Introduction

Colorectal cancer is one of the leading causes of cancer death in North America. Surgery, is the primary curative treatment for these cancers1. However, the colon can be resected in a variety of different configurations, therefore accurate preoperative localization is imperative to facilitate an effective operation. Colonoscopy, performed by an endoscopist (either a surgeon, or a gastroenterologist), is the gold standard for colorectal tumour localization2. However, colonoscopy is also a screening test, and therefore hundreds of colonoscopies are performed before a cancer is diagnosed3. Frequently the endoscopist is not the eventual operating surgeon4. Proper communication is therefore crucial to help surgeons make appropriate treatment decisions based upon information obtained during the initial colonoscopy. Unfortunately, there has been a wide variation in the documentation and communication of colonoscopy findings, including localization and characterization of colorectal cancers and precancerous tumours5. New recommendations to help standardize endoscopic tumour localization for colorectal cancers and polyps have recently been developed. These recommendations were established based upon comprehensive literature review and consensus between 23 Canadian experts5. Previously, endoscopists and surgeons have requested more standardized documentation and localization practices at endoscopy for colorectal tumours6,7,8,9. However, evidence suggests that creating new guidelines is insufficient to substantially impact clinician practices10,11. Meaningful practice changes are more likely when context-specific tailored strategies are used12,13. An important early step in implementation is to understand the perceptions of the end-users14. Our research objective was to identify the barriers and facilitators, as perceived by local gastroenterologists and surgeons, to using new recommendations designed to standardize endoscopic lesion localization for colorectal tumours. Findings from the present research can be used to help develop and pilot test contextualized interventions designed to improve adherence to the recommendations.

Methods

Overview and frameworks

This was a qualitative semi-structured interview study of gastroenterologists and surgeons in a single large central Canadian city (Winnipeg, Manitoba). We used directed content analysis methodology to explore participant perspectives on the new recommendations. This deductive approach to qualitative research is frequently used in health care implementation research and consists of describing phenomena according to participants’ perspectives in simple easily understood terminology. This methodology prioritizes fidelity to participants’ views, and compliments a pragmatic epistemology15,16. Pragmatism and postpositivism were the overlying approaches used to guide this research17.

The Consolidated Framework for Implementation Research (CFIR) was used to guide the creation and analyses of semi-structured interviews13. CFIR is an amalgamation of evidence-based implementation strategies across multiple disciplines, provides a comprehensive framework for assessing multiple aspects of implementation strategies, and has been used in multiple domains of health sciences research18. Guided by CFIR, data from surgeons and gastroenterologist interviews were analyzed separately to identify and compare the perceived facilitators and barriers to implementing the new documentation and tumour marking practices. Lastly, researchers have developed a new tool for matching expert-recommended interventions to barriers according to each CFIR category (CFIR-ERIC)19,20,21,22. Barriers identified in the present research were aligned with the CFIR-ERIC tool to identify possible solutions to overcome barriers to implementing the new documentation and tumour marking practices.

Study setting

This study was conducted with gastroenterologists and surgeons in Winnipeg, Manitoba, Canada. Winnipeg is the largest urban centre (population 800,000 people) in the province of Manitoba (population 1.4 million people), has the only colorectal cancer referral site in this province, and treats 800–900 colorectal cancers annually23,24. All gastroenterologists and surgeons in Winnipeg operate under a single payer publicly funded “fee-for-service” billing model. Each of the six hospitals in Winnipeg has an outpatient endoscopy suite. There are two non-hospital-based endoscopy units. Three of the hospitals have operating rooms for surgeries requiring hospital admission. Winnipeg endoscopist practice patterns have been well-described as having high repeat preoperative endoscopy rates (29%)6,25,26,27 and suboptimal endoscopy report quality25,28.

Study participants and sample

Study sample size was guided by the concept of theoretical sufficiency29. According to this principle, the research process concluded once all relevant categories from the CFIR framework had sufficient illustrative examples. Interview transcripts were analyzed soon after each interview, and mapped to the CFIR framework. Participants did not review their transcripts. Gastroenterologist and surgeon perspectives were analyzed separately, to ensure theoretical sufficiency was achieved in each group.

