Applying the theory of planned behaviour to understand mental health professionals’ intentions to work using a human rights-based approach in acute inpatient settings

Abstract

Background

There has been a shift to implement human rights-based approaches in acute mental health care due to increasing concerns around quality of care. National Health Service (NHS) Trusts have a legal duty to uphold a person’s human rights, therefore it is important to understand what any barriers might be. Using psychological theory may help to develop this understanding.

Aim

To test whether the theory of planned behaviour can be an effective model in understanding mental health professionals’ intentions to work using a human rights-based approach.

Method

Participants were recruited from two NHS Trusts in the North West of England. A cross-sectional, survey design was used to examine mental health professionals’ intentions to use human rights-based approaches.

Results

Multiple regression analyses were performed on the theory of planned behaviour constructs showing that attitude and subjective norm significantly predicted intention. Perceived behavioural control did not add any significant variance, nor any demographic variables.

Conclusion

There could be factors outside of the individual clinician’s control to fully work within a human rights-based framework on acute mental health wards. The theory of planned behaviour offers some understanding, however further development work into measuring human rights outcomes on acute mental health wards is needed.

Keywords:

• Human rights-based approaches
• theory of planned behaviour
• perceived behavioural control
• mental health professionals
• acute mental health care

Introduction

The Human Rights Act 1998 (Great Britain Human Rights Act, 1998) legally protects human rights for those receiving care and treatment in mental health hospitals in the UK, however failings in upholding people’s dignity and rights have been found in these settings (Parliamentary & Health Service Ombudsman, 2018). Human rights should reflect the minimum standard of treatment in relation to physical, psychological and social wellbeing and therefore, all public organisations, including the National Health Service (NHS), must comply with the Human Rights Act 1998. The rights of individuals who are detained under a section of the Mental Health Act (Great Britain Mental Health Act, 1983) can be restricted under principles of protection of health, or public safety, but only if it is deemed proportionate and is the least restrictive option (EQHRIA, n.d.). Unfortunately, quality of care within mental health settings does not always meet these minimum standards (Care Quality Commission, 2018). It is necessary to understand potential causes of these failings to effectively improve services, including any impacting psychological factors. There is currently little research published on how mental health professionals perceive human rights or integrate them into their practice and outcome measures are scarce.

The Department of Health (Department of Health, 2007) published a framework to assist NHS Trusts to implement a human rights-based approach (HRBA) to healthcare (2007), outlining five key principles to better enable good quality and efficient healthcare. When using a HRBA, values such as fairness, respect, equality, dignity and autonomy (FREDA) should be proactively placed at the centre of all clinical decision making (Curtice & Exworthy, 2010). It is assumed that healthcare professionals already possess these core values and therefore the FREDA framework is more easily implemented in healthcare settings. Human rights-based frameworks have been criticised however for being too vague and defined too broadly. Evaluative issues also exist as the concepts do not lend themselves to tangible measurement. Being too conceptual in nature makes it difficult to hold those who violate the approach to account, risking its integrity and usefulness (Batliwala, 2007). Kinderman and Butler (2006) cite these issues as reasons why rights-based frameworks are not fully implemented into public services.

The theory of planned behaviour (Ajzen, 1991) is a psychological theory that proposes that having intention to act is a prerequisite for a behaviour to occur and has a large evidence base in individual health-related behaviour change (Albarracin et al., 2001; Godin & Kok, 1996; Hagger et al., 2002; Hausenblas et al., 1997; Sheeran & Taylor, 1999). The theory outlines three underlying factors that influence intention to act; attitude, subjective norm and perceived behavioural control. Attitude relates to the belief an individual holds towards the action (positive or negative), subjective norm relates to the perceived social pressure from important others (weak or strong), and perceived behavioural control relates to whether an individual believes they have control in carrying out the action (both internal and external control). Perceived behavioural control has been found to contribute uniquely to behaviour, as well as independently predicting intention, providing strength for its addition to the model (Armitage & Conner, 2001).

