Exploring the Intersection of Empathy and Behavior Change Techniques in Telehealth Coaching Messages: A Survey Study

doi:10.21203/rs.3.rs-3648815/v1

Background: The global rise in noncommunicable diseases (NCDs) demands innovative public health strategies. Key NCD risk factors are lifestyle-related, emphasizing the importance of behavior change (BC) interventions. Here, empathy plays a pivotal role by fostering rapport, enhancing patient-provider communication, and promoting an individual's motivation to change. However, the role of empathy in BC interventions and its potential intersection with BC techniques (BCTs) remains understudied, especially in telehealth. This study seeks to investigate this intersection in the context of asynchronous health coaching.

Methods: We conducted a two-round survey with 11 health coaches (HCs) who responded to 10 participant messages from the Bump2Baby and Me trial. In total, we examined 88 HC messages for empathic responses and the presence of techniques from the taxonomies of BC techniques (BCTs), Motivational BCTs (MBCTs), and Self-Determination Theory techniques (SDTTs).

Results: We identified 121 empathic responses that strongly overlapped with MBCTs (79%, n=213) in the form of Praise (47%, n=100) via unconditional positive regard and acknowledging patient perspectives and feelings, and Normalizing (23%, n=49) via neutral support and unconditional regard. While BCTs and SDTTs showed no substantial overlaps with empathic responses, 67% of BCTs elicited participants’ feelings and views about their coaching program.

Conclusions: The role of MBCTs in empathy expressions boosts patient engagement and self-perception, fostering relatedness. BCTs supported a patient-centered approach by empowering patients to articulate and plan their goals, which promotes autonomy. The lack of alignment in other categories suggests that considering empathy and BC techniques in isolation provides an incomplete picture of BC interventions.

Behavior change

empathy

telehealth

coaching

motivational interviewing

self-determination theory

The growing global burden of noncommunicable diseases (NCDs), which account for 7 in 10 deaths worldwide [1], demands novel approaches in public health systems around the world. Diet, physical activity, tobacco, sleep, and alcohol are the key risk factors for NCDs such as diabetes and cardiovascular disease [2]. Because of these high and rising numbers, there is a need for interventions that support individuals at risk of developing these conditions, to shift their lifestyle to help prevent, delay and more effectively manage these conditions. To address this, some countries like the UK have developed national guidelines [3] that require new behavior change (BC) programs to draw on “state of the art” theoretical frameworks and evidence base.

As part of the endeavour to increase replicability of BC and enable consistent synthesis and building an evidence base across studies, Michie and colleagues developed the first Behavior Change Techniques (BCTs) taxonomy or BCTTv1. This is an expert consensus-based approach for use in BC interventions design and evaluation [4] that consists of 93 techniques across 16 domains. The BCTs are the ‘active ingredients’ of distinct interventions that aim to support individuals to make health-related behaviors, such as healthy lifestyle habits (diet and physical activity) [5] or smoking cessation [6].

Whilst this first taxonomy, by Michie, sought to be exhaustive, there have been two additional taxonomies associated with two theoretical frameworks: techniques from a self-determination theory (SDT) perspective referred to as SDTTs [7], and techniques from motivational interviewing (MI) referred to as MITs [8]. According to SDT, the psychological factors and processes behind an individual’s motivation to change their behavior rely on the three main psychological needs: autonomy, competence, and relatedness [9]. MI prioritizes the intrinsic implications as opposed to the extrinsic factors (i.e., rewards and punishments). This perspective aligns with SDT’s emphasis on personally meaningful and relevant reasons to change for the individual [10]. Combining BCTs with SDTs and MITs has the potential to enhance autonomous motivation and amplify the effectiveness of healthcare BC interventions. Examining the interplay among these techniques offers a comprehensive understanding of the interactions between healthcare professionals (HCP) delivering the intervention and their patients. While there is some overlap with the BCT taxonomy, Teixeira et al. [11] investigated elements unique to these frameworks. There are 93 BCTs, 25 MBCTs, and 19 SDTTS. While BCTTv1 techniques are limited to changes in behavior, the MBCTs aim to elicit change in both motivation and behavior. In contrast, SDTTs satisfy the three psychological needs that prompt the motivation to change. When mapped against the BCT taxonomy, there are 7 MBCTs that overlap with BCTs, as shown in Table 1, mostly targeting competence. Regarding overlaps between BCTs and SDTTs, Gillison et al. [7] found overlaps for most of the SDTTS, of which 4 are unique and can be classified according to their SDT target for autonomy (use of non-controlling language, intrinsic goal orientation), competence (task climate), and relatedness (group co-operation).

Table 1. Overlapping MBCTs and BCTS from Teixeira et al. [11]

BCTs	MBCTs	Target
1.1 Goal setting (behavior)	MBCT 17. Assist in setting optimal challenge	Competence
8.7 Graded tasks	MBCT 17. Assist in setting optimal challenge	Competence
1.4 Action planning	MBCT 19. Help develop a clear and concrete plan of action	Competence
1.2 Problem-solving	MBCT 15. Address obstacles for change	Competence
2.2 Feedback on behavior	MBCT 21. Offer constructive, clear, and relevant feedback	Competence
2.3 Self-monitoring of behavior	MBCT 18. Promote self-monitoring	Competence
3.1 Social support	MBCT 14. Prompt identification and seek available social support	Relatedness
13.4 Valued self-identity	MBCT 4. Explore life aspirations and values	Autonomy

While a wealth of research supports the efficacy of evidence-based strategies in BC [5, 12], translating these into large-scale, real-world implementation remains a challenge. These strategies, initially designed to support physical interventions, are now facing an increased demand for understanding their effectiveness in digitally delivered interventions. The rapid advancement of telehealth demands a research force that, in addition to the factors that contribute to a successful BC intervention, assesses any differences that exist in a digital setting.

