data set be made fully available. For more information about our data policy, please see http://journals.plos.org/plosone/s/data-availability.

Upon re-submitting your revised manuscript, please upload your study’s minimal underlying data set as either Supporting Information files or to a stable, public repository and include the relevant URLs, DOIs, or accession numbers within your revised cover letter. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories. Any potentially identifying patient information must be fully anonymized.

Important: If there are ethical or legal restrictions to sharing your data publicly, please explain these restrictions in detail. Please see our guidelines for more information on what we consider unacceptable restrictions to publicly sharing data: http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions. Note that it is not acceptable for the authors to be the sole named individuals responsible for ensuring data access.

We will update your Data Availability statement to reflect the information you provide in your cover letter.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: No

Reviewer #2: No

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Page 7, line 111: If you imputed observations for the dropout observations you should explicitly state that

Page 8, line 132: When empowerment (or continuous PROs) is treated as an outcome, is there a validated threshold for the average difference in score beyond which the results are considered as important? If yes, please use it to interpret the findings

Page 8, line 143: Similar to the previous comment, is there a threshold for the difference in PedsQL 4.0 should regard as important? Is it dependent on baseline values?

Page 9, line 154: You need to evaluate skewness for the distribution of your continuous measurements, if skewed you should report your descriptives in terms median and interquartile range

Page 9, line 156: Large partial eta squared valued may reflect small sampling variability rather effectiveness of empowerment. To demonstrate effectiveness please report mean differences or other similar measures

Page 9, line 167: For continuous outcomes, random effects (including random intercept) accounts for bias in variances and p-values but not biases in fixed effects. Please re-word your sentence to make it clear that you are not referring to fixed effects

Page 10, line 193: By using FIML, you are assuming that the missing observations are missing at random. Please justify how you landed to this assumption

Page 11, line 214: You have 3 time points so the comparison shouldn’t be confined to only 2 points (I’m aware that Table 1 has all 3 time points)

Table 1, Quality of Life: Because of lack of linear trend, I don't find the reported P value and partial eta squared for QoL to be useful in this instance (QoL decreased at T2 but increased at T3). Also, it is a little puzzling that SE decreases over time although there are dropouts in subsequent time points. Even if you imputed, I would expect FIML to account for uncertainty due imputed data

Page 12, line 218: QoL did not decrease consistently. It decreased at T1 and increased from T1 to T2

Pages 13 & 14: How do we interpret beta? Is it standardized mean difference (standardized to have mean 0 and variance 1)? If so, it is hard to tell if large values for beta imply a large difference in means or small variance. It is informative to report mean differences in their natural form along with 95% CI.

Page 13, line 233: The meaning of “effect” seems to differ across the manuscript. Please make sure that you are consistent with the usage of the term “effect”. Also, what you describe as “effect” is simply a measure of compatibility of your model with the null hypothesis

Page 14, line 245: The asterisks in p values seems to suggest the smaller is the p values the larger is the magnitude of effectiveness. Please note, small p-values may merely reflect small SEs, as such, they shouldn't be treated as measures of strength of effectiveness

Page 18, lines 342-349: There is a difference between predicting and evaluating whether empowerment affects PRO and vice versa. Your conclusion suggests your interest was in predicting whereas the research question suggest the interest is in estimating the effects. More clarity is needed on your objectives

Reviewer #2: This manuscript considers secondary analysis of data generated from a randomized clinical trial. The objective here is to examine the direction of effects between patient empowerment and PROs in young subjects with CHD. The design generated a longitudinal study, and the analysis here considers comparing 2 groups. I have the following queries, which, when addressed, will strengthen the analysis.

1. It would be great to provide some ballpark sample size/power estimate, wrt. the desired effect size in mind. This maybe important to replicate this design. For example, 2 groups, each with 70 subjects, were considered. So, what is the resultant power wrt. this sample size, considering the longitudinal design, and the primary response variable (the composite score).

