Cancer treatment and management have become increasingly complex over the past two decades, with therapeutic decisions often based on input from a multidisciplinary team that consists of radiologists, surgeons, pathologists, and oncologists . For patients with suspected or confirmed cancer, clear and thorough recording of diagnostic and surgical procedures and findings support accurate diagnosis and staging. Such recording also facilitates more accurate prognosis estimates, post-operative management, and adjuvant treatment planning. The dominant method of reporting findings from diagnostic tests/procedures, surgery, and pathology examinations is the narrative report, which is a free text, descriptive account of the procedure, suspected or confirmed findings, and proposed treatment. Physicians dictate this report, often through automated telephone systems, and professional transcriptionists transcribe the oral description into a written document that is eventually placed into a patient's medical record. Research has demonstrated that narrative reports inconsistently provide the information required to understand the disease and make informed patient care decisions [2–7].
Another method of reporting, the synoptic report, captures data items in a structured manner and contains only items critical for understanding the disease and subsequent impacts on patient care. There is a spectrum of what is generally considered a synoptic report , from synoptic-like structured templates without scientifically validated elements to sophisticated electronic systems with drop-down menus, discrete data fields, standardized language, automated coding processes, and strong evidentiary basis. A landmark study in the early 1990s, which audited pathology practice patterns at 532 institutions in three countries, found that the one practice associated with completeness of pathology reporting for colorectal cancer specimens was use of a standardized report or checklist . Since that time, researchers have consistently demonstrated that synoptic reports (even paper-based 'checklist' formats) vastly improve the quality of pathology reporting in colorectal [1, 2, 9–13], breast [1, 9, 14–16], lung [1, 17], prostate , pancreatic , melanoma , and hematolymphoid cancers . More recently, synoptic reporting has been shown to improve the quality of surgical reporting for a variety of malignancies, including colorectal , breast , thyroid , and pancreatic cancers , as well as non-malignant operative procedures [24, 25].
Electronic synoptic reporting tools also lead to health system efficiencies compared to the dominant, dictated method of reporting [25–27]. Laflamme et al.  showed that use of synoptic templates accelerated the mean time for a verified surgical report to reach the patient's medical record by 800-fold compared to narrative reporting (28 minutes versus > 14 days, respectively). Moreover, the mean time from the end of the surgery to initiating the report was substantially less when using synoptic templates (0.43 hours) versus dictation (9.7 hours). Similar efficiencies were demonstrated in subsequent studies [26, 27]. In a Canadian study, for example, 97% of synoptic reports were finalized, placed in the patient's medical record, and sent to all health professionals involved in the patient's care within 24 hours of surgery compared to a mean of 90 days for narrative reports . Researchers have also estimated considerable cost-savings through the elimination of transcription services [25, 26].
Beyond improving completeness of reporting and availability/immediacy of reports, synoptic reporting tools have the potential to improve quality of care by integrating practice guidelines/best evidence into report templates [21, 26] and providing an efficient, real-time mechanism to generate data from the diagnostic and peri-operative periods [21, 26, 28, 29]. These data may be used to provide real-time performance feedback to physicians and surgeons as well as enable improved process and outcomes measurement. International jurisdictions are increasingly endorsing synoptic reporting, including actively supporting/funding the implementation of synoptic templates [30–32] and providing commendation status to pathology labs that include a synoptic synopsis of scientifically validated data elements in their reports . In addition, the professional pathology colleges in Canada, US, UK, and Australia have formalized a collaboration to develop common, internationally agreed-upon, standardized cancer reporting protocols .
The synoptic report represents a complex innovation (i.e., new knowledge, tool, or practice) in cancer care, with its implementation and use requiring fundamental shifts in physician behaviour and practice culture  as well as support from the organization (e.g., changes in institutional policies/processes) and larger system (e.g., governance arrangements, integration with health information technology infrastructure). Despite the demonstrated benefits, some physicians have reported reluctance to use synoptic reporting tools, with concerns including lack of flexibility in reporting complex procedures/cases [36, 37], the prospect of being monitored , and discomfort with using information technology [37, 38]. Changing physician reporting practice is a complex undertaking that requires comprehensive approaches at different levels of the health system . This may be particularly true for narrative reporting, a practice that has existed for millennia .
Implementing new practices in healthcare organizations
Knowledge translation (KT) research has largely focused on potentially useful strategies (e.g., opinion leaders, academic detailing, reminder systems) for improving the adoption and uptake of evidence (e.g., clinical practice guidelines) into practice . Most of these strategies fall within the realm of individual-level interventions [39, 42–44], with the target being 'autonomous' clinicians who are deemed to be more-or-less independent in their capacity to assemble and apply knowledge to modify their practices . Despite the sizable amount of literature in this area, however, numerous systematic reviews have been unable to demonstrate which of these strategies work best, or even consistently, across clinical settings [43–45]. Many researchers have emphasized the unpredictable, slow, and haphazard nature of research implementation and use processes, with interventions working some of the time in some situations, but not at other times in seemingly similar situations [41, 42, 46], and the reasons for these differences unclear .