We know that face-to-face and digital interventions to manage or prevent NCDs can have a similar efficacy [13, 14], making it a reliable channel to deliver BC interventions. In this context, some authors have studied telehealth’s interactional features such as therapeutic alliance, participation, and empathy in remote [15] or text-based [16] consultations. Clinical empathy is the ability of a HCP to place themselves in the shoes of the patient to understand and participate in their feelings and emotional state [17]. Although research has shown that an empathic and positive communication greatly influences patient outcomes [18], there is limited evidence on how empathy unfolds in telehealth interactions and its implications for describing and testing BC interventions. To our knowledge, no one has explored this relationship before. While there is a BCT taxonomy, there is no established framework to measure empathy. The most extended approach relies on identifying emotional expressions in a patient’s discourse that their HCP may emphasize with, so called cues and concerns [19] or empathy opportunities (EOs) [20], which are often missed [21]. The empathy appraisal work by Pounds [22] adds a linguistic dimension based on Systemic Functional Linguistics (SFL) to this approach. SFL [23] explores the communicative and meaning potential of language production and structure, and the way it influences how individuals interact in diverse social contexts. Pounds’ empathy appraisal framework outlines seven forms of EOs found in patient’s cues and statements (Table 2).

Table 2. Description and examples of patient’s EOs adapted from Rey Velasco et al [24].

Empathy Opportunity (EO)	Category	Example
EO1	EXPLICIT EXPRESSIONS OF NEGATIVE FEELINGS, such as an emotive behavior or a mental state	‘I cried when I found out’
EO2	IMPLICIT EXPRESSION OF NEGATIVE FEELINGS through reference to a negative experience, such as fear, confusion, anxiety, or sadness	‘It’s been 3 days and I haven’t heard back from my GP’
EO3	EXPLICIT EXPRESSION OF NEGATIVE JUDGEMENT (others or self)	‘She is such an irresponsible person’
EO4	IMPLICIT EXPRESSION OF NEGATIVE JUDGEMENT (others or self)	‘I could have done better’
EO5e	EXPLICIT EXPRESSION OF POSITIVE SELF-JUDGEMENT	‘I am a good father’
EO5i	IMPLICIT EXPRESSION OF POSITIVE SELF-JUDGEMENT	‘I am eating healthier than ever!’
EO6	EXPLICIT EXPRESSION OF NEGATIVE APPRECIATION (things, events, actions).	‘The dinner was so boring’
EO7	IMPLICIT EXPRESSION OF NEGATIVE APPRECIATION (things, events, actions)	‘I am not sure this is something for me’

We combined this framework with an additional layer of SFL: the transitivity system [24]. This system allows us to systematically investigate how speakers construe their experience of the world by accounting for the participants involved, the processes (verbal groups), and the circumstances of this experience [25]. We conducted a pilot study [24] to explore the linguistic structure and meanings of EOs in text messages exchanged between clients and their coaches within a digital health coaching intervention for pregnant women at risk of gestational diabetes [26]. Our findings uncovered valuable insights that contribute to patient EO identification. In our study, most EOs were implicit, with negative ones seeking help and positive ones seeking praise. Enhancing the recognition of such opportunities has the potential to improve both the frequency and quality of empathy expressed by HCPs. The comprehensive understanding of patients’ statements and cues extends to the analysis of health coach (HCs) responses to patient EOs. Pounds described several categories for doctors’ empathic responses (Table 3); however, these categories are presented in isolation and lack an examination of the connection between these responses and each patient EO. We address this research gap in a survey study elsewhere [27] that forms the foundation for this article’s focus on the identification of BC techniques in HC responses. While the importance of empathy in behavior change (BC) interventions has been recognized, such as its role in fostering rapport and promoting the individual’s intention to change [28], the research on empathy and BC has largely evolved independently of each other. Only a systematic review conducted by Smith et al. [29] delved into the behaviors targeted and behavior change techniques (BCTs) employed in HCP' empathy training. However, the integration of BC techniques into the HCPs’ empathic expressions within the actual health interactions remains unexplored. Furthermore, to the best of our knowledge, and despite its acknowledged relevance [30], there are no explorations into the intersection of these approaches in the context of telehealth coaching. These two approaches to the interaction raise the question of to how they might overlap, interact, or complement each other when investigating empathy within BC interventions, and whether their categories are redundant when studying the patient-HCP interaction. In this paper, we will respond to those questions by examining the connections and overlaps between Pounds' empathy appraisal categories for HCP responses and the BC techniques. We expect our findings to enlighten the link between empathy and BC and contribute to the design of future telehealth interventions.

Table 3. Classification of doctors’ verbal empathic expressions [27], adapted from Pounds [22].