2. A one-way repeated measures ANOVA was used to analyze the longitudinal outcomes. ANOVAs depend heavily on Gaussianity assumptions of the response variable. How was that assessed? If failed, alternative methods are needed, such as Friedman's test, maybe needed.

3. Same goes for the RI-CLPM modeling, which assumes Gaussianity . Further, to better promote the RI-CLPM, some comparisons are needed with the basic CLPM. Authors may follow this link if they want to:

https://johnflournoy.science/2017/10/20/riclpm-lavaan-demo/

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

PLoS One. 2022 Nov 10;17(11):e0277267. doi: 10.1371/journal.pone.0277267.r002

Author response to Decision Letter 0

24 May 2022

Reviewer #1

Page 7, line 111: If you imputed observations for the dropout observations you should explicitly state that

Information on how missing data was managed is provided under the “Data analysis” section. For the RI-CLPM missing data was managed through FIML. However, data was not imputed for the one-way repeated measures ANOVA.

Minimal clinically important differences (MCIDs) were calculated for all scales using a distribution-based method, which was established as one standard error of measurement [1]. This approach is relatively stable across populations and considers the precision of the measure [1, 2]. This information has now been included in the “Data analysis” section, as well as in Table 1. Furthermore, the interpretation of this information has been included in the “Results” (See page 10, lines 173-175).

Page 8, line 143: Similar to the previous comment, is there a threshold for the difference in PedsQL 4.0 should regard as important? Is it dependent on baseline values?

As mentioned previously, MCIDs were determined based on a one standard error of measurement.

Page 9, line 154: You need to evaluate skewness for the distribution of your continuous measurements, if skewed you should report your descriptives in terms median and interquartile range

Normality of the data was assessed through histograms and Q-Q plots as well as with the Kolmogorov Smirnov och Shapiro Wilk tests. To counter the potential bias of these approaches, normality was also assessed through the skewness and kurtosis of the data. Acceptable skewness was between -2/+2 and kurtosis between -7/+7 [3]. An assessment of all this data indicated the data was approximately normally distributed. Therefore, the report of means and standard deviations is fitting.

The partial eta square for patient empowerment was 0.02, which is a small effect size [4]. The only outcome that had a large effect size was transition readiness (partial eta square= 0.23). To account for the potential bias of this estimate, we are also reporting MCIDs in Table 1.

This sentence has been rephrased to highlight the fact that the random intercept does not account for biases in fixed effects (See page 11, lines 187-188).

Page 10, line 193: By using FIML, you are assuming that the missing observations are missing at random. Please justify how you landed to this assumption

Indeed, FIML assumes that data are missing at random. In order to confirm this assumption, Little’s missing completely at random (MCAR) test was undertaken [5, 6]. This analysis was non-significant (p=0.128), confirming the data are missing at random. This information has been included in the “data analysis” section.

Page 11, line 214: You have 3 time points so the comparison shouldn’t be confined to only 2 points (I’m aware that Table 1 has all 3 time points)

This was a typographical error that has been corrected. Indeed, the comparison was done for the three time points.

The information reported in table 1 is not the standard error (SE), but rather the standard deviation (SD). Unfortunately, this was a typographical error that was not identified before the submission of the manuscript. This error indeed influences the interpretation of the table and expected changes as the sample size decreases. We have corrected the typographical error in Table 1. It is worth noting, that FIML was only used for the RI-CLPM and not for the one-way repeated measures ANOVAs.

Page 12, line 218: QoL did not decrease consistently. It decreased at T1 and increased from T1 to T2.

The reviewer is correct, and this sentence has been revised to clarify that quality of life decreased initially and later increased again.

While it is possible to report 95% CI, it is not standard practice to report this information. The β coefficient from the cross-lagged relations indicates the extent to which for instance a persons’ change in deviation from their expected quality of life score is predicted by deviations from their expected score on patient empowerment from the previous measurement. The reported coefficients in the manuscript are standardized.