The theory also aims to explain why a behaviour occurred (rather focussing solely on prediction) by identifying antecedents of attitude, subjective norm and behaviour, which determine intentions and action. These are salient beliefs towards the behaviour and include behavioural beliefs (influencing attitudes), normative beliefs (influencing subjective norms), and control beliefs (influencing perceived behavioural control; Ajzen, 1991). Behavioural beliefs are formed through considering the potential outcomes and the evaluation of those outcomes (favourable or unfavourable). Normative beliefs are formed through the perceived social pressures towards an outcome and the levels of motivation to comply with these, and control beliefs are formed through the presence or absence of factors that will/will not lead to the outcome. Control beliefs can be influenced by past experiences (Kan & Fabrigar, 2017).

A meta-analysis conducted on the efficacy of the theory based on 185 independent studies found the theory of planned behaviour accounted for 27% of the variance in behaviour and 39% of the variance in intention; supporting efficacy for the model (Armitage & Conner, 2001). More recently, the theory of planned behaviour has been applied to better understand health professionals’ behaviour (Perkins et al., 2007). Lecomte et al. (2018) explored the impact of mental health clinicians’ attitudes towards implementing cognitive behavioural therapy for psychosis using the theory of planned behaviour highlighting the significant impact of clinicians’ attitudes and perceived social norms on implementation. Organisational barriers were found to inhibit clinicians’ levels of perceived behavioural control. A study to understand the predictors of psychological well-being practitioners’ intention to use cognitive behavioural therapy self-help materials (Levy et al., 2016) found that attitude most strongly predicted intention, with subjective norm and perceived behavioural control also adding significant variance. The full model explained 70% of the variance in intention; considerably higher than previous studies (using the meta-analysis conducted by Armitage & Conner, [2001] as a comparator).

Within the last decade there has been growing criticism of the theory of planned behaviour, with authors arguing that the theory is not supported through experimental tests and relies heavily on self-report measures as observations of behaviour. There is a strong call for retirement of the theory altogether (Sniehotta et al., 2014). Armitage and Conner (2001) report that when behaviour is measured through self-report, the theory of planned behaviour accounts for 11% more of the variance than objective measurement. In response to this criticism, Ajzen (2015) outlines a distinct lack of full understanding of the theory of planned behaviour, suggesting that it can be supported through experimental testing when the studies fully conform to the required standards needed for behaviour change.

Given the lack of assessments measuring adherence to human rights, the theory of planned behaviour offers an opportunity to design a questionnaire using pre-defined guidelines (Francis et al., 2004) that explores intentions to work using a human rights-based approach.

Aims

The aim of this paper is to test whether the determinants of the theory of planned behaviour; attitude, subjective norm and perceived behavioural control, can effectively predict mental health professionals’ intentions to work using a human rights-based approach.

Method

Ethics

Ethical approval was granted by the National Research Ethics Committee (18/NW/0170) and Health Research Authority. The University of Liverpool acted as sponsor for the research (UoL001352).

Design

This study adopted a cross-sectional questionnaire design. Participants comprised mental health professionals working on acute mental health wards in two NHS Trusts in the North West of England.

Construction of the theory of planned behaviour questionnaire

The theory of planned behaviour questionnaire was primarily constructed by the first author in consultation with co-authors of the paper and was based on the published guidelines by Francis et al. (2004) and recommendations of Ajzen (2006). Indirect measures were constructed by first conducting an elicitation study to gather commonly held behavioural, normative and control beliefs from a small sample of the target population. A questionnaire with open ended questions was emailed to five mental health nurses, recruited through convenience sampling. Four replies were received (response rate of 80%) and themes were explored through a content analysis (Krippendorff, 2018), where the first author coded the beliefs that were most recurrent into questionnaire items. Direct measures of attitude, subjective norm and perceived behavioural control were constructed as per examples provided in the guidelines, with a change in content to fit the behavioural outcomes of the study. The same procedure was used to construct generalised intention items. All the items were measured on a 7-point Likert-scale.

Participants

The study recruited from two NHS Trusts specialising in mental health in the North West of England. Participants were recruited from five wards in total.