	Category	Description and examples
RESPONDING TO PATIENT CUES	Expressing explicit understanding or acknowledgement of patients' feelings and views	Formulations including verbs of acknowledgement (‘I understand/see/realise/appreciate…’), adjectival constructions expressing understanding (‘It is clear/apparent… to me’), or alternative formulations (‘It strikes me that…’, ‘I am aware/conscious that…’).
	Sharing the patient’s feelings or views through expressions of agreement (‘emotive empathy’)	Shared feelings (‘I would also… if I were you’), shared feelings through interjection and intonation (‘Oh no!’), shared judgement (‘Yes, your boss could have been more understanding’), shared appreciation (‘Yes, this is a difficult exercise’)
	Expressing acceptance in response to patients’ explicit, implicit or potential negative or positive self-judgement	Unconditional positive regard (or ‘praise’) through: Explicit expression of positive judgement of the patient as a person (‘You are a fantastic mom’) Implicit expression of positive judgement through explicit positive appreciation of the patient’s actions or thoughts (‘It looks like you are making great progress’) Repetition or paraphrasing of patients’ words and avoidance of immediate countering statements or premature reassurance Allowing patients to express their feelings and views fully through minimal responses, nodding, and avoidance of interruption.
		Neutral support (even when approval cannot be granted, withholding judgement): Explicit appreciation of the patient’s behavior, ideas or feelings in terms of their ‘normality’ and ‘acceptability’ (‘It is (completely) normal/not unusual/acceptable…to do/think/feel X’, ‘It is not (at all) surprising/crazy to do/think/feel X’) Explicit expressions of judgement when denying potential negative self-assessment by the patient (‘You are not odd, bad, crazy…for doing/thinking/feeling X…’)
ELICITING PATIENTS’FEELINGS AND VIEWS	Direct elicitation of patients' feelings	Enquiries about mental state (‘How/what do/did you feel/think/expect?’) or emotive behavior (‘How did/do you react?’)
	Indirect elicitation of patients' feelings through:	Questions about potentially negative/critical experiences (‘Did you have a happy childhood?’)
		Invitation of patients' confirmation, rejection or clarification of interpreted affectual states (‘You seemed concerned when I mentioned diabetes’)
		Enquiries about the 'emotive behavior' displayed (‘Why are you crying?’)
	Direct elicitation of patients’ judgement	‘How are you finding your coach?’
	Indirect elicitation of patients’ judgement through:	Enquiries about others or patient's behavior (‘How did your husband react?’)
	Indirect elicitation of patients’ judgement through:	Invitation of patients’ confirmation, rejection or clarification of interpreted views of others or themselves (‘You seem very motivated to start your health programme’)
	Direct elicitation of patients’ appreciation	‘How would you rate your progress so far?’
	Indirect elicitation of patients’ appreciation through:	Enquires about therapy or medication (‘Would you like us to review your goals together?’)
	Indirect elicitation of patients’ appreciation through:	Invitation of patients’ confirmation, rejection or clarification of interpreted view of things, events and actions (‘You do not seem to find that book so useful’)
	Indirect elicitation of patients’ feelings and views through formulations in which feelings and evaluations (potential or real) are attributed to third parties who might find/found themselves in circumstances similar to the patient's	‘Many people in your position would be quite annoyed [affect]/would find this annoying [appreciation].’, ‘My sister went through something similar and she struggled to see the benefits of [affect]/to value [appreciation] that opportunity’

Study participants

This study took place within Liva Healthcare A/S (Liva), a tele-health coaching service. All Liva HC’s hold a health-related degree (primarily registered dietitians, but also nursing, physiotherapy, midwifery and psychology) and receive a state-of-the-art training on the BC framework that is 13 hours long at the time of recruiting. The training program includes BC theories and frameworks such as SDT, MI, and the transtheoretical model of health BC [31] (see Figure 1) and their associated coaching strategies and BCTs. Moreover, where necessary, HCs receive an additional program-specific training, e.g., in the Bump to Baby and Me trial [26], HCs received 27.5 hours of training.

Our recruitment criteria encompassed being a Liva health coach (HC) and offering coaching in English. We recruited participants via a dedicated Slack®[32] network channel utilized by Liva HCs. Among the 35 eligible HC candidates during recruitment, 11 decided to participate and one participant withdrew. To be enrolled in the study, the coaches had to complete a comprehensive survey. One coach did not complete the full survey and was therefore not part of the analysis.

Participant messages

We extracted the 10 survey messages from a sample of 236 participant messages from the Irish and English intervention arms of the Bump2Baby and Me trial that was available when we prepared the survey (August 2022). The 10 messages (Appendix 1) met the following inclusion criteria: representing all the EO categories overall and not being ambiguous, i.e., including enough information to allow the HC to respond to the message in isolation. In preparation for the analysis, we split the sentences into clauses, being a group of words that includes a subject and a verb, [33] as our unit of analysis. Since different clauses express different ideas that the HC can approach, we used the term subEOs when there was more than one clause in the same EO. For example, we find three clauses or subEOs in the following EO2 (Implicit expression of negative feelings): “I'll keep my goals atm the same (10.1)| just not to push my self more for something I can't accomplish (10.2)| and| not having then self let down moments (10.3)". HCs may address all three or focus on the one they consider the most important. Additionally, we split EO5 (Expressions of positive self-judgement) into two subcategories that differentiated its explicit from its implicit form. Table 4 shows an overview of the subEO distribution in the survey messages.