Across the manuscript when we write “effect”, we refer to the influence one variable has over the other. This effect can be for instance the influence patient empowerment has over another PRO in a follow-up measurement or the influence patient empowerment has over itself.

We have revised the manuscript to confirm the use of the word “effect” is consistent throughout the manuscript.

As the reviewer indicates, p-values are not indicators of the effect’s magnitude. P-values are indicators of how likely it is to have found a particular set of results if the null hypothesis is true. If the researcher wants to indicate the magnitude of a significant result, then other measures such as Cohen’s D or eta square are more relevant, since these measures will indicate whether the effect is small, medium or large. Nonetheless, our reporting of the p-values as such is common practice. The more asterisks (*), the smaller the p value, and this should be interpreted as the probabilities of having found these results if the null hypothesis is true, are smaller.

The aim of our study was to examine the longitudinal associations between patient empowerment and other patient-reported outcomes. This examination entailed both, estimating the effects and determining which variable has a predicting effect over the other. Therefore, the information provided in the conclusions is in line with this. We decided to focus on the cross-lagged relations (i.e., predictive effect), because it’s the most relevant finding from the study.

Reviewer #2

This manuscript considers secondary analysis of data generated from a randomized clinical trial. The objective here is to examine the direction of effects between patient empowerment and PROs in young subjects with CHD. The design generated a longitudinal study, and the analysis here considers comparing 2 groups. I have the following queries, which, when addressed, will strengthen the analysis.

It would be great to provide some ballpark sample size/power estimate, wrt. the desired effect size in mind. This maybe important to replicate this design. For example, 2 groups, each with 70 subjects, were considered. So, what is the resultant power wrt. this sample size, considering the longitudinal design, and the primary response variable (the composite score).

The sample size for this study is based on the sample size calculation that was made for a larger study. In the latter, the calculation was based on an improvement of patient empowerment of 5.25 points, an α=0.05 and power=80%. This meant we needed 63 patients in each arm. However, to compensate for drop-outs, 70 patients were included in each arm. This information has been included in the “Study population” section.

A one-way repeated measures ANOVA was used to analyze the longitudinal outcomes. ANOVAs depend heavily on Gaussianity assumptions of the response variable. How was that assessed? If failed, alternative methods are needed, such as Friedman's test, maybe needed.

One of the assumptions for one-way repeated measures ANOVAs is that the residuals should be normally distributed. The residuals represent the difference between each individual observation and the group’s mean from where the observation came from. Hence, the raw data does not need to be normally distributed. The normal distribution of the residuals was evaluated through their skewness and kurtosis. These parameters were selected given that tests such as Shapiro-Wilk and Kolmogorov Smirnov tests are sensitive to sample sizes and histograms, P-P plots and Q-Q plots, while useful, rely heavily on the interpretation of the observer. Cut-off values of -2/+2 for skewness and between -7/+7 for kurtosis were used to determine whether the residuals were normally distributed [3]. While these cut-off values might seem too broad, ANOVAs are analyses that also tolerate certain deviations from the normal distribution.

Same goes for the RI-CLPM modeling, which assumes Gaussianity. Further, to better promote the RI-CLPM, some comparisons are needed with the basic CLPM. Authors may follow this link if they want to:

https://johnflournoy.science/2017/10/20/riclpm-lavaan-demo/

RI-CLPM models assume multivariate normality for continuous outcomes. However, testing for multivariate normality its difficult to assess, given that small deviations from normality are easily detected in large samples and in small samples, power might not be enough to detect them [7]. Therefore, it is often the case that the inspection of each variable (univariate normality) through their skewness and kurtosis is a useful tool [3]. This is important because multivariate normality is not possible without univariate normality. As it was mentioned in a previous response, the variables’ distribution was assessed, and their skewness and kurtosis were within the accepted ranges. However, since there was the risk that our data was not complying with the multivariate normality assumption, as an estimator we used “robust maximum likelihood (MLR)”. This estimator is meant for variables with non-normal distributions [7].