Participants were invited to participate if they; worked in acute adult mental health inpatient services, were employed directly by the NHS Trust, and worked directly with service users. Based on a medium effect size (ƒ² = 0.15) and a power level of 0.8, the minimum number of health professionals required for a multiple regression analysis is 76 as calculated using G*Power software (Faul et al., 2009). Across five wards there were a total of 150 mental health professionals eligible; 76 members of staff participated resulting in a response rate of 50.7%.

Measures

The theory of planned behaviour measure provided to participants consisted of 59 items in total measuring generalised intention, attitude, subjective norm and perceived behavioural control. Examples of items for each variable are shown in Table 1. Generalised intention consisted of 3 items and achieved an acceptable Cronbach’s alpha score (a = 0.75). The attitude measure consisted of 14 items overall; 4 direct measures and 10 indirect measures. The direct measures achieved an acceptable Cronbach’s alpha score (a = 0.77). Subjective norm consisted of 22 items; 4 direct measures and 18 indirect measures. The direct measures were not able to achieve an acceptable Cronbach’s alpha score, even after deletion of items it remained poor (a = 0.56). Due to issues with internal consistency with the perceived behavioural control items the final analysis resulted in 6 items; 2 direct measures and 4 indirect measures. These achieved a Cronbach’s alpha (a) score of 0.64.

Table 1. Example items in the theory of planned behaviour measure.

Generalised intention	I intend to support service users with mental health difficulties using a human rights-based approach
Direct measure of attitude	Supporting a service user with mental health difficulties using a human rights-based approach is: harmful/beneficial
Indirect measure of attitude + Outcome evaluations of attitude	Supporting a service user using a human rights-based approach will help to reduce stigma around their mental health difficulties Reducing stigma for a service user is: Extremely undesirable/Extremely desirable
Direct measure of subjective norm	People who are important to me want to support service users with mental health difficulties using a human rights-based approach
Indirect measure of subjective norm + Motivation to comply with norms	Senior managers would: disapprove/approve: of me supporting service users using a human rights-based approach What senior managers think I should do matters to me: not at all/very much
Direct measure of perceived behavioural control	Whether I support service users using a human rights-based approach or not is entirely up to me
Indirect measure of perceived behavioural control + Power of the factors	I feel equipped with the knowledge in how to use a human rights-based approach My knowledge in using a human rights-based approach makes me: less likely/more likely: to support service users with this approach

Demographic variables were also measured for descriptive purposes including information on job role, amount of years qualified, how many years working in acute mental health services, and training in human rights.

To aid understanding of human-rights approaches and the aim of the study, participants also received an information leaflet outlining the articles of the Human Rights Act (1998) that are most relevant to someone who is an inpatient (see British Institute for Human Rights at https://www.bihr.org.uk).

Procedure

Data were collected in person over a period of 6 months, from October 2018 to April 2019. To maximise participation, paper copies of the questionnaire pack were left on each ward for participants to complete in their own time. The pack included the participant information sheet, consent form, demographic questionnaire, human rights leaflet, and theory of planned behaviour measure. These were placed into a separate envelope once completed which the researcher then collected on return visit. Contact details of the research team were made available.

Data analysis

Data were inputted and analysed using IBM SPSS Statistics Version 24 software. Pearson product moment correlation coefficients were conducted between the direct and indirect measures of attitude, subjective norm and perceived behavioural control to confirm the validity of the indirect measures. Correlations were also performed to explore any relationships between the three direct predictor variables and generalised intention.

The multiple regression analysis was conducted in two stages; firstly, the direct measures of attitude, subjective norm and perceived behavioural control were regressed onto generalised intention. Secondly, the demographic variables were inputted into the model in a separate step as additional predictor variables.