Coach messages

To assess HC responses, we developed two different surveys to be completed in relation to 10 patient messages from the Bump2baby trial [34]. The first round focused on the most appropriate techniques (BCTs, MITs, and SDTTs) that HCs would use in a written response according to their coaching background. We compiled the responses and included the techniques that were selected by ≥50% of HCs for each message to build the second-round survey. In the second-round survey, participants were asked to build a response to the messages that included as many BCT techniques they considered suitable. Additionally, we provided an open-response box in both rounds to allow HCs for comments and suggestions on the techniques. Once all the HCs filled in the survey, we imported their anonymized responses to MAXQDA Analytics Pro 2022, a qualitative analysis software [35]. ERV coded the text segments according to the EO response, empathic verbal expressions, and BC techniques (BCTs, MITs, SDTTs) categories. A second coder provided reliability coding.

Table 4. EOs and subEOs distribution in the patient messages included in the survey [27].

*EO categories*	*Patient messages, n (%)*	*Including subEOs, n (%)*
*EXPLICIT EOS*	19 (21.8)	27 (22.3)
*Positive*	2 (10.5)	2 (7.4)
*EO5e – Self-judgement*	2 (2.2)	2 (1.6)
*Negative*	17 (89.5)	25 (92.6)
*EO1 – Feelings*	7 (7.5)	9 (7.1)
*EO3 – Judgement (others or self)*	4 (4.3)	7 (5.5)
*EO6 – Appreciation (things, events, actions)*	6 (6.5)	9 (7.1)
*IMPLICIT EOs*	68 (78.2)	94 (77.7)
*Positive*	20 (29.4)	29 (16.3)
*EO5i – Self-judgement*	20 (21.5)	29 (22.8)
*Negative*	48 (70.6)	65 (83.7)
*EO2 – Feelings*	14 (15.1)	20 (15.7)
*EO4 – Judgement (others or self)*	13 (14.0)	15 (11.8)
*EO7 – Appreciation (things, events, actions)*	21 (22.6)	30 (23.6)
*TOTAL*	87 (100)	121 (100)

Coding categories

We coded the BC techniques and empathic expressions found in the HC messages obtained from the survey. In preparation for our analysis, we mapped the overlaps identified by Teixeira [11] and Gillison [7] for BC, MI, and SDT techniques to identify those that were unique and applicable to our study setting (written interaction in a health coaching program). In addition, we identified the primary SDT target associated to motivation for each technique as autonomy, competence, or relatedness [7]. Our assessment resulted in 28 techniques (Table 5) for HC messages coding.

Table 5. Relevant BCTs mapped against MI and SDT techniques, as well as their primary target, according to Teixeira [11] and Gillison [7].

Grouping and BCTs	MBCTs	SDTs	Primary target
1.1 Goal-setting (goal)	MBCT 17. Explore barriers and obstacles	SDT6: Structure	Autonomy
1.2. Problem-solving	MBCT 15. Ask permission to provide information or give advice	SDT14: Barrier identification	Competence
1.3. Goal-setting (outcome)		SDT6: Structure	Autonomy
1.4. Action planning	MBCT 19. Assist in setting optimal challenge	SDT6: Structure	Autonomy
1.5. Review behavior goal(s)		SDT12: Provide informational feedback	Competence
1.6. Discrepancy between current behavior and goal
1.7. Review outcome goal(s)		SDT12: Provide informational feedback	Competence
1.9. Commitment
2.3. Self-monitoring of behavior	MBCT 18. Clarify expectations	SDT12: Provide informational feedback	Competence
2.4. Self-monitoring of outcome(s) of behavior		SDT12: Provide informational feedback	Competence
2.6. Biofeedback		SDT12: Provide informational feedback	Competence
3.1. Social support (unspecified)	MBCT 14. Use attentive, reflective listening	SDT17: Encourage social support seeking	Relatedness
4.2. Information about antecedents		SDT3: Provide a rationale SDT13. Provide information	Autonomy Competence
4.3. Re-attribution
5.1. Information about health consequences		SDT3: Provide a rationale SDT13. Provide information	Autonomy Competence
5.4. Monitoring of emotional consequences
5.6. Information about emotional consequences		SDT3: Provide a rationale SDT13. Provide information	Autonomy Competence
7.1. Prompts/cues
8.2. Behavior substitution
8.3. Habit formation
8.4. Habit reversal
8.6. Generalisation of target behavior
8.7. Graded tasks	MBCT 17. Explore barriers and obstacles	SDT11: Provide optimal challenge	Competence
9.1. Credible source
11.2. Reduce negative emotions
11.3. Conserving mental resources
12.1. Restructuring the physical environment
12.2. Restructuring the social environment		SDT7: Emphasise responsibility	Autonomy
12.4. Distraction
12.6. Body changes
13.1. Identification of self as role model
13.2. Framing/reframing
15.1. Verbal persuasion about capability		SDT15: Provide support and encouragement	Competence
15.4. Self-talk
	MBCT 8. Acknowledge and respect perspectives and feelings
	MBCT 9. Encourage asking of questions
	MBCT 10. Show unconditional regard
	MBCT 11. Demonstrate/show interest in the person
	MBCT 12. Use empathic listening
		SDT4: Use of non-controlling language	Autonomy
		SDT5: Intrinsic goal orientation	Autonomy
		SDT10: Task climate	Competence

We applied Pounds’ framework for doctors’ potential empathic verbal expressions to code HC’s empathic responses [22]. This framework has two categories: Responding to patient cues and Eliciting patient’s feelings and views (Table 3). Our criterion for an empathic response coding was that the HC message contained a keyword from the patient EO or subEO. Additionally, we created a “General” category for those empathic responses that did not contain a keyword and/or could be connected to multiple EOs or subEOs. In our Results table, we only show the categories that were present in our dataset to facilitate its interpretation.