Participants in this research were selected purposively to obtain feedback from practitioners working in each endoscopy suite and operating room across the city, including diverse sub-specializations, practice patterns, and career stages. Initial recruitment emails were sent to every gastroenterologist and general surgeon in Winnipeg to ensure all eligible individuals felt invited and included. Follow-up emails were directed at specific groups who were under-represented after initial responses. Participant informed consent was obtained prior to each interview.

Interview guide development and data collection

A semi-structured interview guide was developed (Additional file 1) based upon questions recommended by the CFIR authors and available on http://cfirguide.org/13. The questions were iteratively refined through meetings with the research team including a gastroenterologist, surgeon, knowledge translation expert, and an expert qualitative researcher and psychologist. Questions were piloted with two senior general surgery residents to assess interview flow, length, content clarity and appropriateness. Revisions were made following the pilot interviews, and pilot data were not integrated into the analysis.

CFIR authors recommend that implementation researchers identify which CFIR constructs they will assess in advance based on the relevancy to the study, rather than to attempt to assess every construct at once13. Four of the five CFIR domains aligned with the study objectives. Detailed rationale for which constructs were included are reported in Additional file 2.

Interviews were conducted by video teleconference (Zoom Video Communications, San José, CA), according to local COVID-19 pandemic-related restrictions. Participants were provided with a copy of the recommendations both prior to and at the beginning of the meeting. A previously published visual infographic tool was used to help participants understand and refer to the recommendations5. All interviews were audio-recorded and later transcribed by the primary analyst.

Data analysis

Units of analysis

Data were categorized according to the CFIR constructs separately by provider specialty (i.e., gastroenterologist or surgeon). Within each group, constructs were subsequently categorized as facilitators or barriers according to coded perceptions. Findings from the two groups were then compared using a triangulation process to identify common and contrasting themes between specialties.

Data coding

Interview transcripts were imported into NVivo software for Mac (version 12.2.0; QSR International, Melbourne, Australia) for analysis. Coding was performed in duplicate independently by two researchers (GJ and CEK) using directed content analysis30. This deductive qualitative research approach is best used when an existing theory has previously been established to explain an observed phenomenon30, and has been used previously for analysis of qualitative interviews using the CFIR31. Following this approach, transcripts were coded using a predetermined codebook and inclusion criteria (Additional file 3)13.

Each transcript was first analyzed at the entire transcript level; these were reviewed repeatedly and coded deductively to CFIR constructs according to the codebook. Data were then reviewed at the level of each interview question to check for additional information that was missed during initial coding. After the entire coding process, both analysts met and created a single unifying codebook through consensus.

Construct relative priority

After interview transcripts were coded, participant perspectives were ranked according to whether a CFIR construct was perceived as a barrier or facilitator to implementing the new recommendations. Ranking criteria were adapted from previous work, and which have been used previously to differentiate high from low-performance implementation settings32,33. Ratings were performed separately by both analysts. Ranking criteria were based on level of agreement among study participants’ expressed views, language strength, and concrete examples used to emphasize responses (Additional file 4).

Validation strategies

Multiple strategies were used throughout the research process to ensure results’ validity including triangulation, reporting disconfirming evidence, dialogic engagement, and reflexivity34. Validation strategy details are listed in Additional file 5.

CFIR-ERIC intervention mapping

Researchers have recently developed and refined a tool to align CFIR constructs to the expert recommendations for implementing change (ERIC) framework20,22. Barriers identified in a study setting can be entered into the tool’s algorithm according to the CFIR framework, and subsequently the tool reports a prioritized list of strategies to consider, based upon prior consensus research20,22. The tool also reports the degree of consensus among the experts for each ERIC strategy as a method to address a particular CFIR barrier. According to this framework, strategies that were endorsed by ≥ 50% of the experts are deemed ‘Level 1’ strategies, and strategies that are endorsed by 20–49.9% of the experts are deemed ‘Level 2’ strategies19,20,21,22. The CFIR-ERIC authors suggest selecting a combination of both broadly applicable strategies, with high cumulative endorsement across multiple barrier constructs, in addition to specific strategies (i.e., level 1 strategy applicable to only one barrier)22. To ensure both approaches were addressed in this research, ERIC strategies were identified based on level 1 endorsement for each individual CFIR barrier, in addition to identifying more “general” strategies with high cumulative endorsement across all barrier constructs. ERIC strategies were stratified according to provider specialty.

