Testicular self-examination behaviour, the development and longitudinal evaluation of a theory-based intervention utilizing the theory of planned behaviour: A clustered randomized control trial

Abstract

The incidence of testicular cancer has rapidly increased in recent years. The disease is associated with growing morbidity and mortality among young men. Early detection is essential for less aggressive treatment and a better prognosis. Cancer organizations and health professionals recommend that testicular self-examination (TSE) is performed at least once a month every month, because of its high growth rate (10 to 30 days). TSE is the only independent way of detecting the disease. Research has shown that knowledge of testicular cancer and TSE is exceptionally low among the age group most affected by the disease. The purpose of the thesis was to develop an intervention based on the theory of planned behaviour (TPB). The intervention was developed to promote the performance of TSE and to increase the level of knowledge of the disease and the behaviour. The entire research took multiple phases, for example, an elicitation phase, the questionnaire development, a baseline study, the development of the intervention and the final phase was the delivery and evaluation of the intervention in the main longitudinal study.

The elicitation study identified the underlying beliefs associated with testicular cancer and TSE (N= 50 males; aged 17 to 35 years). Findings showed that 21 underlying beliefs were salient to the respondents. The questionnaire was developed and the baseline study was then conducted (n=280 males; 17 to 35 years) to pilot the questionnaire. The intervention was developed incorporating the findings from the elicitation study. The main study was a longitudinal clustered randomised controlled trial (CRCT). Experimental one, the theory-based intervention was a ‘Multi-media’ PowerPoint presentation plus an ‘Action Cancer’ leaflet. Experimental two was an ‘Action Cancer’ leaflet and control was the third condition. The interventions were evaluated (relative to control) over 12 months (pre-intervention, one month, six months and 12 months). The outcome measure was a self-administered questionnaire based on the components of an extendedly modified TPB and reported TSE intentions and behaviour.

The latent Growth Modelling approach (LGM) was employed to examine the effects of the two interventions (relative to control) and the psychological variables that can explain the observed pattern of change in TSE behaviour over time (Duncan & Duncan, 2004). The LGM also provided information on the inter-individual and intra- individual variability within the sample both at initial status (T2) and behavioural change over the 12 months (T3 and T4). The LGM was found to be an acceptable description of the data.

For the main study, data were obtained from a sample of college students (n=892; Mean age 25.67; SD= 8.14). Preliminary findings showed that the performance of TSE behaviour increased significantly from never to mean score of 1.7 times per month over the 12 months. Frequency findings illustrated that at T1 87% of males never performed TSE and by T4 this percentage reduced to 23%.

LGM analysis indicated that there was significant individual variability in terms of the initial status and slope factors. To endeavour to explain the variability, we found that the theory-based intervention was the strongest predictor of both behaviour and knowledge at T2; however there was no change beyond T2. The finding indicates that the change in behaviour from T1 to T2 stayed at the same frequency of performance overtime. The ‘Action Cancer’ Leaflet alone was a significant predictor of behaviour. However, the ‘leaflet’ had a negative effect on behaviour over time. This finding indicates that the frequency of behavioural performance decreased overtime. The leaflet did not have a significant affect on the level of knowledge at the initial status T2 or overtime. Behavioural intention was found to be a significant predictor of the intercepts for both behaviour and knowledge at T2 but not for behavioural or knowledge change. Whilst, the covariates past behaviour and knowledge were found not to influence intention, they emerged to have a significant influence in behaviour and knowledge at one month follow-up. The LGM showed that instrumental and affective attitude, self-efficacy, moral norm and anticipated regret were important predictors of intention.

A path analysis was then employed to explore the individual mediating effects of the modeled variables with the interventions. The findings from the path analysis suggest that compared to control the leaflet and the theory-based intervention had an effect on behaviour with the ‘Multi-media’ having a stronger effect. Moreover, the control condition had a significant effect on behaviour also. Affective attitude and self efficacy had a significant effect on both the theory-based intervention and the ‘Action Cancer’ leaflet at T2. Subjective norm, moral norm and anticipated regret had a significant influence on the theory-based intervention but not the leaflet at T2. Moreover, none of the interventions or the theory-based variables had a significant influence on behaviour at T3 or T4.

Whilst the current findings demonstrate the significant independent effects of administering a TPB-based questionnaire only, findings showed positive effects for the ‘Multi-media’ intervention long-term, whereas, compared to control the leaflet illustrated a negative effect over time. Current findings demonstrate the demand for any type of intervention to promote TSE behavioural performance; because the prevalence of the behaviour and the level of knowledge were extremely low at T1. The administration of an inexpensive questionnaire may well promote behaviour among the target population. However, the TPB-based intervention offers a promising long-term perspective for future researchers in intervention development.

Date of Award	Oct 2010
Original language	English
Supervisor	Siobhan M McCann (Supervisor) & Sam Murphy (Supervisor)