In reality, many organizational and socio-political (e.g., inter-organizational networks, funding arrangements) factors affect whether individuals in clinical settings actually make changes in their practice [41, 47–50]. Much research has demonstrated the importance of organizational characteristics (e.g., culture, leadership, management support, evaluation/feedback mechanisms, and presence of champions) to implementation efforts in healthcare settings [41, 51–65]. Moreover, many of the defining features of healthcare systems, including the range and diversity of stakeholders, complex governance/resource arrangements, and professional autonomy and specialization of many of its staff, result in many different cultures and norms as well as high levels of interdependency amongst professionals in the system [66, 67].
Consequently, many implementation processes in healthcare organizations will also be characterized by a high degree of interdependency amongst organizational members [68, 69]. Indeed, many innovations introduced in healthcare will require coordinated use by many individuals and professional groups to achieve benefits (electronic medical records are one example). These individuals are situated in organizational relationships wherein the implementation and use of a new tool or practice will ultimately be influenced by many interpersonal processes, including 'coalition building,' rhetoric, and persuasion [70, 71]. Thus, while individual-level interventions are important to change clinical practice, the complex nature of healthcare organizations means individual-level interventions alone cannot change clinical practice in a widespread, sustainable way [39, 48, 72–75].
Understanding the dynamics of innovations in organizations has a long history in management and organizational sciences . Rogers  has conceptualized the innovation-decision process as one that unfolds in distinct stages whereby an organization moves from initial awareness or knowledge of an innovation to eventually successfully integrating the innovation into ongoing processes (or, alternatively, rejecting the innovation). Contrary to this perspective, extensive longitudinal study of innovation processes led Van de Ven et al.  to describe the 'innovation journey' as a non-linear cycle during which ideas are developed (or adapted) and put into practice by people who, through their relationships and negotiations with others, make the changes necessary to implement the innovation within a specific organizational context. They highlight that people and relationships are instrumental to this journey, which is characterized by many divergent and convergent activities wherein the initial idea often leads to multiple ideas/actions, setbacks and delays occur frequently, staff experience high levels of elation and frustration, notions of success change, and new interdependencies are established that affect the wider organization. This broader 'systems' perspective [78, 79] has recently made its way into KT dialogue and debate , challenging our thinking of a linear view of KT (e.g., researcher-push model) and moving us toward one that is much more contextual, relational, and 'living' in nature.
The objective of this study is to examine the key interpersonal, organizational, and system-level factors (hereafter referred to as 'multi-level' factors) that influence the implementation and use of synoptic reporting in three specific cases of cancer care. The interpersonal level relates to the relational aspects at the level of the implementation team/program: e.g., teamwork and team dynamics, communication, partner engagement, coalition building, power dynamics, and use of rhetoric and persuasion to accomplish goals/tasks. The organizational level relates to institutional (i.e., hospital) factors that influence implementation and behaviour change: e.g., organizational culture, leadership, management, intra-organizational relationships, evaluation capacity/mechanisms, implementation policies and practices, infrastructure, and presence of champions. The system level refers to the broader sociopolitical context: e.g., policies such as financial incentives/disincentives, resource and governance arrangements, and inter-organizational norms and networks.
This study involves three initiatives (the cases) in Nova Scotia, Canada, that have implemented a synoptic reporting tool within their departments/programs. The examination of each case will involve answering the following specific research questions:
1. What, if any, common factors affected implementation and use across cases? How was it that these factors 'transcended' the different contexts (setting, timing, and 'actors' involved)?
2. Are there context-specific factors within each case, which were not found in other cases, that affected implementation and use? If so, what are they and what are their specific relationships to the setting, timing, and actors?
The outcome of this study will be a descriptive and explanatory account of the multi-level factors that influence the implementation and use of synoptic reporting in cancer care.
Case study methodology (CSM) [80, 81] will be used to study the three synoptic reporting cases in-depth, explore which factors were barriers or facilitators of implementation and use, examine relationships amongst factors, and uncover which factors appear to be similar (and distinct) across cases. CSM permits the rigorous study of a contemporary phenomenon within its real-life context , and of the complex interactions between the social actors and their actions and environments . Case studies typically focus on 'how' and 'why' questions and explore multiple dimensions of some particular phenomenon. Flyvbjerg  argues that such in-depth study (of real cases in specific contexts) may be pivotal to transitioning from a novice to an expert understanding of the phenomenon.
This complexity means that case study researchers deal with distinct contexts whereby there are more variables of interest than data points. As a result, case studies rely on multiple sources of evidence and benefit from knowledge of the literature and existing theoretical perspectives . The use of multiple sources is vital to CSM, as it permits corroboration (i.e., triangulation) of findings and resultant interpretations . The use of existing theoretical perspectives helps guide data collection and analysis. Without a prior theoretical understanding, researchers risk spending considerable time and effort gathering basic information and 'providing description without meaning' (Hartley, cited in ).