We identified 121 empathic responses in our dataset (n= 88 messages from 9 HCs) in response to patient EOs (n=10 patient messages). While there was no difference in the overall response ratio (calculated by dividing the number of subEOs by the number of not missed subEOs) for implicit (89) and explicit (89) EOs, this value was higher for positive implicit EOs (97) and negative explicit EOs (92). As shown in Table 6, there were 15 subEOs that none of the HCs addressed and which were mainly implicit (n=12, 80% missed subEOs).

Table 6. Overview of empathic responses [27]. For each category, the Response Ratio was calculated by dividing the number of subEOs by the number of not missed subEOs.

	*Missed subEOs* *n (% missed EOs)***	*Response Ratio*
*Explicit EOs (n=27)*	3 (20)	89
Positive (n=2)	1 (7)	50
Negative (n=25)	2 (13)	92
*Implicit EOs (n=94)*	12 (80)	87
Positive (n=29)	1 (7)	97
Negative (n=65)	11 (73)	83
*TOTAL (n=121)*	15 (100)	85

Table 7 shows the overlap between the coded categories overall, whether they were part of an empathic response or not. There are high overlaps between the techniques and Pounds’ categories. 79% (n=213) of MBCTs overlap with the Responding to patient cues coding, 46% (n=91) of BCTs overlap with Eliciting patient feelings and views coding, and 55.8% (n=29) of SDTs overlap with Eliciting patient feelings and views coding.

Table 7. Overlap between categories overall, showed as n, (%).

Techniques	Overlap with Responding patient cues (n=280)	Overlap with Eliciting patient feelings and views (n=147)	With BCTs	With MBCTs	With SDTs	No overlap
BCT (n=198)	13 (6.6)	91 (46)	5 (2.5)	2	2	86 (43.4)
MBCT (n=268)	213 (79.4)	39 (14.2)	2 (0.75)	0	1 (0.4)	13 (4.9)
SDT (n=52)	7 (13.5)	29 (55.8)	2 (3.8)	1 (1.9)	0	13 (25)

We then identified the overlaps between Pounds’ categories and the techniques for the same text segment. Table 8 shows the overlaps between Techniques and Responding to patient cues. In the MBCT category, 47% (n=100) of all identified MBCT overlaps for the same text segment correspond to the Praise category, which is a form of unconditional positive regard, mostly by Acknowledging/respecting patient’s perspectives/feelings (MBCT 8) (n=83, 39% all MBCTs) (“Well done on sustituing unhealthy snacks for fruits!!| That's a great idea”). Similarly, a 23% (n=49) of all MBCTs overlap with Normalizing through an explicit appreciation, which is a form of neutral support, generally by Showing unconditional regard (MBCT 10, n=33, 16%) (“it is okay if every meal or snack isn't as 'healthy' as you may hope,| the occasional treat is fine,| consistency is better than perfection!|”). We also found MBCT10 overlaps in response to patient cues through Praise (n=14, 7%) (“No need to worry about anything else,| you've been busy at being a brilliant mum and bringing your boy into the world!”) and Understanding/acknowledging patient’s feelings and views (n=14, 7%) (“I understand that logging your goals isn't your top priority at the moment.”). The remaining techniques categories, i.e., BCTs and SDTs, showed no substantial overlaps with the Responding to patient cues categories.

Table 9 shows the overlap between Techniques and Eliciting patient feelings and views. Most of these overlaps are related to the therapy or medication (in our study, the coaching program and/or participants’ goals), by eliciting an indirect appreciation. This category overlaps with a 67% of all BCTs, mainly through Goals and planning BCTs such as Problem-solving (BCT 1.2) (n=12, 12%) (“What could be an alternative behavior or activity that you could do when the weather isn't great for a walk?”) and Action planning (BCT 1.4) (n=11, 11%) (“During your busier weeks, have you considered menu planning before the week kicks off?”). The same category overlaps with a 45% (n=17) of all MBCTs, mostly by demonstrating interest in the patient (MBCT 11) (n=11, 29%) (“I have attached the guidelines on physical activity in pregnancy. | You can let me know how you compare!”), and 86% (n=24) of all SDTs, mainly through use of noncontrolling language (SDT 4) (n= 15, 54%) (“For step tracking, could we consider an alternative, such as setting a goal for walking in exercise instead? | What would a good target be in your opinion?”).

Table 8. Overlap between techniques and Pounds’ coding for responding to patient cues in empathic responses. The empty cells represent a 0 value.

Table 9. Overlap between techniques and Pounds’ coding for eliciting patient feelings and views in empathic responses. Percentages are shown for each technique category. The empty cells represent a 0 value.

Our study contributes to a better understanding of the role of empathy in digital BC interventions by uncovering significant overlaps across the coding categories. These findings hold implications for both clinical practice and theoretical frameworks. While substantial overlaps exist between specific categories, the lack of convergence within the remaining categories suggests that interpreting empathy and BC codes in isolation would be an incomplete approach. Instead, these categories may gain a richer meaning and complement each other in the BC intervention if we consider them together.