Participants frequently proposed their own solutions during the interviews. In post-hoc analysis, these opinions were deductively coded to the ERIC framework and compared to those strategies identified via the CFIR-ERIC tool.

Research team and reflexivity

The interviewer and primary analyst (GJ) was a male general surgery resident and master’s in science student throughout this research process. He was previously acquainted with many of the research participants prior to the interviews through his residency training. All attempts were made during the research to minimize the effect of biases these relationships may cause by acknowledging them throughout the research, discussing emerging findings with the research team, and critically examining the effects on the knowledge generated at each interview and during analysis.

Ethics approval and consent to participate

This study was reviewed by the University of Manitoba Heath Research Ethics Board (HREB) for approval prior to data collection (reference number: HS25143, H2021:315). All experiments were performed in accordance with relevant guidelines and regulations. All participants provided informed consent prior to participation.

Conference presentation

Partial results from this work were presented as a poster at Digestive Diseases Week 2022 in San Diego, CA and at the Canadian Surgery Forum 2022 in Toronto, ON, Canada.

Results

Participant demographics

There were 33 surgeons and 19 gastroenterologists who treat colorectal cancers identified as potential participants in Winnipeg during the study period. Of the 52 individuals invited, 11 gastroenterologists and 10 general surgeons participated in the study between October 2021 and January 2022. Participant demographics are shown in Table 1. Individuals participated from every endoscopy suite, hospital, and operating room in the city. Mean interview time was 56 min and 55 s.

Table 1 Participant characteristics (N = 21).
Full size table

CFIR content analysis

Twenty-seven CFIR constructs were assessed and deemed relevant to the research questions. Perceived barriers and facilitators to following the new recommendations are summarized according to construct relative priority rankings in Table 2.

Table 2 CFIR rankings stratified by participant specialty.
Full size table

Both major (n = 4) and total facilitators (n = 11) were more numerous for surgeons compared to gastroenterologists (9 total, 2 major). Gastroenterologists and surgeons had eight net facilitator constructs in common: ‘relative advantage’ (major for surgeons only), ‘adaptability’, ‘trialability’ (major), ‘complexity’ (major for surgeons only), ‘design quality and packaging’, ‘cosmopolitanism (major for gastroenterologists only)’, ‘structural characteristics’ (major for surgeons only), and ‘tension for change’. Uniquely, surgeons identified ‘innovation source’, ‘self-efficacy’ and ‘leaning climate’ as facilitators, whereas gastroenterologists highlighted ‘leadership engagement’. The only universally acknowledged (major) facilitator for both groups was the ability of the recommendations to be trialed prior to full implementation.

Surgeons identified ten barriers whereas gastroenterologists identified nine. All nine gastroenterologist barriers were also identified as barriers for surgeons: ‘external policy and incentives (major)’, ‘organizational incentives and rewards (major)’, ‘available resources (major)’, ‘goals & feedback’ (major for gastroenterologists only), ‘access to knowledge & information’, ‘knowledge & beliefs about the intervention’, ‘self-efficacy’, ‘individual identification with the organization’, ‘evidence strength and quality’, and ‘costs’. The tenth barrier for surgeons, ‘compatibility’, had more mixed perspectives for gastroenterologists. Table 3 provides a summary of barriers and facilitators identified within each construct with exemplar quotations according to gastroenterologists and surgeons.