Results

Participant characteristics

Out of 76 participants, 57 (75.0%) described themselves as female and 18 (23.7%) described themselves as male; one preferred not to answer (1.3%). The majority of participants were aged 25-34 (27, 35.5%), followed by 45-54 (16, 21.1%), 35-44 (15, 19.7%), 16-24 (13, 17.1%) and 55-64 (4, 5.3%); one preferred not to answer (1.3%). The predominant ethnic group reported was White British (61, 80.3%), followed by British Indian (2, 2.6%), White Irish (2, 2.6%) and White Other (2, 2.6%). Nursing assistants were the most common job role represented (29, 38.2%), followed by nurses (27, 35.9%). Thirteen participants (17%) held a role defined as allied health professional. Also represented were psychiatrists (5, 6.6%) and an assistant practitioner (1, 1.3%). Just over half of participants had worked in acute mental health services for 4 years or less (41, 54%) with the biggest majority being in the “less than one year” category (17, 22.4%). Forty participants (52.6%) reported having received internal training on human rights and 10 (13.2%) reported receiving external training.

Theory of planned behaviour analysis

Descriptive statistics and correlation analysis

Descriptive statistics from the theory of planned behaviour questionnaire are presented in Table 2.

Pearson product moment correlation coefficients were calculated between the three predictor variables (attitude, subjective norm, and perceived behavioural control) and intention to test for the assumption of linear correlation as required by multiple regression analysis (as shown in Table 3). Both attitude and subjective norm showed a significant positive correlation with intention (p’s <.01). Perceived behavioural control showed a negative correlation and was not significant with intention (p >.05).

Table 2. Minimum values, maximum values, means and standard deviations from the theory of planned behaviour questionnaire.

Predictor variable (with range of scores)	Minimum	Maximum	Mean	Std. deviation
Direct ATT (1 to 7)	4	7	6.47	0.82
Indirect ATT (−105 to +105)	0	105	87.24	23.17
Direct SN (1 to 7)	2.75	7	5.21	0.90
Indirect SN (−189 to +189)	10	189	118.97	41.17
Direct PBC (1 to 7)	3	7	4.45	0.74
Indirect PBC (−42 to +42)	0	42	29.26	11.95
Intention (1 to 7)	4	7	6.54	0.71

ATT: attitude; SN: subjective norm; PBC: perceived behavioural control.

To confirm whether the indirect items adequately measured the breadth of attitude, subjective norm, and perceived behavioural control, bivariate correlations between the direct and indirect measures were conducted to test for validity. The direct and indirect scores of attitude were significant and positively correlated (r = 0.686, p < .01), as were the subjective norm scores (r = 0.451, p < .01), however the perceived behavioural control scores were not significant and negatively correlated (r = −0.106, p > .05).

Prediction of intention

Two multiple regression analyses were performed for the prediction of intention. Firstly, to test whether the theory of planned behaviour variables of attitude, subjective norm and perceived behavioural control predicted intention to work using a human rights-based approach (results shown in Table 4). Secondly, to test the theory of planned behaviour measures with the additional demographic variables of age, gender, NHS Trust, training in human rights, qualified or non-qualified member of staff, and amount of time worked in acute mental health.

Table 3. Pearson product moment correlation coefficients for the three TPB predictor variables and intention.

	Subjective norm	Perceived behavioural control	Intention
Attitude	0.360**	−0.236*	0.695**
Subjective norm		0.044	0.462**
Perceived behavioural control			−0.211

* Correlation is significant at the 0.05 level; **correlation is significant at the 0.01 level.

Together, the direct measures of attitude, subjective norm and perceived behavioural control significantly predicted generalised intention (F3,72 = 28.271, p < .001) and accounted for 52.2% (adjusted R²) of the overall variance. Attitude was the strongest predictor of the three variables, having the highest standardized coefficient (b = 0.508, p = .001), followed by subjective norm (b = 0.202, p = .042). On its own, perceived behavioural control was not found to be a significant predictor of intention (b = −0.082, p = .466) (Table 4).

Table 4. Regression coefficients for prediction of intention (model 1).

	Unstandardised coefficient (B)	S error (B)	Standardised coefficient ($\beta )$	t	p Value
Direct measure of attitude	0.508	0.078	0.583	6.555	.000
Direct measure of subjective norm	0.202	0.068	0.256	2.956	.004
Direct measure of perceived behavioural control	−0.082	0.080	−0.085	−1.020	.311
F statistic	28.271
R square	0.541
Adjusted R square	0.522

None of the demographic variables included in model two significantly predicted generalised intention. The amount of variance in this model was lower than model one (Adjusted R² = 0.483) and it reduced the influence of subjective norm on intention, becoming non-significant (b = 0.210, p = .053).