A substantial proportion (79.4%) of MBCTs overlapped with Responding to patient’s cues coding, mostly with the praise category (n = 100, 47%). Praise is a positive reinforcement strategy that increases intrinsic motivation by targeting competence and perceived autonomy in the patient, supported by SDT, which translates into better engagement and performance [36]. This contributes to counteract negative feelings, and promote a more positive view of ourselves (self-judgement and self-confidence) [37]. In our study, HCs expressed praise in the form of patient perspectives/feelings acknowledgement or respect (MBCT 8), which validates patients' emotions and supports their efforts to change, e.g., to adopt healthier eating habits (“Well done on thinking on down on sugary things and bread, you're thinking of the best options for your health, congratulations!”). Previous research underlines the pivotal role of this empathic response in constructing therapeutic rapport and fostering an environment where patients feel secure expressing themselves [38]. Similarly, the overlap between showing unconditional regard (MBCT 10), which targets relatedness, and Normalizing through explicit appreciation demonstrate HCs’ efforts to offer neutral support by reassuring patients that occasional deviations from health goals are acceptable (“It is hard to focus on new things when we have a lot of stress going on.”). Additionally, we identified another overlap between showing unconditional regard (MBCT10) and expressing Agreement by sharing feelings and views with the patient. Normalizing and agreeing are two interesting ways to express empathy, since the HC is unlikely to experience the same feelings as the patient but could agree on the patients’ negative appreciation about an item, an action, or an event (“It's very understandable with the poor weather and feeling tired that it's affecting your mood, and consequently, you are snacking more”). Davis [39] identified these as ‘cognitive’ and ‘emotive’ empathy, respectively. He also differentiated two types of emotive empathy: ‘parallel’, by expressing a similar perspective, and ‘reactive’, by describing the same feelings experienced in similar circumstances.

On the other hand, the low overlap between BCTs and Responding to patient cues categories suggests that while these techniques may contribute to the overall intervention, their association with empathic responses is very low. However, we found a significant overlap (n = 65, 67%) between BCTs and Eliciting patient feelings and views about the coaching program or goals category. In this empathic response, the HCs adopt a patient-centered approach to collaborate with patients to articulate and plan their goals, fostering an environment of empowerment and partnership [40]. Similarly, the overlap between MBCTs and Eliciting patient feelings and views coding highlighted HCs’ efforts to demonstrate interest in the patient (MBCT 11, ‘What is your favourite family meal to cook?’). This empathic expression fosters commitment and the sense of being heard by their HC, which aligns with research suggesting that eliciting and validating patients' perspectives plays a vital role in enhancing patient engagement and catalyzing BC [41]. Overall, these findings reinforce empathy's role in boosting patients' self-esteem and motivation, aligning with the principles of SDT that emphasize autonomy, competence, and relatedness [9], particularly within the realm of MBCTs. By acknowledging and respecting patients' feelings, HCs establish a trust and collaboration relationship, which may enhance the effectiveness of BC techniques. There were several cases of overlapping and co-occurring codes across coding categories, which will be the focus of our next research work. Considering that this is the first study of its kind, some codes may be unnecessary or redundant when combining empathy appraisal and BC approaches. However, it's worth considering that the expressions of empathy may influence the effectiveness of BC techniques, especially the BCTs, which showed the lowest overlap. While data from face-to-face doctor-client interactions suggest this influence [42, 43], further exploration is needed to determine whether BCT efficacy is indeed moderated by the presence of empathy in text-based interactions. Nevertheless, the complexity of these relationships needs further exploration, considering variables such as different populations, HC training, and therapeutic contexts.

Ethical approval and consent to participate

The Bump2Baby and Me trial [34], where the authors are authorized researchers, is the source of the dataset we used to obtain the 10 participant messages. The trial received ethical approval from all study sites (Dublin: National Maternity Hospital Ethics Committee ref EC18.2020; Bristol: Wales Research Ethics Committee ref 21/WA/0022; Granada: CEIM/CEI Provincial de Granada; Melbourne: Monash Health Human Research Ethics Committee ref RES-20-0000-892A). To preserve participants’ anonymity, we removed any personal information from the messages.

For the HC survey, we prepared a participant consent form and received approval from The Institutional Ethical Review Board, University of Copenhagen, Department of Psychology. Approval number: IP-IRB / 08072022. All approved applications follow i) the 1975 Helsinki Declaration as revised in 2008 (including §24 on informed consent, and – for clinical studies – §19 on clinical trial preregistration); ii) APA ethical standards in publishing (APA Ethical principles, 2010, Section 8: Research and publication; including 8.14 on data retention and sharing), as well as national ethical guidelines. Informed consent was obtained from all participants. Once a HC consented to participate, they were assigned an ID number that we used throughout the study's duration to ensure data anonymity. We administered the survey using SurveyXact® (Rambøll Group A/S, Copenhagen), and the platform automatically anonymised HCs' replies upon survey completion. Enrolled HCs received a code after completing the survey to claim compensation for their participation. We did not collect any other data from HCs.

Consent for publication

Not applicable.

Availability of data and materials

The data supporting the findings of this manuscript are available upon request. Requests for access to the data can be directed to ERV and will be subject to approval based on compliance with ethical and legal considerations.

Competing interests

ERV and DHL are employees of Liva Healthcare, which provides the coaching services described in the article. TS has no competing interests.

Funding

This research is part of an industrial PhD project sponsored by Innovation Fund Denmark, the University of Copenhagen, and Liva Healthcare. The Impact Diabetes B2B (Bump2Baby and Me) project has received funding from the European Union’s Horizon 2020 research and innovation program (grant agreement no. 847984). The project also acknowledges collaborative funding from the National Health and Medical Research Council (grant no. 1194234).