Table 3 CFIR barriers and facilitators to implementation of the new endoscopic lesion localization recommendations according to gastroenterologists and surgeons.
Full size table

CFIR-ERIC strategy matching

According to the CFIR-ERIC matching tool, strategies to address barriers identified by gastroenterologists and surgeons are displayed in Figs. 1 and 2, respectively. The top four ERIC strategies were identical for both gastroenterologists and surgeons: 1. ‘Conduct educational meetings’, 2. ‘Alter incentive/allowance structures’, 3. ‘Identify and prepare champions’, and 4. ‘Access new funding’. The CFIR-ERIC tool also identified six level 1 strategies (indicated in bold in the figures) to address CFIR barriers. Again, these strategies were identical for both gastroenterologists and surgeons: 1. ‘Conduct educational meetings’; 2. ‘Alter incentive/allowance structures’; 3. ‘Access new funding’; 4. ‘Develop educational materials’; 5. ‘Audit and provide feedback’; and 6. ‘Distribute educational materials’.

Figure 1
figure 1figure 1

ERIC strategies matched to gastroenterologists’ CFIR barriers. Percentages indicate relative expert endorsement of a strategy to address a CFIR barrier according to CFIR-ERIC strategy matching by Waltz et al.22. Level 1 strategies (≥ 50% expert endorsement) displayed in green, and bolded in left column. Level 2 strategies (20–49% expert endorsement) displayed in yellow. Strategies are presented in descending order by cumulative endorsement across CFIR barrier constructs. Cumulative endorsement is the sum of expert endorsements for an ERIC strategy across all identified barriers. Major barrier constructs had universal agreement among participants, minor barriers had mixed perspectives.

Full size image
Figure 2
figure 2figure 2

ERIC strategies matched to surgeons’ CFIR barriers. Percentages indicate relative expert endorsement of a strategy to address a CFIR barrier according to CFIR-ERIC strategy matching by Waltz et al.22. Level 1 strategies (≥ 50% expert endorsement) displayed in green, and bolded in left column. Level 2 strategies (20–49% expert endorsement) displayed in yellow. Strategies are presented in descending order by cumulative endorsement across CFIR barrier constructs. Cumulative endorsement is the sum of expert endorsements for an ERIC strategy across all identified barriers. Major barrier constructs had universal agreement among participants, minor barriers had mixed perspectives.

Full size image

Participant suggestions for implementation

Participants had many suggestions for how they would like to see the new recommendations implemented in their setting. Of the 73 total ERIC constructs, 24 were addressed by at least one participant during the interviews. The number of participants who endorsed a specific ERIC strategy are listed, and compared to percent endorsement according to the CFIR-ERIC strategy tool output in Table 4. The top five participant recommended strategies were: 1. ‘Audit and provide feedback’, “The best way probably would be for someone to have some degree of formalized feedback on their performance, which probably means receiving feedback on some scheduled time interval rather than feedback regarding each individual case.” (Surgeon 7); 2. ‘Change record systems’, “The only other debatable thing which I don't see it happening would be if somehow in EndoVault you actually recorded the endoscopies," (Gastroenterologist 18); 3. ‘Distribute educational materials’, “Place the infographic by the computer to reference during your paperwork.” (Gastroenterologist 18); 4. ‘Conduct educational meetings’, “A five-minute ad right before the next surgery or the next journal club or a five-minute plug before the next the GI Journal Club, right? Those are forums where you're getting enough people coming that you're going to get critical mass.” (Surgeon 14). 5. ‘Promote adaptability’: “Change anything that they perceive as an extra step”.

Table 4 Frequency of ERIC strategies suggested by interview participants compared to recommended strategies according to CFIR-ERIC strategy mapping.
Full size table

Discussion

Various groups have created recommendations to standardize lesion localization techniques9,35,36,37, however, there is large variation in these practices25,27,38,39. New Canadian Delphi consensus recommendations for optimal endoscopic localization of colorectal neoplasms provides a framework to standardize practices between providers5. Guided by the CFIR, the present research identifies across gastroenterologists and surgeons in a major Canadian city: (1) consensus on barriers and facilitators to implementing these new recommendations, and (2) areas with mixed perceptions both within and across study groups.

