Keywords: clinician behavior; intention; habit; multiple behaviors; planning; self-efficacy.
PREDICTORS OF CLINICIAN BEHAVIORS IN MANAGING DIABETES 5There is enduring clinical and policy interest in promoting evidence-based health care.However, the production and dissemination of evidence of clinical best practice does not automatically lead to implementation. The importance of the use of theory to understand clinician behavior and to inform the choice of change interventions to improve care is well acknowledged (Godin, Bélanger-Gravel, Eccles, & Grimshaw, 2008; Grol, Bosch, Hulscher, Eccles, & Wensing, 2007; Michie, Johnston, Francis, Hardeman, & Eccles, 2008). One important step in this process is to identify how well theoretical models predict clinician behaviors. A number of studies have tested the constructs in social cognition models as predictors of clinician behavior. Most of these have been applied to predict a single behavior using a single theory, and on average explain 31% (R 2 range < 0.01 to 0.58) of its variability (Godin et al., 2008). This range has been assumed to be due to variation in behaviors or respondents, or to methodological issues. The present study aimed to test these assumptions directly by investigating relationships between constructs from predominant theories of behavior in the same clinicians, as predictors of multiple clinician behaviors in the context of diabetes care in the primary care setting.With a prevalence of over 5% (NHS_Information_Centre, 2010) complications related to Type 2 diabetes are an important cause of avoidable mortality (Adler et al., 2000).In the United Kingdom, people with diabetes are primarily managed by the integrated activities of primary care teams composed of general practitioners and practice nurses. only. The present study also aimed to test each theory at the clinician and organization (i.e., practice) level.
Theories of Behavior Tested in this StudyThe rationale for selecting the theories investigated in this study was to reflect those tested in the 'PRocess modeling in ImpleMEntation' (PRIME) study, where multiple models were used to predict five clinical behaviors in five different samples of clinicians (Bonetti et al., 2010; Bonetti et al., 2006; Eccles et al., 2007; Grimshaw et al., 2011). The present research, the 'improving Quality in Diabetes (iQuaD)' study, builds on PRIME by aiming to test theories across multiple behaviors within the same clinicians.The Theory of Planned Behavior (TPB) is the most commonly tested theory of health professional behavior (Godin et al., 2008) and suggests that behavior is a function of beliefs that influence intention which, along with perceived behavioral control, determine behavior Clinician behaviors are likely the result of more than health professionals being motivated and feeling capable of engaging in evidence-based behaviors. Some clinician behaviors may be performed in a habitual manner in light of anticipated consequences such that their enactment in everyday practice does not necessarily depen...