Authors’ contributions

ERV designed and performed the study, including gaining ethical approval, recruiting participants, and analysing and synthesising data. DL provided support in drafting and reviewed the final manuscript. TS conceived the research, supported the interpretation and discussion of results, and reviewed the final manuscript. ERV wrote the first draft of the manuscript. All authors approved the final version of the manuscript.

Acknowledgments

We would like to acknowledge the contributions of Assoc. Prof. Zsofia Demjen (University College London) to the reliability coding, data analysis, and interpretation processes of this study. We would also like to acknowledge the contribution of the members of the Impact Diabetes B2B Collaboration Group: Assoc. Prof. Sharleen L. O'Reilly, Prof. Fionnuala M. McAuliffe, Assoc. Prof. Mary Codd, and Assoc. Prof. Ricardo Segurado (University College Dublin), Prof. Helle T. Maindal and Nanna Husted Jensen (Aarhus University), Dr. Anna Davies and Assoc. Prof. Christy Burden (University of Bristol), Prof. Karsten Vrangbæk and Dr. Laura Elina Pirhonen (University of Copenhagen), Assoc. Prof. Alberto Puertas, Dr. Francisca S. Molina, and Dr. María García-Ricobaraza (University of Granada), Katie Angotti, and Signe Boldt Bensen (Liva Healthcare), Prof. Karen J Campbell, Assoc. Prof. Rachel Laws, and Assoc. Prof. Vincent L Versace (Deakin University), Prof. Helena Teede, Assoc. Prof. Cheryce L Harrison, Assoc. Prof. Jacqueline Boyle, and Dr Georgia Soldartis (Monash University).