Importantly, most barriers (9 out of 10) identified were common to both gastroenterologists and surgeons. The CFIR-ERIC strategy-matching algorithm was used to propose externally validated (based upon expert consensus) types of strategies needed to overcome perceived barriers. Study participants also proposed their own implementation strategies. Due to similarities in perceived barriers between specialty groups, top ERIC strategies were identical for both specialties. There was also substantial overlap between expert-recommended strategies, and those suggested by our participants. Combining these approaches allows us to narrow down from a list of 73 ERIC categories into seven context-specific implementation strategies, including: 1. ‘Access new funding’, 2. ‘Altering incentives/allowance structures’, 3. ‘Change record systems’, 4. Educational interventions (i.e., ERIC recommends: ‘Distribute educational materials’, ‘Develop educational materials’ and ‘Conduct educational meetings’), 5. ‘Audit and provide feedback’, 6. ‘Identify and prepare champions’, and 7. ‘Promote adaptability’, The first three strategies in particular address the most common ‘major’ barriers identified by both specialty groups, which stem from a lack of internal and external organizational factors to incentivize compliance with the recommendations, and a lack of key resources needed to follow the recommended practices.

While one strength of the CFIR-ERIC framework is its flexibility, its breadth also makes these recommended strategies relatively non-specific. How these strategies can be utilized in future implementation efforts depends upon budget constraints, logistical considerations, and knowledge translation expert interpretation. Our participants’ suggestions allow us to tailor these recommended strategies into more prescriptive “next-steps,” recognizing that multiple interventions are possible, and these strategies need to be evaluated prospectively to ensure their validity.

One major barrier identified by both gastroenterologists and surgeons was a lack of specific resources required to follow the new recommendations. Therefore, ‘accessing new funding’, recommended by ERIC, is likely essential to any proposed solution. For example, new funding could be used to apply for resources such as modifications to the endoscopy medical record system or increase access to recommended materials (e.g., magnetic endoscope positioning device).

A lack of incentives to encourage compliance with the new recommendations was also a major barrier identified. Altering incentives, (e.g., pay-for-performance) is one of the most frequently studied ERIC strategies and is the subject of two recent systematic reviews. Both reviews identified mixed or inconsistent effects of pay-for-performance, and it is unclear which types of incentives targeted at which individuals are likely to lead to improved care40,41. While altering incentives is an expert-recommended strategy20, others suggest that this strategy is best used in combination with others, as it is unlikely to help overcome systemic barriers that prevent guideline adoption42. Altering incentive/allowance structures was also repeatedly mentioned as a desirable strategy by our participants. One popular suggestion was to provide additional compensation for a tattoo placed and documented exactly as recommended.

While not explicitly recommended according to the CFIR-ERIC strategy matching framework, ‘change record systems’ was a top strategy endorsed by our participants. Participants emphasized that local record systems do not allow for easy documentation of the recommended practices (e.g., tattoo information requires free text input). Furthermore, synoptic reporting has strong efficacy evidence for improved documentation of quality indicators in surgery43,44, diagnostic radiology45, and pathology46. Given the evidence of synoptic reports’ efficacy, changing medical records (i.e., implementing a purpose-specific synoptic report) represents an important strategy to consider locally, although would likely require additional financial resources to implement and maintain.

‘Educational interventions’ are designed to disseminate knowledge about the new recommendations. Educational interventions have been independently associated with increased clinician adherence to guidelines on a recent systematic review and meta-analysis47. However, optimal methods of clinician education to encourage guideline compliance are unknown48. Example strategies proposed by our participants include informational emails, infographic posters in the endoscopy suites, and grand rounds presentations. Combining educational interventions with additional implementation strategies appears to be superior to educational interventions alone in some settings49,50.

‘Audit and feedback’ has strong empirical evidence to support its’ effectiveness20, based upon a large Cochrane systematic review and meta-analysis51. Although the benefits of audit and feedback observed were generally small, and were highly dependent upon the method of feedback used and the baseline performance51. There are many potential aspects of the present recommendations to target for feedback. A common example raised by our participants was tattoo quality. Some endoscopists said they wouldn’t raise a saline bleb, place a 3-quadrant tattoo, or that the volume of injected ink was unimportant. However, previously local surgeons have raised tattoo quality as a major issue for lesion localization38. Prospective evaluation and feedback on these and other recommended practices is one method to address these concerns and provide real-world local data to encourage providers to fall in line with the recommendations.