The CLPM is nested in the RI-CLPM, so comparisons of these models are in occasions presented in the literature. We have decided to provide the results from these comparisons as a Supplementary File. This was done so because the paper focuses on evaluating the longitudinal relationships between patient empowerment and other PROs, rather than the comparison across different structural equation models. We refer to the Supplementary File in page 11, lines 189-190.

References

1. Sedaghat AR. Understanding the Minimal Clinically Important Difference (MCID) of Patient-Reported Outcome Measures. Otolaryngology–Head and Neck Surgery. 2019;161(4):551-60.

2. Crosby RD, Kolotkin RL, Williams GR. Defining clinically meaningful change in health-related quality of life. Journal of Clinical Epidemiology. 2003;56(5):395-407.

3. Byrne BM. Structural equation modeling with AMOS : basic concepts, applications, and programming. Mahwah, N.J.: Mahwah, N.J. : Lawrence Erlbaum Associates; 2001.

4. Lakens D. Calculating and reporting effect sizes to facilitate cumulative science: a practical primer for t-tests and ANOVAs. 2013;4(863).

5. Roderick JAL. A Test of Missing Completely at Random for Multivariate Data with Missing Values. Journal of the American Statistical Association. 1988;83(404):1198-202.

6. Li C. Little's Test of Missing Completely at Random. The Stata Journal. 2013;13(4):795-809.

7. Kline Rex. Principles and practices of strucutal equation modeling. Fourth edition ed. United States: Guilford Publications; 2016.

Attachment

Submitted filename: Response to reviewers.docx

Click here for additional data file.^{(30.3KB, docx)}

PLoS One. doi: 10.1371/journal.pone.0277267.r003

Decision Letter 1

Thomas Phillips

25 Jul 2022

PONE-D-21-31558R1The longitudinal association between patient empowerment and patient-reported outcomes: what is the direction of effect?PLOS ONE

Dear Dr. Acuña Mora,

The manuscript has been evaluated by two reviewers, and their comments are available below.

The reviewers note concerns about the statistical analyses presented and request re-analyses be completed.

Could you please carefully revise the manuscript to address all comments raised?

Please submit your revised manuscript by Sep 05 2022 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

We look forward to receiving your revised manuscript.

Kind regards,

Thomas Phillips, PhD

Staff Editor

PLOS ONE

Journal Requirements:

Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: (No Response)

Reviewer #2: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

Reviewer #1: Yes

Reviewer #2: (No Response)

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: (No Response)

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

Reviewer #1: Yes

Reviewer #2: (No Response)

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

Reviewer #1: Yes

Reviewer #2: (No Response)

**********

6. Review Comments to the Author

Reviewer #1: The authors fully addressed most of the comments. There is one issue on MCID (Page 10, line 172) that require more clarification from the authors. Specifically, they use the standard error (SE) less than 1 to reflect whether the scale is showing MCID. Since SE are sample size dependent, their values might not reflect clinical importance of the scale measurement results. For example, a very large study done in a population with very low QoL may yield a SE less than 1, and using SE one will erroneously conclude that the QoL in this population is high. In fact, SE in such a population is guaranteed to fall below 1 as long as you keep on increasing the sample size. On the other hand, a modest size study done on a population with high QoL (on average) may result in SE greater than 1. Perhaps you should consider using SD instead, as it reflects variability between subject measurements (on average) rather than the variability on a sample statistic.

Reviewer #2: (No Response)

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

**********

PLoS One. 2022 Nov 10;17(11):e0277267. doi: 10.1371/journal.pone.0277267.r004

Author response to Decision Letter 1

18 Aug 2022

Dr. Thomas Philips (Staff Editor)

PLOS ONE

Submission of manuscript

Dear Dr. Philips,

Please find attached the revised version of the manuscript entitled “The longitudinal association between patient empowerment and patient-reported outcomes: what is the direction of effect?”. We would like to resubmit this manuscript for consideration by PLOS ONE.