(WHO) WHO. World health statistics 2023: monitoring health for the SDGs, sustainable development goals. Available at: https://www.who.int/publications/i/item/9789240074323. 2023.
Budreviciute A, Damiati S, Sabir DK, Onder K, Schuller-Goetzburg P, Plakys G, et al. Management and Prevention Strategies for Non-communicable Diseases (NCDs) and Their Risk Factors. Front Public Health. 2020;8:574111.
(NICE) NIfHaCE. National Institute of Health and Care Excellence (NICE). Behaviour change: individual approaches. URL: https://www.nice.org.uk/guidance/ph49. 2014.
Michie S, Richardson M, Johnston M, Abraham C, Francis J, Hardeman W, et al. The behavior change technique taxonomy (v1) of 93 hierarchically clustered techniques: building an international consensus for the reporting of behavior change interventions. Ann Behav Med. 2013;46(1):81-95.
Cradock KA, G OL, Finucane FM, Gainforth HL, Quinlan LR, Ginis KA. Behaviour change techniques targeting both diet and physical activity in type 2 diabetes: A systematic review and meta-analysis. Int J Behav Nutr Phys Act. 2017;14(1):18.
Gainforth HL, Lorencatto F, Erickson K, West R, Michie S. Characterizing clients' verbal statements in behavioural support interventions: The case of smoking cessation. British Journal of Health Psychology. 2016;21(3):600-12.
Gillison FB, Rouse P, Standage M, Sebire SJ, Ryan RM. A meta-analysis of techniques to promote motivation for health behaviour change from a self-determination theory perspective. Health Psychology Review. 2019;13(1):110-30.
Hardcastle SJ, Fortier M, Blake N, Hagger MS. Identifying content-based and relational techniques to change behaviour in motivational interviewing. Health Psychology Review. 2017;11(1):1-16.
Ryan RM, Deci EL. Self-Determination Theory. In: Maggino F, editor. Encyclopedia of Quality of Life and Well-Being Research. Cham: Springer International Publishing; 2020. p. 1-7.
Miller WR, Rollnick S. Meeting in the middle: motivational interviewing and self-determination theory. Int J Behav Nutr Phys Act. 2012;9:25.
Teixeira PJ, Marques MM, Silva MN, Brunet J, Duda JL, Haerens L, et al. A classification of motivation and behavior change techniques used in self-determination theory-based interventions in health contexts. Motivation Science. 2020;6:438-55.
An S, Song R. Effects of health coaching on behavioral modification among adults with cardiovascular risk factors: Systematic review and meta-analysis. Patient education and counseling. 2020;103(10):2029-38.
Wu C, Wu Z, Yang L, Zhu W, Zhang M, Zhu Q, et al. Evaluation of the clinical outcomes of telehealth for managing diabetes: A PRISMA-compliant meta-analysis. Medicine (United States). 2018;97(43).
Berry MP, Sala M, Abber SR, Forman EM. Incorporating automated digital interventions into coach-delivered weight loss treatment: A meta-analysis. Health psychology : official journal of the Division of Health Psychology, American Psychological Association. 2021;40(8):534-45.
Irvine A, Drew P, Bower P, Brooks H, Gellatly J, Armitage CJ, et al. Are there interactional differences between telephone and face-to-face psychological therapy? A systematic review of comparative studies. Journal of Affective Disorders. 2020;265:120-31.
Pounds G, De Pablos-Ortega C. Patient-centred communication in British, Italian and Spanish Ask-the-Expert healthcare websites. Communication and Medicine. 2015;12(2-3):225-41.
Piasecki M. Clinical communication handbook. Blackwell Science. 2003.
Howick J, Moscrop A, Mebius A, Fanshawe TR, Lewith G, Bishop FL, et al. Effects of empathic and positive communication in healthcare consultations: a systematic review and meta-analysis. J R Soc Med. 2018;111(7):240-52.
Piccolo LD, Finset A, Mellblom AV, Figueiredo-Braga M, Korsvold L, Zhou Y, et al. Verona Coding Definitions of Emotional Sequences (VR-CoDES): Conceptual framework and future directions. Patient education and counseling. 2017;100(12):2303-11.
Bylund CL, Makoul G. Examining empathy in medical encounters: an observational study using the empathic communication coding system. Health Commun. 2005;18(2):123-40.
Hsu I, Saha S, Korthuis PT, Sharp V, Cohn J, Moore RD, et al. Providing support to patients in emotional encounters: A new perspective on missed empathic opportunities. Patient education and counseling. 2012;88(3):436-42.
Pounds G. Empathy as ‘appraisal’: A new language-based approach to the exploration of clinical empathy. Journal of Applied Linguistics and Professional Practice. 2011;7(2):145-68.
Halliday MAK, Halliday MAK, Matthiessen CMIM, Matthiessen CMIM. An introduction to functional grammar. 3. ed. / rev. by Christian M.I.M Matthiessen. ed. London: Arnold; 2004.
Rey Velasco E, Pedersen HS, Skinner T. Analysis of Patient Cues in Asynchronous Health Interactions: Pilot Study Combining Empathy Appraisal and Systemic Functional Linguistics. JMIR Formative Research. 2022;6(12):e40058.
Martin JR, Matthiessen CMIM, Painter C. Working with Functional Grammar: Arnold; 1997.
O'Reilly SL, Burden C, Campoy C, McAuliffe FM, Teede H, Andresen J, et al. Bump2Baby and Me: protocol for a randomised trial of mHealth coaching for healthy gestational weight gain and improved postnatal outcomes in high-risk women and their children. Trials. 2021;22(1):963.
Rey Velasco E, Demjen Z, Skinner T. Digital Empathy in Behaviour Change Interventions: A Survey Study on Health Coach Responses to Patient Cues. [Manuscript submitted for publication]. 2023.
Abdulrahman A, Richards D, Ranjbartabar H, Mascarenhas S. Verbal empathy and explanation to encourage behaviour change intention. Journal on Multimodal User Interfaces. 2021;15(2):189-99.
Smith KA, Bishop FL, Dambha-Miller H, Ratnapalan M, Lyness E, Vennik J, et al. Improving Empathy in Healthcare Consultations—a Secondary Analysis of Interventions. Journal of General Internal Medicine. 2020;35(10):3007-14.
Virtanen L, Kaihlanen A-M, Laukka E, Gluschkoff K, Heponiemi T. Behavior change techniques to promote healthcare professionals’ eHealth competency: A systematic review of interventions. International Journal of Medical Informatics. 2021;149:104432.
Prochaska JO, Velicer WF. The Transtheoretical Model of Health Behavior Change. American Journal of Health Promotion. 1997;12(1):38-48.
Johnson HA. Slack J Med Libr Assoc 2018; 106(1): 148-151 2 Slack Technologies Inc: Slack Available at wwwslackcom. 2018.
Crystal D. A Dictionary of Linguistics and Phonetics: Wiley; 2011.
O'Reilly SL, Laws R, Maindal HT, Teede H, Harrison C, McAuliffe FM, et al. A Complex mHealth Coaching Intervention to Prevent Overweight, Obesity, and Diabetes in High-Risk Women in Antenatal Care: Protocol for a Hybrid Type 2 Effectiveness-Implementation Study. JMIR Res Protoc. 2023;12:e51431.
VERBI-Software. MAXQDA 2022 [computer software]. Berlin, Germany: VERBI Software. Available from maxqda.com. 2021.
Koestner R, Zuckerman M, Koestner J. Praise, involvement, and intrinsic motivation. Journal of Personality and Social Psychology. 1987;53:383-90.
Soenens B, Vansteenkiste M. Understanding the complexity of praise through the lens of self-determination theory. 2020. p. 27-35.
Decety J, Fotopoulou A. Why empathy has a beneficial impact on others in medicine: unifying theories. Front Behav Neurosci. 2014;8:457.
Davis MH. Empathy: A social psychological approach. Boulder, CO, US: Westview Press; 1994. x, 260-x, p.
Luxford K. Patient-centred care: improving quality and safety by focusing care: Australian Commission on Safety and Quality in Health Care; 2010.
Street RL, Jr., Makoul G, Arora NK, Epstein RM. How does communication heal? Pathways linking clinician-patient communication to health outcomes. Patient education and counseling. 2009;74(3):295-301.
Moller AC, Merchant G, Conroy DE, West R, Hekler E, Kugler KC, et al. Applying and advancing behavior change theories and techniques in the context of a digital health revolution: proposals for more effectively realizing untapped potential. Journal of behavioral medicine. 2017;40(1):85-98.
Hojat M, Shannon SC, Desantis J, Speicher MR, Bragan L, Calabrese LH. Empathy in medicine: National norms for the jefferson scale of empathy: A nationwide project in osteopathic medical education and empathy (POMEE). Journal of the American Osteopathic Association. 2019;119(8):520-32.

Competing interest reported. ERV and DHL are employees of Liva Healthcare, which provides the coaching services described in the article. TS has no competing interests.

APPENDICES.docx

Exploring the Intersection of Empathy and Behavior Change Techniques in Telehealth Coaching Messages: A Survey Study

Status:

Version 1

Abstract

Figures

INTRODUCTION

METHODS

Study participants

Participant messages

Coach messages

RESULTS

DISCUSSION

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1