‘Identify and prepare champions’ was a top recommended strategy, primarily as it is the highest endorsed ERIC strategy to address cultural barriers in an organization. Participants also mentioned champions as individuals who could continue to spur uptake of the recommendations on an ongoing basis after the initial implementation measures are over. As with many ERIC strategies, the effectiveness of implementation champions to address culture barriers are based primarily on expert opinion, and there is a paucity of evidence to inform the validity of this approach20.

Finally, ‘promote adaptability’ is important, as it reflects the reality that it is currently unknown if every recommendation must be followed to enhance localization and diminish repeat endoscopies, or if instead some recommendations can be ignored, and the desired effect will still occur5. To address this concept, the CFIR introduces the concept of an intervention’s “core components” versus its “adaptable periphery”13. The core components are the aspects of an intervention that must be followed for implementation success, whereas the adaptable components are the optional aspects that may not necessarily be required. The authors of the Delphi consensus recommendations suggest that recommendations with lower consensus could be considered “optional”, whereas those with higher consensus (i.e., consensus from the first Delphi voting round) are more strongly recommended5.

Taking all of these recommendations together, a possible implementation strategy in Winnipeg might include: modification of the endoscopy synoptic reporting system to include items from the new recommendations; purchase of magnetic positioning devices for all endoscopy suites in the city; the provision of additional compensation for a tattoo placed and documented exactly as recommended; a multi-faceted educational intervention including informational emails, infographic posters in the endoscopy suites, and grand rounds presentations; implementation of a systematic audit and feedback strategy targeted at tattoo quality, and compliance with recommended documentation; recruitment of individuals at each endoscopy suite to champion implementation of the new recommendations and maintain enthusiasm; and an adaptation plan where recommendations with lower consensus can be modified while maintaining the “core” highly recommended aspects.

Contributions to the literature

To our knowledge, this is the first study in the literature to use the CFIR to examine barriers and enablers to implementing a new guideline targeted towards gastroenterologists and surgeons to reduce repeat endoscopy. Using this approach, we have applied modern implementation science methodology to identify strategies that may be used to enhance uptake of the recommendations in Winnipeg in the future. Others have attempted to evaluate endoscopy guideline implementation and quality improvement, however, those prior efforts are difficult to compare due to their lack of frameworks, and poor reporting of implementation strategies25,38,52,53. A strength of the current research is that by selecting robust, frequently used frameworks (CFIR and ERIC), we position the present research in the context of a broader body of literature13,20. This process has many benefits. For example, by following a structured theory-based framework, our research can serve as a sort of formula for others to follow suite. While our results are not necessarily applicable to implementation of the new recommendations outside of Winnipeg, the processes used are open to critique, and readily applicable elsewhere54. Our literature review also provides an up-to-date summary of the strengths and limitations of the CFIR and ERIC constructs evaluated. By using a framework, it is also imminently apparent to ourselves, and to other researchers, which aspects of the study setting have been evaluated, and which areas need further research (e.g., the entire ‘process’ domain, and the ‘innovation source’ and the ‘stage of change’ constructs). Had we used an inductive or tacit-knowledge-derived framework, deficient areas may not have been as apparent54.

Another major benefit of using an implementation science framework is that we have built upon the previous advances of others13. For example, our CFIR construct ranking criteria has been used previously on a post-hoc basis to examine factors associated with prior implementation success for weight management33 and hypertension strategies32. We built upon this prior research in multiple ways. First, we expanded up Damschroder and Lowery’s construct ranking system32,33. We modified their system and adapted to the pre-implementation phase, which has never been done previously. We propose using this ranking system as a new way to identify barriers significant enough to warrant selection of ERIC strategies. Previously researchers have selected ERIC strategies according to all CFIR barriers identified by participants, without a method of determining their relative significance22. Others selected ERIC strategies for all CFIR constructs, regardless of whether they were perceived as a barrier or facilitator55. Furthermore, now that we have a baseline assessment, we could also evaluate how perceptions of CFIR constructs in Winnipeg change in response to implementation strategies for our new recommendations.