Discussion

The aim of this paper is to test whether the theory of planned behaviour can be an effective model in understanding mental health professionals’ intentions to work using a human rights-based approach.

Overall, the theory of planned behaviour constructs were able to significantly predict intention, accounting for 52.2% of the variance, with this effect size being comparable to other theory of planned behaviour studies (Armitage & Conner, 2001). Attitude was the strongest predictor of intention with the highest mean, followed by subjective norm. Subjective norm is often found to be the weakest construct within the literature, however, other studies that have applied the theory of planned behaviour to teams of healthcare professionals have also found subjective norm to be a strong predictor of intention (Foy et al., 2007; Hanbury et al., 2011; Ince et al., 2015). Contrary to common findings, perceived behavioural control did not add significant variance to the model. This could have been impacted by issues with internal consistency, which mirrors findings of a similar study aiming to better understand clinicians’ perspectives on the implementation of cognitive behavioural therapy for psychosis (Lecomte et al., 2018). Kraft et al. (2005) also report that considerable variation in internal consistency on the perceived behavioural control measure is not uncommon.

The demographic variables reported are of interest, although did not add any further variance. Only 52.6% of participants identified receiving internal human rights training even though both NHS Trusts offered this as mandatory, raising questions around how staff engage with and implement knowledge from such training packages. A large percentage of participants were new to working in acute mental health (working for less than a year), suggesting either the role attracts newly qualified members of staff, or these settings attract high numbers of staff turnover. The potential impact of this may also be worth exploring further.

Using the theory of planned behaviour has been a useful tool to begin to understand mental healthcare professionals’ intention to use human rights-based approaches, however the understanding is limited. The theory has seen much success in improving individual intention to perform a health-related behaviour where generally, organisational and cultural factors are not present (Ince et al., 2015). The Improved Clinical Effectiveness through Behavioural Research Group (2006) suggest the theory of planned behaviour may not be a sufficient theory-base to apply when there are external influencing factors outside of the individual clinicians’ control. In an exploration of leadership in mental health nursing, Cleary et al. (2011) were able to identify organisational and cultural factors that may alienate mental health professionals and increase perceived lack of control. If public service staff do not feel empowered to implement a human rights approach, or feel a social collective responsibility, then staff may feel psychologically resistant (Kinderman & Butler, 2006). Organisational culture could be an important component of health professionals’ individual perceived levels of power to work within certain guidelines or approaches, alongside individual attitudes and social norms.

It is important to note the intention-behaviour gap that exists within this study due to behaviour not being objectively observed. Given the settings in which this study was taking place and constraints to resources, it was not possible to directly evaluate whether staff were working using human rights approaches. One major constraint to this is the lack of existing measurement tools within the field of human rights in healthcare and the limited resources available to the authors to design a measurement tool of this scale. Alongside this, there were ethical considerations around staff consent to being observed, as well as service user privacy if implementing a direct measurement of behaviour rather than self-report. Acknowledging this intention-behaviour gap however, a separate mixed methods study was conducted exploring service user views on human rights on acute mental health wards (to be published). The data were collected on the same wards at the same time point and although it does not offer a direct measure of behaviour that can be linked to participating staff members, it sought to offer some indication of levels of human rights-based working.

It would be useful to return to certain criticisms around human rights-based approaches, such as definitions being too broad and conceptual, causing issues with evaluation. For a theory of planned behaviour measure, the behaviour is usually stringently defined, and consideration should be given to whether asking participants if they felt in control of using human rights approaches may have been too broad, as per the critique by Batliwala (2007). Using the theory of planned behaviour however, offered an opportunity to design a questionnaire where measures in this area are scarce, and therefore has opened discussions around the determinants of working with human rights amongst mental healthcare professionals. It is clear further research and development work in this area, although complex, is needed. Narrowing the focus of the behavioural outcomes, for example by using the FREDA framework (Curtice & Exworthy, 2010), and working to better define human rights outcomes would be recommended.

