Healthcare systems and professionals fail to deliver the quality of care to which they aspire. Multiple studies internationally have observed evidence to practice gaps demonstrating that 30 to 40 percent of patients do not get treatments of proven effectiveness, and equally discouraging, up to 25 percent of patients receive unnecessary care that is potentially harmful [1
]. Such evidence to practice gaps have significant adverse effects on the health and social welfare of citizens and economic productivity.
Lumbar spine imaging for low back pain in primary care settings is an example of an evidence to practice gap. Low back pain is an extremely common presentation in primary care. However, lumbar spine imaging in patients under 50 years is of limited diagnostic benefit within primary care settings [4
]. Globally, clinical guidelines for the management of low back pain do not recommend routine imaging of patients with low back pain [4
]. Furthermore, standard lumbar spine x-rays (the most common imaging modality used by UK primary care physicians) are associated with significant ionising radiation dosage. Despite this, lumbar spine x-rays are the fourth most common x-ray request from UK primary care physicians [9
], with x-ray referrals continuing at the rate of 7 per 1000 patients per year [10
]. We conducted a trial that found that for the majority of primary care physician requests, case note review could not identify appropriate indications for referral [10
]. The trial also observed a reduction in lumbar spine x-rays of 20 percent without apparent adverse effects following the introduction of educational messages [10
Recognition of evidence to practice gaps has led to increased interest in more active strategies to disseminate and implement evidence. Over the past two decades, a considerable body of implementation research has been developed [11
]. This research demonstrates that dissemination and implementation interventions can be effective, but provides little information to guide the choice or optimise the components of such complex interventions in practice [12
]. The effectiveness of interventions appears to vary across different clinical problems, contexts, and organizations. Our understanding of potential barriers and enablers to dissemination and implementation is limited and hindered by a lack of a 'basic science' relating to determinants of professional and organizational behaviour and potential targets for intervention [14
]. The challenge for implementation researchers is to develop and evaluate a theoretical base to support the choice and development of interventions as well as the interpretation of implementation study results [15
]. Despite recent increased interest in the potential value of behavioural theory to predict healthcare professional behaviour, relatively few studies have assessed this. A recent review by Godin et al
. explored the use of social cognitive models to better understand determinants of health care professionals' intentions and behaviours [16
]. They identified 72 studies that provided information on the determinants of intention, but only 16 prospective studies that provided information on the determinants of behaviour.
The current study, one part of the PRIME (PRocess modelling in ImpleMEntation research) study) [17
], aimed to investigate the use of a number of psychological theories to explore factors associated with primary care physician lumbar spine x-ray referrals. Previous PRIME studies have used similar methods to explore factors associated with primary care physicians' use of antibiotics for sore throats and general dental practitioners' use of routine intra-oral x-rays and preventive fissure sealants [18
]. Variables were drawn from the Theory of Planned Behaviour (TPB) [21
], Social Cognitive Theory (SCT) [22
], Operant Learning Theory (OLT) [23
, Implementation Intentions (II) [24
], Common Sense Self-Regulation Model (CS-SRM) [25
], and Weinstein's Stage Model termed the Precaution Adoption Process (PAP) [26
]. These specific theories, which are described in detail elsewhere [28
], were chosen because they predict behaviour but vary in their emphasis. Some focus on motivation, proposing that motivation determines behaviour, and therefore the best predictors of behaviour are factors that predict or determine motivation (e.g.
, TPB). Some place more emphasis on factors that are necessary to predict behaviour in people who are already motivated to change (e.g.
, II). Others propose that individuals are at different stages in the progress toward behaviour change, and that predictors of behaviour may be different for individuals at different stages (e.g.
, PAP). The specific models used in this study were chosen for three additional reasons. First, they have been rigorously evaluated with patients or with healthy individuals. Second, they allow us to examine the influence on clinical behaviour of perceived external factors, such as patient preferences and organisational barriers and facilitators. Third, they all explain behaviour in terms of variables that are amenable to change.
The objective of this study was to identify those theories and the theoretical constructs that predicted clinical behaviour, behavioural simulation (as measured by the decisions made in response to five written clinical scenarios) and behavioural intention for lumbar spine x-ray referral.