We thank you for the opportunity to revise and resubmit the manuscript. Furthermore, we would like to thank the reviewer for the constructive comment. Enclosed is a letter summarizing the rebuttal on the comment from the reviewer. Corrections and additions made in the manuscript have been marked using the “track changes” option of MS Word.

We hope that we have sufficiently addressed your comments. Thank you for considering this manuscript for publication in the PLOS ONE.

On behalf of all co-authors,

Mariela Acuña Mora.

Reviewer #1

The authors fully addressed most of the comments. There is one issue on MCID (Page 10, line 172) that require more clarification from the authors. Specifically, they use the standard error (SE) less than 1 to reflect whether the scale is showing MCID. Since SE are sample size dependent, their values might not reflect clinical importance of the scale measurement results. For example, a very large study done in a population with very low QoL may yield a SE less than 1, and using SE one will erroneously conclude that the QoL in this population is high. In fact, SE in such a population is guaranteed to fall below 1 as long as you keep on increasing the sample size. On the other hand, a modest size study done on a population with high QoL (on average) may result in SE greater than 1. Perhaps you should consider using SD instead, as it reflects variability between subject measurements (on average) rather than the variability on a sample statistic.

As the reviewer points out, the standard error (SE) is sample size dependent. This means that as the sample size increases, the smaller the SE becomes and vice versa. This characteristic of the SE can have an impact on the way the MCID is calculated and hence, interpreted. While other distribution-based methods to calculate the MCID are available, for instance methods that use the standard deviation (SD), there are also limitations when following this approach. One of the most relevant limitations of using the SD is that the MCID calculation may be sample dependent, and thus the calculation is less generalizable [1, 2]. MCIDs that rely on the SE allow for a higher generalizability of the scores [1, 2]. Another benefit of using the SEM is that it is expressed in the original metric of a measure, which facilitates the interpretation of the results [2].

Despite its limitations, the calculation of the MCID using the SE is frequently used in studies [2, 3, 4] and the available evidence suggests that this approach is reliable for identifying MCID [2]. It is also worth noting that the MCID varies based on the sample and the context. Therefore, it should not be accepted as a universal fact [1]. The interpretation of the MCID should be made with caution, while keeping in mind the limitations of the selected approach. One solution could be to use different approaches when reporting the MCID so the reader can make a comparison of the values based on different methods [1, 5]. We have reflected on this option but decided not do so, because it would bring the discussion on the MCID beyond perspective, and it would distract the reader from the main focus of the article, being the direction of effects. Thank you for your understanding.

References

1. Sedaghat AR. Understanding the Minimal Clinically Important Difference (MCID) of Patient-Reported Outcome Measures. Otolaryngology–Head and Neck Surgery. 2019;161(4):551-60.

2. Wyrwich, K.W., Tierney, W.M. & Wolinsky, F.D. Further Evidence Supporting an SEM-Based Criterion for Identifying Meaningful Intra-Individual Changes in Health-Related Quality of Life. J Clin Epidemiol. 1999; 52(9), pp.861–873.

3. Ousmen, A., Touraine, C., Deliu, N. et al. Distribution- and anchor-based methods to determine the minimally important difference on patient-reported outcome questionnaires in oncology: a structured review. Health Qual Life Outcomes. 2018; 16(228). https://doi.org/10.1186/s12955-018-1055-z

4. Mouelhi, Y. et al. How is the minimal clinically important difference established in health-related quality of life instruments? Review of anchors and methods. Health Qual Life Outcomes. 2020; 18(1), p.136.

5. Crosby RD, Kolotkin RL, Williams GR. Defining clinically meaningful change in health-related quality of life. J Clin Epidemiol. 2003;56(5):395-407.

Attachment

Submitted filename: Response to reviewers Rev2.docx

Click here for additional data file.^{(19.3KB, docx)}