Limitations

There are limitations with the measures used in this study. For the theory of planned behaviour questionnaire, Francis et al. (2004) recommend conducting an elicitation study for the development of the indirect measures with approximately 25 health professionals; here only four mental health nurses participated. Although this is lower than recommended, it is not dissimilar to similar studies (Foy et al., 2007; Hanbury et al., 2011; Janus et al., 2017; Levy et al., 2016).

Two of the questions on the subjective norm component of the theory of planned behaviour measure referred to “psychologists” (“Psychologists do not/do support service users using a human rights-based approach” and “Doing what psychologists do is important to me not at all/very much”). NHS Trust A did not have any psychology provision on its wards and so participants felt unable to provide answers. It would be beneficial to have this knowledge beforehand to tailor the questionnaire accordingly. Both the subjective norm and perceived behavioural control items showed issues with internal consistency, suggesting these variables may not have been constructed as robustly as the attitude measures, or using human rights as a behaviour outcome is too broad. These issues would need attention if the study were to be replicated.

When constructing a theory of planned behaviour study, it is recommended to objectively measure actual behaviour when possible (Armitage & Conner, 2001). Unfortunately, this study was not able to fulfil this, as discussed above, and future consideration should be given on how to directly measure working within a human rights framework.

Paper copies of the questionnaires were left on the wards for participants to complete in their own time. Therefore, consideration should be given to motivational bias and whether the attitudes reported here can be generalised to all acute mental health staff.

Recommendations and clinical implications

The aim of this paper was to test whether the theory of planned behaviour can be an effective model in understanding mental health professionals’ intentions to work using a human rights-based approach. Results suggest that individual attitudes and social norms could predict intention, but participants did not perceive they had behavioural control over human rights outcomes. Overall, mental health professionals’ attitude towards working using a human rights-based approach is positive (mean value of 6.47 out of 7) and therefore interventions could focus on turning positive attitudes into direct action. Drawing on the significance of the subjective norm scores, designing an intervention where individual attitudes can be fostered as shared attitudes within the team may be beneficial. An example of this might be having human rights as the subject of a team away day involving group discussion. Changes in the way supervision is delivered could offer another intervention focussing on subjective norm, particularly if focus is given to the importance of human rights, for example embedding human rights principles as a fixed supervision discussion point. This may then foster a feeling of human rights being important to the individual, senior managers, and the organisation. An intervention with mental health professionals may help to elicit perceived barriers to control over working using human rights-based approaches. A problem-solving approach may help to foster increased feelings of individual control. However, further research into perceived behavioural control is recommended to understand the negative correlation with intention. There are many hypothesised factors within the literature around external factors contributing to the practice of mental health professionals. To explore factors impacting on human rights specifically, qualitative interviews could be conducted initially.

All staff who participated had completed mandatory training in human rights, however only just over half recognised this. Therefore, it would be beneficial to explore how staff engage with training packages and to measure how effectively the theory is linked into practice.

Conclusion

Mental health professionals’ attitudes and perceived social pressures significantly predicted their intention to work using human rights-based approaches on acute mental health wards. Contrary to previous findings, perceived behavioural control did not add any significant variance, suggesting that working within a human rights-based approach feels out of mental health professionals’ direct control. Due to the positive attitudes already held by the participants in this study, increased training to instil individual differences may not be most relevant to produce meaningful change, however thought should be given to how health professionals engage with mandatory training. Measuring health professional’s adherence to working within human rights practices is a complex task given the scarcity of assessment tools and evidence-base. Using the theory of planned behaviour has been useful in highlighting positive attitudes towards human rights but has exposed issues around perceived control which may need further research. Therefore, future developments should focus on effectively defining human rights-based outcomes to offer guidance to mental health professionals and measure adherence to human rights law, with the hope of improving the quality of acute mental health services.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The author(s) reported there is no funding associated with the work featured in this article.