Another unique aspect of our research is that we have expanded implementation science frameworks to a new discipline: endoscopy guideline-implementation. Colon cancer screening has been previously evaluated using implementation science framework56,57,58,59, including the CFIR60,61, but to our knowledge, guidelines for implementation of new endoscopy practices have not been evaluated with any implementation science framework.

A final unique aspect of our research is we have identified a disconnect between what strategies our clinician participants desire compared to those that are recommended by experts. To our knowledge, this is the first study to specifically examine the differences between strategies endorsed by the CFIR-ERIC experts, and those strategies desired by research participants. The CFIR-ERIC strategies are purported to address CFIR barriers according to expert opinion, but as discussed above, to date there is little empirical evidence to support selection of one strategy over another22. Comparison between ERIC strategies, or to those strategies identified by research participants represents an interesting avenue for further research. These comparisons may provide much-needed evidence for how a strategy can be selected in the future. Presumably, participants would be more likely to buy-in to ERIC strategies they specifically endorsed, although there is no evidence to support this yet.

Limitations

Despite the important of our findings, this study design has some important limitations. First, the barriers and enablers identified are specific to the participants and settings evaluated, and should not be interpreted as broadly generalizable. This is not an inadequacy of the present research, rather, it is an inherent characteristic of qualitative descriptive research methodology62. Despite this limitation, the research processes used can be repeated in other settings to guide implementation elsewhere. Due to our use of CFIR, our findings may also be comparable to other settings.

A second limitation is that after our study completion, a new version of the CFIR was developed to reflect ongoing developments in knowledge translation research63. While the new CFIR has some new concepts, our data could be mapped to the new CFIR if necessary, to allow comparisons in future research.

Another limitation is that alternate coding systems or frameworks could have been used. There are hundreds of knowledge translation frameworks described54. We selected the CFIR and ERIC frameworks due to their broad applicability, and good fit for the research questions, methods, and settings. However, another framework could have been selected and possibly led to different results. Even within the CFIR, there are multiple methods of analysis and data coding that are possible64,65. For example, neither the CFIR nor the ERIC framework define what constitutes a significant barrier that is important enough to warrant application of dedicated implementation strategies13,20. In our analysis, we selected “net” barriers as those in need of ERIC strategies, with specific emphasis on ‘major’ barriers. However, there is no compelling evidence to suggest that only these barriers require solutions. In the present research, even net facilitator constructs had some associated barriers. One strategy is to target ERIC strategies to all barriers identified, no matter how infrequent22. Another approach is to target ERIC strategies to overcome barriers and also to amplify facilitators55. Had we followed either of these alternate approaches, we would have identified nearly every CFIR construct as in need of an ERIC strategy, which defeats the purpose of examining local barriers and facilitators a priori to identify targeted strategies.

One final limitation was that our research reflects only the perspectives of surgeons and gastroenterologists who participated. While a significant portion of all gastroenterologists and surgeons in Winnipeg chose to engage in the study (21 out of 52 possible participants), and hailed from diverse practice settings and backgrounds, our findings may not reflect the perceptions of those who chose not to participate. Furthermore, nurses, patients, healthcare administrators, allied health professionals, non-physician policy makers, and managers were excluded. This was done deliberately, as the new recommendations are targeted primarily at physicians, and their perspectives were felt to be key for devising next steps in the implementation process. Now that local gastroenterologist and surgeon perceptions are known, these additional stakeholders should be engaged to ensure their perspectives are considered for subsequent implementation.

Conclusions

This research lays the groundwork for enhancing expert-recommended practices for colorectal lesion localization during colonoscopy in Winnipeg. We identified barriers and enablers from gastroenterologists and surgeons, mapping them to implementation science constructs. Despite some differences, both groups shared many perspectives, allowing us to create a unified list of implementation strategies to overcome barriers. We also compared participant-suggested strategies to those endorsed by implementation experts, forming a list of potentially effective local strategies. Future research should test these strategies' advantages and their impact on endoscopy quality.